P788: FACTORS DRIVING TREATMENT DECISION IN PATIENTS WITH INTERMEDIATE-RISK MYELODYSPLASTIC SYNDROME (MDS): A RETROSPECTIVE ANALYSIS FROM THE GRUPO ESPAÑOL DE SÍNDROMES MIELODISPLÁSICOS SPANISH MDS REGISTRY

Abstract

Background: Current treatment (tx) options for patients (pts) with MDS are limited, although multiple potential txs are in development. The use of the Revised International Prognostic Scoring System (IPSS-R) for diagnosis of MDS supports tx decisions and prognostic assessment. Pts with very low/low-risk (vL/LR) MDS usually receive supportive care (SC) to manage symptoms, whereas those with very high/high-risk (vHR/HR) MDS commonly receive active txs such as hypomethylating agents (HMAs), allogeneic hematopoietic stem cell transplantation (alloSCT), or chemotherapy (CT). There is limited agreement on whether active tx or standard SC should be used for pts with intermediate-risk (IR) MDS. Aims: We aimed to identify factors possibly associated with the real-world use of tx in pts with IR-MDS in the Spanish MDS registry. Methods: This was a retrospective, observational study of data from pts registered in the Grupo Español de Síndromes Mielodisplásicos (GESMD) from January 2008-June 2020. Adult pts with MDS (WHO 2008 classification) who signed an informed consent and had a minimum follow-up of 6 months in the registry were included. Pts with unavailable data on IPSS-R risk or HMA tx were excluded. We collected demographic, clinical, and laboratory data, and tx used, including azacitidine (AZA), CT, alloSCT, SC (eg, transfusions, growth factors) and other tx (eg, immunosuppressors, androgens). Univariate analysis was performed to evaluate the association of potential risk factors, using categorical and continuous variables, with the tx use. We focused our analysis on AZA and SC in pts with IR-MDS ineligible for alloSCT. Odds ratio, 95% confidence interval, and exploratory P value were reported for comparison between tx use. Results: Overall, 4604 pts (median age: 76 y; male: 61%; IPSS-R score vHR: 432, 9.4%; HR: 487, 10.6%; IR: 761, 16.5%; LR: 1795, 39%; vLR: 1129, 24.5%) were included. Pts with IR-MDS were 63% male and similar in number by score and median age at diagnosis compared with the total study population: 3.5 (n=268, 73 y), 4.0 (n=270, 75 y) and 4.5 (n=223, 75 y). Most common treatments received were AZA (n= 241, 31.7%) and SC (n=250, 32.9%); a minority received CT and/or alloSCT (n=84, 11%), while 186 pts (24.4%) did not receive any. AZA was the only HMA used in this population. The median time from diagnosis to start of AZA in this group was 3 months. When comparing AZA with SC use in pts with IR-MDS, univariate analysis (Table 1) showed that gender, age at diagnosis, year of diagnosis, bone marrow and peripheral blood blasts, transfusion dependency, and hematologic parameters at diagnosis (eg, hemoglobin, platelet, and monocyte counts) were associated with tx selection. A multivariate analysis is ongoing for deeper understanding of these data. Image:Summary/Conclusion: These preliminary data show several factors significantly associated with tx use, suggesting that in some pts with IR-MDS, AZA is more likely to be used versus SC when these factors are considered. It is possible that selection bias, misclassification and confounding may have occurred as the registry only collects routine data according to local practices, with variability in the quality and completeness of data across participating centers. Other factors not captured, such as pt characteristics, comorbidities, access to health care, and social/familial support, could have also influenced the tx decision. Further analyses are ongoing to identify the relationship of multiple factors at once and assert the strength of these associations while removing the effect of confounders.