2152 - Inferring the correlation between incidence rates of melanoma and the average tumor-specific epitope binding ability of HLA class I molecules in different populations

Background

Human Leucocyte Antigen (HLA) molecules are encoded by the most polymorphic genes in the human genome. The genetic variation of these genes are considerable across different geographic subpopulations. We hypothesised that this genetic variation might contribute to the risk of melanoma both at population and subject level.

Methods

We developed a cancer risk predictor based on the complete HLA class I genotype of individuals. The HLA-score, used in the predictor describes the ability of the HLA class I alleles of an individual to bind epitopes derived from 48 selected tumor antigens as an indicator of the breadth of the tumor-specific T-cell responses. We collected HLA data for subjects from 20 different geographic regions (ethnic populations) (n = 3278) as well as the corresponding melanoma incidence rates. The average HLA-scores were compared to the incidence rates. We also classified a mixed US population consisting of melanoma and healthy subjects based on their HLA-score.

Results

On population level, we found significant correlation between the incidence rates of melanoma and average HLA-scores in different geographic regions (R² = 0.5005; p < 0.001; n = 20; df = 18). The highest average HLA-scores (range 75-140) were obtained for the Far East Asian and Pacific regions, where the incidence rates are low (0.4-3.4 per 100,000 per year). The lowest average HLA-scores (range 50-90) were obtained in the European and US regions, where the rates are high (12.6-13.8 per 100,000 per year). On subject level, the risk ratio between the riskiest (HLA-score <34) and the most protected groups (HLA-score ≥96) was 5.69 comparing the top and bottom 20% of the HLA-score distribution (p < 0.05). These HLA-score ranges are consistent with the threshold values separating populations with low and high incidence rates of melanoma.

Conclusions

By developing a novel HLA-score determined by autologous HLA allele binding epitopes of tumor antigens, we showed that individuals with HLA allele sets supporting broader tumor-specific T-cell responses have lower risk of developing melanoma. These results imply that the HLA genotype and HLA-score could be used to determine the immunogenetic risk of melanoma.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Treos Bio Zrt.

Funding

Treos Bio Zrt.

Disclosure

L. Molnar: Shareholder / Stockholder / Stock options, Full / Part-time employment: Treos Bio Ltd. J. Toth: Shareholder / Stockholder / Stock options, Full / Part-time employment: Treos Bio Ltd. O. Lorincz: Shareholder / Stockholder / Stock options, Full / Part-time employment: Treos Bio Ltd. Z. Csiszovszki: Shareholder / Stockholder / Stock options, Full / Part-time employment: Treos Bio Ltd. P. Pales: Full / Part-time employment: Treos Bio Ltd. K. Pántya: Shareholder / Stockholder / Stock options, Full / Part-time employment: Treos Bio Ltd. M. Megyesi: Shareholder / Stockholder / Stock options, Full / Part-time employment: Treos Bio Ltd. E. Somogyi: Shareholder / Stockholder / Stock options, Full / Part-time employment: Treos Bio Ltd. E.R. Tőke: Shareholder / Stockholder / Stock options, Full / Part-time employment, Officer / Board of Directors: Treos Bio Ltd. All other authors have declared no conflicts of interest.