Genotype–phenotype correlations in fetuses and neonates with autosomal recessive polycystic kidney disease

The prognosis of autosomal recessive polycystic kidney disease is known to correlate with genotype. The presence of two truncating mutations in the PKHD1 gene encoding the fibrocystin protein is associated with neonatal death while patients who survive have at least one missense mutation. To determine relationships between genotype and renal and hepatic abnormalities we correlated the severity of renal and hepatic histological lesions to the type of PKHD1 mutations in 54 fetuses (medical pregnancy termination) and 20 neonates who died shortly after birth. Within this cohort, 55.5% of the mutations truncated fibrocystin. The severity of cortical collecting duct dilatations, cortical tubule and glomerular lesions, and renal cortical and hepatic portal fibrosis increased with gestational age. Severe genotypes, defined by two truncating mutations, were more frequent in patients of less than 30 weeks gestation compared to older fetuses and neonates. When adjusted to gestational age, the extension of collecting duct dilatation into the cortex and cortical tubule lesions, but not portal fibrosis, was more prevalent in patients with severe than in those with a non-severe genotype. Our results show the presence of two truncating mutations of the PKHD1 gene is associated with the most severe renal forms of prenatally detected autosomal recessive polycystic kidney disease. Their absence, however, does not guarantee survival to the neonatal period. The prognosis of autosomal recessive polycystic kidney disease is known to correlate with genotype. The presence of two truncating mutations in the PKHD1 gene encoding the fibrocystin protein is associated with neonatal death while patients who survive have at least one missense mutation. To determine relationships between genotype and renal and hepatic abnormalities we correlated the severity of renal and hepatic histological lesions to the type of PKHD1 mutations in 54 fetuses (medical pregnancy termination) and 20 neonates who died shortly after birth. Within this cohort, 55.5% of the mutations truncated fibrocystin. The severity of cortical collecting duct dilatations, cortical tubule and glomerular lesions, and renal cortical and hepatic portal fibrosis increased with gestational age. Severe genotypes, defined by two truncating mutations, were more frequent in patients of less than 30 weeks gestation compared to older fetuses and neonates. When adjusted to gestational age, the extension of collecting duct dilatation into the cortex and cortical tubule lesions, but not portal fibrosis, was more prevalent in patients with severe than in those with a non-severe genotype. Our results show the presence of two truncating mutations of the PKHD1 gene is associated with the most severe renal forms of prenatally detected autosomal recessive polycystic kidney disease. Their absence, however, does not guarantee survival to the neonatal period. Autosomal recessive polycystic kidney disease (ARPKD) is characterized by the association of bilateral renal cystic disease and congenital hepatic fibrosis (CHF). ARPKD can involve a wide spectrum of clinical phenotypes, correlated in part with the age at presentation.1Zerres K. Mucher G. Becker J. et al.Prenatal diagnosis of autosomal recessive polycystic kidney disease (ARPKD): molecular genetics, clinical experience, and fetal morphology.Am J Med Genet. 1998; 76: 137-144Crossref PubMed Scopus (190) Google Scholar,2Guay-Woodford L.M. Desmond R.A. Autosomal recessive polycystic kidney disease: the clinical experience in North America.Pediatrics. 2003; 111: 1072-1080Crossref PubMed Scopus (160) Google Scholar The most severe forms of the disease, accounting for about 40% of cases, are detected early during gestation by ultrasonography that shows tremendously enlarged echogenic kidneys and oligohydramnios. If medical pregnancy termination is not performed, most of these infants die in the perinatal period from respiratory insufficiency due to pulmonary hypoplasia. But over half of the patients are detected late during pregnancy or are undetected prenataly and survive the perinatal period. More than half of them require kidney transplantation before reaching 20 years of age, whereas others have preserved renal function until adulthood.1Zerres K. Mucher G. Becker J. et al.Prenatal diagnosis of autosomal recessive polycystic kidney disease (ARPKD): molecular genetics, clinical experience, and fetal morphology.Am J Med Genet. 1998; 76: 137-144Crossref PubMed Scopus (190) Google Scholar, 2Guay-Woodford L.M. Desmond R.A. Autosomal recessive polycystic kidney disease: the clinical experience in North America.Pediatrics. 2003; 111: 1072-1080Crossref PubMed Scopus (160) Google Scholar, 3Zerres K. Rudnik-Schoneborn S. Deget F. et al.Autosomal recessive polycystic kidney disease in 115 children: clinical presentation, course and influence of gender.Acta Paediatr. 1996; 85: 437-445Crossref PubMed Scopus (143) Google Scholar, 4Roy S. Dillon M.J. Trompeter R.S. et al.Autosomal recessive polycystic kidney disease: long-term outcome of neonatal survivors.Pediatr Nephrol. 1997; 11: 302-306Crossref PubMed Scopus (146) Google Scholar, 5Capisonda R. Phan V. Traubuci J. et al.Autosomal recessive polycystic kidney disease: outcomes from a single-center experience.Pediatr Nephrol. 2003; 18: 119-126Crossref PubMed Scopus (263) Google Scholar CHF complications such as bleeding esophageal varices and cholangitis, may occur as soon as early childhood or during adulthood, sometimes years after renal transplantation.6Shneider B.L. Magid M.S. Liver disease in autosomal recessive polycystic kidney disease.Pediatr Transplant. 2005; 9: 634-639Crossref PubMed Scopus (61) Google Scholar ARPKD is caused by mutations in the PKHD1 gene that encodes the protein fibrocystin expressed in the primary cilia and the basal body of renal and bile ducts epithelial cells.7Onuchic L.F. Furu L. Nagasawa Y. et al.PKHD1, the polycystic kidney and hepatic disease 1 gene, encodes a novel large protein containing multiple immunoglobulin-like plexin-transcription-factor domains and parallel beta-helix 1 repeats.Am J Hum Genet. 2002; 70: 1305-1317Abstract Full Text Full Text PDF PubMed Scopus (359) Google Scholar, 8Ward C.J. Hogan M.C. Rossetti S. et al.The gene mutated in autosomal recessive polycystic kidney disease encodes a large, receptor-like protein.Nat Genet. 2002; 30: 259-269Crossref PubMed Scopus (550) Google Scholar, 9Ward C.J. Yuan D. Masyuk T.V. et al.Cellular and subcellular localization of the ARPKD protein; fibrocystin is expressed on primary cilia.Hum Mol Genet. 2003; 12: 2703-2710Crossref PubMed Scopus (248) Google Scholar, 10Bergmann C. Senderek J. Sedlacek B. et al.Spectrum of mutations in gene for autosomal recessive polycystic kidney disease (ARPKD/PKHD1).J Am Soc Nephrol. 2003; 14: 76-89Crossref PubMed Scopus (146) Google Scholar, 11Wang S. Luo Y. Wilson P.D. et al.The autosomal recessive polycystic kidney disease protein is localized to primary cilia, with concentration in the basal body area.J Am Soc Nephrol. 2004; 15: 592-602Crossref PubMed Scopus (117) Google Scholar, 12Zhang M.Z. Mai W. Li C. et al.PKHD1 protein encoded by the gene for autosomal recessive polycystic kidney disease associates with basal bodies and primary cilia in renal epithelial cells.Proc Natl Acad Sci USA. 2004; 101: 2311-2316Crossref PubMed Scopus (129) Google Scholar, 13Wilson P.D. Polycystic kidney disease.N Engl J Med. 2004; 350: 151-164Crossref PubMed Scopus (582) Google Scholar, 14Torres V.E. Harris P.C. Mechanisms of disease: autosomal dominant and recessive polycystic kidney diseases.Nat Clin Pract Nephrol. 2006; 2: 40-55Crossref PubMed Scopus (228) Google Scholar Presently, 303 mutations of PKHD1 have been reported in ARPKD patients (http://www.humgen.rwth-aachen.de/). Several studies have shown that truncating mutations were present in about 55% of the most severe cases—medical pregnancy termination or early demise patients—but only in about 20% of live patients.10Bergmann C. Senderek J. Sedlacek B. et al.Spectrum of mutations in gene for autosomal recessive polycystic kidney disease (ARPKD/PKHD1).J Am Soc Nephrol. 2003; 14: 76-89Crossref PubMed Scopus (146) Google Scholar, 15Bergmann C. Senderek J. Kupper F. et al.PKHD1 mutations in autosomal recessive polycystic kidney disease (ARPKD).Hum Mutat. 2004; 23: 453-463Crossref PubMed Scopus (112) Google Scholar, 16Bergmann C. Senderek J. Schneider F. et al.PKHD1 mutations in families requesting prenatal diagnosis for autosomal recessive polycystic kidney disease (ARPKD).Hum Mutat. 2004; 23: 487-495Crossref PubMed Scopus (57) Google Scholar, 17Bergmann C. Senderek J. Windelen E. et al.Clinical consequences of PKHD1 mutations in 164 patients with autosomal-recessive polycystic kidney disease (ARPKD).Kidney Int. 2005; 67: 829-848Abstract Full Text Full Text PDF PubMed Scopus (200) Google Scholar, 18Sharp A.M. Messiaen L.M. Page G. et al.Comprehensive genomic analysis of PKHD1 mutations in ARPKD cohorts.J Med Genet. 2005; 42: 336-349Crossref PubMed Scopus (72) Google Scholar, 19Losekoot M. Haarloo C. Ruivenkamp C. et al.Analysis of missense variants in the PKHD1-gene in patients with autosomal recessive polycystic kidney disease (ARPKD).Hum Genet. 2005; 118: 185-206Crossref PubMed Scopus (37) Google Scholar, 20Adeva M. El-Youssef M. Rossetti S. et al.Clinical and molecular characterization defines a broadened spectrum of autosomal recessive polycystic kidney disease (ARPKD).Medecine (Baltimore). 2006; 85: 1-21Crossref PubMed Scopus (167) Google Scholar, 21Furu L. Onuchic L.F. Gharavi A. et al.Milder presentation of recessive polycystic kidney disease requires presence of amino acid substitution mutations.J Am Soc Nephrol. 2003; 14: 2004-2014Crossref PubMed Scopus (89) Google Scholar, 22Bergmann C. Kupper F. Dornia C. et al.Algorithm for efficient PKHD1 mutation screening in autosomal recessive polycystic kidney disease (ARPKD).Hum Mutat. 2005; 25: 225-231Crossref PubMed Scopus (45) Google Scholar But no study ever analyzed the relationship between the severity of histological renal and hepatic lesions and the genotype. The aim of our study was to investigate the genotype–phenotype correlations in a cohort of 74 fetuses or early demise patients with ARPKD, by analyzing the relations between the histological severity of renal and hepatic lesions and the type of PKHD1 mutations. According to inclusion criteria, dilatation of medullar collecting ducts, radially oriented, was present in the 73 available kidney specimens. Portal fibrosis and increased number and dilatation of bile ducts were present in the 61 available liver specimens. The severity of renal lesions was evaluated according to five criteria: (i) The extension in the cortex of the dilatation of collecting ducts, which was graded as either absent, or partial if present only in the inner part of the cortex, with preservation of the nephronic zone and the superficial rows of developing nephrons, or global if affecting the whole cortex up to the subcapsular areas (Figure 1). (ii) The presence of degenerative changes of cortical tubules (proximal and distal convoluted tubules) that may be dedifferentiated, atrophic or reduced in number. According to their extension and severity, lesions were classified as grade 1: absent of significant changes or mild tubular lesions mostly seen in the deep cortex; grade 2: moderate but diffuse tubular lesions; grade 3: massive reduction in the number of nephrons with complete disappearance of proximal tubules (Figure 2a to f). (iii) The absence or presence of proximal tubule dilatation with preservation of a normal epithelium (Figure 2g). (iv) The appearance of the glomeruli: they may be normal or present unspecific changes within interstitial fibrotic areas, from retraction to collapse, often surrounded by a thickened Bowman's capsule. The lesions were classified as absent, partial or global (Figure 2a to h). (v) The change in the cortical interstitium: fibrosis or fibro-oedema was either absent, focal, or diffuse (Figure 2i).Figure 2Categorization of cortical lesions. (a–f) Grading of cortical tubule lesions with (a, d): grade 1 (patient 68), despite some degree of autolysis, tubule morphology is preserved; (b, e): grade 2 (patient 6), epithelial dedifferentiation and atrophy of some cortical tubules leading to irregular dilatation; (c, f): grade 3 (patients 70 and 12), cortical tubules are scanty and dedifferentiated. (g) dilatation of proximal tubules without epithelial lesions (patient 72). (a, b, d–h) grading of glomerular lesions with (a, b, d, e): normal glomeruli (patients 68 and 6); (g) modest retraction of the glomeruli (patient 72); (f, h) moderate to complete retraction of the glomeruli (patient 12). (i) focal cortical interstitial fibrosis (patient 15). Histological sagittal sections of kidneys. Hematoxylin and eosin safran staining. Bar—a, b, c=300 μm; d, e, f, g, i=80 μm, h=150 μm. *Proximal convoluted tubule; °distal convoluted tubule. For information on patients, see Supplementary Table S1.View Large Image Figure ViewerDownload (PPT) In the liver, the severity of lesions was categorized according to the importance of portal fibrosis, classified as grade 1: presence of mild fibrosis, only touching the largest portal tracts; grade 2: fibrosis of all portal tracts, whatever their size; grade 3: massive fibrosis leading to marked enlargement of all portal tracts with fibrous extensions bridging adjacent portal areas (Figure 3). Extension of collecting duct dilatation in the cortex was absent in three cases (4%), partial in 35 cases (48%) and global in 35 cases (48%). Cortical tubule lesions were grade 1 in 19 cases (26%), grade 2 in 29 cases (40%), and grade 3 in 25 cases (34%). Proximal tubule dilatation was observed in 26 cases (37%). Glomerular lesions were present in 44 cases (62%) and cortical fibrosis in 24 cases (34%) (Table 1 and Supplementary Table S1). Portal fibrosis was grade 1 in 5 cases (8%), grade 2 in 41 cases (67%), and grade 3 in 15 cases (25%) (Table 1 and Supplementary Table S1).Table 1Relationships between gestational age and severity of renal and hepatic lesionsNGA ≤30 weeksGA >30 weeksP-valueRenal lesions Collecting duct dilatation in the cortex73<10-3 Absent2 (6)1 (3) Partial26 (72)9 (24) Global8 (22)27 (73) Cortical tubule lesions730.02 Grade 113 (36)6 (16) Grade 216 (44)13 (35) Grade 37 (19)18 (49) Proximal tubule dilatation71<10-3 No16 (44)29 (83) Yes20 (56)6 (17) Glomerular lesions71<10-3 Absent20 (56)7 (20) Partial11 (31)9 (26) Global5 (14)19 (54) Cortical fibrosis710.08 Absent28 (78)19 (54) Focal6 (17)9 (26) Diffuse2 (6)7 (20)Portal fibrosis610.002 Grade 13 (12)2 (6) Grade 222 (84)19 (54) Grade 31 (4)14 (40)Abbreviation: GA, gestational age.Results are expressed as numbers (percentages).For definition of lesions: see Results, Renal and hepatic histology. Open table in a new tab Download .doc (.12 MB) Help with doc files Supplementary Table S1 Abbreviation: GA, gestational age. Results are expressed as numbers (percentages). For definition of lesions: see Results, Renal and hepatic histology. The degree of extension of collecting duct dilatation within the cortex was significantly correlated with the severity of cortical tubule lesions, glomerular lesions and cortical fibrosis. Proximal tubule dilatation was the only lesion that was significantly more frequent when collecting duct dilatation in the cortex was absent or partial compared with global (Table 2). There was also a correlation, close to significance, between the extension of collecting duct dilatation within the cortex and the grade of portal fibrosis (Table 2).Table 2Relationships between the degree of extension of collecting duct dilatation in the cortex and the severity of cortical tubule lesions, proximal tubule dilatation, glomerular lesions, cortical fibrosis and portal fibrosisCollecting ducts dilatation in the cortexNAbsentPartialGlobalP-valueCortical tubule lesions73<10-3 Grade 13 (100)16 (46)0 Grade 2018 (51)11 (31) Grade 301 (3)24 (69)Proximal tubule dilatation71<10-3 No014 (41)31 (91) Yes3 (100)20 (59)3 (9)Glomerular lesions71<10-3 Absent3 (100)23 (68)1 (3) Partial011 (32)9 (26) Global0024 (71)Cortical fibrosis71<10-3 Absent3 (100)30 (88)14 (41) Focal04 (12)11 (32) Diffuse009 (26)Portal fibrosis600.09 Grade 11 (50)2 (7)2 (7) Grade 21 (50)22 (79)17 (57) Grade 304 (14)11 (37)Results are expressed as numbers (percentages).For definition of lesions: see Results, Renal and hepatic histology. Open table in a new tab Results are expressed as numbers (percentages). For definition of lesions: see Results, Renal and hepatic histology. The degree of extension of collecting duct dilatation within the cortex, the severity of cortical tubule and glomerular lesions significantly increased with gestational age (GA). A similar trend was observed between the severity of cortical fibrosis and GA. Proximal tubule dilatation was the only lesion that was significantly more frequent when GA was ≤30 weeks compared with >30 weeks. The grade of portal fibrosis also significantly increased with GA (Table 1). The impact of GA on histological lesions is illustrated in Figure 4, showing renal and hepatic lesions in two consecutive affected foetuses of the same family (patient 54), at 27 weeks and 13 weeks of GA, respectively. A total of 128 mutations were detected, four patients (no. 17, 36, 59, and 60) had three mutations (Supplementary Table S2). The detection rate of the mutations was 83.8% (124 mutated alleles out of 148 expected). Among the 75 distinct identified mutations (Supplementary Table S3), 41 (54.6%) were not previously reported according to the PKHD1 mutation database (http://www.humgen.rwth-aachen.de/). At least two mutations were detected in 53 patients (71.6%) and one mutation in 18 patients (24.3%). No mutations could be detected in three patients. Download .doc (.11 MB) Help with doc files Supplementary Table S2 Download .doc (.07 MB) Help with doc files Supplementary Table S3 Seventy-one mutations (55.5%) were truncating with 44 frameshifting small deletions/insertions/duplications, 24 nonsense mutations and three splice mutations with exon skipping, leading to out of frame fusion of exons. Fifty-four mutations (42.2%) were missense mutations. One mutation was a missense mutation associated with an in-frame insertion of six bases in exon 16, one mutation was an in-frame deletion of 42 bases in exon 30 and one mutation was a one-base deletion in exon 67 leading to the change of the last 12 amino acids of the fibrocystin, the abolition of the stop codon and the addition of nine amino acids (Supplementary Table S2, Supplementary Table S3). Parent DNA analysis in the four patients with three mutations showed that the co-inherited mutations were always missense mutations (Supplementary Table S2). We observed eight patients with homozygous mutations, implicating truncating mutations in all except one (Supplementary Table S2). When parents DNAs were available (five cases), the homozygous character of the mutation was confirmed by testing both parents. In the three remaining cases, intragenic SNPs and intra and flanking extragenic microsatellites showed a homozygous pattern (data not shown), indicating that the mutation was either homozygous, or in combination with a large PKHD1 deletion. Mutations were observed all along the gene, scattered in 41 exons. However, three mutations were recurrent: the T36M missense (n=13) in exon 3 and the frameshifts c.5895dupA (L1966fs) (n=9) in exon 36 and c.9689delA (D3230fs) (n=13) in exon 58 (Supplementary Table S2, Supplementary Table S3). Furthermore, the screening of 6 exons (3, 32, 36, 57, 58, and 61) allowed the identification of 51% of mutations. Additional 6 exons (9, 14, 18, 19, 22, and 34) led to a detection rate of 64%. For this analysis, truncating mutations (T) (nonsense, frameshift and splice mutations) and non-truncating mutations (NT) (missense mutations and diverse in-frame modifications) were considered (Supplementary Table S2). According to the fact that the disease is recessive with a loss of function of the fibrocystin protein, we considered genotype as severe when two truncating mutations were present (T/T, 21 patients) and non-severe if at least one NT mutation was present (41 patients including 20 with NT/T, 12 with NT/NT and 9 with NT/non identified (NI) mutations). Patients with T/NI (nine patients) or NI/NI (three patients) mutations were excluded for the genotype–phenotype correlation analysis as they could have undetected severe mutations such as an intragenic or large PKHD1 deletion, not screened in our study. Severe genotype was significantly more frequent in fetuses that underwent medical pregnancy termination at 30 weeks of GA or less (53%) compared to those with GA over 30 weeks (16%, including three medical pregnancy terminations and two post-natal deaths). Conversely, the majority of patients who underwent medical pregnancy termination after 30 weeks of GA or who died after birth had non-severe genotype (Table 3). It is interesting to note that 10 different missense mutations (T36M, I222V, R328G, T1300K, C1692Y, P1791L, R2370G, W2736G, I2957T, and R3240K) were identified among the 16 patients with non-severe genotype who died after birth.Table 3Distribution of gestational age at medical pregnancy termination or death according to genotype (severe or non-severe)GenotypeNNon-severeSevere62N=41N=21P-valueMedical pregnancy termination ≤30 weeks301416 (53)0.006Medical pregnancy termination >30 weeks14113 (21)Post-natal demise18162 (11)Results are expressed as numbers (percentages per row).For definition of genotypes: see Results, Genotype–phenotype correlation analysis. Open table in a new tab Results are expressed as numbers (percentages per row). For definition of genotypes: see Results, Genotype–phenotype correlation analysis. When GA was not taken into account, no correlation was found between the severity of genotype and any of the histological criteria of severity (Supplementary Table S4). However, when adjusted to GA, the degree of extension of collecting duct dilatation in the cortex was significantly more important in the patients with severe genotype compared to those with non-severe genotype. The grade of cortical tubule lesions adjusted to GA was also more important in the patients with severe genotype than in those with non-severe genotype (Table 4). Proximal tubule dilatation, glomerular lesions, cortical fibrosis, and hepatic fibrosis adjusted on GA were not significantly different between the two groups of genotype (Table 4). Download .doc (.04 MB) Help with doc files Supplementary Table S4Table 4Relationships between genotype (severe or non-severe) and renal lesions and portal fibrosis adjusted to gestational ageGenotypeNNon-severeSevereP-valueRenal lesionsCollecting duct dilatation in the cortex610.01 GA ≤30 weeks Absent or partial251411 (44) Global505 (100) GA >30 weeks Absent or partial770 Global24204 (17)Cortical tubule lesions610.07 GA ≤30 weeks Grade 11147 (64) Grade 215105 (33) Grade 3404 (100) GA >30 weeks Grade 1330 Grade 211101 (9) Grade 317143 (18) Proximal tubule dilatation590.44 GA ≤30 weeks No1257 (58) Yes1899 (50) GA >30 weeks No25214 (16) Yes440 Glomerular lesions590.34 GA ≤30 weeks Absent1798 (47) Partial1156 (55) Global202 (100) GA >30 weeks Absent330 Partial871 (13) Global18153 (17) Cortical fibrosis590.85 GA ≤30 weeks Absent241113 (54) Focal532 (40) Diffuse101 (100) GA >30 weeks Absent13121 (8) Focal972 (22) Diffuse761 (14)Portal fibrosis520.69 GA ≤30 weeks Grade 1321 (33) Grade 219109 (47) Grade 3110 GA >30 weeks Grade 1220 Grade 215123 (20) Grade 312102 (17)For definition of lesions: see Results, Renal and hepatic histology.For definition of genotypes: see Results, Genotype–phenotype correlation analysis. Open table in a new tab For definition of lesions: see Results, Renal and hepatic histology. For definition of genotypes: see Results, Genotype–phenotype correlation analysis. This is the first series of ARPKD patients who analyzed the genotype–phenotype correlations via the relations between the severity of renal and hepatic histological lesions and PKHD1 mutations. The severity of the disease in this series of prenatally detected ARPKD was confirmed by the association of severe cortical lesions with the typical dilatation of collecting ducts in the medulla. Extension to the cortex of collecting duct dilatation was observed in 96% of patients and cortical tubule lesions grade 2 or 3 in 74% of the patients. Similarly, portal fibrosis grade 2 or 3 was present in 92% of patients (Table 1 and Supplementary Table S1). It is generally considered that cortical lesions in ARPKD are the consequence of compression by the cystic collecting ducts extending to the cortex. Our series confirms that the severity of cortical tubule lesions, glomerular lesions, and cortical fibrosis were significantly correlated with the degree of extension of collecting duct dilatation within the cortex (Table 2). It is interesting to note that a correlation close to significance was also found between the degree of portal fibrosis and the extension of collecting ducts in the renal cortex, suggesting that renal and hepatic involvement have a parallel severity in the most severe forms of ARPKD detected prenatally. A prerequisite before investigating the correlations between histological phenotype and genotype was to analyze the effect of GA on the severity of histological lesions. Our study confirmed that the degree of extension of collecting duct dilatation within the renal cortex, the severity of cortical tubule lesions, glomerular lesions, cortical fibrosis, and portal fibrosis increased significantly with GA (Table 1), a logical issue, but not demonstrated previously in humans, even in the seminal studies of Osathanonah and Potter.23Osathanondh V. Potter E.L. Pathogenesis of polycystic kidneys. Type 1 due to hyperplasia of interstitial portions of collecting tubules.Arch Pathol. 1964; 77: 466-473PubMed Google Scholar The only histological abnormality, which we observed to be more frequent when GA was ≤30 weeks compared with >30 weeks was the dilatation of proximal tubules (Table 1), which also was more frequent when the extension of collecting duct dilatation within the cortex was absent or partial compared with global (Table 2). Dilatation of proximal tubules has been reported as an early lesion, preceding collecting duct cystic dilatation in human ARPKD fetuses24Nakanishi K. Sweeney Jr, W.E. Zerres K. et al.Proximal tubular cysts in fetal human autosomal recessive polycystic kidney disease.J Am Soc Nephrol. 2000; 11: 760-763PubMed Google Scholar and several mouse models of ARPKD.25Avner E.D. Studnicki F.E. Young M.C. et al.Congenital murine polycystic kidney disease. I. The ontogeny of tubular cyst formation.Pediatr Nephrol. 1987; 1: 587-596Crossref PubMed Scopus (45) Google Scholar, 26Nauta J. Ozawa Y. Sweeney Jr, W.E. et al.Renal and biliary abnormalities in a new murine model of autosomal recessive polycystic kidney disease.Pediatr Nephrol. 1993; 7: 163-172Crossref PubMed Scopus (96) Google Scholar, 27Moyer J.H. Lee-Tischler M.J. Kwon H.Y. et al.Candidate gene associated with a mutation causing recessive polycystic kidney disease in mice.Science. 1994; 264: 1329-1333Crossref PubMed Scopus (303) Google Scholar, 28Ricker J.L. Gattone 2nd, V.H. Calvet J.P. et al.Development of autosomal recessive polycystic kidney disease in BALB/c-cpk/cpk mice.J Am Soc Nephrol. 2000; 11: 1837-1847PubMed Google Scholar, 29Woollard J.R. Punyashtiti R. Richardson S. et al.A mouse model of autosomal recessive polycystic kidney disease with biliary duct and proximal tubule dilatation.Kidney Int. 2007; 72: 328-336Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar Ward et al.9Ward C.J. Yuan D. Masyuk T.V. et al.Cellular and subcellular localization of the ARPKD protein; fibrocystin is expressed on primary cilia.Hum Mol Genet. 2003; 12: 2703-2710Crossref PubMed Scopus (248) Google Scholar have shown that the PKHD1 protein fibrocystin is expressed in embryonic human kidneys not only in the branching ureteric ducts and collecting ducts, but also, with a weaker intensity, in the developing nephrons. In addition, expression of fibrocystin has been shown in human adult12Zhang M.Z. Mai W. Li C. et al.PKHD1 protein encoded by the gene for autosomal recessive polycystic kidney disease associates with basal bodies and primary cilia in renal epithelial cells.Proc Natl Acad Sci USA. 2004; 101: 2311-2316Crossre