Abstract 5465: HRDetect as a predictive score of platinum response in advanced PDAC
Bibliographic record
Abstract
Abstract Background: Predictors of response to DNA damaging agents may expand the population of patients who could benefit from platinum and/or PARP inhibition in pancreatic cancer (PDAC). The HRDetect score derived from whole genome sequencing (WGS) incorporates patterns of substitution base signatures and structural variation and is associated with tumours deficient in homologous recombination repair (HRD). Methods: The prospective COMPASS trial (NCT 02750657) enrolls patients with advanced PDAC prior to systemic treatment for WGS and RNAseq. Choice of combination chemotherapy is at the discretion of the physician. HRD tumours are identified by the presence of a number of known genomic characteristics including evidence of biallelic inactivation of BRCA/PALB2 or the RAD51 family of genes, the presence of signature 3, the number and proportion of indels that are deletions of at least 4 base pairs (bp), and the presence of rearrangement signatures 3 and 5. We ascertained the number of patients with high HRDetect scores (>0.7) and evaluated outcomes and response to chemotherapy including platinum regimens. Results: As of 1st May 2019 204 eligible patients were enrolled. High HRDetect scores (HRDetecthi ) were present in 26/204 (13%). Only half of these cases were represented by HRD tumours including: 9 germline BRCA1/2 or PALB2 cases with biallelic inactivation and 3 tumours with somatic biallelic inactivation of BRCA2, RAD51C and XRCC2. 1 additional HRD case was of unknown aetiology. All HRD cases were HRDetecthi. Two germline carriers of BRCA2 without a second somatic hit were HR intact with low HRDetect scores. Of the remaining 13 non-HRD tumours identified as HRDetecthi, 6 exhibited a tandem duplicator phenotype (TDP) with tandem duplications ranging from 10Kbp to 1Mbp in size. These tumours did not have any other genomic characteristics of HRD. Notably, of the 204 patients included in this analysis, a TDP was present in 15 (7%), of which 40% were HRDetecthi. In patients treated with modified FOLFIRINOX(mFFX) (n=111, *includes 2 patients treated with cisplatin/gemcitabine) the response rate in HRDetecthi patients (n=16) was 56% vs 20% in HRDetectlo patients (n=95) p=0.005. There was no difference in response rates according to HRDetect score in patients treated with gemcitabine-nab-paclitaxel; HRDetecthi (n=7) 14.3% vs. HRDetectlo(n=71) 32.8%, p=0.42. Median OS in all patients receiving mFFX (n=111) was 10.2 months; in patients with HRDetecthi PDAC, median OS was 15.5 mths vs. 9.9 mths in HRDetectlo PDAC, (HR 0.42, 95% CI 0.24-0.87, p=0.01). In all patients (n=78) receiving gemcitabine based combination treatment the median OS was 8.1 mths; in HRDetecthi patients median OS was 11.5mths vs 7.8 mths in HRDetectlo patients (HR 0.94, 95% CI 0.42-2.13, p=0.8). Conclusions: In patients with advanced PDAC with high HRDetect scores, response rate to FFX is over 50% and survival is significantly longer compared to patients with PDAC and low HRDetect scores. HRDetect identifies an additional population of patients with non-HRD PDAC who benefit from DNA damaging agents. Citation Format: Grainne M. O'Kane, Robert E. Denroche, Amy Zhang, Sarah Picardo, Robert C. Grant, MIchael J. Allen, Gun Ho Jang, Yifan Wang, Anna Dodd, Stephanie Rampotar, Shawn Hutchinson, Mustaphe Tehfe, James J. Biagi, Dianne Chadwick, Bernard Lam, Julie Wilson, Faiyaz Notta, Sandra E. Fischer, George Zogopoulos, Steven Gallinger, Jennifer J. Knox. HRDetect as a predictive score of platinum response in advanced PDAC [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5465.
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 | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
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".