The genetic differences between types 1 and 2 in von Hippel-Lindau syndrome: comprehensive meta-analysis

Background Patients with von Hippel-Lindau (VHL) disease are at risk of developing tumors in the eye, brain, kidney, adrenal gland, and other organs based on their gene mutations. The VHL tumor suppressor gene contains pathogenic variants responsible for these events. This meta-analysis aims to investigate the genetic differences among the various types of VHL syndrome and their correlation with the location of mutations (exons and domains) in the VHL gene. Method Papers eligible for publication until September 2023 were identified using the electronic databases of PubMed, Google Scholar, Scopus, and EMBASE. The Random Effect model was utilized to evaluate the genetic differences between type 1 and type 2 VHL syndromes. Results The prevalence of missense mutations (MSs) was found to be 58.9% in type 1, while it was 88.1% in type 2. Interestingly, the probability of observing MSs in type 1 was 0.42 times lower compared to type 2. The mutation hotspots of the VHL gene were R167Q/W, Y98H, R238W, and S65L, respectively. Although type 2 had a high presentation of Y98H and R238W, it did not have a higher S65L than type 1. The analysis demonstrated a statistically significant higher prevalence of truncated mutations (PTMs) in type 1. Among type 1, large/complete deletions (L/C DELs) were found in 16.9% of cases, whereas in type 2 only 3.7%. This difference was statistically significant with a p-value < 0.001. Overall, the probability of identifying mutations in domain 2 compared to domain 1 was found to be 2.13 times higher in type 1 (p-value < 0.001). Furthermore, the probability of detecting exon 1 in comparison with observing exon 2 in type 1 was 2.11 times higher than type 2 and revealed a statistically significant result (p-value < 0.001). The detection of exon 2 was 2.18 times higher in type 1 (p-value < 0.001). In addition, the likelihood of discovering exon 2 compared with others was significantly lower in type 1 compared with type 2 VHL (OR = 0.63, p-value = 0.015). Conclusions We have revealed a comprehensive genetic difference between types 1 and 2 of VHL syndrome. The significant differences in MS, PTMs, L/C DELs, and the location of the mutations between type 1 and type 2 VHL patients in the Asian, European, and American populations emphasize the genetic heterogeneity of the syndrome. These findings may pave the way for the diagnosis, treatment, and further investigation of the mechanisms behind this complex genetic disorder.


Introduction
Von Hippel-Lindau (VHL) syndrome is a genetic disorder that is uncommon and autosomal dominant and affects approximately 1/36,000 individuals.Multiple lesions, including retinal capillary hemangioblastoma (RCH), central nervous system hemangioblastoma (CHB), renal cell carcinoma (RCC), pheochromocytomas (PCC), and as well as pancr eatic cyst (PC), kidney cysts (KC), pancreatic neuroendocrine tumors (PNETs), endolymphatic sac tumor (ELST), epididymal cyst, PGL (paragangliomas), and liver cysts have been reported in affected individuals [1,2].A germline pathogenic variant in the tumor suppressor gene VHL, located on chromosome 3p25-26 [3,4], is found 80% of cases with a family history, while 20% of cases are de novo mutations [5].The VHL gene is responsible for encoding the VHL protein (pVHL), which is a crucial determinant of the hypoxia-inducible transcription factor α (HIF-α) [6].A complex consisting of elongin B, elongin C, and Cul 2 proteins is formed by pVHL.Under normal conditions, hypoxia-inducible factors (HIFs) are degraded by this complex [3].However, when pVHL deficiency occurs, an overexpression of HIF-α leads to unregulated angiogenesis and highly vascularized tumor development [5,7,8].Nonetheless, there are still some uncertainties about these processes, and it is crucial to identify any variants of VHL syndrome to enhance the diagnosis of disease and the precision of the genetic testing.
VHL is classified into two categories: Type 1 and Type 2. Type 2 can be divided into subtypes A, B, and C. VHL disease type 1 is recognized for its low risk of PCC, but it tends to develop CHB, RCH, PC, RCC, and PNETs.On the other hand, the type 2 VHL disease is known for its high risk of PCC and is classified into type 2 A with low risk of RCC, type 2B with high risk of RCC, and type 2 C which only develops PCC (Table 1) [1,2,9,10].
Recently, Belzutifan, a small molecule oral HIF 2 α inhibitor, has been approved by the US Food and Drug Administration for treating adults with RCC, CHB, and PNET, which is linked to VHL disease.It may be restricted to VHL patients with positive genetic tests, and the detection of VHL variants.Therefore, assessing or annotating each variant becomes more important for the treatment strategy of VHL disease [11].
This meta-analysis study aims to decipher the genetic differences between type 1 and type 2 VHL syndromes in a comprehensive way, which will help us understand the typical mutations in the VHL gene that cause VHL syndrome.
The analysis we conducted to study VHL manifestation utilized VHL disease types 1 and 2 (A, B, and C)).We examined mutation types such as missense mutations (MSs), protein-truncating mutations (PTMs) (Nonsense, frameshift, and splice site, and small insertion/deletions (INDELs) of VHL protein), large/complete deletions (L/C DELs), and variant locations (exon and domain) as described in our recent study [12].

Search strategy
The electronic databases PubMed, Scopus, EMBASE, and Google Scholar were searched to find papers eligible by September 2023.The keywords were selected based on MeSH terms or text words.The search syntax was: (von hippel lindau OR VHL OR von Hippel-Lindau OR VHL gene) AND (retinal capillary hemangioma OR RCH OR Retinal hemangioblastoma OR RH) AND (central nervous system OR CNS OR central hemangioblastoma OR CHB OR central nervous system hemangioblastoma OR CNS-HB) AND (pancreatic neuroendocrine tumor OR PNET) AND (Pancreatic cyst OR PC) AND (renal cell carcinoma OR RCC) AND (kidney cysts OR KC) AND (pheochromocytomas OR PCC) AND (Mutations OR Genotype) AND (Types OR Phenotype).

Eligibility criteria
The inclusion criteria were: the study explored the relationship between VHL gene alterations and VHL syndrome; the evaluation involved at least one type of VHL variant or the location of the variants, and papers with available full text in English.The Exclusion criteria were: review articles, case reports, letters, and comments; and the reported information was not sufficient for analysis, as in our previous study [23].

Data extraction
The related data was extracted and evaluated separately by two independent reviewers (FA and SCH).An Excel data sheet was created to obtain the necessary data from the documents.We entered the data for every patient because most articles reported information about people individually.The data collected includes the first author, country, continents (Asia, Europe, and America (Brazil, These findings may pave the way for the diagnosis, treatment, and further investigation of the mechanisms behind this complex genetic disorder. Mexico, and USA)), year of publication, type of variants, type of VHL disease, and the location of the mutations.

Statistical analysis
STATA version 16 was used to conduct the statistical analysis.Odds Ratio (OR) was considered as the effect size.The analysis relevant to primary studies was utilized due to the individual cases were considered in this metaanalysis.The use of random effects model was necessary to draw inferences.The power of this model lies in its ability to analyze correlated data [21].Our meta-analysis considers each publication as a cluster, and the cases of each publication are correlated with each other.The random effect model was employed to examine correlation.The correlation between the subjects was controlled by adding a random parameter.The model below is a typical random effect models: The probability of having VHL phenotypes is represented by π ij , β0 is responsible for the intercepting, X represent the vector of independent variables, β is the parameter that controls independent variables, and random parameter Ui is responsible for the publication random effect.Adding Ui to the model makes the models consider the correlation within each publication [11].The Odd Ratio (OR) was used to calculate the association between types of VHL, variant types, and locations.A P value < 0.05 was considered to be significant.

Results
The process of selecting papers is depicted in Fig. 1.This study included 74 publications that were eligible (Table 2).We analyzed data from 2261 patients according to type of VHL disease.Of these, 1630 patients (72.1%) were classified as Type 1 and 631 patients (28%) were labeled as Type 2. 25% of Type 1 cases had a PTM variant of VHL.On the other hand, of the 631 patients diagnosed with Type 2, MSs were found in 88% (Fig. 2, A).We found that the mutation hotspots of the VHL gene include S65L, R238W, Y98H, and R167Q/W, which have a frequency between 2.4 and 7.2% (Fig. 2, B).The frequency of the mutation hotspots in the VHL gene was shown in Fig. 2, C.
In type 1, the mutation hotspots Y98H and R238W were much lower than in type 2. Also, in type 2, the mutation hotspots S65L was lower than the type 1 (Table 3).
Table 4 displays the correlation between mutations and syndrome types, both globally and by continent.The analysis revealed that the prevalence of MSs was 58.9% in patients with type 1 VHL, while it was 88.1% in patients with type 2. Notably, the probability of observing MSs in type 1 patients was 0.42 times lower compared to type 2 patients, with a statistically significant p-value < 0.001.However, analysis demonstrated a statistically significant higher prevalence of PTMs in type 1 patients (25.3% vs. 8.4%).
The probability of detecting PTMs was 2.01 times higher in type 1 patients compared to type 2 patients (p-value < 0.001).The research revealed that there was a discrepancy in the occurrence of L/C DELs between Type 1 and type 2 patients.In type 1 patients, L/C DELs were found in 16.9% of cases, while in type 2 patients, the occurrence was only 3.7%.Additionally, there was a 2.1 times greater chance of detecting L/C DELs in type 1 patients than in type 2 patients.This difference was statistically significant with a p-value < 0.001.
The analysis of Asian publications and the overall data indicated a significant difference in MSs occurrence between types 1 and type 2 VHL patients.In type 1 patients, MSs were found in 69.1% of cases, while in type 2 patients it was 85.3%.Furthermore, the odds ratio (OR) for MSs in type 1 patients was found to be 0.7 when compared to type 2 patients.The statistical analysis showed a significant difference with a p-value of 0.036.However, no statistically significant correlation was observed between the types of VHL and the occurrence of PTMs or L/C DELs.
The prevalence of MSs was lower in type 1 patients in Europe compared to type 2 patients (55.8% vs. 88.6%).The study showed that there was a 0.33 lower chance of MSs in type 1 patients, which indicates a statistically significant difference (p-value < 0.001).
Analysis revealed that PTMs occurred more frequently in type 1 than in type 2 (21.6% vs. 6.6%).It was found that the likelihood of observing PTMs was 2.72 times higher, demonstrating a statistically significant OR (p-value < 0.001).Furthermore, type 1 patients showed a statistically significant OR of 2.14 (p-value < 0.001) with more frequent detection of L/C DELs (23.1 vs. 4.8%).It seemed that MSs were observed more frequently in type 2 in America compared to type 1 (92.8%vs.49.4%), which indicated a statistically significant OR of 0.36 (p-value = 0.013).In addition, type 1 had a higher rate of L/C DELs than type 2 (15.3% vs. 2.5%).Analysis displayed a statistically significant OR of 2.77 (p-value = 0.002).Type 1 had a higher rate of PTMs than type 2 (35.7% vs. 6%), but there was no significant relation between observing PTMs and types of VHL (Table 4).
Table 5 shows the association between domains and VHL types.Detecting mutations in domain 2 was 2.13 times more likely in type 1 VHL patients than in domain 1 (p-value < 0.001).The probability of detecting domain 2 was 3.75 times higher in type1 VHL patients in Asians compared to domain 1, demonstrating a statistically significant relationship (p-value < 0.001).
Also, when investigating Europeans, the analysis revealed a statistically significant association between finding domain 1 or domain 2 and types of VHL (p-value = 0.012).The probability of occurrence of domain 2 was 3.23 times higher in type 1 patients compared to type 1.There was no significant relation between domains and VHL types in Americans.
Additionally, detecting mutations in exon 1 compared to others was 0.52 lower (p-value = 0.031).Type 1 VHL patients had a significant difference in the occurrence of mutations in either exon 2 or exon 3 compared to type 2. In type 1 patients, the probability of detecting mutations in exon 2 was 2.35 times greater (P-value < 0.001).The chance of observing exon 3 mutations was 0.26 times lower in type 1 patients compared to type 2 patients (p-value < 0.001) (Table 5).There was no other association that was significant (Table 6).
In Europe, the probability of finding mutations in exon 1 rather than exon 2 was significantly lower in type 1 VHL patients compared to type 2 by 0.78 times (p-value = 0.019).Also, detection of mutations in exon 1 was significantly higher in type 1 patients compared to type 2 patients (OR = 2.11, p-value < 0.001).Mutations in exon 1 were 0.76 less likely to occur in type 1 patients than type 2 patients (p-value = 0.023).Moreover, detecting mutations in exon 2 compared to exon 3 was significantly higher in type 1 patients by an OR = 2.74 (p-value < 0.001).Type 1 patients had significantly lower odds of finding mutations exon 3 than type 2 patients, with an OR = 0.35 (p-value < 0.001).No other significant association was found (Table 6).
Type 1 patients had a significant differences in exon 1 compared to exon 2, resulting in an OR of 2.04 (p-value = 0.041) when compared to type 2 patients.A significant OR was observed when comparing the occurrence of exon 1 mutations in type 1 VHL patients with those in type 2. Detecting mutations exon 1 was 0.39 times lower in type 1 patients (p-value = 0.035).The probability of finding exon 2 mutations when compared to other exons in type 1 patients was 0.19 times lower than in type 2 VHL patients, and this result was statistically significant (p-value = 0.023).In addition, type 1 patients had a significantly lower probability of finding mutations in exon 3 than type 2 patients (OR = 0.18, p-value < 0.001) (Table 6).

Discussion
The current classification of VHL type 1 and type 2 primarily relies on clinical observations and the probability of PCC or RCC.Although studies have attempted to establish a connection between VHL genotype and phenotype, their clinical relevance is restricted.This implies that their diagnostic and therapeutic benefits are limited.However, given that patients are susceptible to tumor development throughout their lives, having a predictive outlook on the onset of VHL-related cancers could allow clinicians to devise a more personalized surveillance strategy when presented with a unique mutation.
The presented findings provide valuable insights into the genetics of VHL syndrome, specifically how different mutation types/locations related to the two distinct subtypes of VHL (type 1 and type 2).
In total, we analyzed 2,261 patients with VHL and identified 72.1% type 1 and 28% type 2, similar to the results of Tamura et al. [22].This study revealed that mutations in the VHL gene are most commonly found at R167Q/W, Y98H, R238W, and S65L, respectively.Intriguingly, the mutation hotspots Y98H and R238W were present higher in type 2, especially type 2 A and 2B, which agrees with multiple studies [22][23][24][25][26]. Additionally, S65L was much lower in type 2 than in type 1.It is recommended that patients with mutations Y98H and R238W are likely to have several severe symptoms of the disease.Therefore, it is better to get timely screening to prevent the spread of the disease by this mutation in before they develop other symptoms.this study, type 2 have a greater chance of developing MS than type 1.On the contrary, the probability of finding L/C DELs or PTMs is almost twice as high in type 1 VHL patients, as supported by several findings [8,[27][28][29][30].The disparities between mutation types in type 1 and type 2 VHL patients demonstrate the genetic heterogeneity of VHL syndrome and its correlation with the disease's manifestation [25].
MSs, which often hinder the proper folding of the VHL protein, are more prevalent in type 2 VHL.This trend is true in different geographic regions.Whether the higher predisposition towards RCC or PCC is primarily attributed to the inherent nature of misfolded proteins or to the specific effects of variant MSs with higher prevalence remains uncertain.While it has been indicated that certain MSs in the VHL gene's hotspot region can lead to a less stable protein that can lead to RCC development [31], but it is important to take into account the prevalence of these mutations when evaluating their overall  It appears that the incomplete VHL protein has a less comparatively less detrimental effect on predisposing individuals to RCC and PCC, especially when L/C DELs occur in specific regions of the gene [32][33][34][35].The observation that L/C DELs involving specific gene locations are less likely to result in type 2 VHL may offer an explanation for the different effects of these types of mutations in different populations.
In the Asian population, there was no significant association between L/C DELs and VHL types.Conversely, based on the available reports, L/C DELs are more prevalent in type 1 VHL among American and European populations.This suggests that Asian populations may have L/C DELs that predispose them more to RCC, however, we could not find a higher incidence of L/C DELs in the Asian population with type 2. Compared to other populations, Asians have a significantly lower prevalence of L/C DELs in Type 1.On the contrary, the higher prevalence of PTMs in type 2 VHL of Asians (compared to the other two populations) seems to be the reason why the chance of this genotype having different VHL subtypes is insignificant.Different types of PTMs in the Asian population could explain this finding, but a thorough investigation is needed to ascertain that.
The analysis of genetic mutations in Asian, European, and American populations in the context of VHL syndrome has provided a heterogeneity of VHL-associated mutations, specifically MSs, PTMs, L/C DELs, and their association with different VHL subtypes, which is similar to the study of Dollfus et al. [36] and Fagundes et al. [37].
We also demonstrated that the chance of finding a mutation in the β domain of the VHL gene is twice as high as finding a mutation in the α domain in type 1 VHL.Asian publications had a higher probability of domain mutation in type 1 VHL, while American publications did not have any significant odds of any domain mutation.It has been clarified that type 2 VHL is most often cause by mutations in Elongin C binding found in the α domain of the VHL gene, which is similar to Maxwell et al. [38].On the other hand, mutations in the β domain that affect   HIF are predominantly found type 1 VHL.study's findings confirm the reported VHL phenotype and underlying However, we were unable to determine the specific mutated gene in each domain because of insufficient data.Similar findings were observed when assessing the association between exon mutations and VHL phenotypes.Subgroup analyses by continent revealed that in the Asian population, the observance of exon 1 rather than exon 3 was significantly higher in type 1 VHL patients compared to type 2. Exon 2 mutations were significantly higher in type 1 Asian patients, while exon 3 mutations were significantly lower, suggesting that other exons play a more prominent role.The European population showed significant difference in observing exon 1 in type 1 patients compared to type 2. The likelihood of detecting exon 1 compared to exon 3 was also significantly higher in type 1 European patients.Additionally, type 1 patients were more likely to have exon 2 mutations than exon 3 mutations.
In the American population, significant associations were found between exon 1 and exon 2, with exon 1 being more prevalent among type 1 patients.In type 1 patients, there was significant decrease in the chance of detecting exon 2 compared to other exons.Furthermore, there was a significant difference in exon 3 mutations in type 1 patients compared to type 2.
In summary, mutations in exons 1, 2, and 3 have distinct associations with VHL subtypes, suggesting that the location and nature of genetic alterations within the gene may influence disease presentation.

Conclusion
This study demonstrates that there is a significant variation in the types of mutations that are associated with different VHL subtypes in various populations.The significant differences in MSs, PTMs, L/C DELs, and the location of the mutations between type 1 and 2 VHL patients in the Asian, European, and American populations underscore the genetic heterogeneity of the syndrome.These findings emphasize the importance of considering regional and ethnic variations when studying VHL genetics.
The observed variations in mutation patterns between different populations may have implications for patient diagnosis and management, as well as for future research into the underlying mechanisms of VHL syndrome.Understanding these genetic differences can inform targeted therapies and personalized treatment approaches tailored to the specific mutation profiles of VHL patients in different populations.

Fig. 1
Fig. 1 Flow diagram eligibility criteria of articles on VHL Mutations in VHL

Fig. 2
Fig. 2 Clinical types and VHL variants.(A) Families whose described phenotype does not include PCC (Type1, n = 1630).(B) Families whose described phenotype includes PCC (Type 2, n = 631).(C) The frequency of the mutation hotspots of the VHL gene

Table 1
Classification of VHL syndrome

Table 2
Summary of data collection of VHL mutations in VHL disease of

Table 3
The frequency of mutation hotspots of the VHL gene

Table 4
The correlation between mutations and types of syndrome, both in total and by continent

Table 5
The relation between domains and VHL types

Table 6
The relation between mutation in exons and types of VHL