POS0877 THE EFFECT OF PLATELET INHIBITORS ON DIGITAL ULCERS IN SYSTEMIC SCLEROSIS - A DERIVATION AND VALIDATION EUSTAR STUDY

Abstract

Background:Digital ulcers (DUs) affect half of the patients with systemic sclerosis (SSc) and can be complicated by gangrene and amputation. The direct involvement of platelets in the development of DUs has been suggested by in vitro studies, which encouraged physicians to consider platelet inhibitors as a therapeutic option in the management of DUs. However, until now, there is no clinical study to assess the efficacy of platelet inhibitors for DUs in SSc patients.Objectives:To demonstrate a possible relationship between treatment with platelet inhibitors and the occurrence of DUs at the next follow-up visit in patients with SSc.Methods:This study used prospectively collected data from the European Scleroderma Trials and Research group (EUSTAR) registry. Patients fulfilling the 2013 ACR/EULAR SSc classification criteria with complete longitudinal data on the presence of DUs and platelet inhibitors were included in the analysis. Multiple imputation using a random forest algorithm was implemented to handle missing values.The dataset was split into a derivation and validation cohort. To investigate the response for the binary dependent variable of DUs, a generalized linear mixed model (GLMM) was developed in the derivation cohort and validated using ROC analysis and Brier scores to address discrimination and calibration, respectively.Results:Of 3,463 patients (2,961 in the derivation cohort, 722 in the validation cohort), 453 had current DUs at the baseline and 245 were exposed to platelet inhibitors (table 1).Our GLMM revealed that the exposure to platelet inhibitors is associated with a reduced risk of DUs at the next follow up visit (OR = 0.33, 95% CI = [0.13 to 0.82]). Further factors associated with absence or presence of DUs at the next follow-up visit are shown in figure 1. This confirmed the previously identified risk factors for the presence of DUs, supporting the overall robustness and the validity of our model.The performance was evaluated by ROC curve analysis and showed an AUC = 97.97% (95% CI = [96.93% to 97.67%]) for the derivation cohort and AUC = 77.3% (95% CI = [74.01% to 81.39%]) for the validation cohort, respectively, showing an acceptable discrimination. The Brier score was 0.05 in the derivation cohort and 0.07 in the validation cohort, suggesting a good calibration of the model.Conclusion:Our model, with acceptable discrimination and good calibration, suggests a positive treatment effect of platelet inhibitors on DUs in clinical practice.Table 1.Baseline characteristics of patients before imputationCharacteristicsOverallDerivation setValidation setn3,4632,691772Age (median [IQR])56.00 [47.00, 66.00]56.00 [47.00, 65.00] 57.00 [48.00, 67.00]Disease duration (median [IQR]) 9.00 [4.00, 16.00] 9.00 [4.00, 16.00] 8.00 [4.00, 15.00]Disease subset = Limited cutaneous SSc (%) 1562 (65.2) 1164 (64.6) 398 (66.9)DUs (%): Current 453 (13.1) 378 (14.0) 75 (9.7)DUs (%): Never 1783 (51.5) 1326 (49.3) 457 (59.2)DUs (%): Previously 1227 (35.4) 987 (36.7) 240 (31.1)mRSS (median [IQR]) 5.00 [2.00, 11.00] 6.00 [2.00, 12.00] 4.00 [1.00, 11.00]Joint Contractures = Yes (%) 881 (26.8) 770 (29.4) 111 (16.5)LVEF (median [IQR])62.00 [60.00, 65.00]60.00 [60.00, 65.00] 65.00 [60.00, 67.00]Dyspnea NYHA III and IV (%)300 (9.5)214 (8.6)86 (12.7)Pulmonary hypertension = Yes (%) 244 (10.7) 200 (11.3) 44 (8.4)Lung fibrosis on HRCT = Yes (%) 685 (46.6) 600 (47.7) 85 (39.7)FVC % predicted (median [IQR])97.00 [82.00, 111.00]95.00 [81.00, 110.00]101.00 [85.00, 115.00]Serum creatinine mg/dl (median [IQR]) 0.70 [0.60, 0.90] 0.70 [0.60, 0.90] 0.70 [0.70, 0.90]Anti-Scl-70 positive = Yes (%) 1147 (33.1) 958 (35.6) 189 (24.5)CRP elevation = Yes (%) 639 (21.1) 490 (20.8) 149 (22.1)Platelet inhibitors therapy = Yes (%) 245 (7.1) 206 (7.7) 39 (5.1)Oral anti-coagulants therapy = Yes (%) 53 (1.5) 50 (1.9) 3 (0.4)Disclosure of Interests:None declared