Role of vascular endothelial growth factor polymorphisms (-2578C > A, -460 T > C, -1154G > A, +405G > C and +936C > T) in endometriosis: a case–control study with Brazilians

Endometriosis is a benign estrogen-dependent disease, characterized by the presence and growth of endometrial tissue outside the uterus, and represents one of the most common benign gynecological disorders nowadays [1]. This disease is associated with infertility, severe and incapacitating painful symptoms, including chronic pelvic pain, dysmenorrhea and dyspareunia [2, 3]. It has been estimated that endometriosis affects 10% of women of reproductive age, but the real prevalence may even be higher because it is often not diagnosed due to its heterogeneous clinical manifestation [4]. Endometriosis frequently produces serious effects on professional, social and marital life [5].

Despite many investigations about endometriosis, the pathogenesis of the disease remains unclear, although the predominant theory is that it is due to retrograde menstruation [6]. In addition, endometriotic lesions require an adequate blood supply to survive in their ectopic sites, and angiogenesis represents a crucial step during this process [7]. The development of new blood vessels is a complex dynamic process, which is regulated by a signal sequence of different angiogenic factors. The Vascular Endothelial Growth Factor (VEGF) is one of the most potent angiogenic factors and several authors postulated that it would be involved in the progress of the ectopic lesions in endometriosis [8, 9]. Accordingly, our group demonstrated that VEGF-induced angiogenesis is a critical aspect in the pathophysiology of this disease [10‐12].

VEGF is encoded by the VEGF gene [13], which is polymorphic, with several single nucleotide polymorphisms (SNPs) in regulatory regions [14]. Recently, there is growing interest in investigating if VEGF SNPs may affect the inheritable susceptibility to endometriosis [15‐18]. The results are conflicting, possibly due to the diversity of populations studied and because endometriosis is a heterogeneous disease [15]. In addition, no investigation regarding the susceptibility to endometriosis considered the combined effect of the five most studied VEGF SNPs (-2578C > A, -460 T > C, -1154G > A, +405G > C and +936C > T) in their possible haplotypes.

No significant difference was observed in the mean age between the endometriosis patients (35.8 ± 8.6) and the control group (34.5 ± 6.4). Conversely, BMI, parity, number of spontaneous abortion, infertility and all preoperative endometriosis symptoms were significantly different between the two groups (Table 2). There was a predominance of low or normal BMI values (≤24.9) among endometriosis patients (75.1%), whereas controls had a predominance of overweight or obesity (58.4%), with 3.4% of patients showing morbid obesity (BMI ≥ 40). The distribution of endometriotic patients according to the worst endometriotic lesion was as follows: DIE (151 patients; 83.0%) and not DIE (31 patients; 17.0%).

The VEGF -2578C > A, -460 T > C, -1154G > A, +405G > C, +936C > T SNPs were in HWE in the overall study population and in each group (cases and controls). Figure 1 and Table 3 show, respectively, the minor allelic and genotypic frequencies of the VEGF SNPs. Significant differences in the allele and genotype frequencies were observed between the two groups with respect to the -1154G > A (P = 0.005 and P = 0.01, respectively). By contrast, no significant differences were detected in allele or genotype distribution of the -2578C > A, -460 T > C +405G > C, +936C > T SNPs between endometriosis patients and controls. The analysis of risk associations for the -1154G > A in developing either endometriosis or DIE (Table 4) suggests an approximate 2-fold increased risk for individuals with any variant genotype (GA + AA), or an approximate 6-fold increased risk for individuals with the homozygous variant genotype AA. Although no statistically significant risk association was detected for individuals with the heterozygous variant genotype (GA), a codominance model was inferred for the -1154G > A polymorphism (P_trend = 0.008).

Haplotypes of the VEGF gene were determined for all patients and also for endometriosis cases and controls separately. The results revealed that SNPs -2578C > A, -460 T > C, -1154G > A and +405G > C were in strong linkage disequilibrium, forming a single haploblock, while +936C > T was not linked to the other SNPs (Figure 2). Therefore, haplotype analysis was only conducted between VEGF -2578C > A, -460 T > C, -1154G > A and +405G > C SNPs, and six haplotypes were inferred (Table 5). There was negative risk association for the development of endometriosis for the haplotypes CCGG and ATGG, when compared with the reference haplotype CTGG, either considering all cases, only DIE patients or stages III-IV of endometriosis. In addition, the haplotype ATGG showed negative risk associations for the development of endometriosis when considering all cases or DIE, but not stages III-IV, whereas the haplotype CTGC was protective only for the development of stages III-IV.

The pathogenesis and the molecular mechanisms that underlie the development of endometriosis have troubled investigators through many years, remaining an enigma. Endometriosis is regarded as a complex trait in which genetic and environmental factors contribute to the disease heterogeneous phenotype. Regarding the epidemiological evaluation of the study population, we observed that women with endometriosis have lower BMI and are less frequently obese than control subjects. Our results corroborate previous findings [25‐29], although the reason for inverse correlation between BMI and endometriosis risk is still unclear. It can be hypothesized that genetic factors contributing to endometriosis may also be linked to BMI [30, 31]. Although epidemiological data can be used to better understand the endometriosis, further studies should investigate the genetics, environmental and physiopathological basis of the decreased BMI in women with endometriosis.

Because angiogenesis represents a critical step in the establishment and pathogenesis of endometriosis, this process has been viewed as a potential new target to better define the mechanisms that cause the disease. A large number of studies have observed that VEGF was significantly higher in women with endometriosis, which supported a key role for VEGF in the pathological angiogenesis in endometriosis [9‐11]. Polymorphisms in VEGF may alter protein concentrations, influence the process of angiogenesis and relate to inter-individual variation in the risk of endometriosis. The promoter, and the 5’- and 3’- UTR of the VEGF gene contain key regulatory elements, which contribute to the high variability in VEGF production among tissues [14, 32, 33].

The inheritable susceptibility to endometriosis justifies the growing interest in identifying genetic polymorphisms that could lead to an increased risk or severity of the disease, in order to provide additional support for treatment planning. The present results indicate a positive association between VEGF -1154G > A and the risk of developing endometriosis, which is maintained when considering only the cases of DIE or stages III-IV. Such risk association was not observed previously [34‐36]. Thus, Liu et al. [34] proposed that the -1154AA genotype decreased endometriosis risk compared to the -1154GG genotype, whereas the latter reports showed no difference in the distribution of VEGF -1154G > A genotypes between cases and controls [35, 36]. Nevertheless, recent studies suggest that the VEGF -1154G > A SNP poses an increased risk of recurrent spontaneous abortion [37, 38]. Because such studies did not evaluate the occurrence of endometriosis as a possible cause of the recurrent spontaneous abortions, it cannot be excluded as a confounding factor in the association analyses. In addition, it has been reported that the frequency of the VEGF -1154G > A SNP in Brazilians might be different between individuals self-identified as “Blacks” or “Whites” [39]. The present study did not collect information on race or skin color. However, all individuals came from the same region of Brazil, had similar social backgrounds, and were recruited at two public hospitals, when assigned for laparoscopic procedures, regardless of the therapeutic indication. Therefore, no major racial or color differences is expected between cases and controls, which had equal access to the public health system.

With regards to the other four VEGF SNPs (-2578C > A, -460 T < C, +405G > C, +936C < T), our results suggest no significant effect on the susceptibility to endometriosis. It is noteworthy that our result is in agreement with Zhao and colleagues [40], which suggested no evidence for an association between endometriosis and the VEGF -2578C > A, -460 T < C, +405G > C and +936C < T SNPs, when considered together in a larger number (958 cases and 959 controls) of Australian women. Such findings appear to be corroborated by other studies which evaluated the different VEGF SNPs independently from their effect on the risk of endometriosis in different populations, and found no significant associations with -2578C > A[16, 40], -460 T < C[16, 34, 40‐48], +405G > C[16, 35, 40, 43, 44, 49, 50] or +936C < T[34, 35, 40, 51]. Nevertheless, results from a meta-analysis suggest that the VEGF -2578C > A might be protective for the development of endometriosis [18], whereas +936C > T was pointed as a risk factor [16‐18]. The increased risk of endometriosis for +936C < T was found independently on a single study, although the SNP showed no correlation with VEGF mRNA in endometriosis lesions or VEGF protein levels in peritoneal fluid [44]. Accordingly, Kim and colleagues [51] showed a lack of association between +936C < T genotypes and serum VEGF levels in endometriosis patients and controls.

As far as we know, the present work is the first study to focus on the possible contribution of the five most studied VEGF SNPs (-2578C > A, -460 T > C, -1154G > A, +405G > C and +936C < T) and its haplotypes on the susceptibility of endometriosis. In agreement with previous studies, -2578C > A, -460 T > C, -1154G > A[34] and -2578C > A, -1154G > A, +405G > C[35] were in linkage disequilibrium, while the +936C < T was visibly physically far, and had low LD with the other 4 markers in the gene [34, 35]. Only three studies reported association between VEGF haplotypes and susceptibility to endometriosis; however, the haplotypes with only two [41, 46] or three SNPs [34, 35] were evaluated. In the present study, we observed negative risk associations with the development of endometriosis for the haplotypes CTGC (only for stages III-IV), ATGG (for all cases combined or DIE), and CCGG haplotype (for all conditions). The haplotype ACAG, which was the only one containing the -1154A allele showed a non-significant positive risk association for endometriosis, in all conditions evaluated. Taken together, the results suggest that the effects of VEGF haplotypes in the risk of endometriosis are more significant and clinically relevant than those of each SNP evaluated separately. It is becoming increasingly important to derive data from different populations to build a database which can then be used in future investigations to a better understanding of the genetic and environmental factors affecting risk to development endometriosis.

In conclusion, our findings with VEGF SNPs and endometriosis in Brazilian women indicate a risk association for the polymorphism -1154G > A, and protective effect for the haplotype CCGG. This is the first study to evaluate the combined influence of the five most common VEGF SNPs. Therefore, further studies on the functional relevance of the VEGF polymorphisms and exposure to environmental factors in endometriosis are required to confirm our observations.