Pulmonary hypoplasia is a rare and devastating morbidity among newborns, resulting in mortality up to 70% (1, 2). A pathological feature is an arrest in bronchial growth during fetal growth (3). The pathogenesis of pulmonary hypoplasia is generally categorized into three mechanisms, including thoracic compression, lack of fetal breathing movement, and loss of lung fluid (4). Many causal factors might lead to pulmonary hypoplasia, including intrathoracic masses, oligohydramnios, skeletal malformations, neuromuscular malformations, pleural effusion, cardiac lesions, abdominal wall defects, and chromosomal aberrations (2). Congenital cytomegalovirus (CMV) infection occurs in 0.2–2.0% of live-born infants (5), and is a major non-genetic cause of deafness and childhood neurodevelopmental disabilities (6). Therefore, aiming at improvement of fetal/infantile prognosis, Yamada and collaborators organized Japanese Congenital Cytomegalovirus Infection Immunoglobulin Fetal Therapy Study Group, and commenced multicenterded fetal therapy trial of hype-immunoglobulin injection into the peritoneal cavity for symptomatic congenital CMV infections since 2005. And, we have reported a promising results that 41.7% of symptomatic congenital CMV infections infants whose mothers received fetal therapies had no or minimal sequelae (7). Thus, recently, we have actively performed fetal therapy to prenatally diagnosed symptomatic congenital CMV infections with parental informed consent under approval of the institutional ethics boards of the Kobe University Hospital.

Symptomatic cases manifest various features at birth, including small-for-gestational age, microcephaly, thrombocytopenia, liver dysfunction, retinopathy, abnormal brain images, and abnormal auditory brainstem responses (6–10). However, lung complications are uncommon in congenital CMV infections (11).

We describe two cases of congenital CMV infections in patients who suffered from pulmonary hypoplasia due to massive fetal ascites despite fetal therapy and died soon after birth.

Pulmonary hypoplasia is a rare condition, affecting 9–11/10,000 live births (1). The pathophysiology of pulmonary hypoplasia is a reduction in the number of lung cells, airways, and alveoli, resulting in a decrease in organ size and weight. These factors are highly correlated with insufficient gas exchange (4). Diagnosis of pulmonary hypoplasia is based on clinical findings of respiratory distress occurring almost secondary to other fetal anomalies (1, 10). In addition, low-intensity fetal lung on MRI could be the clues of pulmonary hypoplasia (12). There have been no sufficient data regarding its impact on the neurodevelopmental outcome; however, favorable outcomes at 5 years of ages were reported in congenital diaphragmatic hernia, which is often associated with pulmonary hypoplasia (13). The most feasible pathogenesis of pulmonary hypoplasia in our cases is thoracic compression via massive ascites, which occurred from early pregnancy. The greatest impairment of lung development occurs when compression occurs in the lungs during the last trimester (4). The fetal lungs were highly compressed at this time in our cases. Extrinsic thoracic compression usually occurs in oligohydramnios, where the maternal uterine wall directly compresses the fetal thorax (4). Our case 2 suffered from hydrops fetalis, and Page and Stocker reported this as the cause of pulmonary hypoplasia (14); however, their cases were complicated with large pleural effusions or renal dysplasia; two main causes of pulmonary hypoplasia via thoracic compression or loss of lung fluid. Our case 2 showed only mild pericardial effusions but not pleural effusions, and we considered that those could not affect the lung development significantly as space occupying lesions of the chest. Thus, we believe that not hydrops fetalis itself, but its underlying mechanisms to develop pulmonary hypoplasia is more important. In our cases, the volume of ascites was so large that the fetal abdominal wall might not have been able to extend further, resulting in strong compression of the fetal thorax. Although fetal ascites is an uncommon complication of congenital CMV infections, overt systemic disease might occur as hepatosplenomegaly and ascites in the fetus as a result of liver insufficiency (15).

In lung complications of congenital CMV infections, a few patients suffering from persistent pulmonary hypertension have been reported (16–18). However, their etiology was hypothesized to be interstitial pneumonia or vasculitis, and not pulmonary hypoplasia. Similar to our cases, Stocker reported an autopsy case of congenital CMV infection presenting as massive ascites (28% of total body weight) with secondary pulmonary hypoplasia (26% of expected weight) that was diagnosed at postmortem (19). In his study, CMV inclusion bodies were detected within the nucleus and cytoplasm of the epithelial cells of the bile duct. Although we could not perform the autopsy of the lungs due to unavailability of parental consents, fetal MRI and autopsy imaging could add sufficient information regarding pulmonary hypoplasia.

Despite aggressive neonatal resuscitation, we could not rescue our patients. Additionally, multiple fetal therapies failed to prevent development of pulmonary hypoplasia. Fetal treatment with hyper-immunoglobulin was supposed to have neutralizing effects by inhibiting the replication of CMV, or immunomodulatory effects by decreasing the immune cells associated with the production of inflammatory cytokines which can contribute to immune-mediated fetal damages (20–22). We have previously reported that (1) immunoglobulin fetal therapy could decrease CMV DNA copy numbers in fetal ascites and (2) CMV viral load is weaker in fetal ascites than in amniotic fluid in congenital CMV infections (7). With regard to viral load, CMV-DNA copy numbers in amniotic fluid of our two cases are comparable with those in previous reports where fetal therapy caused a good outcome (7). Therefore, we believe that fetal treatment modality, which consisted of ascites removal followed by administration of immunoglobulin or albumin, was effective in controlling viral load, but not effective for controlling the volume of fetal ascites.

Pulmonary hypoplasia due to fetal ascites is a rare, but an important, cause of neonatal death in congenital CMV infections. Further fetal intervention to control fetal ascites should be considered.

Parental written informed consent was given to present cases. The fetal therapy protocol was approved by the institutional ethics boards of the Kobe University Hospital.

KF, IM, and KN managed the patients, contributed to the conception of the study, and drafted the manuscripts; MM, KT, and MD managed the mothers and performed the fetal therapy; and KI and HY critically reviewed the manuscript. All authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer MV and handling editor declared their shared affiliation.