Safety, Efficacy, and Biomarkers of Response to Azacitidine (AZA) with Nivolumab (Nivo) and AZA with Nivo and Ipilimumab (Ipi) in Relapsed/Refractory Acute Myeloid Leukemia: A Non-Randomized, Phase 2 Study

Background: Preclinically blocking PD-1/PD-L1 pathways enhanced anti-leukemic responses. PD-1 positive CD8 T-cells were increased in the bone marrow (BM) of pts with AML (Daver et al, ASH 2016). Azacitidine up-regulates PD-1 and interferon-gamma signaling (Yang et al., Leukemia 2013).

Methods: Pts were eligible for the AZA+Nivo (cohort 1) if they had relapsed/refractory AML (R/R AML), ECOG ≤ 2, and adequate organ function. The recommended phase 2 dose was established as AZA 75mg/m² Days 1-7 with Nivo 3mg/kg Day 1 and 14. Courses were repeated approximately every 4-5 weeks indefinitely. 70 R/R AML pts have been treated; cohort 1 is closed.

A subsequent cohort of AZA+Nivo+Ipi in R/R AML was opened (cohort 2), with the same eligibility criteria as cohort 1 except that enrollment was restricted to pts with salvage 1 and 2 AML. Ipi 1mg/kg Q6 weeks was added to the established AZA+Nivo schedule. This dosing was based on lung and melanoma dosing for Ipi + Nivo.

Responses in both cohorts included best response within 3 months of therapy initiation.

Results:

Cohort 1: 70 R/R AML pts with median (med) age 70 years (range, 22-90), secondary AML (44%), poor risk cytogenetics (34%), med salvage 2 (range, 1-7) were enrolled. The ORR was 33% including 15 CR/CRi (22%) (4 CR, 11 CRi), 1 PR, and 7 HI maintained on study >6 months (10%) without allogeneic stem cell transplant. Additionally, 6 pts (9%) remained on study with stable disease (SD) >6 months. The remaining 41 pts (58%) had no response. 8-week (wk) mortality was 10%. Patients who achieved a CR/CRi/PR/HI/SD (n=29; 42%) had significantly improved overall survival (OS) compared with NRs (n=41; 58%) (16.1 versus 4.1 months; p<0·001). By univariate analysis (UVA) improved ORR was seen in pts with pretherapy BM blast <20%, circulating WBC </=10,000/mL, ASXL1 mutation, and no prior HMA-based therapy (Table 1). By UVA improved OS was seen in pts who achieved any response or SD (CR/CRi/PR/HI/SD), were salvage 1 status, and had ASXL1 mutation. The med OS with Aza+Nivo compared favorably to historical med OS with AZA-based protocols in salvage 1 or 2 pts at MDACC (6.0 versus 4.7 months; p=0.03)(Fig 1), with the most prominent improvement seen in salvage 1 pts (11.1 versus 4.1 months; P<0.001). Grade 3/4 immune related adverse events (irAEs) were observed in 8 (11%) pts, including pneumonitis (n=3), colitis (n=2), and nephritis, skin rash, and hypophysitis (1 each). The majority (12 of 16; 75%) of these occurred in the first 8 wks after nivo initiation. These responded rapidly to steroids when initiated early.

Flow-cytometry (MFC) was performed on pre- and post-therapy BMAs, after 2 doses (end of cycle 1 or EOC1) and 4 doses (EOC2) of nivol in 19 of 23 responders (R) (CR/CRi/PR/HI) (83%) and 28 of 41 NRs (68%). 36 parameter CyTOF was performed at same time-points in 5R and 5NRs. On pretherapy BMAs, R had a higher frequency of CD3⁺ (p=0.04) and CD8⁺ cells (p=0.09) on MFC (Fig 2A). CD3⁺ in the pretherapy BMA with a cut-off of 13.2%, had a sensitivity of 74% and a specificity of 65% for predicting response. The CD3+ was >13.2% in 26 of 47 pts (55%) who had an evaluable pretherapy BMA. Immunohistochemistry confirmed that CD3+ cells were increased in pretherapy BMs in R. PhenoGraph clustering of all CD3-gated cells revealed 24 meta-clusters (Fig 2B). The frequency of an effector CD8⁺ T cell cluster (C2) expressing CD45RA⁺PD1^loTbet^hiEomes^lo was significantly higher in the pretherapy BMAs of R versus NRs (11.2% versus 2.5%; p=0.002), and was further expanded in R not in NRs in the EOC1, EOC2, and EOC4 BMAs (Fig 2C).

Cohort 2: 20 pts with R/R AML, med salvage 2 (range 1-2), med age 68 (range, 25-83), secondary AML (36%), adverse cytogenetics (25%), and common molecular mutations TP53 in 5, RAS in 4, DNMT3A/ASXL1 in 2 pts each, were enrolled. 14 pts enrolled before May 1, 2018 were evaluable for response: 6 CR/CRp/CRi (43%) (including 2 CR, 2CRp, 2CRi), 2 stable disease > 6 mths, and 6 no response. The 8-wk mortality was 0. Grade 3/4 irAEs noted in 26% of the pts, most common pneumonitis. Med cycles to response of 2 (1-3). The med OS for all pts was not reached and the projected 1-yr OS in R/R AML was encouraging at 58% (Fig 1).

Conclusion: AZA+Nivo was effective in pts with AML in Salvage 1 or those with increased pretherapy CD3⁺ BM infiltrate by MFC or IHC. BM CD3 and CD8 are simple assays to select pts for future trials of AZA+PD1 inhibitors in AML. AZA+Nivo+Ipi had a encouraging CR/CRp/CRi rate (43%) and med OS (NR) in pts with salvage 1/2 AML and is enrolling.

View largeDownload slide

View largeDownload slide

Close modal