Skip to main content
Premium Trial:

Request an Annual Quote

Immunotherapy Response Influenced by Cancer Patients' HLA Genotypes

NEW YORK (GenomeWeb) – Patients' human leukocyte antigen (HLA) genotypes influence how well they respond to immunotherapy, according to a new study.

HLA class I molecules bind peptides from intracellular proteins, including tumor proteins, and present them to CD8+ T cells that kill infected or cancerous cells. Immunotherapies that have been successful in treating some metastatic cancers rely on CD8+ T cells and HLA class I activity to work.

In a study appearing in Science today, researchers from Memorial Sloan Kettering Cancer Center and elsewhere examined whether patients' HLA variants affect their survival. They genotyped more than 1,500 advanced cancer patients who were treated with checkpoint blockage immunotherapies. The Sloan Kettering-led team found that heterozygosity at the HLA-I locus was linked to better overall survival after treatment.

"Our results have important implications for predicting response to [immune checkpoint blockade] and for the design of neoantigen-based therapeutic vaccines," Sloan Kettering's Timothy Chan and his colleagues wrote in their paper.

For their study, the researchers investigated two cancer patient cohorts, both of which were treated with immune checkpoint blockade drugs. One cohort included 369 patients with advanced melanoma or advanced non-small-cell lung cancer who were treated with anti-CTLA-4 or anti-PD-1 therapy, while the other cohort included 1,166 patients with a range of cancers — including melanoma and NSCLC — and who were given anti-CTLA-4, anti-PD-1/PD-L1 therapy, or a combination of those therapies.

Patients' HLA class I genotypes were gauged from normal DNA by either sequencing or an HLA typing assay.

As HLA class I molecules are both highly polymorphic and variable in their peptide-binding region, Chan and his colleagues suspected that people who are homozygous at one or more HLA-I locus would present a less diverse range of tumor-derived antigens. Likewise, they thought patients with more diverse HLA-I molecules would display a greater range of antigens and have increased post-therapy survival.

For both cohorts, the researchers examined HLA-I variation at the HLA-A, HLA-B, and HLA-C genes and found that homozygosity at at least one of those loci was associated with decreased survival. The effect, though, was greater for HLA-B and HLA-C. They noted that this could be because HLA-B is generally expressed at higher levels than HLA-A or HLA-C and because HLA-B binds to a greater range of residues.

Chan and his colleagues also examined the interplay between the number of somatic mutations within a cancer genome — which has also been linked to treatment response — and HLA-I variation. HLA-I homozygosity and low mutation burden was strongly linked to lower survival, as compared to heterozygous patients with high tumor mutation burden, they found.

Loss of heterozygosity at HLA-I loci within tumor genomes was also linked to decreased survival, especially among patients whose tumors had a low mutation load.

Since HLA-I alleles can be categorized into 12 different supertypes, the researchers also divvied their melanoma patients by their supertypes and found that two HLA-B supertypes — B44 and B62 — were associated with differences in survival after anti-CTLA-4 treatment. The link between the B44 supertype and better survival was driven by half a dozen alleles, while the B62 link to decreased survival was driven by one allele, HLA-B*15:01, they found.

The researchers traced the effect of this allele to the peptide-binding groove. In molecular dynamic simulations, they found that in HLA-B*15:01, a molecular bridge was present there, while it was broken or more flexible with other alleles. This bridge could impede the interaction with T-cell receptor and neoantigen recognition.

Because HLA-I genotypes can affect patient outcomes, Chan and his colleagues suggested that they be taken into account in future clinical trials of immunotherapies.