Risk Factors for Thromboembolic Events and Bleeding in Patients With Stomach Cancer: A Cohort Study

BACKGROUND:Stomach cancer is associated with systemic disturbances, including impaired hemostasis, which may resultinthrombosis and bleeding, significantly affectingtheprognosis, qualityoflife, and riskoffatal outcome. AIM:The study aimedtoidentify risk factorsforthrombosis and bleedinginpatients with stomach cancer, assesstheprognostic valueofexisting risk scores usingaretrospective analysisofour own findings, and propose waystoimprove them. METHODS:The study analyzed medical recordsof178 patients with stomach cancer who were treatedatSechenov Universityin2021–2023. The medical records were grouped basedontwo independent variables:thepresenceofthromboembolic events (25 patients with and 150 patients without thromboembolic events) andthepresenceofbleeding (23 patients with and 155 patients without bleeding). The outcomes were assessed by examining imaging findings (ultrasound and computed tomography) specifiedinmedical records, which allowed confirming or ruling out thromboembolic events,aswellasinitial examination findings and discharge summaries following hospitalizationforchemotherapy. Qualitative variables are presentedasmean±standard deviation, whereas categorical variables are presentedasabsolute values and percentages. Intergroup differences were assessed usingtheMann–Whitney and Pearson’s chi-squared tests. Significant associationsforthromboembolic events and bleeding were identified using multivariate regression analysis. The prognostic valueofscores was assessed using ROC analysis (AUC, sensitivity, specificity). SPSS 23.0 was usedforstatistical analysis. RESULTS:Patients with chronic heart failure had an 11-fold greater riskofthromboembolic events than patients without chronic heart failure (odds ratio [OR] 11.12; 95% confidence interval [CI] 3.14–38.74;p=0.001). Chronic kidney disease was associated with more thana5-fold greater risk (OR 5.07; 95%CI 2.02–12.71;p=0.002), and varicose veins with more than an 8-fold greater risk (OR 8.12; 95%CI 2.87–22.94;p=0.001). Adding these factorsto theKhoranascore improved its prognostic value:inROC analysis, AUC was 0.823 comparedto0.615atbaseline. This indicatesahigh discriminatory capabilityof themodified model (AUC0.8). The REACH score showedahigh prognostic valueforassessingtheriskofbleeding (AUC=0.738). CONCLUSION:Adding chronic heart failure, chronic kidney disease, and varicose veinsto theKhorana score improves its prognostic valueforassessingtheriskofthromboembolic events. The REACH score was effectiveinpredictingtheriskofbleeding.