Journal of Perinatology (2014) 34, 203–208 & 2014 Nature America, Inc. All rights reserved 0743-8346/14 www.nature.com/jp

ORIGINAL ARTICLE

Outcomes of extremely preterm infants following severe intracranial hemorrhage AS Davis1, SR Hintz1, RF Goldstein2, N Ambalavanan3, CM Bann4, BJ Stoll5, EF Bell6, S Shankaran7, AR Laptook8, MC Walsh9, EC Hale5, NS Newman9, A Das10 and RD Higgins11 for the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network OBJECTIVE: Severe intracranial hemorrhage (ICH) is an important prognostic variable in extremely preterm (EPT) infants. We examined imaging and clinical variables that predict outcomes in EPT infants with severe ICH. STUDY DESIGN: Retrospective analysis of 353 EPT infants with severe ICH. Outcomes were compared by examining: (i) unilateral vs bilateral ICH; and (ii) presence vs absence of hemorrhagic parenchymal infarction (HPI). Regression analyses identified variables associated with death or neurodevelopmental impairment (NDI). RESULT: Bilateral ICH and HPI had higher rates of adverse outcomes and were independently associated with death/NDI. HPI was the most important variable for infants of lower birth weight, and bilateral ICH for larger infants. For infants surviving to 36 weeks, shunt placement was most associated with death/NDI. CONCLUSION: Bilateral ICH and the presence of HPI in EPT infants with severe ICH are associated with death/NDI, though the importance depends on birth weight and survival to 36 weeks. Journal of Perinatology (2014) 34, 203–208; doi:10.1038/jp.2013.162; published online 26 December 2013 Keywords: intraventricular hemorrhage; neurodevelopmental impairment; extremely low birth weight; cranial ultrasound

INTRODUCTION Neonatologists rely on several prognostic paradigms when counseling families of extremely preterm (EPT) infants about the possibility of adverse outcome. Existing methods utilize antenatal variables,1 illness severity scoring systems,2,3 and early clinical variables.4,5 The diagnosis of severe intracranial hemorrhage (ICH) on cranial ultrasound (cUS) is often considered a key variable to predict adverse outcome. Severe ICH, typically defined as grade III or IV by Papile criteria,6 has been shown to be associated with adverse neurologic outcomes, including moderate- to severecerebral palsy (CP), neurodevelopmental impairment (NDI) and even cognitive delay.7–10 The presence of severe neuroimaging findings, though, is not always predictive of adverse early childhood outcome. A singlecenter study demonstrated that B50% of extremely low birth weight infants with grade III and IV ICH did not have significant cognitive impairment or neurologic abnormality at 20-month follow-up.11 Similarly, in a large cohort of extremely low birth weight survivors, 45% of infants with grade III ICH and 30% of infants with grade IV ICH did not have NDI.4 A more recent report found that, except when accompanied by white matter damage, ICH is not associated with developmental delay.12 One possible explanation for the range in outcomes reported is that many studies use ‘severe ICH’ as a composite predictor that does not distinguish between unilateral and bilateral ICH, and hemorrhagic

parenchymal infarction (HPI),13 typically defined as grade IV ICH, is not differentiated from a grade III ICH, or bleeding that is confined to the ventricles. We hypothesized that specific characteristics of ICH, including bilateral compared with unilateral severe ICH or the presence of HPI, would be important predictors of outcomes of EPT infants with severe ICH. Using data from the Eunice Kennedy Shriver National Institute of Child Health and Human Development Neonatal Research Network (NRN), we compared the outcomes of EPT infants with severe ICH, using bilaterality of ICH and presence of HPI as distinct predictor variables. We further hypothesized that this information, in combination with other clinical factors, could be used in multivariable regression analysis and Classification and Regression Tree (CART) modeling to refine the prediction of outcome in this high-risk group. METHODS This was an analysis of infants o27 weeks gestational age (GA) and born 1/1/2006 to 12/31/2007 at participating NRN centers with severe ICH (see definition below). Infants with major congenital anomalies, including congenital central nervous system defects, complex congenital heart disease and chromosomal anomalies were excluded. All data were collected as part of the prospective NRN Survey of Morbidity and Mortality among very low birth weight infants (401 to 1500 g) and follow-up studies, with research coordinators abstracting data using study definitions as

1 Division of Neonatal and Developmental Medicine, Department of Pediatrics, Stanford University, Palo Alto, CA, USA; 2Division of Neonatal-Perinatal Medicine, Department of Pediatrics, Duke University Medical Center, Durham, NC, USA; 3Department of Pediatrics, University of Alabama at Birmingham, Birmingham, AL, USA; 4Statistics and Epidemiology Unit, RTI International, Research Triangle Park, NC, USA; 5Department of Pediatrics, Emory University School of Medicine, Atlanta, GA, USA; 6Department of Pediatrics, University of Iowa Carver College of Medicine, Iowa City, IA, USA; 7Department of Pediatrics, Wayne State University, Detroit, MI, USA; 8Department of Pediatrics, Brown University, Providence, RI, USA; 9Department of Pediatrics, Case Western Reserve University, Cleveland, OH, USA; 10Statistics and Epidemiology Unit, RTI International, Rockville, MD, USA and 11Pregnancy and Perinatology Branch, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA. Correspondence: Dr AS Davis, Division of Neonatal and Developmental Medicine, Department of Pediatrics, Stanford University, 750 Welch Road, Suite 315, Palo Alto, CA 94304, USA. E-mail: [email protected] Received 14 August 2013; revised 12 November 2013; accepted 18 November 2013; published online 26 December 2013

Severe ICH and outcomes AS Davis et al

204 described. The institutional review board at each center approved these studies.

Study definitions Trained research staff recorded the most severe findings on cUS in the first 28 days of life as interpreted by staff radiologists at each NRN center. Infants with grade III or IV ICH by cUS, based on definitions by Papile6 and as described in previous reports4,7,8 were defined as having severe ICH. For the purposes of the current study, unilateral severe ICH was defined as grade III or grade IV ICH confined to one side of the brain; less severe ICH (grade I or II) or no ICH could have been present on the contralateral side. Bilateral ICH was defined as grade III or grade IV (any combination thereof) on both sides. HPI was defined as having grade IV ICH on at least one side. The diagnoses of bilateral ICH or HPI were not mutually exclusive in this study: infants were classified as having bilateral or unilateral ICH and were separately classified as having HPI or no HPI. The diagnoses of periventricular leukomalacia (PVL), porencephalic cyst and ventriculomegaly were recorded for infants surviving to 36 weeks postmenstrual age (PMA). PVL was diagnosed by the presence of cystic echolucencies in the periventricular white matter on any cUS imaging. Porencephalic cyst was defined as one or more cysts within the cerebral hemisphere that may or may not communicate with the lateral ventricle, excluding subependymal and choroid plexus cysts. Ventriculomegaly was defined as the presence of enlarged ventricles on cranial imaging performed closest to 36 weeks PMA based on local interpretation, or from the cUS in the first 28 days, if later imaging was not performed (seven infants). Additional demographic, maternal and neonatal information was collected from birth until death, hospital discharge or 120 days. Estimated GA was determined by best obstetric estimate. Antenatal steroids were defined as the administration of any corticosteroids to accelerate fetal maturity in the present pregnancy. Infants were classified as small for GA at birth, defined by a birth weight o10th percentile for gender and GA.14 Surfactant treatment was defined as at least one dose of any surfactant. Bronchopulmonary dysplasia was defined as requiring supplemental oxygen at 36 weeks PMA. Postnatal steroid treatment was any steroid given for the prevention or treatment of bronchopulmonary dysplasia. Necrotizing enterocolitis was defined as modified Bell stage IIA or greater.15,16 Severe retinopathy of prematurity (ROP) was defined as stage 3 or greater with ‘plus’ disease. Early-onset sepsis (within 72 h of birth) and late-onset sepsis (after 72 h) were defined by a positive blood culture and antimicrobial therapy for X5 days or when there was intent to treat but the infant died before 5 days of therapy. Seizures were recorded based on treatment with an anticonvulsant for 472 h or when there was intent to treat but the baby died before 72 h. Late death was defined as death after 36 weeks PMA.

Neurodevelopmental follow-up Surviving infants were assessed at 18 to 22 months corrected age using a standardized neuromotor exam and the Gross Motor Function Classification System (GMFCS).17 Examinations were performed by certified examiners who were trained to reliability during a 2 day workshop on neurologic assessment or certified by those who attended the workshop. The Bayley Scales of Infant Development III18 Cognitive and language composite scores were administered by experienced testers, who were annually certified by 1 of 4 gold-standard examiners. CP was defined as a non-progressive central nervous system disorder characterized by abnormal muscle tone in at least one extremity and abnormal control of movement or posture that interfered with or prevented age-appropriate activity. Moderate CP was defined as no ambulation or ambulation only with assistive devices but ability to sit independently or with support (GMFCS level 2 or 3). Severe CP was defined as the inability to ambulate or sit with support (GMFCS level 4 or 5). NDI was defined as any of the following: moderate- to severe-CP with a GMFCS X2, Bayley III Cognitive composite score o70, bilateral blindness (defined as vision o20/200 bilaterally) or permanent bilateral severe hearing loss. Unimpaired was defined as all of the following: Bayley III Cognitive composite score 485, no moderate- to severe-CP, GMFCSo1 and without bilateral blindness or deafness.

Statistical analyses Statistical analyses were performed using SAS statistical software version 9.2 (SAS Institute, Cary, NC, USA) and SPSS AnswerTree software (SPSS, Chicago, IL, USA). Statistical significance was established using a P-valueo0.05. Demographic, perinatal and neonatal characteristics of the Journal of Perinatology (2014), 203 – 208

study population were assessed using descriptive statistical methods. Bivariate analyses were conducted for the outcomes of infants at 18 to 22 months using the w2 test for differences in proportions and the Student’s t-test for continuous variables. Comparisons were performed for the categorical predictors: (1) bilateral vs unilateral ICH and (2) HPI vs no HPI. Multivariable logistic regression models were created using a backward stepwise approach with Po0.20 as the criterion for remaining in the model. The first included all infants and modeled the outcome death or NDI at 18 to 22 months corrected age. Variables initially included in the model were: antenatal steroids, cesarean delivery, multiple birth, rupture of membranes 418 h, antepartum hemorrhage, histologic chorioamnionitis, male sex, birth weight, GA, Small for GA, race/ethnicity, 5-min Apgar o4, cardiopulmonary resuscitation in the delivery room, surfactant, high frequency ventilator in the first 28 days, any HPI and PVL (using cUS results in the first 28 days). Bilateral vs unilateral ICH was forced into the model, and NRN center was entered as a random effect. A second regression analysis was limited to infants surviving to 36 weeks PMA and modeled the outcome of late death or NDI. Variables considered in this model included all above variables as well as the following in-hospital morbidities: patent ductus arteriosus, late-onset sepsis, seizures, bronchopulmonary dysplasia, ROP, postnatal steroids, surgery (for necrotizing enterocolitis, patent ductus arteriosus or ROP), ventriculomegaly, porencephalic cyst, the presence of a ventriculoperitoneal shunt and PVL. CART algorithms were developed using the same patient groups, outcomes and variables as in the logistic regression modeling, except center, including a separate analysis for infants surviving to 36 weeks PMA. CART analysis ranks the importance of predictors through recursive partitioning of variables. For continuous variables such as birth weight, the program selects optimal, rather than empiric, cut-points. For this analysis, the maximum tree depth was empirically set at five levels, with a minimum of 50 observations in the parent (upper) node and 25 observations in each child (lower) node.

RESULTS During the study period, 2218 infants o27 weeks GA were born alive in 20 NRN centers, including 353 infants with severe ICH who had outcome data available (93% follow-up rate, see Supplementary Figure). Of the 507 infants excluded due to lack of cUS, 504 died (473 in the first day of life, 499 in the first week of life). Bilateral ICH was present in 70% and HPI was present in 58% of the study population (see Supplementary Table for additional demographic, perinatal and clinical characteristics of the study population). Infants lost to follow-up (n ¼ 26) were more likely to be female, had significantly higher birth weights, and had higher rates of PVL, ROP, postnatal steroid exposure and surgery compared with those who died or were followed-up at 18 to 22 months. Infants lost to follow-up did not differ in the rates of bilateral vs unilateral ICH or HPI vs no HPI. Unadjusted early childhood outcomes of the study population are shown in Table 1. Infants with bilateral ICH or HPI on at least one side were more likely to have the outcomes of death at any time or death/NDI. However, there was no difference in late death (after 36 weeks PMA) by ICH bilaterality or by presence or absence of HPI. Of the 175 infants who died before 36 weeks PMA, 138 (79%) had respiratory support withdrawn; ICH was listed as a cause of death or withdrawal of care in 50% of cases. There was no significant variation across centers in withdrawal of support. Among infants who died, GA and birth weight did not differ between those who died due to withdrawal of support vs those without withdrawal of support. Bilateral ICH was associated with a higher likelihood of support being withdrawn compared with unilateral ICH (85 vs 60%, unadjusted OR 3.84 (CI 1.76, 8.41), Po0.001). Similarly, infants with HPI were more likely to have support withdrawn compared with those without, but the difference between groups was not statistically significant (83 vs 70%, unadjusted OR 2.01 (CI 0.94, 4.27), P ¼ 0.07). Among survivors to follow-up, NDI, moderate-to-severe CP and Bayley III Cognitive composite score o70 were more common following bilateral ICH or HPI (all Po0.05). Infants with bilateral ICH & 2014 Nature America, Inc.

Severe ICH and outcomes AS Davis et al

205 Table 1.

Early childhood outcomes of study population

Variable

Severe ICH type

HPI present

Unilateral

Bilateral

No

Yes

N (%)

N (%)

N (%)

N (%)

N (%)

353 194 (55) 19 (5) 236 (67)

105 46 (44) 4 (4) 55 (52)

248 148 (60)a 15 (6) 181 (73)a

147 63 (43) 9 (6) 79 (54)

206 131 (64)b 10 (5) 157 (76)b

Infants surviving to 18–22 month follow-up N NDI Unimpairedc Moderate/severe CP Seizures Rehospitalized Tube feedings OFCo10th percentile Bayley III Cognitive o70 Bayley III Cognitive o85 Bayley III Language o70 Bayley III Language o85

159 42 (26) 75 (47) 26 (16) 16 (10) 89 (56) 21 (13) 55 (35) 28 (18) 72 (45) 27 (17) 75 (55)

59 9 (15) 31 (53) 5 (8) 2 (3) 28 (47) 9 (15) 19 (32) 5 (8) 22 (37) 12 (20) 26 (49)

100 33 (33)a 44 (44) 21 (21)a 14 (14)a 61 (61) 12 (12) 36 (36) 23 (23)a 50 (50) 15 (15) 49 (59)

84 16 (19) 44 (52) 8 (10) 8 (10) 44 (52) 5 (6) 28 (33) 10 (12) 36 (43) 15 (18) 43 (57)

75 26 (35)b 31 (44) 18 (24)b 8 (11) 45 (60) 16 (21)b 27 (36) 18 (24)b 36 (48) 12 (16) 32 (52)

Bayley III scores (mean±s.d.) Cognitive composite Score Language composite Score

83 (16) 80 (19)

85 (13) 83 (18)

81 (17) 78 (19)

83 (13) 80 (16)

82 (18) 79 (22)

All children N Death Death after 36 weeks Death/NDI

Abbreviations: CP, cerebral palsy; HPI, hemorrhagic parenchymal infarction; ICH, intracranial hemorrhage; NDI, neurodevelopmental impairment. a Po0.05 for bilateral ICH vs unilateral ICH comparison. bPo0.05 for HPI vs no HPI comparison. cUnimpaired was defined as all of the following: Bayley III Cognitive composite score 485, no moderate-to-severe CP, GMFCS o1, and without bilateral blindness or deafness.

were more likely to have seizures than infants with unilateral ICH. Infants with HPI were more likely to require tube feedings at 18 to 22 months compared with those without. The rate of unimpaired status was higher in infants with unilateral ICH (53%) compared with infants with bilateral ICH (44%), but this result was not statistically significant. The covariates retained by the stepwise regression analyses for the outcomes death/NDI and late death/NDI are shown in Table 2. For all infants, bilateral ICH and HPI were independently associated with death/NDI and had similar magnitudes of effect. For the analysis limited to infants surviving to 36 weeks PMA, the presence of a shunt, bilateral ICH and HPI were associated with late death/NDI, with shunt being the greatest in magnitude. When additional analyses were undertaken to force GA, sex, Small for GA and antenatal steroids into the models, significant variables associated with outcomes remained unchanged (data not shown). CART analysis results are shown in Figure 1, with predictive variables that are more strongly associated with the outcome shown higher on the decision tree. When the entire study population was included in the analysis (Figure 1a), infants with a birth weight p707 g (the optimal cut-point selected by the CART program) were more likely to have the outcome death/NDI. In these smaller infants, HPI was more likely to influence the outcome death/NDI (88 vs 65%), and in those without HPI, male infants had a worse outcome (82 vs 49%). In the larger infants, bilateral ICH was associated with death/NDI (63 vs 38%), and among infants with bilateral ICH, the presence of HPI predicted a worse outcome (76 vs 48%). When the analysis was limited to infants surviving to 36 weeks PMA (Figure 1b), neither bilateral ICH nor the presence of HPI remained in the decision tree. However, the presence of a shunt predicted late death/NDI (70 vs 28%). Among infants without a shunt, the receipt of postnatal steroids and the presence of enlarged ventricles on late US were associated with late death/NDI. & 2014 Nature America, Inc.

Table 2.

Stepwise logistic regression models for the outcome death/

NDI Variable All Children (N ¼ 353) HPI Bilateral ICH 5-min Apgar o4 Surfactant High frequency ventilator Male Birth weight (kg) Cesarean delivery

OR (95% CI)

P value

3.56 3.16 2.81 2.45 1.90 1.81 0.02 0.53

(2.08, (1.80, (1.28, (0.71, (1.07, (1.06, (0.00, (0.31,

6.09) 5.54) 6.17) 8.42) 3.36) 3.10) 0.16) 0.90)

o0.001 o0.001 0.01 0.15 0.03 0.03 o0.001 0.02

Survivors to 36 weeks PMA (N ¼ 178)a Bilateral ICH 2.75 HPI 2.50 Shunt 5.35 Postnatal steroids 3.71 Antepartum hemorrhage 3.10 ROM 418 h 2.52 PVL 2.46 BPD 2.18 Ventricle size enlarged 2.16 GA 0.64

(1.05, (1.02, (1.63, (1.22, (1.19, (0.89, (0.87, (0.86, (0.89, (0.42,

7.20) 6.10) 17.5) 11.2) 8.09) 7.15) 6.96) 5.48) 5.23) 0.96)

0.03 0.045 0.006 0.02 0.02 0.08 0.1 0.07 0.09 0.03

Abbreviations: BPD, bronchopulmonary dysplasia; CI, confidence interval; HPI, hemorrhagic parenchymal infarction; ICH, intracranial hemorrhage; NDI, neurodevelopmental impairment; PMA, postmenstrual age; PVL, periventricular leukomalacia; ROM, rupture of membrane. a Primary outcome for survivors to 36 weeks was late death/NDI.

The predictive abilities of logistic regression and CART analysis were comparable. Logistic regression models correctly classified 76% of infants for the outcome death/NDI and 78% for the Journal of Perinatology (2014), 203 – 208

Severe ICH and outcomes AS Davis et al

206

Figure 1. Results of CART analysis. (a) The most relevant variables for the outcome death/NDI when the entire study population was analyzed. (b) Results for the outcome late death/NDI among infants surviving to 36 weeks.

outcome late death/NDI. Both models had 80% sensitivity. By comparison, the accuracy of the CART model for the outcome death/NDI was 71%, with 72% sensitivity. CART analysis for the outcome late death/NDI correctly classified 72% of infants, with higher specificity (93%) but low sensitivity (31%).

DISCUSSION Abnormalities on head ultrasound in infants born prematurely have been a specific focus of investigation in neonatal research, and more recent studies have begun to assess the differences between unilateral vs bilateral involvement and the presence of HPI in relation to neurologic outcomes. We examined a cohort of EPT infants with severe ICH and demonstrated that bilateral ICH and PI are significantly associated with death/NDI, independent of other demographic and clinical variables. Approximately half of Journal of Perinatology (2014), 203 – 208

infants who survived to early childhood follow-up were classified as unimpaired. Two large cohort studies examined the association between cUS abnormalities and CP. The EPIPAGE study (N ¼ 1954) reported a higher proportion of CP in infants with bilateral cystic PVL (74%) compared with unilateral PVL (35%); a similar result was found for infants with HPI compared with lower grades of ICH.19 In the ELGAN cohort (N ¼ 1053), higher rates of CP were reported with bilateral ICH (28%) compared with those with unilateral findings (9%).20 Increasing severity of HPI has been associated with death, neonatal seizures, and an abnormal neuromotor exam beyond 12 months’ adjusted age.21 Bassan et al.22 assessed 30 preterm infants with HPI at a mean follow up of 30 months corrected age and reported a trend for bilateral HPI to predict hypertonia. Roze et al.23,24 analyzed 54 preterm infants with HPI to determine the contribution of both laterality (for example, right vs left) and & 2014 Nature America, Inc.

Severe ICH and outcomes AS Davis et al

207 bilaterality toward early childhood outcomes. No differences were seen in the rates of death between infants with bilateral vs unilateral HPI. The rate of CP did not differ by bilaterality or laterality in surviving infants up to early school age follow-up. In a single center study by Merhar et al.25 bilaterality did not result in worse outcome except for infants with grade IV ICH. Maitre et al.26 evaluated 69 surviving very low birth weight infants with HPI at 18 to 36 months corrected age and found that those with bilateral HPI had higher rates of CP and lower Bayley II scores compared with those with unilateral HPI. In multivariate modeling, infants with bilateral HPI alone had a 93% probability of Bayley II Mental Development Index o70; by comparison, infants with unilateral HPI required the additional diagnoses of PVL and ROP to achieve a 89% predicted probability of Mental Development Index o70. This underscores the potential for multiple clinical factors in combination to contribute to adverse outcome. Our study expands upon the aforementioned analyses by evaluating a large sample of extremely premature infants with severe ICH, which permitted adjustment for multiple clinical variables that may also contribute to adverse outcome. Our unadjusted analyses showed that the outcomes of those with severe unilateral ICH are better than those with bilateral involvement. This was particularly striking for infants who survived to early childhood follow-up. Logistic regression confirmed the significant association between HPI, ICH bilaterality and outcome in the entire study population as well as in surviving infants, independent of other potentially confounding variables. The models were further validated by the stepwise retention of variables previously known to be associated with poor outcome, such as male sex, lower birth weight and GA, and use of postnatal steroids. CART analysis may be useful for physicians, who are accustomed to clinical algorithms, to use when counseling families of EPT infants with severe ICH about the risk of adverse outcome in the context of other clinical variables. While providing useful prognostic information, though, both logistic regression and CART models should serve to complement clinical decision-making, as their accuracy at B70% indicates that more precise descriptions of severe ICH characteristics and consideration of other currently unmeasured or unmeasurable factors and influences may be needed to improve our predictive abilities. Infants in our study population had high rates of death and withdrawal of support. Bassan et al.21 also reported an association between the severity of HPI and the likelihood to withdraw care. There is potential for bias in any study that reports an outcome of early neonatal death for a variable such as ICH, as clinicians likely use this information to counsel families regarding the continuation of intensive care. For this reason, we separately analyzed infants who survived until 36 weeks PMA, and the independent association of bilateral ICH and HPI persisted for the outcome late death/NDI. As mortality rates after 36 weeks PMA were not different between groups by bilaterality or HPI, this association is explained by observed differences in neurodevelopmental outcomes at 18 to 22 months corrected age. Thus, our data support that neuroimaging variables, including bilateral involvement and presence of HPI, are significantly associated with early childhood neurodevelopmental outcomes among EPT infants with severe ICH. An important limitation of our study was the inability to further characterize the parenchymal involvement in infants classified as having HPI, as detailed information regarding the extent and location of injury was not prospectively collected for the NRN database. This study also relied on a retrospective analysis of neuroimaging data extracted from the medical record. The diagnosis of severe ICH has been shown to have good interreader reliability,27 minimizing the potential for misclassification bias. Arguably, our data may reflect a more ‘real world’ diagnosis of severe ICH, as rigorous, prospective training of central readers was not utilized. & 2014 Nature America, Inc.

Our study highlights the need for future investigations to consider bilateral involvement and HPI as important clinical factors rather than continued use of a categorical definition of ‘severe ICH’. Further, both multivariate regression modeling and CART analysis demonstrated that clinical variables in addition to severe ICH remain important in the prediction of adverse outcome. Our findings support the concept that future studies should distinguish severe ICH with HPI from severe ICH that is limited to the ventricles. Among infants with HPI, the impact of laterality (right vs left), location and extent of involvement on early childhood outcomes should be investigated. Although previous reports suggest that preterm infants with severe ICH constitute a population at extremely high risk for adverse outcome, our data suggest that a substantial proportion of surviving infants do not have serious impairment in early childhood. It is unclear if the same result would be seen at school age follow-up, or whether more subtle deficits would be found. This underscores the need for continued assessment of preterm infants with severe ICH that includes long-term follow-up and improved identification of factors associated with both adverse and favorable outcomes.

CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS We are indebted to our medical and nursing colleagues and the infants and their parents who agreed to take part in this study. Investigators, in addition to those listed as authors, who participated in this study are listed in the Appendix. The National Institutes of Health and the Eunice Kennedy Shriver National Institute of Child Health and Human Development provided grant support for the Neonatal Research Network’s Generic Database Study and Follow-up Study.

REFERENCES 1 Tyson JE, Parikh NA, Langer J, Green C, Higgins RD. Intensive care for prematurity: moving beyond gestational age. New Engl J Med 2008; 358: 1672–1681. 2 Richardson DK, Corcoran JD, Escobar GJ, Lee SK, for the Canadian NICU Network and The Kaiser Permanente Neonatal Minimum Data Set Wide Area Network and The SNAP-II Study Group. SNAP-II and SNAPPE-II: simplified newborn illness severity and mortality risk scores. J Pediatr 2001; 138(1): 92–100. 3 Meadow W, Frain L, Ren Y, Lee G, Soneji S, Lantos J. Serial assessment of mortality in the neonatal intensive care unit by algorithm and intuition: certainty, uncertainty, and informed consent. Pediatrics 2002; 109(5): 878–886. 4 Broitman E, Ambalavanan N, Higgins R, Vohr B, Das A, Bhaskar B et al. Clinical data predict neurodevelopmental outcome better than head ultrasound in extremely low birth weight infants. J Pediatr 2007; 151: 500–505. 5 Ambalavanan N, Baibergenova A, Carlo WA, Saigal S, Schmidt B, Thorpe KE et al. Early prediction of poor outcome in extremely low birth weight infants by classification tree analysis. J Pediatr 2005; 148: 438–444. 6 Papile LA, Burstein J, Burstein R, Koffler H. Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1,500 gm. J Pediatr 1978; 92(4): 529–534. 7 Vohr BR, Wright LL, Poole WK, McDonald SA. for the NICHD Neonatal Research Network Follow-Up Study. Neurodevelopmental outcomes of extremely low birth weight infants o32 weeks’ gestation between 1993 and 1998. Pediatrics 2005; 116: 635–643. 8 Hintz SR, Kendrick DE, Vohr BR, Poole WK, Higgins RD, for the National Institute of Child Health and Human Development Neonatal Research Network. Changes in neurodevelopmental outcomes at 18 to 22 months’ corrected age among infants of less than 25 weeks’ gestational age born in 1993-1999. Pediatrics 2005; 115: 1645–1651. 9 Fily A, Pierrat V, Delporte V, Breart G, Truffert P. EPIPAGE Nord-Pas-de-Calais Study Group. Factors associated with neurodevelopmental outcome at 2 years after very preterm birth: the population-based Nord-Pas-de-Calais EPIPAGE cohort. Pediatrics 2006; 117(2): 357–366. 10 Wood NS, Costeloe K, Gibson AT, Hennessy EM, Marlow N, Wilkinson AR et al. The EPICure study: associations and antecedents of neurological and developmental disability at 30 months of age following preterm birth. Arch Dis Child Fetal Neonatal Ed 2005; 90(2): F134–F140.

Journal of Perinatology (2014), 203 – 208

Severe ICH and outcomes AS Davis et al

208 11 Hack M, Wilson-Costello D, Friedman H, Taylor GH, Schluchter M, Fanaroff AA. Neurodevelopment and predictors of outcomes of children with birth weights of less than 1000 g: 1992-1995. Arch Pediatr Adolesc Med 2000; 154(7): 725–731. 12 O’Shea TM, Allred EN, Kuban KC, Hirtz D, Specter B, Durfee S et al. Intraventricular hemorrhage and developmental outcomes at 24 months of age in extremely preterm infants. J Child Neurol 2012; 27(1): 22–29. 13 de Vries LS, van Haastert IC, Benders MJNL, Groenendaal F. Myth: cerebral palsy cannot be predicted by neonatal brain imaging. Semin Fetal Neonatal 2011; 16: 279–287. 14 Alexander GR, Himes JH, Kaufman RB, Mor J, Kogan MA. United States national reference for fetal growth. Obstet Gynecol 1996; 87(2): 163–168. 15 Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L et al. Neonatal necrotizing enterocolitis: therapeutic decisions based on clinical staging. Ann Surg 1978; 187(1): 1–7. 16 Walsh MC, Kliedgman RM. Necrotizing enterocolitis: treatment based on staging criteria. Pediatr Clin North Am 1986; 33(1): 179–201. 17 Palisano R, Rosenbaum P, Walter S, Russel D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol 1997; 99(4): 214–223. 18 Bayley N. Bayley Scales of Infant and Toddler Development. 3rd edn. Psychological Corp: San Antonio, TX, 2006. 19 Ancel P-Y, Livinec F, Larroque B, Marret S, Arnaud C, Pierrat V et al. Cerebral palsy among very preterm children in relation to gestational age and neonatal ultrasound abnormalities: The EPIPAGE Cohort Study. Pediatrics 2006; 117(3): 828–835.

20 Kuban KC, Allred EN, O’Shea TM, Paneth N, Pagano M, Dammann O et al. Cranial ultrasound lesions in the NICU predict cerebral palsy at age 2 years in children born at extremely low gestational age. J Child Neurol 2009; 24(1): 63–72. 21 Bassan H, Benson CB, Limperopoulos C, Feldman HA, Ringer SA, Veracruz E et al. Ultrasonographic features and severity scoring of periventricular hemorrhagic infarction in relation to risk factors and outcome. Pediatrics 2006; 117(6): 2111–2118. 22 Bassan H, Limperopoulos C, Visconti K, Mayer DL, Feldman HA, Avery L et al. Neurodevelopmental outcome in survivors of periventricular hemorrhagic infarction. Pediatrics 2007; 120(4): 785–792. 23 Roze E, Kerstjens JM, Maathuis CGB, ter Horst HJ, Bos AF. Risk factors for adverse outcome in preterm infants with periventricular hemorrhagic infarction. Pediatrics 2008; 122(1): e46–e52. 24 Roze E, Van Braeckel KNJA, van der Veere CN, Maathuis CGB, Martijn A, Bos AF. Functional outcomes at school age of preterm infants with periventricular hemorrhagic infarction. Pediatrics 2009; 123(6): 1493–1500. 25 Merhar S, Tabangin ME, Meinzen-Derr J, Schibler KR. Grade and laterality of intraventricular haemorrahge to predict 18-22 month neurodevelopmental outcomes in extremely low birthweight infants. Acta Pediatrica 2012; 101: 414–418. 26 Maitre NL, Marshall DD, Price WA, Slaughter JC, O’Shea TM, Maxfield C et al. Neurodevelopmental outcome of infants with unilateral or bilateral periventricular hemorrhagic infarction. Pediatrics 2009; 124(6): e1153–e1160. 27 Hintz SR, Slovis T, Bulas D, Van Meurs KP, Perritt R, Stevenson DK et al. Interobserver reliability and accuracy of cranial ultrasound scanning interpretation in premature infants. J Pediatr 2007; 150(6): 592–596.

Supplementary Information accompanies the paper on the Journal of Perinatology website (http://www.nature.com/jp)

Journal of Perinatology (2014), 203 – 208

& 2014 Nature America, Inc.

Outcomes of extremely preterm infants following severe intracranial hemorrhage.

Severe intracranial hemorrhage (ICH) is an important prognostic variable in extremely preterm (EPT) infants. We examined imaging and clinical variable...
310KB Sizes 0 Downloads 0 Views