KIR and HLA Genotypes Influence Clinical Outcome in Multiple Myeloma Patients Treated with SAR650984 (Anti-CD38) in Combination with Lenalidomide and Dexamethasone

SAR650984 (SAR) is a humanized IgG1 monoclonal antibody that binds selectively to the human CD38 receptor. TCD11863 (TCD) is a phase Ib trial evaluating the combination of SAR with lenalidomide (LEN) and dexamethasone (Dex) in relapsed/refractory multiple myeloma (RRMM) [NCT01749969]. Because natural killer (NK) cells contribute to antibody-dependent cellular cytotoxicity, we hypothesized that immunogenetic factors contributing to NK cell function would influence clinical activity among RRMM patients treated with SAR/LEN/Dex. We therefore prospectively performed KIR and HLA class I genotyping on 31 RRMM patients enrolled in TCD. We used the typing data to stratify patients based on predicted differences in KIR3DL1, HLA-Bw4 receptor-ligand binding affinities. We used the Kaplan-Meier method to estimate probabilities of progression-free survival (PFS) and time to progression (TTP) among all 31 treated patients, with the log-rank test to evaluate differences in survival distributions based on patient KIR and HLA genotype status, the cox proportional hazard model to examine the association of the hazard of failure with genotype status, and the Fischer's exact test to compare overall response rates (ORR) among 28 evaluable patients. P-values are provided for exploratory purpose, since the study was not powered for formal comparison of efficacy variables by genotype subgroup. Of the 31 patients, 12 had a high-affinity KIR3DL1,HLA-B Bw4-80Ile compound genotype (including 3 with both HLA-A and HLA-BBw4-80Ile), and 19 lacked the KIR3DL1,HLA-B Bw4-80Ile genotype, with 7 of these 19 patients possessing a low-affinity KIR3DL1, HLA-Bw4-80Thr genotype (including 1 with HLA-ABw4-80Ile), 7 with a missing ligand or receptor genotype, and 5 with a KIR3DL1, HLA-ABw4-80Ile genotype not containing other HLA-B Bw4 alleles. Presence of a high-affinity KIR3DL1,HLA-B Bw4-80Ile genotype (n=12, 10 evaluable for ORR) was associated with high ORR and prolonged TTP and PFS relative to patients lacking KIR3DL1,HLA-B Bw4-80Ile (n=19, 18 evaluable for ORR; 90% ORR vs. 56%, P=0.10; TTP HR 0.11, [95% CI] 0.08-0.68, P<0.01; PFS HR 0.22, [95% CI] 0.05-0.98, P=0.03). The benefit of KIR3DL1,HLA-B Bw4-80Ile varied with gene dose, such that the proportion of patients remaining on treatment increased with increasing copy number of HLA-B Bw4-80Ile from 0 (n=19) to 1 (n=8) to 2 (n=4) copies, with 21% remaining on treatment at 6 months vs. 50% vs. 100% (TTP P=0.02). Because HLA-A Bw4-80Ile ligands bind KIR3DL1 with different strength and specificity than HLA-B Bw4-80Ile ligands, we separated patients with KIR3DL1, HLA-ABw4-80Ile from patients with KIR3DL1, HLA-BBw4-80Ile and found a specific association of prolonged PFS among patients with a pure high-affinity KIR3DL1,HLA-B Bw4-80Ile genotype (n=9; 7 patients remain on treatment at 6 months, 2 AEs), compared with patients with a pure KIR3DL1, HLA-A Bw4-80Ile genotype (n=5; 2 on treatment at 6 months, 3 PDs), a KIR3DL1, HLA-ABw4-80Ile genotype containing other HLA-B Bw4 alleles (n=4), a pure low-affinity KIR3DL1, HLA-Bw4-80Thr genotype (n=6; 2 on treatment at 6 months, 3 PDs, and 1 AE), and a missing KIR3DL1 ligand or receptor genotype (n=7; 3 on treatment at 6 months, 3 PDs, and 1 AE) [PFS P=0.02]. While the number of prior therapies (2-4 vs. 5-7 vs. 8-12, P=0.02) and ISS stage at the start of treatment (stage I to III, P<0.01) also influenced PFS, possession of KIR3DL1,HLA-B Bw4-80Ile was associated with prolonged PFS within each of these subgroups. In summary, a KIR3DL1, HLA-B Bw4-80Ile genotype predictive of high-affinity NK cell receptor-ligand interactions and potent NK cell licensing correlated with increased ORR and PFS among patients treated with SAR/LEN/Dex, although the small sample size requires validation in future studies.