1

Kyla Ortved, DVM, PhD, DACVS, DACVSMR Assistant Professor of Large Animal Surgery

New Bolton Center, University of Pennsylvania

Septic Joints: Diagnostics, Treatment and Outcomes

Common equine emergency

n 3 Main Causes: 1) Hematogenous spread (foals) 2) Penetrating wound 3) Iatrogenic

n  Less commonly extension of peri-articular infection

Clinical signs n Severe, NWB lameness

– Lameness may be less severe if joint open & draining

n  Joint effusion

n Peri-articular swelling

n Heat, pain on palpation

Pathophysiology n Bacterial contamination of synovial space n  Inoculation dose depends on virulence &

pathogenicity of organism – ~ 105 CFU – Staph spp. require significantly fewer CFU – Adequan significantly decreases CFU needed

(Gustafson et al., 1989)

n Colonization of synovium leads to marked synovial inflammatory response

Hematogenous route of infection

n  Tortuous nutrient vessels (epiphyseal, metaphyseal, periosteal and terminal branches of nutrient artery)

n  Foals can also develop thrombosis/osteonecrosis

Types of Septic Arthritis in Foals n S-Type: Synovial fluid and membrane infected

E-Type: Articular epiphysis also involved

P-Type: Physis of long bone infected +/- joint infected

2

Etiological agents n  Foals

–  Gram –ve: E. coli, Salmonella, Pseudomonas, Enterobacter, Acinetobacter, Proteus, Klebsiella, Citrobacter spp.

–  Rhodococcus (usually immune-mediated arthritis)

n  Adults –  Penetrating wound: Gram –ve > gram +ve –  Iatrogenic: Gram +ve, Strep, Staph aureus

n  Other less common agents –  Mycoplasma –  Fungal

Joint degradation n Inflammatory mediators

– Neutrophil infiltration – Activated synoviocytes & chondrocytes – Bacteria

n Bacterial toxins – Alpha- and gamma-cytolytic toxins

n Degradative enzymes – Hyaluronidase, MMPs, aggrecanases

Diagnostic considerations n Signalment

n History

n Hematology

n Synovial fluid analysis

n  Imaging: radiographs, ultrasound

n Advanced imaging: nuclear scintigraphy, CT, MRI

Diagnosis – Signalment & history

n  Foals (hematogenous) –  Prematurity – High sepsis score – Umbilical abscess –  Partial or complete failure of passive

transfer §  50-88% of foals with septic arthritis have FPT (Meijer et al.

2000)

–  Fever

n Adults –  Penetrating wound

–  Iatrogenic § Injection § Post-op (rare ~1%)

– Nearby infection (such as foot abscess or cellulitis)

–  Adults are not necessarily febrile

Diagnosis – Signalment & history Hematology n CBC

– Normal in adults – Neutrophilia in foals

n Fibrinogen – Mild hyperfibrinogenemia in adults – Moderate hyperfibrinogenemia in foals

n Serum amyloid A – Usually increased

n Blood culture for foals

3

n Site away from open wounds -  Probably OK to sample through inflamed tissue

n Distend joint with sterile fluid to determine communication with wounds

n  Inject joint with antibiotic after sampling

***Always sample joint if you are concerned

about communication or sepsis

Synoviocentesis

Cytology Culture Culture broths

Sample types

Diagnosis - Synovial fluid n Synovial fluid analysis

– Viscosity changes rapidly with sepsis – Cloudy fluid suggests at least 30 * 10^3/dL cells

Diagnosis - Synovial fluid

Normal Inflammation Infection

Total Protein

<2.0 - <2.5 g/dL <2.5 g/dL >4.0 g/dL

WBC <450 to <5000 500 to 20,000 >30,000/dL

Cell Type

<10% neutrophils 10-20% >80% neutrophils

Diagnosis - Synovial fluid n Synovial fluid cytology n Gram stain

– Uncommon to see bacteria – Rapid guidance for antibiotic

selection

Diagnosis – Importance of neutrophils

n Normal: <10% neutrophils; > 90% lymphocytes, monocytes, macrophages

n Septic: >80% neutrophils almost always infected

- <75% neutrophils suggests that the infection is resolving

n Early in infection the neutrophils are usually non-degenerate

4

Diagnosis – Synovial fluid lactate and SAA

n  Increased lactate in synovial fluid > 4.9mmol/L in 66% (Van Pelt, 1974)

n  Increased SAA common but can be normal (Jacobsen et al., 2006)

group 1 (Figure 2). Neither serum nor synovial fluidSAA concentrations in group 4 differed from those ingroup 1. Concentrations of SAA and total protein insynovial fluid samples were significantly correlated (P <0.001; Spearman correlation coefficient, 0.57).

Concentrations of SAA in serum and synovial fluidranged from below detection limit to 402 and 94.5mg/L, respectively, among horses in group 2 (Figures 1and 2). The highest concentrations were detected in ahorse with an open infected fracture of the secondmetacarpal bone in close proximity to the radiocarpal

and intercarpal joints. Two horses in group 2 hadserum and synovial fluid SAA concentrations less thanthe detection limit: in 1 of those horses, arthrocentesishad been performed 2 hours after injury, and the otherhorse had a closed fracture involving the radiocarpaljoint. Two horses (1 with a wound that penetrated tothe radiocarpal joint and 1 with a wound that pene-trated to the talocrural joint) had low initial SAA con-centrations in synovial fluid that decreased furtherduring the course of treatment with joint lavage andadministration of antimicrobials and nonsteroidal anti-inflammatory drugs (Figure 4).

Four horses with active infectious arthritis or ten-ovaginitis had moderate to high SAA concentrations inserum and synovial fluid (Figures 1 and 2). However,

1740 AJVR, Vol 67, No. 10, October 2006

Figure 1—Serum SAA concentrations in 10 healthy horses and21 horses with various classifications of joint disease. Group 1= Healthy control horses. Group 2 = Horses with possible jointcontamination from wounds or fractures communicating with orin close proximity to joints. Group 3 = Horses with infectiousarthritis or infectious tenovaginitis. Group 4 = Horses with non-inflammatory joint diseases (osteoarthritis and osteochondro-sis). Values in columns with different superscript letters (a,b) aresignificantly (P ≤ 0.05) different.

Figure 2—Synovial fluid SAA concentrations in the same horsesas in Figure 1. See Figure 1 for key.

Figure 3—Mean ± SD total protein concentrations in synovialfluid in response to repeated arthrocentesis in 5 horses.Arthrocentesis was performed at time points 0, 4, 8, 12, 24, 48,72, 96, and 144 hours. *Significant (P < 0.05) difference, com-pared with value at the 0 time point.

Figure 4—SAA concentrations in synovial fluid during the courseof treatment of infectious arthritis (square) in 1 horse and possi-ble joint contamination (radiocarpal joint, triangle; talocrural joint,circle) in 2 horses.

06-03-0026r.qxp 9/14/2006 2:53 PM Page 1740

group 1 (Figure 2). Neither serum nor synovial fluidSAA concentrations in group 4 differed from those ingroup 1. Concentrations of SAA and total protein insynovial fluid samples were significantly correlated (P <0.001; Spearman correlation coefficient, 0.57).

Concentrations of SAA in serum and synovial fluidranged from below detection limit to 402 and 94.5mg/L, respectively, among horses in group 2 (Figures 1and 2). The highest concentrations were detected in ahorse with an open infected fracture of the secondmetacarpal bone in close proximity to the radiocarpal

and intercarpal joints. Two horses in group 2 hadserum and synovial fluid SAA concentrations less thanthe detection limit: in 1 of those horses, arthrocentesishad been performed 2 hours after injury, and the otherhorse had a closed fracture involving the radiocarpaljoint. Two horses (1 with a wound that penetrated tothe radiocarpal joint and 1 with a wound that pene-trated to the talocrural joint) had low initial SAA con-centrations in synovial fluid that decreased furtherduring the course of treatment with joint lavage andadministration of antimicrobials and nonsteroidal anti-inflammatory drugs (Figure 4).

Four horses with active infectious arthritis or ten-ovaginitis had moderate to high SAA concentrations inserum and synovial fluid (Figures 1 and 2). However,

1740 AJVR, Vol 67, No. 10, October 2006

Figure 1—Serum SAA concentrations in 10 healthy horses and21 horses with various classifications of joint disease. Group 1= Healthy control horses. Group 2 = Horses with possible jointcontamination from wounds or fractures communicating with orin close proximity to joints. Group 3 = Horses with infectiousarthritis or infectious tenovaginitis. Group 4 = Horses with non-inflammatory joint diseases (osteoarthritis and osteochondro-sis). Values in columns with different superscript letters (a,b) aresignificantly (P ≤ 0.05) different.

Figure 2—Synovial fluid SAA concentrations in the same horsesas in Figure 1. See Figure 1 for key.

Figure 3—Mean ± SD total protein concentrations in synovialfluid in response to repeated arthrocentesis in 5 horses.Arthrocentesis was performed at time points 0, 4, 8, 12, 24, 48,72, 96, and 144 hours. *Significant (P < 0.05) difference, com-pared with value at the 0 time point.

Figure 4—SAA concentrations in synovial fluid during the courseof treatment of infectious arthritis (square) in 1 horse and possi-ble joint contamination (radiocarpal joint, triangle; talocrural joint,circle) in 2 horses.

06-03-0026r.qxp 9/14/2006 2:53 PM Page 1740

Increased SAA in group 3 (septic)

SAA decreased during course of treatment

Diagnosis – Serum SAA n  Serum SAA may be useful to monitor response

to treatment (Haltmayer et al., 2017)

monitoring treatment response in patients with injuriespenetrating synovial structures in a clinical set up. Dataindicates that the course of plasma SAA concentrationreflects the response to treatment well. Patients ofGroup 1 and 3 with a highly favourable response to ini-tial surgical treatment showed a continuous decrease ofSAA concentrations between 48 and 96 h after surgeryas well 6 days after surgery. In contrast in cases with on-going infection (Group 2) different patterns of SAAcould be observed. Two horses (4, 19) showed persist-ently high or even increased levels of SAA after the firstsurgery, one horse (17) showed a very high increase be-tween the first and the second surgery and the otherhorse (7) showed very high initial SAA values and only a

minor decrease before the second surgery. However thishas to be interpreted with care due to the low numberof patients and no conclusion can be drawn from thedata at this point.Persistently high concentrations of SAA have been pre-

viously reported in horses with complications after elect-ive and emergency surgery [14, 25, 26]. In the presentstudy this was considered to be an indication for lack oftreatment response. However, further surgical treatmentwas instigated only with corresponding clinical signs andsynovial fluid analysis and was not based on SAA values.This does not preclude that animals could have potentiallyimproved without additional intervention. However, suchan experimental set up cannot be used in client owned

Fig. 1 a Course of plasma SAA of all patients in Group 2. Peak values could be observed between pre OP (n = 4) and 48 h after surgery (n = 7).Notice the continuous decrease of plasma SAA concentrations after surgery. b Box and whisker plots of course of plasma SAA concentrationduring treatment of patients of Group 1 Explanations: In box-and-whisker plots, central box represents values from lower to upper quartile, middleline represents the median; whiskers extend from minimum to maximum value, excluding outside and far out values which are displayed asseparate points

Fig. 2 Course of plasma SAA in horses of Group 2 Timepoints of surgery are marked (+). In patients 4 and 19 an increase of plasma SAA wasobserved before the following surgical treatment. Patient 7 showed very high initial SAA values with only a minor decrease before the secondsurgical intervention. Patient 17 showed an unusual high increase after the first surgery

Haltmayer et al. BMC Veterinary Research (2017) 13:137 Page 7 of 11

horses. The time interval of 48 h for SAA sample collec-tion in the present study was chosen because SAA showspeak concentration 48 h after onset of infection [13, 18].Concentrations of SAA correspond to the severity of tis-sue damage as well as to the volume of damaged tissue[18] and drop rapidly as soon as synthesis stops [18]. Dueto these unique features of SAA sequential measurementsof plasma SAA concentrations were suggested as a usefulaid in patient management, planning of treatment strat-egies and in decision-making whether to perform surgeryor not [13, 18].Duration of the disease process prior to admission as

well as the stage of the disease process at admission haveto be carefully considered when interpreting SAA concen-trations in a patient [17, 18]. In the current study patientsin Group 3 had normal or very low SAA concentrations(range from 6.7 mg/L to 23.1 mg/L) at the time of admis-sion, despite a positive pressure test and/or neutrophilcounts >90% in synovial fluid. In these patients the re-ported duration between detection of injury by the ownerand hospital admission was <24 h thus below the reportedlag time for SAA increase. Jacobsen et al. [17] also de-scribed one case where the authors assumed that the timefrom injury to sample collection was too short for thelocal inflammatory response to induce hepatic SAA syn-thesis. In this particular case the delay between injury andsample collection was only 2 h. The three patients inGroup 3 support this suggestion [17]. Therefore care hasto be taken when interpreting single SAA values, espe-cially in early stages of disease and results of the currentstudy strongly suggest that SAA values obtained at a singletime point should not be used for diagnostic purposes.Also low-grade infections [14] and well sequestrated

infections [14] were previously reported to cause a re-tarded response in plasma SAA in relation to clinicalsigns. This can be explained by a local inflammatory re-sponse that is not “strong” enough to trigger SAA synthe-sis in the liver and incite a systemic inflammatoryresponse. In contrast Belgrave et al. [14] reported thatSAA values were increased 24 h prior to onset of clinicalsigns in horses with colic and metritis [13] and thereforeSAA could potentially aid in early detection of inflamma-tion. However due to the limited number of patients inGroup 3 no conclusions can be drawn at this point.In the present study most patients showed peak concen-

tration of plasma SAA 48 h after surgery. In all cases sur-gery was performed on the day of hospital admission.Therefore, peak values might have been influenced by thedegree of infection of the injury site and/or the synovialstructure, as well as by surgical trauma [25–27], general an-aesthesia [28, 29] or antimicrobial treatment [30]. Jacobsenet al. [27] reported slightly increased plasma SAA values upto 5 days after arthroscopy (up to 100 mg/L), however theselevels remain much lower than in the cases of the presentstudy [26]. Sanchez-Teran et al. [31, 32] recently suggestedin their study that in healthy adult horses neither arthro-scopic lavage nor through-and-through joint lavage withneedles influenced plasma SAA concentrations 48–120 hafter surgery. Moreover, Pollock et al. [26] described thathealthy horses undergoing an elective surgery respondedsimilar to surgical trauma than horses with elevated plasmaSAA concentrations undergoing emergency surgery. In fourpatients in Group 1 a decrease in plasma SAA could be ob-served between admission and 48 h after surgery (Fig. 1a).These results do not support the suggestion that surgicaltrauma alone led to the typical rise and fall pattern post-

Fig. 3 Illustration of the typical rise and fall pattern of plasma SAA concentrations of patients with injuries <24 h. Peak concentrations are observed48 h after surgery. Notice the relative low pre OP plasma SAA values

Haltmayer et al. BMC Veterinary Research (2017) 13:137 Page 8 of 11

Horses with <24hr history of penetrating wound

responding to 1 surgery

Horses with >24hr history of penetrating wound requiring >1 surgery

Diagnosis - Culture n Culture

– Only ~50% of samples will have positive culture – Positive culture correlate with decreased

prognosis (Taylor et al., 2010)

– Synovial fluid vs. synovium – Culture broth increases rate of positive culture

(Dumoulin et al., 2010)

–  Isolation of a single organism is almost always indicative of infection

n Antimicrobial sensitivity important

Diagnosis - Radiographs n Radiographic lytic changes can occur in

less than 1 week – Most changes are seen at 7-10 days post

infection

n Evaluate for fracture, gas in joint, lytic changes, foreign bodies – 59% of foals have osteomyelitis/bone necrosis

concurrently

One month old Thoroughbred foal who recently developed severe swelling in the right stifle and associated lameness

Foal with septic physitis

Parviainen, 2002

5

Thoroughbred gelding with RF fetlock effusion Thoroughbred gelding that stepped on a nail several hours prior to presentation

Specialty diagnostics n Nuclear scintigraphy

– Can be very useful in subclinical infections

n MRI

– Sensitive for osteomyelitis – Fluid and bone may be affected

Gaschen et al., 2011

Easley et al., 2011

Advanced techniques n  Molecular diagnostics - PCR

–  Detects bacterial/viral DNA (MicroGen-DX, Texas) –  Specific primers (e.g. 16S ribosomal RNA) –  Can be multiplexed (multiple genes in one mix)

n Advantages §  Rapid (<8 hours) §  No interference from antibiotics §  Sensitive (10 CFU/mL)

n Disadvantages §  Equipment §  Contamination by skin or environmental bacteria

Stahl et al., 2000

Yang et al., 2008

Kim et al., 2010

Treatment principles

I.  Sterilize the joint

II.  Facilitate drainage

III.  Restore normal synovial joint viscosity and fluid properties

1. Sterilize the joint n Systemic antimicrobials

– Broad spectrum IV to begin

n Local antimicrobials: –  Intra-articular –  IV regional perfusion –  Intraosseous – Elution from antibiotic beads

6

Intra-articular antimicrobials n High levels directly into joint n Most effective after lavage (vs. in lavage

fluids) n  IA drains can be used to deliver antibiotics

Physiological considerations in septic arthritis

n Hypervascularity due to inflammatory response which influences pharmacokinetics: – Greater rate of transfer of antibiotics from serum

to synovium – More persistence of the drug in the diseased joint

Errecalde et al., 2002

Semi-logarithmic plot of average serum and synovial

concentrations for six arthritic horses after i.v. administration of 40 mg AMX/kg body mass

Regional limb perfusion n  Advantages

–  Reaches infected tissue / joints by diffusion –  Far greater concentration (>10 fold) than systemic IV –  Concentrations remain higher for longer

n  Disadvantages –  May not work if tourniquet not functional –  Venous thrombosis –  Difficult in limb with cellulitis

Regional limb perfusion n  Tourniquet proximal to joint

–  Rubber (Levine et al., 2010), esmarch, pneumatic (Alkabes et al., 2011)

n  Volume of perfusate –  High volume (100ml) may be better if tourniquet proximal

(Oreff et al., 2016)

–  Lower volume 30-60ml for distal tourniquets

n  Appropriate sedation to minimize movement n  Duration 20-30 minutes

Ivester & Adams, 2007

RLP - Antimicrobials n  Amikacin most common n  Cephalosporins, ticarcillin, vancomycin,

enrofloxacin, imipenem n  Dose of antimicrobials

–  1/3 systemic dose? –  2-3g amikacin

RLP - Antimicrobials Antimicrobial Dosage Methods of delivery

Cefazolin 1g All

Ceftazidime/ceftriaxone 2g RLP, IA

Penicillin/ampicillin 1/4 – 1/2 systemic dose RLP, IA

Gentamicin 0.5 – 1g 1/3 systemic dose

All

Amikacin 1-3g All

Ceftiofur 1g RLP, IA

Ticarcillin+clavulanate 500mg – 1g All

Vancomycin 300mg RLP, beads

Enrofloxacin 500mg All

Imipenem 500mg RLP, IA

7

Intraosseous perfusion n  More invasive/painful than IV regional n  Hole drilled into long bone and a cannula is used

to deliver infusion n  Requires tourniquet like IV regional

Ivester et al. 2007

Antibiotic-impregnated materials

n Non-absorbable – Polymethylmethacrylate (PMMA) – Normally need to be removed – Elution of antibiotics is longest (months)

n Absorbable – Collagen – Gelatin – DL lactide-glycolide copolymers – Plaster of Paris – Elution over days to weeks (Haerdi-Landerer et al., 2010)

2. Drainage n Lavage with balanced physiological solution n Arthroscopy n Arthrotomy n Removal of necrotic tissues

2. Drainage n  If significant amount of fibrin or thick synovial

fluid or bony lesion, more aggressive drainage is selected: –  Arthroscopy –  Arthrotomy

Arthroscopy vs. Arthrotomy n Experimentally, both eliminate the infection

equally but arthrotomy was faster n Arthrotomy has increased risk of ascending

infection Bertone et al. 1992

Advantages Disadvantages Lavage - Can be done

anywhere - Less expensive - Can be done standing

- Unable to remove fibrin, infected synovium - Unable to examine joint

Arthroscopy - Excellent exam and removal of fibrin & synovium - Debridement of bony lesion

- Increase costs - General anesthesia

Arthrotomy Best drainage - Ascending infection - Premature closure

8

Benefit of arthroscopic lavage 3. Restoration of joint health n  Anti-inflammatories

–  NSAIDS –  HA –  Bandaging

n  Additional pain control –  Epidurals –  IA anesthetics –  Bandaging –  Support of contralateral limb

Sodium hyaluronate (HA)

n Anti-inflammatory properties

n Decreases: –  Cellular infiltration –  Granulation tissue –  Total glycosaminoglycans loss (p>0.05)

Brusie et al. 1992

Epidural analgesia n Combination of opioid and α-2 agonist

most effective n Single injection or indwelling catheter n Best for hindlimb

Epidural analgesia Table 1Drug regimens used for spinal (epidural/intratechal) anesthesia/analgesia and reported volumes and dosages

Drug Volume (mL) Site of Injection Duration of Effect (Hours) Comments

Single Drug

Lidocaine 0.5%–2%a 5–8 Co1-Co2/L-S 0.75–1.5 Epidural/intratechal, rapid onset

Mepivacaine 2% 5–8 Co1-Co2/L-S 1.5–3 Epidural/intratechal, rapid onset

Ropivacaine 0.1%–0.5% 5–10 Co1-Co2/L-S 3–8 Epidural/intratechal, less ataxia

Bupivacaine 0.1%–0.5%a 5–8 Co1-Co2/L-S 3–8 Epidural/intratechal, more ataxia

Xylazine 0.17 mg/kga 10 Co1-Co2/L-S 1.0–1.5 Epidural/intratechal, sedation/ataxia

Detomidine 20–40 mg/kg 5–10 Co1-Co2/L-S 2–4 Epidural/intratechal, sedation/ataxia

Medetomidine 2–5 mg/kg 10–30 Co1-Co2/L-S 4–6 Epidural/intratechal, sedation/ataxia

Morphine 0.05–0.2 mg/kga 10–30 Co1-Co2/L-S 3–8 Epidural ONLY, useful for CRI mL/h) viaepidural catheter

Methadone 0.1 mg/kg 20 Co1-Co2 3 Epidural, similar to morphine

Tramadol 1 mg/kg 10–30 Co1-Co2 4–5 Epidural, similar to morphine

Ketamine 0.5–2.0 mg/kgb 10–30 Co1-Co2 0.5–1.25 Epidural, some ataxia

Hydromorphone 10–30 Co1-Co2 4–5 Epidural, similar to morphine

Drug Combinations (Balanced Regional Analgesia)

Lidocaine 2 % 1 xylazine 0.17 mg/kg 5–8 Co1-Co2 4–6

Lidocaine 2 % 1 morphine 0.1–0.2 mg/kg 5–8 Co1-Co2 4–6

Bupivacaine 0.125 % 1 morphine0.1–0.2 mg/kg

10–30 Co1-Co2/L-S 8–>12 Also useful for CRI (0.5–2 mL/h) viaepidural catheter

Xylazine 0.17 mg/kg 1 morphine0.1–0.2 mg/kg

10–30 Co1-Co2/L-S !12

Detomidine 10–30 mg/kg 1 morphine0.1–0.2 mg/kg

5–10 Co1-Co2/L-S 24–486–8

Epidural/intratechal moderate painEpidural/intratechal severe pain

Use lower doses/volumes of the less concentrated drugs for intratechal injection. Intratechal (subarachnoid) injection potentially increases the risk of infectionsand inflammation (spinal cord and meninges).

a Author’s preference for most cases.b May cause injury to the spinal cord.

Natalin

i556

Natalini, 2010

Prognosis n Adults

§  ~70% for life & return to athleticism § Decreased with positive bacterial culture § Decreased long-term px with Staph aureus

n Foals §  ~50% for life and athleticism § Decreased with comorbidities including septicemia,

osteomyelitis

9

Worse prognosis in foals vs. adults n Usually associated

with failure of passive transfer & septicemia

n Often multiple joints (racing prognosis decreases)

n Septic osteitis/osteomyelitis or physitis

Neil et al., 2010

Summary n  Septic arthritis is an emergency n  Early, aggressive treatment is best n  Treatment can be prolonged and expensive n  Prognosis

§  Foals 50% §  Adults 70-80%

n  Combination treatment –  Systemic and local antibiotics –  Anti-inflammatories –  Lavage –  Restoration of joint health –  Pain control

Questions?