In most cases, polycythemia vera (PV) diagnosis can easily be made based on WHO 2016 criteria, but the background to “simple” erythrocytosis cases is often unclear. Erythrocytosis can be primary or secondary. The leading cause of primary acquired erythrocytosis is PV with the Jak2 V617F driver mutation. The sporadic primary hereditary erythrocytosis is caused mainly by the erythropoietin receptor gene (EpoR) mutation. Increased erythropoietin (EPO) level can lead to secondary erythrocytosis from different disorders, i.e., chronic lung and heart disease, chronic increase in carbon monoxide due to smoking, sleep apnea, renal cysts and tumors, liver diseases, high-altitude living, and EPO-secreting tumors (cerebellar hemangioblastoma, renal cell carcinoma, hepatocellular carcinoma, or uterine leiomyoma). When these common causes of secondary erythrocytosis are excluded, an inherited cause involving hemoglobin or erythrocyte regulatory mechanisms may be suspected. Several genes of hemoglobin, hemoglobin-stabilization proteins, EPO receptor, or oxygen-sensing pathway enzymes may be affected. These are the von Hippel-Lindau gene (VHL) (the first described), EGLN1/PHD2, EPAS1/HIF2A, EPO, and EPOR genes (1,2). Inherited erythrocytosis comes with isolated red blood cell increase and normal spleen size, and it is generally believed that the risk of clonal evolution does not accompany it. A small subset of cases is associated with pheochromocytoma and paraganglioma formation. Although the incidence of thrombotic events may be higher, the use of regular phlebotomy and aspirin has not been proven (3). The VHL encodes a protein complex component involved in the ubiquitination and degradation of a hypoxia-inducible factor (HIF). This transcription factor plays a central role in the regulation of gene expression by oxygen. Von Hippel-Lindau (VHL) disease is a complex disorder in which increased oxygen demand of tissues results in a complex set of neoplastic disorders: i.e., clear cell renal carcinoma, pheochromocytoma, hemangioblastomas, pheochromocytoma, or pancreas cyst appearing well before the age of forty, and usually without erythrocytosis. The background of this disease is genetic, caused by a series of mutations in the VHL gene, which is positioned on the short arm of chromosome 3 (3p25-26) (4,5). There are over 1500 germline mutations and somatic mutations found or identified in VHL disease. Some rare cases of mutated VHL might induce isolated polycythemia; interestingly, the other tumors are mostly missing in these cases, so it is recommended to check the VHL in polycythemia if etiology is not otherwise explained (6–10). Much less is known about VHL single nucleotide polymorphisms (SNP). They are considered usually innocent, without any clinical significance. However, VHL SNPs are not always innocent. At least two known SNPs are not insignificant on clinical grounds. One example is the rs779805 c.-195G>A polymorphism (genomic location: GRCh38.p13 chr 3, HGVS: NC_000003.12:g.10141653G>A). According to the ClinVar database, this variant is clinically benign. No studies concerning functional characterization have been found in the literature so far. However, this SNP is frequently associated with clear renal cell cancer in Taiwan and Mainland China, although much less in the Caucasian population (11). Somewhat more prostate and colon cancers have been found in this polymorphism as well. VHL rs779805 polymorphism is not more frequent in colon cancer than others (12). The “G” allele frequency of rs779805 was found to be 0.457 according to the results of the 1000 Genome project in a huge sample (n = 5008 sample size). However, in the subgroup analysis of n = 1000 European samples, it was even less, at 0.297. The “G” allele is also the minor allele in the Asian population, similar to the European data. Interestingly, the “G” allele is the major allele in the African population (allele frequency of 0.844 (https://www.ncbi.nlm.nih.gov/variation/tools/1000genomes/). VHL polymorphism rs1642742 has a similar disorder association and geographic distribution. However, no blood count abnormalities were mentioned in these Chinese reports (13).

We have not detected any known disease-causing VHL mutation. However, we found, even to our surprise, among patients referred with erythrocytosis, 61 patients or probands with the same VHL SNP, namely the intron (c.-195G>A) before the 1st exon: i.e., rs779805 G>A, with GA or AA genotype, similar to what had been described in a Chinese population (11,13) without blood count abnormalities (but associations with solid tumors) (Tables 1–3) No other mutation or SNP has been found. We had 38 patients with idiopathic erythrocytosis, 20 of whom required regular phlebotomy. We screened some of the patients’ relatives. Out of 16 relatives, 5 had erythrocytosis as well. We used the VHL mutation results of 51 control cases from the Department of Laboratory Medicine (Table 1). Samples from 78 patients/relatives and 51 healthy controls were examined. Allele frequency was in Hardy-Weinberg equilibrium for both the control and patient group. The G and A allele frequency values found in our patient’s group (G allele 0.46; A allele 0.54) did not differ significantly from the values calculated for the control group (G allele 0.33; A allele 0.67), p = 0.063. The ratio of the minor G allele carriers (i.e., GG homozygotes and GA heterozygotes combined) was significantly higher in the patient’s group (66%) as compared to the control group (51%) (p = 0.026). However, when erythrocytosis and non-erythrocytosis cases were compared, the AA genotype was dominant: AA 0.37 vs. 0.15 (p = 0.038, power: 89.4%). In idiopathic erythrocytosis, the incidence of GG genotype was lowest, usually with a healthy family member. Only 3 out of 17 (18%) have erythrocytosis (Table 5): GG: 0.03, GA: 0.55, AA: 0.37. If there was no erythrocytosis in the proband group, these were 0.35, 0.50, and 0.06, respectively (p = 0.0017, power 100%). Of the 41 GA genotype patients, 21 had erythrocytosis (Table 3). Erythrocytosis and erythrocytosis requiring phlebotomy were more common in patients with VHL rs779805 GA or AA genotypes than the GG genotype. Interestingly, we also found one patient with polycythemia vera (PV) and one with Myelofibrosis (MF), with low EPO levels with GA genotype, and 3 PV patients with normal EPO levels with AA genotype. The Hgb values of each genotype group did not differ significantly. (Tables 3–6). We identified three families where the same SNP (2 AA and 1 GA genotype) was detectable in family members. In these families, there was an accumulation of erythrocytosis and tumors. Interestingly, either erythrocytosis or tumor occurred, but not both. We have found additional cancers in 14 families: clear renal cell cancer (1), colon cancer (5), melanoma (1), a nodular sclerosis Hodgkin lymphoma, (might or might not be coincidental), a single case of non-Hodgkin lymphoma, bony tumors, gynecological cancers, and hemangioma in first degree relatives. Some patients (during family exploration) had hemoglobin levels over 190 g/L under the age of twenty. The case of patient #10 with an AA genotype is exciting. He was diagnosed with erythrocytosis at the age of 4 and has been regularly phlebotomized ever since (more than 500 times). He has even had six trephine biopsies with indifferent results (Table 4). Upon investigating the whole VHL, we did not find any other genetic variation in our patients. For comparison, we examined VHL polymorphism in 51 healthy volunteers. Hemoglobin level at the upper limit of normal (between 150 and 160 g/L) was observed in 12 healthy individuals, and among them, only GA and AA genotypes were found (n = 7 GA, n = 5 AA).

Rare inherited gene disorders may cause secondary erythrocytosis. These include the genes of proteins involved in oxygen-sensing pathways, such as VHL genes. In addition to the many mutations that cause VHL disease, many SNPs have no clinical significance. It has been shown that in real life, not all polymorphisms are benign. Some may come with malignancies (colon, renal, prostate), some with erythrocytosis. Only a few publications implicate VHL SNPs in inherited erythrocytosis. VHL mutations can cause isolated erythrocytosis. Two homozygous mutations of the VHL, Chuvash (VHL; p.R200W) and Croatian (VHL; p.H191D) polycythemias, both in exon 3, have been reported to cause polycythemia but not a tumor (15,16). Other VHL missense mutations lead to pulmonary hypertension and polycythemia (8). The polymorphism rs779805 has been firstly described in patients with Chuvash polycythemia (17) in a context of a “core haplotype” (also conserved in a patient with von Hippel-Lindau disease) (18). This polymorphism has been linked to renal cell cancer in many papers (11,19–22). GA or GG genotype compared to AA is a risk factor for renal cancer. However, the hypothesis is still discussed, notably in other types of tumors, i.e., decreased risk of prostate cancer (12,13), especially among non-smoking, non-drinking patients without of history of prostate cancer (13). Notably, its link with methylation is discussed: the polymorphism “G” is located in the promoter of the VHL. This gene region is rich in CpG dinucleotides, a so-called CpG island with a sequence “CAA-CG-GCC” that can be targeted by methylation (which may silence the gene expression) (22).

Here we collected data from patients referred with elevated hemoglobin of unknown origin. We could find a special VHL polymorphism (rs779805 G>A) without any other known classical VHL mutation in these patients. These patients have typical characteristics: AA or GA genotypes, high normal or above hemoglobin values (all were over 165 g/L), normal leukocyte and platelet counts, no splenomegaly, no canonical MPN mutations, i.e., Jak-2, calreticulin, MPL, and higher Hgb levels in GA and AA genotype groups compared to GG. Aquagenic pruritus is usually absent. Two cases had very mild, atypical itching. We collected data for 4 years, and no MPN transformation occurred, although the follow-up time was relatively short. Bone marrow examinations were performed in some cases, with normal description and analysis results. We used cautious phlebotomies, targeting <0.52 hematocrit (Htc) levels. This is also recommended by the British Society for Hematology (23). This target Htc is higher than what is suggested in PV, but in most cases, it is a non-compliant case and requires less frequent phlebotomy, which improves compliance. We also recommend 100 mg aspirin daily. All of these patients were advised smoking cessation. We performed an ultrasound and simple cancer screening every year, which we do recommend further on.

We recommend VHL gene sequencing in young patients with otherwise unexplained (isolated) polycythemia, especially if there is any sign of familiarity or unusual precipitation of cancer in the family. We extend our family screening more profoundly, gain data on this polymorphism in more young, healthy volunteers, and screen PV patients. We plan careful hematologic and oncologic follow-up of our patients and first-degree relatives.