A novel aGAPSS-based nomogram for the prediction of ischemic stroke in patients with antiphospholipid syndrome

BACKGROUND: Ischemic stroke (IS) is the most common and life-threatening arterial manifestation of antiphospholipid syndrome (APS). It is related to high mortality and severe permanent disability in survivors. Thus, it is essential to identify patients with APS at high risk of IS and adopt individual-level preventive measures. This study was conducted to identify risk factors for IS in patients with APS and to develop a nomogram specifically for IS prediction in these patients by combining the adjusted Global Anti-Phospholipid Syndrome Score (aGAPSS) with additional clinical and laboratory data. METHODS: A total of 478 consecutive patients with APS were enrolled retrospectively. All patients were randomly assigned to the training and validation cohorts. Univariate and multivariate binary logistic analyses were conducted to identify predictors of IS in the training cohort. Then, a nomogram was developed based on these predictors. The predictive performance of the nomogram for the training and validation cohorts was evaluated by determining areas under the receiver operating characteristic curve (AUROC) and creating calibration plots. A decision curve analysis (DCA) was conducted to compare the potential net benefits of the nomogram with those of the aGAPSS. RESULTS: During a mean follow-up period of 2.7 years, 26.9% (129/478) of the patients were diagnosed with IS. Binary logistic regression analysis revealed that five risk factors were independent clinical predictors of IS: age (P < 0.001), diabetes (P = 0.030), hyperuricemia (P < 0.001), the platelet count (P = 0.001), and the aGAPSS (P = 0.001). These predictors were incorporated into the nomogram, named the aGAPSS-IS. The nomogram showed satisfactory performance in the training [AUROC = 0.853 (95% CI, 0.802–0.896] and validation [AUROC = 0.793 (95% CI, 0.737–0.843)] cohorts. Calibration curves showed good concordance between observed and nomogram-predicted probability in the training and validation cohorts. The DCA confirmed that the aGAPSS-IS provided more net benefits than the aGAPSS in both cohorts. CONCLUSION: Age, diabetes, hyperuricemia, the platelet count, and the aGAPSS were risk factors for IS in patients with APS. The aGAPSS-IS may be a good tool for IS risk stratification for patients with APS based on routinely available data.