OC 36.5 - Predictive factors of venous thromboembolism events in multiple myeloma: a French cohort from 2012 to 2021.

Background: Incidence rate of venous thromboembolisms (VTE) in patients with multiple myeloma (MM) remains high despite thromboprophylaxis guidelines based on current risk stratification.

Aims: To identify risk factors of VTE in a cohort of patients diagnosed with MM, to derive a risk prediction model and to compare predictive values of this model with currently recommended risk scores (IMPEDE VTE and SAVED).

Methods: All consecutive patients diagnosed with MM at the Brest University Hospital between 2012 and 2021 were included. Clinical, biological and radiological data at MM diagnosis were collected and VTE events were adjudicated by independent committee. Predictors of VTE were identified in uni- and multivariable analysis using a Cox model and Hazard Ratio (HR) and 95% confidence interval (95%CI) were calculated. ROC curves and Area Under the Curves (AUC) were computed and compared.

Results: Among 259 patients with MM, 39 (15.1%) presented a VTE (11 pulmonary embolism, 19 proximal and 11 distal deep vein thrombosis). VTE occurred most often within 3 months following MM diagnosis (41.3%). In multivariable analysis, predictive factors of VTE were history of VTE (HR 3.9 (95%CI 1.3-13.6), p=0.03), ECOG scale performance status ≥2 (OR: 5.0 (2.2-11.4), p< 0.001) and the absence of thromboprophylaxis (antiplatelet or prophylactic heparin) (HR 7.7, 95%CI 2.5-23.0), p< 0.001). Our prediction model had an AUC of 77.0 (95%CI 67.9-86.6, p< 0.001)) as compared to 67.6 (95%CI 55.6-79.7, p=0.005) for IMPEDE-VTE and 52.0 (95%CI 38.5-65.5) for SAVED score. Our model discriminate between low risk of VTE (score=0, 12/183 (6.2%) VTE) and high risk of VTE (score≥1, 37/99 (37.4%) VTE).

Conclusion(s): We developed a new risk prediction model for VTE in patients with MM, whose performance seems better than currently used scores. External validation and clinical confirmation of this model are warranted.