Abstract B14: Rapid RNA-based histotyping of ovarian carcinomas
Bibliographic record
Abstract
Abstract Background: Ovarian cancer is a series of distinct diseases typically identified by their histopathological appearance as high-grade serous (HGSC; 70% of cases), low-grade serous (LGSC; 5%), endometrioid (ENOCa, 8%), clear cell (CCC, 12%), and mucinous (MC; 5%) carcinomas. Each type has defining molecular events, gene/protein expression patterns, genetic risk factors, sites of origin, and responses to treatment. Gold standard treatment is surgery followed by platinum-taxane chemotherapy despite mounting evidence suggesting CCCs, MCs, and LGSCs are largely platinum-taxane resistant. If outcomes are to be improved, it is critical to adopt a type specific strategy. Retrospective review studies have suggested histotype may be misdiagnosed or omitted in up to 30% of cases. However, pathological diagnosis of histotypes has been greatly refined in recent years and the use of biomarkers as aides is becoming more widespread. Nonetheless a rapid and fully objective classifier of histotypes will undoubtedly improve diagnostic accuracy, especially in the case of pre-surgical biopsies where small amounts of material present a challenge. Methods: Over 1000 ovarian carcinoma samples underwent expert gynecopathological review to establish a gold standard diagnosis for the 5 major carcinoma types. RNA was extracted from FFPE tissues and levels of a pre-selected set of >100 genes were quantified using the NanoString GX system. Cohort was split with ~1/3 set aside for independent validation. Several statistical models were tested to generate a prediction algorithm for histological type including PAM, Random Forest, Lasso, Recursive Partitioning, and Discriminant Analysis. Feature selection methods and prediction error were examined using cross-validation in the train /test series prior to validation in the independent set. Results: Preliminary analysis suggests classification of the 5 major histotypes is possible using NanoString derived RNA expression levels. Accuracy appears to be equivalent to interobserver variation amongst expert gynecopathologist. Conclusions: The NanoString GX platform provides a stable and reproducible platform on which a robust single sample histological type classifier can be established. Our algorithm combined with the NanoString platform provides a rapid, and cost-effective option that does not require modification to current pathology lab tissue processing protocols. Diagnostic prediction require little material and is applicable to pre- and post- surgical specimens where an objective measure is desired to confirm diagnosis or aide in especially challenging cases. Citation Format: Michael S. Anglesio, Aline Talhouk, Steve E. Kalloger, Gholamreza Haffari, Robertson Mackenzie, Martin Cheung, Janine Senz, Christine Chow, Sherman Lau, Maria Intermaggio, Susan J. Ramus, Andreas du Bois, Jacobus Pfisterer, Jessica N. McAlpine, Friedrich Kommoss, Blake Gilks, Stefan Kommoss, David G. Huntsman. Rapid RNA-based histotyping of ovarian carcinomas. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: From Concept to Clinic; Sep 18-21, 2013; Miami, FL. Philadelphia (PA): AACR; Clin Cancer Res 2013;19(19 Suppl):Abstract nr B14.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.004 | 0.003 |
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".