Identification of Candidate Protein Biomarkers for CIN2+ Lesions from Self-Sampled, Dried Cervico–Vaginal Fluid Using LC-MS/MS

SIMPLE SUMMARY: Cervical cancer is the fourth most common cancer among women worldwide, and screening programs increase its detection rate and survivability. The molecular screening of the presence of human papilloma viruses (HPV) as alternatives to physical examinations offers cost-efficient solutions and can be performed on self-collected samples. A persistent infection with HPV is necessary (but not sufficient) to develop cancer, and additional biomarkers are needed to increase the precision. Here, we have analysed protein biomarkers found in self-collected dried cervico–vaginal fluid (CVF) from both controls and women with cervical cancer pre-stages. Our conclusion is that protein biomarkers can be robustly detected from dried CVF and that these can aid in the detection of cervical cancer pre-stages. ABSTRACT: Molecular screening programs for cervical cancer detect the presence of human papilloma virus (HPV) in cell material or vaginal fluids. Persistent infection with high-risk HPV is a necessary pre-requisite, but the majority of infections do not lead to pathological states. Additional biomarkers are needed to increase the specificity of the molecular tests. Here, we have investigated the possibility of detecting protein biomarkers using mass spectrometry from dried self-sampled cervico–vaginal fluid deposited on FTA cards. We found significant intra-individual correlations (p < 2.2 × 10(−16)), although heterogenous protein profiles were obtained between individuals. Out of 3699 proteins found in total, 169 were detected in at least 95% of the samples. Using a discovery/replication design, 18 proteins were found to be significant in the discovery cohort, with higher values in those cases compared to controls. All of these were found to also have higher levels among the cases in the replication cohort, with one protein (DEAD-Box Helicase) remaining statistically significant. Finally, a predictive 7-protein multivariate model was developed with a sensitivity and specificity of 0.90 and 0.55, respectively. Our results demonstrate that robust measurements of protein biomarkers can be obtained from self-sampled dried CVF and that these could be used to predict cervical cancer pre-stages.