Evaluating Mean Corpuscular Volume As Predictor for Erythropoiesis Stimulating Agent Response in Elderly Patients Diagnosed with Myelodysplasia

Abstract

Background: Low-Risk Myelodysplastic Syndrome (LR-MDS) is genetically heterogeneous. Transfusion dependency (TD) adversely impacts overall survival (OS). Erythropoiesis-stimulating agents (ESAs) are first line for anemia and delay onset of TD, improving quality of life. In older LR-MDS patients harboring a larger number of mutations, ESA response is dramatically reduced. Predictors of ESA response include hemoglobin between 8-10 g/dL, del (5q) anomaly, erythropoietin (EPO) level <100 IU/L, low blast count, and transfusion independence (TI). Macrocytosis is not only observed at MDS diagnosis, but it is also associated with inflammasome activation upon mutation acquisition. Prior studies in mice, for example, have shown that TET2 haploinsufficiency leads to larger erythroid size and pyroptosis. Mean Corpuscular Volume (MCV) is the most practical way to detect red cell size modification in MDS patients. There are no prior studies that have investigated the correlation between diagnostic red cell size (MCV) and prediction of response to MDS directed therapy (i.e. ESA, hypomethylating agents). Given the known association between myeloid mutations and red cell size modifications, we investigated the correlation between MCV and ESA response in LR-MDS patients older than 65 years. Methods: After IRB approval, 81 LR-MDS patients diagnosed at Michael E. DeBakey Houston VA were screened. 68/81 (83.9%) patients had data to evaluate ESA response. ESA response was classified by International Working Group (IWG) MDS 2016 criteria. Statistical analysis was performed with SAS software. Relevant variables with ability to predict ESA response including age, hemoglobin, lactate dehydrogenase (LDH), blast percentage, ferritin, EPO level, and TD history were investigated. Logistic regression analysis was used to examine the independent MCV modification effect on ESA response when adjusting for potential confounders selected from our univariate analysis. Results: 26/68 (38.2%) and 42/68 (61.7%) patients were ESA responders and non-responders, respectively. Median MCV among LR-MDS patients exhibiting ESA response was 98.6 fL (79.4-116.3) vs 93.9 fL (74.6-112.8) in non-responders, p=0.04. Median (range) for age, hemoglobin, LDH, blast percentage, ferritin, and EPO among responders and non-responders were 75.7 years (66-91) vs 76 years (66-88), p=0.80; 9.4 g/dL (8-11.1) vs 8.9 g/dL (5.3-13), p=0.24; 159.2 U/L (68-278) vs 203.1 U/L (119-372), p=0.04; 0.89% (0-5) vs 0.92% (0-10), p=0.96; 249 ng/mL (8.2-649) vs 395 ng/mL (9.3-945), p=0.02; and 195 mIU/mL (7.7-925) vs 174 mIU/mL (19-1626), p=0.80. TD history was observed in 50% and 66% of responders and non-responders, respectively, p=0.34. After accounting for potential confounders, MCV remained the only predictor for ESA response among LR-MDS patients, p=0.04, HR=1.12 (95% CI 1.01-1.25). Conclusions: Superior ESA response was observed among LR-MDS patients older than 65 years who exhibited higher MCV at diagnosis. Our data suggests that ESA response is associated with modifications in erythroid cell morphology (i.e. larger red cell size). In LR-MDS patients exhibiting higher MCV, probability of ESA response increases by 12% per 1-unit MCV increase at diagnosis. Although investigation into a "mutation type and/or mutation combination" leading to MCV modification is needed, MCV could discriminate subgroup of elderly LR-MDS patients who exhibit differential ESA response. Disclosures Yellapragada: Novartis: Employment, Other: Spouse Employment ; Celgene: Research Funding; BMS: Research Funding; Takeda: Research Funding.