According to the literature, FGR combined with a polyhydramnios complicates 0.2- 6.0 % of pregnancies [7–9, 15]. In our study we calculated an incidence of 4.6 %. Initial prenatal detection of FGR combined with polyhydramnios was 29.7 weeks. Compared to previously published studies initial detection of both pathologies was 32.7 weeks . As these studies were conducted before 2000, poorer ultrasound technologies and the changed approach of consultation in prenatal tertiary centers might explain the differences [8, 9].
Chromosomal anomalies represented the highest proportion in our study population (41.8 %). Trisomy 18 was found in 73.4 % of fetuses assigned to the CA group, whereas structural chromosomal aberration was seen in 4.7 %. These results are consistent with the analysis of Sickler et al. and Snijders et al., both reporting nearly 40 % showing chromosomal anomalies [8, 15]. Trisomy 18 was also the most frequently detected numerical chromosomal aberration in their studies (66.7 %, 47.1 %), whereas structural aberration were rarely seen .
Further Sickler et al. reported 92 % (n=36) of the examined fetuses showing further anomalies, including skeletal dysplasia in six cases and arthrogryposis in one case . Comparable to our results we also observed further anomalies in 90.9 %. In relation to the recorded anomalies, Sickler et al. described 24 out of 38 cases to have cardiac abnormalities (63.2 %). Differently, congenital diaphragmatic hernia (CDH) were the most common isolated malformations with 41.7% (10/24) in our study. Cardiac malformations were represented with 29.2 % (7/24). The difference in results may be explained by the analysis of the entire study collective by Sickler et al. including also fetuses with chromosomal aberrations, as trisomy 18 . Regarding perinatal mortality rate Sickler et al. estimated it to be 59 % . Whereas other published data, such as Furman et al. found the mortality rate to be 9.6 %, after exclusion of congenital malformations . These results correspond to our data with an overall mortality rate of 64.7 % and 7.1 % after eliminating congenital malformation.
In accordance with the American College of Obstetricians and Gynecologists (ACOG) and the Royal College of Obstetricians and Gynecologists (RCOG), prenatal genetic diagnostic testing was recommended in all our cases, when the combination of both pathologies was observed in the midtrimester or further structural anomalies were present [2, 20]. We found a 9.2 % and 3.4 % incremental yield by CMA and NGS compared to conventional karyotype. This is consistent with the result of Borrell et al., but lower compared to the findings of Hay et al., who described an incremental yield by CMA of 16 % [21–23]. These differences might be explained by several reasons: 1) different inclusion criteria of the study populations; 2) CMA is currently not a coverage of the public health insurance companies in Germany and therefore not routinely performed; 3) not all patients with an unremarkable karyotype had received a CMA. Estimation of the effective additional benefit remains difficult and might be higher. Nevertheless, our findings, indicates that CMA/ NGS should be recommended, especially in cases with FGR combined with polyhydramnios and normal karyotypes, as they might provide incremental yield (up to nearly 13.4 %) of detecting chromosomal abnormalities and change parental counselling.
In our study we were able to demonstrate that prenatal detection of FGR combined with polyhydramnios should lead the investigator to think of different etiological groups. Fetuses assigned to the CA group were identified significantly earlier (p= 0.01) compared to all the other EGs. In addition, we found that an extremely increased AFI combined with a mild FGR (8.0 ± 8.6) and unremarkable Doppler examination should lead the investigator to think of a possible musculoskeletal disorder. In non-anomalous fetuses, FGR and polyhydramnios were detected later (33 weeks) and of a milder severity compared to the other subgroups (p=0.03). Further, we found significant increased Doppler values of the uterine artery (p=0.04). Affected patients were predominantly primigravida, a preeclampsia occurred in 7.1 %. In 50.0 % maternal diseases were able to explain the occurrence of this rare combination. In the remaining 50.0 % fetal neurological impairments were found in 14.3 %, endocrinological impairments in 7.1 % and in 7.1% skeletal abnormality was suspected, although it is uncertain whether they should be better categorized in the MSD group instead.
The strength of our current study is that it represents one of the largest cohort of prenatally diagnosed FGR combined with polyhydramnios and the first evaluation of the spectrum of the different etiologies and examination of them for significant disparities. Further, first time follow up of the survivors was evaluated over a period from 1 to 14 years (mean 5.8 years).
Nevertheless, our study has certain limitations. First, it is a retrospective analysis. Second, the outcome of fetuses assigned to the H group, was requested in a not standardized procedure, so neurodevelopmental follow up was not clearly defined. Third, although measurement of AFI is superior for identification of polyhydramnios and was used for our analysis, it might be influenced by the different investigators and thus our results of the subgroup comparisons .
In conclusion prenatal detection of an FGR combined with polyhydramnios represents a rare phenomenon with an extremely high mortality rate. If non-anomalous fetuses are observed, maternal history should be investigated especially for haematological diseases. An early referral to a perinatal tertiary center should be performed and an invasive testing should be recommended. Further genetic testing (CMA/NGS) should be performed, especially if the karyotype is unremarkable. Detailed prenatal neurosonography might be helpful. Based on our data, these patients should receive long-term follow up, with emphasis on endocrinological and neurodevelopmental assessment. Signs of neurodevelopmental or growth delay should be recognized. As these disorders may affect quality of live, early detection is crucial. Nevertheless, further prospective studies are necessary.