Dehydrated Hereditary Stomatocytosis Presenting as Severe Perinatal Ascites and Cholestasis

Dehydrated hereditary stomatocytosis (DHS) is a rare genetic hemolytic anemia that can be associated with pseudohyperkalemia and/or perinatal edema. The edema can range from simple ascites to severe—and even lethal—hydrops in the fetus or infant (1,2). These findings may present together or individually, and otherwise unexplained perinatal ascites has been associated with a diagnosis of DHS later in life (3). Neonatal hepatitis has also been reported with DHS, a finding which may complicate diagnostic approaches (4). Regardless, recognizing that severe fetal or perinatal ascites with or without liver abnormalities may be caused by DHS can aid in early diagnosis and prevent unnecessary diagnostic studies and interventions. We present the case of a newborn male with severe fetal ascites complicated by cholestasis and concern for significant liver disease. Genetic testing identified a mutation consistent with DHS and guided care. CASE The patient was born at 32 weeks via Cesarian section for refractory congenital ascites and fetal distress. He required multiple paracenteses in utero for ascites with 800 to 2300 mL removed each time. After delivery, he continued to have recalcitrant ascites (Fig. 1A) requiring biweekly paracenteses and eventual peritoneal drain placement. The severity of the ascites required intermittent parenteral nutrition (PN) due to feeding intolerance, and respiratory support including mechanical ventilation. Extensive laboratory and imaging evaluation—including abdominal ultrasound, echocardiogram, and lymphangiogram—did not reveal an etiology for the ascites demonstrated on abdominal computed tomography scan (Fig. 2). Evaluation of the ascitic fluid after transfer to our center on day of life 56 showed clear fluid with a normal triglyceride level (21 mg/dL) and an elevated serum-ascites albumin gradient (serum albumin of 2.2 g/dL [reference range 2.8–5] and an unmeasurable ascitic fluid albumin level of <1.0 g/dL), suggestive of a nonperitoneal source (eg, liver disease). At 2 weeks of age he began developing direct hyperbilirubinemia and elevated transaminases. Peak laboratory values were as follows (reference range included): total bilirubin: 7.0 mg/dL (0.0–1.2), direct bilirubin: 6.1 mg/dL (0.0.–0.4), aspartate aminotransferase: 272 U/L (20–77), alanine aminotransferase: 215 U/L (5–50), gamma-glutamyl transferase: 257 U/L (27–210), international normalized ratio 1.22.FIGURE 1: A, Recalcitrant ascites 1 week after birth before eventual drain placement. B, Patient at 12 months of age with resolution of ascites, but "prune-like" belly remaining.FIGURE 2: Computed tomography (CT) abdomen demonstrating large amount of ascites (thin arrows).The increasing cholestasis and recurrent ascites prompted a liver biopsy at 60 days of age. Histopathology showed bile duct proliferation and plugging, consistent with PN-associated liver disease, but also obstructive cholestatic liver disease (Fig. 3). He had ultrasonographic evidence of a normal biliary tree, and percutaneous gallbladder cholangiogram demonstrated a patent extrahepatic biliary system.FIGURE 3: Liver biopsy demonstrating marked cholestatic liver disease with portal expansion, bile ductular proliferation (thin arrows), and bile duct plugging (thick arrow). Slide courtesy of Eugenio Taboada, MD.We pursued whole genome sequencing and found the patient to be heterozygous for a paternally inherited pathogenic variant, c.6058G>A (p.A2020T) in the PIEZO1 gene, a mutation reported in an affected French family and responsible for autosomal dominant DHS (5,6). This mutation was not associated with perinatal edema or ascites in other relatives. Of note, the patient's potassium levels were within normal range throughout his hospitalization. He, however, did receive a red blood cell transfusion for a hemoglobin level of 8 g/dL at 56 days of life. The anemia did not persist and was considered consistent with an anemia of prematurity, although (in hindsight) his underlying DHS may have factored. Establishing the diagnosis of DHS precluded further evaluation of the patient's ascites, as reported cases self-resolve. Treatment focused on maximizing nutritional status, peritoneal drainage, and diuresis. With time and improved enteral tolerance (and decreased exposure to PN), the patient's cholestasis resolved, and the ascites responded to medical therapy. He was discharged home at 4 months of age. Follow-up at 12 months showed no recurrence of the ascites, but a typical "prune belly-like" appearance to the abdomen (Fig. 1B). The family was counseled on the variable phenotype of PIEZO1 mutations, and the need to monitor for late complications, including anemia, gallstones, and iron overload. DISCUSSION First described by Miller et al in 1971 (7), DHS is an autosomal dominant hemolytic anemia characterized by an altered permeability of the erythrocyte membrane to monovalent cations (2). This membranopathy leads to an osmotic water loss with dehydration and destruction of the red blood cells. The degree of hemolysis associated with DHS is variable, ranging from mild to moderate anemia. As a result of this variability, patients with DHS are often not diagnosed until adulthood when they can present with chronic anemia, recurrent jaundice, gallstone disease, thrombosis after splenectomy, or iron overload with hemosiderosis (8). In infants, the presentation of DHS has been described as a pleiotropic syndrome of perinatal edema, anemia, and pseudohyperkalemia—where erythrocyte potassium content moves extracellularly at room temperature, giving falsely elevated laboratory levels (2,8). Perinatal and fetal ascites can be isolated or occur in concert with other findings of DHS (3,4). The pathophysiology of the ascites remains unclear, but generally resolves within the first few months. Liver disease associated with DHS has been suggested as an etiology of the ascites in infants. The histologic findings reported by Rees et al (4) in 2004 are, however, dissimilar from our patient. In the previous report, the liver biopsy showed evidence of hepatitis, fatty changes, and increased fibrosis. Our patient's liver histology showed moderate inflammation, but was marked by bile duct plugging, bile ductular proliferation, and a lack of significant fibrosis. To our knowledge, this would be only the second reported liver biopsy in a DHS patient. But, it is unlikely the findings are related to DHS, and instead more consistent with PN-associated cholestasis. DHS is associated with missense mutations in the PIEZO1 gene, on the long arm of chromosome 16 (1,2,5). The PIEZO1 transmembrane protein forms mechanosensitive channels expressed in several different cell types, including erythrocytes, endothelial cells, and renal tubular epithelial cells. PIEZO1 mutations result in increased ion channel activity and changes in cellular cation concentration (2,5). Although the correlation between these mutations and the associated clinical findings remains largely unclear, recognition that each finding singly or together may be the only clinical indication of the syndrome can lead to more frequent diagnosis. In the case of our patient, the availability of efficient and cost-effective genetic testing allowed for identification of a clinically significant mutation that—especially if performed earlier—could significantly impact the diagnostic and treatment plan for patients with fetal/perinatal ascites.