Lipoprotein(a) impact on bleeding and wound healing probed through digestive tract conditions - Mendelian Randomization study
M Golawski, M L Lejawa, T O Osadnik, M G Gierlotka, N P Pawlas, J J Jozwiak, M B BanachAbstract
Introduction
Lipoprotein(a) (Lp(a)) is recognized as a causal risk factor for cardiovascular disease, though its physiological role remains unclear and insignificant. Lp(a) is postulated to influence blood coagulation and wound healing, suggesting it might be involved in conditions like gastric ulcers or digestive tract bleeding.
Aims
To evaluate the effect of genetically determined Lp(a) on genetically predisposed conditions related to bleeding or wound healing in the digestive system.
Methods
A two-sample MR study was performed. Summary statistics were extracted from publicly available datasets. A meta-analysis by Mack et al. provided correlations between single nucleotide polymorphisms (SNPs) and Lp(a) levels in a European population (n=13,781). The UK Biobank, through analyses by Sakaue et al., Ben Elsworth et al., and Neale Lab, supplied correlations between SNPs and conditions such as bleeding gums (cases = 60,218; controls = 400,895), gastrointestinal bleeding (cases = 5,235; controls = 355,959), gastric ulcers (cases = 6,293; controls = 467,985), gastroesophageal reflux disorder (cases = 32,957; controls = 434,296), and proton pump inhibitor use (cases = 53,137; controls = 79,230). Variants were restricted to the LPA gene region and SNPs not in linkage disequilibrium (R^2 < 0.01). Palindromic variants were removed, leaving 12 SNPs for analysis. Data adjusted for sex and sex+isoform were utilized. The inverse-variance-weighted mean was used as the primary analysis method, with secondary analyses conducted using the weighted median, mode, and MR-Egger methods. Sensitivity analysis was performed using a leave-one-out approach. MR-PRESSO was used to detect horizontal pleiotropy.
Results
Lp(a) was not correlated with any of the assessed traits (all p>0.05 in primary and secondary analyses). The leave-one-out analysis indicated a borderline significant correlation between sex- and isoform-adjusted Lp(a) and gastric ulcers upon excluding some SNPs, which may be attributed to multiple testing. This correlation also had an MR-PRESSO estimate of p<0.05, suggesting a potential horizontal pleiotropy signal. No other significant signals were found in MR-PRESSO or the leave-one-out analysis (all p>0.05).
Conclusions
The impact of Lp(a) on wound healing and blood coagulation, at least within digestive tract, appears to be minimal in European population, given the lack of significant effects on bleeding gums, gastrointestinal bleeding, or gastric ulcers. The absence of an effect on gastroesophageal reflux disorder and proton pump inhibitor use further suggests no impact on gastric acid or mucus production and sphincter function as alternative causal pathways. If Lp(a) has a primary physiological role, it likely pertains to other areas.Figure 1Figure 2