Background:Inotuzumab ozogamicin (InO) is a humanized anti‐CD22 antibody conjugated to calicheamicin. Upon binding to CD22 on acute lymphoblastic leukemia (ALL) cells, InO is internalized and calicheamicin is released, resulting in DNA double‐strand cleavage and apoptosis. Single agent InO is highly effective in relapsed/refractory (R/R) ALL, having shown clinical activity in a Phase II study. In a Phase III randomized study (INO‐VATE), InO induced a higher response rate and better overall survival than standard of care (SoC).Aims:To assess outcomes by salvage line, for patients (pts) receiving InO vs SoC in the INO‐VATE study.Methods:In this global, open‐label, randomized trial, pts aged ≥18 years with R/R, CD22‐positive, and Philadelphia chromosome (Ph)+ or Ph‐ B‐cell ALL due to receive first (S1) or second salvage (S2) therapy were eligible for enrolment. Pts were randomized 1:1 to receive InO or the investigator's choice of SoC. Pts enrolled were stratified by age (<55 years vs ≥55 years), salvage‐treatment phase (S1 vs S2), and duration of first remission (<12 months vs ≥12 months). A total of 326 pts (InO: 164; SoC: 162) were randomized (intention‐to‐treat population). As per the case report form, pts received either S1 (InO: 111 pts; SoC: 102 pts) or S2 (InO: 51 pts; SoC: 59 pts) treatment. Data as of January 4, 2017 are presented.Results:Baseline characteristics were well balanced between InO and SoC (for both S1 and S2), with some exceptions for S2: compared with InO, a higher proportion of SoC pts were Ph+(17.6% vs 25.4%) and a lower proportion of SoC pts had complex cytogenetics (33.3% vs 16.9%). Responses stratified by S1 vs S2 are presented (Table). Rates of complete remission (CR)/CR with incomplete hematologic recovery (CRi) and rates of MRD negativity (among pts with CR/CRi) were significantly higher for InO compared with SoC, for both S1 and S2 (all P < 0.02). The difference in rates of CR/CRi and MRD negativity between InO and SoC tended to be larger for S1 but these differences were not statistically significant. PFS was significantly longer for InO compared with SoC, for both S1 and S2 (both P < 0.001). OS was longer for InO compared with SoC for S1 and S2; this difference was statistically significant for S1 (P = 0.035) but not S2 (P = 0.353). The proportion of patients proceeding to allogeneic stem cell transplantation (SCT) was significantly higher for InO compared with SoC, for both S1 and S2 (both P < 0.005). In each treatment arm, the proportion of patients proceeding to SCT was higher for S1 than for S2; the differences between S1 and S2 were not statistically significant in either treatment arm.The incidences of several classes of AEs were similar for S1 vs S2, for both InO and SoC; these included grade ≥3 AEs (InO: 91.0% vs 90.2%; SoC: 95.7% vs 98.0%), grade ≥3 thrombocytopenias (InO: 41.4% vs 39.2%; SoC: 60.2% vs 57.1%), and grade ≥3 febrile neutropenias (InO:24.3% vs 29.4%; SoC: 57.0% vs 49.0%). Hepatobiliary disorders (grade ≥3) were less common in S1 vs S2, for both the InO (12.6% vs 25.5%) and SoC (7.5% vs 10.2%) arms. Veno‐occlusive disease (VOD) was less common in S1 than S2 for the InO (9.9% vs 15.7%) arm but not for the SoC arm (2.2% vs 2.0%).Summary/Conclusion:InO is highly effective and provides significant clinical benefit compared with SoC in patients with R/R ALL. Analysis by salvage status indicates that InO provided higher efficacy and better outcomes when utilized as first salvage therapy. Hepatotoxicity was more common in S2 than S1 for both InO and SoC.image