A multi-biomarker proteomic panel from cervical exfoliated cells for non-invasive detection of cervical cancer: A south indian study

Persistent infection with high-risk human papillomavirus (HR-HPV) is the primary cause of cervical cancer; however, HPV DNA testing alone does not reliably predict disease progression. Host protein alterations induced by HPV-mediated epithelial transformation offer potential complementary biomarkers for early detection and improved risk stratification. In this study, cervical exfoliated cell samples were collected from a community-based screening program in South India and analyzed for HR-HPV DNA using the Cobas® 5800 platform. Forty samples were selected for proteomic discovery and categorized into HPV16-positive, HPV18-positive, other high-risk HPV (OHR)-positive, and HPV-negative groups (n = 10 each). Pooled protein extracts from these groups were subjected to proteomic profiling using two-dimensional gel electrophoresis (2D-GE), followed by identification of differentially expressed proteins through matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry. Functional annotation and protein–protein interaction network analyses were performed using gene ontology and the STRING database. Fourteen significantly overexpressed protein spots corresponding to nine proteins were identified in HPV-positive samples, including transferrin (TF), fibrinogen beta chain (FGβ), cytokeratin-13 (KRT13), and other keratin family members. Network analysis revealed two dominant functional clusters: cytoskeletal organization and epithelial differentiation (keratin-associated), and metabolic, inflammatory, and stress-response pathways (TF, FGβ, CALCOCO1-associated). Validation was performed using ELISA on an independent cohort of 84 samples comprising cervical squamous cell carcinoma (SCC, n = 58) and negative for intraepithelial lesion or malignancy (NILM) (n = 26) cases. Among these, 88% (74/84) were HR-HPV positive, while 10 were HPV-negative. TF, FGβ, and KRT13 levels were significantly elevated in disease groups compared to cytologically normal samples (p < 0.000001). Receiver operating characteristic (ROC) analysis showed moderate diagnostic performance for individual markers (CK13: area under the curve [AUC] AUC = 0.747; TF: AUC = 0.704; FGβ: AUC = 0.656). A multivariable logistic regression model integrating all three biomarkers significantly improved diagnostic performance (p < 0.001; Nagelkerke R 2 = 0.322), achieving ∼80% sensitivity and ∼65% specificity. Overall, this study demonstrates that proteomic profiling of cervical exfoliated cells can identify biologically relevant host protein alterations linked to HR-HPV infection and disease progression. The combined biomarker panel (TF, FGβ, KRT13) enhances diagnostic accuracy and supports a minimally invasive, multi-analyte approach for improved triage, risk stratification, and scalable cervical cancer screening in resource-limited settings.