Abstract A040: Improving specificity for ovarian cancer screening using a novel extracellular vesicle-based blood test
Bibliographic record
Abstract
Abstract Introduction: The low incidence of ovarian cancer (OC) in average risk individuals dictates any OC screening method needs to be both highly sensitive and highly specific. Even using the multimodal OC screening strategy combining longitudinal serum CA125 and transvaginal ultrasound, 16% of OCs were not detected and false positives due to benign ovarian tumors, which are 5-10 times more prevalent than OC, remained high. Due to lack of mortality benefit, population screening is still not recommended. We hypothesized that detecting multiple colocalized protein or glycosylation epitopes (PGEs) on single tumor-derived extracellular vesicles (EVs) would increase specificity and sensitivity for detection of OC. Methods: The Mercy Halo™ OC test employs immunoaffinity capture and proximity-ligation qPCR to detect combinations of up to 3 PGE biomarkers to maximize specificity and measures multiple combinations to maximize sensitivity. We selected 3 PGE combinations to form a panel test to distinguish high-grade serous ovarian cancer (HGSC) from benign ovarian tumors and normal patients and compared performance of our test to CA125. We used a case-control training set from university-associated Canadian biobanks comprised of EDTA plasma from 124 healthy controls, 89 HGSC cases (17 Stage I, 35 Stage II, 37 Stage III)(10 BRCA+), and 192 benign adnexal masses to lock down the test, the data interpretation algorithm, and the cut-off between cancer and non-cancer. Performance was verified in an independent blinded diagnostic case-control set of 397 serum samples from 138 healthy controls, 111 benign adnexal masses, 20 borderlines, 66 HGSC and 62 non-HGSC from the UK Ovarian Cancer Population Study. For comparison, a CA125 ELISA was run for all samples in both studies. Results: Using the locked assay, algorithm, and cut-off established in the training cohort, our test showed a specificity in healthy controls of 92.8% (128/138; 95% CI: 0.87-0.96), a sensitivity in HGSC of 97.0% (64/66; 95% CI: 0.90-0.99), and an AUC of 0.92 (95% CI 0.89-0.95) in the verification cohort. We correctly identified 11/13 Stage I and all 54 Stage II/III/IV HGSC cases. Our test also detected 76.8% (63/82) of the borderline and non-HGSC histotypes. CA125 detected 92.4% (61/66) of the HGSC cases and 84.1% (69/82) of the borderlines and non-HGSC. The greatest improvement in test performance relative to CA125 was seen with benign adnexal masses. In the verification cohort 35.1% (39/111) of the benign samples were CA125 positive compared to 21.6% (24/111) with our test. This reduction in false positives was seen across 10 benign histotypes tested. Conclusions: These results demonstrate that our test using colocalized PGEs on EVs can detect OC, especially HGSC, in both plasma and serum with high sensitivity and specificity. The combinations of 5 biomarkers rather than CA125 alone also decreased benign false positives. The test performance suggests that our test may be useful in average and/or high-risk ovarian cancer screening. Future studies will explore this in asymptomatic populations. Citation Format: Emily S. Winn-Deen, Sanchari Banerjee, Daniel Gusenleitner, Laura T. Bortolin, Jonian Grosha, Kelly M. Biette, Karen Copeland, Christopher R. Sedlak, Bilal Hamzeh, Anthony D. Couvillon, Delaney M. Byrne, Peter A. Duff, Lauren T. Cuoco, MacKenzie S. King, Aleksandra Gentry-Maharaj, Sophia Apostolidou, Dawn R. Mattoon, Christine D. Berg, David F. Ransohoff, Steven J. Skates, Usha Menon. Improving specificity for ovarian cancer screening using a novel extracellular vesicle-based blood test [abstract]. In: Proceedings of the AACR Special Conference on Ovarian Cancer; 2023 Oct 5-7; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_2):Abstract nr A040.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".