Early diagnosis and care is achieved but should be improved in infants with Prader-Willi syndrome

Prader-Willi syndrome (PWS) was first described in 1956 by Prader, Labhart and Willi [1]. It is a complex neurodevelopmental genetic disorder which associates severe hypotonia and feeding difficulties with sucking deficit and anorexia in the neonatal period [2]. In the absence of early diagnosis and care, excessive weight gain occurs and several nutritional phases have been described, leading to severe early obesity starting between 3 and 4 years of age [3, 4].

Diagnosis is possible within the first months of life because these infants display severe hypotonia, and neonatologists recognize this sign alone as sufficient to suggest the need for genetic study [5‐7]. Early diagnosis and adequate multidisciplinary care are crucial for the infant’s outcome as they ensure parental guidance and support, including comprehensive advice to prevent obesity, and stimulation of cognitive and adaptive skills [5]. In addition, we very recently reported for the first time the positive effects of an early short course of oxytocin treatment in infants with PWS younger than 5 months [8]. It is likely that the window of opportunity to obtain significant change is quite narrow, particularly regarding early brain plasticity, and delayed diagnosis may mean a missed opportunity for better quality of life in infants with PWS [8].

The genetic defect is a lack of paternal copy expression of some of the genes at the q11–13 locus of chromosome 15. Several genetic subtypes have been reported: deletion (DEL) in about 65% of the cases, maternal uniparental disomy (UPD) in about 30%, imprinting center defect in less than 5%, and rare cases of translocation involving the chromosome 15 q11-q13 region [9, 10]. Genetic diagnosis is now easily available for neonatologists who care for neonates with severe hypotonia.

The aims of this study were to determine the neonatal incidence of PWS in France and address the issues of delayed diagnosis. We also investigated the care of those newborns and infants who were diagnosed soon after birth.

This study was conducted by the French PWS Reference Center (PWSRC), which is coordinated by the Toulouse University Hospital in association with the Neonatology Unit of the Children’s Hospital of Toulouse. We included those infants born in France between 1 January 2012 and 31 December 2013 with a molecular diagnosis of PWS. After several attempts to obtain comprehensive data, all contacts were asked to focus their attention on 2013 to ensure a maximal dataset for this year. The incidence was therefore only calculated for 2013, although the data on infant care were analyzed for 2012 and 2013. The birth incidence was obtained by dividing the number of PWS diagnoses by the number of live births in 2013, according to the French INSEE (National Statistics Institute) database.

Neonatology units and PWS competence centers were contacted through the French Neonatology Society and the French Pediatric Endocrinology and Diabetology Society. Data obtained by the French Prader-Willi Association (FPWA) were also collected. Based on information from the Orphanet website, we also contacted by email all cytogenetic and molecular biology laboratories that perform PWS diagnosis in France to ensure that every prenatal and neonatal infant born during this period was counted. The first three letters of the first and last names and the date of birth identified cases and avoided duplicates.

Neonatologists and pediatricians were asked to report data about their own PWS patients on a clinical report form (CRF) sent by email. The requested information concerned the pregnancy, neonatal period, diagnosis, initial care and family data. Neonates with birth weight and/or height below −2 standard deviation scores (SDS) according to Usher and MacLean’s tables [11] were considered as small for gestational age (SGA). The results are expressed as medians, 10th and 90th percentiles (10pc; 90pc) and/or mean values ± SDS. Statistical analyses were made with Statview software. The Chi2 test compared proportions and the Mann-Whitney U-test compared the quantitative data.

Birth incidence and genetic diagnosis: We show for the first time that the birth incidence of PWS in France in 2013 was about 1/21,000 births. This is higher than the incidences reported in early 2000 in Australia and Belgium [12, 13], which were about 1/27,000 births, but lower than that reported by Lionti in 2014, which was 1/15,830 births between 2003 and 2012 [14].

In our study, the median age at diagnosis was 18 days in 2013. To our knowledge, this is the earliest age reported so far: one month for Smith [12] and 6.5 months for Vogels [13]. Given that the mean delay between blood sampling and results was 10 days, our data indicate that the diagnosis was suspected within the first week of life. The range of values was quite wide, with most patients (90%) diagnosed within the first 3 months and 5 diagnosed later. Diagnosis was delayed for 2 main reasons: either it was not suspected in neonatology or FISH as the first-line genetic test gave false negative results. This last underlines the risk of performing FISH first.

Although our study shows that the mean age at diagnosis was 18 days, which is very good, we also found delayed diagnoses, suggesting that systematic neonatal screening for PWS should be considered to prevent neonates with PWS from being missed. Indeed, the criteria required to set up neonatal screening for PWS have been met: we report a neonatal incidence close to that of congenital adrenal hyperplasia (1/19,000), for which screening has been performed for many years; the diagnostic tools are well known and reliable (methylation profile abnormality positive in 99.99% of cases of PWS); and early diagnosis enables multidisciplinary care to begin quickly, improving prognosis and providing a window of opportunity for early treatment that may alleviate and change the course of the disease [5]. In addition, Bachere et al. [5] noted that early diagnosis and multidisciplinary care reduces the hospital stay and the duration of tube feeding, optimizes screening for endocrine dysfunction, and prevents growth delay and the early onset of obesity.

We confirmed a previous report [15] of a rise in the proportion of UPD, up to 42% compared with older studies. Recently, Beauloye et al. [16] reported a UPD proportion of 51% in children from 1 to 48 months old and, notably, this genetic subtype was associated with advanced maternal age [17]. We found similar data in our study, with maternal age of UPD being 38 years versus 31 years for DEL (p = 0.0001). Moreover, the mean maternal age of the whole cohort was higher than in the general population (32 years for a first child versus 28 years in the general population, according to INSEE data). As already reported [18, 19], the median paternal age was also older for UPD than for DEL cases (38 versus 33 years, p = 0.003).

In our study, the MAP rate was 5.5%, which is twice that reported in the general population (2.9% in 2012 according to the Biomedicine Agency [20]). However, no patient in our study had epigenetic mutation, which is more frequent in MAP-induced imprinting diseases [21].

Challenges for prenatal diagnosis: Incidences of decreased fetal movement (27% in our study) and polyhydramnios (23%) were less frequent than previously reported (88% and 34%, respectively, for Gross et al. [22]; 91% and 36%, respectively, for Geysenbergh et al. [23]). However, even in cases of these prenatal signs, the amniocenteses reported in our study did not lead to the prenatal diagnosis of PWS. Indeed, out of 9 amniocenteses, 4 were performed for polyhydramnios (alone or associated with other signs), while genetic screening for PWS was performed in only one case with FISH, which gave a normal result. No further molecular analysis was made to exclude or confirm the diagnosis. These data suggest that prenatal diagnosis might be optimized by considering PWS testing with methylation profile analysis when polyhydramnios is present.

Challenges for neonatal care: Most neonates needed NGT feeding (84%) for a median of 38 days, as already reported [5]. Only 18/40 infants had a swallowing assessment, and this evaluation should be promoted. It is performed and interpreted by a speech and language expert, who is a critical caregiver for infants with PWS. At the end of hospitalization, only 21/47 infants had a prescription for speech and language therapy, whereas 44/51 infants had prescriptions for physiotherapy and 21/37 for psychomotor therapy. This study underscores the need to improve the evaluation of oral skills and prescription of speech and language therapy for these newborns, in addition to physiotherapy or psychomotor therapy.

Endocrine follow-up: The first visit to a pediatric endocrinologist was made within the first 7 days of life for 84% of the infants diagnosed within the first 3 months. For the whole cohort, the median age at the first visit was 3 months. Endocrine dysfunctions could thus be detected early and GH treatment started within the first year, as recommended [24]. Indeed, 43% of the infants had already started GH treatment at the time of inclusion, with a mean age of 12 months, which is excellent, given the GH effects on muscle and cognitive development in these patients [25].

Our study is the first to assess the birth incidence of PWS in France, at about 1/21,000 births. We confirm the increasing proportion of UPD in conjunction with advanced maternal age. Prenatal and neonatal cases are still sometimes missed, generally because the early signs are not recognized or inappropriate molecular tests are prescribed and may therefore give false negative results. Setting up neonatal PWS screening procedures would prevent these late diagnoses and permit early care and treatment. Comprehensive neonatal care should include a systematic sucking and swallowing assessment in order to optimize feeding and mother-infant interaction.