Abstract
Background: The development of effective cancer vaccines is hindered by extensive tumor heterogeneity and human leukocyte antigen (HLA) diversity, which limit the scalability and population-wide applicability of personalized vaccine approaches. Universal cancer vaccines targeting conserved tumor antigens represent a promising alternative; however, systematic identification of broadly reactive T-cell epitopes remains a major challenge.
Methods and Results: In this study, we employed an immunoinformatic approach to predict and prioritize T-cell epitopes suitable for universal cancer vaccine development. Pan-cancer transcriptomic and proteomic datasets were used to identify tumor-associated antigens consistently overexpressed across multiple cancer types. Both MHC class I and class II T-cell epitopes were predicted from selected antigens and evaluated for HLA binding affinity, promiscuity, and sequence conservation. Population coverage analysis was integrated to estimate global applicability across diverse ethnic groups. The analysis identified a focused set of epitopes derived from conserved oncogenic proteins, including those involved in telomere maintenance, cell cycle regulation, and apoptosis inhibition, which demonstrated broad HLA binding and high predicted population coverage. Cross-cancer expression and immunogenicity assessments further supported the translational rel evance of these candidates.
Conclusion: This immunoinformatic framework demonstrates the feasibility of identifying broadly reactive T-cell epitopes capable of overcoming tumor and population heterogeneity. The findings provide a computa tional foundation for the rational design of universal cancer vaccines and support further experimental vali dation of the identified epitopes as scalable components of next-generation cancer immunotherapy strategies.
To Read or Download the Article PDF