MSACL 2024 Abstract

Introduction: Current clinical high-grade serous ovarian cancer (HGSOC) diagnostic tests rely on measuring serum CA-125 (MUC16) levels to: test women who have vague symptoms of HGSOC, monitor women who have been diagnosed with HGSOC, or monitor women who are at high risk of developing HGSOC. However, this method is not adequately sensitive (78%) or specific (50%) to screen for HGSOC in the general population. Thus, routine screening such as imaging (transvaginal ultrasound, computed tomography scan, and magnetic resonance imaging) and pelvic exams are recommended for women who are at higher risk. The current lack of a reliable screening test for HGSOC in the general population emphasizes the critical need for more robust diagnostic biomarkers.

Objectives: In this study, we sought to develop and validate a multiplexed, targeted parallel reaction monitoring (PRM) assay for the relative quantification of 23 HGSOC protein biomarker candidates in sera. The measurands were 64 tryptic peptides from these candidate diagnostic protein biomarkers. Our goal was to identify a set of proteins with differential levels in the sera of patients with HGSOC compared to sera from healthy women or women with benign ovarian conditions, suggesting that these proteins may play an important role in the pathogenesis of HGSOC and could be developed as diagnostic biomarkers.

Methods: To develop a PRM assay for our target peptides in sera, we followed nationally recognized consensus guidelines for validating fit-for-purpose Tier 2 targeted MS assays. We used a Q-Exactive Plus mass spectrometer coupled with a Vanquish microflow LC system. Blood was collected immediately before surgery from 69 women with an abdominal mass suspected to be ovarian cancer (for the benign ovarian disease, early-stage HGSOC, and late-stage HGSOC cases) or from women with benign non-gynecological health conditions (e.g., eye surgery, hernia repair, hip replacement, and gallbladder removal) to serve as non-cancer controls. The endogenous target peptide concentrations were calculated using calibration curves in non-depleted serum for each target peptide. Assay linearity, LOD, LOQ, and repeatability (intra-day, inter-day, and total) were determined. Receiver operating characteristic (ROC) curves were analyzed to evaluate the diagnostic performance of the biomarker candidates.

Results: Among the 64 peptides corresponding to 23 proteins in our PRM assay, 24 peptides corresponding to 16 proteins passed the assay validation acceptability criteria. The LOD, LOQ, and linearity (represented by R2 values of the linear regression models) of the 64 peptides ranged from 0.22 to 95.15 fmol, 0.67 to 285.41 fmol, and 0.9754 to 0.9997, respectively. A total of 6 of these peptides from insulin-like growth factor-binding protein 2 (IBP2), sex hormone-binding globulin (SHBG), and TIMP metalloproteinase inhibitor 1 (TIMP1) were reliably quantified in sera from our patient cohort. To evaluate the diagnostic performance of individual peptides and combinations of peptides in HGSOC, we determined AUC values from ROC curves for IBP2_LEG, IBP2_LIQ, TIMP1_GFQ, and combinations of the 3 peptides in non-cancerous vs. late-stage HGSOC and benign vs. late-stage HGSOC conditions. The accuracy, sensitivity, specificity, and AUC of CA-125 to discriminate late-stage HGSOC from benign ovarian conditions were 94.3%, 94.4%, 94.1%, and 0.987 respectively. The combination of CA-125, IBP2_LEG, and IBP2_LIQ showed improved accuracy (96.4%) and sensitivity (100%) with lower specificity (93.3%) and AUC (0.985) compared to CA-125 alone. Although the CA-125 values alone had outstanding performance in our sample groups, the addition of our biomarker candidates IBP2_LEG and IBP2_LIQ increased the accuracy and sensitivity to detect late-stage HGSOC while maintaining similar specificity and AUC.

Conclusion: The frontier of clinical diagnostics is expanding to encompass new technologies such as PRM in clinical laboratories. Confirming the results from previously published studies using orthogonal analytical approaches, IBP2 was identified as a diagnostic biomarker candidate based on its significantly increased abundance in the late-stage HGSOC patient sera compared to the healthy controls and patients with benign ovarian conditions. To improve the potential clinical utility of our assay as a screening tool for women in the general population and to increase the specificity of the diagnostic performance of our assay, it would be essential to include sera from women who do not have elevated levels of CA-125. One reason for the inability of candidate biomarkers to achieve clinical utility is their low specificity in large cohorts. Another strategy to improve the diagnostic performance of our PRM assay would be to stratify our patient cohort by menopausal status and to age-match the cases and controls, thereby enhancing the assay’s clinical utility.