lactate dehydrogenase predicts hypoxic ischaemic encephalopathy in newborn infants: a preliminary...

REGULAR ARTICLE

Lactate dehydrogenase predicts hypoxic ischaemic encephalopathy innewborn infants: a preliminary studyMathias Karlsson ([email protected])1,2, Eva Wiberg-Itzel1, Ela Chakkarapani4, Mats Blennow3, Birger Winbladh1, Marianne Thoresen4

1.Department of Clinical Science and Education, Karolinska Institute at Stockholm Soder Hospital, Stockholm, Sweden2.Department of Clinical Research, Karlstad, Sweden3.Institution CLINTEC Huddinge, Karolinska Institutet, Stockholm, Sweden4.Clinical Science at South Bristol (Child Health), University of Bristol, Bristol, UK

KeywordsHypoxic ischaemic encephalopathy, Liver enzymes,Newborn, Prediction

CorrespondenceMathias Karlsson, Centre for Clinical Research,Centralsjukhuset i Karlstad, Hus 73, S-651 85Karlstad, Sweden.Tel: +46 (0) 54 615000 |Fax: +46 54 61 42 98 |Email: [email protected]

Received14 December 2009; revised 22 February 2010;accepted 11 March 2010.

DOI:10.1111/j.1651-2227.2010.01802.x

ABSTRACTBackground: Enzyme leakage as a result of hypoxia-ischaemia induced cell damage

in affected organs is seen together with hypoxic ischaemic encephalopathy (HIE) after peri-

natal asphyxia.Aim: To investigate whether plasma lactate dehydrogenase [LDH], alanine amino-

transferase [ALT] and aspartate aminotransferase [AST] during the first 12 h after birth pre-

dict HIE and adverse neurodevelopment outcome in newborn term infants with intra-

partum signs of foetal distress.Methods: Enzymes were measured within 12 h post partum in newborn infants

with differing degree of HIE (n = 41) and in infants with signs of foetal distress during birth

(n = 205) without HIE (non-HIE group). All infants were randomized into two groups. One

group (n = 123) was used for calculation of cut off limits for the enzymes studied and the

other group (n = 123) was used for calculation of the predictive value of the enzymes for

detection of HIE.Results: Using ROC curves, a cut off level of 1049 U ⁄ L for [LDH] was the best pre-

dictor of HIE (sensitivity 100% and specificity 97%) but also for long term outcome after

HIE.

Conclusion: [LDH] is a good predictor of HIE during the first 12 h after birth. This result is ofclinical interest offering a potential inexpensive and safe prognostic marker in newborn infantswith perinatal asphyxia.

INTRODUCTIONApproximately four million newborn children die annually.Almost one quarter (23%) of these deaths are caused byperinatal asphyxia (1), most of them in developing coun-tries. In addition, perinatal asphyxia causes an even greaternumber of children with neurological sequels. The clinicalsigns following severe perinatal asphyxia, traditionallycalled hypoxic ischaemic encephalopathy (HIE), has a grad-ing of I (mild), II (moderate) or III (severe) (2,3). Infantswith moderate to severe HIE have a risk of death or severehandicap of 24% and 78% respectively (2). In addition,infants with moderate encephalopathy but without majorneurological damage, perform worse on cognitive testswhen tested in early school-age (3). Also, even in the

absence of clinical signs of encephalopathy, infants whowere resuscitated after birth have an increased risk of lowIQ score at 8 years of age (4). While, in most cases, the clini-cal symptoms of moderate and severe HIE are obvious, theneonatal symptoms of mild cerebral injury are more subtlemaking an early precise diagnosis more difficult. Early pre-diction of HIE is needed for selection of newborn infantswho could benefit from neuroprotective treatment likehypothermia.

One major endogenous defence mechanism of a foetusexposed to hypoxia-ischaemia (HI) is based on the ability tocentralize cardiac output to prioritized organs like thebrain, heart and adrenals at the expense of less importantorgans like the liver, lungs, skin and muscles. Multi-organdysfunction (MOD) is a natural consequence of this defencemechanism (5). Injured cells leak intra-cellular enzymes,some of which are easy to measure in plasma e.g. lactatedehydrogenase (LDH), alanine aminotransferase (ALT) andaspartate aminotransferase (AST). Increased levels of theseenzymes have been reported after neonatal asphyxia (6–8).These enzymes originate from different organs with ALT

Abbreviations

ALT, alanine aminotransferase; AST, aspartate aminotransferase;HIE, hypoxic ischaemic encephalopathy; IQR, inter quartilerange; LDH, lactate dehydrogenase; ROC, receiver operatingcharacteristic.

Acta Pædiatrica ISSN 0803–5253

ª2010 The Author(s)/Journal Compilation ª2010 Foundation Acta Pædiatrica/Acta Pædiatrica 2010 99, pp. 1139–1144 1139

being the most liver specific enzyme followed by AST that isalso present in e.g. the myocardium, muscles and erythro-cytes while LDH is present in most body tissues (9).

The substantial rise in LDH, ALT and AST after organdamage (6,10–12) and asphyxia directly after birth (cordblood), resulting from organ damage following asphyxia(7,11) and their different rates of disappearance fromplasma (5–36 h) (13), make these enzymes potential predic-tors of the severity of the hypoxic-ischaemic insult in theperinatal period. Therefore, we aimed to investigatewhether plasma [LDH], [ALT] and [AST] levels during thefirst 12 h after birth predict the development of HIE andlong term adverse neurodevelopment outcome in term new-born infants with intra-partum signs of foetal distress.

PATIENTS AND METHODSPatientsThe study population was a mixed population of infants pre-senting with either intra-partum foetal distress with perina-tal asphyxia or intra-partum foetal distress withoutperinatal asphyxia with a gestational age of more than36 weeks (n = 301). Intra-partum foetal distress wasdefined as a non-reassuring foetal heart rate trace (tachycar-dia, bradycardia, decreased variability, loss of accelerations,late or variable decelerations) and ⁄ or blood or meconiumstained amniotic fluid. These signs were considered an indi-cation for operative interventions (caesarean sections, ven-touse and forceps deliveries) and ⁄ or foetal scalp blood-sampling. A series of enzymes in routine sampling from 24normothermic asphyxiated infants with HIE were collectedat St Michaels Hospital, Bristol, UK between the year 2004and 2008 and infants with signs of intra-partum foetal dis-tress were prospectively included at the delivery departmentat Sodersjukhuset, Stockholm, Sweden between 2004 and2005 (n = 256). Also included were data from infants at theneonatal intensive care unit (NICU) at Huddinge UniversityHospital and Sodersjukhuset, Stockholm, Sweden (n = 21)between the year 2000 and 2001 in a previously performedand published (6) study from our group. The profile forinclusion and exclusion in this study is presented as a flowchart in Figure 1 and described in detail below.

Asphyxia (n = 39) was defined as the presence of: (1)Apgar score £5 at 10 min after birth or, (2) umbilical arterialpH (pHa) <7.00 or base deficit ‡16 mmol ⁄ L or (3) resuscita-tion with more than 10 min of positive pressure ventilationbefore stable spontaneous respiration.

The HIE diagnosis (n = 41) was based on abnormal neu-rology regarding consciousness, muscle tone, poor suck,need for assisted ventilation and seizures (14). Seizures(n = 32) were defined clinically or with amplitude-inte-grated EEG (aEEG, CFM; Olympic Medical, Seattle, WA,USA or Nervus, Cephalon, Norresundby, Denmark). AllHIE infants displayed signs of intra-partum foetal distressand were diagnosed with mild (n = 8), moderate (n = 16)or severe (n = 17) hypoxic ischaemic encephalopathy(HIE). In total 32 of the 41 infants with HIE (78%) fulfilledthe criteria for intra- partum asphyxia (as described above).

The infants with signs of intra-partum foetal distress bornat Sodersjukhuset were included in the non-HIE group ifthey did not develop symptoms of HIE (n = 260). However,no biochemical data were available from 55 infants becauseof haemolysis, sample errors or missing data. In the remain-ing 205 infants in the non-HIE group, 172 infants were dis-charged as healthy while 33 infants had the followingdiagnoses: respiratory distress, major congenital abnormali-ties, anaemia, subgaleal haemorrhage, Rh isoimmunization,jaundice requiring phototherapy, birth weights outside themean ±2 SD, hypoglycaemia, congenital hydronephrosis,metabolic acidosis, low Apgar score and need for positivepressure ventilation at birth.

At 18 months of age surviving infants diagnosed withHIE had a neurological examination, according to clinicalroutine, by certified staff. The non-HIE group was followedup through medical records. Based on these examinationsthe infants were divided into a ‘adverse neurodevelopmentoutcome group’ (death ⁄ severe disability) and a ‘favourableneurodevelopment outcome group’ (normal ⁄ mild disabil-ity).

MethodsThe enzymes were determined with standard procedures bythe accredited routine clinical chemistry laboratories at theparticipating hospitals. Briefly, blood for analysis of LDH,ALT and AST (400 lL) was collected in lithium heparintubes. The blood was centrifuged and analysed according tothe local hospital routine. In the UK the analyses were per-formed in plasma and presented as U ⁄ L whereas in Swedenthe analyses were performed in serum and presented as

HIE- group(n = 41)

Including infants with:

Mild, moderate or severe HIE.

Randomized group 1 for cut off (R-ROC)

HIE (n = 22) Non-HIE (n = 101)Total (n = 123)

Randomized group 2 for prediction (R-Pred)

HIE (n = 19) Non-HIE (n = 104)Total (n = 123)

Non-HIE- group (n = 205)

Including infants with:

Intra partum foetal distress with or without signs of

asphyxia in absence of HIE

Total (n = 301) GA ≥ 36weeksFindings of intra partum foetal distress with or without hypoxic ischaemic encephalopathy (HIE) Data not available (n = 21)

Haemolysis or sample error (n = 34)

246 infants randomized to two groups for model calculation

Figure 1 Profile for inclusion and exclusion in this study. HIE, hypoxic

ischaemic encephalopathy.

LDH predicts HIE Karlsson et al.

1140 ª2010 The Author(s)/Journal Compilation ª2010 Foundation Acta Pædiatrica/Acta Pædiatrica 2010 99, pp. 1139–1144

lkat ⁄ L. For statistical calculations, the latter were con-verted to Units ⁄ Litres (U ⁄ L) by multiplying the result inlkat ⁄ L by 60.

During NICU-treatment, a blood sample for analysis ofLDH, ALT and AST was collected together with routinebiochemistry. In case of repeated analysis these data wererecorded as well. In the control group, blood for LDH, ALTand AST analysis was collected from cord blood immedi-ately after birth.

The study protocol was approved by the Regional EthicalCommittees. In the group of infants that were previouslystudied and reported (6), we had not applied for ethicalapproval for examination at 18 months. Therefore, no datafrom this group of infants (n = 21) were analysed orreported after the first week of life in this study.

Data analysisLinear regression analysis was performed on all 246 infantswith LDH, ALT and AST as the dependent variables anddegree of HIE, Apgar score at 5 min, pHa and base deficitin cord blood (umbilical artery), maternal age, parity, modeof delivery, gender of the infant, birth weight and gestationalage as independent variables. A significance value of 0.05was used when allowing independent factors to enter theregression to find potential relationships between theenzymes and the independent variables.

The enzyme concentrations (all with skewed distribution)were compared between the groups with the Mann–Whit-ney U-test which was performed with Bonferroni’s correc-tion for multiple comparisons. Nominal data wereevaluated using Chi-squared test or Fishers exact test.

The 246 included infants were then randomized into twohalves (15). The first randomized half (R-ROC) was used forcalculation of best cut off values for LDH, ALT, AST andpH for detection of HIE and adverse neurodevelopmentoutcome obtained from receiver operating characteristiccurves (ROC). These cut off values were then applied on thesecond randomized half (R-Pred) for calculation of the pre-dictive values.

RESULTSA total of 246 infants were included in the study with 123infants randomized to the group used for calculation of cutoff limits (R-ROC) and 123 infants to the group used for cal-culation (R-Pred) of the predictive values of the enzymesstudied. Background clinical variables did not significantlydiffer between the two groups.

There was a significant difference in 5 min Apgar scoreand values for pH and base excess (BE) in umbilical arterydirectly after birth between the HIE group and the non-HIEgroup (p < 0.0001, Table 1). Apgar score, pHa or BE didnot differ between asphyxiated infants with no or mild HIEvs infants with moderate to severe HIE.

Seizures were diagnosed and treated in 32 infants and themedian (IQR) time between birth and first seizure was 2.5 h(1.0–6.0). Enzyme analysis was performed before or simul-taneously with the onset of seizures in 13 of these infants.

Seven infants, all with severe HIE, died within 10 days afterbirth.

A significant relationship between the independent vari-ables Apgar score at 5 min (B = )373 (CI )580, )165)) anddegree of HIE (B = 809 (CI 187, 1431)) and the dependentvariable [LDH] was seen in the linear regression analysis(Table 2a). A significant relationship between Apgar scoreat 5 min (B = )38 (CI )48–)27)) and [ALT] and degree ofHIE (B = 372 (CI 285–458) and [AST] was also seen (Table

Table 1 Obstetric and neonatal data for included newborn infants (n = 246)

divided into non-HIE group (n = 205) and infants with mild to severe HIE

(n = 41)

Non-HIE HIE p value

GA (weeks) 39.9 (±1.5) 40.1 (±1.3) 0.461

Birth weight (g) 3623 (±523) 3600 (±676) 0.737

Apgar 5¢<7 6 (3%) 37 (90%)* <0.001

pHa in cord 7.22 (±0.08) 6.99 (±0.21)* <0.001

BEa in cord )7.00 (±3.63) )16.04 (±6.76)* <0.001

Ventilation >10 min 2 (1%) 30 (79%)* <0.001

Females 89 (44%) 16 (43%) 0.927

Maternal age (years) 31.8 (±4.6) 30.8 (±6.5) 0.481

Nullipara 111 (58%) 11 (48%) 0.749

Operative interventions 116 (57%) 17 (46%) 0.313

Foetal indication 71 (64%) 11 (79%) 0.528

FSB acidosis (n = 169) 16 (10%) 1 (50%) 0.192

Foetal indication = Numbers of operative interventions (caesarean sections or

ventouse) on the indication ‘foetal distress’; FSB acidosis = Foetal scalp blood

lactate >4.8 mmol ⁄ L and ⁄ or pH <7.21; GA = Gestational age; pHa and

BEa = First pH and base excess in umbilical artery after birth.

Scalp blood test was performed in 169 cases.

*p < 0.05.

Table 2 Independent variables with a significant relationship to the dependent

variable [LDH] (a), [ALT] (b) and [AST] (c); from a linear stepwise regression

model

Dependantvariable

(a) Lactate dehydrogenase [LDH] <12 h post partum(R = 0.63, R2 = 0.39)

Independentvariables B SE t Sig CI Lower CI Upper

Constant 4180 1028 4.06 0.000 2152 6208

Apgar 5¢ )373 105 )2.58 0.001 )580 )165

HIE 809 315 2.09 0.011 187 1431

Dependantvariable

(b) Alanine aminotransferase [ALT] <12 h post partum(R = 0.44, R2 = 0.19)

Constant 4180 1028 4.06 0.000 2152 6208

Apgar 5¢ )38 5.5 )6.87 0.000 )48 )27

Dependantvariable

(c) Aspartate aminotransferase [AST] <12 h post partum(R = 0.521, R2 = 0.272)

Constant 4180 1028 4.06 0.000 2152 6208

HIE 372 44 8.49 0.000 285 458

The calculation was performed in all 246 infants included in the study. Exclu-

ded independent variables were: pH and base excess in umbilical artery,

maternal age, gravida, para and gender of the infant, birth weight and gesta-

tional age.

Karlsson et al. LDH predicts HIE


2b,c). No other clinical variable showed any significant rela-tionship with the enzymes.

The LDH, ALT and AST activities were significantlyhigher (p < 0.0001) in the HIE group compared with thenon-HIE group (Table 3). Also LDH, ALT and AST dis-criminated significantly between infants with moderate tosevere HIE compared with asphyxiated infants with noneor mild HIE (Table 3). The median time (IQR) betweenbirth and blood-sampling was 0 h in the non-HIE groupand 6.0 (3.8–8.6)h in the HIE group. In infants withrepeated analysis (n = 34) 25 infants increased in LDHactivity between the first and second sample (74%) and nineinfants (26%) showed the highest LDH activity in the firstsample. The median (IQR) time between birth and the sec-ond sample was 26.8 (14.5–36)h.

The best [LDH], [ALT] and [AST] cut off values, calcu-lated in the R-ROC group (n = 123), for prediction of HIEare given in Table 4 together with the predictive values thatwere assessed in the R-Pred group (n = 123). A secondanalysis of the predictive value of [LDH], [ALT] [AST] forprediction of HIE in the group of patients where enzymevalues were obtained simultaneously or before onset of sei-zure was calculated (Table 4). In both cases LDH offeredhigh sensitivity, specificity and predictive value for predic-tion of mild to severe HIE. Thirdly, the follow up data fromexamination at18 months showed poor outcome in eightinfants. The most suitable marker for prediction of adverseneurodevelopment outcome was LDH even if the positivepredictive value was lower compared with prediction ofHIE.

DISCUSSIONThe main finding of this study was that [LDH] predictedHIE with high sensitivity (100%) and specificity (98%). Thiswas also the case in asphyxiated infants where enzyme anal-ysis was performed before the onset of seizures. Also [LDH]was predictive for long term outcome. In this preliminarystudy, data were combined from a Swedish unselectedcohort of term infants with signs of foetal distress (the con-trol group) and a combined cohort of infants from UK andSweden investigated after birth as a result of the presence offoetal distress followed by signs of perinatal asphyxia. Com-bining infants from different centres is necessary because of

the low incidence of HIE in single centres. It is important todocument similarities or differences in the underlying popu-lations. There are several factors supporting the validity ofour findings: The clinical criteria for evaluating the infants(Apgar, HIE) and the clinical laboratory methods (LDH,ALT, AST and pH) are well established. Furthermore, thechanges in the enzyme concentrations are large, makingminor differences in laboratory standards and differenttimes for sampling less important (see below). The predic-tive value of [LDH], [ALT] and [AST] in term newbornswith asphyxia is not yet fully investigated. In a study fromLackmann and Tollner (11) 10 of 49 full term infants devel-oped HIE and the sensitivity and specificity for using AST

Table 3 Median LDH, ALT and AST serum activity with inter quartile range in newborn (<12 h) infants with mild, moderate or severe HIE (HIE group, n = 41) com-

pared with newborn infants (non-HIE group, n = 205) showing signs of foetal distress during birth but without any clinical signs of HIE

HIE group (n = 41) Non-HIE (n = 205) Asphyxia, none to mild HIE (n = 11) Asphyxia, moderate to severe HIE (n = 28)

LDH 2076* (1461–3925)

p < 0.0001

575 (455–702) 1012* (850–1572)

p = 0.001

2693 (1774–4782)

ALT 49* (23–141)

p < 0.0001

10 (7–14) 19.8* (9.6–46.0)

p = 0.007

70.5 (30.0–275.0)

AST 136* (91–413)

p < 0.0001

29 (25–35) 79.1* (27.8–139.4)

p = 0.004

272.9 (119–659)

Median LDH, ALT and AST serum activity with inter quartile range in infants with asphyxia (n = 39) and none to mild HIE (n = 11) compared with infants with mod-

erate to severe HIE (n = 28).

* p<0.05 non-HIE vs. HIE group.

Table 4 Best cut off values for enzymes obtained from a receiver operating

characteristic curve (ROC) presented together with area under the curve (AUC)

and routine arterial pH

LDH ALT AST pHa

All included (HIE n = 19)

AUC 0.98 0.96 0.97

Cut off 1049 U ⁄ L 15 U ⁄ L 68 U ⁄ L 7.05

Sensitivity 100% 95% 85% 60%

Specificity 98% 90% 98% 94%

PV+ 90% 64% 85% 69%

PV) 100% 99% 98% 94%

Pre seizure (Seizures n = 8)

AUC 0.98 0.96 0.97

Cut off 1049 U ⁄ L 16 U ⁄ L 59 U ⁄ L 7.05

Sensitivity 100% 90% 75% 63%


PV+ 83% 47% 60% 71%

PV) 100% 99% 99% 94%

Long term (sequele n = 4)

AUC 0.97 0.91 0.96

Cut off 1176 U ⁄ L 15 U ⁄ L 68 U ⁄ L 7.05

Sensitivity 100% 100% 67% 50%


PV+ 29% 19% 33% 22%

PV) 100% 100% 99% 97%

The cut off values were used for calculation of prediction of hypoxic ischaemic

encephalopathy (HIE) or adverse neurodevelopment outcome (severe sequel

or death at 18 months) in 123 newborn infants (R-pred). ‘Pre seizure’ refer to

the group of infants where enzyme analysis were performed before clinical

examination and assessment of degree of HIE.



or LDH, 24 h post partum, for prediction of HIE were 90%and 71% respectively. The substantially higher predictivevalues seen in our study compared with the Lackmannstudy could be explained by (1) the lower number of infantswith HIE in their study (n = 10) and (2) their later timepoint for blood collection (24 h post partum). Enzymeshave also been studied as a retrospective diagnostic markerof HIE (16).

One limitation in our study is that blood for enzyme anal-ysis was collected later in the asphyxiated infants (0.5–12 hpost partum) than in the control group (0 h). The enzymeactivity in serum increases approximately 40% under nor-mal conditions during the first 12 h after birth in healthynewborns (17). Therefore our study might slightly overesti-mate the enzyme differences between the control and theasphyxia groups. We therefore calculated a new ROC curveincluding only infants with enzyme data before 3 h of ageinstead of 12 h in the moderate to severe HIE group(n = 9). The area (95% Confidence Interval) under the ROCcurve for LDH then increased from 0.98 (0.97–1.00) to 0.99(0.98–1.00), i.e. the specificity of LDH for prediction of HIEis even better if the test is performed closer to birth. Further-more, the increased LDH levels seen at 12 h of life inhealthy infants previously reported (17) is low comparedwith the LDH levels found in the group of infants with HIEin our study, suggesting that the difference in time points forblood collection between the groups does not have a greatimpact on the result in this study.

The present results showing that [LDH] predicts mild,moderate and severe HIE is of direct clinical interest viewedfrom several perspectives:

Firstly, clinical trials suggest long term benefits of mildhypothermia (HT) treatment in infants with moderate andsevere HIE (18–20) when given as soon as possible after theinsult. The importance of the time factor is supported byanimal studies suggesting that the neuroprotective effects ofHT decline with increased delay between the hypoxic-is-chaemic (HI) insult and the start of the HT treatment(21,22). A prognostic tool early after birth would make itpossible to select candidates for hypothermia treatment.Recent data show that infants with asphyxia, even if they donot develop moderate or severe HIE, are at greater risk oflow intelligence quotients (IQs) later in life (4); this raisesthe question whether neuroprotective therapy should beprovided for this group of infants in the future. A biochemi-cal marker for prediction of mild as well as moderate tosevere HIE would be of interest in this perspective. In fact,the result showing a good predictive value of [LDH] beforeonset of seizure increases the potential of this enzyme as a‘biochemical criteria’ for early identification of candidatesfor hypothermia treatment.

Secondly, our results are also of interest in obstetrics.LDH levels correlate to nucleated red blood cells duringlate foetal life, a known indicator of chronic hypoxia (23)and we speculate that measurement of LDH, ALT and ASTin scalp blood during parturition could offer a more predic-tive test for foetal distress than pH and lactate, reflectingcellular damage in organs affected by the hypoxia-

ischaemia, in some cases before the brain is damaged (5,24).The differing dynamic pattern and decline rates of LDH,ALT and AST during the first hours and days after theasphyxia could make it possible to identify foetuses enteringthe delivery with ante-partum hypoxia-ischaemia and pro-vide information whether the asphyxia started before orduring delivery. This study was not designed to investigatethe potential of these enzymes for timing of the asphycticevent. However, the enzyme pattern seen in our study incases with known onset of asphyxia support this theory andresemble that seen in adults after cardiovascular events withknown onset (10,12).

[LDH] had the best predictive value of the three enzymesanalysed. This difference between the enzymes could beexplained by the enzyme distribution in mammals. LDH ispresent in all tissues and is therefore superior as a marker ofcell damage with unknown origin which is the case in theasphyxia process. Multi-organ dysfunction is nearly alwaysseen together with HIE but it could not be predicted whichorgan systems would be affected on an individual basis.Therefore, LDH is suitable as a first Triage tool which couldbe followed by more specific markers and examination fordetermination of the multi-organ dysfunction pattern andits severity.

In conclusion this study reports that [LDH], [ALT] and[AST] during the first 12 h after birth are good predictors ofthe different grades of severity of HIE and long term out-come and might offer an inexpensive and reliable comple-ment to the already existing prognostic tests for evaluatingnewborns with asphyxia. Therefore, these preliminaryresults motivate a prospective study on the subject.

ACKNOWLEDGEMENTSWe are grateful for funding from Research and PublicHealth in Varmland and SPARKS.

COMPETING INTERESTSNone.

References

1. Lawn JE, Cousens S, Zupan J, Lancet Neonatal Survival Steer-ing Team. 4 million neonatal deaths: when? Where? Why?Lancet 2005; 365: 891–900.

2. Peliowski A, Finer NN. Birth asphyxia in the term infant. In:Sinclair JC, Bracken MB, editors. Effective care of the newborninfant. Oxford: Oxford University Press, 1992: 249–79.

3. Lindstrom K, Hallberg B, Blennow M, Wolff K, Fernell E,Westgren M. Moderate neonatal encephalopathy: pre- andperinatal risk factors and long-term outcome. Acta ObstetGynecol Scand 2008; 87: 503–9.

4. Odd DE, Lewis G, Whitelaw A, Gunnell D. Resuscitation atbirth and cognition at 8 years of age: a cohort study. Lancet2009; 373: 1615–22.

5. Jensen A, Garnier Y, Berger R. Dynamics of fetal circulatoryresponses to hypoxia and asphyxia. Eur J Obstet Gynecol Re-prod Biol 1999; 84: 155–72.

Karlsson et al. LDH predicts HIE


6. Karlsson M, Blennow M, Nemeth A, Winbladh B. Dynamics ofhepatic enzyme activity following birth asphyxia. Acta Paediatr2006; 95: 1405–11.

7. Lackmann GM, Tollner U, Mader R. Serum enzyme activitiesin full-term asphyxiated and healthy newborns: enzyme kinet-ics during the first 144 hours of life. Enzyme Protein 1993; 47:160–72.

8. Zanardo V, Bondio M, Perini G, Temporin GF. Serum glu-tamic-oxaloacetic transaminase and glutamic-pyruvic transami-nase activity in premature and full-term asphyxiated newborns.Biol Neonate 1985; 47: 61–9.

9. Rej R. Aspartate aminotransferase activity and isoenzyme pro-portions in human liver tissues. Clin Chem 1978; 24: 1971–9.

10. Gibson PR, Dudley FJ. Ischemic hepatitis: clinical features,diagnosis and prognosis. Aust NZ J Med 1984; 14: 822–5.

11. Lackmann GM, Tollner U. The predictive value of elevation inspecific serum enzymes for subsequent development of hyp-oxic-ischemic encephalopathy or intraventricular hemorrhagein full-term and premature asphyxiated newborns. Neuropedi-atrics 1995; 26: 192–8.

12. Seeto RK, Fenn B, Rockey DC. Ischemic hepatitis: clinical pre-sentation and pathogenesis. Am J Med 2000; 109: 109–13.

13. Reichling JJ, Kaplan MM. Clinical use of serum enzymes inliver disease. Dig Dis Sci 1988; 33: 1601–14.

14. Levene MI, Sands C, Grindulis H, Moore JR. Comparison oftwo methods of predicting outcome in perinatal asphyxia. Lan-cet 1986; 1: 67–9.

15. Picard RR, Berk KN. Data splitting. Am Stat 1990; 44: 140–7.16. Reddy S, Dutta S, Narang A. Evaluation of lactate dehydroge-

nase, creatine kinase and hepatic enzymes for the retrospectivediagnosis of perinatal asphyxia among sick neonates. IndianPediatr 2008; 45: 144–7.

17. Lackmann GM. Reference values for selected enzyme activitiesin serum from healthy human neonates. Clin Biochem 1996;29: 599–602.

18. Azzopardi DV, Strohm B, Edwards AD, Halliday H, Levene M,Thoresen M, et al. Moderate hypothermia to treat perinatal as-phyxial encephalopathy. N Engl J Med 2009; 361: 1349–58.

19. Gluckman PD, Wyatt JS, Azzopardi D, Ballard R, Edwards AD,Ferriero DM, et al. Selective head cooling with mild systemichypothermia after neonatal encephalopathy: multicentrerandomised trial. Lancet 2005; 365: 663–70.

20. Shankaran S, Laptook AR, Ehrenkranz RA, Tyson JE, McDon-ald SA, Fanaroff AA, et al. Whole-body hypothermia for neo-nates with hypoxic-ischemic encephalopathy. N Engl J Med2005; 353: 1574–84.

21. Gunn AJ. Cerebral hypothermia for prevention of brain injuryfollowing perinatal asphyxia. Curr Opin Pediatr 2000; 12:111–5.

22. Karlsson M, Tooley JR, Satas S, Hobbs C, Chakkarapani E,Stone J, et al. Delayed hypothermia as selective head coolingor whole body cooling does not protect brain or body innewborn pig subjected to hypoxia-ischemia. Pediatr Res2008; 64: 74–8.

23. Verspyck E, Gaillard G, Parnet F, Marret S, Marpeau L. Fetallactic dehydrogenase variation in normal pregnancy and incases of severe intra-uterine growth restriction. Prenat Diagn1999; 19: 229–33.

24. Perez R, Espinoza M, Riquelme R, Parer ET, Llanos AJ. Argi-nine vasopressin mediates cardiovascular responses to hypox-emia in fetal sheep. Am J Physiol 1989; 256: R1011–8.



lactate dehydrogenase predicts hypoxic ischaemic encephalopathy in newborn infants: a preliminary...

Documents