Quantitative and qualitative analysis of neoantigens: genomic data integration to identify biomarkers of response to immunotherapy in non-muscle invasive bladder cancer

Abstract

Bladder cancer is the 9th most common cancer in the world and 70-80% of cases correspond to non-invasive bladder cancer (NMIBC). NMIBC treatment is surgical (tumor transurethral scraping) followed by adjuvant immunotherapy. In order to reduce the recurrence risk of the disease, immunotherapy is performed through intra-vesical applications of the tuberculosis vaccine BCG (Bacillus Calmette-Guérin), which results in the elimination of tumor cells through the stimulation of patient's immune response, capable of producing lymphocytes that supposedly recognize tumor-specific antigens (neoantigens) or the bacillus itself internalized by tumor cells. About 30-40% of patients do not respond to BCG treatment and recidivate, while 10-25% progress to invasive forms. To date there are no predictive biomarkers of BCG response implemented in the clinical oncology and the understanding of the molecular mechanisms underlying immunotherapy response is critical for the identification of biomarkers and patients prone to durable benefit. Correlation data between response to NMIBC treatment with BCG (as well as relapse-free or progression-free survival) and tumor mutational load (total number of somatic mutations) are controversial in the literature. For different tumor types, recent works shows that neoantigens with higher homology to infectious disease-related molecules (microbial epitopes) are biomarkers of long-term survivors; however, this analysis has not yet been adopted to explain progression and response to treatment in NMIBC bladder cancer type. Evidence that the progression of NMIBC is related to the specificity of the antigen presentation mechanism is the increased frequency of mutations in MHC (major histocompatibility complex) genes in cases of MIBC with previous treatment with BCG. Our hypotheses are: (i) the amount of neoantigens similar to microbial epitopes may be a predictive marker of response to BCG immunotherapy in NMIBC and (ii) somatic mutations and polymorphic variants in the MHC act respectively as silencers of the immune response against the tumor and decrease their degree of affinity to the tumor peptides, interfering in the recognition of neoantigens. Then, for quantitative and qualitative analysis of neoantigens, we will use bioinformatics tools and employ a mathematical model (neoantigen fitness model) that is based on the degree of similarity of tumor peptides to epitopes of pathogens. Quantitative and qualitative data, together with MHC data, will subsequently be correlated to the different clinical outcomes for these individuals, such as clinical benefit (absence of relapse or progression up to two years after treatment), relapse-free survival and survival progression free. (AU)