?Background: Human leukocyte antigen (HLA) heterozygosity expands the repertoire of recognizable antigens, optimizing the immune response to cellular threats. HLA heterozygosity is associated with reduced risk and severity of infections and may play a role in cancer susceptibility as well, although evidence is mixed. There are 8 classical HLA loci with varying and specific immune functions: HLA-I (HLA-A, HLA-B, HLA-C) and HLA-II (HLA-DPA1, HLA-DPB1, HLA-DQA1, HLA-DQB1, HLA-DRB1). Likely which specific loci are heterozygous may be biologically important for cancer susceptibility, but this is underexplored.
 
Objectives: To investigate the association between HLA heterozygosity and cancer in a diverse cohort.
 
Methods: The Mount Sinai BioMe is a longitudinal cohort and biorepository linking electronic health records with germline whole-exome sequencing data. HLA alleles were inferred using POLYSOLVER with 2-field (HLA-I) or 3-field (HLA-II) resolution. Maximal heterozygosity status was defined as six (HLA-I) or ten (HLA-II) unique alleles. Cancer diagnoses were provided by the Mount Sinai Cancer Registry. Logistic regression analysis was used to investigate the association between HLA heterozygosity and cancer, per each HLA locus, overall and stratified by genetically determined ancestry.
 
Results: 28,180 individuals were successfully HLA-I typed. The mean age at blood sample donation was 51.1 years old, and female sex was slightly more common at 58.4%. Most participants were heterozygous at HLA-I loci; HLA-A (95.8%), HLA-B (91.6%), HLA-C (90.1%), and maximal heterozygosity (83.2%). A smaller subset of participants had HLA-II typing (n=15,976); heterozygosity for these loci was less frequent; HLA-DPA1 (53.4%), HLA-DPB1 (83.5%), HLA-DQA1 (66.2%), HLA-DQB1 (89.2%), HLA-DRB1 (93.4%), maximal heterozygosity (32.4%). There were 2,916 (10.4%) recorded incident cancer cases.
 
In the overall cohort, both HLA-B and HLA-C were associated with decreased odds of any cancer diagnosis (Odds Ratio [OR]: 0.89, 95% Confidence Interval [CI]: 0.78-1.01 and OR: 0.90, 95% CI: 0.79-1.02, respectively), as was being maximally HLA-I heterozygous (OR: 0.91, 95% CI: 0.82-1.01). Among HLA-II loci, only HLA-DQB1 (OR: 0.83, 95% CI: 0.72-0.96) and maximal heterozygosity (OR: 0.89, 95% CI: 0.80-0.99) were statistically significantly protective against cancer risk.
 
The two most common genetically determined ancestry types were African (n=11,136, 39.5%) and European (n=10,269, 36.5%). Among African participants, HLA-B (OR: 0.82, 95% CI: 0.68-1.01), maximal HLA-I (OR: 0.87, 95% CI: 0.75-1.02) and HLA-DPA1 heterozygosity (OR: 0.87, 95% CI: 0.74-1.02) were associated with decreased odds of cancer. For Europeans, HLA-B (OR: 0.81, 95% CI: 0.66-1.01), HLA-C (OR: 0.73, 95% CI: 0.61-0.89) and maximal HLA-I heterozygosity (OR: 0.80, 95% CI: 0.68-0.94) were associated with decreased odds of cancer, as were HLA-DQB1 (OR: 0.76, 95% CI: 0.61-0.97) and maximal HLA-II heterozygosity (OR: 0.79, 95% CI: 0.64-0.96). 
 
Conclusions: In a large, multi-ethnic cohort, HLA heterozygosity appeared to be protective against cancer risk. However, the effects appear to be modulated by ancestry, possibly because of each locus’s unique evolutionary history and biological function. For instance, HLA-B is known to be under the greatest selective pressure from pathogens, so that epidemic events alter allelic frequencies in impacted populations. HLA loci represent promising biomarkers of cancer susceptibility and should be the focus of ancestry-aware, cancer prevention strategies.

Association Between HLA Heterozygosity and Cancer Risk