Baseline characteristics for the overall real-world cohort and the GARNET cohort (before and after matching)
| Overall real-world cohort (n=999) | GARNET (dostarlimab) cohort | ||||
| Before matching (n=153) | Scenario 1 (ESS=18) | Scenario 2 (ESS=62) | Scenario 3 (ESS=67) | ||
| Race, n (%) | |||||
| White | 841 (84.2) | 117 (76.5) | 82.6% | 81.8% | 84.2% |
| Black | 57 (5.7) | 5 (3.3) | 1.1% | 1.2% | 5.7% |
| Other | 78 (7.8) | 8 (5.2) | 3.3% | 4.6% | 7.8% |
| Unknown | 23 (2.3) | 23 (15.0) | 13.0% | 12.4% | 2.3% |
| Age, n (%) | |||||
| <65 years | 428 (42.8) | 75 (49.0) | 38.4% | 50.2% | 57.1% |
| ≥65 years | 571 (57.2) | 78 (51.0) | 61.6% | 49.8% | 42.9% |
| ECOG PS*, n (%) | |||||
| 0 | 320 (32.0) | 61 (39.9) | 36.2% | 39.4% | 37.1% |
| 1 | 181 (18.1) | 92 (60.1) | 63.8% | 60.6% | 62.9% |
| Unknown | 498 (49.8) | 0 | 0 | 0 | 0 |
| Histology†, n (%) | |||||
| Endometrioid | 424 (42.4) | 121 (79.1) | 42.4% | 42.4% | 42.4% |
| Non-endometrioid | 575 (57.6) | 30 (19.6) | 57.6% | 57.6% | 57.6% |
| Unknown | 0 | 2 (1.3) | 0 | 0 | 0 |
| FIGO stage†, n (%) | |||||
| I/II | 221 (22.1) | 65 (42.5) | 28.2% | 36.8% | 22.1% |
| III/IV | 778 (77.9) | 88 (57.5) | 71.8% | 63.2% | 77.9% |
| Tumor grade†, n (%) | |||||
| 1/2 | 274 (27.4) | 102 (66.7) | 27.5% | 45.9% | 40.0% |
| 3/4 | 389 (38.9) | 46 (30.1) | 38.9% | 50.4% | 56.2% |
| Unknown | 336 (33.6) | 5 (3.3) | 33.6% | 3.7% | 3.8% |
| Number of prior platinum-based therapies‡, n (%) | |||||
| 0 | 0 | 2 (1.3)§ | 0 | 0 | 0.7% |
| 1 | 999 (100) | 128 (83.7) | 100% | 100% | 76.9% |
| 2+ | 0 | 23 (15.0) | 0 | 0 | 22.4% |
| Surgery, n (%) | |||||
| Yes | 815 (81.6) | 135 (88.2) | 69.7% | 89.6% | 81.6% |
| No | 184 (18.4) | 18 (11.8) | 30.3% | 10.4% | 18.4% |
Scenario 1: Matching variables are histology, grade, and number of prior platinum-based therapies; scenario 2: Matching variables are histology and number of prior platinum-based therapies; scenario 3: Matching variables are race, histology, stage at initial diagnosis, and surgery (ECOG PS was not included as a matching variable due to patients with unknown status). Percentages for matching variables for each scenario are highlighted in bold. Differences in baseline characteristics between cohorts were measured using the Chi-squared test; if at least one cell was ≤5, then Fisher’s exact test was used.
*Status at index.
†at initial diagnosis.
‡in the recurrent/advanced setting.
§the prior therapies that these patients received were not counted as a line of therapy per trial protocol; these patients were excluded from this analysis.
ECOG PS, Eastern Cooperative Oncology Group Performance Status; ESS, effective sample size; FIGO, International Federation of Gynecology and Obstetrics.