IL-1RAP Is a Targetable Surface Antigen in Relapsed/Refractoy AML

Context: Relapsed/Refractory (R/R) AML is an unmet medical need. Novel AML immunotherapies require surface antigens that have broad expression, both at the inter-individual level and intra-individual (blasts and leukemic stem cells [LSCs] levels, with minimal expression on healthy HSPCs. IL-1RAP, the accessory co-receptor of the interleukin 1, 32, and 33 receptors, is a promising target antigen with broad expression on AML cells at diagnosis (Boissel et al. EHA 2024 #P453, De Dreuzy et al. AACR 2024 #6322) and has been suggested as a potential Leukemic Stem Cell (LSC) marker (Askmyr et al. Blood 2013). Its expression in R/R AMLs eligible for early-stage trials is unknown. Objective: To determine IL-1RAP expression on blasts and LSCs in a prospective cohort of R/R AMLs. Methods: Peripheral Blood (PB) or Bone marrow (BM) from relapsed/refractory (excluding molecular relapse) AML patients ≥18 years old accrued to the ALFA PPP registry (NCT04777916) were collected and prospectively assessed for IL-1RAP expression (clone B-R58) in a 10-marker panel on an Ique3 high-throughput cytometer (Sartorius). Flow analyses were done on Kaluza (Beckman). Results: From 26 July 2023 till 11 June 2024, we analyzed 38 patients (M/F 19/19, median age 67y), with primary refractory (26%) or relapsed (74%) AML, including 25% therapy-related and 17% secondary and 58% de novo AML. NPM1, FLT3 and IDH1/2 mutations were present in 25%, 29% and 17% of patients, respectively. Most (79%) had been previously exposed to intensive Cx, 58% to venetoclax and 21% had a previous allo-HCT. Peripheral blood (55%) or bone marrow (45%) samples (median blast % before gradient separation: 34%) were processed. On a median of 20508 gated blasts (IQR 7796-31858), the median proportion of IL-1RAP+ cells was 24% (IQR 11-55, range 1-99) and the median Antibody Binding Capacity (ABC) was 970 (IQR 514-3722, range 35-3722). Based on IL-1RAP ABC distribution in CD45high/SSClow lymphocytes used as internal controls (median 38, IQR 16-96), we defined ABC ≥200 as the definition of IL-1RAP positivity in blasts; 34 (89%) patients had IL-1RAP+ blasts. In univariable analyses, neither sample source (PB vs BM, p=0.06) nor processing method (fresh vs thawed, p=0.28) influenced IL-1RAP ABC on blasts. Neither age, sex, ontogeny nor disease status (relapse vs primary refractory), history of intensive Cx, hypomethylating agents, venetoclax or allo-HCT had impact on IL-1RAP ABC on blasts (all p>0.05). Ten (26%) pts had flow-defined monocytic differentiation. These pts had comparable IL-1RAP ABC in gated blasts to those without monocytic differentiation (p=0.17). Median IL-1RAP ABC of blasts was 1724 in NPM1mut pts vs 856 in NPM1wtpts (p=0.04) and 1878 vs 814 in FLT3mut vs FLT3wt pts (p=0.004) while IDH1/2 mutations had no impact (p=0.91). In a multivariable linear regression followed by backward variable selection and accounting for sample source, NPM1 and FLT3 status, after backward variable selection, only FLT3 status influenced IL-1RAP ABC (p=0.04). All (100%) FLT3mut pts had an ABC ≥ 200 vs 81% of FLT3wt pts (p=0.5). This was in agreement with FLT3-ITD being the main genetic biomarker associated with increased IL-1RAP expression (FC=1.68, q-val <10-20) in a public transcriptomic dataset of 1224 AML patients (Severens, Leukemia 2024). A median of 2560 (IQR 542-9960) CD34+/CD38-/(CD97|TIM3|CLL1)+ LSCs were acquired. In the 29 pts with at least 200 gated LSCs, median IL-1RAP ABC in LSCs was 2,176 (IQR 1231-3059), representing a strong enrichment compared to the total blast population (p=0.006). IL-1RAP ABCs were significantly higher in FLT3mut vs FL3wt pts (p=0.01). All other demographic, treatment history or genetic variables had no impact on IL-1RAP ABC in LSCs. Conclusion. IL-1RAP is expressed on blasts in 89% of relapsed/refractory AML patients regardless of disease ontogeny and treatment history. IL-1RAP is over-expressed in phenotypically-defined LSCs compared to blasts. FLT3mut pts have higher IL-1RAP expression in both blasts and LSCs. This data supports the use of IL-1RAP as a target antigen in R/R AMLs. The safety and clinical activity of the first in class, 3rd generation autologous IL-1RAP-directed CAR T cell therapy (CCTx-001) will be tested in the RESOLVE AML-001 study (NCT06281847).