Hydrops fetalis (HF) is an infrequent foetal disease. Depending on its pathophysiology, there are two types of HF: immune HF and non-immune HF (NIHF), the latter of which may be secondary to multiple structural or physiological abnormalities. Due to the generalised use of anti-D gamma globulin prophylaxis, nearly 90% of cases diagnosed at present have a non-immune aetiology, with an approximate prevalence of 1/1700 to 1/3000 pregnancies.1,2

It is a severe disease with a poor prognosis, with a reported perinatal mortality that ranges between 50% and 98%.3–5 The prenatal and neonatal prognosis, however, vary depending on the underlying aetiology, the possibility of initiating intrauterine treatment, the gestational age at the time of diagnosis and the severity of HF.

Some authors have attempted to describe the prenatal and postnatal prognosis of antenatally diagnosed HF based on the underlying aetiology,6–8 but there is scarce evidence analysing the prognosis of HF overall and based on the possibility of providing IUT. Similarly, few studies have assessed long-term survival in infants product of such pregnancies. Knowing the foetal prognosis and neonatal outcomes depending on the underlying aetiology is essential in order to counsel families in the event of a diagnosis of HF.

In consequence, the objective of our study was to analyse medium- and long-term obstetric and perinatal outcomes for pregnancies with an antenatal diagnosis of HF, with an immune or non-immune aetiology, in our hospital, and their association with the aetiology of HF and the delivered IUT. We also sought to identify prenatal and perinatal outcome predictors that could be useful for antenatal counselling.

We conducted a retrospective observational study in pregnant women that received a prenatal diagnosis of HF between January 1, 2011 and December 31, 2021 who received prenatal and postnatal care at our hospital.

The sample included the pregnancies with a prenatal diagnosis of HF and the live neonates product of these pregnancies if the prenatal and postnatal care was delivered at our hospital during the study period, whether the infants had hydrops at the time of birth or not.

We excluded newborns for whom the data were incomplete due to transfer to a different facility in the prenatal or postnatal period. We also excluded pregnant women with an isolated diagnosis of ascites, pleural effusion or hydrothorax that did not meet the criteria for HF.

Hydrops fetalis was defined as the detection on the foetal ultrasound examination of subcutaneous oedema associated with effusion in a serous cavity or an abnormal fluid collection in two or more serous cavities (pericardium, pleura or peritoneum).

We defined intrauterine treatment (IUT) as any medical or surgical intervention performed during pregnancy with the aim of treating the underlying cause of hydrops or ameliorate its impact on the foetus. The definition comprehended pharmacological treatment (eg antiarrhythmic medication), transfusion of blood products, foetoscopic procedures (such as laser coagulation of placental anastomoses or serous cavity aspiration and drainage) as well as foetal surgery. The aetiologies that can be treated medically are foetal tachyarrhythmias and anaemias. Laser photocoagulation or umbilical cord occlusion was performed in monochorionic pregnancies that meet the criteria for severe twin-twin transfusion syndrome with significant impact on one of the foetuses. Foetal surgery was contemplated in the case of malformations or tumours that could benefit from IUT, and palliative treatment of HF consisted of placement of drainage lines or aspiration of cavities in foetuses in cases in which the fluid accumulation had a significant impact on the foetus.

Neonatal death was defined as death between 0 and 28 days post birth. We defined morbidity as any significant disease, sequela or impairment beyond 1 year post birth. The definition of cardiac morbidity included chronic heart failure, haemodynamically significant structural heart disease, arrhythmia requiring prescription of pharmacological treatment or electrophysiology procedures at discharge or pulmonary hypertension at the time of discharge. The definition of neurologic morbidity included psychomotor delay, moderate to severe motor impairment, hearing loss, blindness, attention-deficit hyperactivity disorder, language disorder, dysphagia or epilepsy. Lastly, the definition of nephrological morbidity included hypertension requiring pharmacological treatment and chronic kidney disease. The followup after discharge took place within the period under study, lasting between 1 and 10 years.

We reviewed the health records of each patient through the electronic health record system of the hospital, collecting obstetric data (maternal characteristics, trimester at diagnosis, ultrasound findings, indication for and type of invasive prenatal test, cause of HF, detected genetic changes, need of IUT), data about the birth (gestational age at birth, birth weight, presence of hydrops at birth, characteristics of the delivery and type of resuscitation required, need of cavity drainage and need of admission in the neonatal intensive care unit [NICU]) and data on neonatal outcomes (type and duration of respiratory support, need of cavity drainage, parenteral nutrition or dialysis, length of stay in the NICU, comorbidities in the medium and long term, death and timing of death in days post birth).

All the information was collected in a database built with the software SPSS, version 25.0. We made an initial descriptive analysis to determine the frequencies, measures of central tendency and measures of dispersion. In the inferential analysis, we used different tests depending on the nature of the variables (χ2 test, Fisher exact test, Student t-test, Mann-Whitney U de Mann-Whitney). We considered P values of less than .05 statistically significant.

The study was approved by the research ethics committee of the hospital (PR[AMI] 46/2023).

Our study showed that HF is associated with a high mortality overall, and particularly with a high prenatal mortality. In addition, the prevalence of postnatal morbidity among live newborns was also substantial. We also found that survival was, overall, higher in pregnancies with HF due to causes for which IUT was available, on account not only of a larger proportion of live births but also of a greater survival at 1 year post birth compared to cases of HF that could not be treated in utero.

The current literature on HF is scarce and the results of different studies are often contradictory. When it came to prenatal mortality, our findings were consistent with most other case series, in which the proportion of live births is of approximately 20%–25%, with the exception of a few series, like the one published by Meng et al.,6 in which the proportion of live births was higher. There is substantial variation between studies in the reported survival of live newborns, which ranges from 25% to 90.4%, and was 70.7% in our cohort.6–8. These differences probably stem from the heterogeneity of the studies, as most only include cases of NIHF, and significant differences in the laws regarding termination of pregnancy in the countries where the studies have been conducted, which could affect study outcomes.

The most frequent underlying causes in our series were chromosomal disorders and gene changes, followed by infection and cardiopathies (including tachyarrhythmias and structural heart diseases). These findings are consistent with the existing literature, although some studies have reported higher proportions of cardiac involvement and lower proportions of genetic disorders.8–10 The latter could be related to advances in genetic and molecular diagnostic techniques in recent years in Spain, which have improved the detection of monogenic disorders associated with NIHF. In 2020, Quinn et al.11 published a systematic review in which they identified 131 genes for which there was strong evidence supporting an association with NIHF. In our sample, single gene changes were identified in 21 patients, most of them variants previously associated with NIHF in the literature.

We also ought to highlight that, despite the advances in diagnostic and genetic techniques,12 there is still a significant percentage of idiopathic HF cases. In our cohort, the underlying aetiology was not identified in 1 out of 7 patients, a percentage that was lower compared to the previous literature (17.8%–34.7%).1,6,13 Differences between centres in the available diagnostic methods could explain the heterogeneity of the results.

In our study, we found a significantly greater percentage of elective termination of pregnancy in pregnant women given a prenatal genetic disorder diagnosis compared to those who did not receive such a diagnosis. Our study is the first to analyse this association and underscores the importance of improving prenatal aetiological diagnosis and the early diagnosis of genetic causes to improve prenatal counselling.

As regards prenatal treatment, one fourth of the patients in our study could receive IUT during the pregnancy, most frequently cavity drainage, intrauterine transfusions and antiarrhythmic medication, findings that were also consistent with the previous literature.1,8 An important finding in our study is that the overall outcomes were better in patients who received IUT compared to those who did not, which is particularly relevant in relation to the prenatal counselling offered to the families.

The causes of HF associated with the greatest survival, foetal as well as neonatal, were infection and arrhythmia. Every published series, including our study, has identified chromosomal and genetic disorders as the aetiologies with the poorest obstetric outcomes, due in part, as noted above, to the greater percentages of ETP and miscarriage associated to these causes. On the other hand, the causes associated with the poorest neonatal outcomes were structural heart diseases and pulmonary hypoplasia, with a survival of less than 20% at 1 year post birth, which was consistent with previous studies.14,15

The published evidence on the postnatal management and morbidity after discharge is scarce and does not allow comparisons between studies. Meng et al.6 reported a higher mortality in their case series compared to ours (77.2% versus 58.6%). Another finding worth highlighting was the considerable long-term morbidity observed in diseases associated with a lower neonatal mortality, such as intrauterine infections, arrythmias and congenital anaemias. This was particularly relevant in the case of intrauterine infections, as two thirds of the patients with this history in our study had significant neurologic morbidity in the long term. However, these findings were not statistically significant due to the limited sample size.

Among the main strengths of our study, its large sample, one of the largest case series published to date analysing obstetric and neonatal outcomes of immune or non-immune HF and the prognosis based on the possibility of delivering IUT. It is also one of the first series analysing both survival and morbidity in these patients in the medium to long term. In addition, our hospital is a reference centre for the prenatal treatment of HF, thereby minimising the risk of selection bias intrinsic to retrospective studies.

However, there are also limitations to our study, chief of which is its retrospective design. In addition, while the sample was large, the size of the groups corresponding to each of the underlying causes of HF was small, limiting the statistical power of the results and the generalization to other populations. On the other hand, the low incidence of HF made it necessary to include patients managed over a long time, and therefore there may have been significant changes both in obstetric and neonatal care during the study period.

Hydrops fetalis is a rare foetal condition associated with a high morbidity and mortality and with a heterogeneous aetiology. Since it is an infrequent and complex condition, it is essential that it is managed with a multidisciplinary approach in tertiary care hospitals to continue advancing the development of techniques allowing an early aetiological diagnosis, offer highly specialised obstetric and neonatal care and conduct prospective studies in larger allowing a more detailed analysis of obstetric outcomes, neonatal survival and long-term morbidity and mortality based on the aetiology of HF, all for the ultimate purpose of improving prenatal counselling for families in the event of diagnosis of HF.

This research project did not receive specific financial support from funding agencies in the public, private or not-for-profit sectors.

The authors have no conflicts of interest to declare.