P25: CARFILZOMIB-BASED TREATMENTS COULD BE ADDED TO IMPEDE-VTE SCORE TO BETTER PREDICT VENOUS THROMBOEMBOLIC RISK IN MM PATIENTS

Abstract

Introduction: Venous thromboembolic (VTE) risk in Multiple Myeloma (MM) patients treated with novel agents can be very high, particularly with lenalidomide combined with high dose steroid. Guidelines do not recommend a validated Risk Assessment Model (RAM) to precisely identify patients with high VTE risk, in order to personalize the thromboprophylactic strategy. The IMWG score, that is widely used, is not actually performing well, as VTE events continue to occur in about 15% of MM patients. We aim to test IMPEDE-VTE score, which seems to be the most effective RAM in MM, in our real life MM population, trying to eventually implement it with emergent new risk factors. Patients and methods: In our database, we prospectively collected data from 393 MM patients, treated from 2010 to 2021. One-hundred-thirteen (29%) patients were >75 years old, 45% were frail according to simplified Frailty Score, 17% had abnormal LDH level, 19% had renal failure and 4% had a BMI ≥35. ISS stage 3 was scored in 27% of patients and 31% of patients had high-risk cytogenetics. Sixty-six percent of patients received Immunomodulatory (IMIDs)-based therapy (lenalidomide in 32% of patients). Monoclonal antibodies (MoAbs) were used in 8% of patients (MoAbs associated to lenalidomide in 75%) and carfilzomib in 10% of patients (carfilzomib associated to lenalidomide in 51%). Patients were stratified per IMPEDE-VTE score: high, intermediate and low risk were formed by 10%, 41% and 49% of patients, respectively. Eleven percent of patients had a central venous catheter (CVC), and thromboprophylaxis was chosen according to the IMWG criteria (ASA or none 74%, anticoagulants 26%). We have finally analysed the impact of therapeutic risk factors on thromboembolic events’ incidence. Results: VTE events’ rate in our MM population was 16%, the majority of them occurring during the first three months from the starting of therapy. None of patient-related risk factors (age, CCI, Frailty Score, LDH value), nor disease-related risk factors (ISS, cytogenetics, renal failure), except for bone disease, was significant in univariate analysis for increasing VTE risk. On the contrary, CVC, IMPEDE-VTE score, lenalidomide and carfilzomib treatments retained statistical significance. Analysing therapy-related risk factors for VTE risk, we have found only a trend between the group treated with IMIDs-based (thalidomide, lenalidomide, pomalidomide) and that with non-IMIDs-based therapy (p=0.095). Nevertheless, patients treated with lenalidomide had higher VTE risk compared with patients treated without lenalidomide (13% vs 7% at 6 months; p<0.001). MoAbs-based therapy concurred to increase VTE risk, but only when associated to lenalidomide (p<0.001). On the contrary, carfilzomib monotherapy significantly increased VTE risk (p<0.001), regardless the association with lenalidomide. Stepwise Cox regression analysis selected IMPEDE-VTE score (LR-IR vs HR: 7% vs 16% at 6 months; p=0.032) and carfilzomib treatment (24% vs 7% at 6 months; p< 0.001) as independent significant risk factors for VTE. Conclusions: Despite our analysis being retrospective, we confirmed IMPEDE-VTE as a predictive score of VTE risk in a real life MM population. However, our analysis found carfilzomib-based therapy an independent risk factor from IMPEDE-VTE score. We attempt to prospectively study the “carfilzomib-based therapy” parameter in order to implement IMPEDE-VTE score, aiming to better stratify VTE risk in MM patients.