Autosomal recessive polycystic kidney disease (ARPKD) involves cystic renal collecting ducts and biliary dysgenesis. Perinatal presentation is typical and 30-50% of affected neonates die due to respiratory insufficiency. In ARPKD, early nephrogenesis proceeds normally. Initial cysts involve medullary, then cortical collecting ducts, with increased renal echogenicity and oligohydramnios. The latter rarely occurs before 20 weeks gestational age(GA). The ARPKD locus maps to chromosome (chr) 6p12 and linkage analysis indicates that a single gene is involved (Genomics, in press). Haplotype-based prenatal diagnosis of ARPKD is feasible in “at risk” families.

We performed prenatal genetic analysis in a family with presumed ARPKD. Three previous siblings had enlarged, echogenic kidneys in utero. In two, autopsies were performed and ARPKD was the final diagnosis. In the test fetus, the chr 6p haplotypes were distinct from the two previous siblings and the fetus was predicted to be unaffected. However, by 22 wks GA, enlarged, echogenic kidneys were detected and the pregnancy was terminated. Our review of the fetal autopsy revealed marked cystic dilatation of the renal collecting ducts as well as small glomerular cysts. The nephrogenic zone was thin and interrupted. Medullary development was poor. Re-review of the three previous cases revealed oligohydramnios before 20 wks GA in two of the three; similar renal histopathology in one sibling at 22 wks GA; and no biliary dysgenesis in the other sibling at 33 wks GA. Taken together, the clinical and pathologic data in this sibship were not consistent with ARPKD and this novel R-PKD was not linked to chr 6p. Given the phenotypic similarities between this family's R-PKD and several mouse PKD models, we extended our haplotype analyses to evaluate the homologous human intervals for the mouse cpk (chr 2p25-24), orpk (chr 13q12), and bpk (chr 10q21) PKD loci. These analyses excluded linkage between this family's R-PKD and human cpk or orpk. In contrast, all three affected siblings studied shared the same chr 10q21 haplotypes. This association was significant with a chi-squared value of 9.33 (df 2; P = 0.0094).

We conclude: 1) this family has a novel R-PKD phenotype; 2) haplotype-based testing excludes linkage to the chr 6p ARPKD locus and implicates the human homologue of the mouse bpk PKD gene on chr 10q21. These data provide the first evidence that homologous genes may operate in common molecular pathways in human PKD and mouse PKD models.