Editor—Autosomal dominant polycystic kidney disease (ADPKD) is a multisystem disease with intrafamilial phenotypic heterogeneity. Cerebral aneurysms develop in 10-20% of patients with ADPKD.1 2 Aneurysm rupture may precede the development of hypertension or renal manifestations of the disease.2 Two ADPKD loci (PKD1 andPKD2) map to chromosomes 16p and 4q, respectively. Reports of unlinked pedigrees have suggested the existence of a third unmapped locus.3-7 We report a three generation white family with an ADPKD like disease, unlinked to thePKD1 or PKD2 loci (fig 1, table 1). Other genetic causes of renal cysts, such as tuberous sclerosis and von Hippel-Lindau disease, have been excluded clinically and linkage to the known loci for these diseases is unlikely. This family supports the case for a third locus for ADPKD, but differs from the previously described unlinked pedigrees by its presentation with cerebral aneurysm rupture.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Pedigree with autosomal dominant polycystic kidney disease.

Ethical approval for this study was granted by the local research ethics committee. The family was identified because three members had sustained a subarachnoid haemorrhage. All family members who were older than 18 were invited for examination and screening for intracranial aneurysms by magnetic resonance angiography (MRA). There was no clinical evidence of tuberous sclerosis, von Hippel-Lindau disease, or autosomal recessive polycystic kidney disease (diseases associated with cerebral aneurysm formation). MRA scans that were suspicious for cerebral aneurysm were followed up by conventional cerebral angiography. In view of advanced age, MRA screening was not offered to I.2, but it was noted that she had previously suffered a CVA of undetermined aetiology. II.2 and II.4 had both sustained a subarachnoid haemorrhage as a result of rupture of an anterior communicating artery aneurysm (ACoA), and II.6 had a ruptured posterior communicating artery aneurysm (PCoA). All three were normotensive with normal serum urea and creatinine concentrations when they presented with subarachnoid haemorrhage. There was no past history of ADPKD within the family. Subjects III.8, III.9, III.10, and III.12 declined examination.

Subjects meeting recognised criteria for diagnosis of polycystic kidney disease were classed as affected.8 9 II.1, although designated normal, declined MRA and ultrasound investigation but was happy to provide blood. II.2 and II.6 did not meet these criteria. II.2 at the age of 57 had no renal cysts but was hypertensive, had a ruptured cerebral aneurysm, and had a son with renal cysts. II.6 had marked hepatic cystic disease, was also hypertensive, and had a ruptured cerebral aneurysm. In addition, her son, III.7, had renal cysts and was mildly hypertensive at the age of 19 years. Both II.2 and II.6 were designated affected because they both had many phenotypic characteristics of the disease, had an affected parent, and also had affected offspring. Absence of renal cysts in a subject with ADPKD proven by linkage has been described previously.8

Linkage analysis was carried out. Microcrosatellite markers flanking the known loci ofPKD1/TSC2,PKD2, and TSC1were chosen.10-13 CI-46 was used as a marker for theVHL locus.14 Results were analysed in two separate fashions using the LINKAGE (v 5.04) program.15 First, we tested the hypothesis that this family had classical ADPKD linked to either thePKD1 or PKD2 loci using recognised liability classes for ADPKD.9 At risk but unaffected family members were assigned to three liability classes, less than 10 years, 20-25 years, and greater than 25 years, with penetrances of 0.22, 0.66, and 0.85, respectively.9Linkage to the PKD1 andPKD2 loci was also assessed using an affecteds only analysis. Multipoint lod scores were consistently less than −2.0 across thePKD1/TSC2 andPKD2 loci using both methods. The disease is therefore not linked to these loci. Further linkage analysis was restricted to affected subjects only, as there are no data on the penetrance of ADPKD unlinked to PKD1 andPKD2.

We considered linkage to other loci associated with renal cystic disease or intracranial aneurysm formation. Multipoint lod scores plotted atTSC1/HHT1(<−1.5) andELA1/HHT2(<−2.5) suggested non-linkage. Two point lod scores for theVHL locus (<−2.0 within 0.08 cM from CI-46) also showed non-linkage.

Since subjects I.1 and II.9 were non-smokers who died at an early age from emphysema, we considered the diagnosis of α1-antitrypsin deficiency. Serum α1-antitrypsin levels and phenotypes were recorded in I.2, II.2, II.4, II.6, and III.11. Results lay within the normal range and all subjects typed had the MM phenotype.

There are previous reports of unlinked pedigrees.3-7Daoust et al 3 reported a two generation family with six affected subjects. Although only one member in this pedigree showed evidence of renal insufficiency, affected subjects had renal, hepatic, splenic, and ovarian cysts and were hypertensive. Cerebral aneurysms were not reported in this pedigree.3 These authors reported negative lod scores atTSC1 and VHL. In the pedigree reported by de Almeida et al,4 there were 25 subjects available for linkage analysis with 12 affected subjects in four generations. The phenotypic characteristics of all pedigree members were not presented. Affected subjects were noted to meet recognised ultrasonographic diagnostic criteria, to be variously hypertensive, and to show slow progression to renal replacement therapy. The authors specifically noted the absence of cerebral aneurysms or cerebral haemorrhage in the pedigree.4 TSC1 andVHL were not excluded by linkage analysis.

In conclusion, this family has an inherited condition similar to classical autosomal dominant polycystic kidney disease, with vertical transmission of disease. This is characterised by renal cysts, hepatic cysts, hypertension, and cerebral aneurysms. Within the pedigree there are no affected males of sufficient age to have allowed male to male transmission to be shown, so the possibility of X linked inheritance cannot be excluded. The family is not linked to thePKD1/TSC2 orPKD2 loci and supports the case for a third locus for ADPK-like conditions. This depends on us assuming that affected members of this pedigree are suffering from the consequences of the same monogenic disease mutation; it is possible (but unlikely) that the pedigree is a rare coincidental clustering of ADPKD-like phenotypes and that the condition is behaving as an autosomal dominant trait. The maximum and mean simulated lod scores for this pedigree at θ=0 were 2.9 and 1.6, respectively, Thus, the available samples will not be able to show linkage at a lod score >3.