Von Hippel-Lindau Syndrome

A number sign (#) is used with this entry because von Hippel-Lindau syndrome (VHL) is caused by heterozygous mutation in the VHL gene (608537) on chromosome 3p25.

Evidence suggests that variation in the cyclin D1 gene (CCND1; 168461) on chromosome 11q13 may modify the phenotype.

Homozygous or compound heterozygous mutations in the VHL gene cause familial erythrocytosis-2 (ECYT2; 263400).

Description

Von Hippel-Lindau syndrome (VHL) is a dominantly inherited familial cancer syndrome predisposing to a variety of malignant and benign neoplasms, most frequently retinal, cerebellar, and spinal hemangioblastoma, renal cell carcinoma (RCC), pheochromocytoma, and pancreatic tumors.

Neumann and Wiestler (1991) classified VHL as type 1 (without pheochromocytoma) and type 2 (with pheochromocytoma). Brauch et al. (1995) further subdivided VHL type 2 into type 2A (with pheochromocytoma) and type 2B (with pheochromocytoma and renal cell carcinoma). Hoffman et al. (2001) noted that VHL type 2C refers to patients with isolated pheochromocytoma without hemangioblastoma or renal cell carcinoma. McNeill et al. (2009) proposed that patients with VHL syndrome caused by large VHL deletions that include the HSPC300 gene (C3ORF10; 611183) have a specific subtype of VHL syndrome characterized by protection from renal cell carcinoma, which the authors proposed be named VHL type 1B.

Nordstrom-O'Brien et al. (2010) provided a review of the genetics of von Hippel-Lindau disease.

Clinical Features

The cardinal features of von Hippel-Lindau syndrome are angiomata of the retina and hemangioblastoma of the cerebellum. Hemangioma of the spinal cord has also been observed. Pheochromocytoma occurs in some patients. The combination of hypertension with angioma may lead to subarachnoid hemorrhage. Hypernephroma-like renal tumors occur in some patients. Polycythemia may be due to either the hemangioblastoma of the cerebellum or the hypernephroma. Hemangiomas of the adrenals, lungs, and liver, and multiple cysts of the pancreas and kidneys, have been observed in some instances.

Although there were many earlier reports of this syndrome (see HISTORY), Melmon and Rosen (1964) introduced the term 'von Hippel-Landau' syndrome and described a large kindred with multiple features of the disorder. The condition of arteriovenous aneurysm of retina and midbrain with facial nevus, described by Bonnet et al. (1938) and by Wyburn-Mason (1943), is of uncertain relationship to this condition. Metastatic renal cancer occurs in some instances (Kranes and Balogh, 1966). Goldberg and Duke (1968) examined the eyes of an affected 51-year-old black male whose mother died of cerebellar tumor at age 26 years. The same case was described by McKusick (1961). In addition to the association of tumors of the brain and adrenal medulla that occurs in neurofibromatosis and in von Hippel-Lindau disease, cerebellar tumors sometimes produce paroxysmal hypertension similar to that of pheochromocytoma. Urinary catecholamines are normal in such cases (Cameron and Doig, 1970).

Cysts and 'hypernephroid' tumors of the epididymis have been described in VHL patients (Grossman and Melmon, 1972). Male patients may have papillary cystadenoma of the epididymis, an unusual tumor that is bilateral when it occurs in von Hippel-Lindau disease and is not familial when unilateral (Price, 1971). The experience of Lamiell (1987) differed, however; 7 of 21 affected males in 1 kindred had an epididymal mass and 5 of these were unilateral. Tsuda et al. (1976) observed the occurrence of bilateral papillary cystadenoma of the epididymis in 3 brothers with VHL syndrome. Bilateral papillary cystadenomas of the broad ligament, presumably of mesonephric origin, is the probable homologous tumor of the female (Erbe, 1978).

In von Hippel-Lindau disease with pheochromocytoma, Atuk et al. (1979) reported hypercalcemia which was corrected in all by removal of the tumor. In several patients, pheochromocytoma antedated development of retinal lesions. Fishman and Bartholomew (1979) described 3 related patients with striking pancreatic involvement. One had exocrine pancreatic insufficiency. In an extensively affected kindred, Fill et al. (1979) found renal cell carcinoma in 16 of 42 cases and carcinoma of the pancreas in 4 of 42.

Griffiths et al. (1987) found reports of 6 patients with von Hippel-Lindau syndrome, pheochromocytoma, and islet cell tumor. A further 11 patients showed pheochromocytoma and islet cell tumor. No patient with von Hippel-Lindau syndrome had a carcinoid tumor, which is a feature of neurofibromatosis with pheochromocytoma (see 162200). No cases of neurofibromatosis had islet cell tumor. In Cardiff, Wales, 20 patients with cerebellar hemangioblastoma were seen between 1972 and 1985. In 8 of these, Huson et al. (1986) subsequently established the diagnosis of von Hippel-Lindau disease. Although the diagnosis had not previously been considered, in retrospect, 7 of the 8 cases were known to be at risk for the syndrome.

Jennings et al. (1988) demonstrated the usefulness of family studies in determining asymptomatic lesions requiring treatment, such as renal cell carcinoma. They also reported the occurrence of a spermatic cord mesenchymal hamartoma in this disorder. Lamiell et al. (1989) identified 43 affected members in a large kindred, which was exceptional for absence of pheochromocytoma and erythrocythemia, for more renal and pancreatic cysts and malignancies, and for somewhat fewer eye or central nervous system lesions. Bilateral renal adenocarcinoma was found presymptomatically in 5 young subjects who had bilateral nephrectomy and hemodialysis. Three survived long-term after renal transplants. Five members of the family had pancreatic malignancy.

Horbach et al. (1989) suggested that the combination of adrenal pheochromocytoma and ipsilateral renal cell carcinoma may represent a forme fruste of von Hippel-Lindau disease.

Neumann and Wiestler (1991) found a striking tendency for familial clustering of particular VHL features. Both angiomatosis retinae and hemangioblastoma of the CNS occurred in most families, whereas the occurrence of renal lesions and/or pancreatic cysts was mutually exclusive with pheochromocytoma. The authors interpreted these findings to indicate that the VHL locus is complex, with the existence of different mutations in different families or the occurrence of additional genetic lesions that cooperate with the VHL gene on chromosome 3p. They suggested a linear sequence of features as follows: pheochromocytoma, angiomatosis retinae, hemangioblastoma of the CNS, renal lesions, pancreatic cysts, and epididymal cystadenoma.

In the course of an evaluation of 41 families with this disorder from the United States and Canada, Glenn et al. (1991) found 1 large family with a distinctive phenotype: the most common disease manifestation was pheochromocytoma occurring in 57% (27 of 47) of affected members; few (4 of 47) had symptomatic spinal or cerebellar hemangioblastomas; no affected family member had renal cell carcinoma or pancreatic cysts. Genetic analysis demonstrated, however, that the disorder in this family was linked to the same markers found to be linked to typical VHL. The observations are clearly relevant to the descriptions of families with 'pure' pheochromocytoma (171300); they may be instances of allelism at the VHL locus.

Keeler and Klauber (1992) described renal cell carcinoma in a 16-year-old boy, probably the youngest reported example of hypernephroma in VHL disease.

Lenz et al. (1992) demonstrated that norepinephrine-producing adrenal pheochromocytoma in von Hippel-Lindau disease can produce the clinical syndrome of hypertension associated with severe hypokalemia and hyperreninemic hyperaldosteronism. The hyperreninemic hyperaldosteronism was rapidly improved by beta-blockade and was completely reversed by tumor removal. Kerr et al. (1995) described hemangioblastoma of the optic nerve in a 27-year-old woman with von Hippel-Lindau syndrome, the tenth such reported case.

Davies et al. (1994) described a 65-year-old woman who was an obligate carrier of the gene for von Hippel-Lindau disease. Her father, 2 brothers, 2 sisters, and 3 sons had hemangioblastomas and renal carcinomas. Careful examination of the woman showed only a small benign renal cyst. Such cysts are very common in the general population. Therefore, obligate gene carriers may not exhibit any features of the disease beyond the age of 60 years.

Using a VHL register set up in the northwest of England in 1990 containing information on 83 affected persons, Maddock et al. (1996) studied population statistics, clinical features, age at onset, and survival. The mean age at onset of first symptoms was 26.25 years, with cerebellar hemangioblastoma being the most common presenting manifestation (34.9% of cases). The mean age at diagnosis of VHL was 30.87 years. Overall, 50 patients (60.2%) developed a cerebellar hemangioblastoma, 34 (41%) a retinal angioma, 21 (25.3%) a renal cell carcinoma, 12 (14.5%) a spinal hemangioblastoma, and 12 (14.5%) a pheochromocytoma. Mean age at death was 40.9 years with cerebellar hemangioblastoma being the most common cause (47.7% of deaths). In addition to the 83 clinically affected subjects, Maddock et al. (1996) identified 3 obligate carriers who were considered to be lesion free on extensive screening tests. In the regionally based cancer registry, 14% of all CNS hemangioblastomas were found to occur as part of VHL, but investigations for VHL in apparently sporadic cases appeared to have been limited.

Endolymphatic sac tumors (ELSTs) are highly vascular, benign, but locally aggressive neoplasms of the endolymphactic system that often destroy the surrounding temporal bone. They are very rare and generally occur sporadically, but occur with increased frequency in patients with VHL. Manski et al. (1997) found MRI evidence of 15 ELSTs in 13 (11%) of 121 patients with VHL, but in none of 253 patients without evidence of VHL (P less than 0.001). Clinical findings in these 13 patients included hearing loss in 13, tinnitus in 12, vertigo in 8, and facial paresis in 1. Mean age at onset of hearing loss was 22 years (range, 12 to 50 years).

Lonser et al. (2004) described 3 cases of von Hippel-Lindau disease that illustrated the following features of endolymphatic sac tumors: morbid hearing loss due to a radiologically undetectable microscopic tumor in the endolymphatic sac or duct; initial symptoms caused by hemorrhage, endolymphatic hydrops, or both; an origin in the endolymphatic duct or sac; and molecular evidence of an association with von Hippel-Lindau disease. Complete surgical resection of the endolymphatic sac tumors is curative and can be performed with the preservation of hearing and the alleviation of vestibular symptoms.

Butman et al. (2007) reported 35 VHL patients with ELSTs; 3 had bilateral tumors. Mean age at symptom onset was 31 years (range, 11 to 63 years). In additional to hearing loss, tinnitus, and vertigo, other features included aural fullness, aural pain, and facial nerve weakness. Detailed CT and MRI studies showed that 7 (18%) ears had otic capsule invasion, which was always associated with hearing loss. Tumors with otic capsule invasion were larger (2.2 cm) than those without capsule invasion (1.2 cm). However, there was not a significant association between tumor size and hearing loss. Intralabyrinthine hemorrhage was detected in 79% of ears with sudden hearing loss. Butman et al. (2007) concluded that hearing loss associated with ELSTs can result from otic capsule invasion, intralabyrinthine hemorrhage, or endolymphatic hydrops.

James (1998) tabulated reports of 4 women with VHL and broad ligament papillary cystadenoma published between 1988 and 1994 (Gersell and King, 1988; Funk and Heiken, 1989; Korn et al., 1990; Gaffey et al., 1994; Karsdorp et al., 1994) and added a fifth case. These are mesosalpinx cysts, which are the equivalent of epididymal cysts in the male. The cysts were unilateral in at least 3 of the 5 cases. They occurred along the full course of the mesonephric duct, in the mesosalpinx close to the ovary, over the uterine tubes, and near the vaginal fornix in a remnant of the Gartner duct (the female counterpart to the duct of the epididymis). At least 3 of the patients had multiple renal cysts and bilateral renal cell carcinoma. In the patient reported by Korn et al. (1990), screening for VHL after the papillary cystadenomas were diagnosed revealed pancreatic cysts and lesions of the cerebellum and kidney; renal cell carcinoma was diagnosed during follow-up surgery. The unilateral cyst in the patient reported by Gaffey et al. (1994) was preceded by the finding of a middle ear papillary tumor. The combined presentation of mesonephric cystadenoma and ear tumor was noted in reports of epididymal cysts in men (Price, 1971). Gaffey et al. (1994) suggested that the ear tumor and the adnexal tumor may represent 'major visceral manifestations of VHL.' (Nomenclature: According to the VHL Family Alliance, a genetic support group, the approved terminology is 'adnexal papillary cystadenoma of probable mesonephric origin,' abbreviated APMO (Graff, 1998).)

Fukino et al. (2000) described a Japanese VHL family in which 2 of the 3 affected members developed acute occlusive hydrocephalus that necessitated emergency surgery for ventricular shunting or drainage. In both cases, the occlusion of the cerebrospinal canal was caused by cerebellar hemangioblastoma. The 2 patients with hydrocephalus were sisters aged 8 and 19 at the time of development of obstructive hydrocephalus. They inherited VHL from their mother, who also suffered from cerebellar hemangioblastoma requiring surgery as well as from retinal angiomas.

McCabe et al. (2000) described the clinical features, association with von Hippel-Lindau disease, and visual acuity outcomes of patients with juxtapapillary capillary hemangioma, on or adjacent to the optic nerve. Because of their location, a hamartoma on the lesions could potentially be misdiagnosed as papilledema, papillitis, choroidal neovascularization, or choroiditis. Endophytic, exophytic, and sessile forms were described. On long-term follow-up, visual acuity generally worsened. Patients with VHL and juxtapapillary hemangioma more often presented at a younger age, had tumors with an endophytic growth pattern, and had bilateral, multiple tumors. Tumor treatment with laser photocoagulation resulted in variable visual acuity outcomes in the patients reported.

Raja et al. (2004) reported that external beam radiotherapy (EBRT) was a useful option in the treatment of retinal hemangiomas secondary to VHL disease that progressed despite standard therapy. EBRT led to improvement of visual acuity, reduction in tumor volume, and stabilization of retinal detachment in most patients treated.

Eisenhofer et al. (2001) examined the mechanisms linking different biochemical and clinical phenotypes of pheochromocytoma in MEN2 (171400) and VHL to underlying differences in the expression of tyrosine hydroxylase (TH; 191290), the rate-limiting enzyme in catecholamine synthesis, and of phenylethanolamine N-methyltransferase (PNMT; 171190), the enzyme that converts norepinephrine to epinephrine. Signs and symptoms of pheochromocytoma, plasma catecholamines and metanephrines, and tumor cell neurochemistry and expression of TH and PNMT were examined in 19 MEN2 patients and 30 VHL patients with adrenal pheochromocytomas. MEN2 patients were more symptomatic and had a higher incidence of hypertension (mainly paroxysmal) and higher plasma concentrations of metanephrines, but paradoxically lower total plasma concentrations of catecholamines, than VHL patients. MEN2 patients all had elevated plasma concentrations of the epinephrine metabolite metanephrine, whereas VHL patients showed specific increases in the norepinephrine metabolite normetanephrine. The above differences in clinical presentation were largely explained by lower total tissue contents of catecholamines and expression of TH and negligible stores of epinephrine and expression of PNMT in pheochromocytomas from VHL than from MEN2 patients.

Taouli et al. (2003) discussed abdominal imaging findings, including pictorial images, from more than 150 patients with VHL syndrome. The most common findings were renal and pancreatic masses.

In a prospective study of 406 VHL patients from 199 families seen at 1 institution in a 12-year period, Chew (2005) found that 205 of the patients had ocular involvement. Patients with complete deletion of the VHL gene were less likely to have ocular involvement than those with partial deletion, missense, or nonsense mutations (9% vs 45%; p less than 0.0001). Chew (2005) identified a previously unreported ocular feature, retinal neovascularization, in 17 patients. Chew (2005) also found 11 cases of intraorbital/intracranial hemangioblastoma, previously reported to be rare in VHL, accounting for 5.3% of all vision-threatening lesions in this study group.

In surgically excised retinal hemangioblastomas associated with von Hippel-Lindau disease, Liang et al. (2007) demonstrated high levels of VEGF (192240) and CXCR4 (162643) mRNA and protein but low levels of CXCL12 (600835). Increased expression of VEGF and CXCR4 was also detected in more active hemangioblastomas.

Binderup et al. (2016) performed a retrospective analysis of a national cohort study that included 52 VHL mutation carriers. The analysis spanned a total of 799 person-years. From birth to time of the report, 581 manifestations were diagnosed during 2,583 examinations in the study subjects. The rate of new tumor development varied significantly with age and was highest at 30 to 34 years (0.4 new tumors/year). Tumor location further influenced the rate. The risk of retinal tumors was highest in subjects during the teenage years but was highest for cerebellar tumors in subjects during their 30s. Truncating VHL mutation carriers had a significantly higher manifestation rate compared with missense mutation carriers (hazard ratio = 1.85, 95% confidence interval 1.06-3.24, p = 0.031). The authors concluded that the rate of new manifestation development is not constant throughout the life span of VHL patients. They recommended careful retinal surveillance during the teenage years and increased cerebellar surveillance in adulthood.

Pathogenesis

Interfamilial differences in predisposition to pheochromocytoma in VHL reflect allelic heterogeneity such that there is a strong association between missense mutations and risk of pheochromocytoma. Prowse et al. (1997) investigated the mechanism of tumorigenesis in VHL tumors to determine whether there were differences between tumor types or classes of germline mutations. They studied 53 tumors (30 renal cell carcinomas, 15 hemangioblastomas, 5 pheochromocytomas, and 3 pancreatic tumors) from 33 patients (27 kindreds) with VHL. Overall, 51% of 45 informative tumors showed LOH at the VHL locus. In 11 cases, it was possible to distinguish between loss of the wildtype and mutant alleles, and in each case the wildtype allele was lost. LOH was detected in all tumor types and occurred in the presence of both germline missense mutations and other types of germline mutation associated with a low risk of pheochromocytoma. Intragenic somatic mutations were detected in 3 tumors (all hemangioblastomas) and in 2 of these could be shown to occur in the wildtype allele. Their study provided the first example of homozygous inactivation of the VHL gene by small intragenic mutations in this type of tumor. Hypermethylation of the VHL gene was detected in 33% (6 of 18) of tumors without LOH, including 2 renal cell carcinomas and 4 hemangioblastomas. Prowse et al. (1997) stated that although hypermethylation of the VHL gene had been reported previously in nonfamilial RCC and although methylation of tumor-suppressor genes had been implicated in the pathogenesis of other sporadic cancers, this was the first report of somatic methylation in a familial cancer syndrome. Herman et al. (1994) observed hypermethylation of the VHL gene in 19% of sporadic RCCs. Versteeg (1997) provided a general discussion of aberrant methylation in cancer.

By using comparative genomic hybridization (CGH), Lui et al. (2002) characterized the genetic profiles of 36 VHL-related pheochromocytomas. They found loss of chromosome 3 or chromosome 11 in 34 tumors (94%) and 31 tumors (86%), respectively. There was significant concordance of deletions in chromosomes 3 and 11, suggesting that they are involved in 2 different but necessary and complementary genetic pathways. The loss of chromosome 11 appeared to be specific for VHL-related pheochromocytoma as it was not present in any of the 10 VHL-related CNS hemangioblastomas studied and was significantly less common when compared with sporadic and MEN2-related pheochromocytomas. The authors stated that this was the first report of a novel consistent genetic alteration that is selected and specific for VHL-related pheochromocytoma.

Population Genetics

Maher et al. (1991) estimated the point prevalence of heterozygotes in East Anglia to be 1 in 53,000, with an estimated birth incidence of 1 in 36,000 live births. Reproductive fitness was 0.83. Direct and indirect estimates of the mutation rate were 4.4 per million gametes per generation and 2.32 per million gametes per generation, respectively. No significant association was found between parental age or birth order and new mutations. In the Freiburg district of Germany, Neumann and Wiestler (1991) calculated the prevalence of this disorder to be 1 in 38,951.

Maddock et al. (1996) reported on a VHL register set up in the northwest of England in 1990. There was information on 83 affected persons. In addition, the effectiveness of the screening program used and the occurrence of CNS hemangioblastomas in the general populations were examined. The diagnostic point prevalence of heterozygotes in the region was 1 in 85,000 persons, with an estimated birth incidence of 1 in 45,500 live births. The mutation rate was estimated directly to be 1.4 x 10(-6)/gene/generation (1 in 714,200).

Wu et al. (2012) identified mutations in the VHL gene in 12 (75%) of 16 Chinese probands with clinically diagnosed VHL syndrome. PCR-direct sequencing detected 12 mutations, 1 of which was novel, in 12 patients (75%). Use of universal primer quantitative fluorescent multiplex PCR (UPQFM-PCR) enabled detection of 2 large deletions in 2 (12.5%) patients. The 2 remaining patients carried atypical variations in the VHL gene that could not definitively be called pathogenic. Nine (56.3%) probands did not have a family history of the disorder, suggesting a high frequency of de novo mutations among Chinese patients. Clinically, 15 families were classified as type 1 (without pheochromocytoma) and 1 as type 2 (with pheochromocytoma). The most common manifestations were CNS hemangioblastoma, clear cell renal cell carcinoma, and pancreatic cysts and tumors. Combining this information with previous reports of Chinese VHL patients indicated that the clinical features and spectrum of VHL mutations among the Chinese are comparable to those found in large-scale investigations from other countries.

Diagnosis

Seizinger et al. (1991) pointed out that visceral cysts of the kidney, pancreas, and epididymis occur not only as features of VHL but also in the general population, and that the presence of such cysts, unaccompanied by other more typical lesions such as retinal and cerebellar hemangioblastoma, may represent a major diagnostic problem. The application of flanking markers for the VHL gene for presymptomatic diagnostic testing confirmed that epididymal cysts are indeed not suitable as a diagnostic criterion. The genetic studies suggested that VHL with or without pheochromocytomas is caused by defects within the same gene. Renal cell carcinoma occurs as part of VHL; a second more proximal region of chromosome 3, 3p14.2, is responsible for 'pure familial renal cell carcinoma' (144700).

Webster et al. (2000) calculated the likelihood of VHL in an individual presenting with a single ocular angioma conditional upon the age of presentation, results of DNA analysis, family history of VHL, and results of systemic screening, and produced a risk estimate table for individuals with combinations of these variables.

Hes et al. (2003) noted that in the presence of a positive family history, VHL disease can be diagnosed clinically in a patient with at least 1 typical VHL tumor. Typical VHL tumors are retinal, spinal, and cerebellar hemangioblastoma; renal cell carcinoma; and pheochromocytoma. Endolymphatic sac tumors and multiple pancreatic cysts suggest a positive carriership in the presence of a positive VHL family history because they are uncommon in the general population. In contrast, renal and epididymal cysts occur very frequently in the general population and are, as sole manifestations, not reliable indicators for VHL disease. In patients with a negative family history of VHL-associated tumors, a diagnosis of VHL disease can also be made on the basis of 2 or more hemangioblastomas or a single hemangioblastoma in association with a visceral manifestation (e.g., pheochromocytoma or renal cell carcinoma). Hes et al. (2003) suggested the following criteria for eligibility for VHL gene mutation analysis: a patient with classic VHL disease (meeting clinical diagnostic criteria) and/or first-degree family members; a person from a family in which a germline VHL gene mutation has been identified (presymptomatic test); a VHL-suspected patient, i.e., one with multicentric tumors in 1 organ, bilateral tumors, 2 organ systems affected, or 1 VHL-associated tumor at a young age (less than 50 years for hemangioblastoma and pheochromocytoma or less than 30 years for renal cell carcinoma); or a patient from a family with hemangioblastoma, renal cell carcinoma, or pheochromocytoma only.

Mutation Analysis

Using DNA polymorphic markers, Glenn et al. (1992) studied 16 families with VHL disease. Of 48 asymptomatic persons at risk of developing this illness because of an affected parent or sib, DNA polymorphism analysis predicted that 9 were carriers of the disease gene and 33 had the wildtype allele. The test was not informative in 6 persons. All 9 persons predicted to carry the VHL gene had evidence of occult disease on clinical examination. There was no clinical evidence of VHL disease in 32 of 33 persons predicted to carry the wildtype allele.

Richards et al. (1993, 1994) found that large germline deletions could be detected by Southern analysis and pulsed field gel electrophoresis in 19% and 3% of VHL patients, respectively.

To determine whether the pheochromocytoma-associated syndromes VHL and MEN2 play a role in the development of thoracic functioning paragangliomas, Bender et al. (1997) analyzed germline DNA from 5 unselected patients with this tumor for mutations in the genes that predispose to VHL and MEN2. Molecular and clinical data revealed that 3 (60%) had VHL, with 2 different germline mutations of the VHL gene, but no individual was affected by MEN2. Two of these 3 patients with VHL did not show any additional VHL-associated lesions. Bender et al. (1997) suggested that VHL should be considered in the differential diagnosis of thoracic pheochromocytoma, and that in VHL patients suspected of a catecholamine-secreting tumor, thoracic localization should be considered if an adrenal pheochromocytoma cannot be detected.

Pack et al. (1999) stated that the reported frequency of detection of VHL germline mutations had varied from 39 to 80%. Stolle et al. (1998) found that a quantitative Southern blotting procedure improved this frequency. Pack et al. (1999) reported the use of fluorescence in situ hybridization as a method to detect and characterize VHL germline deletions. They reexamined a group of VHL patients shown previously by SSCP and sequencing analysis not to harbor point mutations in the VHL locus. They found constitutional deletions in 29 of 30 VHL patients in this group, using cosmid and P1 probes that covered the VHL locus. They then tested 6 phenotypically normal offspring from 4 of these VHL families: 2 were found to carry the deletion and the other 4 were deletion-free. In addition, germline mosaicism of the VHL gene was identified in 1 family. Thus, FISH was found to be a simple and reliable method to detect VHL germline deletions and to be practically useful in cases where other methods of screening fail to detect abnormalities in the VHL gene.

Hes et al. (2000) performed mutation analysis of the VHL gene in 84 patients presenting with a single CNS hemangioblastoma and 4 with multiple hemangioblastomas, but no other features of VHL. A VHL germline mutation was found in 3 of 69 (4.3%) of those with single hemangioblastomas presenting at less than 50 years of age (3 of 84 (3.6%) in total) and 2 of the 4 patients with multiple hemangioblastomas. A VHL mutation was found in a 44-year-old woman presenting with a single cerebellar hemangioblastoma, in 4 clinically unaffected relatives, and in 2 single cases presenting at 29 and 36 years. Hes et al. (2000) recommended that in addition to conventional clinical and radiologic investigations, VHL mutation analysis be offered to those presenting with CNS hemangioblastomas before the age of 50 years.

Sgambati et al. (2000) presented 2 cases of VHL mosaicism. In each of 2 families, standard testing methods (Southern blot analysis and direct sequencing) identified the germline mutation in the VHL gene of the offspring, but not in their clinically affected parent. Additional methods of analysis of the affected parents' blood detected the VHL gene mutation in a portion of their peripheral blood lymphocytes. In one case, detection of the deleted allele was by FISH, and, in the second case, a 3-bp deletion was detected by conformational sensitive gel electrophoresis and DNA sequencing of cloned genomic DNA. Sgambati et al. (2000) concluded that mosaicism in VHL is important to search for and recognize when an individual without a family history of VHL has VHL. Patients diagnosed without family histories of the disease have been reported in as many as 23% of kindreds. Identification of individuals potentially mosaic for VHL will affect counseling of families, and these individuals should themselves be included in clinical screening programs for occult disease.

Cytogenetics

Kiechle-Schwarz et al. (1989) found rearrangements resulting in partial or total trisomy of chromosome 3p in 3 cell clones from pheochromocytomas derived from patients with von Hippel-Lindau syndrome.

Kovacs and Kung (1991) analyzed the DNA from 28 nonpapillary renal cell carcinomas arising in 2 patients with VHL disease. They used both karyotypic and RFLP analyses for evidence of allelic recombination on chromosomes 3 and 5. Two distinct breakpoint clusters were identified, each associated with different karyotypic alterations. The first type involves a breakpoint at chromosome 3p13, with the common nonreciprocal translocations occurring between 3p and 5q or 1q, resulting in the net loss of 3p and gain of 5q or 1q segments. In the second form of translocation, which was less common, the breakpoint on 3p and on the partner chromosome is near the centromere at bands p11 or q11. This kind of rearrangement was observed in 10 tumors with a nonrandom involvement of chromosome 3. In each case the derivative chromosome carrying the translocated 3p segment was preferentially eliminated from the tumor cells.

Mapping

Go et al. (1984) found no linkage of VHL with any of 31 marker loci by studying 41 affected persons among 220 descendants of an ancestral couple. Wells et al. (1987) found linkage with a segregating minisatellite band at a recombination fraction of 0.15 (lod score of about 3.0). In studies of 9 families, Seizinger et al. (1988) found linkage with RAF1 (164760) on chromosome 3; the combined maximum lod score was 4.38 at theta = 0.11 (1 lod unit confidence interval, the approximate equivalent of 95% confidence interval, of theta = 0.04 to 0.23). The finding of recombination between VHL and RAF1 indicated that the site of the VHL mutation is not in the RAF1 gene.

Seizinger et al. (1989) reported a multipoint linkage analysis of the VHL gene using a battery of polymorphic markers in the region 3p26-p25. Vance et al. (1990) confirmed the linkage of VHL to 3p. Hosoe et al. (1990) concluded that VHL is located between RAF1 (3p25) and D3S18 (3p26).

Maher et al. (1990) confirmed the assignment to 3p by linkage studies in 12 British families. They suggested that the VHL locus is telomeric to the THRB gene (190160). No evidence for genetic heterogeneity was found. Seizinger et al. (1991) reported the location of the VHL locus at 3p26-p25 on the basis of studies in 28 pedigrees comprising 164 affected persons. By multipoint linkage analysis, Maher et al. (1991) demonstrated flanking markers. They found no evidence of locus heterogeneity; families with and without pheochromocytoma showed linkage to D3S18 with which no recombination was observed (maximum lod score = 6.6 at theta = 0.0; CI, 0.00-0.06). By multipoint linkage analysis, Richards et al. (1993) narrowed the target region for isolation of the VHL disease gene by positional cloning techniques to a 4-cM interval between D3S1250 and D3S18.

Richards et al. (1993) constructed a long-range physical map around D3S601, which had been shown to be tightly linked to the VHL locus, and screened 91 VHL patients from 80 kindreds for germline rearrangements using pulsed field gel electrophoresis. Two patients were found to have germline deletions within this region, approximately 120 kb and 50 kb, respectively, telomeric to D3S601. The results established the position of the VHL locus with respect to D3S601, refined its localization to a small region of approximately 50 kb, and excluded the plasma membrane Ca(2+)-transporting ATPase-2 gene (108733) as the site of the VHL mutation.

Molecular Genetics

In 28 of 221 kindreds with von Hippel-Lindau syndrome, Latif et al. (1993) identified rearrangements of the VHL gene. Eighteen of these rearrangements were due to deletion in the VHL gene: 1 of these was an in-frame 3-nucleotide deletion (608537.0001).

In 55 of 94 unrelated VHL kindreds, Crossey et al. (1994) identified 40 different mutations in the VHL gene. The 2 most frequent mutations were arg238-to-gln (608537.0005) and arg238-to-trp (608537.0003), which were detected in 5 and 4 unrelated kindreds, respectively.

Ciotti et al. (2009) identified mutations in the VHL gene in 9 (100%) of 9 unrelated families and in 16 (88.9%) of 18 isolated patients presenting with the classic phenotype of VHL syndrome. VHL mutations were also found in 2 (66.7%) of 3 patients who met the diagnostic criteria for VHL syndrome, but who also had multiple cerebellar hemangioblastomas. Of those with mutations, 6 (22%) of 27 were found to have complete or partial deletions of the gene. No VHL mutations were found in 13 additional patients who did not meet the full diagnostic criteria of the disorder, but who had some suggestive features.

Modifiers of VHL

To assess the influence of variation in CCND1 (168461) on the retinal, renal, and central nervous system (CNS) manifestations of von Hippel-Lindau disease (193300), Zatyka et al. (2002) genotyped 118 patients for the codon 242 G-A SNP (168461.0001). The number of retinal angiomas was significantly higher in individuals harboring the G allele compared with AA homozygotes (p of 0.04). Possession of 1 or more G alleles was associated with earlier diagnosis of CNS hemangioblastoma by almost 2-fold, although the difference did not attain statistical significance (p of 0.05). A similar analysis for onset of renal cell carcinoma showed no evidence of an association with CCND1 genotype.

In a retrospective analysis of 123 patients from 55 families with VHL, including 13 with complete germline deletion of the VHL gene and 42 with partial gene deletions, Maranchie et al. (2004) observed a paradoxically lower prevalence of renal cell carcinoma in those with complete gene deletions. RCC occurred more frequently in patients with partial germline VHL deletions relative to complete deletions (48.9% vs 22.6%, p = 0.007). This striking phenotypic dichotomy was not seen for cystic renal lesions or for CNS (p = 0.22), pancreas (p = 0.72), or pheochromocytoma (p = 0.34). Deletion mapping demonstrated that development of RCC had an even greater correlation with retention of HSPC300 (C3ORF10; 611183), located within the 30-kb region of 3p immediately telomeric to the VHL gene (52.3% vs 18.9%, p less than 0.001), suggesting the presence of a neighboring gene or genes critical to the development and maintenance of RCC.

Cascon et al. (2007) found that 6 of 8 VHL patients without RCC had large germline deletion of the VHL gene including deletion of HSPC300. In contrast, 9 of the 10 with RCC had retention of the HSPC300 gene. Analysis of 9 sporadic RCC tumors showed that all retained an HSPC300 allele. Cascon et al. (2007) concluded that loss of the HSPC300 gene confers protection against renal clear cell carcinoma.

Genotype/Phenotype Correlations

Although pheochromocytoma occurs in only about 7% of VHL patients, marked interfamilial differences are often observed. Examining the relationship between VHL gene mutations and phenotype in 65 VHL kindreds, Crossey et al. (1994) found that large deletions or intragenic mutations predicted to cause a truncated protein were found in 36 of 53 families without pheochromocytoma but in only 2 of 12 families with pheochromocytoma (P less than 0.01). Of 12 families with pheochromocytoma, 10 had missense mutations compared with 13 of 53 kindreds without pheochromocytoma (P less than 0.001). In particular, the arg238-to-trp and arg238-to-gln mutations were associated with a high risk (62%) of pheochromocytoma.

Chen et al. (1995) identified germline mutations in 85 of 114 VHL families (75%). They found that the types of mutations responsible for VHL without pheochromocytoma (VHL type 1) differed from those responsible for VHL with pheochromocytoma (VHL type 2). Microdeletions/insertions, nonsense mutations, or deletions were found in 56% of families with VHL type 1; missense mutations accounted for 96% of those responsible for VHL type 2. Specific mutations in codon 238 accounted for 43% of the mutations responsible for VHL type 2 (see 608537.0003-608537.0005).

Zbar et al. (1996) performed germline mutation analysis in 469 VHL families from North America, Europe, and Japan. Germline mutations were identified in 300 (63%) of the families tested; a total of 137 distinct intragenic germline mutations were detected. Most (124 of 137) of the mutations occurred in 1 or 2 families; a few occurred in 4 or more families. In this large series, it was possible to compare the effects of identical germline mutations in different populations. Germline VHL mutations produce similar cancer phenotypes in Caucasian and Japanese VHL families. Germline VHL mutations were identified that produced 3 distinct cancer phenotypes: (1) renal carcinoma without pheochromocytoma, (2) renal carcinoma with pheochromocytoma (e.g., 608537.0003), and (3) pheochromocytoma alone (e.g., 608537.0012). Zbar et al. (1996) provided a catalog of VHL germline mutations with associated phenotype information.

In a patient with von Hippel-Lindau syndrome due to a 505T-C transition (608537.0009) in the VHL gene, Schimke et al. (1998) found a secretory carotid body paraganglioma, the first such instance; a nonfunctional malignant carotid body tumor had been described in a patient with VHL by Hull et al. (1982).

Gallou et al. (1999) analyzed the occurrence of RCC in VHL families, based on the nature of the VHL mutations. They observed RCC in at least 1 member of the VHL families in 77% of cases with mutations leading to truncated proteins, and in 55% of cases with missense mutations (P less than 0.05). Thus, mutations resulting in truncated proteins may carry a higher risk of RCC in VHL patients.

Bradley et al. (1999) described a family with VHL disease and a mutation in the VHL protein (608537.0017). Of 13 affected individuals, 7 had renal cell carcinoma and 1 had pheochromocytoma. The authors contrasted this family with 2 families reported by Chen et al. (1996) that had a mutation at the same position but causing a different amino acid change (608537.0012). In these families, 19 of 22 affected individuals had pheochromocytoma and none had renal cell carcinoma. Bradley et al. (1999) concluded that different amino acid changes at the same position can cause very distinct clinical phenotypes.

Hes et al. (2000) described 5 VHL families in which direct sequencing of the coding region of the VHL gene failed to identify the family-specific mutation. Further molecular analysis revealed deletions involving the VHL gene in each of these families. In 4 families, partial deletions of 1 or more exons were detected by Southern blot analysis. In the fifth family, FISH analysis demonstrated the deletion of the entire VHL gene. The data supported the previously established observation that families with a germline deletion have a low risk for pheochromocytoma. Further unraveling of the genotype-phenotype correlations in VHL disease revealed that families with a full or partial deletion of the VHL gene exhibited a phenotype with a preponderance of central nervous system hemangioblastoma.

Friedrich (2001) reviewed genotype/phenotype correlations in von Hippel-Lindau syndrome.

Hoffman et al. (2001) noted that a type 2C VHL mutant, L188V (608537.0014), which had been associated with a pheochromocytoma-only phenotype (and had been shown to retain the ability to promote HIF (603348) ubiquitylation), retained the ability to suppress cyclin D1 (CCND1; 168461) expression, suggesting that loss of VHL-mediated suppression of cyclin D1 is not necessary for pheochromocytoma development in VHL disease. Other studies had suggested that (1) genetic modifiers influence the phenotypic expression of VHL disease (Webster et al., 1998); and (2) polymorphic variation at the CCND1 codon 242 A/G SNP (168461.0001) may influence cancer susceptibility or prognosis in some situations. Therefore, Zatyka et al. (2002) analyzed the relationship between CCND1 genotype and phenotypic expression of VHL disease. They found an association between the G allele and multiple retinal angiomas (p = 0.04), and risk of central nervous system hemangioblastoma (p = 0.05). The findings suggested that a variety of HIF-independent mechanisms may contribute to the tumor suppressor activity of the VHL protein and that polymorphic variation at one VHL protein target influences the phenotypic expression of VHL disease.

In a study of 573 individuals with VHL syndrome from 200 unrelated families, Ong et al. (2007) found that age at disease onset was significantly earlier, and age-related risks of retinal angiomas and RCC were higher in individuals with nonsense or frameshift mutations compared to those with deletions or missense mutations. The results also confirmed the association of pheochromocytomas with missense mutations, particularly those that resulted in surface amino acid substitutions.

Wong et al. (2007) characterized the germline mutations found in 335 patients with VHL disease associated with retinal capillary hemangioblastomas (RCHs) and sought to establish genotype-phenotype correlations between genotype category (amino acid substitutions, protein-truncating mutations, and complete deletions) and ocular phenotype. The prevalence of RCHs was lowest (14.5%) among patients with complete deletions; the overall prevalence of retinal angiomatosis was 37.2%. Genotype category had no correlation with unilaterality or bilaterality of ocular disease or with the number or extent of peripheral RCHs. The prevalence of RCHs at the juxtapapillary location was lower among patients with protein-truncating mutations than in patients with amino acid substitutions. Complete deletions were associated with the highest mean visual acuity.

Franke et al. (2009) identified germline deletions in the VHL gene ranging from 0.5 to 250 kb in 54 families with VHL syndrome. In 28 of these families, at least 1 additional gene was deleted including FANCD2 (227646), HSPC300 (C3ORF10;611183), and IRAK2 (603304). The precise breakpoints were determined in 33 index patients. Of the 66 breakpoints, 90% occurred in Alu elements, indicating that Alu-mediated recombination is a major mechanism for germline deletions of the VHL gene. Among all 54 families with VHL syndrome resulting from germline