Oropharyngeal microbiota and probiotic treatment in hospitalised patients

Anna Tranberg

DOCTORAL DISSERTATIONby due permission of the Faculty of Medicine, Lund University, Sweden.

To be defended at Belfragesalen, BMC, Klinikgatan 32, Friday the 12th of February at 9.15 am

Faculty opponent Professor Emeritus Ola Winsö

OrganizationLUND UNIVERSITY

Author(s)

Document name

Date of issue 12 February, 2021

Sponsoring organization

Title: Oropharyngeal microbiota and probiotic treatment in hospitalised patients

AbstractThe human body exists in mutualistic balance with a large range of microbiota. In hospitalised or ill patients, this balance can be disturbed. Opportunistic bacteria may thrive in areas they do not normally inhabit, for example in the oropharyngeal tract. The pathogenesis of the most frequent and costly hospital-associated infection; pneumonia is multifactorial, one of the key components being micro-aspiration of pathogenic microbes from the oropharynx into the lower respiratory tract.Probiotics, for example Lactiplantibacillus plantarum may help restore the microbiological balance and immunological function in the human respiratory and gastro-intestinal tract.Our aim in this thesis was to1) describe the oropharyngeal microbiota in patients in different settings:a) during the first day in the intensive care unit (study I), b) at admission and throughout hospitalisation in wards for surgical and medical care (study I, II and III) and c) throughout the course of advanced head and neck surgery (study IV) and2) investigate, in randomized trials, if daily oral administration of Lactiplantibacillus plantarum 299 and 299v during the hospitalisation could postpone or diminish the appearance of a disturbed orpoharyngeal microbiota in ward patients (study III) and in patients undergoing advanced head and neck surgery (study IV).A secondary aim in studies II-IV was to see if a disturbed oropharyngeal microbiota could be associated with healthcare-associated infections, mainly pneumonia (II-III) and surgical site infections (IV).As a first part of study III, we performed in vitro experiments to see if Lactiplantibacillus plantarum 299 and 299vcould inhibit growth of the seven most frequent pathogens found in the ICU cultures.In the clinical studies, we used classic cultivation techniques by oropharyngeal swabs. Samples were always taken at admission and during hospitalisation oropharyngeal swabs were obtained every third day. In the in vitro experiments, agar-overlay and co-culture techniques were used.Our results showed an early and significantly increased incidence of disturbed oropharyngeal microbiota in both ward and ICU patients, compared to healthy controls in the society. Proton pump inhibitor use before and at admission to a hospital ward or intensive care unit was a strong risk factor for colonisation of gut microbiota in the oropharyngeal tract in all our studies. In the ward patient cohorts the risk of developing a disturbed oropharyngeal microbiota increased with length of stay and antibiotics > 24 hours before admission. In study II there was an association between proton pump inhibitor intake and hospital-acquired pneumonia. In the in vitro experiments, Lactiplantibacillus plantarum 299 and 299v showed clear inhibition of all seven pathogens, the inhibition being mainly pH-dependent. In the randomized clinical trials, daily lactobacilli treatment did not influence the incidence of developing a disturbed oropharyngeal microbiota, nor affect the incidence of healthcare-associated infection.In conclusion, we showed that a disturbed oropharyngeal flora is common in hospitalised patients, with proton pump inhibitor /treatment being the most prominent risk factor. The lactobacilli inhibited pathogen growth in vitro, but had no effect on the oropharyngeal microbiota in vivo, nor did lactobacilli influence the incidence of hospital- acquired infection.

Key words

Classification system and/or index terms (if any)

Supplementary bibliographical information Language

ISSN: 1652-8220 ISBN: 978-91-8021-018-8

Recipient’s notes Number of pages 94 Price

Security classification

I, the undersigned, being the copyright owner of the abstract of the above-mentioned dissertation, hereby grant to all reference sources permission to publish and disseminate the abstract of the above-mentioned dissertation.

Signature Date 2021-01-07

Oropharyngeal microbiota and probiotic treatment in hospitalised

patients

Anna Tranberg

SupervisorLisa I Påhlman

Co-supervisorsBengt Klarin and Carolina Samuelsson

Coverphoto by Anna Tranberg

Photo on the back by Kajsasfoto.se

Copyright pp 1-95 Anna Tranberg

Paper 1 © Wiley

Paper 2 © Wiley -Open access

Paper 3 © Wiley - Open access

Paper 4 © submitted

Lund University Faculty of MedicineDepartment of Clinical Sciences Anesthesiology and Intensive Care

ISBN 978-91-8021-018-8ISSN 1652-8220Lund University, Faculty of Medicine Doctoral Dissertation Series 2021:12

Printed in Sweden by Media-Tryck, Lund University, Lund 2021

To my family

"Concerning the wonderful structure of things in nature, investigated by Microscope”

- by German Jesuit priest and scholar Athanasius Kircher 1646

Table of Contents

Abstract...........................................................................................................9List of publications.......................................................................................11Abbreviations...............................................................................................12

Introduction...........................................................................................................13Historical background...................................................................................13Healthcare-associated infections..................................................................14Pathogenesis of Hospital-associated pneumonia..........................................19The microbiome............................................................................................21Probiotics......................................................................................................27Probiotic effect in different body parts and conditions................................28Lactiplantibacillus plantarum 299 and 299v................................................35

Aims of the studies................................................................................................39Study I...........................................................................................................39Study II.........................................................................................................39Study III........................................................................................................39

Materials and methods.........................................................................................41Ethics............................................................................................................41Study populations.........................................................................................41Oropharyngeal sample collection.................................................................44Microbiological procedures and definitions.................................................44In vitro experiments......................................................................................46Statistical analysis.........................................................................................48

Results....................................................................................................................51Study I...........................................................................................................51Study II.........................................................................................................55Study III........................................................................................................57Study IV........................................................................................................61

Discussion...............................................................................................................67

Main conclusions...................................................................................................73Study I...........................................................................................................73Study II.........................................................................................................73Study III........................................................................................................73Study IV........................................................................................................73

Populärvetenskaplig sammanfattning................................................................75Bakgrund......................................................................................................75

Acknowledgements...............................................................................................79

References..............................................................................................................83

9

AbstractThe human body exists in mutualistic balance with a large range of microbiota. In hospitalised or ill patients, this balance can be disturbed. Opportunistic bacteria may thrive in areas they do not normally inhabit, for example in the oropharyngeal tract. The pathogenesis of the most frequent and costly hospital-associated infection; pneumonia is multifactorial, one of the key components being micro-aspiration of pathogenic microbes from the oropharynx into the lower respiratory tract. Probiotics, for example Lactiplantibacillus plantarum may help restore the microbiological balance and immunological function in the human respiratory and gastro-intestinal tract. Our aim was to

1) describe the oropharyngeal microbiota in patients in different settings: a) during the first day in the intensive care unit (study I), b) at admission and throughout hospitalisation in wards for surgical and medical care (study I, II and III) and c) throughout the course of advanced head and neck surgery (study IV) and

2) investigate, in randomized trials, if daily oral administration of Lactiplantibacillus plantarum 299 and 299v during the hospitalisation could postpone or diminish the appearance of a disturbed oropharyngeal microbiota in ward patients (study III) and in patients undergoing advanced head and neck surgery (study IV).

A secondary aim in studies II-IV was to see if a disturbed oropharyngeal microbiota could be associated with healthcare-associated infections, mainly pneumonia (II-III) and surgical site infections (IV).

As a first part of study III, we performed in vitro experiments to see if Lactiplantibacillus plantarum 299 and 299v could inhibit growth of the seven most frequent pathogens found in the ICU cultures. In the clinical studies, we used classic cultivation techniques by oropharyngeal swabs. Samples were taken at admission and during hospitalisation oropharyngeal swabs were obtained every third day. In the in vitro experiments, agar-overlay and co-culture techniques were used. Our results showed an early and significantly increased incidence of disturbed oropharyngeal microbiota in both ward and ICU patients, compared to healthy controls in society. Proton pump inhibitor use before and at admission to a hospital ward or intensive care unit was a strong risk factor for colonisation of gut microbiota in the oropharyngeal tract in all our studies. In the ward patient cohorts, the risk of developing a disturbed oropharyngeal microbiota increased with length of stay and antibiotics > 24 hours before admission. In study II there was an association between proton pump inhibitor intake and hospital-acquired pneumonia. In the in vitro experiments, Lactiplantibacillus plantarum 299 and 299v showed clear inhibition of all seven pathogens, the inhibition being mainly pH-dependent. In the randomized clinical trials, daily lactobacilli treatment did not influence the

10

incidence of developing a disturbed oropharyngeal microbiota, nor affect the incidence of healthcare-associated infection.

In conclusion, we showed that a disturbed oropharyngeal flora is common in hospitalised patients, with proton pump inhibitor /treatment being the most prominent risk factor. The lactobacilli inhibited pathogen growth in vitro, but had no effect on the oropharyngeal microbiota in vivo, nor did lactobacilli influence the incidence of hospital-acquired infection.

11

List of publicationsThis thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I. Proton pump inhibitor medication is associated with colonisation of gut microbiota in the oropharynx. Tranberg A1, Thorarinsdottir HR1, Holmberg A, Schött U, Klarin B. Acta Anaesthesiol Scand. 2018 Jul;62(6):791-800. doi: 10.1111/aas.13094. 1These authors contributed equally.

II. Disturbance in the oropharyngeal microbiota in relation to antibiotic and proton pump inhibitor medication and length of hospital stay. Tranberg A, Samuelsson C, Klarin B. APMIS. 2020 Sep 27. doi: 10.1111/apm.13087.

III. Efficacy of Lactiplantibacillus plantarum 299 and 299v against nosocomial oropharyngeal pathogens in vitro and as an oral prophylactic treatment in a randomized, controlled clinical trial. Tranberg A, Klarin B, Johansson J, Påhlman L. MicrobiologyOpen 2020 Dec 7. doi: 10.1002/mbo3.1151

IV. Effect of Lactiplantibacillus plantarum treatment on oropharyngeal pathogens and postoperative infection in patients undergoing extensive head and neck surgery: a randomized controlled pilot trial. Tranberg A, Påhlman L, Samuelsson C, Klarin B. Submitted.

12

AbbreviationsBMI body mass indexCFU Colony-forming unitCI confidence intervalHAI healthcare-associated infectionHAP healthcare-associated pneumoniaICU intensive care unitLAB lactic acid bacteriaLRT lower respiratory tractMDR multidrug-resistantNV-HAP non-ventilator healthcare-associated pneumonia OR odds ratioPPI proton-pump inhibitorRCT randomized controlled trialRR risk ratioSCFA Short-chain fatty acidsSSI surgical site infectionVAP ventilator-associated pneumoniaURT upper respiratory tractUTI urinary tract infectionLp299 Lactiplantibacillus plantarum 299Lp299v Lactiplantibacillus plantarum 299v

13

Introduction

Historical backgroundThe currently accepted and proved “germ theory of disease”, states that microorganisms or “germs” may lead to disease. Basic models of modes of infection consistent with the “germ theory” can be found in several old dynasties, for example written down in the Mosaic Law from the ancient near east and in scripts from ancient India (600 years BC), where the Indian physician Sushruta theorized: "Leprosy, fever, consumption, diseases of the eye, and other infectious diseases spread from one person to another by sexual union, physical contact, eating together, sleeping together, sitting together, and the use of same clothes, garlands and pastes.”, in his book Sushruta Samhita (1).

Two thousand years later, in 1674, the Dutch businessman and scientist Antonie van Leeuwenhoekis is said to have been the first to see and describe bacteria using his self-made lenses and microscopes. He called the living organisms he saw: "animalcules", meaning "little animals". He was able to describe the difference in appearance of the “animacules” in different sorts of water and from different parts of the human body. Antonie van Leeuwenhoekis is by many considered as the true “Father of Microbiology", although 20 years earlier, the Jesuit priest and scholar Athanasius Kircher investigated the blood of plague victims under the microscope. He noted the presence of "little worms" or "animalcules" in the blood and concluded that the disease was caused by microorganisms. He was the first to attribute infectious disease to a microscopic pathogen. Kircher also proposed hygienic measures to prevent the spread of disease, such as isolation, quarantine, burning clothes worn by the infected and wearing facemasks to prevent the inhalation of germs. It was Kircher who first proposed that living beings could enter and exist in the blood.

In the 19th century the beginning of modern microbiology and medicine was created by the experiments and conclusions made by Louis Pasteur, Ignaz Semmelweis and Joseph Lister among many others. Louis Pasteur’s experiments on spoiled milk, treated with heating (pasteurization), and his fermentation studies on wine and beer showed that the spoilage/fermentation of the liquid required particles from the air (and not the air itself). These classic experiments were performed in swan neck bottles and the results lead Pasteur to believe that micro-organisms could be responsible for infecting animals and humans and to cause disease. These beliefs triggered the start of his experiments on immunology and vaccines. Pasteur had

14

many rivals and many contemporary scientists declared to have been deceived by him. Among them was Antoine Béchamp who claimed to have been the first to show the role of microorganisms in fermentation. Béchamp and Pasteur became lifelong rivals, Béchamp also questioned the “germ theory”; denying that bacteria could invade a healthy animal and cause disease. Béchamp instead believed that disease occurs when the host or the internal environment of the body becomes favorable to pathogenic organisms. Béchamp’s dispute with Pasteur lead to his dishonouration in the medical world, and his life-long work faded into scientific obscurity.

While the “germ theory and disease” and Louis Pasteur’s contributions to the field still stands, I believe that the fate of Béchamp was somewhat unfair. Some of his ideas, experiments and results have been verified in modern microbiology. Many studies, including the ones constituting this thesis, show that it may not be enough with the presence of a pathogenic bacteria to cause disease or infection. For the colonisation to become an infection certain host factors or weaknesses are predisposing. Concurrent illness, medication, a compromised immune system and/or alterations in the individual’s normal microbiome contribute to allow an existing potential pathogen to become invasive.

Healthcare-associated infectionsDefinition and implicationsThe world health organisation (WHO) defines a healthcare-associated infection as follows: “Healthcare-associated infection (HAI), also referred to as "nosocomial" or "hospital" infection, is an infection occurring in a patient during the process of care in a hospital or other health care facility which was not present or incubating at the time of admission” (2). WHO further regards HAIs (the most frequent adverse event in health care) being a costly complication in health care facilities around the world, entailing prolonged hospitalisation, and an increase in morbidity and mortality in the patients afflicted (3). Whilst the exact global burden of HAI is unknown, estimated prevalence rates are between 5.7%-19.1% among low and middle-income countries and 5.7–7.5% in high- income countries (4). According to the European Centre for Disease Prevention and Control (ECDC), each year around 3.2 million patients are infected with healthcare-associated infections following exposure in healthcare facilities across the European Union and a total of 37 000 of them die as a direct consequence. The ECDC has in several point prevalence surveys shown that the prevalence of patients with at least one HAI in the European Union/European Economic Area (EU/EEA) sample was 5.9% (country range: 2.9–10.0%) in 2017 (5). HAIs also contribute to a global increase in antimicrobial resistance of 36%

15

during 2016-2017, due to prolonged and broad antibiotic treatment, and longer hospital stays (5).

Regarding healthcare economics, HAIs impose a huge burden on healthcare economy, exact figures depending on type of HAI, country and resources (6). In the United States, the Centre of Disease control and prevention state that on any given day, about one in 31 hospital patients has at least one healthcare-associated infection. While there is considerable variability in the costs of HAI, the low-cost estimates of $5.7 to $6.8 billion annually are still substantial when compared to the cost of inpatient stays for other medical conditions (7, 8).

According to the ECDC the most reported types of HAI in Europe in 2017 were respiratory tract infections (21.4% pneumonia and 4.3% other lower respiratory tract infections), urinary tract infections (18.9%), surgical site infections (18.4%), bloodstream infections (10.8%) and gastro-intestinal infections (8.9%), with Clostridium difficile infections accounting for 44.6% of the latter or 4.9% of all HAI (5).

In Sweden, The Swedish Association of Local Authorities and Regions (SALAR), has applied the Institute for Healthcare Improvement (IHI) Global Trigger Tool for Measuring Adverse Events (9), to measure HAIs. This project began in 2008 and the last report showed a national decrease in prevalence of HAIs from 5.2% to 4.4% between the years 2013-2018. SALAR ascribes this decrease to the implementation of preventive strategies (10). Still, 57 000 Swedes are annually afflicted by a HAI (whereof 1300 patients die), and SALAR estimates that 30-50% of the HAIs could be avoided. For the society HAIs lead to 2.2 billion SEKS in extra costs/year for the healthcare system and a doubling of the median length of hospitalisation (>10 days) (10, 11). The most frequent HAIs in SALAR´s report was urinary tract infection (UTI), surgical site infection (SSI) and pneumonia. The three nosocomial infections with the highest mortality rates in Sweden during this period were sepsis (36% of all deaths), pneumonia (31% of all deaths) and SSI (13% of all deaths).

Ventilator-associated pneumoniaHAIs are much more frequent among patients admitted to an Intensive care unit (ICU) than among those admitted to other hospital wards, affecting up to 50 % of all ICU patients according to the global “Extended Study on Prevalence of Infection in Intensive Care III” (EPIC III) from 2020 (12). The most common HAI in the ICU is ventilator-associated pneumonia (VAP) (13), defined as pneumonia occurring more than 48–72 hours after endotracheal intubation and mechanical ventilation. Data on prevalence of VAP vary. A review from 2013 stating that VAP occurs in 9–40% of intubated patients (14), however, lack of a gold standard definition leads to both under- and overdiagnosis (15), and the recent EPIC III prevalence study showed higher numbers than earlier estimated (12). VAP is associated with increased mortality and morbidity for the ICU patient (16) and is the leading cause

16

of death from nosocomial infections in critically ill patients. Mortality rates are the highest in immunocompromised, surgical and elderly patients (17). Similar to most HAIs, VAP is associated with increased hospital stay and healthcare costs, and increased risk of infection with multidrug-resistant (MDR) pathogens. The highest risk of developing VAP is during the first week of ventilator therapy, and the time of onset of nosocomial pneumonia depends on the aetiology, chosen antimicrobial treatment and outcome. Previously, VAP was categorised as either early onset (arising after < 5days) or late-onset VAP, but new reports have challenged this classification (17). More recent studies have found comparable aetiologies in patients with early onset or late onset VAP (except for S. aureus) (18). This may be related to the worldwide rise in MDR pathogens and illustrates that the local ICU ecology is a very important risk factor for acquiring MDR pathogens, irrespective of the length of intubation. Despite decades of intense work and studies striving to diminish the risk of developing VAP, the condition still remains one of the major threats to modern intensive care. Preventive strategies such as “VAP care bundles” seem to be effective and has been shown to lower the mortality from 20-30% to 9%. These bundles include 1) identifying early signs of infection, especially during the first week of ventilation, 2) obtaining an invasive or non-invasive culture sample and 3) early antibiotic treatment (15). Other parts of preventive “VAP care bundles” include placing the patient in a semi-recumbent position, oral care, daily assessment of sedation needs and possibility for spontaneous breathing. Subglottic suctioning devices also reduce the risk of VAP (15, 19).

Table 1. From Christine A'Court and Christopher S Garrard, “Nosocomial pneumonia in the intensive care unit: mechanisms and significance” in Thorax 1992 (22). Reprinted after permission by the publisher, licence number 4954760526540

17

In the ECDC report: “Healthcare-associated infections acquired in intensive care units - Annual Epidemiological Report 2016 [2014 data]”, the most frequently isolated microorganisms in ICU-acquired pneumonia were Pseudomonas aeruginosa followed by Staphylococcus aureus, Klebsiella species. and Escherichia coli. See Table 1 for the results from three studies with culture samples attained by bronchoscopy or bronchiolar lavage, from critically ill patients in 1992 (20). The same pathogens are still responsible for most VAPs in 2020, albeit with alarmingly increasing prevalence of MDR pathogens (21).

In recent years several expert groups have sought to define, classify and create guidelines regarding treatment regimens for MDR pathogens, elucidating resistance profiles in Staphylococcus aureus, Enterococcus species, Enterobacteriaceae (other than Salmonella and Shigella), Pseudomonas aeruginosa and Acinetobacter species, bacteria often responsible for healthcare-associated infections and prone to multidrug resistance (17, 22).

Nonventilator hospital-associated pneumoniaNonventilator hospital-associated pneumonia (NV-HAP) is globally emerging as the most common and lethal HAI (8, 23). In an article by Ewan et al from 2017, it is described as “the Cinderella” disease: NV-HAP receiving little attention in hospital or research despite great costs and high prevalence (24). The definition of NV-HAP is pneumonia that develops in patients that have been admitted to the hospital for >48 hours (usually the incubation period is at least 2 days).

There are several reasons for the lack of studies, focus and guidelines regarding NV- HAP. Early infection prevention programs and guidelines, dating from the 1970- 80’s focused on what was considered the most common nosocomial infections then: SSIs, device-and catheter-associated infections and VAP. Many hospitals and countries have well-established surveillance and prevention strategies regarding these HAIs, which has led to a clear decrease in prevalence of several of these nosocomial infections (8, 25).

Almost all knowledge and studies regarding HAP come from ICU environments. Globally, national quality improvement initiatives have implicitly encouraged hospitals to focus on VAP, in particular by repeatedly proposing public reporting of VAP but not of NV-HAP (26). National institutes for healthcare improvement have requested prevention bundles for VAP but not for NV-HAP. The focus on standardised surveillance and prevention of the above-mentioned HAIs, has over the years changed the scenery regarding nosocomial infections. In a multistate point-prevalence survey organized by the Centres for Disease Control and Prevention (CDC) and reported in 2018, the investigators found that the majority of pneumonias occur in non-ventilated patients (8). In 2018, the Pennsylvania Patient Safety Authority published a survey including all acute-patient HAPs from 2013-

18

2016. The consensus was that both NV-HAP and VAP continue to be problematic for acute-care patients in Pennsylvania. However, NV-HAP affects more patients, contributes to more deaths, and increases costs more than VAP (23). Their recommendation regarding NV-HAP was that obligatory surveillance, education of nursing staff and preventive care bundles should be instigated promptly. In a study by Baker et el from 2018, they showed that NV-HAP is not only a disease afflicting the old and frail, but it occurs in all patient groups, in all hospitals. Furthermore, they showed that NV-HAP is associated with significant costs, morbidity and mortality, and recommend that NV-VAP should be elevated to the same level of prevention as device-related infections (25). They recommend strict application of NV-VAP preventive care bundles. In a study from Magill et al regarding changes in prevalence of HAIs in US hospitals, the 2018 statistics show a significant decline in overall HAI prevalence (8). However, the decline pertains to UTIs and SSIs, whereas the prevalence of HAP and Clostridium difficile infection increased. Also, in the European guidelines for management of HAP and VAP, NV-HAP is discussed as an understudied and undermanaged HAI, occurring in 5-20 cases per 1000 patients. Regarding the Swedish recommendations for preventing HAIs, written by the National Board of Health and Welfare (Socialstyrelsen) and published on the Public Health Agency’s (Folhälsomyndigheten) website, the information covers a broad number of HAIs. However, with respect to HAP, the data and prevention recommendations mainly focus on VAP (27).

Together these observations suggest the possibility that there is a major gap in current surveillance and prevention programs regarding NV-HAP. There is also a need for huge multi-centre studies, requiring a large sample size to evaluate effect of preventive measures (28). There are a few recent studies focusing on surveillance and possible prevention bundles for NV-HAP (25, 29).

As with VAP, there is no golden standard for diagnosis of NV-HAP. Achieving a culture sample from the lower respiratory tract is too invasive for this patient category. Sputum cultures or oropharyngeal swabs are currently the most common ways to identify a pathogen. In an Italian study looking at risk factors for NV-HAP, the investigators were able to identify the probable causative agent in 40% of the cases, the most frequent pathogen being Streptococcus pneumoniae, followed by Enterobacter species, Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae and Pseudomonas aeruginosa (30). Regarding recommendations for antibiotic treatment of a bacterial NV-HAP, it is impossible to give other than generalised treatment recommendations. Each hospital carries its own specific ecological and resistance patterns due to previous and current policies. It is clear that recommendations for initial empiric antimicrobial treatment should take into account (acknowledge a sufficient flexibility to allow modifications according to) the local and unique pathogen colonisation, which is the result of preventive and treatment strategies, and the type of patient and bacterial population treated.

19

Pathogenesis of Hospital-associated pneumonia

The healthy upper and lower respiratory systemThe upper respiratory tract (URT) includes the nostrils, nasal passages, paranasal sinuses, the nasopharynx and oropharynx, and the portion of the larynx above the vocal cords, and serves to humidify, clean and heat the air that enters the lower respiratory tract (LRT), which consists of the portion of the larynx below the vocal cords, the trachea, bronchi, bronchioli and alveoli (were the gas exchange occurs). Through advances in microbiology over the last decades, we now know that different parts of the respiratory system are inhibited by specific communities of bacteria and fungi, which vary from person to person, and between continents. These communities are called microbiomes, and are influenced by environmental factors and intrinsic factors, and live in a mutualistic balance with the host (31). The microbiomes interplays with our immune system, and probably plays a greater role to our health than we can fathom (32). Together, the microbiomes living in our body (mainly in the gastro-intestinal tract) contain a diverse genome that is 10-fold our own. Together we constitute a kind of superorganism, dependent on each other for survival.

Normally, all people micro-aspirate daily. Micro-aspiration means that small quantities of liquid/saliva from the oropharynx spills down into the lungs. Studies show that the microbiome of the adult lungs reflects the microbiome of the oropharynx, and not the other microbiomes of the URT or mouth (33). However, since the healthy LRT is a master of reducing pathogen growth, the quantity of micro- organisms is much lower in the lungs than in the oropharynx (34). The LRT has many features that makes the lungs a difficult place for bacteria to survive in, and their existence there is more transient than actual colonisation. The bronchial tree is unique as the mucus flow is upwards, thanks to cilia moving unwanted elements cranially. Further down beyond the bronchioles, the alveoli bubbles constitute a huge surface where gas exchange occurs. The alveoli are lined with a thin single-cell squamous epithelial coated with surfactant which is a lipoprotein complex, containing free fatty acids and proteins known for their antibacterial properties (35). Furthermore, the alveoli protect themselves from colonisation by alveolar macrophages, a phagocyte residing on the surface of the alveoli. All in all, in the healthy subject the upper and lower respiratory tract live in a balance consisting of a healthy oropharyngeal microbiota, normal levels of aspiration and open functioning lungs with several protective immunological features in place (33, 35).

Pathogenesis behind HAPThe development of nosocomial pneumonia is multifactorial. In a recent study on critically ill patients, a disturbed oropharyngeal microbiota, collapsed lungs and an

20

affected immune system contributed to an early increased bacterial burden in the lungs (cultures obtained by mini-broncho alveolar lavages) and was predictive of acute respiratory distress syndrome (ARDS) and a worse clinical outcome (36). In a study from 1969, the investigators showed similar results; that a disturbed oropharyngeal microbiota in hospitalised patients correlated with the “level of sickness” in the patient (37). However, there are also distinct risk factors associated with the development of both VAP and NV-HAP. Some of these depend on patient characteristics present at admission (medication, chronic disease, smoking and alcohol habits) and are consequently not easy to modify or change. Other risk factors are hospital-associated events such as: surgery, lack of oral care, fasting, supine position, increased micro-aspiration due to sedatives, feeding tubes etc. In a mini-review by Wu et al from 2019, they show that old age, male gender, smoking, increased ventilation time, decreased level of consciousness, supine contribute to an increased risk for developing VAP(38). The increase of possible pathogens that enter the lower respiratory tract together with an altered ecosystem in the alveoli, entailing local inflammation and unaerated areas, make it easier for pathogens to thrive (35). In a study from 2019, the investigators show that patient position, proton-pump inhibitors (PPIs), opioid treatment and use of vasopressors plus antibiotic therapy indirectly alter the patient’s microbiome and thereby further increase the risk of developing VAP, see Figure 1. These risk factors are often considered in the ICU “ventilation care bundles”, that are developed to prevent VAP. Regarding NV-HAP, Sopena et al published an article in 2014 concerning risk factors for developing HAP outside of the ICU. They found that hospital admission in the previous month, malnutrition, depressed consciousness, chronic renal failure, anaemia (haemoglobin <10 g/dL), charlson comorbidity index ≥3 and thoracic surgery increased the risk for NV-HAP (39). In our study published in 2020, we showed that PPI medication and receiving antibiotics before hospitalisation were associated with an increased risk for NV-HAP (40).

The gut-lung axis may play a role in critically ill patients or in other patients where the gut is affected (for example by abdominal surgery). Some studies proclaim that a “leaky gut “, meaning increased gut permeability, may lead to microorganisms from the gut entering the lung and thereby influencing the lungs’ immune defence. This can also work the other way, i.e., the lung sends proinflammatory and immunomodulatory signals to the gut (for example in ARDS) (32).

21

Figure 1. Factors that may alter the microbiome in both gut and respiratorytract in critically ill patients, and in turn increse the risk for VAP. Reprinted by permission from Moron et al (41).

The microbiome

The oropharyngeal microbiomeThe oropharynx constitutes the pharynx in the upper respiratory tract, located behind the oral cavity. It includes the back third of the tongue, the soft palate, the side and back walls of the throat, and the tonsils (see Figure 3).

22

Figure 3: The oropharynx This work by Cenveo is licensed under the Creative Commons Attribution 3.0 United States

The oropharynx is lined with non-keratinized stratified squamous epithelium and the phylum that inhabits the oropharynx differs significantly from that in the adjacent oesophagus, nasal and mouth in the healthy adult. However, the phyla of the oropharynx resemble that of the saliva (34, 35). Taxonomically, bacteria are classified in the following order as phyla, classes, orders, families, genera, and species. According to a study using16S rRNA microarrays on seven healthy adults, the majority of the oropharynx contains microbiota from these five phyla in descending order: Firmicutes, Proteobacteria, Bacteroidetes, Actinobacteria and Fusobacteria (42). Since in our studies, the majority of the participants are over middle-aged to old, it was interesting to learn that the phyla detected in a study looking at frail older patients (median age 80) was similar, albeit with enhanced domination by Firmicutes (48,3%), showing mainly as an increase in Streptococci (43). Old people also suffer from a gradual loss in species variability compared to younger patients, and there has been speculation whether this loss of diversity influences the risk of infection. Also, with age, the microbiota of the nasopharynx more and more resembles the microbiota of the oropharynx (44).

Stepping down to the more familiar genus level, one can start out by emphasizing that, while there is a great resemblance in phyla and genus between individuals, the quantitative differences can be vast (45). In a study performing oropharyngeal swabs on 158 healthy students using 16S rRNA sequencing, the results corroborate with the results from the above-mentioned studies (45). The study shows that the oropharynx harbours a very diverse assembly of predominately anaerobic microbiota, containing both gram-positive and gram-negative bacteria. Among the

23

158 students, the quantity of the total amount of bacteria consisting of streptococci varied from 2.4-66%. The most common genera in the oropharynx in that study was Streptococcus, Veillonella, Rothia, Prevotella and Actinomyces (45)(Figure 4). The study also points out that even if the different parts of the URT, the oesophagus, and the mouth harbour distinctly different microbiomes, there is a continuum of ecological interactions between all these milieus.

Figure 4: Results from 16S rRNA sequencing from oropharyngal swabs from 158 healthy students Reprinted according to Creative Commons Attribution 4.0 International License, and permission by the authors Retchless et al, , https://doi.org/10.1038/s41598-020-57450-8

24

The oral microbiomeThe oral cavity is part of the gastro-intestinal tract (Figure 5). The microbial composition of the oral cavity is extensively investigated due to easy accessibility for sampling. The oral cavity contains the second largest community of microorganisms in the body (after the colon) (46). It has two principally different niches with respect to inhabiting microorganisms: the soft oral mucosa and the teeth with the soft tissue surrounding them. Dental plaques consist of organised biofilms formed on the teeth. Biofilms can be described as slime layers consisting of one or many different bacteria, where the bacteria organise themselves into an interactive and functional community. Several studies, predominantly from nursing-home settings, have linked poor dental hygiene to dental biofilm aspiration and an increased risk of pneumonia. When it comes to inflammation due to dental plaques formed near the gingival margin: gingivitis, it may untreated progress to the chronic inflammatory disease periodontitis, which in turn may lead to local bone destruction (47). The microbiome on the teeth and the soft tissue around the teeth (gingiva) are also associated with development of dental caries and periodontal disease. Especially an abundance of Lactobacilli and Streptococcus mutans have been implicated in the formation of caries, but the current belief is that the development of caries is more multifactorial than earlier understood (47).

Figure 5: The oral cavity Wikimedia Commons

25

To summarize, it is a disturbance in the oropharyngeal microbiota with subsequent micro-aspiration that is thought to be the route for pathogens to enter the LRT and contribute to the development of HAP (33). Poor oral and dental care can further increase the risk of developing HAP via aspiration of dental plaques and/or via a general disturbance of oral microbiota which influence the composition of the saliva.

Our microbiome formation starts through the mouthAround birth it is by oral contamination (by vaginal delivery and breastfeeding) that the formation of our entire internal microbiome starts, and at one year of age the microorganisms predominantly found in the mouth are Streptococcus salivarius, Lactobacilli, Actinomyces, Neisseria and Veillonella (46). The maturation and colonisation of the human microbiome is believed to be achieved around the age of five (44). Some studies theorize that it is in this period of life (from birth- five years old) that it is easiest to modulate the gastro-intestinal microbiome, the composition of which influences the risk to develop different diseases such as allergies, diabetes, obesity, irritable bowel syndrome, cancer, cardiovascular disease and depression (47, 48).

The human microbiomeThe human microbiome project (HMP), was initiated by the United States National Institutes of health to better understand and define the human microbiome. The first phase of the project (HMP1) was launched in 2008, and finished in 2013. The second phase of the project, the Integrative Human Microbiome Project (iHMP) was launched in 2014 with the aim of elucidating the roles of microbes in health and disease. The HMP1 characterized the microbial communities from 300 healthy individuals, at five sites on the human body: the nasal passages, the oral cavity, the skin, the gastrointestinal tract, and the urogenital tract. 16S rRNA sequencing was used to identify the bacterial communities in order to elucidate whether there is a core healthy microbiome. The 16S rRNA sequencing was complemented on a smaller scale through selective extensive whole genome sequencing, all information which is accessible at https://hmpdacc.org/hmp/. In a review from 2018, Gilbert et al discuss the diversity and complexity of our microbiome, and what factors influence its composition and change. Age, diet, exercise, antibiotics, food and occupation can all lead to transient or permanent changes in our microbiome at different sites (49). The microbiome consists of bacteria, virus, archaea and fungi. Together they contain thousands of genes, interacting with our body, and by far exceeding the human genome in diversity and flexibility. The current belief is that in our body there is 1.3 bacteria residing per one human cell (50), albeit with great individual variation. As mentioned in the section above, our microbiome seems to

26

influence how our body functions and responds in a variety of ways, both in sickness and in health. The current knowledge in this field is thoroughly described in a state- of-the-art review by Young in BMJ 2017 (51), where the author defines the nomenclature surrounding this field and describes current knowledge about how our microbiome contribute to our health, and how a disturbed microbiome may contribute to certain diseases. He describes a mutualistic relationship where, for example, in the gut, the microbiota co-metabolises bile salts and acids in “communication” with the host. The gut microbiome also influences epithelial responses and systemic responses in regard to the development and activity of the host immune system, and also helps to protect us from colonisation of potential pathogens (see Figure 6). However, if the microbiome is altered, disturbed or dysfunctional, there are convincing evidence for an increased risk for infectious diseases such as Clostridium difficile infection and pneumonia. The gut microbiome is altered in inflammatory bowel diseases as compared to healthy controls, and the same is true for asthma and chronic obstructive pulmonary disease. A disturbed gut microbiome is similarly associated with metabolic disorders such as obesity and diabetes, where the microbiome is believed to influence the host energy metabolism (51).

Figure 6 Microbiome-host interactions. Reprinted with permission by the author Vincent Young and BMJ (51)

27

ProbioticsProbiotics are defined by the International Scientific Association for Probiotics and Prebiotics (ISAPP) as live microorganisms that, when administered in adequate amounts, confer a health benefit to the host (52). Prebiotics are compounds in food that induce the growth or activity of beneficial microorganisms such as bacteria and fungi (53). Pro-and prebiotics can be given as a combination, synbiotics, where the selected prebiotic helps growth and survival of the probiotic it is combined with (54). As an example, Health Canada has accepted the following bacterial species, when delivered in food at a level of 1 × 109 colony forming units (CFU) per serving, as probiotics for which nonstrain-specific claims might be made: Bifidobacterium (adolescentis, animalis, bifidum, breve and longum) and Lactiplantibacillus (acidophilus, casei, fermentum, gasseri, johnsonii, paracasei, plantarum, rhamnosus and salivarius) (55).

Recent reviews suggest that some of the positive effects of probiotics are more general, and more widespread among different probiotic strains, species and genera, than earlier believed. Probiotics display a variety of antimicrobial properties, for example the production of ammonia, lactic acid, free fatty chains, hydrogen peroxide and bacteriocins (56). However, many traits are strain specific, and it is therefore important to identify a specific probiotic by its full nomenclature, including species, subspecies and (if appropriate) by an alphanumeric title. Strain-specific effects may be diverse due to different antimicrobial properties and their effect on different physiological systems, such as neurological, immunological and endocrinological. Other strains may produce specific bioactive substances (52, 57). See Figure 7.

Figure 7: Possible distribution of mechanisms among probiotics. Reprinted by permission from Hill et al (52). SFCA= Short-chain fatty acid

28

Probiotic effect in different body parts and conditions

Oropharyngeal tractIn 2002, Stjernquist-Desatnik published a pilot study where she inspected if Lactobacillus plantarum DSM 9843 (also known as Lp299v) has the ability to persist on the tonsillar surface. The rationale for the study was the notion that eradication of the normal throat microbiota might be one cause for treatment failure of group A streptococcal tonsillitis. The investigator believed that oral administration of Lactobacillus plantarum DSM 9843 might stimulate a return to a normal oropharyngeal microbiota, which would encourage the growth of, for example, alfa streptococci with inhibitory activity against group A streptococci. The conclusion was that Lactobacillus plantarum DSM 9843 persisted on the tonsillar surface on healthy test subjects (58).

The above study inspired Klarin et al to perform a randomized clinical trial to see if the burden of VAP-associated pathogenic microbes in the oropharynx could be reduced. Critically ill mechanically ventilated patients received either standard oral cavity sanitisation care using chlorhexidine or oral cavity care with cotton swabs drained in Lactobacillus plantarum 299 (Lp299) (DSM 6595) (61).

The study by Klarin et al clarified that Lp299 becomes established in the oropharynx after twice daily application of the lactobacilli-soaked cotton swabs, (by detection via classical cultivation). The results showed no difference in the incidence of new pathogens in the oropharynx nor in the tracheal aspirates between the groups. Thus, Lp299 had an effect comparable to that of chlorhexidine in reducing colonisation of pathogens in the oropharynx, but without the possible adverse effects associated with chlorhexidine use in critically ill patients (59, 60).

Oral cavityIn the last decades, several studies have been conducted to elucidate if administration of probiotics could improve gingivitis and/or periodontal health. So far, different studies have shown conflicting results. One recent randomized controlled trial (RCT), where the intervention group received daily application of Lactobacillus reuteri for 12 weeks resulted in reduced probing pocket depths but had no suppressive effect on the investigated periodontal pathogens, as compared to the control group (61). In an RCT from 2020, a tablet containing 108 CFUs of the probiotic Weissella cibaria CMU was administered once daily in the intervention group, in a cohort of patients without periodontitis. Controls were given tablets without probiotics. After 8 weeks of treatment, bleeding on probing from the base of the gingival surface was significantly reduced in the intervention group. There

29

was also an improvement in microbial environment, with a reduction in Fusobacterium nucleatum in subgingival plaques (62). F. nucleatum is normally present in the oral cavity, and is known to participate in plaque formation.

GastrointestinalIn an overview by Cremon et al from 2018, the authors summarise the probiotic beneficial effect from a gastro-intestinal point of view (57). The authors emphasise that although the evidence behind many beneficial effects from probiotics are well proven, the actual mechanisms remain mainly unidentified. Probiotics can increase resilience against pathogens, improve epithelial function by for example strengthening the tight junctions in the epithelial barrier of the intestine. Probiotics are also known to produce short-chain fatty acids (SCFAs). SCFAs are produced in the intestine by microbial fermentation of dietary fibres and consist of less than six carbon atoms. SCFAs are the major source of energy for colonocytes (the colonal epithelial cells), and the SCFA butyrate is especially important for colonocyte function and gastrointestinal function (57). Probiotics also increase mucus secretion, and influence inflammatory responses by increasing secretory IgA levels, reduce cytokine production and neutralize carcinogens (63, 64).

Probiotics have well established health effects on several gastrointestinal diseases (57). In acute diarrhoea in children, a probiotic mixture of Lactobacillus rhamnosus and the yeast Saccharomyces boulardii shortens the duration of symptoms (65) with one day, and is a recommended treatment by The European Society for Paediatric Gastroenterology Hepatology and Nutrition (ESPGHAN). Regarding the condition of antibiotic-associated diarrhoea, several RCTs show solid evidence in favour of using probiotics against this condition (66). Probiotics is a recommended treatment of antibiotic-associated diarrhoea in the guidelines from the World Gastroenterology Organisation Global Guidelines “Probiotics and prebiotics” from 2017. The sub-category Clostridium difficile infection may also be alleviated by the administration of probiotics, evidence is gathering but the optimal probiotic is not yet defined (67, 68). Considering the use of probiotics in Irritable Bowel Syndrome there is evidence of a small but significant effect (69). In patients suffering from Inflammatory Bowel disease (IBS), the role for prebiotics is uncertain concerning both induction and maintenance of remission of ulcerative colitis and Crohn's disease (70-72).

SurgeryIn 2018, Wu et al published a meta-analysis of a few decades’ worth of trials on the possible effect of perioperative supplementation with pro- and/or synbiotics on clinical outcome in surgical patients. The conclusion was that “for surgical patients,

30

perioperative supplementation with pro-/synbiotics is effective in preventing or controlling SSI and other infectious complications. Perioperative pro-/synbiotics might also be associated with fewer side effects, lower hospital cost and better quality of life. Current evidence indicated that perioperative synbiotics supplementation is preferred and recommended as an adjunct in surgical patients” (73). In another meta-analysis from 2020, Chowdhury et al (74) analysed 34 RCT’s and 2700 patients, and made conclusions similar to those of Wu et al. According to the authors, postoperative infections represents one third of all sepsis cases. In Chowdhury’s review, they found that both perioperative probiotics and synbiotics significantly reduced postoperative infections in elective abdominal surgery and that treatment with synbiotics also leads to a decreased length of stay. In both reviews, the overall incidence of severe adverse effects was low. Most of the trials analysed in both reviews consisted of pre- or perioperative treatment with different strains of Lactobacillus and Bifidobacterium (75). In a slightly older review from 2011, Jeppsson et al showed a three-fold reduction in SSI in upper gastrointestinal and liver transplantation surgery (63), when the patients were given probiotics perioperatively. In contrast to the other reviews, they found that perioperative probiotics had little effect in reducing SSI in patients undergoing colon resection.

Critically ill and trauma patientsIt may be that changes in the gut microbiome play an integral part of the vicious circle some critically ill patients fall into, eventually leading to multiple organ failure and increased mortality. Analysis of stool samples from critically ill patients using 16S rRNA gene amplification has shown evidence of extreme microbial dysbiosis during critical illness (76). In a study performed by Shimizu et al, faecal samples from critically ill patients were cultured to evaluate bacterial diversity. Patients with a lower/depleted bacterial diversity had a significantly increased incidence of enteritis, pneumonia and mortality due to multi-organ dysfunction compared to patients with diverse stool patterns (77). These findings are consistent with the established role of intestinal microorganisms in the maintenance of immune system response, proliferation of colonocytes, integrity of the epithelium and production of SCFAs. Critical illness involves decreased intestinal motility and reduced oxygen delivery which likely contributes to the development of intestinal dysbiosis. Broad-spectrum antibiotics and lack of enteral feeding also reduce host microbial diversity in these patients (76). In a study by Yamada et al, they showed that concentrations of faecal SCFAs in critically ill patients decreased significantly over a 6-week period (81). Several studies have investigated if pro- or synbiotics or fibre-containing enteral feeding could help restore the loss of microbial diversity and restore the function of the colonic epithelial cells and maintain the normal gut barrier(78, 79), showing promising results regarding a reduction in multi organ failure and diarrhea. Maintenance of high levels of SCFAs may improve

31

gastrointestinal integrity in critically ill patients. There are currently only a few studies, with promising results regarding decreasing gut permeability, that have investigated the possibilities of giving fibre-containing feeds, and/or probiotics to this patient population (67, 80).

In a review from Manzanares et al from 2016 regarding probiotic effect on critical illness, it was concluded that probiotic treatment is associated with a significant reduction in overall infection, including VAP. No effect was shown on in-hospital mortality, length of stay in the ICU or incidence of diarrhoea. They investigated 30 trials, and in the four trials using L. plantarum there was a greater reduction in infection, although subgroup analyses showed no significant difference compared to the other probiotics (81). In a 2020 review, Moron et al discussed the importance of choosing the right probiotic or synbiotic and argued that it should consist of microorganisms that produce or stimulate the production of SCFA, especially butyrate (41). See Figure 8. Wullt et al showed in 2007 that Lp299v stimulates the production of SCFAs and butyrate in patients with recurrent Clostridium Difficile- associated diarrhea (82).

Figure 8. Composition and functions of the intestinal microbiome in critically ill patients compared to healthy individuals. Reprinted by permission from Moron et al (41)

32

There are a few studies focusing on the effect of probiotics on VAP. In an RCT from 2010, 146 patients needing mechanical ventilation were randomized to receive either Lactobacillus rhamnosus GG or placebo twice daily. The intervention group showed a significant reduction in incidence of VAP (83). Another RCT from 2016, including 235 mechanically ventilated patients, also showed a significant reduction in VAP (from 50% to 36%) in the intervention group that received a probiotic capsule containing live Bacillus subtilis and Enterococcus faecalis three times daily (84). Two subgroups of ICU patients have an increased risk of developing VAP; trauma patients and neurosurgical patients, maybe due to long-term deep sedation compared to other ICU patients. A small RCT from 2006 showed that 15 days of administration of a combination of four different probiotics plus fibres (Synbiotic 2000Forte®) to critically ill, mechanically ventilated, multiple trauma patients, resulted in significantly reduced rates of infection, severe sepsis and mortality. VAP was the most common infection and the incidence of VAP was significantly reduced in the treatment arm. Length of stay in the ICU and days with mechanical ventilation were also significantly reduced compared to placebo (85).

An interesting RCT performed on 4500 infants in rural India was published in Nature 2017. The children were allocated to receive either a combination of L. Plantarum ATCC-202195 plus fructooligosaccharide or placebo throughout their first week of life. The report showed a significant reduction in sepsis and death (even if the total number of deaths were very low) and the greatest effect was seen on sepsis (86, 87). See Figure 9.

Figure 9: Effect of synbiotic treatment on sepsis and other morbidities in the first 60 days of life. Reprinted with permission.

33

These results should have implications for our thinking about the use of antibiotics and probiotics. Trials have shown that the “western way” of handling sepsis (following the Surviving Sepsis guidelines) actually leads to increased death in the treatment arms in young African patients (88, 89).

NV-HAPDespite numerous studies concerning VAP and probiotics, very few, if any, studies have been performed to see whether probiotics can reduce the risk for nosocomial pneumonia in ward patients. Since NV-VAP is the most frequent and costly complication in hospitalised patients, the lack of studies is surprising, especially considering the promising results with probiotic treatment on incidence of VAP. In a review from 2014 the authors discuss the pathogenesis behind HAP and the possible role for probiotics various types of respiratory tract infections in hospitalised patients (90). Regarding pneumonia outside the ICU, they refer to three studies on a mouse model with pneumococcal pneumonia, with promising results when treating with an assortment of probiotic strains. In an RCT from Egypt on children admitted to the ICU with pneumonia, most children were not on mechanical ventilation and may be considered as NV-HAP. The intervention group received oral administration of Lactobacillus rhamnosus GG throughout hospitalisation and showed a significant decrease in death, total days of antibiotic treatment and need for mechanical ventilation (91).

Considering the role of probiotics and prevention of upper respiratory tract infections (both bacterial and viral), a Cochrane analysis from 2015 and several well-performed RCTs suggest a positive role for probiotics (92-94).

CancerThere is emerging evidence that probiotic treatment may be an effective adjunct in the treatment of cancer and that it can help alleviate symptoms associated to chemo- and radiotherapy. In a review, Scott et al (2018) presented effects from probiotics that are very similar to those in the critically ill patient (95). The review puts more emphasis on immunomodulatory mechanisms mediated from the gut, like enhancement of natural killer cell activity and the anti-inflammatory and antiproliferative effects of SCFAs. It also brings up the subject of whether probiotics should be considered as possible preventive or treatment adjuncts, or both. Zhang et al showed in a meta-analysis from 2019 that regular intake of fermented dairy foods was associated with a lower overall risk of cancer (96). According to a review by Cheng et al, the bacterium Fusobacterium nucleatum has been implicated as a risk factor for various types of cancer, including colorectal cancer, oesophageal cancer, gastric cancer, head and neck squamous cell carcinoma, and pancreatic

34

cancer (97). The authors also refer to a study on patients with colorectal cancer where the administration of Lactobacillus acidophilus NCFM and Bifidobacterium lactis Bl-04 altered patients’ microbial profile, by increasing the number of butyrate-producing bacteria, while decreasing the abundance of colorectal cancer- associated genera, including Fusobacterium and Peptostreptococcus (98). Lactobacillus species is recommended as an adjunct to prevent diarrhoea after radio/chemo therapy in patients with pelvic cancer in the guidelines written by the Multinational Association of Supportive Care in Cancer (MASCC)(99).

Figure 10 summarizes the possible mechanisms and health effects of probiotics. There are several other patient groups, with for example neurological, psychiatric and autoimmune disease, where probiotics may serve as a beneficial adjunct in therapy or prevention, but presenting them here is beyond the scope of this background.

Figure 10: Schematic representation of the main mechanisms of action through which prebiotics and probiotics exert their health effects and potential clinical target. Reprinted with permission from Cremon et al (57).

35

Lactiplantibacillus plantarum 299 and 299vLactiplantibacillus plantarum- previously named Lactobacillus plantarum (L. plantarum) is a species of bacteria within the Lactobacillus family, which in turn is part of the greater group of lactic acid bacteria (LAB). Bacteria within the LAB- group are Gram-positive, catalase-negative, and possess the ability to ferment carbohydrates to lactic acid (100). LAB compromise several bacterial genera, and is mainly represented by Lactobacillus, Leuconostoc, Pediococcus, Lactococcus, and Streptococcus. Lactobacilli inhabit many environments, and the niche they populate may differ between species. L. plantarum are found in dairy, meat, vegetable, animals and humans (100). In humans, L. plantarum inhabit mucosa, especially in the gastrointestinal tract, and play an important role in maintaining the acidic environment in the vagina.

L. plantarum are rod-shaped bacilli with rounded ends, generally 0.9–1.2 μm wide and 3–8 μm long, occurring singly, in pairs or in short chains. L. plantarum has one of the largest genomes known among the lactic acid bacteria and its plastic genome makes it a very flexible and versatile species (100). It is estimated to be able to grow between pH 3.4 and 8.8, and in the temperature range 12 - 40 °C (101).

Figure 11: Lactiplantibacillus plantarum 299v. Reprinted by permission by Probi AB.

36

In food and especially dairy products, probiotics is a well-known component to help conserve produce, and L. plantarum was one of the first and most common probiotics used in food conserving around the world (97).

Like other probiotics, L. plantarum have several bacteriostatic or bactericidal properties. In addition to its LAB characteristic to produce lactic acid, L. plantarum has the ability to produce bacteriocins (102). Bacteriocins are small proteins or peptides produced by bacteria and have an antagonistic or bactericidal effect on organisms closely related to the producing bacteria (103-107). This is not unique forL. plantarum, but is a trait of many different bacterial strains, also outside of the LAB-group. The bacteriocins are commonly produced during the exponential phase of the bacterial growth curve, or at times of stress (103). Bacteriocins from different species within the LAB-group can differ, but in general they have been divided into four different classes; class I, II, III and IV. Class I- bacteriocins have a low molecular weight (<5kDa), whereas class II and III are larger, 10kDa or >30 kDa, respectively. Class II bacteriocins are usually heat-stabile, in contrast to class III that are usually heat-labile. Bacteriocins produced by L. plantarum species are also referred to as plantarocins, and have different traits and modes of mechanism. The current knowledge of bacteriocin classification and application is presented in a recent overview by Zimina et al (108). As an interesting complement on the topic of bacteriocins, a review also from 2020 presents the possible role of bacteriocins as a part of future antibiotic treatment (109).

Studies using Lactiplantibacillus plantarum 299 and 299vIn 2017, Seddik et al wrote a review over Lactobacillus plantarum and its probiotic and food potentialities. In the beginning of the review, they clarify that L. plantarum “has the qualified presumption of safety (QPS) status from the European Food Safety Authorities (EFSA) and the “Generally Recognized As Safe” (GRAS) status from the US Food and Drug Administration (US FDA)” (100). This safety stamp is crucial as it has enabled this probiotic to be easy to study and use in foods.

There are many Lactobacillus strains, all with different properties. The rationale behind choosing Lp299 and Lp299v as strains suitable for studies of potential beneficial effects on humans was born in the 1990s, were the authors administered 19 different Lactobacillus strains to healthy volunteers for 10 days. They then obtained rectal and jejunal biopsies one day and 11 days after the end of the treatment. At day one and day 11 post treatment, five Lactobacillus strains could be identified in the mucosal biopsies, and on day 11 Lp299 and Lp299v were clearly the dominant strains (110). This study confirmed that Lp299 and Lp299v had an ability to adhere to intestinal mucosa, and were considered the best suitable strains for further research on potential beneficial effects from Lactobacillus. A mentioned above, Wullt et al have shown that Lp299v stimulates the production of

37

SCFAs and

38

butyrate in patients (82). Today, Lp299 and Lp299v have become one of the most studied and commercially used probiotics.

Bellow follows a few examples of medical conditions where treatment with L. plantarum 299 or 299v have shown to have a significantly beneficial effect in human trials:

Cholesterol lowering effects (111, 112)

Irritable Bowel Syndrome (113-116)

Diarrhoea prevention/reduction in Clostridie difficile infection (67, 117- 119)

Coronary artery disease (120)

Major depression (121)

Postoperative infection in major abdominal surgery (75, 122)

39

40

Aims of the studies

Study ITo establish whether there was a disturbance in the oropharyngeal microbiota in patients within 24 hours of admission to either ICU or ward, and to explore differences in culture results between ICU, ward and control populations. The secondary aim was to explore patient characteristics associated with the occurrence of a disturbed oropharyngeal microbiota.

Study IITo investigate the appearance of a disturbed oropharyngeal microbiota in ward patients during hospitalisation and to explore patient characteristics associated with such a disturbance. The secondary aim was to see if there was a correlation between the development of a disturbed oropharyngeal microbiota and hospital-associated pneumonia.

Study IIITo investigate how Lp299 and Lp299v affect growth of nosocomial oropharyngeal pathogens in vitro and to evaluate the possibility to influence the oropharyngeal microbiota in vivo when these lactobacilli were administered prophylactically in hospitalised ward patients.

Study IVTo study if and how treatment with Lp299 and Lp299v affects the oropharyngeal microbiota in patients undergoing extended head and neck cancer surgery, during their hospital stay. The secondary aim was to study if administration of Lp299/299v could reduce the risk for postoperative infections.

41

42

Materials and methods

Details of materials and methods used in the present thesis are given in the separate studies.

EthicsAll clinical studies were approved by the Regional Ethics Committee in Lund, Sweden. Written informed consent, including the permission to collect and publish anonymous data from ward patients and healthy controls were obtained before inclusion in study I-IV. Regarding the ICU patients in study I, written informed consent, including the permission to collect and publish anonymous data, was obtained from the patients or their relatives before they entered the study. In all studies, the patients were given the opportunity to opt out at any time.

It is an ethical dilemma to include ICU patients in clinical studies due to their frequent lack of possibility to give an informed written consent before inclusion. In those cases, as written above, the written consent was given by a relative (we tried to inform and receive a consent from the patient as soon as possible after inclusion). The same dilemma occurred when including a patient who is scheduled for surgery the next day. The patient might not be able to give their full attention and focus on the information about the study and what it implies. We have considered this aspect in all studies. We believe that the oropharyngeal swabs may cause some discomfort, but no pain. Some patients found the intervention mixture ill-tasting, but otherwise we did not cause the patient harm or put them in danger.

Study populationsAll our studies are clinical trials, performed on patients after their informed written consent. As such we have followed Good Clinical Practice (GCP) standards, for example using deidentified standard case report forms (123). One of our two research nurses is a certified monitor of clinical studies and aided in ensuring the practice of respectable research in an ethical and safe manner.

43

Study I: Patient selection and patient dataBetween May 2010 and April 2016 patients (> 18 years) were included using the following inclusion criteria for the three different study groups: (1) critically ill patients admitted to the hospital’s ICU (mixed surgical and medical ICU), requiring mechanical ventilation for at least 24 h, (2) patients admitted to acute medical or surgical wards, not requiring intensive care and (3) controls in the community who were neither hospitalised nor medicated with antibiotics during the previous 2 months.

Patient data was recorded in a standardised form. Patient characteristics are listed in Paper I. The oropharyngeal samples were collected within 24 h of admission to the hospital (ward patients) or intensive care unit (ICU patients). Oropharyngeal swabs were collected on the respective inclusion sites for controls.

Study II: Patient selection and patient dataBetween February 2014 to February 2017, patients were enrolled in the study using the following inclusion criteria: age ≥ 18 years; possible to obtain the first oropharyngeal swab (OPS) within 24 h of hospital admission; expected LOS of > 72 h. The exclusion criterion was hospitalisation in the preceding 14 days. The patients were identified and enrolled at nine different wards, and patients who changed wards during their stay were still eligible to remain in the study, and swabs were collected according to protocol. A standardized case report form was used to record the patient data which are presented in Paper II. The first OPS was collected after inclusion on hospital admission (day 1) and the procedure was repeated on day3 and approximately every fourth day thereafter throughout the patient’s entire LOS. We chose to conduct this study on ward patients, since very few studies have focused on this most common population of hospitalised patients. With our background from the ICU and infectious medicine, we saw the potential to extrapolate data from the ICU studies, seek out differences between ward-patients and ICU-patients and contribute to increased awareness of aetiology and treatment of NV-HAP.

Study III: Patient selection, patient data and randomizationBetween September 2014 to May 2019, patients were included and randomized, using the following inclusion criteria: age ≥18 years, obtaining the first oropharyngeal swab (OPS) within 24 hours of hospital admission, and an expected length of stay of >72 hours. The patients were admitted to medical, surgical or orthopaedic wards. A standardized case report form was used to record patient data, which are presented in Paper III. The randomization was performed directly after inclusion via sealed envelopes at a 1:1 ratio. The randomization was blinded to

44

recruiters, staff and patients. Patients received either a combination of 1010 CFU Lp299 and 1010 CFU Lp299v together with 3 grams of maltodextrin, or placebo consisting of only 3 grams of maltodextrin. Both lactobacilli and placebo were manufactured and generously provided by Probi AB, Lund, Sweden, and delivered in identical freeze-dried sachets labelled “A” and “B”, respectively. The sachets were kept in a -80°C freezer until use. In the ward, the sachets were kept at 4°C for a maximum of five days. Viability controls of the lactobacilli in the sachets were performed yearly throughout the study period, analysing sachets stored at both - 80°C and 4°C. Before administration to the patient, the contents of the sachets were resuspended in 15 mL of sterile water, allowing revival of the potential lactobacilli for 20-40 minutes before administering the mixture to the patient. Patients received the assigned mixture twice daily throughout the hospital stay, with instructions to gargle the mixture as long as possible and then swallow. OPSs were taken at inclusion (day 1), on day 3, and thereafter approximately every fourth day.

Our motive for this study was that probiotics have shown beneficial against VAP. Even if the risk factors and aetiology are not the same regarding VAP and NV-HAP, there is strong reason to believe that probiotics could be a valuable adjunct also among ward patients, to prevent NV-HAP, and maybe also other nosocomial infections (SSIs and Clostridium difficile infection).

Study IV: Patient selection, patient data and randomizationBetween September 2012 and November 2016 at Skåne University Hospital in Lund, Sweden, patients were included and randomized, using the following inclusion criteria: patients aged ≥18 years scheduled for an elective hospital admission for extensive head and neck cancer surgery lasting a minimum of 1.5 hours and demanding tracheal intubation. The randomization was performed on the day before surgery via sealed envelopes at a 1:1 ratio. A block size of six was used to stratify the randomization based on the treatment regimen in order to prevent uneven distribution of treatment regimens between the intervention group and the routine group. The intervention dose of probiotics, regardless of administration route, consisted of 1010 colony-forming units (CFU) Lp299 and 1010

CFU Lp299v. Small plastic sachets containing this above-mentioned amount of freeze-dried Lp299 and Lp229v together with 3 g of maltodextrin were manufactured and generously provided by Probi AB (Lund, Sweden), and stored at -80°C until use. In the ward, the sachets were kept at 4°C for a maximum of seven days. Viability controls of the lactobacilli in the sachets were performed yearly throughout the study period, analysing sachets stored at both -80°C and 4°C. Before administration to the patient, the contents of the sachets were re-suspended in 15 mL of sterile water, allowing revival of the lactobacilli for 20-40 minutes before administering the mixture to the patient. The routine group received standard care. A standardized

45

case report form was used to record patient data and events during the hospitalisation, which are presented in Paper IV.

In the studies concerning probiotics and surgery, it seems that probiotics have their most clear valuable effect in upper abdominal surgery (and liver transplantation) (63). We therefore believe that probiotics may be beneficial to patients undergoing advanced intraoral head and neck surgery, especially since this patient selection often receives a tracheostomy, has damaged local tissue in the URT (as well as a disturbed microbiota), and stay in the hospital for a long time receiving broad- spectrum antibiotics. Both risk for SSI, NV-VAP and Clostridium difficile infection might be diminished by probiotic treatment, in this patient group.

Oropharyngeal sample collectionSterile Nylon® flocked swabs with 1 mL liquid Amies medium (ESwabTM 480C, COPAN Diagnostics Inc., Murrieta, CA, USA), were used to collect the samples. A tongue depressor was used to gain access to and visualize the pharynx. The swab was inserted into the posterior pharynx and rubbed over both the tonsillar pillars and the posterior oropharynx, carefully avoiding touching the tongue, teeth and gums. The sample was then transported to the Department of Clinical Microbiology for cultivation analyses.

Microbiological procedures and definitionsThe OPSs were processed by extended microbiological procedures at the Department of Clinical Microbiology, Skåne University Hospital in Lund. The laboratory is accredited by the accreditation body (SWEDAC) designated by the Swedish government and is formally recognised as competent according to European and international standards.

For bacteria cultivation, the sampling media was inoculated on five types of agar plates (three selective, one differentiating, and one non-selective). All plates were produced in-house, sometimes using commercially available media components (5% horse blood, hematin agar, and Uriselect 4 agar), as listed below:

1) agar with 5% horse blood (LabM, Heywood, Lancashire, UK) supplemented with 10 mg/L colistin and 15 mg/L nalidixic acid with an optochin disc (selective);

2) agar with 5% horse blood supplemented with 2 mg/L gentamicin and 25 mg/L nalidixic acid for Gram-positive cocci including Streptococcus pneumoniae (selective);

46

3) hematin agar (Oxoid™, Thermo Science, Basingstoke, UK) supplemented with 300 mg/L bacitracin for fastidious Gram-negative rods including Haemophilus influenzae (selective)

4) Uriselect 4 agar (Bio-Rad Laboratories, Copenhagen, Denmark) supplemented with 10 mg/L vancomycin for non-fastidious Gram-negative rods (differentiating)

5) hematin agar with a colistin disk (non-selective)

Figure 12: Example of a culture result showing normal oropharyngeal microbiota. Agar plate with top half consisting of agar with 5% horse blood (LabM, Heywood, Lancashire, UK) supplemented with 10 mg/L colistin and 15 mg/L nalidixic acid with an optochin disc (selective, and bottom half with hematin agar with a colistin disk (non-selective. Photo taken by the author at the Department of Clinical Mirobiology in Lund.

The plates were inspected for growth after 16 and 40 hours of aerobic, anaerobic, or CO2 incubation at 35–37°C. If an inspection result was ambiguous at 40 hours, the plate was incubated for an additional 24 hours to obtain a more definite result. Species identification of bacteria was performed using matrix-assisted laser desorption/ionisation time-of-flight (MALDI-TOF) mass spectrometry (MALDI Biotyper Microbial Identification System, Bruker, Boston, MA, USA), using software FlexControl 3.4 and MALDI Biotyper (MBT) Compass 4.1, with MBT Compass Library, DB-7854 MSP (Bruker, Bremen, Germany).

Cultivation and differentiation of Candida spp. was based on colony appearance on CHROM Candida agar (CHROMagar, Hägersten, Sweden) after 48 hours of incubation at 35°C.

For a sample to be considered representative of ‘oropharyngeal microbiota’, bacterial species normally found in the oropharynx were required to grow on the non-selective haematin plate as determined by visual inspection by an experienced

47

senior microbiologist and in accordance with standard practice (124). Figure 12 is a photo of a plate showing growth of normal oropharyngeal microbiota. In the oropharynx, the genera most commonly found are Streptococcus, Prevotella, Capnocytophaga, Rothia, Campylobacter, Veillonella, Neisseria, and Haemophilus (34, 124), followed by a large group of less common genera. The plates were subsequently inspected for signs of a disturbed oropharyngeal microbiota, which required growth of species not normally found in the oropharyngeal cavity or overgrowth of normal oropharyngeal species and/or overgrowth of yeast species. Samples were classified as disturbed oropharyngeal microbiota when there was growth of species not normally found in the oral cavity and/or overgrowth of normal oropharyngeal microbiota on the selective and differentiating plates. Samples with a disturbed oropharyngeal microbiota were divided with the aid of both an infectious disease specialist and the Department of Microbiology into three subclasses

1) gut pathogens,

2) respiratory tract pathogens and

3) yeast.

As far as we know, we are the first group to use this subclassification. The rationale for the subclassification being that maybe different pathogenic mechanisms lies behind the development of a disturbed oropharyngeal microbiota depending on the subclass behind it. We speculated that a disturbed oropharyngeal microbiota consisting of gut pathogens might be due to fasting, feeding tubes, sedation, PPI medication, while maybe broad-spectrum antibiotics and lack of hygiene standards in the personal caring for the patient leads to an overgrowth of respiratory pathogens in the oropharynx.

In vitro experimentsBacterial strainsLp299 and Lp299v were provided by Probi AB, Lund, Sweden. Reference strains of the pathogens Escherichia coli, Staphylococcus aureus, Enterococcus faecalis, Enterococcus faecium, Klebsiella pneumoniae. Pseudomonas aeruginosa and Enterobacter cloacae were purchased from Culture Collection, University of Gothenburg, Sweden. Clinical isolates came from the Department of Clinical Microbiology at Skåne University Hospital, Sweden. These pathogens were chosen for testing against lactobacilli in study III, as they were the most frequently found pathogens in the oropharyngeal cultures from Study I. Therefore, we also tried to find clinical isolates from the same hospital (ICU and ward), with success regarding some strains. As mentioned above, each country and each hospital have its own unique assortment of strains and resistance patterns, and we wanted to consider this

48

when testing the strains in vitro and then treating the patients with Lp299 and Lp299v.

Growth conditionsLp299 and Lp299v were grown in De Man-Rogosa-Sharpe (MRS) broth (Merck, Darmstadt, Germany) and on MRS agar. E. faecalis, E. faecium and S. aureus were cultured in Todd Hewitt (TH) broth (BectonDickinson, Sparks, MD, USA) and agar, whereas E. coli, K. pneumoniae, P. aeruginosa and E. cloacae were grown in Luria Bertani (LB) broth (Sigma-Aldrich, St. Louis, MO, USA) and agar. All strains were cultured at 37°C under aerobic conditions (21% oxygen, 5% CO2).

Agar overlayOvernight cultures of Lp299 and Lp299v were washed in Phosphate-buffered saline (PBS) and adjusted to final concentrations of approximately 2 × 109

CFUs/mL. Varying amounts of Lp299 or Lp299v were added to 8 mL of warm (42–45°C) MRS agar, and poured into Petri dishes. Control plates contained no lactobacilli. After solidification, this bottom agar was incubated at 37°C overnight. A second layer of agar (24 mL), suited for the pathogen, was then cast on top of the MRS agar. Overnight cultures of the pathogens were diluted in PBS, and 10 μL drops of the dilutions were seeded on the top agar. After overnight incubation at 37°C, growth of the pathogen was assessed. The experiments were repeated twice using reference strains and once with clinical isolates of the pathogen.

Inhibitory activity of Lp299 and Lp229vCo-cultures of lactobacilli and pathogens were grown in a mixed broth consisting of 25% (v/v) MRS and 75% (v/v) TH or LB broth. These proportions provided good growth conditions for both lactobacilli and pathogens. Overnight cultures of Lp299, Lp299v and pathogen strains were washed and adjusted to bacterial suspensions of 2 × 109 CFU/mL, and 50 μL of the pathogen and 500 μL of Lp229 or Lp299v were added to 10 mL of mixed broth. As a control, 50 μL of the pathogen suspension was grown in mixed broth in the absence of lactobacilli. The co-cultures were incubated for 5 hours at 37°C. Before and after incubation, a small aliquot of each sample was diluted in PBS and plated on 15-20 mL agar suitable for the pathogen. After incubation overnight at 37°C, colonies of the pathogens were counted, and growth of lactobacilli was ensured. Experiments were performed in triplicate.

49

Antibacterial activity of Lp299 and Lp299v supernatantsOvernight cultures of Lp229 or Lp299v were pelleted by centrifugation, and the supernatants were sterile filtered through a 0.22 μm Millipore Express® PES Membrane Filter. The cell-free supernatants were then either pH-neutralized with 1M NaOH to a pH of 5.4 (corresponding to the natural pH of MRS broth); heat- treated at 99°C for 5 minutes; or incubated with pepsin (Sigma-Aldrich, St. Louis, MO, USA), proteinase K (Thermo Scientific, Waltham, MA, USA) or trypsin (Sigma-Aldrich, St. Louis, MO, USA) at a final concentration of 1 mg/mL for 2 hours at 37°C. The samples were thereafter heated to 99°C for 5 minutes to eliminate the protease activity. The inhibitory effect of the supernatants was tested against E. cloacae and S. aureus. These two were chosen as they were clearly inhibited by Lp299 and 299v and they differ in gram staining and natural habitat. 2.5 mL of pH- neutralized, heat-treated or protease-treated supernatant was added to 7.5 mL of TH or LB to obtain a mixed broth. Untreated supernatant, sterile MRS broth and MRS adjusted to pH 4.2 with acetic acid were included for comparison. Pathogens were washed and diluted as described in the co-culture experiment, and 50 μL of pathogen solution was added to the mixed broths. As a control, pathogens were incubated with Lp299 or Lp299v in a mixed broth with sterile MRS. Samples were plated on agar before and after incubation for 5 hours as described above. After incubation overnight at 37°C, colonies of the pathogens were counted, and growth of lactobacilli was ensured when relevant. Experiments were performed in triplicate.

Statistical analysisStatistical analyses were performed using IBM SPSS Statistics 26 for Windows (IBM Corp., Armonk, NY, USA). P < 0.05 was considered significant, and all statistical tests were two tailed.

Study I-IV Patient characteristicsContinuous variables were presented as median, minimum, and maximum values. Dichotomous variables were presented as number and as percentage of total number. Group wise comparison was done using Fisher’s exact test for categorical variables and Mann–Whitney U-test for continuous variables.

Study IUnivariable logistic regression analyses was used to analyse associations between the dependent variables and independent variables. A stepwise multivariable

50

logistic regression analysis was used to determine independent variables' possible influence on outcome.

Study IIUni- and multi-variable logistic regression analyses were performed for those patients who developed a disturbed oropharyngeal microbiota during hospitalisation (excluding the patients that had a disturbed microbiota at admission). Univariable logistic regression was used to analyse the association between independent variables and the dependent variable. Thereafter, a multivariable logistic regression model using the three strongest predicting variables from the univariable analysis was constructed in which one additional potential explanatory variable was added to determine whether the model improved or did not improve by including a fourth variable. Fisher’s exact test was used to assess the relationship between potential risk factors and HAP. Cox regression analyses of the data gave no additional information regarding the development of a disturbed oropharyngeal microbiota and its correlation to LOS.

Study IIIInhibitory effects in vitro were analysed using student’s paired T-test. Univariable Poisson regression was used to analyse the association between the patients’ predicting variables and the intervention they were randomized to, the dependent variable. A multivariable Poisson regression model using the two strongest predicting variables from the univariable analysis was constructed, in which one additional potential explanatory variable was added to determine whether the model improved or did not improve by including a third variable. A Kaplan-Meyer analysis was performed to test for differences between the placebo and the lactobacilli group regarding ‘time to first disturbed oropharyngeal swab’. Fisher’s exact test was used to assess the relationship between intervention group and nosocomial infection rate.

Study IVKaplan-Meier estimate was applied to analyse potential postponement in occurrence of disturbed microbiota between the routine group and the intervention group. A generalized linear mixed model (GLMM) analysis was used to analyse the potential difference in occurrence of disturbed microbiota at different set observation time intervals, between the two groups. An unadjusted analysis was first performed, after which we added three potentially predicting variables, one at a time, to see how they influenced the model. The predicting variables we added were: PPI medication, median values of number of days on antibiotics and observation time (in days).

51

52

Results

Study IIn this clinical observational, 487 individuals were included. From the ICU, 217 patients were included whereof 167 patients were prospectively included from May 2010 to February 2016, and 50 patients were retrospectively included from previously published material (125). A total of 193 ward patients were prospectively included from January 2014 to December 2016. The 77 control patients were prospectively included from March 2014 to April 2016, and consisted of 30 members of Skåne’s Fire Department (Räddningstjänsten Syd, Lund, Sweden) and 47 members of the Senior Citizens Organisation (Pensionärernas Riksorganisation, Lund, Sweden). See Figure 13.

Figure 13: Individuals included in Study I.

53

The oropharyngeal samples were collected within 24 h of admission to the hospital (ward patients) or intensive care unit (ICU patients), and the results show that a disturbed oropharyngeal microbiota was significantly more common among ICU and ward patients compared to controls (62.2% vs. 1.3%, P < 0.001 and 10.4% vs. 1.3%, P = 0.010), respectively. Colonisation of gut microbiota in the oropharynx was also significantly more common among ICU patients compared with ward patients or controls (26.3% vs. 4.7%, P < 0.001 and 26.3% vs. 1.3%, P < 0.001), respectively. The occurrence of gut microbiota in the ward patients was not significantly different from that seen in controls (4.7% vs. 1.3%, P = 0.29). See Figure 14 and Table 2 for the microbiological results for all three groups.

Figure 14: The oropharyngeal cultures from the three study groups.

54

Table 2: Microbial isolation from oropharyngeal cultures in the three study groups.

OrganismControls Ward patients ICU patients All together

n (%) n (%) n (%) n (%)Respiratory tract pathogens

Staphylococcus aureus 4 (17.4) 26 (14.0) 30 (14.3)Haemophilus influenzae 6 (26.1) 6 (3.2) 12 (5.7)Streptococcus group B 4 (2.2) 4 (1.9)Moraxella catarrhalis 3 (1.6) 3 (1.4)Beta-haemolytic streptococcus group A 2 (1.1) 2 (1.0)Beta-haemolytic streptococcus group G 1 (4.3) 1 (0.5) 2 (1.0)Streptococcus pneumoniae 1 (0.5) 1 (0.5)Staphylococcus epidermidis 1 (0.5) 1 (0.5)

Gut floraEscherichia coli 2 (8.7) 12 (6.5) 14 (6.7)Enterobacter species 1 (4.3) 11 (5.9) 12 (5.7)Enterococcus faecium 11 (5.9) 11 (5.2)Enterococcus faecalis 1 (4.3) 10 (5.4) 11 (5.2)Klebsiella species 1 (4.3) 9 (4.8) 10 (4.8)Pseudomonas aeruginosa 1 (100) 1 (4.3) 7 (3.8) 9 (4.3)Stenotrophomonas maltofilia 8 (4.3) 8 (3.8)Citrobacter species 2 (8.7) 2 (1.1) 4 (1.9)Serratia marcenes 3 (1.6) 3 (1.4)Proteus species 2 (8.7) 1 (0.5) 3 (1.4)Morganella morganii 1 (0.5) 1 (0.5)ESBL 1 (0.5) 1 (0.5)

YeastCandida species 2 (8.7) 63 (33.9) 65 (31.0)

Other origin 3 (1.5) 3 (1.4)Total number of pathogens cultured overall 1 (100) 23 (100) 186 00) 210 (100)

The percentage of positive cultures is related to the total number of positive cultures for all bacterial species in each group. ESBL, Extended spectrum beta-lactamase-producing Enterobacteriaceae.

We analysed the determinants of disturbed oropharyngeal and gut microbiota in the oropharynx using univariable and stepwise multivariable analysis (Table 3a and 3b). We found no significant factors associated with disturbed oropharyngeal microbiota in the ward and ICU patients (Table 3a and b). Restricting the analyses to growth of gut microbiota in the oropharynx, PPI use emerged as the strongest independent predictor in ward and ICU patients in the multivariable analysis (Table 3a and 3b; OR 4.82 and 2.27, respectively). In the ICU group, higher BMI was independently associated with the presence of gut microbiota in the oropharynx (OR 1.08). Diabetes, in contrast, was associated with lower occurrence of gut microbiota in the oropharynx (OR 0.34) in the ICU group. The length of hospital stay (number of days) prior to ICU admission was significantly associated with the presence of gut microbiota in the oropharynx (OR 1.05). Antibiotic therapy ≥ 24 h before inclusion was not associated with changes in microbiota in either the ward or ICU patients. An analysis of risk factors in the control group was not possible due to the low occurrence of disturbed microbiota and gut microbiota (n = 1, 1.3%; Table 2).

55

Table 3a and 3b. Analysing potential predictors of abnormal flora in the oropharynx in (a) ward patients (b) ICU patients.

Factor Univariable analysis Multivariable analysisOR 95% CI P value OR 95% CI P value

(a)Abnormal oropharyngeal flora

Age 1.029 0.991–1.070 0.140 – – –Male gender 0.867 0.342–2.197 0.763 – – –Current or ex-smoker 1.551 0.599–4.018 0.366 – – –Alcohol consumption > 2x per week 0.000 0.000 0.999 – – –PPI use ≥ 24 h before culture sampling 1.551 0.599–4.018 0.366 – – –Antibiotic use ≥ 24 h before culture sampling 1.859 0.488–7.076 0.363 – – –BMI 0.989 0.912–1.072 0.780 – – –Diabetes 0.215 0.028–1.664 0.141 – – –

Gut flora in the oropharynxAge 1.039 0.979–1.102 0.211 – – –Male gender 0.854 0.222–3.281 0.818 – – –Current or ex-smoker 1.829 0.473–7.066 0.382 – – –Alcohol consumption > 2x per week 0.000 0.000 0.999 – – –PPI use ≥ 24 h before culture sampling 4.815 1.162–19.96 0.030 4.815 1.162–19.96 0.030Antibiotic use ≥ 24 h before culture sampling 3.000 0.574–15.67 0.193 – – –BMI 0.939 0.828–1.066 0.331 – – –Diabetes 0.551 0.067–4.557 0.581 – – –

(b)Abnormal oropharyngeal flora

Age 1.006 0.987–1.025 0.543 – – –Male gender 1.483 0.847–2.596 0.168 – – –Current or ex-smoker 0.694 0.384–1.255 0.227 – – –Alcohol consumption > 2x per week 1.154 0.466–2.856 0.757 – – –PPI use ≥ 24 h before culture sampling 1.585 0.872–2.881 0.131 – – –Antibiotic use ≥ 24 h before culture sampling 1.107 0.632–1.937 0.723 – – –BMI 1.009 0.964–1.055 0.712 – – –Diabetes 0.797 0.412–1.542 0.500 – – –APACHE II 1.013 0.975–1.054 0.505 – – –Days in the hospital before ICU admission 1.024 0.991–1.059 0.150 – – –

Gut flora in the oropharynxAge 1.028 1.004–1.053 0.022 – – –Male gender 0.912 0.495–1.680 0.767 – – –Current or ex-smoker 0.981 0.514–1.873 0.953 – – –Alcohol consumption > 2x per week 1.55 0.617–3.891 0.351 – – –PPI use ≥ 24 h before culture sampling 1.767 0.948–3.295 0.073 2.272 1.022–5.054 0.044Antibiotic use ≥ 24 h before culture sampling 1.226 0.657–2.287 0.522 – – –BMI 1.047 0.998–1.098 0.062 1.083 1.021–1.148 0.008Diabetes 0.625 0.287–1.363 0.238 0.338 0.121–0.943 0.038APACHE II 1.038 0.993–1.084 0.098 – – –Days in the hospital before ICU admission 1.047 1.015–1.079 0.004 1.052 1.015–1.090 0.005

The table shows results of the univariable and multivariable logistic regression analyses. In the multivariable analysis, blanks (–) are factors not significantly associated with changes in the oropharyngeal flora after stepwise regression. PPI, proton pump inhibitor; BMI, body mass index; APACHE II, Acute Physiology and Chronic Health Evaluation II; OR, odds ratio; CI, Confidence Interval.

56

Study IIIn this clinical observational study, 134 patients met all inclusion criteria and contributed a total of 466 OPSs. The median number of OPSs per patient was 3 (2– 11), and all OPSs were representative of oropharyngeal flora. The patient characteristics are presented in Table 2 in Paper II.

Figure 15 presents the microbiological results for the 42 patients who presented with any type of disturbed oropharyngeal microbiota on any of the sampling occasions. Each horizontal row corresponds to a unique patient’s observation time. By following the row from left to right it is possible to see OPS changes over time. Using the colour and species key the figure shows the subclass and pathogen for each disturbed OPS. The upper part of the figure shows the patients not receiving antibiotics during hospitalisation and the lower part shows the patients receiving antibiotics during hospitalisation. We found that the longer the hospital stay, the greater the proportion of collected OPSs with a disturbed microbiota. The majority of OPSs that were collected after day 12 and showed a disturbed oropharyngeal microbiota were subclassed as gut microbiota (Figure 15).

Figure 15: Oropharyngeal swab (OPS) culture results for the 42 patients who had at least one OPS sample with a disturbed microbiota during their hospitalisation.

57

In 119 patients (89%), the first OPS (at admission) was normal. In this group, the univariable analyses showed that antibiotics given before and during hospitalisation predicted development of a disturbed oropharyngeal microbiota (see Table 3 in Paper II). In the multivariable analyses, antibiotics during hospitalisation was the only variable close to being statistically significant for the occurrence of a disturbed oropharyngeal microbiota in this group (P = 0.052). Restricting the univariable analyses to colonisation of gut pathogens showed that PPI medication and antibiotics before hospitalisation were associated with an increased risk of colonisation of gut pathogens, whereas acute hospital admission was associated with a decreased risk, see Table 4.

Table 4. Univariable logistic regression analysis. Positive for gut flora during hospitalisation. Total number of participants 119.

Variable Number of patients OR (95% CI) PAge > 70 62 0.91 (0.33–2.47) 0.846Gender, male 61 1.23 (0.45–3.36) 0.693Smoker 33 0.71 (0.22–2.34) 0.572BMI > 35 11 2.32 (0.55–9.76) 0.249Diabetes 22 2.66 (0.87–8.11) 0.086Alcohol > 2 times/week 73 2.49 (0.77–8.10) 0,129Received antibiotics before hospitalisation 10 4.52 (1.13–18.1) 0.033Proton pump inhibitor medication 39 4.10 (1.44–11.6) 0.008Antibiotics during hospitalisation 60 2.21 (0.77–6.34) 0.141Acute admission to hospital 81 0.23 (0.082–0.66) 0.006Ability to walk two flights of stairs 80 1.32 (0.43–4.01) 0.625Cortisone medication 15 2.34 (0.65–8.37) 0.192Lung disease at admission 24 1.16 (0.34–3.90) 0.814Abdominal surgery during hospitalisation 27 1.90 (0.64–5.67) 0.247

In the best-fitting multivariable regression model, ongoing PPI medication was associated with an increased risk of colonisation of gut pathogens, and acute hospital admission was associated with a decreased risk of developing such a disturbed microbiota. See Table 5.

Table 5. Multivariable logistic regression analysis. Positive for gut flora during hospitalization. Total number of participants 119.

Variable Number of patients OR (95% CI) PReceived antibiotics before hospitalization 10 2.08 (0.41–10.4) 0.375Proton pump inhibitor medication 39 3.12 (1.03–9.44) 0.044Antibiotics during hospitalisation 60 2.10 (0.60–7.28) 0.243Acute admission to hospital 81 0.23 (0.074–0.72) 0.012

58

Both PPI medication and antibiotics before hospitalisation were associated with an increased risk of acquiring HAP. See Table 6.

Table 6. Pneumonia as hospital-acquired infection, n=5. a Fisher’s Exact test

Variable

Pneumonia as hospital-associated infection. yes= 5, no= 114

Panumber (%) number (%)Received antibiotics before hospitalization 4 (80.0) 6 (5.3) <0.001

Proton pump inhibitor medication 4 (80.0) 35 (30.7) 0.039Antibiotics during hospitalisation 5 (100.0) 55 (48.2) 0.057Acute admission to hospital 5 (100.0) 78 (68.4) 0.654

Study III

Lp299 and Lp299v inhibit growth of bacterial pathogens in vitroThe inhibitory effect of Lp299 and Lp299v on other bacteria was first tested in an agar overlay assay, where varying concentrations of lactobacilli were grown in a bottom MRS agar, and the pathogens were seeded on a top agar. Under these conditions, all experiments showed complete absence of pathogen growth compared to control plates without lactobacilli. Both clinical isolates and the corresponding reference strains were tested and gave the same clear results.

Next, pathogens and lactobacilli were co-cultured in broth to further characterize the inhibitory effect. In this experimental set-up, both Lp299 and Lp299v significantly inhibited growth of S. aureus, E. cloacae, K. pneumoniae, E. coli, E. faecium and P. aeruginosa. For two pathogens, E. cloacae and K. pneumoniae, almost complete eradication of the pathogens was seen, as the numbers of colonies were close to zero after incubation. E. faecalis was significantly inhibited by Lp299v, but in co-culture with Lp299 the inhibition did not reach statistical significance (P = 0.12). See Figure 16.

59

Figure 165: Co-culture of different pathogens with Lp299 or Lp299v resulted in significant growth inhibition for all pathogens except Enterococcus faecalis co-incubated with Lp299. Growth of the pathogen alone in the absence of lactobacilli served as control. *P < 0.05

The antibacterial activity of Lp299 and Lp299v is pH dependentTo investigate the mechanism behind the growth-inhibitory effects of lactobacilli,S. aureus was incubated in a mixed broth containing cell-free supernatants from overnight cultures of Lp299 and Lp299v. The pH of the supernatants was 4.1 and 4.0, respectively. Figure 16 shows that the supernatants significantly inhibited growth of S. aureus to the same extent as co-incubation with live bacteria. When the pH of the supernatants was elevated to 5.4, corresponding to the pH of MRS broth, the inhibitory effect was abolished, and S. aureus grew equally well as in the control. Further, MRS broth made acidic to pH 4.0 (same pH as the supernatants) significantly inhibited S. aureus growth to the same extent as the supernatants from both Lp299 and Lp299v (see Figure 17). The same results were obtained when the experiment was performed with E. cloacae.

60

Figure 17: The inhibitory effect of Lp299 and Lp299v on S. aureus is pH dependent. The bars with an asterisk above them indicate a significant growth inhibition compared to the control with S. aureus grown in normal mixed broth without lactobacilli. SN = supernatant. Neu = neutralized. *P < 0.05.

To examine the possible role of plantaricins secreted by Lp299 and Lp299v, E. Cloacae was incubated with Lp299 and Lp299v supernatants that had been heat- treated to denature the protein content, or pre-incubated with the proteinases pepsin, proteinase K or trypsin to digest proteins in the supernatants. All of the pre-treated supernatants showed the same clear growth inhibition as untreated supernatants, whereas the controls incubated with MRS broth showed expected growth of the pathogen during the incubation time (see Paper III, Appendix Figure A2).

Taken together, the overall conclusion of these experiments is that the inhibitory effect was mainly pH dependent.

Randomized controlled trialIn this RCT 117 patients met all inclusion criteria and contributed a total of 337 OPSs. The median number of OPSs per patient in both groups were 3. All OPSs were representative of oropharyngeal microbiota. The baseline patient characteristics and hospitalisation characteristics of the placebo and lactobacilli groups were similar and are presented in Table 1 in Paper III.

Figure 18 presents the microbiological results for the 27 patients showing any type of disturbed oropharyngeal microbiota on any sampling occasion. Each horizontal row corresponds to a patient’s observation time, and by following the row from left to right it is possible to see OPS changes over time. Using the colour and species

61

key the figure shows the subclass and pathogen for each disturbed OPS. The upper part of the figure shows the patients receiving placebo and the lower part shows the patients receiving lactobacilli.

Figure 18: Oropharyngeal swab (OPS) culture results for the 27 patients who had at least one OPS sample with a disturbed microbiota during their hospitalisation.

In 104 patients (89%), the first OPS at admission was normal. We analysed results from these patients using univariable and multivariable Poisson analyses. The univariable analyses showed that treatment with lactobacilli yielded an RR of 0.96 (CI 0.36–2.55, P = 0.94) for acquiring a disturbed oropharyngeal microbiota during hospitalisation. Both univariable and multivariable analyses showed that treatment with proton pump inhibitor (PPI), cortisone or antibiotics during hospitalisation could be associated with an added risk of developing a disturbed microbiota during hospitalisation. See table 7.

62

Table 7. Poisson regression analysis for developing a disturbed oropharyngeal microbiota during hospitalisation in 104 subjects with a normal microbiota at admission, (yes n = 14, no n = 90). The number of patients receiving Lactobacilli in this analysis was 53, the control group consisted of 51 patients.

Univariable RR (95% CI)

P Multivariable RR (95% CI)

P

Lactobacilli group 0.96 (0.36‒2.55) 0.938Lactobacilli

Diabetes1.41 (0.49‒4.08) 0,528

0.92 (0.33‒2.51)

1.43 (0.48‒4.31)

0.864

0.523Lactobacilli

PPI2.85 (1.10‒7.39) 0.031

1.10 (0.41‒2.96)a

2.89 (1.08‒7.76)a

0.849a

0.035a

Lactobacilli

Cortisone medication4.22 (1.65‒10.8) 0.0026

1.13 (0.44‒2.90)

4.32 (1.68‒11.1)

0.805

0.0025Lactobacilli

Antibiotics before hospitalisatiom2.59 (0.48‒13.9) 0.267

0.94 (0.36‒2.46)

2.62 (0.48‒14.2)

0.893

0.264Lactobacilli

Antibiotics during hospitalisation3.14 (1.19‒8.30) 0.021

1.17 (0.44‒3.13)

3.23 (1.18‒8.84)

0.753

0.023Lactobacilli

Acute admission0.83 (0.29‒2.43) 0.739

0.97 (0.37‒2.55)

0.83 (0.29‒2.41)

0.952

0.738athe maximum number of step-halvings was reached but the log-likelihood value cannot be further improved. Output for the last iteration is displayed. RR = risk ratio, PPI = proton pump inhibitor

Kaplan-Meyer analyses did not show postponement of a development of a disturbed oropharyngeal microbiota, even if there was slight tendency to later development in the treatment group, but the difference was not significant.

Concerning the risk of developing nosocomial infection during hospitalisation, the difference between the two groups did not reach significance. See Table 1 in Paper III.

Study IVIn this randomised controlled trial total, 36 patients completed the study according to protocol. Table 1 in paper IV shows patient characteristics and type of surgical intervention, with no significant difference between the intervention and routine groups. Table 2 in Paper IV shows hospital events for both groups, including median values regarding perioperative infection, observation time, duration of antibiotic therapy, number of OPSs per patient (=6) and number of TASs (=3), There was no significant difference between intervention and routine groups. All cultured OPSs were representative of oropharyngeal microbiota. Figure 1 presents the microbiological OPS results for each patient and each sampling time. Each row represents a patient and a coloured block means that an OPS has been obtained on the day indicated on the top row. The different colours of the blocks represent

63

different groups of pathogens growing in the OPS. The top half of the patient rows in the figure consists of the control group, and the bottom 18 patients have received intervention. See Figure 19.

Figure 19: Oropharyngeal swab (OPS) culture results for all patients (n=36).

64

65

Using Kaplan-Meier estimate for the comparison of time to first disturbed OPS culture, no significant difference regarding postponement of disturbed oropharyngeal microbiota was found between the routine and the intervention group (overall comparisons; p = 0.944). There is a tendency towards longer median time for occurrence of a disturbed OPS in the Lp299/299v group (8 days versus 7 days in the routine group), but this difference was not significant. See Figure 20.

-

Figure 20: Kaplan-Meier estimate including all 36 patients. The patients are divided into Lp299/299v and routine treatment. Following the curve x-axis timeline, patients are excluded after appearance of a disturbed microbiota in the OPS

Table 8 present the GLMM analyses. There was no significant difference between the groups, in occurrence of disturbed microbiota at the different set observation time intervals, not in the unadjusted nor in the three analyses adjusted for PPI medication, number of days with antibiotic treatment and observation time. On day 2, in the unadjusted analyses, there was a tendency to a higher incidence of disturbed OPSs in the routine group compared to the intervention group, but the difference did not reach significance (P= 0.098).

Table 8. Generalized linear mixed model (GLMM) of the cohort.

Day Culture GLMM

Unadjusted Adjusted for PPI medication

Adjusted for days with antibiotic treatment

Adjusted for observation time (in days)

Routine number (%)

Lactobacilli number (%)

OR (95%)Confidence interval

P OR (95%)Confidence interval

P OR (95%)Confidence interval

P OR (95%)Confidence interval

P

1Normal 13 (87) 16 (94) 0.47

(0.030−7.16) 0.580 0.48(0.028−8.36) 0.612 0.43

(0.028−6.62) 0.543 0.49(0.031−7.82) 0.611

Disturbed 2 (13) 1 (6)

2Normal 8 (61) 14(87) 0.18

(0.024−1.39) 0.098 0.19(0.024−1.50) 0.113 0.22

(0.027−1.82) 0.157 0.19(0.024−1.51) 0.114

Disturbed 5 (39) 2 (13)

3-4Normal 7 (70) 10 (71) 0.79

(0.15−4.08)0.773

0.79(0.15−4.09)

0.7771.03

(0.19−5.60)0.976

0.77(0.15−3.95)

0.755Disturbed 3 (30) 4 (29)

5-6Normal 2 (40) 9 (82) 0.19

(0.021−1.78) 0.145 0.19(0.018−1.98) 0.162 0.21

(0.023−2.00) 0.173 0.21(0.022−1.97) 0.169

Disturbed 3 (60) 2 (18)

7-8Normal 5 (56) 5 (62) 0.54

(0.085−3.38) 0.501 0.53(0.090−3.09) 0.475 0.58

(0.092−3.65) 0.557 0.54(0.085−3.44) 0.511

Disturbed 4 (44) 3 (38)

9-11Normal 6 (50) 4 (44) 1.04

(0.20−5.38)0.961

1.04(0.19−5.67)

0.9681.22

(0.23−6.61)0.814

1.06(0.20−5.56)

0.944Disturbed 6 (50) 5 (56)

12-14Normal 4 (50) 2 (40) 1.15

(0.18−7.29)0.884

1.26(0.19−8.33)

0.8091.21

(0.18−8.21)0.843

1.16(0.18−7.66)

0.876Disturbed 4 (50) 3 (60)

65

66

67

Discussion

In this thesis, we investigated the oropharyngeal microbiota in hospitalised patients, and in the two interventional studies (III) and (IV), we explored whether the probiotics Lactiplantibacillus plantarum 299 and 299v could postpone or diminish the development of a disturbed oropharyngeal microbiota. In study III and IV, we also investigated if twice daily treatment with Lp299 and 299v could influence the occurrence of hospital-associated infections.

In study I, we examined the oropharyngeal microbiota in (1) critically ill patients admitted to the hospital’s ICU, (2) patients admitted to wards and (3) healthy controls in the community. As we believe that a disturbed oropharyngeal microbiota increases the risk for developing nosocomial pneumonia, by increased aspiration of pathogens, our finding could add to providing adequate preventive measures (for patients displaying risk factors at admission). The study from 1969 by Johanson et al, showed that a disturbed oropharyngeal microbiota correlated best with the degree of sickness in the patient (37). In our study we showed that 62% of the ICU patients had a disturbed oropharyngeal microbiota within 24 hours of admission, and among the ward patients 10% had a disturbed oropharyngeal microbiota already within 24 hours of admission to hospital. Among the controls one participant out of 77 (1.3%) had a disturbed oropharyngeal microbiota. With the aid of the Department of Clinical Microbiology and specialists in infectious diseases, we divided the OPS results into two groups; 1) a disturbed microbiota caused by overgrowth of already locally existing microbiota or 2) a disturbed microbiota exhibiting pathogens not normally present in the oropharynx, but originating from the gut below the ventricle. Restricting the uni- and multivariable logistic regression analyses to growth of gut microbiota in the oropharynx, PPI use emerged as the strongest independent predictor in both ward and ICU patients in the multivariable analysis (OR: 4.82 and 2.27, respectively). In the ICU group, BMI >35 and length of hospital stay (number of days) prior to ICU admission were also independently associated with the presence of gut microbiota in the oropharynx. Diabetes, in contrast, was associated with lower occurrence of gut flora in the oropharynx (OR 0.34) in the ICU group. It is well established that PPI medication is associated with an increased risk for VAP (126-128).

Our findings might contribute to understanding the pathogenesis behind the increased risk for VAP among patients on PPI treatment. By greatly lowering the pH in the ventricle, PPI enables gut pathogens to establish themselves in the

68

oropharynx and subsequently in the lungs. A higher BMI may further increase the risk by making the patient difficult to mobilize into a semi-recumbent position, and by increased abdominal pressure. In many ICUs, the patients are put on PPI medication as a part of an “admission package”. The patient then remains on PPI medication despite successful enteral feeding or being a “low risk” patient for developing bleeding ulcers. Our results support other studies that suggest daily evaluation of the need for continuing PPI medication, or a change to another other type of ulcer prophylaxis (129). We also show that a disturbed oropharyngeal microbiota is present in more than half of patients that are admitted to an ICU, and in 10% of ward patients. A serious disturbance in the patient’s microbiota has been shown to be linked to a poor outcome (37, 130). The limitations of this study include that the some of the data are over 10 years old, and that the definition of a normal oropharyngeal microbiota might have changed during this period. However, the Department of Clinical Microbiology have confirmed that there has been no change in their definition of a normal oropharyngeal microbiota during the last 20 years (in the south of Sweden). In our study the occurrence of MDR pathogens is surprisingly low in in a global perspective but agrees with finding that show a slower emergence of antibiotic resistance in Sweden, compared to Europe (131).

In study II, we studied 134 patients admitted to different wards. To our knowledge, we are the first to have performed consecutive sampling in a mixed cohort of hospitalised ward patients As in Study I, we found that around 10% (11%), had a disturbed oropharyngeal microbiota within 24 hours of admission. Univariable logistic regression analyses of the entire cohort showed that antibiotic treatment > 24 hours before hospital admission and antibiotic treatment > 24 hours during hospitalisation predicted a disturbed oropharyngeal microbiota. In the multivariable analyses only antibiotic treatment > 24 hours during hospitalisation could be associated with an increased risk for developing a disturbed oropharyngeal microbiota. These findings are hardly surprising since we know that antibiotic treatment (especially lasting several days), affects the existing microbiota (132, 133). These patients may also have had reasons for receiving antibiotic treatment that increase the risk for both a disturbed oropharyngeal microbiota and HAP, for example remaining bedridden for a prolonged period, needing reoperation and enduring prolonged fasting and pain. As in Study I, we divided the OPS results into two groups; 1) a disturbed microbiota caused by overgrowth of already locally existing microbiota or 2) a disturbed microbiota caused by gut pathogens. Restricting the univariable analyses to colonisation of gut pathogens revealed that PPI medication and antibiotics > 24 hours before hospitalisation were associated with an increased risk of colonisation of gut pathogens, whereas acute hospital admission was associated with a decreased risk. In the best-fitting multivariable regression model, ongoing PPI medication was associated with an increased risk of colonization of gut pathogens, and acute hospital admission was associated with a decreased risk of developing such a disturbed microbiota. The risk of developing

69

HAP was increased among patients treated with PPI and/or antibiotics before their hospital admission, but the numbers were small (N = 5).

To summarize, the first two observational studies showed that ongoing PPI medication increases the risk for developing a disturbed oropharyngeal microbiota consisting of gut pathogens, and that PPI was also associated with an increased risk for developing HAP. These observations corroborate previous studies showing that PPI medication in addition to being a risk factor for VAP, also contributes to changes in oropharyngeal microbiota (134). We are the only group to have used the subcategorization of respiratory tract pathogens, yeast and gut pathogens when studying changes in oropharyngeal microbiota. Based on our results, we believe that it is rational to differentiate between types of pathogens, when trying to understand potential causes and initiating interventions. We think that patients that were admitted acutely had a relatively low risk for developing a disturbed oropharyngeal microbiota, since they may have been quite healthy, except for having a hip fracture or acute appendicitis, in contrast to the patients admitted for elective surgery, in whom underlying chronic disease may have affected their general health for an extended period of time. The biggest limitation of this study is the lack of following a larger cohort for a longer period of time. Changes in oropharyngeal microbiota and development of HAP are complications that take time to develop. Having the study from 1969 (37) as a reference as well as underestimating the efficacy of modern Swedish healthcare, we misjudged both the needed size of the cohort and the efficient early hospital discharge (only a small portion of our cohort were hospitalised for more than a week). Considering the insights gained in Study I and II, we believe that there is an incentive for the establishment of interventional ‘respiratory care bundles’, designed for ward patients, which may include intentional discontinuation of PPI treatment, vigorous physiotherapy, vigilant antibiotic stewardship and treatment adjuncts such as probiotics to help maintain a healthy microbiota.

In study III, we found that Lp299 and Lp299v significantly inhibited in vitro growth of nosocomial pathogens commonly found in the oropharyngeal tract of hospitalised patients. In the randomized controlled trial, no difference between the intervention group and the placebo group could be found regarding changes in the oropharyngeal microbiota or in the occurrence of nosocomial infections. The study confirmed already known risk factors for development of a disturbed oropharyngeal microbiota (135). The ability of lactobacilli to inhibit pathogen growth has been shown before (136). This has also been specifically shown for Lp299v, when Hutt et al. demonstrated its antagonistic effect on Salmonella enterica and Helicobacter pylori (137). In the present study, we show for the first time that Lp299 and Lp299v inhibit in vitro growth of pathogens known to cause nosocomial respiratory tract infections. Notably, over half of the ICU patients were colonized with at least one of the seven investigated pathogens as early as 24 hours after admission to our ICU (Study I) (138). The L. plantarum species are shown to have a high production of lactic acid

70

compared to others in the lactic acid bacteria group, and a relatively small production of, for example, hydrogen peroxide and carbon dioxide (136). In our study, the acidic environment produced by the lactobacilli was essential for inhibiting the in vitro growth of the studied pathogens. However, other factors may also be involved. For example, it has been shown that an acidic pH is necessary for other inhibitory mechanisms to be activated. Several studies on plantaricins show that they are activated at a pH <5 (107, 139)). Although we were unable to demonstrate a plantaricin effect in our study, we cannot rule out a possible role of plantaricins in our strains due to the overwhelming effect of acidic pH. Further experiments are required to determine the possible presence and requisites for activity of bacteriocins in Lp299 and Lp299v.

In the clinical trial, we could not show that oral administration of Lp299 and Lp299v prevents or delays the occurrence of a disturbed oropharyngeal microbiota in non- ICU hospitalised patients compared to placebo. In agreement with our two earlier observational studies (Studies I and II) (40, 138) we found that 11% of the patients had a disturbed oropharyngeal microbiota at admission, and that an increasing proportion of the patients developed a disturbed oropharyngeal microbiota during their hospitalisation. Aside from confirming previously reported findings that treatment with PPI and antibiotics are risk factors for disturbed oropharyngeal microbiota, (135, 138), this study showed that oral cortisone was strongly associated with a disturbed oropharyngeal microbiota (P = 0.0025 in the multivariable Poisson regression), which has not been shown before. It is well known that steroid inhalation treatment increases the risk of developing oral candidiasis, but in our study, none of the patients who were on cortisone treatment developed a disturbed microbiota consisting of candida species (, and only a third was simultaneously on PPI medication).

This study is unique, as it focuses on ward patients. The idea of giving hospitalised patients probiotics is tempting, in many ways. Probiotics are harmless, inexpensive and may have the possibility of reducing antibiotic use by helping to maintain or restore the patient’s normal microbiota When swallowed and digested, the probiotics also influence the intestinal tract immune system, for example by stimulating the production of SCFAs (82). As in study II, the greatest weakness in the clinical trial is that our power calculation was based on old studies with longer hospital stays (37). Consequently, the study was underpowered, and we can therefore not fairly rule out our hypothesis that treatment with lactobacilli can decrease or delay the incidence of a disturbed oropharyngeal microbiota during hospitalisation. An additional explanation is that changes in the oropharyngeal microbiota take time, and that a potential contribution to the development of pneumonia takes even longer. Even if Lp299 and Lp299v showed clear growth inhibition of pathogens in vitro, we might need longer treatment times and longer local exposure to be able to influence the oropharyngeal microbiota in patients. Thus, it cannot be excluded that a study involving a larger patient population and

71

more intense administration of a combination of probiotics would show an effect on a clinically meaningful level.

In study IV, a randomized controlled pilot trial on patients undergoing extensive head and neck cancer surgery, there was neither a difference in the occurrence of disturbed oropharyngeal microbiota nor in post-operative infection in patients receiving Lactiplantibacillus plantarum 299 and 299v as compared to the control group that received standard care. Our findings disagree with previous results from abdominal surgery cohorts where perioperative treatment with pro and/or synbiotics reduced the risk for postoperative infection, shortened antibiotic treatment, and attenuated a systematic inflammatory response (73, 74, 140, 141) In a review by Jeppsson et al, it is concluded that probiotics play an important role in conjunction with surgery of the upper gastro- intestinal tract, with a three-fold reduction in postoperative infections following liver transplantation or upper abdominal surgery (63). In our study, the Kaplan-Meier estimate showed a small tendency towards longer median time for occurrence of a disturbed OPS in the Lp299/299v group (8 days versus 7 days in the control group), but this difference was not significant. Similarly, in the GLMM analyses there was no significant difference in occurrence of disturbed microbiota between the groups at the different time intervals. On day two, in the unadjusted analyses, there was a tendency to a more disturbed OPS in the control group compared to the intervention group, but the difference did not reach significance (P= 0.098).

The Lp299/299v treatment may have had beneficial effects that were not explored in our study, such as a restored and normalised gut microbiome and effect on the systemic inflammatory response. In their meta-analysis on surgical patients Wu et al conclude that probiotics lowered hospital costs and improved “Gastro-intestinal Quality of Life” (73). Albeit ambitious, our study has (several) shortcomings. The most evident is the small study cohort. Even at our specialised tertiary hospital, we perform this type of surgery only two-four times a month, and it was not possible to continue recruiting patients for more than four years. With a larger cohort, we might have been able to achieve clearer and maybe even beneficial results regarding the effect of probiotics. In our statical analysis there are tendencies leaning towards a beneficial effect of the intervention in both occurrence and postponement of a disturbed oropharyngeal microbiota. (The study could also have been better designed in order to avoid criticism for potential bias.) Ideally, the treatment should have blinded and the control group should have received a placebo-treatment of similar physical appearance as the probiotic mixture given to the intervention group. (Another shortcoming is not having measured secondary outcomes related to gastrointestinal function.)

In conclusion, despite our mainly negative results in the interventional studies constituting this thesis, I do believe that probiotics have a place as an important adjunct in treatment in certain categories of ward patients (for example in those with long hospital stays, antibiotic exposure and PPI medication), and that “HAP-

72

prevention care bundles” should include probiotics. Probiotics are inexpensive and harmless - nosocomial infections and prolonged hospitalisation are costly.

Regarding critically ill patients and patients undergoing abdominal or upper abdominal/oral surgery, I believe that probiotics may contribute to help maintain a healthy microbiota and help prevent SSIs, Clostridium difficile infection, VAP and maybe aid to prevent an increase in gut permeability by helping the colonocytes maintain their integrity and tight junctions.

Is Lp299v and Lp299 the right choice? Are the administration solutions and dosages optimal? Probably a combination of several strains of probiotics would be better? These are questions that are important and need to be further elucidated. Further, larger studies are needed and warranted because increased use of antibiotics and hospital complications should be avoided as far as possible.

73

Main conclusions

Study IThis study demonstrates an early and significantly increased incidence of disturbed oropharyngeal microbiota in both ward and ICU patients as compared to controls in the community. The incidence of disturbed oropharyngeal microbiota was also significantly higher in ICU patients as compared to ward patients. In critically ill ICU patients, early oropharyngeal colonisation by gut pathogens was common. PPI medication was the strongest independent factor associated with the presence of gut pathogens in both ward and ICU patients.

Study IIIn this study, a significant proportion of ward patients developed a disturbed oropharyngeal microbiota during their hospitalisation. In patients with a long LOS, gut pathogens most commonly caused the disturbance. The risk of developing HAP was increased among patients treated with PPI and/or antibiotics before their hospital admission PPI medication was a strong independent risk factor for colonisation of gut pathogens in the oropharyngeal tract.

Study IIIThis study shows that Lp299 and Lp299v inhibit in vitro growth of commonly found nosocomial pathogens in the oropharynx. Oral administration of Lp299 and Lp299v to non-ICU patients did not reduce the risk of disturbed oropharyngeal microbiota or nosocomial infection.

Study IVIn this randomized controlled pilot trial on patients undergoing extensive head and neck cancer surgery, there was neither a difference in the occurrence of disturbed oropharyngeal microbiota nor in post-operative infection in patients receiving Lp299 and Lp299v as compared to the control group that received standard care.

74

75

Populärvetenskaplig sammanfattning

BakgrundMänniskan lever i symbios med tusentals olika mikrober (bakterier och svampar), som framför allt finns i vårt tarmsystem, men även på huden, i luftvägarna och i underlivet. Bakterierna i de olika delarna i vår kropp påverkar hur vi mår på en mängd olika sätt, och vi är troligen mer samberoende av våra bakterier än vi tidigare förstått. En störning i bakterieuppsättningen kan öka risken för vissa sjukdomar och påverka vårt immunsystem. De mikrober som förekommer i en människa kallas för den individens mikrobiom. Mikrobiomet ser olika ut i kroppens olika delar. Tex så består mikrobiomet i tjocktarmen av helt andra bakterier än de som finns på vår hud eller i luftvägarna.

Hos sjukhusvårdade och framförallt kritiskt sjuka patienter kan mikrobiomet i det bakre svalget (orofarynx) förändras. Istället för streptokocker, och andra för orofarynx normala bakterier, har bakterier som inte bör finnas där, t ex bakterier från tjocktarmen, etablerats eller så har en bakteriesort tagit all plats och näring och tryckt undan bakterier, som bör finnas för att mikrobiomet ska vara i balans. Sjukdom medför att många sjukdomsalstrande bakterier (patogener) blir mer aggressiva då kroppen är stressad och) intaget av mat är litet eller obefintligt. Detta märks särskilt på uppsättningen av bakterier i tarmen hos kritiskt sjuka patienter på intensivvårdsavdelning.

Probiotika, definieras av WHO som "levande mikroorganismer som, när de administreras i adekvata mängder, ger en gynnsam hälsoeffekt”. Det finns flera olika sorters probiotika (”snälla bakterier”) som på olika sätt kan påverka människors hälsa i positiv riktning. Bland annat kan probiotiska bakterier hjälpa till att upprätthålla och återställa ett rubbat mikrobiom. Det finns vetenskapliga belägg för att probiotika minskar risken för tex sårinfektion hos opererade patienter, diarré efter antibiotikabehandling samt lunginflammation hos intensivvårdspatienter som respiratorbehandlas.

Vårdrelaterade infektioner är en allvarlig komplikation till sjukhusvård, och sjukhusförvärvad lunginflammation är en av de vanligaste och farligaste infektioner man kan drabbas av som inneliggande patient. De flesta studierna om diagnostik, riskfaktorer och behandling av sjukhusförvärvad lunginflammation, är genomförda på intensivvårdspatienter. Väldigt få studier finns på vårdavdelningspatienter, trots

76

att antalet patienter som får lunginflammation är mycket större hos dessa än hos intensivvårdspatienter. Lunginflammation uppkommer när bakterier från bakre svalget tar sig ner i lungorna, och lungornas försvarsmekanismer, som normalt sätt klarar av att göra sig av med oönskade bakterier, inte fungerar. Risken för att bakterier ska ta sig ner i lungorna och orsaka lunginflammation ökar när man är fastande, ligger ner, har ont eller inte orkar hosta eller djupandas som vanligt.

I avhandlingens två första studier (I-II) undersökte vi hur mikrobiotan i bakre svalget (de bakterier som växer fram på odlingsplattor när man tagit ett svalgprov) ser ut redan första dygnet dygn på en vård- eller intensivvårdsavdelning. I delstudie 2 följde vi vårdavdelningspatienter under hela deras vårdtid och tog regelbundna prov från bakre svalget, för att se om mikrobiotan i svalget ändras under sjukhusvistelsen. Vi samlade in information om patienternas hälsotillstånd och medicinering, och noterade antibiotikabehandling, operationer och andra sjukhushändelser. Vi noterade även hur många av patienterna som utvecklade en vårdrelaterad infektion under vårdtiden.

Analyserna visade att 10% av patienterna på vårdavdelning, och 60% av patienterna på intensivvårdsavdelning, hade en störd mikrobiota i bakre svalget redan under det första dygnet. I kontrollgruppen från friska personer ute i samhället hade 1,3 % en störd mikrobiota i bakre svalget. Hos de över 100 patienterna på vårdavdelning ökade förekomsten av en störd mikrobiota i bakre svalget med tiden på sjukhus. I våra analyser letade vi efter specifika riskfaktorer till uppkomst av en störd mikrobiota i bakre svalget. Vi delade upp odlingsresultaten i två grupper: 1) de med överväxt av normalt förekommande bakterier och 2) de med tarmbakterier i bakre svalget. Hos de med tarmbakterier i svalget upptäckte vi att den starkaste riskfaktorn var medicinering med protonpumpshämmare (PPI). Denna koppling såg vi hos både vårdavdelnings- och intensivvårdspatienter. PPI är en medicin mot problem med gastrit eller magsår. PPI sänker syranivån (pH) i magsäcken och möjliggör för bakterier från tarmen att kunna ta sig upp i svalget. En annan riskfaktor för rubbad mikrobiota i bakre svalget var antibiotikabehandling före inläggning. Detta kan förklaras med sjukdomspåverkan redan vid inläggningen, samt att antibiotika förstör det normala mikrobiomet. I studie II såg vi även ett samband mellan utvecklandet av lunginflammation och användandet av PPI och antibiotika. Detta stärker vår tes att en rubbad mikrobiota i bakre svalget kan bidra till uppkomsten av lunginflammation, och att man vid användandet av PPI och antibiotika ska vara medveten om för- och nackdelar samt alternativa åtgärder, både på vårdavdelning och inom intensivvården. Förutom ett strikt förhållningssätt till antibiotikaanvändning bör man hos riskgrupper, även på vårdavdelning, arbeta mer preventivt med att förebygga infektioner. På intensivvårdsavdelningar har man ofta särskilda ”care bundles”, dvs behandlingspaket, som innebär dagligt ställningstagande till viss medicinering (PPI, antibiotika, probiotika); aktiv sjukgymnastik och mobilisering, samt andningsträning, för att motverka uppkomst av lunginflammation. Liknande behandlingspaket anpassade till

77

vårdavdelningssituationen, skulle kunna bidra till att få ner frekvensen av vårdrelaterade infektioner och minska kostnader för både patient och samhälle.

De två sista studierna (III-IV) är interventionsstudier, så kallade randomiserad kontrollerade studier (förkortat RCT efter engelskans randomized controlled trial). I dessa studier har patienterna slumpmässigt (randomiserat) valts till en grupp som fått probiotika eller till en kontrollgrupp. I studie III undersökte vi drygt hundra vårdavdelningspatienter som fick gurgla och svälja probiotika 2 gånger om dagen under hela vårdtiden. Samtidigt tog vi regelbundna odlingar från bakre svalgväggen (precis som i studie I-II). I studie IV, en pilotstudie bestående av 36 patienter från Öron-Näsa-Hals avdelningen, ingick patienter som genomgått stor långvarig kirurgi i huvud- och halsregionen (oftast på grund av munhålecancer). Syftet med dessa två behandlingsstudier var att se om probiotika kunde minska eller skjuta upp förekomsten av störd mikrobiota i bakre svalget och minska risken för vårdrelaterade infektioner, särskilt lunginflammation och sårinfektion. Tidigare studier har visat att probiotikabehandling kan minska uppkomsten av lunginflammation hos intensivvårdspatienter samt minska risken för sårinfektion. Särskilt patienter som genomgått kirurgi i övre delen av tarmsystemet har drabbats av betydligt färre sårinfektioner, när de erhållit probiotika före och under operationen. Inga tidigare studier finns på probiotikas eventuellt positiva effekter avseende infektioner hos patienter på vårdavdelning eller hos patienter som genomgår stor kirurgi i mun och svalg.

Det finns många olika probiotika, med olika egenskaper, Vi valde att använda probiotikastammarna Lactiplantibacillus plantarum 299 och 299v, som ingår i familjen mjölksyrabakterier. De är välkända probiotika, som kvarstår i tarmen trots antibiotikabehandling. De är säkra att använda och har använts kommersiellt under lång tid. De har bevisad effekt mot antibiotikaorsakad diarré (Clostridium difficile infektion), kolon irritabele, kranskärlssjuka och depression samt minskar postoperativ infektion vid stor bukkirurgi.

Studie III består av två delar, dels interventionsstudien på vårdavdelningspatienter, dels en inledande laborativ del, där vi undersökte huruvida Lactiplantibacillus plantarum 299 och 299v kunde hindra eller minska tillväxten hos sju av de bakterier vi oftast fann på de odlingsplattor som visade en störd mikrobiota från bakre svalget (studie I). Vi samodlade mjölksyrebakterierna med de sju potentiellt sjukdomsalstrande bakterierna och fann att Lactiplantibacillus plantarum 299 och 299v hindrar eller avstannar tillväxt hos alla de undersökta sju bakterierna. Vidare laborativa studier visade att den tillväxthämmande effekten var mest beroende av att Lactiplantibacillus plantarum 299 och 299v sänkte pH till den grad att de andra bakterierna inte trivdes. Lactiplantibacillus plantarum 299 och 299v producerar mjölksyra men även andra syror, speciellt när bakterierna är i tillväxtfas. Laktobaciller kan också producera bakteriocider, vilka är små proteiner (peptider), som kan döda bakterier som liknar dem själva. I våra väldigt begränsade försök lyckades vi inte påvisa bakterocinproduktion hos Lactiplantibacillus plantarum 299

78

och 299v, men det utesluter inte att den finns där. Tillväxthämningen som uppkom av den surgjorda miljön var så dominerande att den möjligen dolde andra hämmande mekanismer.

Trots våra laborativa resultat och tidigare, av andra utförda studier som visat att probiotikabehandling medför minskning av lunginflammation hos intensivvårdspatienter samt minskad frekvens av sårinfektion hos patienter som genomgått bukkirurgi, kunde vi inte visa att behandlingen påverkade mikrobiomet i svalget eller risken för infektion. I studie III, på vårdavdelningspatienterna, fanns ingen skillnad i uppkomst av en störd mikrobiota i bakre svalget under vårdtiden, och inte heller en skillnad mellan intervention och placebogrupp när det gäller sjukhusförvärvad lunginflammation eller andra vårdrelaterade infektioner. I studie IV, vår pilotstudie, fann vi heller ingen skillnad mellan gruppen som erhöll probiotika och gruppen som erhöll normal vård. En tendens fanns till fördröjning av uppkomst av en störd mikrobiota i bakre svalget, och dag två efter operationen var det fler i kontrollgruppen som fick en störd mikrobiota i bakre svalget, men skillnaden uppnådde inte riktigt statistisk signifikans.

Tyvärr tror vi att en dominerande förklaring till att vi inte kunde påvisa någon skillnad i våra interventionsstudier är att vi inte hade tillräckligt stora patientgrupper, och att patienterna i studie III inte låg inlagda så länge som vi hade förutspått när vi fastställde studieprotokollet. Svensk sjukvård är idag extremt effektiv och även många äldre åkte hem tidigare än vi antagit att de skulle göra.

Vårdrelaterad infektion är en vanlig komplikation till sjukhusvård. För patienterna innebär det lidande, förlängd sjukhusvistelse och antibiotikaanvändning. För samhället innebär varje vårdrelaterad infektion ökade kostnader och ökad risk för uppkomst av antibiotikaresistens. Probiotika är billiga, ofarliga och har i många studier visat sig kunna påverka vårt mikrobiom på ett positivt sätt, vilket är viktigt eftersom ett normalt fungerande mikrobiom hjälper kroppens immunförsvar och motståndskraft.

I våra kliniska studier kunde vi inte påvisa någon effekt av probiotika, vilket skulle kunna tillskrivas komplexiteten i sjukdomssituationen och ett begränsat patientantal. Det finns starka indicier för att probiotika har gynnsamma effekter och ytterligare studier med ett större patientmaterial är angelägna.

79

Acknowledgements

This thesis has required the aid and interest from many people. Thank you to all of you. I would especially like to express my gratitude to:

Bengt Klarin, the brain behind the project. You have taught me to love (and sometimes hate) clinical research with your limitless enthusiasm and a firm belief in the project. Thank for you for introducing me to the fascinating field of probiotics and the importance of our microbiome.

Lisa Påhlman, my supervisor in this thesis. I admire you to no end. You are wise, pragmatic, cool and an experienced researcher. Beside all that, you are also a clinician and a mother, and you have given me the security and help I have needed to be able to finalize this thesis.

Carolina Samuelsson, my co-supervisor. Your energy is contagious, amazing and seemingly limitless. Thank you for being there when I needed help the most, and for teaching me how to write a scientific paper. And thank you for reminding me of continuing to be curious to learn new things.

Mikael Bodelsson, the head of department at the Institution for Clinical Sciences at Lund University, and the professor of Anaesthesia and Intensive Care at Skåne University Hospital. Thank you for your support throughout the years. You are the man behind the scene, when it comes to my thesis. When I have been stuck, you have nudged me in the right direction on several occasions.

Gunilla Islander, my mentor. You are the epiphany of how doctors used to be, well- educated and multitalented far beyond the scopes of medicine. Thank you for inspiring me to become an anaesthesiologist, for thinking “out of the box”, and for being there always, if I have needed you. You are one of a kind.

Hulda Thorarinsdottir, my friend and colleague. I am grateful that we wrote our first paper together, which was tougher than I thought, and it was nice not be alone all those evenings we spent revising the article. You are kind, funny and very hard working. I admire your dedication and I am thankful to have you as a friend.

Anna Holmberg, co-author of the first paper. Thank you for helping two unknowing anaesthesiologists in our first steps into the world of microbiology. As a bonus we share the same interest for outdoor sports, even if you are a bit more hard-core than I am .e :.:.

80

Ulf Schött, professor at our department and co-author of the first paper. You have not been my official supervisor nor my mentor, but you have felt like one. Always supportive and interested and eager to help or just have a relaxing chat. Thank you.

To past and present heads of the Department of Intensive and Perioperative Care at Skåne University in Lund; Eva Ranklev, Görel Nergelius, Bengt Roth, Marie Martinsson, Carolina Samuelsson and Anders Rehn. To my past and present immediate managers, Ingrid Östlund and Dag Winstedt. Thank you for letting research being an integral part of our department, and for supporting me in completing this thesis.

Anne Adolfsson and Susann Schrey, our invaluable research nurses, who have helped me collect data for the clinical studies. I am forever grateful for your help, and for your commitment and friendship.

Julia Johansson, Gisela Hovold and the rest of the crew at BMC, B14 for helping me with the in vitro part of my thesis. I really enjoyed our time together in the lab.

Maria Celander, medical laboratory scientist at the Department of Clinical Microbiology for helping med with my many questions about cultivation methods, and for showing me how your top-of-the-line laboratory works.

Helene Jacobsson, “my statistician”, from” Kliniska Studier Sverige – Forum Söder”, who has helped me with the statistical analysis in all four papers. Thank you for invaluable help and nice talks throughout the years.

Thank you Probi AB for graciously producing and supplying me with both probiotic and placebo sachets.

I have the pleasure of working at a Department full of nice (on the right side of crazy) people who are passionate about their job, thank you all for being you and creating such a stimulating environment.

Shiva, Torunn and Anna L, you are great friends and have all helped me in different ways.

Karin and Nadine, Linda, Kees-Jan, Linda, Patric and Liza -thank you for your friendship and advice.

My Pink Ladies, wherever you are -together for ever (as the song goes).

My parents Ann and Kalle, for always being there for me. There are not enough words, you mean the world to me.

My brother Mattias and his family. Thank you for nice talks about life, gadgets and science. I am happy you are here.

Fabian, my eldest son and the sunshine of my life. You have always been a kind, smart and annoyingly funny guy, and I think you are the one who has helped me the most with this thesis, with your sound advice and positive attitude.

81

Albin and Isak, my two younger sons. Thank you for letting me write this thesis this year. Albin, you are kind and wise. Isak, you are adorable and a rascal. Thank you for reminding me every day what is most important in life.

Mats, my husband and companion. Thank you for your endless support and love.

82

83

References

1. Bhishagratna KL. An English translation of the Sushruta samhita, based on original Sanskrit text. Edited and published by Kaviraj Kunja Lal Bhishagratna. With a full and comprehensive introd., translation of different readings, notes, comperative views, index, glossary and plates. Calcutta; 1916.

2. WHO | The burden of health care-associated infection worldwide [Internet].: World Health Organization [cited Nov 17, 2020]. Available from: http://www.who.int/infection-prevention/publications/burden_hcai/en/.

3. Saleem Z, Godman B, Hassali MA, Hashmi FK, Azhar F, Rehman IU. Point prevalence surveys of health-care-associated infections: a systematic review. Pathog Glob Health. 2019 06;113(4):191-205.

4. Allegranzi B, Nejad SB, Combescure C, Graafmans W, Attar H, Donaldson L, et al. Burden of endemic health-care-associated infection in developing countries: systematic review and meta-analysis. The Lancet. 2011 /01/15;377(9761):228-41.

5. Suetens C, Latour K, Karki T, Ricchizzi E, Kinross P, Moro ML, et al. Prevalence of healthcare-associated infections, estimated incidence and composite antimicrobial resistance index in acute care hospitals and long-term care facilities: results from two European point prevalence surveys, 2016 to 2017. Euro Surveill. 2018 November 01;23(46):10.2807/1560,7917.ES.2018.23.46.1800516.

6. Economic evaluations of interventions to prevent healthcare-associated infections – literature review [Internet].; 2017-04-19T14:00:00+0200 [updated 2017-04- 19T14:00:00+0200; cited Nov 17, 2020]. Available from: https://www.ecdc.europa.eu/en/publications-data/economic-evaluati ons- interventions-prevent-healthcare-associated-infections.

7. HAI Data | CDC [Internet].; 2019 [updated -04-16T07:35:01Z/; cited Nov 17, 2020]. Available from: https://www.cdc.gov/hai/data/index.ht ml .

8. Magill SS, O'Leary E, Janelle SJ, Thompson DL, Dumyati G, Nadle J, et al. Changes in Prevalence of Health Care-Associated Infections in U.S. Hospitals. N Engl J Med. 2018 November 01;379(18):1732-44.

9. IHI Global Trigger Tool for Measuring Adverse Events | IHI - Institute for Healthcare Improvement [Internet]. [cited Nov 17, 2020]. Available from: http://www.ihi.org:80/resources/Pages/Tools/IHIGlobalTriggerToolforMeasu ringAEs.aspx.

10. Vårdrelaterade infektioner [Internet]. [cited Nov 17, 2020]. Available from:https://webbutik.skr.se/sv/artiklar/vardrelaterade-infektioner-2.html.

84

11. Vårdrelaterade infektioner [Internet]. [cited Nov 17, 2020]. Available from:https://webbutik.skr.se/sv/artiklar/vardrelaterade-infektioner-3.html.

12. Vincent J, Sakr Y, Singer M, Martin-Loeches I, Machado FR, Marshall JC, et al. Prevalence and Outcomes of Infection Among Patients in Intensive Care Units in 2017. JAMA. 2020 -4-21;323(15):1478-87.

13. Vincent JL, Bihari DJ, Suter PM, Bruining HA, White J, Nicolas-Chanoin MH, et al. The prevalence of nosocomial infection in intensive care units in Europe. Results of the European Prevalence of Infection in Intensive Care (EPIC) Study. EPIC International Advisory Committee. JAMA. 1995 Aug 23-30;274(8):639-44.

14. Barbier F, Andremont A, Wolff M, Bouadma L. Hospital-acquired pneumonia and ventilator-associated pneumonia: recent advances in epidemiology and management. Curr Opin Pulm Med. 2013 May;19(3):216-28.

15. Edwardson S, Cairns C. Nosocomial infections in the ICU. Anaesthesia and intensive care medicine. 2019 Jan;20(1):14-8.

16. Melsen WG, Rovers MM, Groenwold RHH, Bergmans, Dennis C. J. J., Camus C, Bauer TT, et al. Attributable mortality of ventilator-associated pneumonia: a meta- analysis of individual patient data from randomised prevention studies. Lancet Infect Dis. 2013 Aug;13(8):665-71.

17. Torres A, Niederman MS, Chastre J, Ewig S, Fernandez-Vandellos P, Hanberger H, et al. International ERS/ESICM/ESCMID/ALAT guidelines for the management of hospital-acquired pneumonia and ventilator-associated pneumonia: Guidelines for the management of hospital-acquired pneumonia (HAP)/ventilator-associated pneumonia (VAP) of the European Respiratory Society (ERS), European Society of Intensive Care MediciESICM), European Society of Clinical Microbiology and Infectious Diseases (ESCMID) and Asociacion Latinoamericana del Torax (ALAT). Eur Respir J. 2017 September 10;50(3):10.1183/13993003.00582,2017. Print 2017 Sep.

18. Restrepo MI, Peterson J, Fernandez JF, Qin Z, Fisher AC, Nicholson SC. Comparison of the bacterial etiology of early-onset and late-onset ventilator-associated pneumonia in subjects enrolled in 2 large clinical studies. Respir Care. 2013 July 01;58(7):1220-5.

19. Lim K, Kuo S, Ko W, Sheng W, Chang Y, Hong M, et al. ScienceDirect. Journal of environmental sciences (China). 2017(10):125-6.

20. A'Court C, Garrard CS. Nosocomial pneumonia in the intensive care unit: mechanisms and significance. Thorax. 1992 June 01;47(6):465-73.

21. Antimicrobial resistance and consumption remains high in the EU/EEA and the UK, according to new ECDC data [Internet].; 2020-11-18T13:00:00+0100 [updated 2020- 11-18T13:00:00+0100; cited Dec 16, 2020]. Available from: https://www.ecdc.europa.eu/en/news-events/antimicrobial-resistan ce-and- consumption-remains-high-press-release.

22. Magiorakos A-, Srinivasan A, Carey RB, Carmeli Y, Falagas ME, Giske CG, et al. Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance. Clinical microbiology and infection. 2012 Mar;18(3):268-81.

85

23. A Second Breadth: Hospital-Acquired Pneumonia in Pennsylvania, Nonventilated versus Ventilated Patients | Advisory [Internet]. [cited Nov 24, 2020]. Available from: http://patientsafety.pa.gov:80/ADVISORIES/Pages/201809_NVHAP.aspx.

24. Ewan V, Hellyer T, Newton J, Simpson J. New horizons in hospital acquired pneumonia in older people. Age Ageing. 2017 /05/01;46(3):352-8.

25. Baker D, Quinn B. Hospital Acquired Pneumonia Prevention Initiative-2: Incidence of nonventilator hospital-acquired pneumonia in the United States. Am J Infect Control. 2018 01;46(1):2-7.

26. Klompas M. Hospital-Acquired Pneumonia in Nonventilated Patients: The Next Frontier. Infect Control Hosp Epidemiol. 2016 07;37(7):825-6.

27. Att förebygga vårdrelaterade infektioner - Ett kunskapsunderlag — Folkhälsomyndigheten [Internet]. [cited Nov 25, 2020]. Available from: http://www.folkhalsomyndigheten.se/publicerat- material/publikationsarkiv/a/att-forebygga-vardrelaterade-infektioner-ett- kunskapsunderlag/.

28. Mitchell BG, Russo PL, Cheng AC, Stewardson AJ, Rosebrock H, Curtis SJ, et al. Strategies to reduce non-ventilator-associated hospital-acquired pneumonia: A systematic review. Infection, Disease & Health. 2019 Nov;24(4):229-39.

29. Quinn B, Baker DL, Cohen S, Stewart JL, Lima CA, Parise C. Basic nursing care to prevent nonventilator hospital-acquired pneumonia. J Nurs Scholarsh. 2014 January 01;46(1):11-9.

30. Sopena N, Heras E, Casas I, Bechini J, Guasch I, Pedro-Botet ML, et al. Risk factors for hospital-acquired pneumonia outside the intensive care unit: a case-control study. Am J Infect Control. 2014 Jan;42(1):38-42.

31. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R, Gordon JI. The Human Microbiome Project. Nature (London). 2007;449(7164):804-10.

32. Santacroce L, Charitos IA, Ballini A, Inchingolo F, Luperto P, De Nitto E, et al. The Human Respiratory System and its Microbiome at a Glimpse. Biology (Basel). 2020 October 01;9(10):10.3390/biology9100318.

33. Bassis CM, Erb-Downward JR, Dickson RP, Freeman CM, Schmidt TM, Young VB, et al. Analysis of the upper respiratory tract microbiotas as the source of the lung and gastric microbiotas in healthy individuals. mBio. 2015 Mar 03,;6(2):e00037.

34. Gao Z, Kang Y, Yu J, Ren L. Human pharyngeal microbiome may play a protective role in respiratory tract infections. Genomics Proteomics Bioinformatics. 2014 Jun;12(3):144-50.

35. Huffnagle GB, Dickson RP, Lukacs NW. The respiratory tract microbiome and lung inflammation: a two-way street. Mucosal immunology. 2017 Mar;10(2):299-306.

36. Dickson RP, Schultz MJ, van der Poll T, Schouten LR, Falkowski NR, Luth JE, et al. Lung Microbiota Predict Clinical Outcomes in Critically Ill Patients. Am J Respir Crit Care Med. 2020 March 01;201(5):555-63.

86

37. Johanson WG, Pierce AK, Sanford JP. Changing pharyngeal bacterial flora of hospitalized patients. Emergence of gram-negative bacilli. N Engl J Med. 1969 November 20;281(21):1137-40.

38. Wu D, Wu C, Zhang S, Zhong Y. Risk Factors of Ventilator-Associated Pneumonia in Critically III Patients. Frontiers in pharmacology. 2019;10:482.

39. Sopena N, Heras E, Casas I, Bechini J, Guasch I, Pedro-Botet ML, et al. Risk factors for hospital-acquired pneumonia outside the intensive care unit: a case-control study. Am J Infect Control. 2014 January 01;42(1):38-42.

40. Tranberg A, Samuelsson C, Klarin B. Disturbance in the oropharyngeal microbiota in relation to antibiotic and proton pump inhibitor medication and length of hospital stay. APMIS. 2020 Sep 27,.

41. Moron R, Galvez J, Colmenero M, Anderson P, Cabeza J, Rodriguez-Cabezas ME. The Importance of the Microbiome in Critically Ill Patients: Role of Nutrition. Nutrients. 2019 Dec 07,;11(12).

42. Lemon K. Comparative Analyses of the Bacterial Microbiota of the Human Nostril and Oropharynx. mBio. 2010;1(3).

43. Ewan VC, Reid WDK, Shirley M, Simpson AJ, Rushton SP, Wade WG. Oropharyngeal Microbiota in Frail Older Patients Unaffected by Time in Hospital. Frontiers in cellular and infection microbiology. 2018;8:42.

44. Whelan FJ, Verschoor CP, Stearns JC, Rossi L, Luinstra K, Loeb M, et al. The loss of topography in the microbial communities of the upper respiratory tract in the elderly. Ann Am Thorac Soc. 2014 May;11(4):513-21.

45. Retchless AC, Kretz CB, Rodriguez-Rivera LD, Chen A, Soeters HM, Whaley MJ, et al. Oropharyngeal microbiome of a college population following a meningococcal disease outbreak. Sci Rep. 2020 January 20;10(1):632-8.

46. Deo PN, Deshmukh R. Oral microbiome: Unveiling the fundamentals. J Oral Maxillofac Pathol. 2019;23(1):122-8.

47. Zhang Y, Wang X, Li H, Ni C, Du Z, Yan F. Human oral microbiota and its modulation for oral health. Biomed Pharmacother. 2018 March 01;99:883-93.

48. Mascitti M, Togni L, Troiano G, Caponio VCA, Gissi DB, Montebugnoli L, et al. Beyond Head and Neck Cancer: The Relationship Between Oral Microbiota and Tumour Development in Distant Organs. Frontiers in cellular and infection microbiology. 2019 Jun 26,;9:232.

49. Gilbert JA, Blaser MJ, Caporaso JG, Jansson JK, Lynch SV, Knight R. Current understanding of the human microbiome. Nat Med. 2018 04 10,;24(4):392-400.

50. Sender R, Fuchs S, Milo R. Are We Really Vastly Outnumbered? Revisiting the Ratio of Bacterial to Host Cells in Humans. Cell. 2016 Jan 28,;164(3):337-40.

51. Young VB. The role of the microbiome in human health and disease: an introduction for clinicians. BMJ. 2017 Mar 15,;356:j831.

52. Hill C, Guarner F, Reid G, Gibson GR, Merenstein DJ, Pot B, et al. Expert consensus document. The International Scientific Association for Probiotics and Prebiotics consensus statement on the scope and appropriate use of the term probiotic. Nat Rev Gastroenterol Hepatol. 2014 August 01;11(8):506-14.

87

53. Hutkins RW, Krumbeck JA, Bindels LB, Cani PD, Fahey G, Goh YJ, et al. Prebiotics: why definitions matter. Curr Opin Biotechnol. 2016 February 01;37:1-7.

54. Pandey KR, Naik SR, Vakil BV. Probiotics, prebiotics and synbiotics- a review. J Food Sci Technol. 2015 Jul 22,;52(12):7577-87.

55. Accepted Claims about the Nature of Probiotic Microorganisms in Food [Internet].; 2009 [updated -04-16; cited Dec 5, 2020]. Available from: https://www.canada.ca/en/health-can ada/ services/foo d -nutrition/food- labelling/health-claims/accepted-claims-about-nature-probiotic- microorganisms-food.html.

56. de Vrese M, Schrezenmeir J. Probiotics, prebiotics, and synbiotics. Adv Biochem Eng Biotechnol. 2008;111:1-66.

57. Cremon C, Barbaro MR, Ventura M, Barbara G. Pre- and probiotic overview. Current Opinion in Pharmacology. 2018;43:87-92.

58. Stjernquist-Desatnik A. Persistence of Lactobacillus plantarum DSM 9843 on Human Tonsillar Surface after Oral Administration in Fermented Oatmeal Gruel: A Pilot Study. Acta Oto-Laryngologica. 2000 January 1,;120(543):215-9.

59. Klarin B, Adolfsson A, Torstensson A, Larsson A. Can probiotics be an alternative to chlorhexidine for oral care in the mechanically ventilated patient? A multicentre, prospective, randomised controlled open trial. Crit Care. 2018 10 28,;22(1):272.

60. Tran K, Butcher R. Chlorhexidine for Oral Care: A Review of Clinical Effectiveness and Guidelines. Ottawa (ON): Canadian Agency for Drugs and Technologies in Health; 2019.

61. Laleman I, Pauwels M, Quirynen M, Teughels W. A dual-strain Lactobacilli reuteri probiotic improves the treatment of residual pockets: A randomized controlled clinical trial. J Clin Periodontol. 2020 January 01;47(1):43-53.

62. Kang MS, Lee DS, Lee SA, Kim MS, Nam SH. Effects of probiotic bacterium Weissella cibaria CMU on periodontal health and microbiota: a randomised, double- blind, placebo-controlled trial. BMC Oral Health. 2020 September 02;20(1):243-2.

63. Jeppsson B, Mangell P, Thorlacius H. Use of probiotics as prophylaxis for postoperative infections. Nutrients. 2011 05;3(5):604-12.

64. Orrhage K, Sillerström E, Gustafsson JA, Nord CE, Rafter J. Binding of mutagenic heterocyclic amines by intestinal and lactic acid bacteria. Mutat Res. 1994 Dec 01,;311(2):239-48.

65. Search | ESPGHAN [Internet]. [cited Dec 7, 2020]. Available from:http://www.espghan.org/search-site.

66. Probiotics for the prevention and treatment of antibiotic-associated diarrhea: a systematic review and meta-analysis [Internet].: JAMA; 2012 [updated 05/09/; cited Dec 7, 2020]. Available from: https://pubmed-ncbi-nlm-nih- gov.ludwig.lub.lu.se/22570464/?dopt=Abstract.

67. Klarin B, Wullt M, Palmquist I, Molin G, Larsson A, Jeppsson B. Lactobacillus plantarum 299v reduces colonisation of Clostridium difficile in critically ill patients treated with antibiotics. Acta Anaesthesiol Scand. 2008 Sep;52(8):1096-102.

88

68. Goldenberg JZ, Yap C, Lytvyn L, Lo CK, Beardsley J, Mertz D, et al. Probiotics for the prevention of Clostridium difficile‐associated diarrhea in adults and children. Cochrane Database of Systematic Reviews. 2017(12).

69. Quigley EM, Fried M, GKA, Khalif I, Hungin AP, Lindberg G, et al. World Gastroenterology Organisation Global Guidelines Irritable Bowel Syndrome: A Global Perspective Update September 2015. J Clin Gastroenterol. 2016 October 01;50(9):704-13.

70. Iheozor-Ejiofor Z, Kaur L, Gordon M, Baines PA, Sinopoulou V, Akobeng AK. Probiotics for maintenance of remission in ulcerative colitis. Cochrane Database of Systematic Reviews. 2020(3).

71. Kaur L, Gordon M, Baines PA, Iheozor-Ejiofor Z, Sinopoulou V, Akobeng AK. Probiotics for induction of remission in ulcerative colitis. Cochrane Database of Systematic Reviews. 2020(3).

72. Limketkai BN, Akobeng AK, Gordon M, Adepoju AA. Probiotics for induction of remission in Crohn's disease. Cochrane Database of Systematic Reviews. 2020(7).

73. Wu XD, Liu MM, Liang X, Hu N, Huang W. Effects of perioperative supplementation with pro-/synbiotics on clinical outcomes in surgical patients: A meta-analysis with trial sequential analysis of randomized controlled trials. Clin Nutr. 2018 April 01;37(2):505-15.

74. Chowdhury AH, Adiamah A, Kushairi A, Varadhan KK, Krznaric Z, Kulkarni AD, et al. Perioperative Probiotics or Synbiotics in Adults Undergoing Elective Abdominal Surgery: A Systematic Review and Meta-analysis of Randomized Controlled Trials. Ann Surg. 2020 06;271(6):1036-47.

75. Rayes N, Hansen S, Seehofer D, Müller AR, Serke S, Bengmark S, et al. Early enteral supply of fiber and Lactobacilli versus conventional nutrition: a controlled trial in patients with major abdominal surgery. Nutrition. 2002 Jul-Aug;18(7-8):609-15.

76. Melissa Latorre Suneeta Krishnareddy Daniel E Freedberg. Microbiome as mediator: Do systemic infections start in the gut? World journal of gastroenterology : WJG. 2015;21(37):10487-92.

77. Shimizu K, Ogura H, Tomono K, Tasaki O, Asahara T, Nomoto K, et al. Patterns of Gram-stained fecal flora as a quick diagnostic marker in patients with severe SIRS. Dig Dis Sci. 2011 June 01;56(6):1782-8.

78. Beale RJ, Sherry T, Lei K, Campbell-Stephen L, McCook J, Smith J, et al. Early enteral supplementation with key pharmaconutrients improves Sequential Organ Failure Assessment score in critically ill patients with sepsis: outcome of a randomized, controlled, double-blind trial. Crit Care Med. 2008 January 01;36(1):131-44.

79. O'Keefe SJ, Ou J, Delany JP, Curry S, Zoetendal E, Gaskins HR, et al. Effect of fiber supplementation on the microbiota in critically ill patients. World J Gastrointest Pathophysiol. 2011 December 15;2(6):138-45.

80. McNaught CE, Woodcock NP, Anderson AD, MacFie J. A prospective randomised trial of probiotics in critically ill patients. Clin Nutr. 2005 April 01;24(2):211-9.

89

81. Manzanares W, Lemieux M, Langlois PL, Wischmeyer PE. Probiotic and synbiotic therapy in critical illness: a systematic review and meta-analysis. Crit Care. 2016 Aug 19,;19:262.

82. Wullt M, Johansson Hagslatt ML, Odenholt I, Berggren A. Lactobacillus plantarum 299v enhances the concentrations of fecal short-chain fatty acids in patients with recurrent clostridium difficile-associated diarrhea. Dig Dis Sci. 2007 September 01;52(9):2082-6.

83. Morrow LE, Kollef MH, Casale TB. Probiotic prophylaxis of ventilator-associated pneumonia: a blinded, randomized, controlled trial. Am J Respir Crit Care Med. 2010 Oct 15,;182(8):1058-64.

84. Zeng J, Wang C, Zhang F, Qi F, Wang S, Ma S, et al. Effect of probiotics on the incidence of ventilator-associated pneumonia in critically ill patients: a randomized controlled multicenter trial. Intensive Care Med. 2016 Jun;42(6):1018-28.

85. Kotzampassi K, Giamarellos-Bourboulis EJ, Voudouris A, Kazamias P, EleftheriadisE. Benefits of a synbiotic formula (Synbiotic 2000Forte) in critically Ill trauma patients: early results of a randomized controlled trial. World J Surg. 2006 October 01;30(10):1848-55.

86. Panigrahi P, Parida S, Nanda NC, Satpathy R, Pradhan L, Chandel DS, et al. A randomized synbiotic trial to prevent sepsis among infants in rural India. Nature. 2017 August 24;548(7668):407-12.

87. Panigrahi P, Parida S, Nanda NC, Pradhan L, Chandel DS, Baccaglini L, et al. Corrigendum: A randomized synbiotic trial to prevent sepsis among infants in rural India. Nature. 2018 January 11;553(7687):238.

88. Andrews B, Muchemwa L, Kelly P, Lakhi S, Heimburger DC, Bernard GR. Simplified severe sepsis protocol: a randomized controlled trial of modified early goal-directed therapy in Zambia. Crit Care Med. 2014 November 01;42(11):2315-24.

89. Maitland K, Kiguli S, Opoka RO, Engoru C, Olupot-Olupot P, Akech SO, et al. Mortality after fluid bolus in African children with severe infection. N Engl J Med. 2011 June 30;364(26):2483-95.

90. Alexandre Y, Le Blay G, Boisramé-Gastrin S, Le Gall F, Héry-Arnaud G, Gouriou S, et al. Probiotics: A new way to fight bacterial pulmonary infections? Médecine et Maladies Infectieuses. 2014 January 1,;44(1):9-17.

91. El-Wakeel* N, El-Wakeel A, El-Assmy M, Barakat G, Soliman W. Role of probiotics in prevention of hospital acquired pneumonia in Egyptian children admitted to Pediatric Intensive Care Unit of Mansoura University Childrens Hospital. AJMR. 2016 /08/21;10(31):1240-7.

92. Hao Q, Dong BR, Wu T. Probiotics for preventing acute upper respiratory tract infections. Cochrane Database of Systematic Reviews. 2015(2).

93. Hatakka K, Savilahti E, Pönkä A, Meurman JH, Poussa T, Näse L, et al. Effect of long term consumption of probiotic milk on infections in children attending day care centres: double blind, randomised trial. BMJ. 2001 Jun 02,;322(7298):1327.

90

94. Hojsak I, Snovak N, Abdović S, Szajewska H, Mišak Z, Kolaček S. Lactobacillus GG in the prevention of gastrointestinal and respiratory tract infections in children who attend day care centers: A randomized, double-blind, placebo-controlled trial. Clinical Nutrition. 2010 June 1,;29(3):312-6.

95. Pre-, pro- and synbiotics in cancer prevention and treatment—a review of basic and clinical research [Internet].; 2018 [updated -09-05; cited Oct 14, 2020]. Available from: http://ecancer.org/en/journal/article/869-pre-pro-and-synbiotics-in- cancer-prevention-and-treatment-a-review-of-basic-and-clinical-research.

96. Zhang K, Dai H, Liang W, Zhang L, Deng Z. Fermented dairy foods intake and risk of cancer. Int J Cancer. 2019 05 01,;144(9):2099-108.

97. Cheng WY, Wu CY, Yu J. The role of gut microbiota in cancer treatment: friend or foe? Gut. 2020 October 01;69(10):1867-76.

98. Hibberd AA, Lyra A, Ouwehand AC, Rolny P, Lindegren H, Cedgård L, et al. Intestinal microbiota is altered in patients with colon cancer and modified by probiotic intervention. BMJ Open Gastroenterol. 2017;4(1):e000145.

99. Lalla RV, Bowen J, Barasch A, Elting L, Epstein J, Keefe DM, et al. MASCC/ISOO clinical practice guidelines for the management of mucositis secondary to cancer therapy. Cancer. 2014 May 15;120(10):1453-61.

100. Seddik HA, Bendali F, Gancel F, Fliss I, Spano G, Drider D. Lactobacillus plantarum and Its Probiotic and Food Potentialities. Probiotics Antimicrob Proteins. 2017 June 01;9(2):111-22.

101. Lactobacillus plantarum. 2020 -12-06T01:28:48Z.102. Olasupo NA. Bacteriocins of Lactobacillus plantarum strains from fermented

foods. Folia Microbiol (Praha). 1996;41(2):130-6.103. da Silva Sabo S, Pérez-Rodríguez N, Domínguez JM, de Souza Oliveira, Ricardo

Pinheiro. Inhibitory substances production by Lactobacillus plantarum ST16Pa cultured in hydrolyzed cheese whey supplemented with soybean flour and their antimicrobial efficiency as biopreservatives on fresh chicken meat. Food Res Int. 2017 09;99(Pt 1):762-9.

104. Adebayo FA, Afolabi OR, Akintokun AK. Antimicrobial properties of purified bacteriocins produced from Lactobacillus casei and Lactobacillus fermentum against selected pathogenic microorganisms. ; 2014.

105. Prabhurajeshwar C, Chandrakanth RK. Probiotic potential of Lactobacilli with antagonistic activity against pathogenic strains: An in vitro validation for the production of inhibitory substances. Biomed J. 2017 Oct;40(5):270-83.

106. Tagg JR, Dajani AS, Wannamaker LW. Bacteriocins of gram-positive bacteria. Bacteriol Rev. 1976 Sep;40(3):722-56.

107. Lin T, Pan T. Characterization of an antimicrobial substance produced by Lactobacillus plantarum NTU 102. Journal of Microbiology, Immunology and Infection. 2019 Jun;52(3):409-17.

91

108. Zimina M, Babich O, Prosekov A, Sukhikh S, Ivanova S, Shevchenko M, et al. Overview of Global Trends in Classification, Methods of Preparation and Application of Bacteriocins. Antibiotics (Basel). 2020 August 28;9(9):10.3390/antibiotics9090553.

109. Simons A, Alhanout K, Duval RE. Bacteriocins, Antimicrobial Peptides from Bacterial Origin: Overview of Their Biology and Their Impact against Multidrug- Resistant Bacteria. Microorganisms. 2020 Apr 27,;8(5).

110. Johansson ML, Molin G, Jeppsson B, Nobaek S, Ahrne S, Bengmark S. Administration of different Lactobacillus strains in fermented oatmeal soup: in vivo colonization of human intestinal mucosa and effect on the indigenous flora. Appl Environ Microbiol. 1993 January 01;59(1):15-20.

111. Naruszewicz M, Johansson M, Zapolska-Downar D, Bukowska H. Effect of Lactobacillus plantarum 299v on cardiovascular disease risk factors in smokers. Am J Clin Nutr. 2002 Dec;76(6):1249-55.

112. Bukowska H, Pieczul-Mróz J, Jastrzebska M, Chełstowski K, Naruszewicz M. Decrease in fibrinogen and LDL-cholesterol levels upon supplementation of diet with Lactobacillus plantarum in subjects with moderately elevated cholesterol. Atherosclerosis. 1998 Apr;137(2):437-8.

113. Niedzielin K, Kordecki H, Birkenfeld B. A controlled, double-blind, randomized study on the efficacy of Lactobacillus plantarum 299V in patients with irritable bowel syndrome. Eur J Gastroenterol Hepatol. 2001 Oct;13(10):1143-7.

114. Nobaek S, Johansson ML, Molin G, Ahrné S, Jeppsson B. Alteration of intestinal microflora is associated with reduction in abdominal bloating and pain in patients with irritable bowel syndrome. Am J Gastroenterol. 2000 May;95(5):1231-8.

115. Ducrotté P, Sawant P, Jayanthi V. Clinical trial: Lactobacillus plantarum 299v (DSM 9843) improves symptoms of irritable bowel syndrome. World journal of gastroenterology. 2012 Aug 14,;18(30):4012.

116. Dudzicz S, Adamczak M, Więcek A. Clostridium Difficile Infection in the Nephrology Ward. Kidney and Blood Pressure Research. 2018 Jan;42(5):844-52.

117. Wullt M, Hagslatt ML, Odenholt I. Lactobacillus plantarum 299v for the treatment of recurrent Clostridium difficile-associated diarrhoea: a double-blind, placebo- controlled trial. Scand J Infect Dis. 2003;35(6-7):365-7.

118. Zheng Y, Lu Y, Wang J, Yang L, Pan C, Huang Y. Probiotic properties of Lactobacillus strains isolated from Tibetan kefir grains. PLoS One. 2013 July 22;8(7):e69868.

119. Kujawa-Szewieczek A, Adamczak M, Kwiecień K, Dudzicz S, Gazda M, Więcek A. The Effect of Lactobacillus plantarum 299v on the Incidence of Clostridium difficile Infection in High Risk Patients Treated with Antibiotics. Nutrients. 2015 -12- 04;7(12):10179-88.

120. Malik M, Suboc TM, Tyagi S, Salzman N, Wang J, Ying R, et al. Lactobacillus plantarum 299v Supplementation Improves Vascular Endothelial Function and Reduces Inflammatory Biomarkers in Men With Stable Coronary Artery Disease. Circ Res. 2018 October 12;123(9):1091-102.

92

121. Rudzki L, Ostrowska L, Pawlak D, Malus A, Pawlak K, Waszkiewicz N, et al. Probiotic Lactobacillus Plantarum 299v decreases kynurenine concentration and improves cognitive functions in patients with major depression: A double-blind, randomized, placebo controlled study. Psychoneuroendocrinology. 2019 February 01;100:213-22.

122. Rayes N, Seehofer D, Hansen S, Boucsein K, Müller AR, Serke S, et al. Early enteral supply of lactobacillus and fiber versus selective bowel decontamination: a controlled trial in liver transplant recipients. Transplantation. 2002 Jul 15,;74(1):123- 7.

123. Vijayananthan A, Nawawi O. The importance of Good Clinical Practice guidelines and its role in clinical trials. Biomed Imaging Interv J. 2008 January 01;4(1):e5.

124. Retchless AC, Kretz CB, Rodriguez-Rivera LD, Chen A, Soeters HM, Whaley MJ, et al. Oropharyngeal microbiome of a college population following a meningococcal disease outbreak. Sci Rep. 2020 01 20,;10(1):632.

125. Klarin B, Molin G, Jeppsson B, Larsson A. Use of the probiotic Lactobacillus plantarum 299 to reduce pathogenic bacteria in the oropharynx of intubated patients: a randomised controlled open pilot study. Crit Care. 2008;12(6):R136.

126. Eom CS, Jeon CY, Lim JW, Cho EG, Park SM, Lee KS. Use of acid-suppressive drugs and risk of pneumonia: a systematic review and meta-analysis. CMAJ. 2011 February 22;183(3):310-9.

127. Alshamsi F., Belley-Cote E., Cook D., Almenawer S.A., Alqahtani Z., Perri D., Thabane L., Al-Omari A., Lewis K., Guyatt G., Alhazzani W. Efficacy and safety of proton pump inhibitors for stress ulcer prophylaxis in critically ill patients: A systematic review and meta-analysis of randomized trials. Critical Care. 2016;20:120.

128. Herzig SJ, Howell MD, Ngo LH, Marcantonio ER. Acid-suppressive medication use and the risk for hospital-acquired pneumonia. JAMA. 2009 May 27;301(20):2120-8.

129. Stress Ulcer Prophylaxis • LITFL • CCC Gastroenterology. 2019 -01- 30T10:51:42+00:00.

130. Dickson RP, Schultz MJ, van der Poll T, Schouten LR, Falkowski NR, Luth JE, et al. Lung Microbiota Predict Clinical Outcomes in Critically Ill Patients. Am J Respir Crit Care Med. 2020 Jan 24,.

131. Mölstad S, Löfmark S, Carlin K, Erntell M, Aspevall O, Blad L, et al. Lessons learnt during 20 years of the Swedish strategic programme against antibiotic resistance. Bull World Health Organ. 2017 -11-01;95(11):764-73.

132. Sullivan Å, Edlund C, Nord CE. Effect of antimicrobial agents on the ecological balance of human microflora. The Lancet Infectious Diseases. 2001 /09/01;1(2):101- 14.

133. Nord CE, Heimdahl A, Kager L. Antimicrobial induced alterations of the human oropharyngeal and intestinal microflora. Scand J Infect Dis Suppl. 1986;49:64-72.

134. Rosen R, Hu L, Amirault J, Khatwa U, Ward DV, Onderdonk A. 16S Community Profiling Identifies Proton Pump Inhibitor Related Differences in Gastric, Lung, and Oropharyngeal Microflora. The Journal of Pediatrics. 2015 /04/01;166(4):917-23.

93

135. Frandah W, Colmer-Hamood J, Mojazi Amiri H, Raj R, Nugent K. Oropharyngeal flora in patients admitted to the medical intensive care unit: clinical factors and acid suppressive therapy. J Med Microbiol. 2013 May 01;62(Pt 5):778-84.

136. Annuk H, Shchepetova J, Kullisaar T, Songisepp E, Zilmer M, Mikelsaar M. Characterization of intestinal lactobacilli as putative probiotic candidates. J Appl Microbiol. 2003;94(3):403-12.

137. Hütt P, Shchepetova J, Lõivukene K, Kullisaar T, Mikelsaar M. Antagonistic activity of probiotic lactobacilli and bifidobacteria against entero- and uropathogens. J Appl Microbiol. 2006 -06;100(6):1324-32.

138. Tranberg A, Thorarinsdottir HR, Holmberg A, Schott U, Klarin B. Proton pump inhibitor medication is associated with colonisation of gut flora in the oropharynx. Acta Anaesthesiol Scand. 2018 July 01;62(6):791-800.

139. Song D, Zhu M, Gu Q. Purification and characterization of Plantaricin ZJ5, a new bacteriocin produced by Lactobacillus plantarum ZJ5. PLoS ONE. 2014;9(8):e105549.

140. Arumugam S, Lau CS, Chamberlain RS. Probiotics and Synbiotics Decrease Postoperative Sepsis in Elective Gastrointestinal Surgical Patients: a Meta-Analysis. J Gastrointest Surg. 2016 June 01;20(6):1123-31.

141. Rayes N, Seehofer D, Theruvath T, Mogl M, Langrehr JM, Nüssler NC, et al. Effect of Enteral Nutrition and Synbiotics on Bacterial Infection Rates After Pylorus- preserving Pancreatoduodenectomy: A Randomized, Double-blind Trial. Annals of Surgery. 2007 July;246(1):36–41.

Paper I

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

791

Proton pump inhibitor medication is associated with colonisation of gut flora in the oropharynxA. Tranberg1,† , H. R. Thorarinsdottir1,†, A. Holmberg2, U. Scho€tt1 and B. Klarin1

1Division of Intensive and Perioperative Care, SkOane University Hospital, Lund, Sweden2Division of Infection Medicine, Department of Clinical Sciences, Lund University, Lund, Sweden

CorrespondenceA. Tranberg, Division of Intensive and Perioperative Care, SkOane University Hospital, Getingev€agen 4, SE-22185 Lund, SwedenE-mail: anna_tranberg@yahoo.com

†These authors contributed equally.

Conflict of interestThe authors have no conflicts of interest.

FundingThis work was funded by Swedish Region SkOane0s research grants.

Submitted 18 January 2018; accepted 31January 2018; submission 29 September 2017.

CitationTranberg A, Thorarinsdottir HR, Holmberg A, Scho€tt U, Klarin B. Proton pump inhibitor medication is associated with colonisation of gut flora in the oropharynx. Acta Anaesthesiologica Scandinavica 2018

doi: 10.1111/aas.13094

Background: The normal body exists in mutualistic balance with a large range of microbiota. The primary goal of this study was toestablish whether there is an imbalance in the oropharyngeal flora early after hospital or ICU admittance, and whether flora differs between control, ward and critically ill patients. The secondary goal was to explore whether there are patient characteristics that can be associated with a disturbed oropharyngeal flora.Methods: Oropharyngeal cultures were obtained from three dif-ferent study groups: (1) controls from the community, (2) wardpatients and (3) critically ill patients, the two latter within 24 h after admittance.Results: Cultures were obtained from 487 individuals: 77 con- trols, 193 ward patients and 217 critically ill patients. Abnormalpharyngeal flora was more frequent in critically ill and ward patients compared with controls (62.2% and 10.4% vs. 1.3%, P < 0.001 and P = 0.010, respectively). Colonisation of gut flora in the oropharynx was more frequent in critically ill patients com- pared with ward patients or controls (26.3% vs. 4.7% and 1.3%, P < 0.001 and P < 0.001, respectively). Proton pump inhibitor medication was the strongest independent factor associated with the presence of gut flora in the oropharynx in both ward and criti- cally ill patients (P = 0.030 and P = 0.044, respectively).Conclusion: This study indicates that abnormal oropharyngealflora is an early and frequent event in hospitalised patients andmore so in the critically ill, compared to controls. Proton pump inhibitor medication is associated with colonisation of gut flora in the oropharynx.

Normally, different groups of bacteria inhabit clearly defined parts of the human body in a mutually beneficial balance. Morbidity and medication can alter this balance and disturb the normal flora. This is routinely verified via

the appearance of pathologic bacteria detected with traditional culture techniques. With mod- ern molecular techniques, low concentrations of organisms can be detected earlier, so asymp- tomatic colonisation of, for example, the lower

ORIGINAL ARTICLE

Editorial commentThis study shows that abnormal oropharyngeal flora is a common and early feature among criti- cally ill patients, particularly among those that are treated with proton pump inhibitors before admission to hospital. This could be a factor in subsequent occurrence of nosocomial infections.

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

792

respiratory tract by bacteria from the oropharynx could represent an early subclinical phase before overt signs of pneumonia.1

Bacterial pneumonia is a classic complication to hospitalisation and the most frequent of all nosocomial infections.2 The pathogenesis behind nosocomial pneumonia is described as a process that starts at hospitalisation. Hospital- associated events such as medication with opi- oids, antibiotics and proton pump inhibitors (PPI), surgery, fasting, insertion of tubes and supine position have all been implicated in the pathogenesis, wherein bacteria from the diges- tive system spread into the oropharynx, fol- lowed by micro-aspiration, leading to colonisation of the lungs.3–5

Most of the studies examining risk factors and nosocomial pneumonia have been performed in an intensive care unit (ICU) milieu. Further- more, these studies describe a significant link between the use of proton pump inhibitors (PPIs) and pneumonia.6,7 The pathophysiology behind how PPI increases the risk of hospital- acquired pneumonia (HAP) is the subject of on- going debate. Several investigators have shown that PPIs elevate pH in the ventricle, thereby disrupting the protective acid barrier between gut and oropharynx, allowing gut flora to popu- late the oropharyngeal tract.8,9 The risk of HAP may also be increased via other mechanisms, for example, elevating the pH of the oropharynx and directly disturbing the local bacterial flora, facilitating overgrowth of pathogenic bacteria.10

In a study from 1969, the investigatorsshowed a change in the oropharyngeal flora of hospitalised patients.11 The patients were divided into three groups, and the more severe the physical illness, the more common it was to find Gram-negative bacilli in the oropharyngeal tract. In that study, a change in oropharyngeal flora was already present in the first culture taken within 36 h of admission. In a more recent study, abnormal oropharyngeal flora was present on day 1 in patients admitted to the intensive care unit (ICU).12 In this study, PPI was a significant risk factor for disturbed oropharyngeal flora. These two studies indicate that changes in oropharyngeal flora may be pre- sent before patient admission to hospital.

Prevention of pneumonia could possibly beenhanced by early identification of risk factors

or patient groups at risk. HAP and ventilator- associated pneumonia (VAP) are costly both for patients and for the health care system, increas- ing mortality, morbidity and length of hospital stays.

The primary goal of this study was to estab- lish whether there is an imbalance in the oropharyngeal flora early after hospital or ICU admittance, and whether there is a difference in flora between ICU, ward and control subjects. The secondary goal was to explore whether there are patient characteristics that can be asso- ciated with a disturbed oropharyngeal flora.

Materials and methodsThis clinical observational study was carried out at Skane University Hospital, Lund, Sweden and at two tertiary hospitals in the Southern Regions of Sweden (Centralsjukhuset, Kris- tianstad and Hallands sjukhus, Halmstad). The study protocol was reviewed and approved by the Regional Ethical Review Board, Lund, Swe- den (protocol 2016/933). Informed consent, including the permission to collect and publish anonymous data, was obtained from the patients or their relatives before they entered the study.

PatientsWe included patients, 18 years or older into three study groups: (1) critically ill patients admitted to the hospital’s ICU (mixed surgical and medical ICU), requiring mechanical ventila- tion for at least 24 h, (2) patients admitted to acute medical or surgical wards, not requiring intensive care and (3) controls in the community who were neither hospitalised nor medicated with antibiotics during the previous 2 months.

The ICU patients consisted of 217 patients;167 patients were prospectively included from May 2010 to February 2016 and 50 patients were retrospectively included from earlier pub- lished material (collected from January 2004 to March 2007).13 The same inclusion criteria were applied to all ICU patients. The ward patients consisted of 193 patients who were prospec- tively included from January 2014 to December 2016. The 77 control patients were prospectively included from March 2014 to April 2016, and consisted of 30 members of Skane’s Fire

A. TRANBERG ET AL.

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

793

Department (R€addningstj€ansten Syd, Lund, Sweden) and 47 members of the Senior Citizens Organization (Pension€arernas Riksorganisation, Lund, Sweden).

Patient dataThe following patient data were recorded in a standardised form: age, sex, BMI, cause of hospi- tal or ICU stay, smoking status, alcohol consump- tion, occurrence of diabetes, antibiotic and/or PPI use ≥ 24 h prior to culture sampling. Acute Phys- iology and Chronic Health Evaluation (APACHE) II scores, as a measure of severity of illness, and number of days in the hospital before ICU admis- sion were collected for all ICU patients.

Oropharyngeal swabs were taken from the oropharynx behind the posterior tonsillar pillar from all patients. The samples were collected within 24 h of admission to the hospital (ward patients) or intensive care unit (ICU patients). Oropharyngeal swabs were collected on the respective inclusion sites for controls. Apart from the collection of oropharyngeal swabs, hos- pitalised patients received standard hospital care based on the diagnosis and clinical decisions of the responsible physician.

Microbiological proceduresAll oropharyngeal swabs were processed in the same way, using extended microbiological pro- cedures, as is standard for oropharyngeal swabs from our ICU patients. Samples were inoculated on three different selective, one differentiating and one non-selective agar plate as follows: (1) agar plate with 5% horse blood (LabM, Hey- wood, Lancashire, UK) supplemented with 10 mg/l colistin and 15 mg/l nalidixic acid, (2) agar plate with 5% horse blood supplemented with 2 mg/l gentamicin and 25 mg/l nalidixic acid for Gram-positive cocci including S. pneu-moniae, (3) Hematin agar plate (OxoidTM, ThermoScience, Basingstoke, UK) supplemented with 300 mg/l bacitracin for fastidious Gram-negative rods including Haemophilus influenzae (selective),(4) Uriselect 4 agar (Bio-Rad Laboratories, Copenhagen, Denmark) supplemented with 10 mg/l vancomycin for non-fastidious Gram- negative rods (differentiating) and (5) Hematin

agar with a colistin disk (non-selective).

PPI AND CHANGES IN ORAL FLORA

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

794

All plates were manufactured in house. Plates were inspected for growth after 16 and 40 h of aerobic, anaerobic or CO2 incubation at 35– 37°C. Bacterial species identification was per- formed using matrix-assisted laser desorption/ ionisation time-of-flight (MALDI-TOF) mass spectrometry (MALDI Biotyper Microbial Iden- tification system, Bruker, Boston, MA, USA). Differentiation of Candida species was based on colony appearance on CHROM Candida agar (CHROMagar, H€agersten, Sweden) after 48 h of incubation at 35°C.

DefinitionsWe classified specimens with growth of at least two species of bacteria usually found in the oral cavity as having normal oral flora. Specimens with species not normally found in the oral cav- ity (pathogens or gut flora) or overgrowth of normal oral flora were classified as having abnor- mal oral flora. Specimens with species not nor- mally found in the oral cavity and originating in the gut were classified as having gut flora. All culture results were reviewed by a microbiolo- gist to ensure the correct identification of gut flora.

Statistical analysesResults were expressed as median (range) for continuous variables and number (percentage) for categorical variables. Group wise comparison was done using Fisher0s exact test for categorical variables and Mann–Whitney U-test for continu- ous variables. Univariable logistic regression analyses were used to evaluate patients’ charac- teristics that could be associated with changed oropharyngeal flora. Multivariable logistic regression analyses were thereafter applied to determine independent factors associated with the presence of abnormal flora or abnormal gut flora in the oropharynx. The same eight vari- ables (age, gender, smoking status, alcohol con- sumption, PPI use ≥ 24 h before culture sampling, antibiotic use ≥ 24 h before culture sampling, BMI and occurrence of diabetes) were tested in both ward and ICU patients. In the ICU patients, we added APACHE II score and ‘days in the hospital before ICU admission’ into the analyses. A P value < 0.05 was considered

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

795

significant and all statistical tests were two- tailed. We performed all analyses using SPSS 24 (SPSS Inc, Chicago, IL, USA). Power calcula- tions for the primary endpoint (abnormal oropharyngeal flora) indicated a power of 0.80 with the above selected sample size. Prior data indicated an abnormal flora with a probability of 0.02 among controls and a probability of 0.15 was expected among patients.10

ResultsIn this clinical observational study, we included a total of 487 individuals: 77 controls, 193 ward patients and 217 ICU patients. The baseline characteristics of the patients are shown in Table 1. The three study groups differed signifi- cantly in their baseline characteristics as expected. Smoking, high BMI and antibiotic use≥ 24 h before culture sampling was significantlymore common among the ICU patients as com- pared to the controls and ward patients. Declared alcohol consumption was significantly lower and median age was significantly higher in the ward patients as compared to the controls and ICU patients. Proton pump inhibitor use≥ 24 h before culture sampling was significantlymore common among ICU and ward patients as compared to controls. Regarding the hospi- talised patients, 47% of the ICU patients and 9% of the ward patients had undergone antibi- otic therapy ≥ 24 h before culture sampling. Among the ICU patients, cefotaxime was the

most commonly used antibiotic, followed by piperacillin/tazobactam. Among the ward patients, flucloxacillin and clindamycin were the most common antibiotics used although types of antibiotics used varied more and were less often broad spectrum as compared to the ICU patients.

The causes of hospital or ICU admission are shown in Table 2. Among the ward patients acute or elective abdominal or orthopaedic sur- gery were the most common causes of admis- sion. Sepsis, respiratory failure and cardiovascular disease (insufficiencies, arrests and emergency aneurysms) were the most com- mon causes of admission among the ICU patients. The ICU patients stayed for a median of 1 day (range 0–227) in the hospital before ICU admission. Fifty-two per cent (n = 111) of the ICU patients were admitted to the ICU within 24 h of hospital admission, 74% (n = 159) stayed in the hospital < 5 days before ICU admission.

Oropharyngeal cultures were completed for all487 patients. The missing data on risk factors were as follows: Age 0%, gender 0%, BMI 4%, smoking status 8%, diabetes 7%, alcohol con- sumption 8%, PPI use 4%, antibiotic use 4%, APACHE score and days in the hospital before ICU admission 1% (only ICU patients).

The results from the oropharyngeal cultures in the three study groups are shown in Figure 1 and Tale 3. Abnormal oropharyngeal flora was significantly more common among ICU and

Table 1 Patient characteristics (n = 487).

Variable Controls (n = 77) Ward patients (n = 193) ICU patients (n = 217)

Age 69 (26–85) 73 (22–97)* 66 (20–90)Gender, male 35 (46) 100 (52) 122 (56)BMI 26 (16–44) 26 (15–43) 27 (17–63)*Current or ex-smoker 20 (31) 60 (31) 89 (46)*Diabetes 6 (9) 35 (18) 50 (25)Alcohol consumption > 2x per week 8 (12) 2 (1)* 24 (13)PPI use ≥ 24 h before culture sampling 11 (17) 60 (31) 76 (36)Antibiotic use ≥ 24 h before culture sampling 0 (0) 18 (9) 99 (47)*APACHE II score – – 22 (1–45)Days in the hospital before ICU admission – – 1 (0–227)

Data are presented as medians (ranges) or numbers (valid percentages). BMI, body mass index; ICU, intensive care unit; PPI, proton pump inhibitor; APACHE II, Acute Physiology and Chronic Health Evaluation II. *Variable differed significantly as compared to both other groups

A. TRANBERG ET AL.

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

796

(P < 0.05).

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

797

ward patients as compared to controls (62.2% vs. 1.3%, P < 0.001 and 10.4% vs. 1.3%,P = 0.010). Abnormal oropharyngeal flora wasmore frequent in ICU patients than in ward patients (62.2% vs. 10.4%, P < 0.001). Colonisa- tion of gut flora in the oropharynx was also sig- nificantly more common among ICU patients compared with ward patients or controls (26.3% vs. 4.7%, P < 0.001 and 26.3% vs. 1.3%,P < 0.001). The occurrence of gut flora in theward patients was not significantly different from that seen in controls (4.7% vs. 1.3%, P = 0.29).

Several different species grew in the oropha- ryngeal cultures in the three study groups (Table 3), and about 32% of the cultures were polymicrobial. Overall, the organisms most fre- quently isolated from the oropharyngeal cul- tures were candida species (31%). The most frequently isolated respiratory pathogens were Staphylococcus aureus (14.3%), H. influenzae (5.7%)and Group B streptococci (1.9%). Gut flora waspresent in 41.4% of abnormal oropharyngeal cultures, and the most frequently isolated gut pathogens were Escherichia coli (6.7%), Enterobac- ter species (5.7%), Enterococcus faecium (5.2%), Enterococcus faecalis (5.2%), Klebsiella species(4.8%), Pseudomonas aeruginosa (4.3%) and Steno-trophomonas maltophilia (3.8%).

Multi-resistant bacteria were isolated from only one patient, containing extended spectrum beta-lactamase (ESBL)-producing Enterobacteri- aceae. Only one patient in the control group had abnormal oropharyngeal flora, the gut pathogenP. aeruginosa.

We analysed the determinants of abnormal oropharyngeal and gut flora in the oropharynx using univariable and stepwise multivariable

analysis (Table 4a and b). We found no signifi- cant factors associated with abnormal oropha- ryngeal flora in the ward and ICU patients (Table 4a and b).

Restricting the analyses to growth of gut flora in the oropharynx, PPI use emerged as the strongest independent predictor in ward and ICU patients in the multivariable analysis (Table 4a and b; OR 4.82 and 2.27, respec- tively). In the ICU group, higher BMI was also independently associated with the presence of gut flora in the oropharynx (OR 1.08). Diabetes, in contrast, was associated with lower occur- rence of gut flora in the oropharynx (OR 0.34) in the ICU group. The length of hospital stay (number of days) prior to ICU admission was significantly associated with the presence of gut flora in the oropharynx (OR 1.05). Antibiotic therapy ≥ 24 h before inclusion was not associ- ated with changes in flora in either the ward or ICU patients. An analysis of risk factors in the control group was not possible due to the low

Fig. 1. Results of the oropharyngeal cultures obtained from the three study groups.

Table 2 Cause of hospital or ICU admission.

Cause of admission ICU patients (n = 217) Cause of admission Ward patients (n = 193)

Sepsis, septicaemia 50 (23.4) Acute orthopaedic surgery 57 (29.5)Cardiovascular 49 (22.9) Abdominal surgery 48 (24.9)Respiratory insufficiency 51 (23.8) Elective orthopaedic surgery 36 (18.7)Abdominal 30 (14.0) Respiratory disease 21 (10.9)Trauma 14 (6.5) Cardiovascular disease 15 (7.8)

PPI AND CHANGES IN ORAL FLORA

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

798

Coma 8 (3.8) Infection 5 (2.6)Other 12 (5.6) Other 11 (5.7)

Data are presented as number (valid percentage); ICU, intensive care unit.

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

799

Controls Ward patients ICU patients All togetherOrganism n (%) n (%) n (%) n (%)

Respiratory tract pathogensStaphylococcus aureus 4 (17.4) 26 (14.0) 30 (14.3)Haemophilus influenzae 6 (26.1) 6 (3.2) 12 (5.7)Streptococcus group B 4 (2.2) 4 (1.9)Moraxella catarrhalis 3 (1.6) 3 (1.4)Beta-haemolytic streptococcus group A 2 (1.1) 2 (1.0)Beta-haemolytic streptococcus group G 1 (4.3) 1 (0.5) 2 (1.0)Streptococcus pneumoniae 1 (0.5) 1 (0.5)Staphylococcus epidermidis 1 (0.5) 1 (0.5)

Gut floraEscherichia coli 2 (8.7) 12 (6.5) 14 (6.7)Enterobacter species 1 (4.3) 11 (5.9) 12 (5.7)Enterococcus faecium 11 (5.9) 11 (5.2)Enterococcus faecalis 1 (4.3) 10 (5.4) 11 (5.2)Klebsiella species 1 (4.3) 9 (4.8) 10 (4.8)Pseudomonas aeruginosa 1 (100) 1 (4.3) 7 (3.8) 9 (4.3)Stenotrophomonas maltofilia 8 (4.3) 8 (3.8)Citrobacter species 2 (8.7) 2 (1.1) 4 (1.9)Serratia marcenes 3 (1.6) 3 (1.4)Proteus species 2 (8.7) 1 (0.5) 3 (1.4)Morganella morganii 1 (0.5) 1 (0.5)ESBL 1 (0.5) 1 (0.5)

YeastCandida species 2 (8.7) 63 (33.9) 65 (31.0)

Other origin 3 (1.5) 3 (1.4)Total number of pathogens cultured overall 1 (100) 23 (100) 186 (100) 210 (100)

occurrence of abnormal flora and gut flora (n = 1, 1.3%; Table 3).

DiscussionThis study demonstrates an early and signifi- cantly increased incidence of abnormal pharyn- geal flora in both ward and ICU patients as compared to controls in the community. The incidence of abnormal oropharyngeal flora was also significantly higher in ICU patients as com- pared to ward patients. In critically ill ICU patients, early oropharyngeal colonisation by gut flora was common. Proton pump inhibitor use was the strongest independent factor associ- ated with the presence of gut flora in both ward and ICU patients.

The detection of abnormal gut flora in the oropharynx of ICU patients corroborates the results of Johanson et al. from 1969.11 They

showed an increased incidence of Gram-nega- tive bacilli in oropharyngeal swabs from ward (19%) and moribund patients (38%) within 96 h of admission. The ward patients consisted of patients from the orthopaedic surgery service and the moribund patients were selected from the medical service. In the present study, the frequency of gut flora is lower among ward patients (4.7% to 19%) than in the work of Johanson et al. although our definition of gut flora includes not only Gram-negative bacilli but also other gut microbes (Table 3). One explanation for this difference could be the varying time frames for the collection of oropharyngeal swabs (24 h vs. 96 h from hos- pital admission), as studies have shown a shift towards Gram-negative bacilli with an increased length of stay.3 However, there are also other important differences between these two studies. Over the past 50 years, medicine

A. TRANBERG ET AL.

The percentage of positive cultures is related to the total number of positive cultures for all bacterial species in each group. ESBL, Extendedspectrum beta-lactamase-producing Enterobacteriaceae.

Table 3 Microbial isolation from oropharyngeal cultures in the three study groups.

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

800

Table 4 Analysing potential predictors of abnormal flora in the oropharynx in (a) ward patients (b) ICU patients.

Univariable analysis Multivariable analysis

Factor OR 95% CI P value OR 95% CI P value

(a)Abnormal oropharyngeal flora

G

(b)Abnormal oropharyngeal flora

Age 1.006 0.987–1.025 0.543 – – –Male gender 1.483 0.847–2.596 0.168 – – –Current or ex-smoker 0.694 0.384–1.255 0.227 – – –Alcohol consumption > 2x per week 1.154 0.466–2.856 0.757 – – –PPI use ≥ 24 h before culture sampling 1.585 0.872–2.881 0.131 – – –Antibiotic use ≥ 24 h before culture sampling 1.107 0.632–1.937 0.723 – – –

BMI 1.009 0.964–1.055 0.712 – – –Diabetes 0.797 0.412–1.542 0.500 – – –APACHE II 1.013 0.975–1.054 0.505 – – –Days in the hospital before ICU admission 1.024 0.991–1.059 0.150 – – –

Gut flora in the oropharynxAge 1.028 1.004–1.053 0.022 – – –Male gender 0.912 0.495–1.680 0.767 – – –Current or ex-smoker 0.981 0.514–1.873 0.953 – – –Alcohol consumption > 2x per week 1.55 0.617–3.891 0.351 – – –PPI use ≥ 24 h before culture sampling 1.767 0.948–3.295 0.073 2.272 1.022–5.054 0.044Antibiotic use ≥ 24 h before culture sampling 1.226 0.657–2.287 0.522 – – –

BMI 1.047 0.998–1.098 0.062 1.083 1.021–1.148 0.008Diabetes 0.625 0.287–1.363 0.238 0.338 0.121–0.943 0.038APACHE II 1.038 0.993–1.084 0.098 – – –Days in the hospital before ICU admission 1.047 1.015–1.079 0.004 1.052 1.015–1.090 0.005

The table shows results of the univariable and multivariable logistic regression analyses. In the multivariable analysis, blanks (–) are factors not significantly associated with changes in the oropharyngeal flora after stepwise regression. PPI, proton pump inhibitor; BMI, body mass index; APACHE II, Acute Physiology and Chronic Health Evaluation II; OR, odds ratio; CI, Confidence Interval.

has changed, as is reflected in the younger ages and the higher mortality rate seen for the patients in 1969 as compared to those in the present study (Table 1). Similarly, microbiolog- ical procedures have evolved. The older

clientele in our study show the shift towards an older population receiving advanced health care in Sweden.

The secondary goal of this study was to explore whether certain patient characteristics

PPI AND CHANGES IN ORAL FLORA

Age 1.029 0.991–1.070 0.140 – – –Male gender 0.867 0.342–2.197 0.763 – – –Current or ex-smoker 1.551 0.599–4.018 0.366 – – –Alcohol consumption > 2x per week 0.000 0.000 0.999 – – –PPI use ≥ 24 h before culture sampling 1.551 0.599–4.018 0.366 – – –Antibiotic use ≥ 24 h before culture sampling 1.859 0.488–7.076 0.363 – – –BMI 0.989 0.912–1.072 0.780 – – –Diabetes 0.215 0.028–1.664 0.141 – – –

ut flora in the oropharynxAge 1.039 0.979–1.102 0.211 – – –Male gender 0.854 0.222–3.281 0.818 – – –Current or ex-smoker 1.829 0.473–7.066 0.382 – – –Alcohol consumption > 2x per week 0.000 0.000 0.999 – – –PPI use ≥ 24 h before culture sampling 4.815 1.162–19.96 0.030 4.815 1.162–19.96 0.030Antibiotic use ≥ 24 h before culture sampling 3.000 0.574–15.67 0.193 – – –BMI 0.939 0.828–1.066 0.331 – – –Diabetes 0.551 0.067–4.557 0.581 – – –

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

801

were correlated with disturbed oropharyngeal flora. In our multivariable regression model, we could not identify any independent factors asso- ciated with abnormal oropharyngeal flora although such associations have been found in other studies.11,12 The definition of abnormal oropharyngeal flora includes a wide range of pathogens of various origins (e.g. candida, gut bacteria, contamination from caregivers, or an overgrowth of the patient’s own flora) and dif- ferent pathogenesis and hence, the risk factors associated with their appearance likely differ as well. The heterogeneity in the oral flora may explain the lack of any significant factors being associated with abnormal flora.

When analysing the presence of gut flora inthe oropharyngeal tract via the multivariable regression model (Table 4a and b), PPI use emerged as a strong independent risk factor among both ward and ICU patients. Several studies have shown that PPI use increases the risk of nosocomial pneumonia.6,9,14,15 The result of this study reinforces the hypothesis that PPI use increases the risk of pneumonia by chang- ing the oral flora to harbour gut bacteria which then may be microaspirated into the lungs. Many ICU studies have focused on choosing the ‘right’ antacid medication, to find a treatment that minimises risk of gastrointestinal bleeding while not increasing the risk for VAP during the ICU period.7,16 Those studies have investi- gated risk events for VAP that occur in the ICU (e.g. supine position, gastrointestinal feeding, abdominal surgery, intubation time, APACHE scores and type of antacid etc.), as opposed to our study that focuses on patient characteristics present before hospitalisation or advanced ICU care. Most of our knowledge concerning nosoco- mial pneumonia and its associated risk factors is based on the above-mentioned ICU studies.

Only a handful of publications have studiedthe risk factors associated with nosocomial pneumonia among ward patients,3 and few studies have investigated the risk factors present before hospitalisation. Some early studies have showed that PPI use increases the risk of devel- oping abnormal oropharyngeal flora.17 Our results confirm those of these early studies and offer additional insight into the pathogenesis of nosocomial pneumonia and may help to deter- mine which patients are at an increased risk of

this costly complication. Patients admitted to the hospital and/or ICU that are on daily PPI medication have a higher risk of gut flora colonising the oropharyngeal tract although gut flora colonisation can occur without previous PPI medication.

Nosocomial pneumonia is often considered as a sign of degree of sickness in the patient, where a weakened host immune system is unable to fight off pathogenic bacteria that are flourishing in places they normally would not reside, due to disease or hospital events.

Aside from PPI use, our analysis showed that additional factors were associated with oropha- ryngeal gut flora in ICU patients. Higher BMI emerged as a significant factor. One probable explanation is a higher incidence of gastro-oeso- phageal reflux disease (GERD) and lower sphincter tonus in obese patients (BMI > 30 kg/ m2), due to a distorted pressure gradient across the gastro-oesophageal junction,18,19 giving gut bacteria access to the oropharynx. A recent study12 showed that an increase in BMI signifi- cantly reduced the frequency of abnormal flora in the oropharyngeal tract. However, this was with respect to all abnormal flora and not specifically gut flora. Our results did not show an increased risk of abnormal flora, only an increased risk of gut flora, with increasing BMI, which supports the explanation suggested above.

Furthermore, in the ICU group, diabetesreduced the frequency of gut flora in the oropharynx. Diabetes is associated with an increased risk of developing oral candidosis, and several contributing mechanisms are possi- ble, such as oral pH, salivary production and neural dysfunction.20 Could these changes or an increased incidence of oral candida infection influence the gut flora0s ability to thrive in the oral cavity?

An increased length of stay in the hospital prior to ICU admission was also associated with gut flora colonisation, but not with abnormal flora in our study. These patients were hospi- talised for a few days, receiving medical treat- ment and/or undergoing surgery, frequently having some sort of complication/aggravation that led to a transfer to the ICU. The 1969 study concluded that the severity of sickness in the patients was the factor that best correlated with

A. TRANBERG ET AL.

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

802

the emergence of a Gram-negative oropharyn- geal flora.11 This finding probably explains the association with length of stay seen in our study.

Some of the data from the ICU patients are up to 12 years old. One might assume that the nor- mal oropharyngeal flora has changed in this population over this time period, with antibiotic prescription and resistance, immigration and travelling bringing new bacteria into the equa- tion. However, the microbiology department performing our analyses state that the definition of a normal oropharyngeal flora has remained the same during this period of time in the south of Sweden. In our study, only one participant had multi-resistant bacteria in the oropharynx, a surprising finding when bacterial resistance is an increasing problem worldwide. Antibiotic consumption has gone down in the last decade in Sweden, especially prescription and con- sumption in the primary care facilities.21 This change in treatment has diminished the devel- opment of resistance in the community.

We did not find that antibiotic use prior tohospitalisation and/or ICU admission was asso- ciated with a higher incidence of abnormal oropharyngeal flora in the studied population. This agrees with the study by Johansson from 1969 and a more recent study by Frandah et al. 2015, showing that antibiotic therapy was not associated with a higher incidence of abnormal oropharyngeal flora. A likely explanation is that the oropharyngeal flora is to a lesser extent and more slowly affected by both digested and intra- venous antibiotics than the flora in the intestines.22–24

There are other important limitations to this study. In our multivariable regression analyses, we did not include all of the potential factors involved in developing abnormal oropharyngeal flora. For example, other investigators have observed that renal failure reduces the fre- quency of abnormal flora.12 Similarly, we did not adjust for chronic lung disease, which has been shown to decrease the clearance of patho- gens from the lungs.25 Other factors that may be relevant but were not evaluated include degree of physical fitness and patient comorbidities such as cancer or immunosuppressive therapy.

Several factors concerning sampling and anal-ysis are difficult to control. Recent tooth

brushing or water consumption has not always been evaluated before sampling. Microbiological procedures also differ between institutions, making comparison with other studies difficult. Finally, although there is clear association between patient factors and changes in the oropharyngeal flora, we cannot claim causality between the two.

This study has shown that an abnormal oropharyngeal flora is a common and early fea- ture among hospitalised patients, especially in critically ill patients. It is well established that PPI use during hospitalisation is a factor associ- ated with nosocomial pneumonia. We show that PPI use before admission to a hospital ward or an intensive care unit is a strong factor associ- ated with colonisation of gut flora in the oropharyngeal tract. Recognition of this will help to take appropriate measures – such as choosing an appropriate antacid, oropharyngeal cultures, patient positioning, physiotherapy, oral decontamination and treatment with probiotics already at admission.

AcknowledgementsWe would like to thank Ann-Cathrine Petersson at the Microbiology Department at Skane University Hospital in Lund, for her invaluable help and input during this project.

References1. Robinson J. Colonization and infection of the

respiratory tract: what do we know? Paediatr Child Health 2004; 9: 21–3.

2. Suetens C, Hopkins S, Kolman J, Ho€gberg LD. Point prevalence survey of healthcare-associated infections and antimicrobial use in European long- term care facilities, May–September 2010. Luxembourg: European Centre for Disease Prevention and Control, Publications Office, 2013.

3. Di Pasquale M, Aliberti S, Mantero M, Bianchini S, Blasi F. Non-intensive care unit acquired pneumonia: a new clinical entity? Int J Mol Sci 2016; 17: 287–304.

4. A’Court C, Garrard CS. Nosocomial pneumonia in the intensive care unit: mechanisms and significance. Thorax 1992; 47: 465–73.

5. Stoutenbeek CP, van Saene HK, Miranda DR, Zandstra DF. The effect of selective

PPI AND CHANGES IN ORAL FLORA

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

803

decontamination of the digestive tract on colonisation and infection rate in multiple trauma patients. Intensive Care Med 1984; 10: 185–92.

6. Alshamsi F, Belley-Cote E, Cook D, Almenawer SA, Alqahtani Z, Perri D, Thabane L, Al-Omari A, Lewis K, Guyatt G, Alhazzani W. Efficacy and safety of proton pump inhibitors for stress ulcer prophylaxis in critically ill patients: a systematic review and meta-analysis of randomized trials. Crit Care 2016; 20: 120–32.

7. MacLaren R, Reynolds PM, Allen RR. Histamine-2 receptor antagonists vs proton pump inhibitors on gastrointestinal tract hemorrhage and infectious complications in the intensive care unit. JAMA Intern Med 2014; 174: 564–74.

8. Eom CS, Jeon CY, Lim JW, Cho EG, Park SM, Lee KS. Use of acid-suppressive drugs and risk of pneumonia: a systematic review and meta-analysis. CMAJ 2011; 183: 310–9.

9. Laheij RJ, Sturkenboom MC, Hassing RJ, Dieleman J, Stricker BH, Jansen JB. Risk of community- acquired pneumonia and use of gastric acid- suppressive drugs. JAMA 2004; 292: 1955–60.

10. Bonten MJ, Gaillard CA, de Leeuw PW, Stobberingh EE. Role of colonization of the upper intestinal tract in the pathogenesis of ventilator-associated pneumonia. Clin Infect Dis 1997; 24: 309–19.

11. Johanson WG, Pierce AK, Sanford JP. Changing pharyngeal bacterial flora of hospitalized patients. Emergence of gram-negative bacilli. N Engl J Med 1969; 281: 1137–40.

12. Frandah W, Colmer-Hamood J, Mojazi Amiri H, Raj R, Nugent K. Oropharyngeal flora in patients admitted to the medical intensive care unit: clinical factors and acid suppressive therapy. J Med Microbiol 2013; 62: 778–84.

13. Klarin B, Molin G, Jeppsson B, Larsson A. Use of the probiotic Lactobacillus plantarum 299 to reduce pathogenic bacteria in the oropharynx of intubated patients: a randomised controlled open pilot study. Crit Care 2008; 12: R136.

14. Herzig SJ, Howell MD, Ngo LH, Marcantonio ER. Acid-suppressive medication use and the risk for hospital-acquired pneumonia. JAMA 2009; 301: 2120–8.

15. Williams C, McColl KE. Review article: proton pump inhibitors and bacterial overgrowth. Aliment Pharmacol Ther 2006; 23: 3–10.

16. Bateman BT, Bykov K, Choudhry NK, Schneeweiss S, Gagne JJ, Polinski JM, Franklin JM, Doherty M, Fischer MA, Rassen JA. Type of stress ulcer

prophylaxis and risk of nosocomial pneumonia in cardiac surgical patients: cohort study. BMJ 2013; 347: f5416.

17. Driks MR, Craven DE, Celli BR, Manning M, Burke RA, Garvin GM, Kunches LM, Farber HW, Wedel SA, McCabe WR. Nosocomial pneumonia in intubated patients given sucralfate as compared with antacids or histamine type 2 blockers. The role of gastric colonization. N Engl J Med 1987; 317: 1376–82.

18. Pandolfino JE, El-Serag HB, Zhang Q, Shah N, Ghosh SK, Kahrilas PJ. Obesity: a challenge to esophagogastric junction integrity. Gastroenterology 2006; 130: 639–49.

19. Friedenberg FK, Xanthopoulos M, Foster GD, Richter JE. The association between gastroesophageal reflux disease and obesity. Am J Gastroenterol 2008; 103: 2111–22.

20. Soysa NS, Samaranayake LP, Ellepola AN. Diabetes mellitus as a contributory factor in oral candidosis. Diabet Med 2006; 23: 455–9.

21. Mo€lstad S, Lo€fmark S, Carlin K, Erntell M, Aspevall O, Blad L, Hanberger H, Hedin K, Hellman J, Norman C, Skoog G, StOalsby-LundborgC, Tegmark Wisell K, AOhr'en C, Cars O. Lessonslearnt during 20 years of the Swedish strategic programme against antibiotic resistance. Bull World Health Organ 2017; 95: 764–73.

22. Zaura E, Brandt BW, Teixeira de Mattos MJ, Buijs MJ, Caspers MP, Rashid MU, Weintraub A, Nord CE, Savell A, Hu Y, Coates AR, Hubank M, Spratt DA, Wilson M, Keijser BJ, Crielaard W. Same exposure but two radically different responses to antibiotics: resilience of the salivary microbiome versus long-term microbial shifts in feces. MBio 2015; 6: e01693–15.

23. Edlund C, Bergan T, Josefsson K, Solberg R, Nord CE. Effect of norfloxacin on human oropharyngeal and colonic microflora and multiple-dose pharmacokinetics. Scand J Infect Dis 1987; 19: 113–21.

24. Nord CE, Heimdahl A, Lundberg C, Marklund G. Impact of cefaclor on the normal human oropharyngeal and intestinal microflora. Scand J Infect Dis 1987; 19: 681–5.

25. Cabello H, Torres A, Celis R, El-Ebiary M, Puig de la Bellacasa J, Xaubet A, Gonz'alez J, Agust'ı C, Soler N. Bacterial colonization of distal airways in healthy subjects and chronic lung disease: a bronchoscopic study. Eur Respir J 1997; 10: 1137–44.

A. TRANBERG ET AL.

Acta Anaesthesiologica Scandinavica 62 (2018) 791–800

ª 2018 The Acta Anaesthesiologica Scandinavica Foundation. Published by John Wiley & Sons Ltd

804

Paper II

APMIS © 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology .DOI 10.1111/apm.13087

Disturbance in the oropharyngeal microbiota in relation to antibiotic and proton pump inhibitor medication and

length of hospital stay

ANNA TRANBERG, CAROLINA SAMUELSSON and BENGT KLARIN

Division of Intensive and Perioperative Care, Skane University Hospital, Lund, Sweden

The microbiota of the oropharyngeal tract normally comprises a large variety of bacteria that helps maintain a balanced local environment with regard to aspects such as saliva pH and orodental health (1). Illness and medication can disturb this balance and thereby allow pathogens to colonize the oropharyngeal tract or facilitate the overgrowth of certain other species (2, 3). Examples of such opportunistic pathogens are gut bacteria, bacteria from the upper respiratory tract, and yeast species, as well as any combination of these three groups. Presence of oropharyngeal pathogens, even at low numbers, can be identified by conventional culture techniques. Microaspiration of these pathogens can lead to colonization of the lower respiratory tract and increase the risk of nosocomial pneumonia (NP) (4, 5). Colonization of the oropharyngeal tract

Received 24 February 2020. Accepted 21 September 2020

by gut bacteria is associated with general severity of illness (3) and with proton pump inhibitor (PPI) medication; the latter relation has been shown in intensive care cohorts (6) as well as in non-ICU cohorts at hospital admission (7). PPI use has been associated with a distorted gut microbiota, which in turn can lead to development of enteric infections in general and Clostridium di cile ffi infection specifi- cally (8, 9).

PPI use also increases the risk of developing NP (10, 11), hospital-acquired pneumonia (HAP) and the more serious ventilator-associated pneumonia (VAP) (12–14). The incidence of NP is as high as 5–10 cases per 1000 hospitalizations, and NP is the leading cause of mortality due to hospital-acquired infections (15). An NP diagnosis increases hospital length of stay (LOS) by 7–10 days, and VAP pro- longs mechanical ventilation times as well as ICU LOS (16). Also, hospital costs have been shown to

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, 1which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

JOURNAL OF PATHOLOGY, MICROBIOLOGY AND IMMUNOLOGY

Tranberg A, Samuelsson C, Klarin B. Disturbance in the oropharyngeal microbiota in relation to antibiotic and proton pump inhibitor medication and length of hospital stay. APMIS. 2020.

The aim of this study was to investigate the appearance of a disturbed oropharyngeal microbiota during hospitalization and explore the patient characteristics that maybe associated with such a disturbance. Oropharyngeal swabs were col- lected from 134 patients at hospital admission and every 3–4 days thereafter. The samples were cultivated to determine the presence of a disturbed microbiota, which, in turn, was subcategorized into respiratory tract pathogens, gut micro- biota and yeast species. Demographics, medical history data and hospitalization events were compared. The percentage of disturbed oropharyngeal microbiota increased significantly with length of stay (LOS). Receiving antibiotic treatment during the hospitalization tended to be associated with a disturbed microbiota (OR 2.75 [0.99–7.60]). Proton pump inhibitor (PPI) medication and receiving antibiotics before hospitalization were associated with the development of a disturbed oropharyngeal microbiota with colonization of gut pathogens (OR 3.49 [1.19–10.2] and OR 4.52 [1.13–18.1], respectively), while acute hospital admission was associated with a lower risk of colonization of gut pathogens (OR:0.23 [0.074–0.72]). The risk of developing a disturbed oropharyngeal microbiota increased with LOS in hospitalized patients. PPI medication and receiving antibiotics before hospitalization were independent risk factors for developing oropharyngeal colonization of gut pathogens.

Key words: Oropharyngeal microbiota; PPI; hospitalization; antibiotics; nosocomial infection.

Anna Tranberg, Anna Tranberg, Department of Intensive and Perioperative Care, Skane University Hospital, Getinge- vagen 4, 221 84 Lund, Sweden. e-mail: anna.tranberg_lindqvist@med.lu.se

2 © 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology

TRANBERG et al.

rise by a factor of 5 in non-ICU patients with a diagnosis of HAP (17). Therefore, early detection of a disturbed oropharyngeal microbiota, along with preservation of a normal oropharyngeal microbiota, may be a way to reduce the occurrence of NP to the benefit of the patient, the health care system and society.

The primary aim of the present study was to investigate the appearance of a disturbed oropharyn- geal microbiota during the period of hospitalization. Our second objective was to identify the hospitaliza- tion events and/or patient characteristics associated with the development of a disturbed oropharyngeal microbiota during hospitalization. Given its primary and secondary objectives, the study also has the potential of describing a baseline from which poten- tial interventions to maintain a normal oropharyn- geal microbiota can be hypothesized.

MATERIALS AND METHODS

This clinical study was conducted at Skane University Hospital, Lund, Sweden, using an observational, longitu- dinal and comparative approach.

Ethical approval

The study protocol was reviewed and approved by the Regional Ethical Review Board, Lund, Sweden (no. 2013/ 764). Informed consent, including permission to collect and publish anonymous data, was obtained from all patients at enrolment.

Study population

During the period 6 February 2014 to 1 February 2017, 145 patients were enrolled in the study using the following inclusion criteria: age ≥ 18 years; possible to obtain the first oropharyngeal swab (OPS) within 24 h of hospital admission; expected LOS of> 72 h. The exclusion criterion was hospitalization in the preceding 14 days. Patient enrolment occurred during a 3-year period with the aim of including as many patients as possible during this time with the resources at hand. Being an observational and non-interventional study, no a priori power calculations were performed.

The patients were identified and enrolled at nine di erentff wards: a medical high-dependency unit, a surgical high-de- pendency unit, two orthopaedic wards, two surgical wards and three internal medicine wards. Patients who changed wards during their stay were still eligible to remain in the study, and swabs were collected according to protocol.

Patient data

A standardized case report form was used to record the following patient data: age, gender, smoking status, alco- hol consumption, physical fitness, body mass index (BMI), diabetes, ongoing systemic cortisone treatment, PPI

medication, ongoing antibiotic treatment initiated >24 h before admission (referred to as ‘antibiotics before hospi- talization’), lung disease at admission and whether the patient admission was acute or planned. The following hospitalization events were recorded: antibiotic treatment lasting >24 h during hospitalization (referred to as ‘antibi- otics during hospitalization’), abdominal surgery during hospitalization, and occurrence of hospital-acquired pneu- monia. Patients admitted for elective orthopaedic surgery who only received a perioperative three-dose regime of either cloxacillin or clindamycin were not classified as ‘an- tibiotics during hospitalization’, since that prophylactic perioperative treatment lasted for <24 h, the rationale being the vast scope of evidence, which suggests that short perioperative antibiotic prophylaxis does not markedly influence the oropharyngeal milieu (18). Definitions of variables and abbreviations are presented in Table 1.

Oropharyngeal sample collection

Sterile Nylon® flocked swabs with 1 mL liquid Amies medium (ESwabTM 480C, COPAN Diagnostics Inc., Mur- rieta, CA, USA), were used to collect the samples. A ton- gue depressor and a flashlight were used to gain access to and visualize the pharynx. The swab was inserted into the posterior pharynx and rubbed over both the tonsillar pil- lars and the posterior oropharynx, carefully avoiding touching the tongue, teeth and gums. The sample was then transported to the Department of Microbiology for culti- vation analyses.

The first OPS was collected within 24 h of the hospital admission (day 1) and the procedure was repeated on day 3 and approximately every fourth day thereafter through- out the patient’s entire LOS. In all other respects, the patients received standard care according to their diag- noses and the clinical decisions of the responsible physi- cians.

Microbiological procedures

The OPSs were processed by extended microbiological procedures at the Department of Clinical Microbiology, Skane University Hospital in Lund. The laboratory is accredited by the accreditation body (SWEDAC) desig- nated by the Swedish government and is formally recog- nized as competent according to European and international standards.

For bacteria cultivation, sampling media were inocu- lated on five types of agar plates (three selective, one dif- ferentiating and one non-selective). All plates were produced in-house, sometimes using commercially avail- able media components (5% horse blood, haematin agar and Uriselect 4 agar), as listed below:

1. Agar with 5% horse blood (LabM, Heywood, Lancashire, UK) supplemented with 10 mg/L colistin and 15 mg/L nalidixic acid with an optochin disc (selective)

2. Agar with 5% horse blood supplemented with 2 mg/L gentamicin and 25 mg/L nalidixic acid for Gram-positive cocci including S. pneumonia (selective)

© 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology 3

OROPHARYNGEAL MICROBIOTA AND HOSPITALISATION

Table 1. Definitions of variables (short designations within parentheses) Variable DefinitionAge dichotomous yes/no, >70 yearsGender dichotomous yes/no, maleSmoking status dichotomous yes/no, current or ex-smokerAlcohol consumption dichotomous yes/no, alcohol intake> 2 times/weekPhysical fitness dichotomous yes/no, ability to climb two flights of stairsBody mass index (BMI) dichotomous yes/no, BMI> 35Occurrence of diabetes (diabetes) dichotomous yes/noOngoing systemic cortisone treatment (cortisone treatment)

dichotomous yes/no

PPI medication dichotomous yes/noOngoing antibiotic treatment initiated> 24 h before admission (antibiotics before hospitalisation)

dichotomous yes/no

Lung disease at admission dichotomous yes/noAcute admission to hospital (acute admission) dichotomous yes/noAntibiotic treatment for> 24 h during hospitalization (antibiotics during hospitalization)

dichotomous yes/no

Hospital-acquired pneumonia dichotomous yes/noGastrointestinal surgery during hospitalization (abdominal surgery)

dichotomous yes/no

3. Haematin agar (OxoidTM, Thermo Science, Bas- ingstoke, UK) supplemented with 300 mg/L bacitracin for fastidious Gram-negative rods including H. influenzae (selective)

4. Uriselect 4 agar (Bio-Rad Laboratories, Copen- hagen, Denmark) supplemented with 10 mg/L vancomycin for non-fastidious Gram-negative rods (di erentiating)ff

5. Haematin agar with a colistin disc (non-selective)

The plates were inspected for growth after 16 and 40 h of aerobic, anaerobic or CO2 incubation at 35–37 °C. If an inspection result was ambiguous at 40 h, the plate was incubated for an additional 24 h to obtain a more definite result. Species identification of bacteria was performed using matrix-assisted laser desorption/ionization time-of- flight (MALDI-TOF) mass spectrometry (MALDI Bio- typer Microbial Identification System, Bruker, Boston, MA, USA).

Cultivation and di erentiation of ff Candida spp. were based on colony appearance on CHROM Candida agar (CHROMagar, H€agersten, Sweden) after 48 h of incuba- tion at 35 °C.

Definitions of culture findings

For a sample to be considered representative of ‘oropha- ryngeal microbiota’, several bacterial species normally found in the oropharynx cavity were required to grow on the non-selective haematin plate as determined by visual inspection by experienced senior microbiologists. In the oropharynx, the genera (the taxonomic rank above spe- cies) most commonly found are Streptococcus, Prevotella, Capnocytophaga, Rothia, Campylobacter, Veillonella, Neis- seria and Haemophilus (19, 20), followed by a large group of less common genera. The plates were subsequently inspected for signs of a disturbed oropharyngeal micro- biota, which required growth of species not normally

found in the oropharyngeal cavity or overgrowth of nor- mal oropharyngeal microbiota and/or overgrowth of yeast species.

Samples with a disturbed oropharyngeal microbiota were further assigned to one of three categories: gut microbiota, respiratory tract pathogens or yeast species (see Fig. 1), to aid the analysis, improve understanding of the underlying pathogenesis, and elucidate the results. This classification system with three categories has been used previously (7) and was developed in collaboration with senior microbiologists at the Department of Clinical Microbiology, and senior consultant specialists in Infec- tious Disease at the Skane University Hospital in Lund.

Statistical analyses

Continuous variables were presented with median, mini- mum and maximum values. Dichotomous variables were presented as number and as percentage of total number. For subjects with a normal oropharyngeal microbiota at inclusion, a univariable logistic regression was used to analyse the association between the patients’ characteris- tics (predicting variables) and the development of any type of disturbed oral microbiota (dependent variable), and also, specifically, the development of a disturbed oral microbiota with colonization of gut pathogens.. There- after, a multivariable logistic regression model using the three strongest predicting variables from the univariable analysis regarding colonization of gut pathogens was con- structed in which one additional potential explanatory variable was added to determine whether the model improved or did not improve by including a fourth vari- able. Fisher’s exact test was used to assess the relationship between potential risk factors and HAP.

Statistical analyses were performed using IBM SPSS Statistics 22 for Windows (IBM Corporation, Armonk, NY, USA). Odds ratios (ORs) are presented with a 95% confidence interval (CI). P < 0.05 was regarded as signifi- cant, and all statistical tests were two tailed.

4 © 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology

TRANBERG et al.

Sampling day number 1 3-4 5-8 9-12 13-16 17-20 21-24 25-28 29-32 33-Number of OPSs´ analysed 134 128 87 55 26 15 9 6 3 3

Percentage with disturbed microbiota 11% 13% 15% 22% 8% 33% 22% 50% 100% 67%1 Sa Hi2 Pa Pv3 Sa4 Hi5 Kp Ea Ec6 Hi

Colour and Species Key 7 Hi Hi8 Sa Sa 9 Ko 10 Hi Hi Hi Sa 11 Ec Ko 12 Hi Hi13 Hi14 K15 Kp Hi16 Ko Ko17 Ef 18 C Kp19 Ef20 Cf Cf Ef21 C22 Sa 2324 Ea Ea 25 C26 Efs Efs 27 C C28 A29 C3031 Ecl C32 Sa Hi Sa Hi 33 Efs C34 Efs35 Ef36 Ec Ec Ef Ef37 Ck Ck Ck Ec Efs Efs Ck Efs Ck Efs38 Ecl Ecl Ef C Efs Er39 Ecl Ecl40 Efs41 C42 Efs Ec C Ef K

Fig 1. Oropharyngeal swab (OPS) culture results for the 42 patients who had at least one OPS sample with a disturbed microbiota during their hospitalization. Each horizontal bar represents the duration of a patient’s hospitalization, and the colour indicates the OPS result for each sampling occasion (day number): yellow = normal, blue = respiratory pathogens, terracotta = gut microbiota, green = yeast species for each sampling time point (sampling day number). The figure presents the number of OPSs collected/analysed and the percentage of OPSs with a disturbed microbiota for the total cohort at each sampling time point. The patients are dichotomized according to whether they did or did not receive antibiotics for>24 h during the hospitalization.

RESULTS

Initially, 145 patients were included and provided a first OPS sample. Eleven patients were subsequently excluded due to non-adherence to protocol, 10 due to <2 OPSs during the hospital stay, and one due to inclusion >24 h after hospital admission. Thus,134 patients met all inclusion criteria and con- tributed a total of 466 OPSs. The median number of OPSs per patient was 3 (2–11), and all OPSs were representative of oropharyngeal microbiota.

The baseline patient characteristics and hospital- ization characteristics of the cohort are presented in Table 2. Forty-five patients (34%) were treated

with PPI before and during their hospitalization. Twelve patients (9%) had antibiotic treatment before hospitalization, most of them with flu- cloxacillin or clindamycin. Sixty-six (49%) received antibiotics during hospitalization. Sixteen patients (12%) had ongoing systemic cortisone treatment at admission. Ninety-two (69%) of the 134 patients were acutely admitted to the hospital, and the most common reason for hospitalization in those cases was acute or elective abdominal or orthopaedic sur- gery.

Figure 1 presents the microbiological results for the 42 patients who presented with any type of dis- turbed oropharyngeal microbiota on any of the

Ef EfsAntibiotics during hospitalisation (below line)

C C

No antibiotics during hospitalisation (above line)

Respiratory Tract pathogens Haemophilus influenzae Staphylococcus aureus

Hi Sa

Gut pathogens Acinetobacter species Citrobacter freundii Citrobacter koseri Enterobacter aerogenes Enterobacter cloacae Enterococcus faecalis Enterococcus faecium Enterococcus raffinosus Escherichia coli Klebsiella untyped Klebsiella oxytoca Klebsiella pneumoniaePseudomonas aeruginosa Proteus vulgaris

ACf Ck Ea Ecl Ec Efs

Er Ec K

Ko Kp

YeastCandida albicans

CaNormal oropharyngeal microbiota

© 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology 5

OROPHARYNGEAL MICROBIOTA AND HOSPITALISATION

Table 2. Table–>Patient characteristicsVariable Patient cohort, n = 134Age, years 72 (23–97)Gender, male 69 (51%)Smoking status 38 (28%)Alcohol consumption 77 (57%)Physical fitness 85 (63%)Body mass index 26 (15–43)Diabetes 22 (16%)Cortisone treatment 16 (12%)Proton pump inhibitor medication 45 (34%)Antibiotics before hospitalization 12 (9%)Lung disease at admission 29 (22%)Acute admission 92 (69%)Antibiotics during hospitalization 66 (49%)Hospital-acquired pneumonia 6 (4.5%)Abdominal surgery 29 (22%)See Table 1 for definitions of the variables. Age and BMI are presented as median (range). The other variables are pre- sented as number (valid percentages).

sampling occasions. Approximately 22% of these OPSs had polymicrobial results (i.e. they met the definitions of more than one of the three subclasses of disturbed microbiota). We found that the longer the hospital stay, the greater the proportion of col- lected OPSs with a disturbed microbiota. The majority of OPSs that were collected after day 12 and showed a disturbed oropharyngeal microbiota were subclassed as gut microbiota (Fig. 1).

In 119 patients (89%), the first OPS (at admis- sion) was normal. In this group, the univariable analyses showed that antibiotics given before and during hospitalization predicted development of a disturbed oropharyngeal microbiota (Table 3). In the multivariable analyses, antibiotics during hospi- talization were the only variable close to being sta- tistically significant for the occurrence of a disturbed oropharyngeal microbiota in this group (P = 0.052). Restricting the univariable analyses to colonization of gut pathogens showed that PPI medication and antibiotics before hospitalization were associated with an increased risk of coloniza- tion of gut pathogens, whereas acute hospital admission was associated with a decreased risk (Table 4). In the best-fitting multivariable regres- sion model, ongoing PPI medication was associated with an increased risk of colonization of gut patho- gens, and acute hospital admission was associated with a decreased risk of developing such a dis- turbed microbiota (Table 5).

Both PPI medication and antibiotics before hos- pitalization were associated with an increased risk of acquiring HAP (Table 6).

Cox regression analyses of the data gave no addi- tional information regarding the development of a disturbed oropharyngeal microbiota and its correla- tion to LOS.

DISCUSSION

In this study, we found that a significant propor- tion of a mixed patient cohort admitted to hospital developed a disturbed oropharyngeal microbiota during their hospitalization, and that the propor- tion of included patients who had a disturbed microbiota increased with the length of hospital stay. In patients with a long LOS and a disturbed oropharyngeal microbiota, gut species most com- monly caused the disturbance. The risk of develop- ing HAP was increased among patients treated with PPI and/or antibiotics before their hospital admis- sion, but the numbers were small (N = 5). PPI medication was an independent risk factor for colo- nization of gut pathogens. The subgroup acutely admitted to hospital had a lower risk of acquiring gut microbiota disturbance.

The oropharyngeal bacterial microbiota has a strong impact on the microenvironment in the lower respiratory tract, presumably due to continu- ous microaspiration of bacteria into the lungs (5). Consequently, a disturbed oropharyngeal micro- biota may be a precursor to pneumonia.

In this study, we identified PPI medication as a risk factor for colonization of gut pathogens in the oropharynx and also as a risk factor for developing HAP. These observations corroborate previous studies showing that PPI medication was associated with the development of VAP and HAP (10, 13), and also changes in oropharyngeal microbiota (21). Ongoing antibiotic treatment initiated> 24 h before admission was a risk factor for occurrence of any type of disturbed microbiota during hospitalization. It is well known that antibiotics change the normal microbiota, primarily in the large intestine but also in the oropharyngeal and respiratory tract (22–25).

6 © 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology

TRANBERG et al.

Table 3. Univariable logistic regression analysis of occurrence of a disturbed oropharyngeal microbiota during hospitalization in 119 subjects with a normal

Table 4. Univariable logistic regression analysis of colonization of gut pathogens in the oropharyngeal swabs during hospitalization in 119 subjects with a normal

microbiota at admission microbiota at admissionVariable n OR (95% CI) P value Variable n OR (95% CI) P valueAge, years 0.99 (0.97–1.02) 0.654 Age, years 0.99 (0.96–1.02) 0.426

≤70 57 1.20 (0.51–2.83) 0.683 ≤70 57 0.91 (0.33–2.47) 0.846>70 62 >70 62

Gender GenderFemale 58 1.52 (0.64–3.63) 0.346 Female 58 1.23 (0.45–3.36) 0.693Male 61 Male 61

Smoking status SmokerNo 86 0.69 (0.25–1.89) 0.469 No 86 0.71 (0.22–2.34) 0.572Yes 33 Yes 33

Alcohol consumption Alcohol consumptionNo 46 2.10 (0.81–5.46) 0.127 No 46 2.49 (0.77–8.10) 0.129Yes 73 Yes 73

Physical fitness Physical fitnessNo 39 0.78 (0.32–1.92) 0.592 No 39 1.32 (0.43–4.01) 0.625Yes 80 Yes 80

Body mass index 119 Body mass index≤35 108 2.11 (0.57–7.84) 0.264 ≤35 108 2.32 (0.55–9.76) 0.249>35 11 >35 11

Diabetes DiabetesNo 97 2.35 (0.86–6.40) 0.096 No 97 2.66 (0.87–8.11) 0.086Yes 22 Yes 22

Cortisone treatment Cortisone treatmentNo 104 2.63 (0.84–8.23) 0.095 No 104 2.34 (0.65–8.37) 0.192Yes 15 Yes 15

Proton pump inhibitor medication Proton pump inhibitor medicationNo 80 2.36 (0.98–5.69) 0.056 No 80 4.10 (1.44–11.6) 0.008Yes 39 Yes 39

Antibiotics before hospitalization Antibiotics before hospitalizationNo 109 3.95 (1.05–14.9) 0.042 No 109 4.52 (1.13–18.1) 0.033Yes 10 Yes 10

Lung disease at admission Lung disease at admissionNo 95 0.87 (0.29–2.61) 0.808 No 95 1.16 (0.34–3.90) 0.814Yes 24 Yes 24

Acute admissionNo 38 0.49 (0.20–1.19) 0.116

Acute admissionNo 38 0.23 (0.082–0.66) 0.006

Yes 81 Yes 81Antibiotics during hospitalization Antibiotics during hospitalization

No 59 2.95 (1.17–7.43) 0.021 No 59 2.21 (0.77–6.34) 0.141Yes 60 Yes 60

Abdominal surgery Abdominal surgeryNo 92 1.62 (0.61–4.26) 0.330 No 92 1.90 (0.64–5.67) 0.247Yes 27 Yes 27

There are indications that antibiotic-induced micro- biota changes arise more slowly in the oropharynx as compared with other parts of the gastrointestinal tract (23, 26). If a patient undergoes antibiotic ther- apy before being hospitalized, the e ect of ff that treatment on the oropharyngeal microbiota may therefore occur during the subsequent hospital stay. Ongoing antibiotic treatment> 24 h prior to admis- sion may also be a marker of overall fragility and ‘degree of illness’. As stated by Johanson et al. (3), ‘the appearance of a Gram-negative bacillary microbiota in our patients correlated best with a clinical impression of the degree of illness’.

Those admitted acutely for hospital care in our cohort had a reduced risk of oropharyngeal colo- nization of gut pathogens. This suggests that there may be di erences between patients admit-ff ted for elective procedures and those acutely admitted with respect to characteristics not accounted for in this study. A potential explana- tion is that electively admitted patients’ general health as a whole was undermined by a chronic disorder, whereas a greater proportion of those receiving emergency care were quite healthy prior to their hospitalization.

© 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology 7

OROPHARYNGEAL MICROBIOTA AND HOSPITALISATION

Table 5. Multivariable logistic regression analysis of colonization of gut pathogens in the oropharyngeal swabs during hospitalization in 119 subjects with a normal microbiota at admissionVariable n OR (95% CI) P valueProton pump inhibitor

No 59 2.10 (0.60–7.28) 0.243Yes 60

Antibiotics before hospitalizationNo 109 2.08 (0.41–10.4) 0.375Yes 10

Acute admission

No 38 0.23 (0.074–0.72) 0.012Yes 81

Antibiotics during hospitalizationNo 59 2.10 (0.60–7.28) 0.243Yes 60

Table 6. Hospital-acquired pneumonia (n = 5)Variable Hospital-acquired

pneumoniaP value1

No(n = 114)

Yes(n = 5)

no. (%) no. (%)Proton pump inhibitor medication

No 79 (69.3) 1 (20.0) 0.039Yes 35 (30.7) 4 (80.0)

Antibiotics before hospitalizationNo 108 (94.7) 1 (20.0) <0.001Yes 6 (5.3) 4 (80.0)

Acute admissionNo 36 (31.6) 2 (40.0) 0.654Yes 78 (68.4) 3 (60.0)

Antibiotics during hospitalizationNo 59 (51.8) 0 (0.0) 0.057Yes 55 (48.2) 5 (100.0)

1Fisher’s exact test.

Our results are important. To our knowledge, we are the first to have performed consecutive sam- pling in a mixed cohort of hospitalized ward patients. We are the only group to have used the subcategorization of respiratory tract pathogens, yeast and gut pathogens when studying and describing microbiota changes over time. In an ear- lier study by our group, which involved intensive care patients as well as ward patients, PPI treat- ment was associated with the gut-pathogen colo- nization subclass of disturbed oropharyngeal flora (7). Those results strengthened our belief that it is rational to di erentiate between di erent types ff ff of oropharyngeal microbiota disturbances when trying to understand potential causes and cures. Our pre- sent results support this notion, since it appears that the pathogenesis of overgrowth of already existing microbiota and yeasts di ers from ff the pathogenesis of gut flora colonization.

The vast majority of previous research into pathophysiology, prevention and treatment of HAP comes from ICU environments and may therefore not be fully applicable or relevant to a general ward cohort setting (27). Only a few previous studies have investigated the incidence of HAP and HAP- associated risk factors outside an ICU setting (28– 30). Whereas these studies analysed risk factors associated with HAP, our study aimed at identify- ing potential pre-stages to HAP in a relatively large, non-selected, consecutive cohort of ward patients, which has not been done before.

Our findings reinforce the need for vigilance in the care of patients who have risk factors associ- ated with the development of a disturbed oropha- ryngeal microbiota. Patients admitted with ongoing antibiotic treatment and/or PPI medication may well benefit from more aggressive physiotherapy aimed at maximizing lung aeration and minimizing aspiration; cough training, basic lung expansion therapy and an upright position are all relatively easily accomplished prophylactic measures. Careful consideration of whether there is a real or a relative indication for PPI treatment is also called for, as is stringent and optimized antibiotic stewardship to avoid unnecessary and potentially harmful antibi- otics. Administration of probiotics may also pro- mote and/or restore a normal microbiota (31–34), but its potential needs further exploration.

Our study has limitations. Investigation of a lar- ger cohort for a longer period would have been preferable, enabling us to also evaluate whether a subset of our cohort developed clinical symptoms associated with, for example, pneumonia. Modern Swedish healthcare, however, is characterized by very early hospital discharges, and only a small proportion of our cohort was hospitalized more than a week, which reduced the possibility of iden- tifying long-term alterations in the oropharyngeal microbiota and any late clinical deterioration. Fur- thermore, classic cultivation techniques have the disadvantage of not disclosing slow-growing anaer- obes or non-cultivable bacterial species. Hence, we cannot know if such species were present in our cohort. The rationale for using accredited classic cultivation techniques is that we have striven to align with current clinical practice, where OPS sam- pling routinely provides the treating clinician with guidance in choosing the right treatment for the patient when an emerging airway infection is sus- pected. While molecular genomic techniques might have enhanced our ability to detect and identify specific pathogens in our samples (35, 36), these techniques introduce other sources of errors and are generally more suitable when searching for specific culprit pathogenic bacteria or when

8 © 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology

TRANBERG et al.

analysing samples that ideally should not host a microbiota, for example, cerebral spinal fluid. There are, however, some well-performed recent studies where deep sequencing of the amplified 16S rRNA gene has been used to show oropharyngeal micro- biota changes associated with PPI medication and antibiotic therapy (21, 24, 25). Their results are in line with our results, but it needs to be emphasized that the study populations di er, since thoseff studies were performed on out-of-hospital patients with chronic disease undergoing long-term PPI/antibiotic treatment.

Based on the results of this clinical observational study, we conclude that the oropharyngeal micro- biota seems to undergo changes during hospitaliza- tion. We note that risk factors for disturbance of the oropharyngeal microbiota and for the develop- ment of HAP include antibiotic exposure, length of hospital stay and the use of PPI medication. Con- sidering the insights gained in this investigation, it would be interesting to pursue this research area by determining whether di erent types of interventionsff can attenuate the aforementioned processes. Inter- ventional ‘care bundles’ could potentially include intentional discontinuation of PPI treatment, vigor- ous physiotherapy, vigilant antibiotic stewardship and probiotic treatment.

ACKNOWLEDGEMENTS

Thank you Maria Celander, Department of Clinical Microbiology in Skane University Hospital, for valuable help regarding interpretation and understanding of the results and methodology.

CONFLICTS OF INTEREST

The authors declare that there are no conflicts of interest.

FUNDING

This work was funded by research grants from Swedish Region Skane, with no specific grant from any other funding agency.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

REFERENCES

1. Dewhirst FE, Chen T, Izard J, Paster BJ, Tanner ACR, Yu W, et al. The human oral microbiome. J Bacteriol 2010;192:5002–17.

2. Jia G, Zhi A, Lai PFH, Wang G, Xia Y, Xiong Z, et al. The oral microbiota – a mechanistic role for sys- temic diseases. B Dent J 2018;224:447–55.

3. Johanson WG, Pierce AK, Sanford JP. Changing pha- ryngeal bacterial flora of hospitalized patients: emer- gence of gram-negative bacilli. N Engl J Med 1969;281:1137–40.

4. Robinson J. Colonization and infection of the respira- tory tract: what do we know? Paediatr Child Health 2004;9:21.

5. Hu nagle GB, Dickson RP, Lukacs NW. The respira-ff tory tract microbiome and lung inflammation: a two- way street. Mucosal Immunol 2017;10:299–306.

6. Frandah W, Colmer-Hamood J, Mojazi Amiri H, Raj R, Nugent K. Oropharyngeal flora in patients admit- ted to the medical intensive care unit: clinical factors and acid suppressive therapy. J Med Microbiol 2013;62(Pt 5):778–84.

7. Tranberg A, Thorarinsdottir HR, Holmberg A, Schott U, Klarin B. Proton pump inhibitor medication is associated with colonisation of gut flora in the oropharynx. Acta Anaesthesiol Scand 2018;62:791– 800.

8. Freedberg DE, Toussaint NC, Chen SP, Ratner AJ, Whittier S, Wang TC, et al. Proton pump inhibitors alter specific taxa in the human gastrointestinal microbiome: a crossover trial. Gastroenterology 2015;149:883–85.e9.

9. Imhann F, Bonder MJ, Vich Vila A, Fu J, Mujagic Z, Vork L, et al. Proton pump inhibitors a ect the ff gut microbiome. Gut 2016;65:740–48.

10. Herzig SJ, Howell MD, Ngo LH, Marcantonio ER. Acid-suppressive medication use and the risk for hos- pital-acquired pneumonia. JAMA 2009;301:2120–28.

11. Bateman BT, Bykov K, Choudhry NK, Schneeweiss S, Gagne JJ, Polinski JM, et al. Type of stress ulcer prophy- laxis and risk of nosocomial pneumonia in cardiac surgi- cal patients: cohort study. BMJ 2013;347:f5416.

12. Eom CS, Jeon CY, Lim JW, Cho EG, Park SM, Lee KS. Use of acid-suppressive drugs and risk of pneu- monia: a systematic review and meta-analysis. CMAJ 2011;183:310–9.

13. MacLaren R, Reynolds PM, Allen RR. Histamine-2 receptor antagonists vs proton pump inhibitors on gastrointestinal tract hemorrhage and infectious com- plications in the intensive care unit. JAMA Intern Med 2014;174:564–74.

14. Alshamsi F, Belley-Cote E, Cook D, Almenawer SA, Alqahtani Z, Perri D, et al. E cacy and safety of ffi pro- ton pump inhibitors for stress ulcer prophylaxis in critically ill patients: a systematic review and meta- analysis of randomized trials. Crit Care 2016;20:120.

15. Flanders SA, Collard HR, Saint S. Nosocomial pneu-monia: state of the science. Am J Infect Control 2006;34:84–93.

16. Nair GB, Niederman MS. Nosocomial pneumonia: lessons learned. Crit Care Clin 2013;29:521–46.

17. C akir Edis E, HatipoVglu ON, Yılmam I_, Su€t N. Eco-nomic burden of nosocomial pneumonia in non-inten- sive care clinics. Tuberk Toraks 2015;63:8–12.

© 2020 The Authors. APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Medical Microbiology and Pathology 9

OROPHARYNGEAL MICROBIOTA AND HOSPITALISATION

18. Sullivan A, Edlund C, Svenungsson B, Emtestam L, Nord CE. E ect of perorally administered pivmecilli-ff nam on the normal oropharyngeal, intestinal and skin microflora. J Chemother 2001;13:299–308.

19. Retchless AC, Kretz CB, Rodriguez-Rivera LD, Chen A, Soeters HM, Whaley MJ, et al. Oropharyngeal microbiome of a college population following a meningococcal disease outbreak. Sci Rep 2020;10:1–9.

20. Human pharyngeal microbiome may play a protective role in respiratory tract Infections. Elsevier Enhanced Reader. Available at: https://reader.elsevier.com/reade r/sd/pii/S1672022914000485?token=77716CA9F33E4 83FF561FBFBCEE480B139B61AB962CEC4A3C9C B54E926A48A318B20A440E5BB42401E26 F1A303716B42. Accessed 20 Jul 2020.

21. Rosen R, Hu L, Amirault J, Khatwa U, Ward DV, Onderdonk A. 16S community profiling identifies pro- ton pump inhibitor related di erences in gastric, ff lung, and oropharyngeal microflora. J Pediatr 2015;166:917–23.

22. Zaura E, Brandt BW, Teixeira de Mattos MJ, Buijs MJ, Caspers MPM, Rashid M-U, et al. Same expo- sure but two radically di erent responses toff antibi- otics: resilience of the salivary microbiome versus long-term microbial shifts in feces. MBio 2015;6: e01693–15.

23. Rashid M, Weintraub A, Nord CE. E ect of ff new antimicrobial agents on the ecological balance of human microflora. Anaerobe 2012;18:249–53.

24. Choo JM, Abell GCJ, Thomson R, Morgan L, Waterer G, Gordon DL, et al. Impact of long-term erythromycin therapy on the oropharyngeal micro- biome and resistance gene reservoir in non-cystic fibrosis bronchiectasis. mSphere 2018;3:e00103.

25. Dos Santos LOPES, Santiago GUIDO, Brusselle G, Dauwe K, Deschaght P, Verhofstede C, et al. Influ- ence of chronic azithromycin treatment on the compo- sition of the oropharyngeal microbial community in patients with severe asthma. BMC Microbiol 2017; 17:109.

26. Edlund C, Nord CE. Manipulation of the oropharyn- geal and intestinal microflora by norfloxacin: microbi- ological and clinical aspects. Scand J Infect Dis Suppl 1988;56:14–21.

27. Di Pasquale M, Aliberti S, Mantero M, Bianchini S, Blasi F. Non-intensive care unit acquired pneumonia: a new clinical entity? Int J Mol Sci 2016;17(3):287.

28. Sopena N, Heras E, Casas I, Bechini J, Guasch I, Pedro- Botet ML, et al. Risk factors for hospital-acquired pneu- monia outside the intensive care unit: a case-control study. Am J Infect Control 2014;42:38–42.

29. Ewan V, Hellyer T, Newton J, Simpson J. New hori- zons in hospital acquired pneumonia in older people. Age Ageing 2017;46:352–58.

30. Thompson DA, Makary MA, Dorman T, Pronovost PJ. Clinical and economic outcomes of hospital acquired pneumonia in intra-abdominal surgery patients. Ann Surg 2006;243(4):547–52.

31. de Vrese M, Schrezenmeir J. Probiotics, prebiotics, and synbiotics. Adv Biochem Eng Biotechnol 2008;111:1–66.

32. Fooks LJ, Gibson GR. Probiotics as modulators of the gut flora. Br J Nutr 2002;88(Suppl 1):39.

33. Klarin B, Wullt M, Palmquist I, Molin G, Larsson A, Jeppsson B. Lactobacillus plantarum 299v reduces colonisation of Clostridium di cile in critically ffi ill patients treated with antibiotics. Acta Anaesthesiol Scand 2008;52:1096–102.

34. Kujawa-Szewieczek A, Adamczak M, Kwiecien, K, Dudzicz S, Gazda M, Wiezcek A. The e ectff of Lacto- bacillus plantarum 299v on the incidence of Clostrid-ium di cile infection in high risk patients treated withffi antibiotics. Nutrients 2015;7(12):10179–88.

35. Sontakke S, Cadenas MB, Maggi RG, Diniz PP, Bre- itschwerdt EB. Use of broad range16S rDNA PCR in clinical microbiology. J Microbiol Methods 2009;76:217–25.

36. Harris KA, Hartley JC. Development of broad-range 16S rDNA PCR for use in the routine diagnostic clin- ical microbiology service. J Med Microbiol 2003;52: 685–91.

Paper III

ORIG INAL AR T I C L E

Efficacy of Lactiplantibacillus plantarum 299 and 299v against nosocomial oropharyngeal pathogens in vitro and as an oral prophylactic treatment in a randomized, controlled clinical trial

Anna Tranberg1 | Bengt Klarin1 | Julia Johansson1 | Lisa I. Påhlman2,3

1Division of Intensive and Perioperative Care, Skåne University Hospital Lund, Lund, Sweden2Division of Infectious Diseases, Skåne University Hospital Lund, Lund, Sweden3Wallenberg Centre for MolecularMedicine, Lund University, Lund, Sweden

CorrespondenceAnna Tranberg Lindqvist, Division of Intensive and Perioperative Care, Skåne University Hospital, Getingevägen 4, SE- 22185 Lund, Sweden.Email: anna.tranberg_lindqvist@med.lu.se

Funding informationRegion Skåne, Grant/Award Number: Doktorand-2019-0144 and Doktorand-2020-0459

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.© 2020 The Authors. MicrobiologyOpen published by John Wiley & Sons Ltd.

MicrobiologyOpen. 2020;00:e1151. https://doi.org/10.1002/mbo3.1151

www.MicrobiologyOpen.com | 1 of 12

Received: 23 September 2020 | Revised: 1 December 2020 | Accepted: 7 December 2020 D0I: 10.1002/mbo3.1151

AbstractBackground: Disturbance in the oropharyngeal microbiota is common in hospi- talized patients and contributes to the development of nosocomial pneumonia. Lactiplantibacillus plantarum 299 and 299v (Lp299 and Lp299v) are probiotic bacteria with beneficial effects on the human microbiome.Aim: To investigate how Lp299 and Lp299v affect the growth of nosocomial oro- pharyngeal pathogens in vitro and to evaluate the efficacy in vivo when these probi- otics are administered prophylactically in hospitalized patients.Methods: The in vitro effect of Lp299 and Lp299v on nosocomial respiratory tract pathogens was evaluated using two methods, the co-culture and agar overlay. In the clinical study, patients were randomized to orally receive either probiotics or placebo twice daily during their hospital stay. Oropharyngeal swabs were analyzed at inclusion and every fourth day throughout hospitalization.Findings: All tested pathogens were completely inhibited by both Lp299 and Lp299v using the agar-overlay method. In the co-culture experiment, Lp299 and Lp299v sig- nificantly (p < 0.05) reduced the growth of all pathogens except for Enterococcus faeca- lis co-incubated with Lp299. In the clinical study, daily oral treatment with Lp299 and Lp299v did not influence the development of disturbed oropharyngeal microbiota or nosocomial infection. Proton pump inhibitors, antibiotics, and steroid treatment were identified as risk factors for developing disturbed oropharyngeal microbiota.Conclusions: Lp299 and Lp299v inhibited pathogen growth in vitro but did not affect the oropharyngeal microbiota in vivo. The ClinicalTrials.gov Identifier for this study is NCT02303301.

1 | INTRODUC TION

Probiotics are defined by the International Scientific Association for Probiotics and Prebiotics (ISAPP) as live microorganisms that, when administered in adequate amounts, confer a health benefit on the host (Hill et al., 2014). Different probiotics display a vari- ety of antimicrobial properties, for example, the production of ammonia, lactic acid, free fatty chains, hydrogen peroxide, and bacteriocins (de Vrese & Schrezenmeir, 2008). Lactiplantibacillus plantarum 299 and 299v (Lp299 and Lp299v) are probiotic bac- teria within the lactic acid bacteria group. L. plantarum can se- crete bacteriocins, also called plantaricins (Adebayo et al., 2014; Prabhurajeshwar & Chandrakanth, 2017; Seddik et al., 2017), with inhibitory effects on, for example, oral Streptococcus mutans (Hasslöf et al., 2010). A recent review (Simons et al., 2020) pres- ents the possible role of bacteriocins as a part of future antibiotic treatment.

A disturbance in the microbiome of the oropharynx, defined as an overgrowth of normally existing species and/or establishment of new potential pathogens, has been shown to indicate the degree of sickness in the host, and to be associated with increased mortal- ity in the intensive care unit (ICU) and non-ICU patients (Dickson et al., 2020; Johanson et al., 1969). The microbiome of the orophar- ynx and that of the lower respiratory tract resemble each other, probably due to the microaspiration of oropharyngeal microbiota (Bassis et al., 2015). Microaspiration of disturbed oropharyngeal microbiota has been shown to play a part in the complex patho- genesis behind the development of pneumonia (Bahrani-Mougeot et al., 2007; Huffnagle et al., 2016) Therefore, a large number of studies have investigated the effects of decontamination of the oropharynx (using chlorhexidine or local antibiotics) or of adminis- tration of probiotics, with varying results (Bo et al., 2020; Gu et al., 2012; Karacaer et al., 2017; Klarin et al., 2008; Morrow et al., 2010; Wang et al., 2013; Weng et al., 2017). These studies have been car- ried out in ICU settings or pediatric populations, diminishing the occurrence of respiratory disease as well as antibiotic-associated diarrhea (Hatakka et al., 2001; Ling et al., 2019; Niveen et al., 2016). Because of the heterogeneity in the populations studied, general- ized conclusions and recommendations about probiotic benefits are difficult to present.

In this study, we investigated whether Lp299 and Lp299v can reduce or inhibit the growth of nosocomial pathogens in vitro and in vivo. Seven pathogens were selected for the in vitro study, due to their frequent appearance in oropharyngeal cultures in ICU and non-ICU patients according to previous studies (Klarin et al., 2018; Tranberg et al., 2018). The randomized controlled trial aimed to study whether oral administration with lactobacilli could prevent or delay the occurrence of disturbed oropharyngeal microbiota in non- ICU hospitalized patients.

2 | MATERIAL S AND METHODS

2.1 | In vitro study

2.1.1 I Bacterial strains

Lp299 and Lp299v were provided by Probi AB, Lund, Sweden. Reference strains of the pathogens Escherichia coli

(CCUG 24), Staphylococcus aureus (CCUG 1800), Enterococcus

faecalis (CCUG 19916) and E. faecium (CCUG 542), Klebsiella

pneumoniae (CCUG 225), Pseudomonas aeruginosa (CCUG 551) and Enterobacter cloacae (CCUG 6323) were purchased from Culture Collection, University of Gothenburg, Sweden. Clinical isolates came from the Department of Clinical Microbiology at Skåne University Hospital, Sweden.

2.1.2 I Growth conditions

Lp299 and Lp299v were grown in De Man-Rogosa-Sharpe (MRS) broth (Merck, Darmstadt, Germany) and on MRS agar. E. faecalis,E. faecium, and S. aureus were cultured in Todd Hewitt (TH) broth (Becton Dickinson) and agar, whereas E. coli, K. pneumoniae, P.

aer- uginosa, and E. cloacae were grown in lysogeny broth (LB) (Sigma- Aldrich, St. Louis, M0, USA) and agar. All strains were cultured at 37°C under aerobic conditions (21% oxygen, 5% C02).

2.1.3 I Agar overlay

0vernight cultures of Lp299 and Lp299v were washed in Phosphate- buffered saline (PBS) and adjusted to final concentrations of approx- imately 2 × 109 colony-forming units (CFU)/ml. Varying amounts (4 x 104, 4 x 105, and 4 x 106 CFU, respectively) of Lp299 or Lp299v were added to 8 ml of warm (42-45°C) MRS agar and poured into Petri dishes. Control plates contained no lactobacilli. After solidifica- tion, this bottom agar was incubated at 37°C overnight. The second layer of agar (24 ml), suited for the pathogen, was then cast on top of the MRS agar. Overnight cultures of the pathogens were diluted 1:1000, 1:10,000, and 1:100,000 in PBS, and 10 μl drops of the di- lutions were seeded on the top agar. After overnight incubation at 37°C, the growth of the pathogen was assessed. Experiments were repeated twice using reference strains and once with clinical isolates of the pathogen.

2.1.4 I Inhibitory activity of Lp299 and Lp229v

Co-cultures of lactobacilli and pathogens were grown in a mixed broth consisting of 25% (v/v) MRS and 75% (v/v) TH or LB broth.

TRANBERG ET AL.

2 ot 12

|

These proportions provide good growth conditions for both lac- tobacilli and pathogens. 0vernight cultures of Lp299, Lp299v, and pathogen strains were washed and adjusted to bacterial suspensions of 2 x 109 CFU/ml, and 50 μl of the pathogen and 500 μl of Lp229 or Lp299v were added to 10 ml of mixed broth. As a control, 50 μl of the pathogen suspension was grown in mixed broth in the absence of lactobacilli. The co-cultures were incubated for 5 hours at 37°C. Before and after incubation, a small aliquot of each sample was di- luted in PBS and plated on 15-20 ml agar suitable for the pathogen. After incubation overnight at 37°C, colonies of the pathogens were counted, and the growth of lactobacilli was ensured. Experiments were performed in triplicate.

2.1.5 I Antibacterial activity of Lp299 and Lp299v

supernatants

0vernight cultures of Lp229 or Lp299v were pelleted by cen- trifugation, and the supernatants were sterile filtered through a0.22 μm Millex®- GP, Millipore Express® PES Membrane Filter(Merck Millipore Ltd). The cell-free supernatants were then either pH-neutralized with 1 M Na0H to a pH of 5.4 (corresponding to the natural pH of MRS broth); heat-treated at 99°C for 5 minutes, or incubated with pepsin (Sigma-Aldrich), proteinase K (Thermo Scientific), or trypsin (Sigma-Aldrich) at a final concentration of 1 mg/ml for 2 hours at 37°C. After that, the samples were heated to 99°C for 5 minutes to eliminate the protease activity. The inhib- itory effect of the supernatants was tested against E. cloacae andS. aureus. These two were chosen as they were inhibited by Lp299 and 299v, and they differ in Gram staining and natural habitat.2.5 ml of pH-neutralized, heat-treated, or protease-treated super- natant was added to 7.5 ml of TH or LB to obtain a mixed broth. Untreated supernatant, sterile MRS broth, and MRS adjusted to pH 4.2 with acetic acid were included for comparison. Pathogens were washed and diluted as described in the co-culture experi-ment, and 50 μl of pathogen solution was added to the mixedbroths. As a control, pathogens were incubated with Lp299 or Lp299v in a mixed broth with sterile MRS. Samples were plated on agar before and after incubation for 5 hours as described above. After incubation overnight at 37°C, colonies of the pathogens were counted, and the growth of lactobacilli was ensured when relevant. Experiments were performed in triplicate.

2.2 | Randomized controlled trial

2.2.1 I Study population

Patients were enrolled in the study between 2014 and

2019 at the University Hospital in Lund using the following inclusion criteria: age�18 years, obtaining the first oropharyngeal swabs (0PS) within 24 hours of hospital admission, and an expected length of stay of more than 72 hours. Exclusion criteria were respiratory infection and prior

| 3 ot TRANBERG ET AL.

hospitalization within two weeks. The patients were enrolled by the investigators, research nurses, or medical students. The patients were admitted to medical, surgical, or orthopedic wards. A stand- ardized case report form was used to record patient data.

We based our approximative power calculation on a study from 1969, where oropharyngeal cultures had been analyzed throughout the hospitalization in ward patients (Johanson et al., 1969). From that study, we estimated that a sample size of 75 patients in each group would be sufficient to show a significant difference between the groups in the occurrence of disturbed microbiota in the oropharynx. The patients were younger but much sicker in the study from 1969, and patients that would now be admitted to the ICU were treated in gen- eral wards. The study from 1969 was observational, and the above fac- tors put together made it difficult to make an exact power calculation.

2.2.2 I Randomization

The randomization was performed directly after inclusion via sealed envelopes at a 1:1 ratio. The randomization was blinded to recruit- ers, staff, and patients.

2.2.3 I Intervention

Patients received either a combination of 1010 CFU Lp299 and 1010 CFU Lp299v with 3 grams of maltodextrin or a placebo con- sisting of only 3 grams of maltodextrin. Both lactobacilli and placebo were manufactured and generously provided by Probi AB, Lund, Sweden, and delivered in identical freeze-dried sachets labeled "A" and "B," respectively. The sachets were kept in a -80°C freezer until use. In the ward, the sachets were kept at 4°C for a maximum of five days. Viability controls of the lactobacilli in the sachets were per- formed yearly throughout the study period, analyzing sachets stored at both -80°C and 4°C. Before administration to the patient, the contents of the sachets were resuspended in 15 ml of sterile water, allowing the revival of the potential lactobacilli for 20-40 minutes before administering the mixture to the patient. Patients received the assigned mixture twice daily throughout the hospital stay, with instructions to gargle the mixture as long as possible and then swal- low. 0PSs were taken at inclusion (day 1), on day 3, and after that approximately every fourth day. In all other respects, the patients received standard care.

2.2.4 I Microbiological procedures and definitions

The 0PSs were processed by extended microbiological proce- dures at the Department of Clinical Microbiology, Skane University Hospital in Lund. The laboratory is accredited by the accreditation body (SWEDAC) designated by the Swedish government and is for- mally recognized as competent according to European and interna- tional standards.

For bacteria cultivation, sampling media were inoculated on five types of agar plates (three selective, one differentiating, and one nonselective). All plates were produced in-house, sometimes using commercially available media components (5% horse blood, hematin agar, and UriSelect 4 agar), as listed below:

1. Agar with 5% horse blood (LabM, Heywood) supplemented with 10 mg/L colistin and 15 mg/L nalidixic acid with an optochin disk (selective);

2. Agar with 5% horse blood supplemented with 2 mg/L gentamicin and 25 mg/L nalidixic acid for Gram-positive cocci including Streptococcus pneumoniae (selective);

3. Hematin agar (0xoid™, Thermo Science) supplemented with 300 mg/L bacitracin for fastidious Gram-negative rods including Haemophilus influenzae (selective)

4. UriSelect 4 agar (Bio-Rad Laboratories) supplemented with10 mg/L vancomycin for non-fastidious Gram-negative rods (differentiating)

5. Hematin agar with a colistin disk (nonselective).

The plates were inspected for growth after 16 and 40 hours of aerobic, anaerobic, or C02 incubation at 35-37°C. If an inspection result was ambiguous at 40 hours, the plate was incubated for an additional 24 hours to obtain a more definite result. Species iden- tification of bacteria was performed using matrix-assisted laser de- sorption/ionization time-of-flight (MALDI-T0F) mass spectrometry (MALDI Biotyper Microbial Identification System, Bruker), using software FlexControl 3.4 and MALDI Biotyper (MBT) Compass 4.1, with MBT Compass Library, DB-7854 MSP (Bruker, Bremen, Germany).

Cultivation and differentiation of Candida spp. were based on col- ony appearance on CHR0M Candida agar (CHR0Magar, Hägersten, Sweden) after 48 hours of incubation at 35°C.

For a sample to be considered representative of "oropharyngeal microbiota," bacterial species normally found in the oropharynx were required to grow on the nonselective hematin plate as deter- mined by visual inspection by an experienced senior microbiologist and following standard practice (Retchless et al., 2020). Samples were classified as disturbed oropharyngeal microbiota when there was a growth of species not normally found in the oral cavity and/ or overgrowth of normal oropharyngeal microbiota on selective and differentiating plates. Samples with disturbed oropharyngeal micro- biota were divided into three subclasses (see Figure 4): gut patho- gens, respiratory tract pathogens, and yeast.

2.2.5 I Statistical analyses

Inhibitory effects in vitro were analyzed using Student's paired t- test. In the randomized controlled trial, continuous variables were presented as median, minimum, and maximum values. Dichotomous variables were presented as numbers and as a percentage of the

total number. For subjects with a normal oropharyngeal microbiota

TRANBERG ET AL.

4 ot 12

|

at inclusion, a univariate Poisson regression was used to analyze the association between the patients’ characteristics (predicting variables) and the intervention they were randomized to (dependent variable). Thereafter, a multivariate Poisson regression model using the two strongest predicting variables from the univariate analysis was constructed, in which one additional potential explanatory vari- able was added to determine whether the model improved or did not improve by including a third variable. A Kaplan-Meyer analysis was performed to test for differences between the placebo and the lactobacilli group regarding "time to first disturbed oropharyngeal swab." Fisher's exact test was used to assess the relationship be- tween the intervention group and nosocomial infection rate.

Statistical analyses were performed using IBM SPSS Statistics 26 for Windows (IBM Corp., Armonk, NY, USA). 0dds ratios (0Rs) are presented with a 95% confidence interval. p < 0.05 was considered significant, and all statistical tests were two-tailed.

3 | RESULTS

3.1 | Lp299 and Lp299v inhibit the growth of

bacterial pathogens in vitro

The inhibitory effect of Lp299 and Lp299v on other bacteria was first tested in an agar-overlay assay, where varying concentrations of lactobacilli were grown in a bottom MRS agar, and the pathogens were seeded on a top agar. Under these conditions, all experiments showed a complete absence of pathogen growth compared to con- trol plates without lactobacilli. Both clinical isolates and the cor- responding reference strains were tested and gave the same clear results. See Table A1.

Next, pathogens and lactobacilli were co-cultured in broth to fur- ther characterize the inhibitory effect. In this experimental set-up, both Lp299 and Lp299v significantly inhibited the growth of S. au- reus, E. cloacae, K. pneumoniae, E.

coli, E. faecium, and P. aeruginosa. For two pathogens, E. cloacae and K. pneumoniae, almost complete eradication of the pathogens was seen, as the number of colonies was close to zero after incubation. E. faecalis was significantly inhib- ited by Lp299v, but in co-culture, with Lp299 the inhibition did not reach statistical significance (p = 0.12). See Figure 1.

3.2 | The antibacterial activity of Lp299 AND Lp299v IS pH-dependent

To investigate the mechanism behind the

growth-inhibitory ef- fects of lactobacilli, S. aureus was incubated in a mixed broth con- taining cell-free supernatants from overnight cultures of Lp299 and Lp299v. The pH of the supernatants was 4.1 and 4.0, respec- tively. Figure 2 shows that the supernatants significantly inhibited the growth of S. aureus to the same extent as co-incubation with live bacteria. When the pH of the supernatants was elevated to 5.4, corresponding to the pH of MRS broth, the inhibitory effect

FI G U R E 1 Co-culture of different pathogens with Lp299 or Lp299v resulted in significant growth inhibition for all pathogens exceptEnterococcus faecalis co-incubated with Lp299. The growth of the pathogen alone in the absence of lactobacilli served as control. *p < 0.05

FI G U R E 2 The inhibitory effect of Lp299 and Lp299v on S. aureus is pH-dependent. S. aureus was incubated with Lp299, Lp299v, orcell-free supernatants from overnight cultures of the lactobacilli. To explore the role of pH, S. aureus was also incubated with pH-neutralized supernatants and with acidified MRS broth. The bars with an asterisk above them indicate a significant growth inhibition compared to the control with S. aureus grown in a mixed broth without lactobacilli. SN =supernatant. Neu =neutralized. *p < 0.05

was abolished, and S. aureus grew equally well as in the control. Further, MRS broth made acidic to pH 4.0 (same pH as the super- natants) significantly inhibited S. aureus growth to the same extent as the supernatants from both Lp299 and Lp299v (see Figure 2). The same results were obtained when the experiment was per- formed with E.

cloacae (see Figure A1).To examine the possible role of plantaricins secreted by

Lp299 and Lp299v, E. Cloacae was incubated with Lp299 and Lp299v su- pernatants that had been heat-treated to denature the protein content, or pre-incubated with the proteinases pepsin, proteinase K or trypsin to digest proteins in the supernatants. All of the pre- treated supernatants showed the same clear growth inhibition as untreated supernatants, whereas the controls incubated with MRS

broth showed expected growth of the pathogen during the incuba- tion time (see Figure A2).

Taken together, the overall conclusion of these experiments is that the inhibitory effect was mainly pH-dependent.

3.3 | Randomized controlled trial

Between the 18th of September 2014 to 1st of May 2019, 135 pa- tients were included and randomized. Eighteen patients were ex- cluded due to non-adherence to protocol. Thus, 117 patients met all inclusion criteria and contributed a total of 337 0PSs. (See Figure 3 for the C0NS0RT diagram). The median number of 0PSs per patient

| s ot TRANBERG ET AL.

TA B L E 1 Descriptive statistics of patients (n = 117)

Age, years 76 (22 96) 76 (36 97) 0.926b

Body mass index 26 (18 40) 26 (17 40) 0.311b

Diabetes 15 (26%) 8 (14%) 0,108a

Both univariate and multivariate analyses showed that treatment with proton pump inhibitor (PPI), cortisone, or antibiotics during hospitalization could be associated with an added risk of developing disturbed microbiota during hospitalization (Table 2).

Kaplan-Meyer analyses were performed to determine whether treatment with Lp299 and Lp299v could delay the development of disturbed microbiota. There was a slight tendency to later develop- ment of disturbed microbiota in the treatment group, but the differ-

Alcohol intake>2 times/week

Proton pump

19 (33%)

12 (21%)

21 (37%)

19 (32%)

1.000a

0.345a

ence was not significant.Concerning the risk of developing a nosocomial

infection during hospitalization, the difference between the two groups did not reach

inhibitor significance. In the treatment group, 4/58 patients (7%) developed a

Able to walk two 33 (57%) 33 (57%) 0.490a nosocomial infection, while the incidence in the placebo group was

flights of stairs 10/59 (17%, p = 0.153), with no obvious difference in the severity of in-

Cortisone medication

3 (5%) 6 (10%) 0.490a fection between the groups (Table 1). The causes of nosocomial infec-

Unplanned admission

45 (78%) 42 (71%) 0.526a tion were urinary tract infections, wound infections, and pneumonia.

Antibiotics >24 2 (3%) 1 (1.7%) 0.619a

hours beforehospitalization 4 | DISCUSSION

Prophylactic antibiotics perioperatively

Length of hospital stay (days)

36 (62%) 31 (53%) 0.352a

8 (3 32) 7 (3 32) 0.260c

In this combined laboratory and clinical study, we found that Lp299 and Lp299v significantly inhibited in vitro growth of nosocomial pathogens commonly found in the oropharyngeal tract of hospi- talized patients. In the randomized controlled trial, no difference between the intervention group and the placebo group could be found regarding changes in the oropharyngeal microbiota or the oc- currence of nosocomial infections. The study confirmed the already known risk factors for the development of disturbed oropharyngeal microbiota (Frandah et al., 2013).

Data are presented as median (range) or number (percentage).aFisher's exact test, exact sig. (2-sided). bIndependent samples t-test, sig. (2-tailed). cMann-Whitney U test, exact sig. (2-tailed).

in both groups was 3. All 0PSs were representative of oropharyn- geal microbiota. The baseline patient characteristics and hospitali- zation characteristics of the placebo and lactobacilli groups were similar and are presented in Table 1.

Figure 4 presents the microbiological results for the 27 pa- tients showing any type of disturbed oropharyngeal microbiota on any sampling occasion. Each horizontal row corresponds to a patient's observation time, and by following the row from left to right it is possible to see 0PS changes over time. Using the color and species key, the figure shows the subclass and pathogen for each disturbed 0PS. The upper part of the figure shows the pa- tients receiving placebo and the lower part shows the patients receiving lactobacilli.

In 104 patients (89%), the first 0PS at admission was normal. We analyzed results from these patients using univariate and multivari- ate Poisson analyses. The

univariate analyses showed that treatment with lactobacilli yielded an RR of 0.96 (CI 0.36-2.55, p = 0.94) for ac- quiring disturbed oropharyngeal microbiota during hospitalization.

TRANBERG ET AL.

6 ot 12

|

pPlacebo(n =

Lactobacilli(n = 58)

Variable

Gender, male 28 (48%) 26 (44%) 0.712a

Current or ex-smoker 23 (40%) 24 (41%) 1.000a

20 (34%) 0.309a14(24%)

Antibiotics >24 hours during hospitalization

0.236c

3 (2 6)3 (2 8)Oropharyngeal swabs

Nosocomial infection 4 (6.9%) 10 (17%) 0.153a

The ability of lactobacilli to inhibit pathogen growth has been shown before (Annuk et al., 2003). This has also been specifically shown for Lp299v when Hutt et al. demonstrated its antagonistic ef- fect on Salmonella enterica and Helicobacter pylori (Hutt et al., 2006). Furthermore, a study on oral care in ICU patients showed that Lp299 could be identified in the oropharynx in all patients given the study product (Klarin et al., 2008), indicating that the lactobacilli remain in the oropharynx after oral administration. In this study, we show for the first time that Lp299 and Lp299v inhibit in vitro growth of pathogens known to cause nosocomial respiratory tract infections. Notably, 60% of ICU patients are colonized with at least one of the seven investigated pathogens as early as 24 hours after admission to the ICU (Tranberg et al., 2018). The L. plantarum species have a high production of lactic acid compared to others in the lactic acid bacteria group, and a relatively small production of, for example, hy- drogen peroxide and carbon dioxide, which is typical for this group of facultatively heterofermentative lactobacilli (Annuk et al., 2003). In our study, the acidic environment produced by the lactobacilli was essential for inhibiting the in vitro growth of the pathogens under study. However, other factors may also be involved. For example, it has been shown that an acidic pH is necessary for other inhibitory mechanisms to be activated. Several studies on plantaricins show that they are activated at a pH <5 (Lin & Pan, 2019; Song et al., 2014). Although we were unable to demonstrate a plantaricin effect in our

Allocated to lactobacilli (n = 70) Received allocated intervention (n = 59) Did not receive allocated intervention (did not

receive intervention by staff) (n = 11)

Allocation Allocated to placebo (n = 65)

Received allocated intervention (n = 59) Did not receive allocated intervention by staff

(n = 6)

Lost to follow-up (n = 0)

Discontinued intervention (did not like taste of mixture) (n = 1)

Follow-Up Lost to follow-up (n = 0)

Discontinued intervention (n = 0)

Analysed (n = 58) Excluded from analysis (n = 0)

Analysis Analysed (n = 59)

Excluded from analysis (n = 0)

FI G U R E 3 The C0NS0RT flow diagram. C0NS0RT (Consolidated Standards of Reporting Trials) diagram demonstrating the progressthrough the phases of the randomized trial of two groups

study, we cannot rule out a possible role of plantaricins in our strains due to the overwhelming effect of acidic pH. Further experiments are required to determine the possible presence and requisites for the activity of bacteriocins in Lp299 and Lp299v.

In the clinical trial, we could not show that the oral administra- tion of Lp299 and Lp299v prevents or delays the occurrence of dis- turbed oropharyngeal microbiota in non-ICU hospitalized patients compared to placebo. In agreement with earlier findings (Tranberg et al., 2018), we found that 11% of the patients had disturbed oro- pharyngeal microbiota at admission and that an increasing propor- tion of the patients developed disturbed oropharyngeal microbiota during their hospitalization (see Figure 3). We also confirmed previ- ously reported findings that treatment with PPI and antibiotics were risk factors for disturbed oropharyngeal microbiota (Frandah et al., 2013; Tranberg et al., 2018). In this study, oral cortisone was strongly

associated with disturbed oropharyngeal microbiota (p = 0.0025 in the multivariate Poisson regression analyses shown in Table 2, and p = 0.0026 in the univariate Poisson regression), which has not been shown before. The contribution of steroid inhalation treatment to the risk of developing oral candidiasis is well known. In our study, none of the patients who were on cortisone treatment developed disturbed microbiota consisting of Candida species, and only the third was simultaneously on PPI medication.

This study is unique, as it focuses on ward patients. Most previ- ous studies on disturbed oropharyngeal microbiota and its possible contribution to nosocomial pneumonia have focused on intensive care patients. The idea of giving hospitalized patients probiotics is tempting, in many ways. Probiotics are harmless, inexpensive, and may reduce antibiotic use by restoring the patient's microbi- ota toward being healthier and more normal. When swallowed and

| 7 ot TRANBERG ET AL.

Enrollment

Excluded (n = 27) Not meeting inclusion criteria (n = 2) Declined to participate (n = 25) Other reasons (n = 0)

Assessed for eligibility (n = 162)

Randomized (n = 135)

Sampling day number Day 1 Day 3-4 Day 5-8 Day 9-12 Day 13-16 Day 17-20 Day 21-24Number of samples analyzed 117 113 64 24 14 4 1Percentage with disturbed microbiota 11 14 11 17 7 100 0

Color and Species KeyRespiratory Tract pathogensHaemophilus influenzae HiStaphylococcus aureus

Sa Streptococcus beta group G StBGGut pathogensCitrobacter freundii CfEnterobacteraerogenes EaEscherichia coli EcEnterobacter cloacae EclEnterococcus faecium EfEnterococcus faecalis EfsKlebsiellauntyped KKlebsiellaoxytoca KoKlebsiella pneumoniae KpMorganella morganii MmPseudomonas aeruginosa PaProteus mirabilis PmProteus vulgaris PvSerratia marcescens SmYeastCandida albicans CaCandida glabrata CgCandida tropicalis Ct

Normal oropharyngeal microbiota

FI G U R E 4 0ropharyngeal swab (0PS) culture results for the 27 patients who had at least one 0PS sample with disturbed microbiota during their hospitalization. Each horizontal bar represents the patient's observation time, and the colored bars indicate an 0PS culture result for each sampling time: yellow = normal, blue = respiratory pathogens, terracotta = gut microbiota, gray = yeast species. The top row describes the time frames within which the 0PS was obtained. The second and third row shows the number of 0PSs collected/analyzed and the percentage of 0PSs with disturbed microbiota for the total cohort at each sampling time point. The patients are divided according to whether they received probiotics or placebo during hospitalization

Univariate RR 95 CI) p

MultivariateRR 95 CI) p

Lactobacilli group 0.96 (0.36 2.55) 0.938Lactobacilli group 0.92 (0.33 2.51) 0.864Diabetes 1.41 (0.49 4.08) 0.528 1.43 (0.48 4.31) 0.523Lactobacilli 1.10 (0.41

2.96)a0.849a

PPI 2.85 (1.10 7.39) 0.031 2.89 (1.08 7.76)a

0.035a

Lactobacilli 1.13 (0.44 2.90)

0.805

Cortisone 4.22 (1.65 10.8) 0.0026 4.32 (1.68 11.1) 0.0025Lactobacilli 0.94 (0.36

2.46)0.893

Antibiotics before

hospitalization

2.59 (0.48 13.9) 0.267 2.62 (0.48 14.2) 0.264

Lactobacilli 1.17 (0.44 3.13)

0.753

Antibiotics during

hospitalization

3.14 (1.19 8.30) 0.021 3.23 (1.18 8.84) 0.023

Lactobacilli 0.97 (0.37 2.55) 0.952Unplanned admission

0.83 (0.29 2.43) 0.739 0.83 (0.29 2.41) 0.738

athe maximum number of step-halvings was reached but the log-likelihood value cannot be further improved. 0utput for the last iteration is displayed. RR =risk ratio, PPI =proton pump inhibitor.

TA B L E 2 Poisson regression analysis for developing disturbed oropharyngeal microbiota during hospitalization in 104 subjects with a normal microbiota at admission, (yes n = 14, no n = 90). The number of patients receiving Lactobacilli in this analysis was 53, the control group consisted of 51 patients

TRANBERG ET AL.

a ot 12

|

Kp Ea Ec Hi

Hi K

Kp

SaHi

Pa Pv Efs

SaEc Ko

Hi Hi

Placebo given to patients below line Ca Kp Cf Ef

CaCf

Ec Probiotics given to patients above line StBG Sa Hi

Hi

EcEcMmSm PmPm

Hi Es Hi

Cg Ct CaCaCa Pm Hi Ecl

Cg

Ca

EcEc Hi

Pat

ient

s w

ith d

istu

rbed

mic

robi

ota

durin

g

digested, the probiotics also influence the intestinal tract immune system. New connections between the composition of the gut mi- crobiota and a wide range of diseases such as irritable bowel syn- drome and depression have emerged in the last decade (Didari et al., 2015; Wallace & Milev, 2017).

An important weakness of the clinical trial is that our power calculation was based on older studies with longer hospital stays (Johanson et al., 1969). Consequently, the study was underpow- ered, and we can therefore unfairly

rule out our hypothesis that treatment with lactobacilli can decrease or delay the incidence of disturbed

oropharyngeal microbiota during hospitalization. An ad- ditional explanation for the lack of effect of probiotic

treatment is that changes in the oropharyngeal microbiota take time and a po- tential contribution to the

development of pneumonia even longer. Even if Lp299 and Lp299v showed clear growth inhibition on patho- gens in

vitro, we might need longer treatment times and longer local exposure to clinically be able to influence the

oropharyngeal micro- biota. Thus, it cannot be excluded that a study involving a larger pa- tient population and

more intense administration of a combination of probiotics would show an effect on a clinically meaningful

level. In conclusion, this study shows that Lp299 and Lp299v inhibit in vitro growth of commonly found

nosocomial pathogens in the oro- pharynx. 0ral administration of Lp299 and Lp299v to non-ICU pa- tients

did not reduce the risk of disturbed oropharyngeal microbiota

or nosocomial infection.

E THIC S S TATEMENTThe study protocol was approved by the Regional Ethical Review Board, Lund, Sweden (no. 2013/764). Written informed consent, including permission to collect and publish anonymous data, was ob- tained from the patients at inclusion.

ACKNOWLEDG MENTSWe wish to thank research nurses Anne Adolfsson and Susann Schrey for invaluable help, and Gisela Hovold for excellent technical assistance. We also wish to thank Probi AB for generously producing both the probiotic and placebo sachets. This work was funded by research grants from Swedish Region Skane (Doktorand-2019-0144 and Doktorand-2020-0459).

CONFLIC TS OF INTERE S TNone declared.

AUTHOR CONTRIBUTIONAnna Tranberg: Conceptualization (equal); Data curation (equal); Formal analysis (lead); Funding acquisition (lead); Investigation (lead); Methodology (supporting); Project administration (equal); Visualization (lead); Writing-original draft (lead); Writing-review & editing (equal). Bengt Klarin: Conceptualization (equal); Data curation (equal); Formal analysis (supporting); Funding acquisition (support- ing); Investigation (supporting); Methodology (supporting); Project administration (equal); Visualization (supporting); Writing-

original draft (supporting); Writing-review & editing (equal). Julia Johansson:

| 9 ot TRANBERG ET AL.

Conceptualization (supporting); Data curation (equal); Formal analysis (equal); Funding acquisition (supporting); Investigation (equal); Methodology (supporting); Project administration (equal); Visualization (supporting); Writing-original draft (supporting); Writing-review & editing (equal). Lisa I. Påhlman: Conceptualization (supporting); Investigation (supporting); Methodology (equal); Project administration (equal); Supervision (equal); Visualization (supporting); Writing-original draft (equal); Writing-review & editing (equal).

DATA AVAIL ABILIT Y S TATEMENTThe datasets generated and/or analyzed during the in vitro part of the study are available in the figshare repository at https://doi. org/10.6084/m9.figshare.13312118.v1 and https://doi.org/10.6084/ m9.figshare.13311929.v1. The datasets generated and/or analyzed during the RCT part of the study are not publicly available due to pa- tient confidentiality but are available from the corresponding author on reasonable request. The study is registered at ClinicalTrials.gov, with the clinical trial number: NCT02303301: https://clinicaltrials. gov/ct2/show/NCT02303301

ORCIDAnna Tranberg https://orcid.org/0000-0002-7685-2395 Bengt Klarin https://orcid.org/0000-0002-3531-8188 Lisa I. Påhlman https://orcid.org/0000-0001-6366-2309

REFE RENCE SAdebayo, F. A., Afolabi, 0. R., & Akintokun, A. K.

(2014). Antimicrobial properties of purified bacteriocins produced from Lactobacillus casei and Lactobacillus fermentum against selected pathogenic microorganisms. British Journal of Medicine and Medical Research, 4(18), 3415-3431. https://doi.org/10.9734/bjmmr/2014/8584

Annuk, H., Shchepetova, J., Kullisaar, T., Songisepp, E., Zilmer, M., & Mikelsaar, M. (2003). Characterization of intestinal lactobacilli as putative probiotic candidates. Journal of Applied Microbiology, 94(3), 403-412. https://doi.org/10.1046/j.1365-2672.2003.01847.x

Bahrani-Mougeot, F., Paster, B., Coleman, S., Barbuto, S., Brennan, M., Noll, J., Kennedy, T., Fox, P., & Lockhart, P. (2007). Molecular anal- ysis of oral and respiratory bacterial species associated with venti- lator-associated pneumonia. Journal of Clinical Microbiology, 45(5), 1588-1593. https://doi.org/10.1128/jcm.01963-06

Bassis, C., Erb-Downward, J., Dickson, R., Freeman, C., Schmidt, T., Young, V., Beck, J., Curtis, J., & Huffnagle, G. (2015). Analysis of the upper respiratory tract Microbiotas as the source of the lung and gastric microbiotas in healthy individuals. MBio, 6(2), https:// doi.org/10.1128/mBio.00037-15

Bo, L., Li, J., Tao, T., Bai, Y., Ye, X., Hotchkiss, R., Kollef, M., Crooks, N., & Deng, X. (2020). Probiotics for preventing ventilator-associated pneumonia. The Cochrane Database of Systematic Reviews.

de Vrese, M., & Schrezenmeir, J. (2008). Probiotics, prebiotics,

and synbiot-ics. Food Biotechnology, 1-66.

https://doi.org/10.1007/10_2008_097 Dickson, R., Schultz, M., van der Poll, T., Schouten, L., Falkowski, N., Luth,

J., Sjoding, M., Brown, C., Chanderraj, R., Huffnagle, G., Bos, L., de Beer, F., Bos, L., Claushuis, T., Glas, G., Horn, J., Hoogendijk, A., van Hooijdonk, R., Huson, M., . Witteveen, E. (2020). Lung microbiota predict clinical outcomes in critically ill patients. American Journal of Respiratory and Critical Care Medicine, 201(5), 555-563. https://doi. org/10.1164/rccm.201907-1487oc

Didari, T., Mozaffari, S., Nikfar, S., & Abdollahi, M. (2015). Effectiveness of probiotics in irritable bowel syndrome: Updated systematic review

with meta-analysis. World Journal of Gastroenterology, 21(10), 3072. https://doi.org/10.3748/wjg.v21.i10.3072

Frandah, W., Colmer-Hamood, J., Mojazi Amiri, H., Raj, R., & Nugent,K. (2013). 0ropharyngeal flora in patients admitted to the medi- cal intensive care unit: clinical factors and acid suppressive ther- apy. Journal of Medical Microbiology, 62(5), 778-784. https://doi. org/10.1099/jmm.0.053066-0

Gu, W., Wei, C., & Yin, R. (2012). Lack of Efficacy of Probiotics in Preventing Ventilator-Associated Pneumonia. Chest, 142(4), 859-868. https://doi.org/10.1378/chest.12-0679

Hasslöf, P., Hedberg, M., Twetman, S., & Stecksen-Blicks, C. (2010). Growth inhibition of oral mutans streptococci and candida by com- mercial probiotic lactobacilli - an in vitro study. BMC Oral Health, 10(1), https://doi.org/10.1186/1472-6831-10-18

Hatakka, K., Savilahti, E., Pönkä, A., Meurman, J. H., Poussa, T., Näse, L., Saxelin, M., & Korpela, R. (2001). Effect of long-term consumption of probiotic milk on infections in children attending day care cen- tres: double blind, randomised. BMJ, 322(7298), 1327. https://doi. org/10.1136/bmj.322.7298.1327

Hill, C., Guarner, F., Reid, G., Gibson, G., Merenstein, D., Pot, B., Morelli, L., Canani, R., Flint, H., Salminen, S., Calder, P., & Sanders,M. (2014). The International Scientific Association for Probiotics and Prebiotics consensus statement on the scope and appropri- ate use of the term probiotic. Nature Reviews Gastroenterology & Hepatology, 11(8), 506-514. https://doi.org/10.1038/nrgas tro.2014.66

Huffnagle, G., Dickson, R., & Lukacs, N. (2016). The respiratory tract microbiome and lung inflammation: A two-way street. Mucosal Immunology, 10(2), 299-306. https://doi.org/10.1038/ mi.2016.108

Hutt, P., Shchepetova, J., Loivukene, K., Kullisaar, T., & Mikelsaar, M. (2006). Antagonisticactivityofprobioticlactobacilliandbifidobacteriaagainst entero- and uropathogens. Journal of Applied Microbiology, 100(6), 1324-1332. https://doi.org/10.1111/j.1365-2672.2006.02857.x

Johanson, W., Pierce, A., & Sanford, J. (1969). Changing Pharyngeal Bacterial Flora of Hospitalized Patients. New England Journal of Medicine, 281(21), 1137-1140. https://doi.org/10.1056/NEJM196911202812101

Karacaer, F., Hamed, I., Özogul, F., Glew, R., & Özcengiz, D. (2017). The function of probiotics on the treatment of ventilator-associated pneumonia (VAP): facts and gaps. Journal of Medical Microbiology, 66(9), 1275-1285. https://doi.org/10.1099/jmm.0.000579

Klarin, B., Adolfsson, A., Torstensson, A., & Larsson, A. (2018). Can probi- otics be an alternative to chlorhexidine for oral care in the mechan- ically ventilated patient? A multicentre, prospective, randomised controlled open trial. Critical Care, 22(1), https://doi.org/10.1186/ s13054-018-2209-4

Klarin, B., Molin, G., Jeppsson, B., & Larsson, A. (2008). Use of the pro- biotic Lactobacillus plantarum 299 to reduce pathogenic bacteria in the oropharynx of intubated patients: a randomised controlled open pilot study. Critical Care, 12(6), R136. https://doi.org/10.1186/cc7109

Lin, T., & Pan, T. (2019). Characterization of an antimicrobial sub- stance produced by Lactobacillus plantarum NTU 102. Journal of Microbiology, Immunology and Infection , 52(3), 409-417. https://doi. org/10.1016/j.jmii.2017.08.003

Ling, Z., Liu, X., Guo, S., Cheng, Y., Shao, L., Guan, D., Cui, X., Yang, M., & Xu, X. (2019). Role of probiotics in Mycoplasma pneumoniae pneumonia in children: A short-term pilot project. Frontiers in Microbiology, 9, https://doi.org/10.3389/fmicb.2018.03261

Morrow, L., Kollef, M., & Casale, T. (2010). Probiotic Prophylaxis of Ventilator-associated Pneumonia. American Journal of Respiratory and Critical Care Medicine, 182(8), 1058-1064. https://doi. org/10.1164/rccm.200912-18530C

Niveen, E., Angi, E., Mohamed, E., Ghada, B., & Wafaa, S. (2016). Role of probiotics in prevention of hospital acquired pneumonia in Egyptian children admitted to Pediatric Intensive Care Unit of Mansoura University Children's Hospital. African Journal of Microbiology Research, 10(31), 1240-1247. https://doi.org/10.5897/ ajmr2016.8085

Prabhurajeshwar, C., & Chandrakanth, R. (2017). Probiotic potential of Lactobacilli with antagonistic activity against pathogenic strains: An in vitro validation for the production of inhibitory substances. Biomedical Journal, 40(5), 270-283. https://doi.org/10.1016/j. bj.2017.06.008

Retchless, A., Kretz, C., Rodriguez-Rivera, L., Chen, A., Soeters, H., Whaley, M., & Wang, X. (2020). 0ropharyngeal microbiome of a college population following a meningococcal disease outbreak. Scientific Reports, 10(1), https://doi.org/10.1038/s41598-020- 57450-8

Seddik, H., Bendali, F., Gancel, F., Fliss, I., Spano, G., & Drider, D. (2017). Lactobacillus plantarum and Its Probiotic and Food Potentialities. Probiotics and Antimicrobial Proteins, 9(2), 111-122. https://doi. org/10.1007/s12602-017-9264-z

Simons, A., Alhanout, K., & Duval, R. (2020). Bacteriocins, antimicrobial peptides from bacterial origin: Overview of their biology and their impact against multidrug-resistant bacteria. Microorganisms, 8(5), 639. https://doi.org/10.3390/microorganisms8050639

Song, D., Zhu, M., & Gu, Q. (2014). Purification and characterization of plantaricin ZJ5, a new bacteriocin produced by Lactobacillus plan- tarum ZJ5. PLoS One, 9(8), e105549. https://doi.org/10.1371/journ al.pone.0105549

Tranberg, A., Thorarinsdottir, H., Holmberg, A., Schött, U., & Klarin, B. (2018). Proton pump inhibitor medication is associated with col- onisation of gut flora in the oropharynx. Acta Anaesthesiologica Scandinavica, 62(6), 791-800. https://doi.org/10.1111/aas.13094

Wallace, C., & Milev, R. (2017). The effects of probiotics on depres- sive symptoms in humans: A systematic review. Annals of General Psychiatry, 16(1), https://doi.org/10.1186/s12991-017-0138-2

Wang, J., Liu, K., Ariani, F., Tao, L., Zhang, J., & Qu, J. (2013). Probiotics for preventing ventilator-associated pneumonia: a systematic review and meta-analysis of high-quality randomized controlled trials. PLoS One, 8(12), e83934. https://doi.org/10.1371/journal.pone.0083934

Weng, H., Li, J., Mao, Z., Feng, Y., Wang, C., Ren, X., & Zeng, X. (2017). Probiotics for preventing ventilator-associated pneumonia in me- chanically ventilated patients: a meta-analysis with trial sequen- tial analysis. Frontiers in Pharmacology, 8, https://doi.org/10.3389/ fphar.2017.00717

TRANBERG ET AL.

10 ot 12

|

How to cite this article: Tranberg A, Klarin B, Johansson J, Påhlman LI. Efficacy of Lactiplantibacillus

plantarum 299 and 299v against nosocomial oropharyngeal pathogens in vitro and as an oral prophylactic treatment in a randomized, controlled clinical trial. MicrobiologyOpen. 2020;00:e1151. https://doi.org/10.1002/mbo3.1151

APPENDIX 1

800E.cloacae growth inhibition 761,94

700

600

500

400 389,79

300

200

100

0

100,00

Growth in % in five hours

0,25 0,0016,20 30,24

0,00

–100

MRS=control

Lp299 Lp299v SN Lp299 SN Lp299v Neu SNLp299

Neu SN Lp299v

MRS pH 4

Figure A1 The inhibitory effect of Lp299 and Lp299v on E. cloacae is pH-dependent. E. cloacae was incubated with Lp299, Lp299v, or cell- free supernatants from overnight cultures of the lactobacilli. To explore the role of pH, E. cloacae was also incubated with pH-neutralized supernatants and with acidified MRS broth. There is a significant difference in five-hour growth between the control (E. cloacae grownin 25% MRS broth) and all other groups (p < 0.05). There is also a significant difference in the five-hour growth between neutralizedsupernatants and all other groups (p < 0.05). SN =supernatant. Neu =neutralized

Figure A2 0vernight cultures of E. cloacae were allowed to incubate with pure supernatant (SN); boiled SN; and SN treated with the proteinases pepsin, proteinase K, and trypsin. Both Lp299 and Lp299v SN were included in the experiment. The results show that all of the pre-treated supernatants showed the same clear growth inhibition as untreated supernatants, whereas the control (E. cloacae incubated with MRS broth) showed expected growth of the pathogen during the incubation time

| 11 ot TRANBERG ET AL.

0

Average growth in 5 hours

0,02 0,00 0,27 0,00 0,00 0,00 0,00 0,00 0,00 0,25

20

40

60

80

E. cloacae incubated with proteinase-treated supernatants100

100,00

120

Gro

wth

in %

com

paed

to c

ontro

lG

row

th (%

of c

ontro

l)

Results from the agar overlay

(RS)E. cloacae (RS) 0 0 0 0 0 0P. aeruginosa (RS) 0 0 0 0 0 0P. aeruginosa (CI) 0 - - 0 - -

Abbreviations: CI, Clinical isolate; RS, Reference strain.Growth described as 0 = no growth, or 1= growth - = experiment not performed

TRANBERG ET AL.

12 ot 12

|Lp299 Lp299v TABLE A1

experiment

4 × 104 4 × 105 4 × 106

4 × 104 4 × 105 4 × 106

CFU CFU CFU CFU CFU CFU

S. aureus (RS) 0 0 0 0 0 0S. aureus (CI) 0 - - 0 - -E. faecalis (RS) 0 0 0 0 0 0E. faecalis (CI) 0 - - 0 - -E. faecium (RS) 0 0 0 0 0 0E. faecium (CI) 0 - - 0 - -E. coli (RS) 0 0 0 0 0 0E. coli (CI) 0 - - 0 - -

--0--0K. pneumoniae(CI)