PERSONALPRACTICE - adc.bmj.com · The recommended starting dose of dexam-ethasone is in the region of05-1 mg/kg/day, which is a relatively large dose to weight ratio.'5-25 As a lot

Archives of Disease in Childhood 1993; 68: 330-336

PERSONAL PRACTICE

The effectiveness and side effects ofdexamethasone in preterm infants withbronchopulmonary dysplasia

PCNg

The incidence of bronchopulmonary dysplasia(BPD) among ventilated infants is estimatedto be between 4 and 40% depending on gesta-tion but undoubtedly the highest incidence, inexcess of 70%, occurs in infants weighing1000 g or less at birth.' The incidence of BPDmay further increase in the coming decade asadvances in neonatal intensive care enableclinicians to save even smaller, lower gesta-tion, and more critically ill infants. BPD isalready a major cause of mortality and longterm morbidity. A significant proportion ofinfants will be oxygen dependent for a pro-longed period of time. Growth delay andneurodevelopmental deficit have also beenreported with increasing frequency.' Thus,effective therapeutic measures are urgentlyrequired.Dexamethasone, a potent, long acting

steroid with almost exclusive glucocorticoidproperty, was intensively studied and appearedpromising in improving pulmonary function.This article examines the efficacy and adverseeffects of dexamethasone treatment in pre-term infants with BPD.

Academic Departmentof Child Health,St Bartholomew'sHospital MedicalCollege, London

Correspondence to:Dr P C Ng,Special Care Baby Unit,St Bartholomew's Hospital atHomerton, Homerton Row,London E9 6SR.

Mechanism of actionSeveral modes of action have been proposedto explain the association between steroidtreatment and improvement in lung function.They include an increase in surfactant synthe-sis2 3; enhancement of ,-adrenergic activity4;stimulation of antioxidant production2; stabili-sation of cell and lysosomal membrane5;breakdown of granulocyte aggregates withimprovement in the pulmonary microcircula-tion6 7; inhibition of prostaglandin and leuko-trine synthesis8 9; removal of excess lungwater'0; and suppression of the cytokine medi-ated inflammatory reaction in the lung." Thelatter two theories are of particular interestand may provide an explanation for the rapidchanges in lung function.Ariagno et al demonstrated a marked

steroid induced reduction in lung water con-

tent in newborn mice raised in an 80% oxygenenvironment.'0 They showed that the decreasein lung water was due to a loss of fluid fromthe interstitial tissue of the lung with no

reduction in the thickness of the blood-gasbarrier. Sahebjami et al have also shown thatsteroids can accelerate the resolution of pul-monary oedema and improve alveolar volume

either by increasing the rate of reabsorption ordecreasing the formation of alveolar fluid exu-date."2 Further support for the lung fluid the-ory comes from a recent study that demon-strated that diuresis occurred within 48 hoursof the commencement of dexamethasone inpreterm infants with BPD. 13 This diureticphase correlated closely with the period ofimproved lung function.Another plausible explanation may relate

to the ability of corticosteroids to suppressthe inflammatory process induced by baro-trauma and oxygen toxicity. The cytokine,tumour necrosis factor-oa (TNF-x), was notpresent in the bronchopulmonary secretionsof ventilated preterm infants until after fourdays of postnatal age."I Thereafter, the con-centration increased progressively and thehighest concentrations were found in babieswith BPD who required prolonged ventila-tory and oxygen support. It was of interestto note that the time course of TNF-a pro-duction correlated well with the pattern ofinflammatory cell migration into the lungs.The influx of polymorphonuclear neu-trophils and activated alveolar macrophagesin turn triggered an inflammatory reactionleading to pulmonary fibrosis.'4 The concen-trations of TNF-ax and inflammatory cellsfrom bronchopulmonary lavage were dra-matically decreased within 24 hours of start-ing steroid treatment." Dexamethasonedecreased the synthesis of the cytokine andlimited the polymorphonuclear cell andmacrophage migration to the inflamed area.Hence, it has the ability to restrain the alve-olar macrophage induced alveolitis and pre-vent the full activation of the inflammatorycascade.

Despite intensive research, the exactmechanism to explain the action of dexam-ethasone has not been fully delineated. Asthe pathogenesis of BPD is complex andmultifactorial, it is likely that more than onemechanism is responsible for the acute andrapid improvement of pulmonary functionseen with steroid. The overall pictureappears to be that dexamethasone probablyexerts its effect by reducing the tracheo-bronchial alveolar inflammatory responseand pulmonary oedema, thereby facilitatinggas exchange, airway patency, and improvinglung compliance which enables successfulextubation.

330

on January 12, 2020 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.68.3_Spec_N

o.330 on 1 March 1993. D

ownloaded from

http://adc.bmj.com/

331The effectiveness and side effects of dexamethasone in preterm infants with bronchopulmonary dysplasia

Table I Use of steroids in BPD: prospective/controlled studies

Authors Study design No of Results Complicationspatients

Mammel et al Prospective, double blind, randomised, 6(1983)15 placebo controlled, crossover,

sequential analysis

Avery et al(1985) 16

Ariagno et al(1987)17

Cummings et al(1989)18

Harkavy et al(1989)'9

Ohlsson et al(1989)20

Noble-Jamiesonet al (1989)23

Kazzi et al(1990)21

Singhal et al(1990)22

CollaborativeDexamethasoneTrial Group(1991)24

Prospective, double blind, randomised,placebo controlled, sequential analysis

Prospective, double blind, randomised,placebo controlled




Prospective, double blind, randomised,placebo controlled (non-ventilatordependent infants)

Prospective double blind, randomised,placebo, controlled (oral steroids)


Multicentre, prospective, double blind,randomised, placebo controlled

14(7 treated,7 controls)


36(13 treatedon 42 days'course, 12treated on18 days'course,11 controls)



18(9 treated,9 control)



285(94 treatedventilatordependent,49 treatednon-ventilatodependent,94 controlsventilatordependent,48 controlsnon-ventilatodependent)

>L Alveolar-arterial oxygen gradientt Peak inspiratory pressureL Fractional inspiratory oxygen.1 Ventilatory rateT Weaning from respiratorT Dynamic pulmonary complianceT Weaning from respirator

T Total respiratory system complianceT Weaning from respirator

Treated v controlMean airwaypressure

I Ventilatory rate

42 days treated v18 days treatedand controls

T Weaning fromrespirator

T Weaning fromsupplementaloxygen

Improve neuro-developmentaloutcome

Tendency to Iduration ofhospitalisation

I Oxygen requirement (in first 10 daysof treatment

T Weaning from respiratorT Weaning from respirator

Oxygen requirement(1 week after steroid but did notshorten the time weaning to room air)

I Oxygen requirement (day 8 and17 after initiation of treatment)

T Weaning from respiratorNo improvement in lung function

T Weaning from respirator (tendencyto 4 duration of hospitalisation andI duration of supplemental oxygen)

)r

)r

Major problems with sepsis (subacutebacterial endocarditis, septicaemia,cytomegalovirus infection, pneumonia)

Hypertension, hyperglycaemia, sepsis,longer duration of hospitalisation intreated infants

Hypertension, hyperglycaemia,Gram negative sepsis, gastrointestinalhaemorrhage

Pulmonary airleaks, adrenocorticalsuppression

Hyperglycaemia, delay in weight gain

Hypertension, glucosuria, leukocytosis,bradycardia

Leukocytosis, weight loss, periventricularechodensity (± cyst formation)

Hypertension, glucosuria

Hypertension

Hypertension, hyperglycaemia

Increase=T, decrease=l.

EfficacyThere are 10 published prospective, ran-

domised, controlled trials concerning the use

of dexamethasone in BPD. Eight were fromNorth America and involved relatively smallnumbers of patients.'5-22 Two were from theUK,23 24 of which one was a large internat-ional multicentre study organised by theCollaborative Dexamethasone Trial Group inOxford.24 The findings of these trials are sum-

marised in table 1.There is now ample evidence to show that

dexamethasone improves lung function andcompliance in BPD infants within 72 hours ofcommencement16 17 and in turn facilitatesweaning from assisted ventilation and extuba-tion.'5-21 24 However, with respect to the dura-tion of supplement oxygen treatment andlength of hospital stay, the impact of dexa-methasone is less certain. Only one report

demonstrated a significant decrease in theduration of supplemental oxygen requirementwhen a prolonged 42 day course was given.'8Although the collaborative trial group also

reported a favourable trend,24 such a claimwas not substantiated by other randomisedtrials. Two reports showed a positive trend infavour of a shorter duration of hospitalisationin the steroid treated group, but the differencewas not statistically significant.'8 24

Perhaps the most important claim for bene-ficial long term effects of steroids in BPD wasmade by Cummings et al, who showed a sig-nificant improvement in the neurodevelop-mental outcome of 6 and 15 months in infantstreated with a 42 day course of dexametha-sone.18 As this is the only randomised studythat provides data on neurodevelopmentaloutcome and involved only a small number ofpatients, a definite conclusion cannot bedrawn as yet. A three year follow up assess-ment from the collaborative trial group willhopefully provide us with some answers in thenear future.None of the randomised trials performed to

date have managed to show an improvementin the mortality rate of steroid treated infantscompared with controls.15-24


http://adc.bmj.com

/A

rch Dis C


o.330 on 1 March 1993. D

ownloaded from

http://adc.bmj.com/

332

Treatment strategyUnfortunately, there are no universally agreedguidelines on the most appropriate indication,timing, dosage, and regimen of dexametha-sone to use as different clinical trials havetheir own experimental design and outcomemeasures. So far, only a general trend wasemerged.

(1) INDICATIONSDexamethasone is only of proved value ininfants who develop BPD and have difficultyin weaning from the ventilator.5-21 24 At pre-sent, there is not enough evidence to extendits recommendation to infants who are oxygendependent but are not ventilated. Although inthe latter group, a transient reduction in oxy-gen requirement has been demonstrated, theoverall duration of oxygen dependency wasnot shortened.23 24

Babies who show continuous and steadyimprovement or suffer a temporary setback inventilation, especially due to sepsis, must notbe hastily put on dexamethasone. It is alsoessential that every effort is taken to identifyall treatable causes that may prevent success-ful weaning before any baby is commenced onsteroid treatment. Factors like anaemia andmetabolic derangements must be corrected;haemodynamically significant patent ductusarteriosis must be treated; sepsis should beexcluded by C reactive protein and blood cul-tures; and plasma caffeine concentrationsmust be maintained well within the therapeu-tic range of facilitate diaphragmatic functionand to prevent apnoea. Only after exclusion ofall these treatable causes should dexametha-sone be considered as the next option.

(2) TIMINGIn most studies infants were started on steroidtreatment between the second and fourthweek of postnatal age.'6-22 Dexamethasonegiven during that period is proved to be ofvalue with regard to weaning from the ventila-tor. Interestingly, recent evidence suggeststhat dexamethasone given much earlier for thetreatment of respiratory distress syndrome onday 1 is also effective in improving pulmonarystatus and facilitating extubation.25 By mini-mising barotrauma and oxygen induced toxi-city in the initial period, steroids are potentiallyable to reduce the incidence of lung injury.Furthermore, we have recently demonstratedthat TNF-ax in endotracheal aspirate onlyappears in significant quantity after day 4 ofpostnatal age"I and perhaps this represents themost appropriate timing for starting steroidtreatment to suppress pulmonary inflamma-tion. Another advantage of starting dexam-ethasone treatment on day 4 is to allow theexclusion of congenital infection, which is attimes impossible to differentiate from respira-tory distress syndrome. A double blind ran-domised study is currently underway at ourinstitution to investigate the effect of day 4steroid on ventilation parameters and itsimpact on the incidence and severity of BPD.

(3) DOSAGE AND DURATION OF TREATMENTThe recommended starting dose of dexam-ethasone is in the region of 05-1 mg/kg/day,which is a relatively large dose to weightratio.'5-25 As a lot of undesirable effects ofcorticosteroids are dose dependent, such asprotein catabolism, adrenal axis suppressionetc, using a lower starting dose will certainlybe advantageous if it can produce the desiredanti-inflammatory effect. Unfortunately, notrial has so far made any direct comparison ofdifferent dosages. At present, the only recom-mendation is to start dexamethasone treat-ment at about 05-0-6 mg/kg/day but almostcertainly in future this dosage can be morefinely tuned according to the severity of lungdisease in individual babies.

Unlike the starting dose, the dose taperingschedule and duration of treatment differ sub-stantially between different centres.'5-24 Sofar, only Cummings et al have studied theeffectiveness of dexamethasone treatment inrelation to the length of treatment.'8 Theyconcluded in their small cohort of patientsthat the prolonged 42 day course was superiorto an 18 day course or placebo with respect topulmonary and neurodevelopmental outcomes.My experience with the use of dexametha-

sone in BPD indicated that prolonged depen-dency on the drug is not unusual. With ashort 10 day course of treatment similar to theone used by the multicentre study,24 a signifi-cant proportion of infants required to bereventilated after the cessation of the course oftreatment and a second course is frequentlyneeded at a later stage. Therefore we modifiedour course to a three week dose taperingregimen starting with an initial dose of0-6 mg/kg/day for the first week, reducingto 0 3 mg/kg/day for the second week, and0 15 mg/kg/day for the final week. Theincidence of extubation using this threeweek course is comparable with longercourses,'5 16 18 and our observation is alsosupported by studies that use a similar lengthof treatment.'9-21As few data have been published to indicate

the most appropriate and effective regimen, Ican only recommend a dose tapering treat-ment period of no less than three weeks inorder to prevent relapse and with a view toextend the duration to six weeks if furtherstudies are able to confirm the finding ofCummings et al. In principle, the most impor-tant rule governing the use of steroids is to usethe minimum dose and the shortest durationof treatment possible to achieve the desiredanti-inflammatory action.

(4) SECOND COURSE OF STEROIDTo date, no scientific trial has focused on theuse of a second course of dexamethasone inBPD. The following data is based on personalclinical experience and observation over a fouryear period, from 1987 to 1990. Sixty ninevery low birthweight infants were commencedon dexamethasone. Seven (10%) died whileon the first three week course of steroid. Ofthe 14 (20%) infants who received a secondthree week course of treatment, six were

Ng


http://adc.bmj.com

/A

rch Dis C


o.330 on 1 March 1993. D

ownloaded from

http://adc.bmj.com/

The effectiveness and side effects ofdexamethasone in preterm infants with bronchopulmonary dysplasia

ventilator dependent when the course startedand eight were in headbox oxygen greaterthan 40%. In the ventilated group, dexam-ethasone was obviously given as a rescue mea-sure after all forms of conventional treatmenthad failed. All six infants responded initiallywith a reduction in ventilation and oxygenrequirement but only four continued to showadequate improvement to permit extubation.The other two infants continued to deterio-rate and required increasing ventilatory sup-port. They eventually died because of BPDand respiratory failure. As to the unventilated,oxygen dependent group, dexamethasone wasgiven to minimise supplemental oxygenrequirement. A transient beneficial effect wasobserved in the first week but in most casesoxygen requirement increased as the dose ofsteroid was reduced. All infants in this groupsurvived.As this is a clinical observation, and

involves only a small number of infants, itdoes not allow a firm conclusion to be drawnwith regard to the use of a repeat course ofdexamethasone. However, a 67% success rateof extubation in the ventilated group isencouraging. I have no hesitation in trying asecond course of steroid when an infantremains stuck on the ventilator while all othertherapeutic options have already beenattempted.

The adverse effects of dexamethasoneAlthough the side effects of corticosteroidshave been extensively investigated in olderchildren and adults, they have not been com-

Table 2 Complications of corticosteroid treatment

1. Central nervous system

2. Ophthalmological

3. Cardiovascular

4. Respiratory5. Gastrointestinal

6. Renal

7. Musculoskeletal

8. Haematological and immunological

9. Metabolic and endocrine

10. Psychological

Pseudotumour cerebriGross motor developmental retardation*Abnormalities of electroencephalography*New ultrasonographically diagnosed echodensities*GlaucomaPosterior subcapsular cataractsReactivation of dormant herpes keratitisExacerbation of fungal and bacterial eye infectionsRetinopathy of prematurity*Hypertension*Cardiac hypertrophy*Sustained bradycardia*Pneumothrax*Peptic ulcer, haemorrhagePerforation*PancreatitisHepatomegaly with fatty infiltrationNephrocalcinosis*NephrolithiasisCalciuria*UricosuriaMyopathy and muscle wasting*Osteoporosis and poor bone mineralisationFractureAvascular bone necrosisIncrease total white cell count*Neutrophilia*MonocytopeniaLymphopeniaEosinopeniaErythropoiesisThrombocytosisPurpuraImmunosuppressionInfection*Somatic growth failure*Hyperglycaemia*Hypokalaemic alkalosisSodium retentionLipolysisProteolysis*Increased gluconeogenesisHypothalamo-pituitary-adrenal axis suppression*PsychosisIrritability*

*Side effects of dexamethasone reported in preterm infants.

prehensively and systematically examined inpreterm infants. Most randomised trials haveconcentrated on the efficacy of the drug withthe question of safety inadequately addressed.Despite the sporadic nature of reportingadverse effects, an impressively long list ofcomplications has already been accumulated(table 2). Those highlighted by the asteriskrepresent the side effects of dexamethasonethat have been reported in preterm infants.

(1) CENTRAL NERVOUS SYSTEM COMPLICATIONSThe most worrying complication is the recentanimal studies concerning central nervous sys-tem developments when corticosteroids areused during the perinatal period. Experimentswith newborn rodents indicate that corticos-teroids can cause a substantial decrease inbrain weight26 27 and cerebral and cerebellarDNA content.27 28 It can also cause delay incortical dentritic branching,26 and severelydisrupt the process of postnatal glial cell for-mation.27 These experiments clearly demon-strate that perinatal corticosteroid treatment iscapable of exerting a permanent and irre-versible effect on neural cell division anddifferentiation.

Despite many serious sequelae noted in fetaland newborn animal studies, there are rela-tively few data regarding the adverse effect ofcorticosteroids in human brain development.The effects of antenatal dexamethasone treat-ment on neurodevelopmental outcomes wereassessed by the collaborative group on antena-tal steroid treatment.29 No detectable differ-ences with regard to head circumference orneurodevelopmental abnormalities could befound at 9, 18, and 36 months that could beattributed to the effect of antenatal dexam-ethasone treatment. In fact, neurodevelopmentwas more strongly influenced by socio-economic background. Other investigatorswho used corticosteroid in the early postnatalperiod also found no difference with regard toneurodevelopmental outcome in the treatedinfants.30 31 One of the most recent reportssuggested that an extended 42 day course ofdexamethasone was superior to a shorter 18day course or placebo in terms of neuro-developmental outcome at 6 and 15 months.'8Published evidence on neurodevelopmentaloutcome in human newborn suggest that thereis no major risk associated with the use of cor-ticosteroid treatment in the neonatal period.

(2) OPHTHALMOLOGICAL COMPLICATIONSIn one animal study it was suggested that dex-amethasone treatment might increase the riskof retinopathy of prematurity by causing rapidchanges in oxygenation.32 Although twohuman studies support such claims,33 34 thereare at least four other trials that failed todemonstrate an increase in the incidence orseverity of this problem. 18 19 21 31

(3) CARDIOVASCULAR COMPLICATIONSHypertension secondary to sodium retention

333


http://adc.bmj.com

/A

rch Dis C


o.330 on 1 March 1993. D

ownloaded from

http://adc.bmj.com/

334

is a common problem in dexamethasonetreated infants and has been reported in themajority of studies.16-18 20-22 24 31 As differentcriteria for the definition of significant hyper-tension have been used, the exact incidenceremains unknown. It is reassuring, however,that in the majority of cases no treatment isrequired.A lesser known side effect of corticosteroids

on the heart is their potential to cause cardiacmuscle hypertrophy.35 36 Intraventricular sep-tum and left ventricle free wall thickness werereported to be increased after two weeks'treatment by dexamethasone.35 The hyper-trophy of these structures appeared to resolvespontaneously by six weeks when the steroidwas stopped.35 No differences were noted inthe ejection fraction.35 This phenomenonappears to be completely reversible.There was also a report suggesting that dex-

amethasone could induce sinus bradycardia inpreterm infants.20 This claim was, however,unsubstantiated by other workers.

(4) RESPIRATORY COMPLICATIONSIt seems almost contradictory to suggest thatdexamethasone given to ventilator dependentpremature infants for improving their pul-monary function can give rise to pulmonarycomplications. An increase in the incidence ofpulmonary air leaks soon after the start ofdexamethasone treatment was probablyrelated to a sudden improvement of pul-monary compliance, leading to overdistensionand rupture.'8 This complication is potentiallyavoidable by vigilant monitoring so that ascompliance improves, ventilation pressure canbe decreased appropriately.

(5) GASTROINTESTINAL COMPLICATIONSGastrointestinal complications that have beenassociated with corticosteroid treatmentinclude peptic ulcer leading to haemorrhageand perforation, pancreatitis, and fatty infil-tration of the liver.

I recently encountered three cases of gastro-duodenal perforation in preterm infantstreated with dexamethasone.37 The overallincidence is in the region of 2-3%. It wasalmost always preceded by haematemesis ormelaena. In all cases, prompt resuscitationwith abdominal paracentesis and early surgicalintervention were performed in order toensure a favourable outcome.

(6) RENAL COMPLICATIONSRenal complications of corticosteroids are rareand include nephrocalcinosis, nephrolithiasis,uricosuria, and calciuria.

Isolated cases of renal calcification havebeen identified in premature infants whoreceive dexamethasone and frusemide treat-ment38; both are known to be calciuric drugs.As none of the randomised trials lookedspecifically for its pattern of occurrence orincidence, the true frequency is not known.One can only presume that infants treated

with dexamethasone, especially in conjunctionwith frusemide, are at an increased risk ofdeveloping renal calcification. No data areavailable on their long term renal function.

(7) MUSCULOSKELETAL COMPLICATIONSSteroid myopathy, with or without musclewasting, affects principally the proximal mus-culature of the limbs. A substantial increase inplasma amino acid concentration39 and asimultaneous increase in urinary excretion of3-methylhistidine,40 soon after dexamethasonewas started, was most likely to be due toendogenous myofibrillar protein catabolism.Clinical myopathy in preterm neonates is dif-ficult, if not impossible, to detect and no suchcase has been reported in the literature so far.Although I have undoubtedly seen poor mus-cle bulk in babies who have received dexam-ethasone, I have not encountered any babywho developed frank muscle weakness orparalysis.

Osteopenia and bony fracture are some-times seen in preterm infants. It is difficult toevaluate the role of steroids in these events asother confounding factors such as immobilisa-tion secondary to the use of muscle relaxants,nutritional rickets, catabolism induced bystress or sepsis, and the use of calciuric drugsmay all, in part, be responsible for the poormineralisation of the bony matrix.

(8) HAEMATOLOGICAL AND IMMUNOLOGICALCOMPLICATIONSCorticosteroids are capable of substantiallyincreasing the total white cell count by caus-ing marked neutrophilia.41 Steroid inducedneutrophilia is due to a shift of neutrophilsfrom the marginating pool to the circulatingpool, an acceleration of the release of matureneutrophils from the bone marrow, and areduction of the egress of neutrophils fromblood to inflammatory sites. Other haemato-logical effects include an enhancement of ery-thropoiesis, thrombocytosis, monocytopenia,lymphopenia, and eosinopenia.The anti-inflammatory and immunosup-

pressive effects of corticosteroids have beenextensively investigated. The complexity ofthe immune system in association with widevariation in response depending on the animalspecies42 43 gives rise to much controversy inassessing the effect of steroids on the humanimmune function. The current view is thatthis category of drug has the capability todivert the trafficking, circulation, and avail-ability of inflammatory cell populations to theinflammatory site and possibly in therapeuticdose modify their functions in vivo.42 43Interestingly, the anti-inflammatory potencyof a particular steroid has little influence on itsimmunosuppressive effect.44 Dexamethasone,which is one of the most potent corticos-teroids, has comparatively little immunosup-pressive effect.44 Although earlier workersreported a high incidence of sepsis in preterminfants treated with dexamethasone,'5 45 morerecent trials suggested that the use of

Ng


http://adc.bmj.com

/A

rch Dis C


o.330 on 1 March 1993. D

ownloaded from

http://adc.bmj.com/

The effectiveness and side effects of dexamethasone in preterm infants with bronchopulmonary dysplasia

dexamethasone was not associated with a sig-nificant increase in the incidence or change inthe pattern of infection. 18 21 24 31 46

(9) METABOLIC AND ENDOCRINECOMPLICATIONSElectrolyte disturbances such as hypokalaemicalkalosis and sodium retention are due to themineralcorticoid effect of dexamethasone.

Hyperglycaemia is probably among themost commonly reported complications ofdexamethasone treatment in pretermbabies. 17 19-21 24 25 31 45 It is easily controllablewith insulin or when the steroid dosage andglucose load are reduced. So far, hyper-glycaemia has not caused any measurablemorbidity. 19A reduction in insulin activity results in an

increase in fatty acid output from adipose tis-sue and a surge in the total serum triglyceridesand cholesterol concentration. Abnormal fatdeposition with cushingoid features hasundoubtedly been spotted in babies whoreceive dexamethasone, although it has notbeen widely reported.

Corticosteroids are known to enhanceprotein catabolism. Babies treated with dex-amethasone showed a generalised and sub-stantial increase in amino acid concentra-tion,39 with a coincidental increase in urinaryexcretion of 3-methylhistidine, which in theabsence of meat protein in the diet sug-gested endogenous myofibrillar proteincatabolism.39 40 This phenomenon appearsto be dose related. Tyrosine and phenylala-nine, however, behaved differently and didnot demonstrate an increase in plasmaconcentration during steroid treatment. Therisk of hyperphenylalaninaemia toxicity istherefore minimal.Growth impairment has been reported in

preterm infants during dexamethasone treatmentfor BPD.18 19 23 25 Interruption of growth in thisearly period is potentially critical for it may havea permanent influence on the 'programming' ofsomatic growth in later life. Recent data suggestthat the effect of growth retardation is likely to betransient and normal growth pattern resumesafter dexamethasone is stopped.'9 So far, therehas not been any long term follow up assessmentto monitor somatic growth in these babies intotheir early childhood.

Perhaps the most feared side effect of corti-costeroid treatment is hypothalamo-pituitary-adrenal axis suppression. The adrenal glandsare undoubtedly suppressed during the treat-ment period but this effect appears to be tran-sient. Within a month after steroid treatmentis discontinued, most babies appear to haveregained their adrenal responsiveness.47 Noneof the babies show any signs of clinical or bio-chemical adrenal insufficiency.47

(10) PSYCHOLOGICAL DISTURBANCES'Irritable' behaviour was observed in a smallproportion of infants with BPD within thefirst few days of dexamethasone treatment.This pattern of behaviour promptly revertedto normal after a reduction in dosage.3'Although conceivably steroid related moodchanges and psychosis could have occurred inpreterm infants as in adults, interpretation ofsuch subjective signs must be viewed withcaution as other confounding factors such asrespiratory discomfort, pain, and gastritismight have contributed to their occurrence.

Balancing benefits and risksAs with any new treatment, the decision whetherto start an infant on dexamethasone rests entirelyon the perception of the benefits and risks of thetreatment (table 3). A clinician may put moreemphasis on one factor as opposed to the othersand finalise his decision one way, while anotherclinician may take an opposite view. So far, thereare no absolute guidelines on which babies totreat or when to treat them. Furthermore, theoptimum dosage of dexamethasone, the mostdesirable schedule, and duration of treatmnenthave not been fully verified. Hence, the aboverecommendations are, to a certain degree,dependant upon subjective impression and per-sonal experience of the drug.

In my opinion, dexamethasone has a definiterole in the management of BPD by shorteningthe duration of mechanical ventilation. The verylimited number of intensive care cots can, there-fore, be used to care for more sick babies. Thistranslates into a substantial financial saving forthe health service per baby treated. The elimina-tion of the ventilator also means greater freedomfor parents to handle their baby, thus promotingclose parent-child bonding. Furthermore, bymaking significant therapeutic progress, themorale of the caregivers working under thestressful conditions of the neonatal unit is boost-ed. Above all, the discontinuation of positivepressure ventilation removes the risk of ventila-tion related complications such as pulmonary airleaks, pneumonia, and further barotrauma,which acts as a continuing stimulus for theinflammatory cascade.With regard to the negative side of steroid

treatment, none of the randomised trials per-formed to date has managed to show animprovement in the mortality rate of steroidtreated infants compared with controls. Someof the side effects that have been reported inpreterm infants are undoubtedly over exagger-ated. In particular, the early reports on theincidence of severe infections'5 45 have nowbeen challenged by later trials that report noappreciable increase in frequency nor anychange in the pattern of occurrence. 18 21 24 31 46The total white cell and neutrophil count are

Table 3 The benefits and risks of dexamethasone treatment

Benefits v Risks1 Shortening the duration of assisted ventilation 1 Hypothalamo-pituitary-adrenal axis suppression2 ?Shortening the duration of supplemental oxygen treatment 2 Enhanced catabolism, hyperaminoacidaemia, and somatic

growth failure3 ? Shortening the duration of hospitalisation 3 Gastroduodenal perforation4 ? Favourable neurodevelopmental outcome 4 Hypertension

5 Hyperglycaemia

335


http://adc.bmj.com

/A

rch Dis C


o.330 on 1 March 1993. D

ownloaded from

http://adc.bmj.com/

336 Ng

affected by dexamethasone and are not reliablepredictors of infection but plasma C reactiveprotein concentration is unaffected andremains a good indicator for monitoring neona-tal sepsis.

In the past few years, even the less commonand infrequent side effects of steroids such asgastrointestinal perforation are increasinglybeing recognised.37 Therefore, neonatal clini-cians must be equipped with the knowledgeand must be prepared to encounter even thelesser known and rare complications of dex-amethasone treatment.

Perhaps, in future, the use of more specifictreatments such as monoclonal antibody toTNF-x and inhaled corticosteroids willemerge to be better treatment for BPD, withfewer side effects. Until more is known aboutthe safety and beneficial effects of systemicsteroids in this group of patients, in particulartheir influence on long term neurodevelop-mental and pulmonary outcome, indiscrimi-nate use should be prohibited. I believe theprudent course is to restrict their use to venti-lator dependent infants with moderate tosevere BPD.

1 Northway Jr. WH. Bronchopulmonary dysplasia: then andnow. Arch Dis Child 1990;65:1076-81.

2 Frank L, Lewis PL, Sosenko IRS. Dexamethasone stimula-tion of fetal rat lung antioxidant enzyme activity in paral-lel with surfactant stimulation. Pediatrics 1985;75:569-74.

3 De Lemos RA, Shermeta DW, Knelson JH, Kotas R,Avery AE. Acceleration of appearance of pulmonary sur-factant in the fetal lamb by administration of corticos-teroids. Am Rev Respir Dis 1970;102:459.

4 Barnes P, Jacobs M, Roberts J. Glucocorticoids preferen-tially increase fetal alveolar beta-adrenoceptors: auto-radiographic evidence. Clin Res 1983;31:70A.

5 Wilson JW. Treatment or prevention of pulmonary cellulardamage with pharmacologic doses of corticosteroid. SurgGynecol Obstet 1972;134:675-81.

6 Skubitz KM, Gaddock PR, Hammerschmidt DE, AugustJT. Corticosteroids block binding of chemotactic peptideto its receptor on granulocytes can cause disaggregationof granulocyte aggregates in vitro. J Clin Invest 1981;68:13-20.

7 Kusajima K, Wax SD, Webb WR. Effects of methylpred-nisolone on pulmonary microcirculation. Surg GynecolObstet 1974;139:1-5.

8 Gryglewski RJ, Panczenko B, Korbut R, Grodzinska L,Ocepkiewicz A. Corticosteroids inhibit prostaglandinrelease from profused lungs of sensitised guinea pigs.Prostaglandins 1975;10:343-5.

9 Hong SCL, Levine L. Inhibition of arachadonic acidrelease from cells as the biochemical action of anti-inflammatory corticosteroids. Proc Natl Acad Sci USA1976;73: 1730-4.

10 Ariagno RL, Sweeney TE, Northway WH. Dexamethasonedecreases lung water without affecting lung compliance ininfant mice raised in 80% oxygen environment. Am RevRespirDis 1986;133:A108.

11 Murch SH, MacDonald TT?, Wood CBS, Costeloe KL.Tumour necrosis factor in the bronchoalveolar secretionsof infants with the respiratory distress syndrome and theeffect of dexamethasone treatment. Thorax 1992;47:44-7.

12 Sahebjami H, Jacob G, Massaro D. Influence of corticos-teroid on recovery from oxygen toxicity. Am Rev RespirDis 1974;110:566-71.

13 Gladstone IM, Ehrenkranz RA, Jacobs HC. Pulmonaryfunction tests and fluid balance in neonates with chroniclung disease during dexamethasone treatment. Pediatrics1989;84: 1072-6.

14 Clement A, Chadelat K, Sardet A, Grimfeld A, TournierG. Alveolar macrophage status in bronchopulmonarydysplasia. Pediatr Res 1988;23:470-3.

15 Mammel MC, Green TP, Johnson DE, Thompson TR.Controlled trial of dexamethasone therapy in infants withbronchopulmonary dysplasia. Lancet 1983;i: 1356-8.

16 Avery GB, Fletcher AB, Kaplan M, Brudno DS.Controlled trial of dexamethasone in respirator-depen-dent infants with bronchopulmonary dysplasia. Pediatrics1985;75:106-1 1.

17 Ariagno RL, Sweeney TJ, Baldwin RB, Inguillo D, MartinD. Dexamethasone effects on lung function in 3 week oldventilator dependent preterm infants. Am Rev Respir Dis1987;135:A125.

18 Cummings JJ, D'Eugenio DB, Gross SJ. A controlled trial

of dexamethasone in preterm infants at high risk forbronchopulmonary dysplasia. N Engl Jf Med 1989;320:1505-10.

19 Harkavy KL, Scanlon JW, Chowdhry PK, Grylack U.Dexamethasone therapy for chronic lung disease in venti-lator dependent infants: a controlled trial. Jf Pediatr1989;1 15:979-83.

20 Ohlsson A, Calvert S, Hosking M, Shennan A.Randomised controlled trial of dexamethasone treatmentin very low birthweight infants with ventilatory dependentchronic lung disease. Pediatr Res 1989;25:225A.

21 Kazzi NJ, Brans YW, Poland RL. Dexamethasone effectson the hospital course of infants with bronchopulmonarydysplasia who are dependent on artificial ventilation.Pediatrics 1990;86:722-7.

22 Singhal N, McMillan DD, Stephure DK. Effects ofACTHand dexamethasone (DXM) on infants with bronchopul-monary dysplasia (BPD). Pediatr Res 1990;27:226A.

23 Noble-Jamieson CM, Reger R, Silverman M.Dexamethasone in neonatal chronic lung disease: pul-monary effects and intracranial complicatios. Eur JPediatr 1989;148:365-7.

24 Collaborative Dexamethasone Trial Group. Dexamethasonetherapy in neonatal chronic lung disease: an internationalplacebo-controlled trial. Pediatrics 1991 ;88:421-7.

25 Yeh TF, Torre JA, Rastogi A, Anyebuno MA, Pildes RS.Early postnasal dexamethasone therapy in prematureinfants with severe respiratory distress syndrome: a dou-ble blind controlled study. J. Pediatr 1990; 117: 273-82.

26 Shapiro S. Some physiological, biochemical andbehavioural consequences of neonatal hormone adminis-tration: cortisol and thyroxin. Gen Comp Endocrinol1968;10:214-28.

27 Howard E, Benjamins JA. DNA, ganglioside and sulfatidein brains of rats given corticosterone in infancy, with anestimate of cell loss during development. Brain Res1975;92:73-87.

28 Weichsel ME. The therapeutic use of glucocorticoid hor-mones in the perinatal period: potential neurologic haz-ards. Ann Neurol 1977;2:364-6.

29 Collaborative Group on Antenatal Steroid Therapy. Effectsof antenatal dexamethasone administration in the infant:long term follow up. J Pediatr 1984; 104:259-67.

30 Gunn T, Reece ER, Metrakos K, Cole E. Depressed Tcells following neonatal steroid treatment. Pediatrics198 1;67:61-7.

31 Ferrara TB, Couser RJ, Hoekstra RE. Side effects andlong-term follow-up of corticosteroid therapy in very lowbirthweight infants with bronchopulmonary dysplasia.J Perinatol 1989;10: 137-42.

32 Uemura Y, Akiya S, Aihara Y, Tsuchiya S. Steroidinduced retinopathy. Proceedings of the conference onretinopathy of prematurity. Columbus, OH: RossLaboratories, 1981;1:162-81.

33 Mammel MC, Fiterman C, Coleman M, Boros SJ. Shortterm dexamethasone therapy for bronchopulmonary dys-plasia: acute effects and 1 year follow-up. Dev PharmacolTher 1987;10: 1-11.

34 Asztalos EV, Ohlsson A, Shennan AT, Smith J. Is dexam-ethasone treatment (Dxt) a risk factor for retinopathy ofprematurity (ROP)? Results of a case control study.PediatrRes 1991;29:201A.

35 Werner JC, Sicard RE, Hansen TWR, Solomon E, CowettRM, Oh W. Untoward cardiac effects of dexamethasonetherapy for bronchopulmonary dysplasia (BPD). PediatrRes 1991;29:249A.

36 Israel BA, Sherman FS, Guthie RD. Cardiac hypertrophyassociated with the use of dexamethasone in prematureinfants. Pediatr Res 199 1;29:218A.

37 Ng PC, Brownlee KG, Dear PRF. Gastroduodenal perfo-ration in preterm babies treated with dexamethasone forbronchopulmonary dysplasia. Arch Dis Child 1991 ;66:1164-6.

38 Kamitsuka MD, Peloguin D. Renal calcification after dex-amethasone in infants with bronchopulmonary dysplasia.Lancet 1991;337:626.

39 Ng PC, Brownlee KG, Kelly EJ, Henderson MH, Smith M,Dear PRF. Changes in the plasma aminogram of parenter-ally fed infants treated with dexamethasone for broncho-pulmonary dysplasia. Arch Dis Child 1992;67:1193-50.

40 Brownlee KG, Ng PC, Henderson MH, Smith M, GreenJH, Dear PRF. The catabolic effect of dexamethasone inthe preterm baby. Arch Dis Child 1992;67:1-4.

41 Evans JCH, Ng PC, Dear PRF. Dexamethasone in bron-chopulmonary dysplasia. Arch Dis Child 1988;63:221.

42 Fauci AS, Dale DC, Balow JE. Glucocorticosteroidtherapy: mechanisms of action and clinical considera-tions. Ann Intern Med 1976;84:304-15.

43 Kaeberle ML. Effect of steroids on immunologic function.Prog Clin Biol Res 1984;161:325-35.

44 Langhoff E, Ladefoged J. Relative immunosuppressivepotency of various corticosteroids measured in vitro. EurJClin Pharmacol 1983;25:459-62.

45 Pomerance J, Puri AR. Treatment of severe neonatal pul-monary dysplasia with dexamethasone. Pediatr Res 1982;16:359A.

46 Ng PC, Thomson M, Dear PRF. Dexamethasone andinfection in preterm babies: a controlled study. Arch DisChild 1990,65:54-8.

47 Ng PC, Blackburn ME, Brownlee KG, Buckler JMH, DearPRF. Adrenal response in very low birthweight babiesafter dexamethasone treatment for bronchopulmonarydysplasia. Arch Dis Child 1989;64:1721-6.


http://adc.bmj.com

/A

rch Dis C


o.330 on 1 March 1993. D

ownloaded from

http://adc.bmj.com/

Documents

PERSONALPRACTICE - adc.bmj.com · The recommended starting dose of dexam-ethasone is in the region of05-1 mg/kg/day, which is a relatively large dose to weight ratio.'5-25 As a lot