New Insights into the Genetic Association of Lipids and Lipid‐Modifying Target Genes with Pancreatitis

Pancreatitis is a complex and progressive inflammatory disease of the pancreas, associated with a high risk of morbidity and mortality. It is well known that pancreatitis is a disease progression from the initial episode of acute pancreatitis (AP) to chronic pancreatitis (CP) via recurrent AP (RAP). However, little is known about identifying at risk of recurrences and progression or about pharmacological interventions to prevent these outcomes. While hypertriglyceridemia is a common cause of AP, with pancreatitis secondary to it typically presenting as an episode of AP or RAP, and rarely as CP. While it has been clearly established that the risk of AP correlates positively with triglyceride (TG) levels [1], and that familial chylomicronemia syndrome is extensively documented as a cause of pancreatitis spanning AP, RAP, and CP [2], a comprehensive genetic investigation of these causal links is still warranted. Furthermore, there is no clear casual effect evidence for lipid-lowering drugs (i.e., statins or fibrates) in reducing the likelihood of pancreatitis. Therefore, we performed Mendelian randomization (MR) analysis to assess the causal associations between lipid levels and AP, CP, alcohol-induced AP and CP (AAP and ACP) using data from the FinnGen Consortium. Drug-target MR and colocalization analyses were further conducted to investigate the potential impact of 15 lipid-modifying targets on these pancreatitis subtypes. To ensure the validity of instrumental variables, a positive control analysis was conducted in parallel to test causal effect of selected druggable targets on coronary artery disease, which has a well-established causal association (see Supplementary Material for detailed methodology). In the primary univariable MR analyses, we found that genetically proxied increases in TG levels were associated with a higher risk of all pancreatitis subcategories. In contrast, genetically proxied increases in high-density lipoprotein cholesterol (HDL-C) levels were causally associated with a decreased risk of AP, AAP, and CP. To account for potential confounding factors, we performed multivariable MR analyses. Specifically, we adjusted for genetically predicted cholelithiasis in AP and for genetically predicted alcohol consumption and smoking in CP. We also conducted multivariable MR analyses for AAP and ACP after adjusting for genetic liability to alcohol consumption. Our results were highly consistent across both univariable and multivariable analyses, showing that genetically proxied levels of TG and HDL-C are causally associated with pancreatitis, whereas no significant association was found between either total cholesterol or low-density lipoprotein cholesterol and any pancreatitis subcategory. In the drug-target MR analyses, we attempted to evaluate effects of 15 lipid-modifying targets on pancreatitis, namely, HMGCR, NPC1L1, PCSK9, APOB, LDLR, ABCG5, ABCG8, ACLY, ANGPTL3, MTTP, PPARA, PPARD, PPARG, APOC3, and LPL. For TG (Table 1), higher genetically proxied lipid levels mediated by APOB increased the risk of CP (odds ratio [OR] = 2.057, 95% confidence interval [CI] 1.290–3.280, p = 0.002) and ACP (OR = 2.236, 95% CI 1.163–4.297, p = 0.016). Likewise, a potential association of APOC3 with CP (OR = 1.366, 95% CI 1.111–1.679, p = 0.003) and ACP (OR = 1.442, 95% CI 1.080–1.926, p = 0.013) was identified. For HDL-C (Table 1), targeting PPARG appeared to decrease the risk of AP (OR = 0.325, 95% CI 0.143–0.740, p = 0.007), AAP (OR = 0.045, 95% CI 0.006–0.363, p = 0.004), CP (OR = 0.295, 95% CI 0.100–0.871, p = 0.027), and ACP (OR = 0.187, 95% CI 0.041–0.852, p = 0.030), while targeting APOB and APOC3 appeared to be associated with reduced risk of CP (APOB: OR = 0.459, 95% CI 0.262–0.805, p = 0.007; APOC3: OR = 0.580, 95% CI 0.364–0.924, p = 0.022), and ACP (APOB: OR = 0.417, 95% CI 0.190–0.916, p = 0.029; APOC3: OR = 0.412, 95% CI 0.214–0.972, p = 0.008). The posterior probability of colocalization between TG and CP in the APOC3 gene region was 79% conditional on the presence of a causal variant for the outcome (Table 1). To enhance the robustness of our findings, we conducted replication analyses using independent datasets from AP (UK Biobank), as well as East Asian and European consortia for both AP and CP. Replication of the APOB-mediated effect on TG levels revealed a significant association with CP (European) (OR = 2.192, 95% CI 1.100–4.370, p = 0.026) (Table 1), consolidating our primary finding. Although the association for APOC3 did not reach statistical significance in these replication datasets, the LPL-related TG levels, which lacked a significant association in our primary analysis, demonstrated a clear association with AP in the East Asian (OR = 2.800, 95% CI 1.578–4.966, p = 0.000) and European (OR = 1.233, 95% CI 1.001–1.519, p = 0.049) datasets (Table 1). This newly confirmed association with LPL provides crucial evidence supporting its established clinical link with AP. The drug-target analyses reinforce our core findings by supporting the causal role of specific lipid metabolism pathways in pancreatitis. These analyses collectively establish distinct causal mechanisms through which APOC3, APOB, and LPL contribute to pancreatitis risk. First, APOC3 elevates systemic concentrations of TG-rich lipoproteins by inhibiting LPL, the principal enzyme for TG clearance. Second, APOB levels serve as a direct proxy for the number of circulating TG-rich lipoprotein particles; genetically elevated APOB indicates an increased particle count, thereby intensifying the total TG load delivered to the pancreas. Recent meta-analysis and randomized clinical trials of APOC3 mRNA-targeted therapeutics, including volanesorsen [3], plozasiran [4], and olezarsen [5], have confirmed their efficacy in both decreasing serum TG levels and lowering the risk of AP. These findings provide strong translational support for our genetic results. In summary, our findings indicate that higher serum TG and lower serum HDL-C causally increase the risk of pancreatitis. The drug-target analyses suggest that targeting APOB, APOC3, and LPL are promising strategies to reduce the risk of multiple forms of pancreatitis, a finding that warrants further investigation. WH and QX are joint senior authors. LY, XY, and WH designed the study. XY, WH, and QX obtained funding and supervised students. XY and WH drafted the manuscript. LY and XY performed the analyses. RS had important intellectual input. WH and RS critically revised the manuscript. All authors read and approved the final version of the article. The authors thank the participants and investigators for providing publicly available summary statistics. This study was supported by Sichuan Science and Technology Plan Projects—Key Research and Development Program (awarded to WH); National Nature Science Foundation of China (No. 82304985 to YX; No. 82270672 to WH; No. 82274321 to QX). The authors have nothing to report. The authors declare no conflicts of interest. All summary statistics analyzed during this MR study are publicly available and sourced from established genome-wide association study consortia. The primary data for pancreatitis subtypes were sourced from the FinnGen consortium (accessible at https://r10.finngen.fi/), while data for serum lipids, replication analyses, and the coronary artery disease control were retrieved from the Global Lipids Genetics Consortium (https://csg.sph.umich.edu/willer/public/glgc-lipids2021/) and the IEU GWAS database (https://gwas.mrcieu.ac.uk/). Comprehensive datasets, including univariable/multivariable MR results, external validation data, and summary statistics for lipid-modifying targets (MR and colocalization), are deposited in Zenodo at https://doi.org/10.5281/zenodo.17707905. No novel software was developed for this study; all analyses were performed using publicly available R packages. The R scripts used to generate the results are available from the corresponding author upon reasonable request. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.