To examine the relationship between chorionicity and neurodevelopmental outcomes in very low birth weight twins at 3 years of age.
A retrospective cohort study of 3538 twins who were admitted to 91 tertiary perinatal centers in the Neonatal Research Network of Japan between 2003 and 2012.
In a comparison of the followed-up 796 monochorionic twins and 786 dichorionic twins, the overall rate of neurodevelopmental impairment was higher in monochorionic twins; specifically, the rate of disability in the language-social area of the Japanese standardized developmental test was higher in monochorionic twins than in dichorionic twins (adjusted odds ratio, 1.55; 95% confidence interval, 1.07–2.25; P = 0.02).
Chorionicity was associated with neurodevelopmental outcome (especially for language and social development) in a large cohort of very low birth weight twins who had a high rate of perinatal morbidity and neurodevelopmental impairment.
Twin gestation is associated with an increased risk of perinatal morbidity and mortality compared with single gestation. This is mainly attributed to higher rates of low birth weight (BW) and earlier gestational age (GA) at delivery, which are also the major risk factors for neurodevelopmental impairment (NDI) . Additionally, other unique characteristics of twin gestation such as fetal growth restriction, maternal complications, and chorionicity [1,2,3,4] can affect neurodevelopmental outcome.
Twins are classified as monochorionic (MC) or dichorionic (DC) on the basis of chorionicity. MC twins have a higher risk of perinatal morbidities than DC twins owing to vascular anastomoses of the placental surface and/or placental territory discordancy. According to these structural features, MC twins are at risk for complications associated with hemodynamic instability such as twin-to-twin transfusion syndrome (TTTS), selective intrauterine growth restriction, and intrauterine death of a co-twin [5,6,7]. Several recent studies have also shown that MC twins have a higher incidence of adverse outcomes, including neurodevelopmental disability, than DC twins [7, 8]; however, previous studies have primarily evaluated term and late-preterm infants who had a relatively low risk for NDI.
In this study, we investigated the potential association between type of chorionicity and neurodevelopmental outcome in a large cohort of very low birth weight (VLBW) twins at 3 years of age.
We performed a retrospective cohort study of twin infants who were registered in the Neonatal Research Network of Japan (NRNJ), a multicenter registry established in 2003 with a grant from the Ministry of Health, Labor, and Welfare. This database records infants with BW ≤ 1500 g who are born in or transferred to participating hospitals within 28 days of birth. The database included information on maternal and neonatal characteristics, neonatal intensive care unit (NICU) interventions, morbidity, mortality, and follow-up data. These data were collected from each hospital, as previously described [9,10,11]. We analyzed data from twin infants admitted to tertiary NICUs at 91 hospitals between 2003 and 2012.
To assess the effect of growth discordance between twin infants on long-term outcomes, eligible twin pairs met the following criteria: (1) identical infant characteristics (GA, birth year, birth hospital), (2) identical mother characteristics (age, fetal number, chorionicity, gravidity, parity, antenatal steroid administration, presence of diabetes mellitus, presence of pregnancy-induced hypertension, and other maternal morbidities), (3) different birth order, and (4) identical sex in cases of MC infants. The BW discordance rate was calculated using the following formula: (larger twin weight − smaller twin weight)/larger twin weight × 100. The prenatal diagnosis of TTTS was performed in accordance with standard ultrasound criteria . Chorionicity was determined by ultrasound assessment or gross examination and histopathological examination. There was no information on TTTS treatment or the diagnosis of amnionicity in the database. Twin pairs with one or both stillborn infants were not included in this study as they were not eligible for registration in the NRNJ. Infants with congenital abnormalities, twins of unknown chorionicity, infants who were born alive but died in the delivery room, and infants who did not match as twin pairs were excluded.
We used the following clinical definitions. Small for gestational age was defined as a BW below the 10th percentile of the population reference BW . Clinical chorioamnionitis was defined by clinical findings such as maternal fever, leukocytosis, and local pain. Antenatal steroid use was defined as the administration of any dose of a corticosteroid to accelerate fetal maturity. Non-reassuring fetal status was diagnosed on the basis of persistent bradycardia or recurrent decelerations in heart rate. Preterm premature rupture of membranes was defined as rupture of the fetal membranes before the onset of regular uterine contractions before 37 weeks of gestation. Respiratory distress syndrome was diagnosed on the basis of clinical and radiological findings. Chronic lung disease was diagnosed when an infant received supplemental oxygen at 36 weeks postmenstrual age. Intraventricular hemorrhage was diagnosed on sonography, using the classification proposed by Papile . Necrotizing enterocolitis was defined in accordance with a Bell’s classification stage of II or greater . Sepsis was defined as culture-proven septicemia or bacteremia. Cystic periventricular leukomalacia was diagnosed using cranial sonography or magnetic resonance imaging at 2 weeks of age or later.
All infant-related information was collected anonymously and dissociated from individual data. This study was approved by the internal review board of Tokyo Women’s Medical University. The database was registered as UMIN000006961. All participating hospitals are listed in the acknowledgments.
Neurodevelopmental assessments were performed at 3 years of chronological age in each participating hospital in accordance with a comprehensive follow-up protocol [10, 16, 17]. Cerebral palsy (CP) was defined as a non-progressive and non-transient central nervous system disorder characterized by abnormal muscle tone in at least one extremity and abnormal control of movement and posture . Hearing impairment was defined on the basis of the use of a hearing aid. Visual impairment was defined as blindness in one or both eyes. Cognitive function was assessed using the Kyoto Scale of Psychological Development (KSPD) test, which was standardized for Japanese children in 2001 , and the mean ± standard deviation developmental quotient (DQ) was 100.6 ± 13.4. The KSPD enabled an independent assessment of postural-motor, cognitive-adaptive, and language-social areas. The full-scale DQ was strongly correlated with the corresponding composite Bayley III score. Developmental delay was defined as a DQ score < 70, which is equivalent to a Bayley III cognitive score of < 85 . If the KSPD was not administered, physicians at each hospital estimated the presence of developmental delay using a clinical examination (designated as “clinical delay”). NDI was defined by any of the following: CP, full-scale DQ < 70, visual impairment, hearing impairment, or clinical delay.
The baseline characteristics of the MC and DC twin groups were compared using independent Student’s t tests or Fisher’s exact tests with Bonferroni correction, as appropriate. Odds ratios (ORs) and 95% confidence intervals (95% CIs) were obtained from logistic regressions using the DC twin group as a reference. On the basis of a previous study, we considered the following variables as confounders of NDI: GA, BW, sex, Apgar score at 5 min, growth discordance rate, non-reassuring fetal status, place of birth (in-born or out-born), maternal age, parity, gestational diabetes mellitus, pregnancy-induced hypertension, preterm premature rupture of the membrane, mode of birth, clinical chorioamnionitis, and antenatal steroid use. A two-sided P value < 0.05 was considered to indicate statistical significance in all tests, and all statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University), which is a graphical user interface for R (The R Foundation for Statistical Computing, version 3.0.2) .
A total of 7119 twin infants were enrolled from 91 tertiary perinatal centers during the study period. Figure 1 shows the outline of subject selection for this analysis. We excluded 494 infants owing to unknown chorionicity (n = 492) and unreliable data (n = 2). In the remaining 6625 infants, we restricted our analysis to infants assumed to be from twin pairs (n = 3696); of these, 17 (0.5%) were born alive but died in the delivery room and 141 (3.8%) had congenital abnormalities and were subsequently excluded. Thus, our total sample included 3538 infants (1767 MC twins and 1771 DC twins). In the total sample, 269 infants died (146 (0.8%) MC twins and 123 (0.7%) DC twins) and 3269 were eligible to participate in the follow-up assessment. Follow-up data at 3 years were available for 1582 infants (48.4% of survivors). Table 1 summarizes the characteristics of infants who were followed-up and lost to follow-up in each chorionicity. The follow-up rates in the MC and DC groups were 49.1% and 47.7% (P = 0.421), respectively. In both groups, followed-up infants had lower BW, earlier GA, and a higher rate of primipara than infants who were not followed-up. On comparison of data from MC twins to those from DC twins who were followed up at 3 years, both GA and BW were not found to be significantly different between groups. BW discordance, the incidence of small-for-gestational age, non-reassuring fetal status, and neonatal morbidities were significantly higher in MC twins than in DC twins. Mothers of DC twins were more likely to be older, primipara, and to have preterm premature rupture of membranes and chorioamnionitis. Among MC twins, 201 (25.3%) infant cases were complicated by TTTS.
Table 2 shows the follow-up outcomes of study infants at 3 years of age. The KSPD was administered to 1169 infants (73.9% of followed-up infants), and the rate of overall DQ < 70 was non-significantly higher in MC twins than in DC twins (20.8% vs. 17.1%, P = 0.109). In the subscale analysis of KSPD, the rate of infants with language-social DQ < 70 was significantly higher in MC twins than in DC twins (crude OR 1.39, 95% CI 1.02–1.90, P = 0.037; adjusted OR 1.55, 95% CI 1.07–2.25, P = 0.02). However, this difference was not observed in comparisons of DC twins with MC twins without TTTS (Supplemental Table 1).
Among infants who did not undergo KSPD testing (279 infants), 39 (14%) were assumed to have clinical delay (22 (15.8%) MC twins and 17 (12.1%) DC twins). Of these, 20 infants had neither CP nor visual/hearing impairment (8 MC infants and 12 DC infants). Of 1497 infants assessed, 92 (12.2%) MC twins and 72 (9.7%) DC twins had CP. The incidence of visual and hearing impairment was 0.9%. Among infants who were fully evaluated (including KSPD testing), a total of 143 (24.2%) MC twins and 114 (19.7%) DC twins had NDI (crude OR 1.30, 95% CI 0.984–1.720, P = 0.065). Overall, 358 (23.9%) of 1582 infants had NDI. Although there was a higher rate of adverse neurological complications (except for visual impairment) in MC twins compared to DC twins, there was no statistically significant between-group difference in the NDI rate (196 (24.6%) MC twins and 162 (20.6%) DC twins; crude OR 1.25, 95% CI 0.989–1.590, P = 0.062; adjusted OR 1.21, 95% CI 0.917–1.60, P = 0.178).
We examined the effect of chorionicity on long-term neurodevelopmental outcomes in VLBW twins registered in the NRNJ between 2003 and 2012. We did not identify a significant effect of chorionicity on composite NDI outcomes at 3 years of age; however, MC twins tended to have a higher rate of neurological disability than DC twins, with a significantly higher rate of abnormal language-social DQ on the KSPD even after adjusting for several potential confounders.
Several studies have shown that MC twins have a higher prevalence of neurologic morbidity than DC twins because of disorders unique to MC placentation. Minakami et al.  reported 1-year follow-up data for 44 MC and 164 DC twins and an increased risk of adverse outcomes in MC twins, primarily due to TTTS. Similarly, Adegbite et al.  found that neurologic morbidity at 2 years of age was sevenfold higher in preterm MC twins than in DC twins as a result of chronic TTTS, discordant BW, and co-twin death. Moreover, MC twins were generally delivered earlier and had lower BW, and these studies did not adjust for perinatal confounding factors such as GA. In contrast to these studies, Hack et al.  reported that there was no significant between-group difference in neurodevelopmental outcome in a prospective cohort study of 182 MC twins and 189 DC twins matched for BW and GA; however, the cohort included more mature infants, most of whom had normal developmental status. In our study, we examined a cohort of VLBW twins with a relatively high rate of NDI (20–25%) and observed a tendency for a higher rate of neurodevelopmental disability in MC twins than in DC twins, particularly for the language-social area of the KSPD.
In a study assessing the effect of intrauterine growth on cognitive outcome in term and near-term MC twins, prenatal growth restriction was significantly and negatively associated with verbal intelligent quotient . In a study of single preterm infants , exposure to suboptimal nutrition during perinatal life was associated with cognitive impairment, especially in language-based skills. These reports suggest that neural areas associated with verbal and language functions are vulnerable to suboptimal nutrition. Given the observation of a significant higher incidence of disability in the area of language and socialization in MC twins than in DC twins, it can be hypothesized that the differences were related to unequal placental sharing (e.g., selective intrauterine growth restriction or TTTS), consistent with higher BW discordance and incidence of small-for-gestational age in MC twins than in DC twins. In line with this hypothesis, subgroup analysis with the exception of TTTS infants suggested that there was no statistical difference in the outcome between MC and DC twins.
TTTS is a serious complication in MC twins that is associated with high perinatal morbidity and mortality. Accordingly, studies have examined the utility of therapeutic interventions such as fetoscopic laser photocoagulation (FLP). Previous studies in Japan have shown that FLP for TTTS improves survival without major neurological complications . Additionally, a Japanese study of VLBW twins found that, in a sample of 55 MC twins, 26 (47.3%) had TTTS and 11 (20.0%) were treated with FLP . While we did not investigate the effectiveness of FLP for TTTS or its effects on our outcomes, given the unavailability of data, future studies should consider the effects of FLP treatment on MC–DC twin comparisons.
Intrauterine death of a co-twin is a complication specific to MC twins. In a systematic review of 17 studies, the risk of NDI in surviving co-twins was greater for MC twins than for DC twins (OR 4.07, 95% CI 1.32–12.51) . In our study, the NRNJ database did not include information about co-twin death. Additionally, we only included twin pairs with both infants live-born and registered them in the database in order to assess the effects of growth discordance. Therefore, we did not include any surviving co-twin in our cohorts, precluding our ability to assess the influence of co-twin death on long-term neurological outcomes. This may partly account for our inability to detect a significant difference in composite NDI outcomes between the MC and DC twin groups.
Our study had several important limitations. First, there was a relatively low rate of follow-up in our cohort (48.4%). With regard to the assessment of long-term outcomes, loss to follow-up may have resulted in some selection bias, as followed-up infants were more likely to be born earlier to primiparous mothers than those lost to follow-up, leading to an overestimation of NDI at 3 years of age. Second, our data set lacked baseline characteristics associated with developmental state such as parental education and socioeconomic status. These factors are known to have effects on cognitive outcomes and are important to consider as potential covariates when examining long-term outcomes. Third, we assumed twin pair status on the basis of the most reliable combination of infants and maternal characteristics; however, we cannot exclude the possibility that some twin pairs were inferred incorrectly in this study.
In conclusion, we demonstrate that VLBW twins are at increased risk of neonatal morbidities related to prematurity and have adverse neurological outcomes in 20–25% of cases. We identified a non-significantly higher rate of overall NDI in MC twins than in DC twins, and found that MC twins were significantly more likely to have verbal disabilities and impaired socialization. These findings suggest that chorionicity is related to neurological development at 3 years of age in VLBW twins.
Institutions enrolled in the study of the Neonatal Research Network, Japan, were as follows: Sapporo City General Hospital, Asahikawa Kosei General Hospital, Engaru-Kosei General Hospital, Kushiro Red Cross Hospital, Obihiro-Kosei General Hospital, Tenshi Hospital, NTT Higashinihon Sapporo Hospital, Nikko Memorial Hospital, Nayoro City General Hospital, Sapporo Medical University, Asahikawa Medical University, Aomori Prefectural Central Hospital, Iwate Medical University, Iwate Prefectural Ofunato Hospital, Iwate Prefectural Kuji Hospital, Iwate Prefectural Ninohe Hospital, Sendai Red Cross Hospital, Akita Red Cross Hospital, Tsuruoka Municipal Shonai Hospital, Yamagata University, Yamagata Prefectural Central Hospital, Fukushima Medical University, Takeda General Hospital, Fukushima National Hospital, Tsukuba University, Tsuchiura Kyodo Hospital, Ibaraki Children’s Hospital, Dokkyo Medical University, Jichi Medical University, Ashikaga Red Cross Hospital, Gunma Children’s Medical Center, Kiryu Kosei General Hospital, Fuji Heavy Industries Health Insurance Society Ota Memorial Hospital, Gunma University, Saitama Children’s Medical Center, Nishisaitama-chuo National Hospital, Saitama Medical University Saitama Medical Center, Kawaguchi Municipal Medical Center, Jichi Medical University Saitama Medical Center, Asahi General Hospital, Chiba Kaihin Municipal Hospital, Kameda Medical Center, Tokyo Women’s Medical University Yachiyo Medical Center, Juntendo University Urayasu Hospital, Tokyo Metropolitan Children’s Medical Center, Tokyo Women’s Medical University, Aiiku Hospital, Nihon University Itabashi Hospital, National Center for Global Health and Medicine, Tokyo Medical University, Teikyo University, Showa University, Japan Red Cross Medical Center, National Center for Child Health and Development, Tokyo Metropolitan Otsuka Hospital, Toho University, Tokyo Metropolitan Bokuto Hospital, Tokyo Jikei Medical University, Tokyo Medical and Dental University, Saint Luku’s International Hospital, Juntendo University, Sanikukai Hospital, Katsushika Red Cross Hospital, Yokohama Rosai Hospital, Yokohama City University Medical Center, St. Marianna University School of Medicine Hospital, Kanagawa Children’s Medical Center, Tokai University, Kitazato University, Odawara Municipal Hospital, Nippon Medical School Musashi Kosugi Hospital, Saiseikai Yokohamashi Tobu Hospital, National Hospital Organization Yokohama Medical Center, Yamanashi Prefectural Central Hospital, Nagano Children’s Hospital, Shinshu University, Iida Municipal Hospital, National Hospital Organization Shinshu Ueda Medical Center, Saku General Hospital, Niigata University, Niigata Prefectural Central Hospital, Niigata Municipal Hospital, Nagaoka Red Cross Hospital, Koseiren Takaoka Hospital, Toyama Prefectural Central Hospital, Toyama University, Ishikawa Medical Center for Maternal and Child Health, Kanazawa Medical University, Kanazawa Medical Center, Fukui Prefectural Hospital, Fukui University, Gifu Prefectural General Medical Center, National Hospital Organization Nagara Medical Center, Takayama Red Cross Hospital, Seirei Hamamatsu Hospital, Shizuoka Saiseikai Hospital, Shizuoka Children’s Hospital, Hamamatsu Medical University, Numazu Municipal Hospital, Yaizu City Hospital, Fujieda Municipal General Hospital, Nagoya Red Cross Daini Hospital, Nagoya University, Nagoya Red Cross Daiichi Hospital, Toyohashi Municipal Hospital, Nagoya City West Medical Center, Anjo kosei Hospital, Tosei General Hospital, Komaki Municipal Hospital, TOYOTA Memorial Hospital, Okazaki Municipal Hospital, Konan Kosei Hospital, National Mie Central Medical Center, Ise Red Cross Hospital, Yokkaichi Municipal Hospital, Otsu Red Cross Hospital, Shiga University of Medical Science Hospital, Nagahama Red Cross Hospital, Uji Tokushukai Hospital, The Japan Baptist Hospital, Kyoto University, Kyoto Red Cross Daiichi Hospital, National Maizuru Medical Center, Fukuchiyama City Hospital, Kyoto Prefectural University of Medicine Hospital, Kyoto City Hospital, Mitsubishi Kyoto Hospital, Yodogawa Christian Hospital, Osaka Medical Center and Research Institute for Maternal and Child Health, Osaka University, Takatsuki General Hospital, Kansai Medical University, Osaka City General Hospital, Osaka City Sumiyoshi Hospital, Aizenbashi Hospital, Toyonaka Municipal Hospital, National Cerebral and Cardiovascular Center, Kitano Hospital, Saiseikai Suita Hospital, Chifune Hospital, Bellland General Hospital, Rinku General Medical Center, Osaka Red Cross Hospital, Yao Municipal Hospital, Osaka General Medical Center, Osaka City University, Hyogo Prefectural Kobe Children’s Hospital, Kobe University, Kakogawa West City Hospital, Saiseikai Hyogoken Hospital, Kobe City Medical Center General Hospital, Hyogo College of Medicine Hospital, Himeji Red Cross Hospital, Toyooka Public Hospital, Hyogo Prefectural Awaji Medical Center, Nara Medical University, Wakayama Medical University, Tottori Prefectural Central Hospital, Tottori University, Shimane Prefectural Central Hospital, Matsue Red Cross Hospital, Kurashiki Central Hospital, Tsuyama Central Hospital, Kawasaki Medical School Hospital, National Hospital Organization Okayama Medical Center, Okayama Red Cross Hospital, Hiroshima City Hiroshima Citizens Hospital, Hiroshima Prefectural Hospital, Hiroshima University, Tsuchiya General Hospital, National Hospital Organization Kure Medical Center, Yamaguchi University, Yamaguchi Grand Medical Center, Tokushima University, Tokushima Municipal Hospital, Kagawa University, National Hospital Organization Kagawa Children’s Hospital, Matsuyama Red Cross Hospital, Ehime Prefectural Central Hospital, Kochi Health Science Center, St. Mary’s Hospital, National Kyushu Medical Center, Kurume University, Kitakyushu Municipal Medical Center, University of Occupational and Environmental Health, Fukuoka University, Kyushu University, Iizuka Hospital, National Hospital Organization Kokura Medical Center, National Hospital Organization Saga Hospital, National Hospital Organization Nagasaki Medical Center, Kumamoto City Hospital, Kumamoto University, Oita Prefectural Hospital, Almeida Memorial Hospital, Nakatsu Municipal Hospital, Miyazaki University, National Hospital Organization Miyakonojo Medical Center, Kagoshima City Hospital, Imakiire General Hospital, Okinawa Prefectural Nanbu Medical Center & Children’s Medical Center, Okinawa Prefectural Chubu Hospital, Naha City Hospital, and Okinawa Red Cross Hospital