Initial results from a phase 2 study of cladribine, cytarabine, and granulocyte-colony stimulating factor with gemtuzumab ozogamicin (CLAG-GO) for the treatment of patients with relapsed/refractory acute myeloid leukemia

Abstract

Abstract Introduction A significant proportion of patients with acute myeloid leukemia (AML) have either primary induction failure or disease relapse after response to induction chemotherapy. Gemtuzumab ozogamicin (GO), consisting of the antitumor antibiotic calicheamicin conjugated to an anti-CD33 antibody, is currently approved for newly diagnosed and relapsed/refractory (R/R) CD33+ AML. In vitro, GO has an additive effect with several cytotoxic agents, including cytarabine (Tanaka M, Anticancer Res 2009). In addition, granulocyte-colony stimulating factor (G-CSF) priming appears to increase CD33 expression and enhance the cytotoxic effect of GO (Rutella S, Exp Hematol 2006 & Leone G, Haematologica 2004). We report the results of the first stage of a single-arm phase II study evaluating efficacy and safety of cladribine, cytarabine, and G-CSF with GO (CLAG-GO) in patients with R/R acute myeloid leukemia (NCT04050280). Methods This is a single-center, single-arm phase II trial of CLAG-GO in adult patients with AML with either persistent disease or relapsed after intensive chemotherapy. Eligible patients received G-CSF 300 mcg subcutaneously daily on days 0-5, cladribine 5 mg/ m2 and cytarabine 2000 mg/m2 (dose reduced to 1000 mg/m2 for serum creatinine ≥ 1.5 mg/dL) intravenously daily on days 1-5, and GO 3 mg/m2 intravenously over 2 hours on days 1 and 4. GO doses were not capped. CD33-PGx6 score (Chauhan L, JCO Precis Oncol 2019) was performed by genotyping the 6 functional single nucleotide polymorphisms (SNPs) reported in CD33. Patients achieving complete remission (CR) or complete remission with incomplete count recovery (CRi) could receive a single cycle of consolidation therapy using the same CLAG-GO regimen, followed by maintenance with GO monotherapy (2 mg/m2 on day 1 of a 28-day cycle) for up to 8 additional cycles. Per investigator discretion, patients could proceed to allogeneic hematopoietic stem cell transplantation (alloSCT) at any time after response was obtained. Primary objectives were to assess the efficacy of CLAG-GO as a salvage regimen, as well as safety and tolerability. Secondary endpoints included measurable residual disease (MRD) status at the end of induction, consolidation and maintenance, duration of response, and overall survival. Exploratory objectives included rate and safety of alloSCT and correlation of clinical response with karyotype, baseline CD33 expression, and CD33-PGx6 score. The study followed Simon's two-stage ‘minimax’ design, with 19 patients planned in the first stage. If there were at least 7 responders, an additional 20 patients would be enrolled in the 2nd stage. Results Among the first 19 enrolled patients, median age was 56 years (range 20-66), and 12 (63%) were male. Median number of prior therapies was 1 (range 1 to 3). Median blasts at baseline were 54% (range 8-96%), and median baseline CD33 expression was 99% (range 45-100%). The composite CR rate at the end of induction was 42% (n=8), of which 36% achieved CR and 5% CRi, all of whom had undetectable MRD. Seven patients (37%) received consolidation, and 6 patients (32%) proceeded to alloSCT; median time to alloSCT (from day 0) was 144 days (range 117-185). With a median follow up of 51 months, median OS (not censored for alloSCT) was 9.53 months (95% CI: 2.78-16.23) and 6 patients (31.6%) remained alive. Median duration of response was not reached. As expected, the most common grade 3 and 4 adverse events were hematologic: febrile neutropenia (95%), thrombocytopenia (79%), and anemia (47%). No patients developed veno-occlusive disease either during study treatment or after alloSCT, and 30-day mortality was 0. CD33-PGx6 analysis showed that patients with low score (implying lower CD33 levels) did not respond and were MRD positive (n=4), as compared to 50% of those (n=7) within the high score group (n=14). Conclusions In the initial stage of this phase II study, CLAG-GO had a comparable response rate to other intensive salvage regimens for R/R AML. The addition of GO did not seem to increase toxicity. Responses were deep and durable, and allowed patients to proceed to alloSCT. CD33-PGx evaluation suggested association with poor response in patients with low score, though sample size was limited. The pre-specified number of responses was met, and enrollment in the study continues, with an additional 20 patients planned.