Serum proteomic-based analysis identifying autoantibodies against PRDX2 and PRDX3 as potential diagnostic biomarkers in nasopharyngeal carcinoma

Background: Nasopharyngeal carcinoma (NPC) is a major head and neck cancer with high occurrence in Southeast Asia and southern China. We aimed to identify autoantibodies that may contribute to early detection of NPC. Methods: We used serological proteome analysis to identify candidate autoantibodies against tumor-associated antigens. Levels of autoantibodies and Epstein-Barr virus capsid antigen-IgA (VCA-IgA) were measured by ELISA in 129 patients with NPC and 100 normal controls. We employed receiver operating characteristics to calculate diagnostic accuracy. Results: Sera from patients with NPC yielded multiple spots, two of which were identified as PRDX2 and PRDX3. Levels of serum autoantibodies against PRDX2 and PRDX3 were significantly higher for patients with NPC than for normal controls (P < 0.01), respectively. Combined detection of autoantibodies against PRDX2 and PRDX3 and VCAIgA provided a high diagnostic accuracy in NPC (an area under the curve (AUC) of 0.811 (95% CI 0.753-0.869), 66.7% sensitivity, and 95.0% specificity). This combination maintained diagnostic performance for early NPC with AUC value of 0.754 (95% CI 0.652-0.857), 50.0% sensitivity, and 95.0% specificity. Conclusions: This study reports autoantibodies against PRDX2 and PRDX3 identified by a proteomic approach in sera from NPC patients. Our findings suggest that autoantibodies against PRDX2 and PRDX3 may serve as supplementary biomarkers to VCA-IgA for the screening and diagnosis of NPC.