Role of prothrombin 19911 A>G polymorphism, blood group and male gender in patients with venous thromboembolism: Results of a German cohort study

Consecutive patients with a first symptomatic VTE were recruited between December 2008 and 2018 whether or not prothrombotic risk factors were present. Screening of thrombophilia was performed in 1012 adolescent and adult patients [14 to < 60 years] and 902 healthy blood donors [18 to < 61 years] from the same catchment area.

Corresponding to the following criteria patients were included into the study: (i) age from puberty status > Tanner 2 [12] to 60 years at first VTE onset, (ii) objectively confirmed thrombosis and re-thrombosis with established imaging procedures such as Duplex- or Doppler- ultrasonography, computed tomography or magnetic resonance tomography for venous thrombosis and spiral computed pulmonary angiography or lung perfusion scintigraphy for pulmonary embolism. Patients were excluded from the study when they suffer from (i) central line-associated VTE, VTE linked to malignancy, antiphospholipid syndrome, inherited antithrombin-, protein C- or protein S-deficiency or (ii) when they were lost to follow-up or did not abide by the follow-up visits.

The primary study objective was to determine the individual VTE risk of the F2 mutation at rs3136516 in Caucasian VTE patients [n = 1012] compared to healthy blood donors [n = 902] adjusted for (i) common ITs, e.g. F5 at rs6025 and F2 at rs1799963, age at onset and blood group [1‐3, 13, 14]. The secondary study objective was to investigate the time to VTE recurrence after withdrawal of antithrombotic therapy, adjust for the above listed potential confounders.

Following a first VTE onset (in- and outpatients) and during regular follow-up visits [3–6 months (adolescents), 9–12 months, 2 years (adults)] we evaluated and re-evaluated patient’s disease history and possible risk factors causing VTE, such as the use of oral contraceptives, pregnancy, obesity, immobilization, surgery and smoking. Clinical data collection also included laboratory test results, antithrombotic therapy, including adherence to anticoagulation and duration and family history of VTE. Adolescent and adult patients were treated independently from the underlying IT risk factors according to the latest antithrombotic VTE therapy guideline. (Online Supplement) [15‐17].

Details on blood sample collection are depicted in detail in the online supplement.

With written or oral parental consent mutations in F2 at rs3136516, F5 at rs6025 and F2 at rs1799963 as well as circulating levels of antithrombin, coagulation inhibitors, d-dimer concentration as well as lupus anticoagulants and antiphospholipid antibodies were investigated with standard laboratory techniques at VTE onset and were repeated during routine follow up visits [12, 14]. Especially, the analyses of protein C and protein S were performed at least three months after the index event and/or withdrawal of vitamin-K-antagonists and deficiency states of antithrombin, protein C or protein S were confirmed if values repeatedly persisted below the age-related reference ranges [14]. Criteria for the hereditary nature of a hemostatic defect were its presence in at least one further first or second-degree family member and/or the identification of a causative gene mutation.

Single nucleotide polymorphism [SNP] Genotyping for F2 A19911G (LRG_551t1:c.1726-59G>A, rs3136516) was performed by TaqMan PCR. The assay includes two allele specific TaqMan probes containing distinct fluorescent dyes and highly specific PCR primers to detect the target (Assay ID: C__11661574_10, ThermoFisher Scientific, RRID:SCR_018060). ABI 7900HT Fast Real-Time PCR System was used for analysis. Of note, the specific snip applied (rs3136516) does not show a significant deviation from Hardy Weinberg equilibrium in the healthy cohort of blood donors (p = 0.55).

Statistical analyses were performed using the MedCalc^® software bvba (version 16.4.3, Ostend, Belgium, RRID:SCR_015044) and StatView 5 software packages (SAS Institute Inc., RRID:SCR_017411). Continuous variables were expressed as mean (± standard deviation) or median [minimum–maximum] values and categorical data were expressed as counts and percentages. The Wilcoxon–Mann–Whitney-U-test investigated differences in continuous non-normal distributed variables between two independent groups. The Chi-square-test was used to analyze differences in binary and categorical data between for the primary study aim Odds ratios [OR] and their corresponding 95% confidence intervals [CI] were calculated using an univariable logistic regression model comparing patients with healthy controls (blood donors). The homozygous GG genotype and the heterozygous AG were compared with the AA wildtype (F2 at rs3136516). In order to evaluate an independent contribution to the risk of VTE and to adjust for further potential confounders (common IT, age at onset, blood group) [1‐3, 13, 14] these variables were entered into a backward multivariable logistic regression model. In addition, to test for possible statistical interactions between F5 rs6025 and F2 rs313616 genotypes a multiplicative scale was incorporated in the model. Per “a priori” definition this backward model included variables with a p value of ≤ 0.2 and, vice versa, removed variables from the model if p was > 0.21. For the secondary study objective, i.e. the time to recurrence, we calculated the probability of VTE-free survival as a function of time utilizing the method of Kaplan and Meier (univariable analysis). The log rank test was used to test for differences in recurrence-free survival between groups. On the basis of previous reports (data presented in the Online Suppl. Tables 1 and 2) minimum sample size calculation for the comparison of two proportions was performed assuming a type I error of 0.05 and a type II error of 0.20 (sampling comparison of proportions) [18]. Patients were withdrawn from the survival analysis (censored cases) either at death unrelated to VTE recurrence or at loss to follow-up using data of the last clinical follow-up visit. In order to evaluate an independent contribution to the risk of recurrent VTE (dichotomous variables: recurrence “yes” versus “no recurrence”) and to adjust for further potential confounders (gender, common IT, age at first VTE onset, blood group) the hazard ratio (HR) together with 95% confidence intervals (CI) were estimated from Cox’s proportional hazards model. To further test the relationship between independent and dependent variables, the likelihood ratio test was performed. To test the proportional hazard assumption, one of the prerequisites for applying the Log-rank test or the Cox regression model fit, we used the cox.zph procedure in the package ‘survival’ (v3.1.8; Therneau TM, 2019) in R 3.4.3 [19]. A p value > 0.05 indicates no violation of the proportional hazard assumption. The recurrence rates were calculated as the number of recurrent events per 100 person-years.

The underlying multicenter cohort study was approved by the medical ethics committee of the University of Münster & Kiel [B304/16], Germany and written informed consent was provided in all cases prior to study participation.

Compared with the AA wildtype the GG genotype (rs3136516) increases the risk of VTE with an OR of 1.39 (CI 1.04–1.73) and 1.21 (CI 0.97–1.51) in carriers of the AG variant. In addition, in the cohort investigated the OR of F5 carriers compared to healthy blood donors was 6.35 (CI 4.59–8.77) and 2.44 (1.47–4.05) in patients with F2 (rs1799963). Combinations between the homozygous F2 GG genotype at rs3136516 and F5 at rs6025 did not significantly influence the thrombotic risk when comparing patient with controls (OR 0.56; CI 0.29–1.09) and combinations between the factor 2 rs1799963 GA/AA variants and the homozygous GG (rs3136516) genotype were neither detected in cases nor in healthy controls, respectively. BG “non-O” was significantly more often found in patients with VTE compared to BG “O” (OR 3.77, CI 3.04–4.69). Furthermore, increasing age per year at first VTE onset enhances the VTE risk significantly by an OR of 1.01 (CI 1.02–1.04).

Backward logistic regression showed that compared with the AA wildtype the GG genotype (rs3136516) was independently associated with a diagnosis of VTE with an OR of 1.48 (95% CI 1.06–2.06) and with the OR of 1.45 (95% CI 1.1–1.92) in carriers in which the AG variant was genotyped. In addition, the OR of F5 in patients compared to healthy blood donors was 5.68 (95% CI 3.94–8.20) and 2.38 (95% CI 1.35–4.20) in patients with F2 (rs1799963). BG “non-O” was significantly more often found in patients with VTE compared to BG “O” (OR 2.74; 95% CI 2.11–3.55). Again, as depicted in univariable analysis increasing age per year at first onset retained its significance also in multivariate analysis (OR 1.03; 95% CI 1.02–1.04). Due to the “a priori” defined statistical cut-off p-values combinations between F5 rs6025 and F2 rs3135616 (multiplicative scale) were removed from the multivariable model and were not included in the final analysis.

Within a mean time of 51.2 months (standard deviation: 43.5), recurrent VTE events (provoked 77%; locations depicted in Table 1) occurred in 178 out of 1012 patients (17.6%) following the first VTE onset (Fig. 1): 26.5% in men and 13.1% in women (p < 0.001). Gender associated incidence rates per 100 patient years (Table 4) were 4.3 (95% CI 3.4–5.3) in men and 1.9 (95% CI 1.6–2.4) in women (p < 0.0001). Further details of thrombotic locations and AC prior VTE recurrence are described in the online supplement material.

Whereas no significant associations were found between the risk of recurrent VTE and individuals classified as carriers of (i) the F2 mutation at rs3136516 “GG” subjects (OR 1.41; 95% CI 0.9–2.16) or in patients carrying (ii) the F2 rs179963 “GA” genotype (OR 1.81; 95% CI 0.87–3.76), results of the logistic regression model demonstrates that (iii) heterozygous F5 carriers had an increased OR of 1.86 (95% CI 1.22–2.83) to suffer from recurrent VTE with a further increase in patients additionally carrying the F2 GG/GA variant (rs3136516): OR 3.34 (95% CI 1.62–6.89). Patients with at least any IT compared to patients with no IT showed an increased odds ratio of 1.7 (95% CI 1.2–2.3) for a recurrence of VTE, while in patients with combined ITs the OR increased to 2.4 (95% CI 1.7–3.4). Of note, there was no significant association between the presence or absence of provoking risk factors and inherited thrombophilia (p = 0.93). The recurrence-free survival is depicted in Fig. 2: 32.7% of patients carrying combined ITs compared to 15.6% without IT developed a second thromboembolic event within 36 months (min–max: 1–120; log rank test: p < 0.0001). Multivariable analysis including time to recurrence demonstrates that male gender [hazard ratio (HR) 2.3; 95% CI 1.7–3.0] and combined ITs (HR 2.11; 95% CI 1.5–3.0) showed a significant impact on the risk of VTE recurrence. Noteworthy, in this specific data set, gender, F5, F2 rs3136516, F2 rs179963 and blood group did not violate the proportional hazard assumption. Remarkably, age, both as continuous term or grouped into three equal sized groups did violate the proportional hazard assumption and was thus not included in the multivariable analysis. Due to the “a priori” defined statistical cut-off p values the comparison between BG O and non-O was removed from the multivariate model. Incidence rates of recurrence per 100 patient years are depicted in detail in Table 4.