Phase 2 Study of Combination of Cytarabine, Idarubicin, and Nivolumab for Initial Therapy of Patients with Newly Diagnosed Acute Myeloid Leukemia

Background: Blocking PD-1/PD-L1 pathways enhances anti-leukemia responses by enabling T-cells in murine models of AML (Zhang et al, Blood 2009). PD-1 positive CD8 T-cells are increased in bone marrow (BM) of pts with AML (Daver et al, AACR 2016). PD1 inhibition has shown activity in AML (Berger et al, Clin Cancer Res 2008). We hypothesized that addition of nivolumab to an induction regimen of ara-C and idarubicin may prolong relapse-free survival (RFS) and overall survival (OS); this study was designed to determine the feasibility of this combination.

Methods: Pts with newly diagnosed acute myeloid leukemia (by WHO criteria; ≥20% blasts) and high risk MDS (≥10% blasts) were eligible to participate if they were 18-65 yrs of age and had adequate performance status (ECOG ≤3) and organ function (LVEF ≥ 50%; creatinine ≤ 1.5 g mg/dL, bilirubin ≤ 1.5 mg/dL and transaminases ≤ 2.5 times upper limit of normal). Treatment included 1 or 2 induction cycles of ara-C 1.5 g/m² over 24 hours (days 1-4) and Idarubicin 12 mg/m² (days 1-3). Nivolumab 3 mg/kg was started on day 24 ± 2 days and was continued every 2 weeks for up to a year. For pts achieving complete response (CR) or CR with incomplete count recovery (CRi) up to 5 consolidation cycles of attenuated dose ara-C and idarubicin was administered at approximately monthly intervals. Eligible pts received an allogeneic stem cell transplant (alloSCT) at any time during the consolidation or thereafter.

Results: 3 pts with relapsed AML were treated at a run-in phase with a dose of nivolumab 1 mg/kg without specific drug-related toxicity. Subsequently, 32 pts (median age 53 yrs; range, 26-65) were treated as above including 30 with AML (24 de novo AML, 2 therapy-related AML, 3 secondary AML and 1 therapy-related secondary AML) and 2 high risk MDS. Pre-treatment genetic risk by ELN criteria was 11 adverse, 16 intermediate, and 5 favorable, including 2 FLT3 -ITD mutated, 5 NPM1 mutated, and 7 TP53 mutated. All 32 pts were evaluable for response and 23 (72%) achieved CR/CRi (19 CR, 4 CRi). The 4-week and 8 week mortality was 6% and 6%. The median number of doses of nivolumab received was 6 (range, 0-13); one pt did not receive nivolumab due to insurance issues. 9 pts underwent an alloSCT. After a median follow-up of 8.3 mths (range, 1.5-17.0) the median RFS among the responding pts has not been reached (range, 0.1 - 15.8 mths) and the median OS has not been reached (range 0.5-17.0 mths).

Grade 3/4 immune mediated toxicities have been observed in 5 pts and include rash, pancreatitis, and colitis. Other grade 3/4 toxicities thought to be potentially related to nivolumab include cholecystitis in one pt. 9 pts proceeded to an alloSCT. Donor source was matched related in 2, matched unrelated in 6 and haplo-identical in 1 pt. Conditioning regimen was Fludarabine plus busulfan-based in 8, and fludarabine plus melphalan in 1 pt. 4 pts developed graft versus host disease (GVHD)(grade I/II in 3, grade III/IV in 1), which responded to treatment in 3.

Multicolor flow-cytometry studies are conducted by the Immunotherapy Platform on baseline (prior to first dose of nivolumab) and on-treatment BM aspirate and peripheral blood to assess the T-cell repertoire and expression of co-stimulatory receptors and ligands on T-cell subsets and leukemic blasts, respectively. The baseline BM was evaluated on 23 of the 32 evaluable pts, including 18 responders and 5 non-responders. Pts who achieved a CR/CRi had a trend of higher frequency of live CD3⁺ total T cell infiltrate as compared to non-responders in the baseline BM aspirates (Fig 1A). We evaluated expression of immune markers on T cell subsets: CD4 T effector cells [Teff]: CD3⁺CD4⁺CD127l^o/+Foxp3^-, CD4 T regulatory cells [Treg]: CD3⁺CD4⁺CD127^-Foxp3⁺, and CD8 T cells. At baseline, BM of non-responders had significantly higher percentage of CD4 T effector cells co-expressing the inhibitory markers PD1 and TIM3 (p<0.05) and a trend towards higher percentage of CD4 T effector cells co-expressing PD1 and LAG3 compared to responders (Fig 1B). Co-expression of TIM3 or LAG3 on PD1⁺ T cells have been shown to be associated with an exhausted immune phenotype in AML (Zhou et al., Blood 2011).

Conclusion: Addition of nivolumab to ara-C and anthracycline induction chemotherapy is feasible and safe in younger pts with AML. Among the pts proceeding to alloSCT the risk of GVHD is not significantly increased.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal