Abstract 5135: Antitumor activity of dostarlimab by PD-L1 and tumor mutation burden (TMB) in patients (pts) with mismatch repair deficient and proficient (dMMR and MMRp) tumors in the GARNET trial
Bibliographic record
Abstract
Abstract Background: Dostarlimab is a programmed death 1 (PD-1) inhibitor approved as monotherapy in pts with dMMR recurrent/advanced endometrial cancer (EC) that has progressed on or following treatment with a platinum-containing regimen or solid tumors that have progressed on or following prior treatment, with no satisfactory alternative treatment options. Here we report on a post-hoc analysis of antitumor activity by PDL1 expression and TMB in pts with dMMR and MMRp solid tumors in the GARNET trial. Methods: GARNET (NCT02715284) is a phase 1, multicenter, open-label, single-arm study of dostarlimab in pts with advanced/recurrent solid tumors. Three expansion cohorts enrolled pts based on MMR status: dMMR (A1) and MMRp (A2) advanced/recurrent EC, and dMMR non-EC solid tumors (F). Pts received dostarlimab 500 mg IV Q3W for 4 cycles, then 1000 mg IV Q6W until progression or discontinuation. TMB and PDL1 were exploratory biomarkers. TMB status was determined by Foundation One test; TMB-high (TMB-H) was defined as ≥10 mutations/Mb. PDL1 expression was determined by combined positive score (CPS) by Ventana assay; PDL1-high (PDL1-H) was defined as CPS ≥1. The study was not powered to assess antitumor activity within subgroups. Results: TMB-H and PDL1-H were common in dMMR solid tumors; PDL1-H was observed in 39.4% of MMRp EC tumors (Table). Objective response rate (ORR) was higher in pts with TMB-H/PDL1-H tumors (55.6% for all cohorts, combined; Table). Safety for each cohort was previously reported.1 Conclusions: PDL1-H and TMB-H were frequently observed in the dMMR EC and non-EC cohorts, regardless of tumor type; PDL1-H was also prevalent in MMRp EC tumors. Although not powered to assess antitumor activity, ORR by BICR per RECIST v1.1 was higher in pts with TMB-H and PDL1-H solid tumors. Across cohorts, dMMR status was predictive of response. 1.Andre T, et al. Ann Oncol 2021;32(suppl 5):S829-S866. 991P. Table A1(dMMR EC)N=103 F(dMMR non-EC)N=106 A1+F(dMMR combined)N=209 A2(MMRp EC)N=142 A1+A2+F (Total)N=351 Biomarker distribution, n (%) TMB High 85 (82.5) 79 (74.5) 164 (78.5) 9 (6.3) 173 (49.3) Low 13 (12.6) 9 (8.5) 22 (10.5) 129 (90.8) 151 (43.0) Unknown 5 (4.9) 18 (17.0) 23 (11.0) 4 (2.8) 27 (7.7) PD-L1 High 56 (54.4) 52 (49.1) 108 (51.7) 56 (39.4) 164 (46.7) Low 23 (22.3) 17 (16.0) 40 (19.1) 45 (31.7) 85 (24.2) Unknown 24 (23.3) 37 (34.9) 61 (29.2) 41 (28.9) 102 (29.1) ORR by BICR per RECIST v1.1, n/N (%, 95% CI)a Overall 46/103(44.7, 34.9–54.8) 41/106(38.7, 29.4–48.6) 87/209(41.6, 34.9–48.6) 19/142(13.4, 8.3–20.1) — TMB-L/PDL1-L (L/L) 1/5(20.0, 0.5–71.6) 1/3 (33.3, 0.8–90.6) 2/8(25.0, 3.2–65.1) 2/43(4.7, 0.6–15.8) 4/51(7.8, 2.2–18.9) TMB-L/PDL1-H (L/H) 2/5(40.0, 5.3–85.3) 1/2(50.0, 1.3–98.7) 3/7(42.9, 9.9–81.6) 7/50(14.0, 5.8–26.7) 10/57(17.5, 8.7–29.9) TMB-H/PDL1-L (H/L) 5/17(29.4, 10.3–56.0) 3/14(21.4, 4.7–50.8) 8/31(25.8, 11.9–44.6) 0/1(0, 0–97.5) 8/32(25.0, 11.5–43.4) TMB-H/PDL1-H (H/H) 29/50(58.0, 43.2–71.8) 22/43(51.2, 35.5–66.7) 51/93(54.8, 44.2–65.2) 4/6(66.7, 22.3–95.7) 55/99(55.6, 45.2–65.5) aOnly those pts with both known TMB status and known CPS were included in ORR calculations Citation Format: Thierry André, Susana Banerjee, Dominique Berton, Susan L. Ellard, Begoña Jimenez, Vanessa Samouëlian, Lucy Gilbert, Valentina Boni, Xinwei Han, Grace Antony, Jennifer Veneris, Ana Oaknin. Antitumor activity of dostarlimab by PD-L1 and tumor mutation burden (TMB) in patients (pts) with mismatch repair deficient and proficient (dMMR and MMRp) tumors in the GARNET trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5135.
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) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.003 | 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".