Genetic Determinants of Hypercoagulability in Complement and Hemostatic Systems Shape the Thrombotic Risk of Paroxysmal Nocturnal Hemoglobinuria

Thrombotic events (TEs) are the main cause of morbimortality in PNH. Despite a monogenic etiology, PNH displays remarkable heterogeneity as to the hemolytic and thrombotic phenotypes, suggesting the influence of modifying factors. Complement inhibitors introduced another level of variability. The CR1 HindIII polymorphism1 and rare CFH variants2 seem to impact the response to anticomplement therapies, illustrating growing evidence on the role of germline variants as a source of heterogeneity. Clone size, activity parameters, and PIGA clonal burden are predictors of TEs recently validated in our large registry of 267 patients3, but the understanding of the genetic underpinnings of thrombosis in PNH is lagging.

Common and rare variants exacerbating complement activation are associated with thrombotic microangiopathies (TMA); while those in the hemostatic system known as thrombophilias mainly predispose to venous TE (VTE). We hypothesized that germline variants in both complement/hemostatic genes associated with hypercoagulability aggravate the phenotype of PNH, with a focus on the risk of TEs. Consequently, we performed sequencing-based analysis in 109 AA/PNH patients. We screened low frequency (MAF<5%) P/LP/risk variants in 32 complement and in 5 thrombophilia genes. We also assessed a set of risk polymorphisms in complement genes predisposing to TMA, and the minor thrombophilias PTG20210A, FV Leiden and other F5 variants linked to APC resistance. Evidence from public genetic association and functional studies was integrated in our algorithm for enrichment in gain of function variants. In addition, using aggregated data from the All of Us Research Program, we applied the same genetic screening on the following WGS sets: healthy controls (HCs n=66008); VTE (n=12128); TMA (n=33); AA/PNH (n=324).

In our cohort, we identified 15 low frequency complement variants (CFH n=8, C3 n=2, CD55 n=2, CD46 n=1, CFB n=1, THBD n=1), and 13 thrombophilia variants (F5 n=7, PROS1 n=3, SERPINC1 n=2, F2 n=1). The genetic burden was 14% for complement variants and 12% for thrombophilic defects. In All of Us validation sets, the burden of complement variants was 17% in AA/PNH, 18% in HCs, 17% in VTE, 33% in TMA. The burden of genetic thrombophilias was 10% in AA/PNH, 10% in HCs, 15% in VTE, 13% in TMA. As expected, vs HCs, complement defects were increased in TMA (P=.04), and genetic thrombophilia in VTE (P<.01), validating our approach to establish genotype-phenotype correlations. Analysis of the exploratory and validation cohorts did not show a significantly higher burden of these defects in AA/PNH vs HCs. The CR1 polymorphism and a CFH allele in LD with a risk haplotype for TMA were present in 17% and 28% of our cohort, respectively, having a similar frequency in the All of Us sets analyzed.

We then investigated the clinical impact of these variants in our well annotated exploratory cohort. Despite individual allele frequencies comparable to HCs, combinations of CR1/CFH risk alleles and complement variants seemed overall more frequent in hemolytic PNH (14 vs 3% in AA). Indeed, risk alleles plus low frequency variants co-occurred in 6 patients; 5 (83%) of whom had hemolytic PNH and TEs, suggesting a synergy for this germline configuration. Laboratory parameters revealed that the complement/hemostatic genetic burden positively correlated with the plasma levels of D dimer (P=.05), supporting increased hypercoagulability. Within our cohort, 38% PNH patients developed TEs. The combination of common and/or low frequency variants increased the TE rate (no variant: 17%, single variant: 23%, multiple variants: 41% P<.01). Genetic thrombophilia seemed also increased in recurrent TEs (40% vs 11% in cases with no recurrence). In multivariable analysis, adjusting by sex, age and clone size, the complement/hemostatic genetic burden was confirmed as an independent predictor for TEs (HR 2.6 P=.02).

In conclusion, genetic determinants of variability in the complement/hemostatic systems shape the thrombotic risk and phenotype of PNH. Altogether, risk alleles and rare variants triggering/amplifying these cascades seem to have synergistic effects with the PNH clone, precipitating the development of hemolytic PNH and TEs caused by the loss of CD55/CD59. Genotype-functional correlations and assessment of the impact of these variants in the response to complement inhibitors are needed.

Acknowledgments

NIH All of Us Program

AAMDSIF