7042 Background: PEV+AZA has been studied in higher-risk MDS/CMML and AML, with encouraging efficacy and an acceptable safety profile without added myelosuppression. This pooled analysis was performed to evaluate the impact of PEV exposure on safety and efficacy. Methods: Data from three studies (NCT01814826, NCT02782468 and NCT02610777) were used in the PEV exposure–safety analyses, including ≥ grade 3 neutropenia (NEU3), febrile neutropenia (FN), ≥ grade 3 thrombocytopenia, ≥ grade 3 alanine aminotransferase elevation, ≥ grade 3 aspartate aminotransferase elevation and ≥ grade 3 treatment-emergent adverse event (TEAE3), in pts with higher-risk MDS/CMML and AML who received PEV+AZA. Data from NCT02610777 were used for exposure–efficacy analyses, including overall survival (OS), event-free survival (EFS), complete response (CR) and CR+partial response (PR), in pts with higher-risk MDS/CMML who received PEV+AZA. The exposure metrics for individual pts were derived from a previously developed population pharmacokinetic model with pooled data from eight phase 1/2 studies. PEV exposure–safety relationships for the toxicity endpoints, exposure–CR and exposure–CR+PR, were estimated by logistic regression. Age, sex, race, baseline Eastern Cooperate Oncology Group (ECOG) Performance Status score and disease type were evaluated as covariates. Cox proportional-hazards models were used to evaluate the PEV exposure–survival for higher-risk MDS/CMML, with age, sex, baseline ECOG PS score, Revised International Prognostic Scoring System score (IPSS-R) and disease type as potential covariates. Results: In total, 135 pts (median age, 74 years; male, 64%; Caucasian, 82%) and 41 pts (median age, 74 years; male, 76%; Caucasian, 90%; median IPSS-R, 5.5) were included in PEV exposure–safety and exposure–efficacy analyses, respectively. PEV exposure was significantly related to the incidence of NEU3 ( p = 0.003), FN ( p = 0.02) and TEAE3 ( p = 0.02), supporting PEV dose reductions for pts with treatment-related toxicities. Relationships between PEV exposures and CR, CR+PR, EFS or OS indicated consistent clinical benefit across ranges of PEV exposure following a starting dose of 20 mg/m2. Conclusions: The association between exposure and safety supports PEV dose reductions for pts with treatment-related toxicities. The exposure–efficacy analyses indicated consistent clinical benefit across ranges of PEV exposure following a starting dose of 20 mg/m2. These results support a favorable benefit–risk profile of the 20 mg/m2 PEV dose on days 1, 3 and 5 in combination with AZA 75 mg/m2 for 7 days in 28-day cycles. Clinical trial information: NCT01814826 , NCT02782468 , NCT02610777.