Introduction
Cancer treatment has been a global challenge. As part of the current standard treatment, immune checkpoint inhibitors (ICIs) therapy benefits the survival of a considerable number of patients across diverse cancer types (Havel et al., 2019; Hellmann et al., 2018; Samstein et al., 2019; Keenan et al., 2019). However, as the prognosis and response of ICI therapies vary greatly among patients, the need for effective biomarkers to distinguish potential benefit population is urgent (Havel et al., 2019; Hellmann et al., 2018; Keenan et al., 2019; Chan et al., 2019). Current immunotherapy medication guidelines mainly focus on TMB level and PD-L1 expression, ignoring the role of antigen presentation in the activation of anti-tumor immune response (Hellmann et al., 2018; Samstein et al., 2019; Chan et al., 2019; Rizvi et al., 2018). The human leukocyte antigen (HLA) complex is the most polymorphic gene system in humans, which encode the heavy chain in the hetero-dimer of MHC-I (major histocompatibility complex class I) molecules (Parham et al., 1989; Klein & Sato, 2000). As essential components involved in tumor antigen presentation, MHC-I molecules present tumor neoantigens to the cell surface for recognition by T cell receptors and activate immune response (Bjorkman et al., 1987; Parham & Ohta, 1996; McGranahan et al., 2017). By interrupting tumor antigen recognition, tumor cells with defected antigen presentation may escape from anti-tumor immune clearance and survive, which in turn affects the efficacy of ICI therapies (Campoli & Ferrone, 2008; Hicklin et al., 1999; Hiraki et al., 2004; Mehta et al., 2007).
Several studies have revealed the significance of HLA-I LOH in innate immunity. In particular, loss of alleles in HLA-C may inhibit immune clearance through the activation of NK cells (Thielens et al., 2012). In addition, multiple clinical researches have shown that HLA-I LOH is widespread in a variety of cancers and HLA-I LOH combined with traditional biomarkers could present better ICI therapies prognostic ability (Shim et al., 2020; Montesion et al., 2020). Shim et al. (2020) proposed that HLA-corrected TMB is a better predictor of ICI therapies than TMB alone. Montesion et al. (2020) pointed out that HLA-I LOH combined with TMB was a better factor for patient stratification in NSCLC. In addition, prior studies have revealed that the genotype of HLA-I alleles affects the response to immune checkpoint inhibitors (Chowell et al., 2018).
However, these previous studies were mainly based on western populations, while the incidence of HLA-I LOH in Chinese pan-cancer population remains unclear. In addition, the incidence of HLA-I LOH varies due to the differences in chromosomal instabilities among cancer types, highlighting the importance of HLA-I LOH studies in large population pan-cancer cohorts (Montesion et al., 2020). To illuminate this uncharted area, we analyzed the genomic variations in a 1638 Chinese pan-cancer cohort to delineate the HLA-I LOH landscape together with clinical and genomic features of those patients who harbored HLA-I LOH. In prior studies, the precise HLA-I LOH detection requires whole-exome sequencing (WES), which is not conducive in large-scale clinical applications due to the high cost of WES. Our study evaluated the reliability of HLA-I LOH analysis using the LOHHLA algorithm (McGranahan et al., 2017) based on a 1021-gene panel sequencing results. Exon 2, exon 3, and bilateral introns of HLA-A/B/C genes were well covered in the 1021-gene panel. We aimed to find the complex role of HLA-I LOH in tumor evolution which may provide valuable information for clinical practice and follow-up ICIs therapy research.