Polycystic Kidney Disease 4 With Or Without Polycystic Liver Disease

A number sign (#) is used with this entry because polycystic kidney disease-4 with or without polycystic liver disease (PKD4) is caused by homozygous or compound heterozygous mutation in the PKHD1 gene (606702), which encodes fibrocystin/polyductin, on chromosome 6p12.

Description

PKD4 is an autosomal recessive polycystic kidney disease (ARPKD) characterized by enlarged, echogenic kidneys with fusiform dilatation of the collecting ducts. Most patients progress to end-stage renal disease (ESRD), but at varying ages. Patients also have liver disease consisting of dilated biliary ducts, congenital hepatic fibrosis (CHF), and portal hypertension (Caroli disease). The most typical disease expression occurs in neonates and includes a history of oligohydramnios, massively enlarged kidneys, and the 'Potter' sequence with pulmonary hypoplasia that leads to respiratory insufficiency and perinatal death in approximately 30% of affected newborns (summary by Hartung and Guay-Woodford, 2014).

For a discussion of genetic heterogeneity of polycystic kidney disease, see PKD1 (173900).

Clinical Features

Ward et al. (2002) briefly reviewed clinical features and classification of autosomal recessive polycystic kidney disease (ARPKD). The disease presentation of ARPKD is highly variable. In infancy, the disease results in significantly enlarged echogenic polycystic kidneys, with pulmonary hypoplasia resulting from oligohydramnios as a major cause of morbidity and mortality. Liver involvement is detectable in approximately 45% of infants and is often the major feature in older patients. The pathologic findings of collecting-duct ectasia in the kidney and ductal-plate malformation in the liver indicates that the basic defect of ARPKD may be a failure of terminal differentiation in the collecting-duct and biliary systems. The variable clinical presentation led to earlier separation into different clinicopathologic groups: perinatal, neonatal, infantile, and juvenile, depending on the age of presentation and the severity of renal and liver disease (Blyth and Ockenden, 1971), suggesting different genetic entities; see HISTORY. Subsequently, evidence of intrafamilial phenotypic variability (Kaplan et al., 1988; Deget et al., 1995) and genetic linkage studies (Zerres et al., 1994; Guay-Woodford et al., 1995) suggested that a less rigid subdivision is indicated, with allelic, rather than genetic, heterogeneity explaining much of the observed variability (Zerres et al., 1998).

It has long been recognized that the age distribution of cases of polycystic kidneys has 2 peaks, one at birth and one between ages 30 to 60 years. Furthermore, most cases with the later peak show the familial pattern of an autosomal dominant (173900). Kaariainen (1987) collected information in Finland on 82 children treated during the years 1974 to 1983 for polycystic kidney disease. The frequency was about 1 per 8,000 births. Early lethal disease was present in 51, whereas 31 survived for over 28 days. The children came from 69 families. They were divided by family studies into 3 groups: autosomal dominant polycystic kidney disease in 11 families, autosomal recessive polycystic kidney disease in 14, and sporadic cases in 44 families. In 3 of the 'dominant' families, 2 or more sibs had manifestations of PKD neonatally. The majority of the grandparents of the children with the recessive or the sporadic form were born in the same sparsely populated areas in northern, central, and eastern Finland, which suggested that most of the sporadic cases were actually recessive PKD.

One of the notable features of both dominant and recessive PKD in humans is the variability of the phenotype, with respect to both disease progression and extrarenal manifestations. While a significant component of this variance is probably due to genetic heterogeneity, Kaplan et al. (1989) documented marked variability of clinical disease even within kindreds. Kaplan et al. (1988) presented evidence contradicting the view that perinatal, neonatal, infantile, and juvenile forms of autosomal recessive polycystic kidney disease represent 4 discrete entities. They described an instructive family in which a male infant presented at birth with very large cystic kidneys together with portal fibrosis and ductule proliferation in the liver, bilateral pulmonary atelectasis, large atrial septal defect, and a ventricular septal defect. Death occurred at 18 hours. An older sib presented at the age of 16 years with no symptoms because the mother wanted reassurance that the daughter had no condition similar to that of the deceased sib. Blood pressure was normal. The liver and spleen were enlarged. Ultrasonography, radiologic studies, and liver biopsy showed renal tubular ectasia and congenital hepatic fibrosis. Thus, the perinatal and juvenile forms of Blyth and Ockenden (1971) are likely to be opposite ends of the phenotypic spectrum of a single entity.

Zerres et al. (1984) gave a comprehensive review of all types of cystic kidney. They stated that evidence of so-called congenital hepatic fibrosis is 'indispensable for the diagnosis of ARPKD.' Hartung and Guay-Woodford (2014) noted that although histologic biliary abnormalities are a universal feature of ARPKD, clinical expression varies widely. In various cohorts, the proportion of ARPKD patients with imaging findings of liver disease has ranged from 45 to 90%.

Gross cystic dilatation of the intrahepatic biliary tree is usually called Caroli disease (congenital biliary ectasia, or nonobstructive intrahepatic bile duct dilatation; see 600643); its frequent association with ARPKD is well established (Bernstein et al., 1975). Caroli disease is a rare cause of chronic cholestasis and hepatolithiasis in young adults. Ros et al. (1993) reported results of treatment with ursodeoxycholic acid (UDCA) in 12 patients with Caroli disease and intrahepatic stones. The duodenal bile of these patients contained cholesterol crystals which suggested that the stones were cholesterol rich. UDCA led to sustained clinical remission, return to normal liver function, and dissolution of intrahepatic stones on ultrasound in all patients. Tsuchida et al. (1995) described Caroli disease in a brother and sister whose asymptomatic father was shown to have the same disease by CT scan. They suggested that the mode of inheritance in this family and in many other instances is autosomal dominant, not autosomal recessive. The sister had undergone laparotomy at the age of 5 years because of hepatomegaly and mottled radiopacities shown by cholangiography. With medical management she remained in good health for the next 21 years. Mottled radiopacities of the hepatic parenchyma were demonstrated by cholangiography in her 9-year-old brother at the time the diagnosis of Caroli disease was made in the sister. He remained asymptomatic, however, until hematemesis due to esophageal varices occurred 21 years later at the age of 30 years. While their healthy father was at that time shown to have the same disease by CT scan, the mother of the affected sibs had a completely normal intrahepatic biliary tree by intravenous cholangiography. Presumably, there was no cystic involvement of the kidneys in any of the 3 affected members of the family. If Caroli disease is defined simply as polycystic dilatation of intrahepatic bile ducts with hepatic fibrosis, it is likely that not all cases have this disorder as part of either infantile or adult polycystic kidney disease. It is possible that some cases represent a distinct entity of isolated hepatic fibrosis with cysts and that some of these represent a distinct genetic entity in its own right (see 600643). However, Adeva et al. (2006) concluded that relatively isolated congenital hepatic fibrosis and/or Caroli disease are part of the ARPKD spectrum.

Deget et al. (1995) observed 42 children from 20 sibships with ARPKD, pro- and retrospectively, over a mean period of 3.7 years. Using the subclassification of Blyth and Ockenden (1971), they assigned 12 patients to the perinatal, 9 to the neonatal, 13 to the infantile, and 8 to the juvenile category. In 11 of the 20 families, different subtypes were observed among affected sibs; in 7 families, affected sibs belonged to adjacent subgroups, while major intrafamilial differences were observed in 4 families. The phenotypic manifestations ranged from stillbirths to mildly affected adults, while intrafamilial variability of the clinical picture was generally small with multiple allelism as the most likely genetic explanation. Age at death showed gross variation in 8 sibships. Sex influence could not explain the differences in clinical course between sibs.

Blickman et al. (1995) reported the sonographic changes in renal function seen on long-term follow-up of children who had the initial diagnosis of autosomal recessive polycystic kidney disease made in the neonatal period. The evaluation involved 14 children with biopsy evidence of the disease; 9 children who survived the neonatal period were followed up for a mean of 13 years (range, 5-19 years) after diagnosis. They found that kidney size as seen on sonograms did not continue to increase despite the patients' linear growth and maintained normal renal function. Rather, a decrease in kidney size and change in echogenicity occurred, producing a pattern similar to that seen on sonograms of patients with autosomal dominant polycystic kidney disease but without the marked increase in kidney size that occurs in that entity. The authors thought that the changing cystic pattern on follow-up sonograms may have been the reason that previous descriptions have varied and why a decrease in size may not herald deteriorating renal function.

Coffman (2002) commented on the elucidation of autosomal recessive polycystic kidney disease by Ward et al. (2002). They published photographs of gross and microscopic views of a specimen from a patient with ARPKD showing the typical pattern of fusiform, cylindrical channels occupying most of the kidney parenchyma and representing massively dilated terminal branches of collecting ducts. A photomicrograph showed the radially oriented, fusiform cysts characteristic of ARPKD.

Bosch et al. (2003) described clinical and renal function stabilization in a 3-year-old Turkish female who survived neonatal onset without hepatic fibrosis, thereby illustrating that prognostication should be cautious even with severe neonatal presentation.

Guay-Woodford and Desmond (2003) analyzed the largest single group of patients with ARPKD (The ARPKD Clinical Database) which was divided into a younger cohort of 166 patients born after January 1, 1990 (median age 5.4 years, 45.8% detected prenatally) and an older cohort of 43 patients born before 1990 biased for long-term survival (median age 14 years, 5.6% detected prenatally). They found a slower rate of disease progression in the older cohort, as assessed by age of ARPKD diagnosis, as well as age of diagnosis of clinical morbidities. Neonatal ventilation was strongly predictive of mortality as well as an earlier age of diagnosis in those who developed systemic hypertension and chronic renal insufficiency. However, for those who survived the perinatal period, the long-term prognosis for survival was much better than generally perceived; 1- and 5-year survival was approximately 90% for those who survived the first month of life. Portal hypertension was not correlated with age at diagnosis, and only a small subset of patients developed clinically significant periportal fibrosis.

Bergmann et al. (2005) examined the clinical course of 164 neonatal survivors out of 186 ARPKD patients from 126 unrelated families. The mean observation period was 6 years (range, 0-35 years), and the 1- and 10-year survival rates were 85% and 82%, respectively. Chronic renal failure was first detected at a mean age of 4 years, with actuarial renal survival rates of 86% at 5 years, 71% at 10 years, and 42% at 20 years. All but 6 patients (92%) had a kidney length on or above the 97th centile for age. About 75% of the study population developed systemic hypertension. Sequelae of congenital hepatic fibrosis and portal hypertension developed in 44% of patients and were related to age. Positive correlations observed between renal and hepatobiliary-related morbidity suggested uniform disease progression rather than organ-specific patterns. Bergmann et al. (2005) noted that missense changes were more frequently observed among patients with a milder clinical course, whereas chain-terminating mutations were more commonly associated with a severe phenotype.

Adeva et al. (2006) commented that the autosomal recessive form of polycystic kidney disease had generally been considered an infantile disorder with typical presentation of greatly enlarged echogenic kidneys detected in utero or within the neonatal period, often resulting in neonatal demise; however, there was an increasing realization that survivors often thrived into adulthood with complications of ductal plate malformation, manifesting as congenital hepatic fibrosis and Caroli disease, becoming prominent. Adeva et al. (2006) retrospectively reviewed the clinical records, and where possible performed PKHD1 mutation screening, in patients diagnosed with ARPKD or congenital hepatic fibrosis at the Mayo Clinic in the period 1961 to 2004. Of 133 cases reviewed, 65 were considered to meet the diagnostic criteria, with an average duration of follow-up of 8.6 +/- 6.4 years. ARPKD was present in 55 cases and 10 had isolated congenital hepatic fibrosis with no or minimal renal involvement. The patients were analyzed in 3 groups categorized by the age at diagnosis: less than 1 year (22 patients), 1 to 20 years (23 patients), and more than 20 years (20 patients). The presenting feature in the neonates was typically associated with renal enlargement, but in the older groups more often involved manifestations of liver disease, including hepatosplenomegaly, hypersplenism, variceal bleeding, and cholangitis. During follow-up, 22 patients had renal insufficiency and 8 developed end-stage renal disease (ESRD), most from the neonatal group. Liver disease was evident on follow-up in all diagnostic groups but particularly prevalent in those diagnosed later in life. In all, 12 patients died, 6 in the neonatal period, but 86% of patients were alive at 40 years of age. The likelihood of being alive without ESRD differed significantly between the diagnostic groups with 36%, 80%, and 88% survival in the 3 groups, respectively, 20 years after the diagnosis. Considerable intrafamilial phenotypic variability was observed. Mutation analysis was performed in 31 families and at least 1 mutation was detected in 25 (81%), with 76% mutant alleles detected in those cases. Consistent with the relatively mild disease manifestations in this group of patients, most of the mutations were missense (79%) and no case had 2 truncating changes. Mutations were detected in all diagnostic groups, indicating that congenital hepatic fibrosis with minimal kidney involvement can result from PKHD1 mutation. No mutations were detected in 6 cases.

Biochemical Features

Gupta et al. (2000) described a child with ARPKD whose levels of urinary basic fibroblast growth factor (FGF2; 134920) were markedly elevated. Expression of this growth factor has been shown to be increased in a mouse model of human polycystic kidney disease (Nakamura et al., 1993). The child had rapid kidney enlargement following heparin administration for an aortic clot; following right nephrectomy, the urinary levels of FGF2 further increased in response to compensatory growth of the remaining kidney. The authors suggested that urinary FGF2 measurement may serve as a noninvasive marker for the progression of cystic renal disease.

Mapping

Zerres et al. (1994) mapped the gene for autosomal recessive polycystic kidney disease to chromosome 6p21-cen by linkage analysis and Mucher et al. (1994) refined the assignment to 6p21.1-p12. No evidence of genetic heterogeneity was found. The study included 36 families (12 multiplex and 24 simplex) with a total of 49 affected children. The mapping of ARPKD to 6p21-cen was performed in a cohort of families with mostly milder phenotypes. To determine whether severe perinatal ARPKD also maps to 6p, Guay-Woodford et al. (1995) analyzed the segregation of 7 microsatellite markers from this interval in 22 families with a severe phenotype. In most of the affected infants, ARPKD was documented by histopathology. Linkage was confirmed and the ARPKD region was refined to a 3.8-cM interval in 6p21.1-p12. Taken together, the findings indicated that 'despite the wide variability in clinical phenotypes, there is a single ARPKD gene.'

At The Jackson Laboratory, Janaswami et al. (1997) reported a mouse model for polycystic kidney disease; the kat(2J) mutation (so named for the combination of kidney, anemia, and testes phenotype) arose on the C57BL/6J strain. The various pleiotropic effects produced by the mutation were facial dysmorphism, dwarfism, male sterility, anemia, and progressive polycystic kidney disease. Janaswami et al. (1997) mapped the kat(2J) mutation to chromosome 8, using F2 animals obtained from a cross between the parent strain and another subspecies. During this study, it was noted that genetic background had profound effects on the disease phenotype of the mutant F2 individuals. Similar observations had been reported in other murine models such as cpk, pcy, and jck. The most important distinguishing features of KAT2J compared with the other recessive mouse models that are influenced by genetic environment were the broader, systemic nature of the disease and its slower progression. Upadhya et al. (1999) described the genetic mapping of modifier loci on chromosomes 1, 2, and 19 that had a pronounced effect on disease progression in the F2 progeny. The 2 quantitative traits that were used in mapping these modifiers were kidney weight and hematocrit. The chromosome 1 modifier had a significant effect on both traits, whereas the effects of chromosome 2 and chromosome 19 modifiers were apparent only for hematocrit and kidney weight, respectively. In addition to these loci with large main effects, they observed loci on chromosomes 4, 10, and 14 that modulated the effects through apparent epistatic interactions.

Exclusion Studies

In 11 kindreds with autosomal recessive polycystic kidney disease, Wirth et al. (1987) excluded close linkage with the marker 3-prime HVR, which is closely linked to the autosomal dominant form; thus, allelism of the 2 forms was excluded. Ramsay et al. (1988) studied 12 South African families and 8 British families with the autosomal recessive form of polycystic kidney disease. They also found no linkage with the 3-prime HVR marker of the alpha-globin gene cluster.

By linkage analysis in 19 families with autosomal recessive polycystic kidney disease, Zerres et al. (1994) demonstrated that the causative gene is not at the PKD2 locus (613095) on chromosome 4.

Diagnosis

Schmidt et al. (1982) reported a successful experience with prenatal diagnosis of kidney disease by ultrasonography in 23 families.

Zerres et al. (1998) noted that prenatal diagnosis of ARPKD using fetal sonography can be unreliable, especially in early pregnancy. They examined the feasibility and reliability of haplotype-based prenatal testing in pregnancies 'at risk' for ARPKD. In a 27-month period they received 258 inquiries regarding prenatal evaluation and completed analyses in 212 families. At the time of report, 65 prenatal analyses had been performed in 57 families. In 45 of the 57 requesting families, the index children were deceased and DNA was extracted from paraffin-embedded tissue. Homozygosity for the disease-associated haplotypes was found in 18 fetuses. In 12 of these fetuses, pathoanatomical examination demonstrated typical ARPKD changes consisting of dilated collecting ducts and the characteristic hepatic ductal plate malformation. In 2 fetuses, these changes were detected as early as 13 weeks of gestational age. Forty-three fetuses were either heterozygous or homozygous for a nondisease-associated haplotype and all those who were born were phenotypically unaffected at birth. In 4 cases, no genotypic prediction was possible because a recombination event occurred between the flanking markers. Three of these pregnancies were terminated and necropsy of the fetuses confirmed ARPKD; 1 fetus was carried to term and showed no abnormalities at birth. An absolute prerequisite for prenatal diagnosis by this method is an accurate diagnosis of ARPKD in previously affected sib(s).

Population Genetics

Zerres et al. (1998) estimated the incidence of ARPKD at 1 in 20,000.

Ramsay et al. (1988) stated that although incidence figures were not available, casual observation indicated that autosomal recessive PKD is more common in Afrikaans-speaking families in South Africa than in other South African populations or in populations elsewhere in the world (Thomson and Isdale, 1984). It was considered to be the result of founder effect and, despite variability in clinical presentation (Isdale et al., 1973), it was considered to be a homogeneous gene defect in the Afrikaans-speaking population.

In Spain, Martinez-Frias et al. (1991) found a frequency of infantile polycystic kidney disease of 1.41 per 100,000 live births. The total frequency of well-recognized autosomal recessive syndromes was 10.3 per 100,000 live births, giving a total carrier frequency of 1:49.

Molecular Genetics

In a genetic analysis of a rat with recessive polycystic disease (Pkd), Ward et al. (2002) found an orthologous relationship between the rat locus and the ARPKD region in humans and identified a candidate gene. The mutation was characterized in the rat, and screening of the 66 coding exons of the human ortholog (PHKD1; 606702) in 14 probands with ARPKD revealed 6 truncating and 12 missense mutations (see, e.g., 606702.0001-606702.0006); 8 of the affected individuals had compound heterozygous mutations.

Bergmann et al. (2003) stated that 29 different PKHD1 mutations had been described. They reported mutation screening in 90 ARPKD patients and identified mutations in 110 alleles, a detection rate of 61%. Thirty-four of the detected mutations had not previously been reported. Mutations were found to be scattered throughout the gene without evidence of clustering at specific sites. Approximately 45% of the changes were predicted to truncate the protein. All missense mutations were nonconservative, with the affected amino acid residues found to be conserved in the murine polyductin ortholog. One recurrent mutation, T36M (606702.0001), was thought to represent a mutation hotspot and was found in a variety of populations. Two founder mutations, R496X (606702.0007) and V3471G (606702.0008), comprised approximately 60% of PKHD1 mutations in the Finnish population. All patients carrying 2 truncating mutations displayed a severe phenotype with perinatal or neonatal demise.

Animal Model

Mandell et al. (1983) described a congenital polycystic kidney mutation (cpk) in the mouse which was thought to be a useful model of autosomal recessive polycystic kidney disease in the human. Cowley et al. (1987) studied MYC (190080) oncogene expression in a mouse model of autosomal recessive PKD. Exceedingly high levels of MYC mRNA were observed in cystic kidneys, suggesting that abnormal MYC expression may play a role in the pathogenesis of this form of PKD. (The Myc gene is on mouse chromosome 15.) Davisson et al. (1991) assigned the cpk locus to mouse chromosome 12 by study of the inheritance of this mouse mutation in crosses segregating a Robertsonian translocation between chromosomes 12 and 14. From homology of synteny, they suggested that the homologous gene in humans may be located in either 2p25-p23 or 7q22-q31. Simon et al. (1994) mapped a phenotypically similar cpk/cpk mutation in the mouse to a region of chromosome 12 homologous to the 2p25-p24 region in the human. The mouse mutation causes bilateral cystic dilatation of the renal collecting tubules and leads to rapidly progressive renal insufficiency in affected homozygotes.

B6 cpk/cpk homozygous mice do not express the biliary ductal plate malformation (DPM) as part of the cpk phenotype. However, homozygous mutants from outcrosses to other strains, e.g. DBA/2J (D2), CD-1, BALB/c, and Mus mus castaneus (CAST), express the DPM. Guay-Woodford et al. (2000) found that all F2-generation cpk/cpk pups from intercrosses of B6 with either CAST or D2 expressed both the typical renal cystic disease and the DPM, although the severity was quite variable. Genetic analysis of focal biliary cysts in both aged D2 +/cpk and F1 heterozygotes demonstrated loss of heterozygosity at the cpk interval, thus supporting a loss-of-function model for biliary cysts. The authors concluded that the cpk allele contains an inactivating mutation which disrupts tubuloepithelial differentiation in the kidney and biliary tract; expression of the biliary lesion is modulated by genetic background; and the specific biliary phenotype is determined by whether loss of function of the cpk gene occurs as a germline or a somatic event.

The variability of the PKD phenotype documented in human (Kaplan et al., 1989) is also evident in mouse models for the disorder. Moyer et al. (1994) generated a line of transgenic mice that contained an insertional mutation causing a phenotype similar to human autosomal recessive polycystic kidney disease. Homozygotes displayed a complex phenotype that included bilateral polycystic kidneys and an unusual liver lesion resembling that in the human disorder and involving primarily the bile ducts and ductules. Through use of the transgene as a molecular marker, Moyer et al. (1994) cloned and characterized the mutant locus. A candidate polycystic kidney disease gene was identified whose structure and expression were directly associated with the mutant locus. A cDNA derived from this gene predicted a peptide containing a motif that was originally identified in several genes involved in cell cycle control. The protein contained 10 copies of an internally repeated 34-amino acid sequence referred to as the tetratricopeptide repeat (TPR). By the study of somatic cell hybrids and interspecific backcross analyses, the mutant locus was mapped to mouse chromosome 14 between the T-cell receptor alpha (see 186880) and Rb (614041) loci. The chromosomal assignment ruled out allelism with pcy (Nagao et al., 1995) and cpk, polycystic kidney mutations that map to chromosomes 9 and 12, respectively. Complementation testing with the bpk mouse also ruled out allelism with that locus. Therefore, it appeared likely that the insertional mutation in the transgenic mouse represented a previously uncharacterized PKD mutation.

Atala et al. (1993) described the mouse mutation 'juvenile cystic kidneys' (jck), which has an intermediate phenotype relative to cpk and pcy. In jck homozygous mice, enlarged kidneys can be palpated between 5 and 7 weeks of age. Focal renal cysts are evident by histologic analysis at 3 days after birth, and cystic change proceeds progressively thereafter. No histologic abnormalities are found in any other organs, including the liver. By linkage studies, Iakoubova et al. (1995) mapped the jck locus to mouse chromosome 11, thus confirming that it is nonallelic with cpk and pcy. The severity of polycystic kidney disease in intercross progeny between C57BL/6J and DBA/2J was significantly more variable than that found in the parental C57BL/6J strain, suggesting that a modifier locus or loci introduced from DBA/2J affected expression of jck. Two regions, one from DBA/2J on chromosome 10 and a second from C57BL/6J on chromosome 1, were found to be associated with inheritance of a more severe PKD phenotype. The inheritance of a C57BL/6J-related locus in association with severe disease showed a maximal quantitative trait locus (QTL) analysis lod score of 16.8.

Liu et al. (2002) showed that a mutation in the Nek8 gene (609799) causes the jck phenotype in mice.

The cpk/cpk mouse model of autosomal recessive PKD has been extensively studied in the C57BL mouse strain. The lack of extrarenal pathology appeared to make it a less than completely satisfactory model for human infantile PKD. Gattone et al. (1996) backbred the cpk gene onto CD1 mice. The development of PKD in CD1 mice homozygous for the cpk gene appeared to be slightly more rapid but otherwise comparable to that seen in inbred C57BL/6J mice. Extrarenal manifestations of the cpk gene were evident in the CD1 strain and included cysts of pancreatic, common bile, and major hepatic ducts. Intrahepatic bile ducts also showed focal dilations. Older heterozygous CD1-cpk/+ mice developed renal (proximal tubular) cysts and prominent liver cysts. Thus Gattone et al. (1996) concluded that genetic background appears to influence the expression of the cpk gene and the expression in the CD1 strain is more similar to that in the human. The occurrence of cystic lesions in human heterozygotes should be investigated.

Omran et al. (2001) showed homology between the pcy locus on mouse chromosome 9 and the human NPHP3 locus (see 608002). Woo et al. (1997) identified 2 major modifier loci of polycystic disease progression in mice with the pcy mutation. The study was prompted by the fact that unlike the uniform disease progression in inbred animals, polycystic kidney disease progression in human families can be highly variable. Modifier genes were identified by study of an intercross between DBA/2-pcy/pcy and Mus m. castaneus. The investigators observed large differences in PKD severity in the cross. In addition, 23 of 800 phenotypically normal mice were pcy/pcy genotypically. One modifier locus, MOP1, was located on mouse chromosome 4 and a second, MOP2, on mouse chromosome 16.

A common feature of the development of ARPKD in humans and mice is a distention of the renal collecting tubules caused by localized proliferation and aberrant secretion of epithelial cells. The expanding structures develop into cysts that are filled with fluid containing biologically active ligands for the epidermal growth factor receptor (EGFR; 131550), such as EGF and TGF-alpha. The EGFR, normally localized at the basolateral surfaces of the collecting tubule epithelium, becomes mislocalized to the apical surface on the cells lining cystic structures. This mislocalization of EGFR is a common end point associated with several different forms of polycystic kidney disease that are initiated by mutations in different genes. To determine whether the increased activity of EGFR in the affected kidneys is a functional event that is directly part of the disease pathway of renal cyst formation, Richards et al. (1998) used a genetic approach to introduce a mutant EGFR with decreased tyrosine kinase activity into a murine model of ARPKD. They found that the modified form of EGFR could block the increase in EGFR-specific tyrosine kinase activity that normally accompanies the development of renal cysts, and this correlated with an improvement in kidney function and a substantial decrease in cyst formation in the collecting ducts. The results suggested that drugs that target the tyrosine kinase activity of EGFR may have therapeutic potential in ARPKD.

Bukanov et al. (2006) reasoned that the dysregulated cell cycle may be the most proximal cause of cystogenesis in murine models of autosomal recessive polycystic kidney disease, and that intervention targeted at this point could provide a significant therapeutic benefit for polycystic kidney disease. Bukanov et al. (2006) showed that treatment with the cyclin-dependent kinase (CDK; see 116953) inhibitor R-roscovitine does indeed yield effective arrest of cystic disease in jck and cpk mouse models of PKD. Continuous daily administration of the drug was not required to achieve efficacy; pulse treatment provided a robust, long-lasting effect, indicating potential clinical benefits for a lifelong therapy. Molecular studies of the mechanism of action revealed effective cell-cycle arrest, transcriptional inhibition, and attenuation of apoptosis. Bukanov et al. (2006) found that roscovitine is active against cysts originating from different parts of the nephron, a desirable feature for the treatment of autosomal dominant polycystic kidney disease, in which cysts form in multiple nephron segments. Bukanov et al. (2006) concluded that their results indicate that inhibition of CDK is an effective approach to the treatment of polycystic kidney disease.

History

Early Classification Schemes for Polycystic Kidney Disease

Three types of cystic kidneys in newborns, infants and children were distinguished by Lundin and Olow (1961). In type I the kidneys are oversized and spongy. The liver and pancreas may show fibrosis and/or cystic change. 'Potter's face' ('squashed' nose, micrognathia, large, floppy, low-set ears) is present in most or all. The Potter face resembles that of a child with his face pressed to a window pane. Lundin and Olow (1961) found 9 cases of type I among 21 sibs. When these figures were treated by the method of Weinberg, the corrected figure of 6 affected in 27 sibs was arrived at (a satisfactory agreement with the ratio expected of a recessive trait). Type II also has large kidneys but is characterized by more abundant connective tissue than in type I. Type III has hypoplastic kidneys. In type II, familial aggregation has been observed, but the evidence for recessive inheritance is not complete.

Carter (1974) summarized a clinicopathologic study by Blyth and Ockenden (1969). Childhood polycystic disease fell into 4 classes according to age of onset, clinical course, proportion of renal tubules involved, and degree of hepatic fibrosis. All four groups, termed perinatal, neonatal, infantile and juvenile, were thought to be recessive. The type was consistent within any one family. Occasionally, the 'adult' dominant form presented in childhood.

Potter (1972) referred to type I cystic kidney as tubular gigantism. This was found in only 2 infants (brothers) among 110,000 born at her hospital. She stated further: 'Neither the pulmonary hypoplasia often responsible for death of infants with renal agenesis nor the facies characteristic of absence of intrauterine renal function occur in these infants.' Potter (1972) referred to type II cystic kidney as early ampullary inhibition and indicated that it is not inherited. It may be unilateral. Potter facies and early death occur when it is bilateral. Potter (1972) referred to type III cystic kidney as combined ampullary and interstitial abnormality. This is the variety that occurs in adults (and occasionally presents symptoms in childhood) and is known as 'polycystic kidneys.' Potter's type IV cystic kidney is that produced by intrauterine urethral obstruction. Obviously Potter's numbering system differs sharply from that of other writers cited here.

Among the 10 surviving children of a Druze couple related as second cousins, Naveh et al. (1980) observed a son with congenital hepatic fibrosis (CHF) and congenital heart disease, a daughter with only congenital hepatic fibrosis, and a second daughter with only congenital heart disease. Three other sibs probably had a small ventricular septal defect, and another probably had mild pulmonary valve stenosis. Shunt was performed in each sib with CHF to relieve portal hypertension and hypersplenism. By electron microscopy, hepatocytes showed giant mitochondria with large laminar inclusions. The propriety of classifying this under the forms of polycystic kidney can be questioned; only about half the cases of CHF have cystic disease of the kidneys. The Potter renofacial syndrome is, of course, not a nosologic entity but rather the consequence of severe oligohydramnios which can result from any of many congenital abnormalities of the kidney or urinary tract.