Pediatric Academic Societies and Asian Society for Pediatric Research 2008 Joint Meeting Platform Session Neonatal Medicine: Clinical Trials I Sunday, May 4 7:00am-9:00am Hawaii Convention Center, Ballroom B

Prevalence of cerebral palsy at 3-5 years of age who participated in an RCT for the prevention of intraventricular hemorrhage by indomethacin in Japanese extremely low birthweight infants Shinya Hirano, Masanori Fujimura, Satoshi Kusuda, Hirofumi Aotani. On behalf of the Neonatal Research Network Japan Disclosure: The trial vials (indomethacin and placebo) were manufactured by Banyu, a subsidiary of Merck & Co. Inc., NJ, USA. The research grant from Dept. of Maternal and Child Health, Ministry of Health and Welfare, Japan. Any real or apparent conflict of interest related to the content of the presentation has been resolved. It is our pleasure to present the results of first trial compiled by the Neonatal Research Network Japan. Any real or apparent conflict of interest related to the content of the presentation has been resolved.

BACKGROUND: We reported in 2005 that the prophylactic indomethacin (IND) was effective to reduce the incidence of severe IVH in Japanese ELBW infants in a large RCT (n=419) Adjusted odds ratio (95% CI) 0.37 (0.18-0.77) p =0.007 Adjusted for institution, gestational age, sex, in/out-born, Apgar score at 1 min HYPOSESIS: The reduced incidence of severe IVH by prophylactic IND in neonatal period may result in the reduction of cerebral palsy (CP) at 3-5 years of age We reported in SPR 2005 that the prophylactic indomethacin was effective to reduce the incidence of severe IVH (Grade three and four) in Japanese extremely low birthweight infants in a large randomized controlled trial. The adjusted odds ratios was 0.37, From this result, we hypothesized that the reduced incidence of severe IVH by prophylactic indomethacin in neonatal period may result in the reduction of cerebral palsy at three to five years of age.

SUBJECTS: Infants with birthweight 400-1000 g INTERVENTION : INDOMETHACIN TRIAL SUBJECTS: Infants with birthweight 400-1000 g INTERVENTION : Starting within 6 hours of birth 3 doses of IND or placebo were given with 6 hours continuous intravenous infusion every 24 hours IND was given at the dose of 0.1 mg/kg-wt/dose OUTCOMES : Grade 3 and 4 IVH within 7 days assessed by US. CP at 3-5 years of age SUBJECTS of study were infants with birthweight between 400-1000 g The intervention was to start within six hours of birth, three doses of indomethacin or placebo were given with six hours continuous intravenous Infusion every twenty four hours Indomethacin was given at the dose of 0.1 mg/kg-wt/dose. OUTCOMES are Grade three and four IVH within 7 days, assessed by Ultrasound scan. And cerebral palsy at three to five years of age.

OBJECTIVE: To compare the prevalence of CP at 3-5 years between the IND and placebo groups in a follow-up study of a large RCT in Japan The OBJECTIVE of this study is , To compare the prevalence of cerebral palsy at 3 to 5 years of age between the indomethacin and placebo groups

718 infants assessed for eligibility 50 : Excluded 67 : Not approached 601 infants eligible Stratified factors Institution 　 Gestational Age Sex In-born / Out-born Apgar at 1min 132 : No consent 469 infants were assigned randomly 235 IND group 234 placebo group Seven hundred eighteen infants were assessed for eligibility. And four hundred sixty nine infants were assigned randomly to the trial. two hundred thirty five infants were assigned to the indomethacin group, and two hundred thirty four were to placebo group. Either before or after discharge from the NICU twenty two infants died in indomethacin group and twenty nine infants in Placebo group . One hundred ninety one children in Indomethacin group, and one hundred eighty seven children in placebo group were assessed for CP in the routine follow-up clinic by qualified pediatricians. The follow up rate was 90.4 % as a whole. Death : 22 Drop out, No data : 22 Death : 29 Drop out, No data : 18 191 378 / 418 (90.4%) 187

RESULTS : Baseline Characteristics IND (N=235) Placebo (N=234) p value Birth weight(g) 775.2±130.6 784.0±139.8 NS Apgar score at 5 min 6.7 ±1.9 6.8 ±1.9 Primipara 119 (50.6%) 109 (46.6%) Singleton / Multiple 189 / 46 188 / 46 Vaginal / C-section 64 / 171 68 / 166 Antenatal steroids 103 (43.8%) 102 (43.6%) This table shows the baseline characteristics of infants and mothers for two groups. There was no evidence of differences between two groups in birthweight, Apgar score at 5 min, and Obstetric characteristics, between two groups . Mean±1SD NS: not significant

RESULTS: Outcomes IND Placebo odds ratio p value Death 22 / 235 (95% CI) p value IVH 3 or 4 <7 days 16 / 235 (6.8%) 32 / 234 (13.7%) 0.37 (0.18-0.77) 0.007 Death 22 / 235 (9%) 29 / 234 (12.4%) 0.75 (0.4-1.4) 0.34 CP 31 / 235 (13.2%) 32 / 234 (13.7%) 0.95 (0.54-1.69) 0.82 CP or death 53 / 235 (22.5%) 61 / 234 (26.0%) (0.53-1.29) 0.37 This slide shows the result. Grade 3 or 4 IVH within 7 days was significantly reduced in IND group. Up to the time of final follow up, there were 22 deaths in IND group and 29 in Placebo group. The number of children who either died or showed the sign of CP were 53 in indomethacin and 61 in placebo group. There is no evidence of difference.

Severity of IVH and Prevalence of CP IND Placebo This slide shows the percentage of CP in relation with the severity of IVH. Each pink bar shows the ratio of CP with number of cases and blue is without CP. In IND group on the left, there is a significant correlation between the grade of IVH and the percentage of CP. The severer the IVH, the higher the rate of CP. There is a same trend in the Placebo group, and there is no evidence of difference in the prevalence of CP between the IND and Placebo groups in relation with the severity of IVH. It is well known that periventricular leukomalacia is one of the major cause of CP. Therefore we specified each case of PVL on this figure in the next slide.. CP Without CP

Severity of IVH , PVL, and Prevalence of CP IND Placebo Here the PVL cases are shown with highlighted bars. In IND group with IVH nil or grade 1, there were 6 cases of PVL. In placebo group there were 5 cases. There were no PVL in IVH grade 2 and 3, and in IVH grade4, 3 cases of PVL all together. There could be seen no bias in the distribution of PVL In indomethacin or placebo groups. We therefore can conclude from these data thatthe distributions of CP both in IND and Placebo groups may be similar in relation with the severity of IVH. If this is true we may wonder why the reduction of IVH did not result in the reduction of CP? CP Without CP with PVL

Subgroup Analysis as for “CP or Death” IND Placebo Odds ratio (95% CI) p value Male 25 / 105 31 / 105 0.75 (0.38-1.44) 0.34 Female 28 / 130 40 / 129 0.90 (0.48-1.69) 0.74 400-599g 7 / 22 18 / 27 0.23 (0.05-0.89) 0.015 600-899g 39 / 164 34 / 150 1.06 (0.60-1.86) 0.81 900-999g 7 / 49 9 / 57 0.88 (0.25-2.95) 0.82 GW 22-23w 9 / 23 15 / 22 0.30 (0.07-1.19) 0.051 GW 24-26w 35 / 136 39 / 135 0.85 (0.48-1.51) 0.56 GW 27w- 9 / 76 7 / 77 1.34 (0.41-4.49) 0.57 Outborn 6 / 45 13 / 44 0.36 (0.10-1.19) 0.06 Inborn 47 / 190 48 / 190 0.97 (0.59-1.59) 0.91 Apgar score 1min <3 28 / 97 33 / 96 0.77 (0.40-1.49) 0.41 Apgar score 5min <3 5 / 15 4 / 14 1.25 (0.19-8.38) 0.78 antenatal steroids 19 / 103 18 / 102 1.05 (0.48-2.29) 0.88 RDS 48 / 182 43 / 187 1.19 (0.72-1.98) 0.45 We then made subgroup analysis wondering if we could find any subgroups who might have had a favorable effect on “CP or Death” by the use of IND. In Birthweight 400-599g, IND was effective to reduce CP or death. For those who are 22-23 weeks of gestation, and outborn, the IND group shows the trend to reduce CP or death. Then we analyzed the birthweight subgroups.

Incidence of IVH 3, 4 and “CP or Death “ Stratified by Birth weight IND Placebo Odds ratio(95% CI) 400-599g 2 /22 11 / 27 7 / 22 18 / 27 0.14 (0.01-0.84) 0.23 (0.05-0.89) 600-899g 12 / 164 19 / 150 39 / 164 34 / 150 0.58 (0.25-1.32) 1.06 (0.60-1.86) 900-999g 2 / 49 2 / 57 7 / 49 9 / 57 1.17 (0.08-16.7) 0.88 (0.25-2.95) This slide shows the incidence of IVH grade three and four, and CP or death for each birth weight groups. In 400-599g group the number of CP or death in the IND group was 7 in twenty 22, whereas it was 18 in 27 in the placebo group. The odds ratio is 0.23 and the IND is effective to reduce CP or death. In other birthweight groups there could be seen no effect.

Incidence of IVH 3, 4 and “CP or Death “ Stratified by Gestational Weeks IVH 3, 4 CP or Death IND Placebo Odds ratio (95% CI) 22-23w 2 /23 11 / 22 9 / 23 15 / 22 0.10 (0.01-0.58) 0.30 (0.07-1.19) 24-26w 13 / 136 19 / 135 35 / 136 39 / 135 0.65 (0.28-1.45) 0.85 (0.48-1.51) 27w- 1 / 76 2 / 77 9 / 76 7 / 77 0.5 (0.09-9.84) 1.34 (0.41-4.49) On the gestational weeks of birth, in 22-23 weeks group, the indomethacin was seem to be effective to reduce IVH grade three or four and CP or death. From these results In most premature group of babies; that is either 400-599g birthweight or 22-23 weeks gestation, indomethacin is shown to be effective to reduce both severe IVH and CP or death at 3-5 years of age.

and Gestational Age Groups IVH 3, 4 and Gestational Age Groups N= 5 40 70 86 116 152 P=0.0034 We showed this figure in the previous SPR in 2005. There can be seen a trend that the smaller the gestation the more the preventive effect of Indomethacin to IVH grade 3 or 4. It could be that infants above 25 weeks the indomethacin is not necessary for the prevention of IVH.

and Birthweight Groups Cerebral Palsy and Birthweight Groups N= 49 314 106 odds ratio=0.23 (0.05-0.89) In this presentation we have shown the evidence that the use of prophylactic IND could be indicated only for those who are born between 400-599g.

CONCLUSIONS: For the whole study subjects, there was no evidence of difference in the number of infants with “CP or death” by the prophylactic use of IND compared with placebo The prevalence of CP in each grade of IVH showed no evidence of difference between IND and placebo groups The prevalence of PVL showed no evidence of difference in IND and placebo groups In a subgroup analysis, the prevalence of “CP or death” was significantly lower in infants with birthweight 400-599g in IND group than placebo group Application of prophylactic IND to the smallest group of ELBW 400-599g is shown as effective in preventing the development of CP at 3-5 years We thank Naohiro Yonemoto and Rintaro Mori for kind advice for statistical analysis. In CONCLUSION : 1. For the whole study subjects, there was no evidence of difference in the number of infants with “CP or death” by the prophylactic use of IND compared with placebo. 2. The prevalence of CP in each grade of IVH showed no evidence of difference between IND and placebo groups. 3. The prevalence of PVL showed no evidence of difference in IND and placebo groups. 4. In a subgroup analysis, the prevalence of “CP or death” was significantly lower in infants with birthweight 400-599g in IND group than placebo group. 5. Application of prophylactic indomethacin to the smallest group of Extremely Low Birth Weight 400-599g is shown as effective in preventing the development of CP at 3-5 years We thank Naohiro Yonemoto and Rintaro Mori for kind advice for sta tistical analysis THANK YOU VERY MUCH FOR YOUR ATTENTION.