Abstract 2287
Background
HLA class I-dependent immune activity is linked to autoimmune diseases, and HLA class I-dependent CD8+ T cells are required for immune checkpoint blockade (ICB) anti-tumor activity. It is unknown if HLA class I is predictive of toxicity to ICB.
Methods
100 patients (pts) with mixed solid tumors received single agent P (anti-PD-1) 200 mg IV Q 3 weeks in the investigator-initiated Phase II trial (INSPIRE study, NCT02644369). Germline whole exome sequencing (WES) of peripheral blood mononuclear cells was analyzed using the Illumina HiSeq2500 platform. Consensus HLA class I alleles were predicted from WES using HLAminer and HLAVBSeq. Using univariate Fisher’s exact test and logistic models adjusting for HLA features, heterozygosity of HLA-A, -B and -C, individual HLA alleles and HLA haplotype dimorphism at positions −21M and -21T of the HLA-A and -B leader sequence were analyzed as predictors of: 1) toxicity defined as ≥ Gr 2 immune-related adverse events (irAE) with at least possible attribution to P; and 2) clinical benefit (CBR) defined as either partial response or stable disease lasting ≥ 6 cycles of P.
Results
In the overall cohort of 100 pts, the frequency of irAE and CBR from P was 21% and 25%, respectively. Thus far, 99 patients had their HLA class I genotype determined. Univariate analysis showed heterozygosity of HLA-A, -B and -C, compared to homozygosity of at least one HLA locus, was not predictive of toxicity (≥ Gr 2 irAE 16.7% vs 27.3%, p = 0.29) but did trend to less response (CBR 19.7% vs 36.4%, p = 0.088). Individual heterozygosity of HLA-A, -B or -C, and HLA-A and -B haplotype dimorphism was not predictive of either toxicity or response. HLA-A*02 allele showed a trend to toxicity (≥ Gr 2 irAE 26.8% vs 11.6%, p = 0.079). A pertinent exploratory toxicity model is summarized in the table.Table: 95P
| Toxicity | |
|---|---|
| A*01 and A*02 (n = 7) | 4(57%) |
| A*01 and no A*02 (n = 13) | 2(15%) |
| A*02 and no A*01 (n = 49) | 11(22%) |
| Neither A*01 nor A*02 (n = 30) | 3(10%) |
Table Fisher exact p-value=0.0503
Conclusions
This study is the first to assess the association between HLA class I genotype and toxicity to P. There is a possible association of HLA-A*01 and HLA-A*02 with toxicity to P.
Clinical trial identification
Trial protocol number: NCT02644369.
Legal entity responsible for the study
Lillian Siu.
Funding
Merck.
Editorial Acknowledgement
Disclosure
P. Bedard: Research funding: Bristol-Myers Squibb, Sanofi, AstraZeneca, Genentech/Roche, Servier, GlaxoSmithKline, Novartis, SignalChem, PTC Therapeutics, Nektar, Merck, Seattle Genetics. A. Spreafico: Support for clinical trials: Merck, Novartis, BMS. A. Razak: Research funding: Merck, BMS, Pfizer, Karyopharm, Deciphera, Eli Lilly, Boehringer, Boston Biomedicals, Roche, Novartis and Genentech; Consultancy: Eli Lilly, Eisai, Boeringher Ingelheim, Merck. P. Ohashi: Consultancy/advisory: Symphogen Inc., Providence Pharmaceuticals, Inc., Baxalta US. Inc., Lion Biotechnologies, Inc. T. Pugh: Honoraria: Merck, Prosigna, Chrysalis Biomedical Advisors; Consulting: DynaCare Research. Funding: Boehringer Ingelheim; Patents, royalties; other intellectual property: Hybrid-capture sequencing for determining immune cell clonality, combined hybrid-capture DNA sequencing for disease detection. L.L. Siu: Advisory board and funding for clinical trials: Merck. All other authors have declared no conflicts of interest.