Table 3.
Baseline characteristics before and after IPTW in the CheckMate 238 and real-world cohorts
| Before IPTW | After IPTWa | |||||||
|---|---|---|---|---|---|---|---|---|
| CheckMate 238 (n = 369) |
Real-world (n = 451)b |
Standardized differencec | P value | CheckMate 238 (n = 353.3) |
Real-world (n = 416.5) |
Standardized differencec | P value | |
| Patient characteristics | ||||||||
| Age at index date, years, mean (SD)d | 54.7 (13.4) | 62.0 (14.0) | 0.529 | < 0.001 | 58.2 (13.5) | 59.2 (15.1) | 0.069 | 0.449 |
| Sex | ||||||||
| Male | 216 (58.5%) | 282 (62.5%) | 0.082 | 0.275 | 59.5% | 59.5% | 0.001 | 0.988 |
| Female | 153 (41.5%) | 169 (37.5%) | 40.5% | 40.5% | ||||
| Race | ||||||||
| White or missing | 345 (93.5%) | 393 (87.1%) | 0.216 | < 0.01 | 90.6% | 88.7% | 0.064 | 0.512 |
| Non-white | 24 (6.5%) | 58 (12.9%) | 9.4% | 11.3% | ||||
| Disease characteristics | ||||||||
| Disease stage at initial diagnosis (per AJCC-8) | ||||||||
| IIIA/IIIB | 120 (32.5%) | 156 (34.6%) | 0.044 | 0.583 | 34.0% | 32.6% | 0.031 | 0.730 |
| IIIC/IIID | 249 (67.5%) | 295 (65.4%) | 66.0% | 67.4% | ||||
| Time from surgical resection to index date, months, mean (SD) | 2.2 (0.6) | 1.4 (0.5) | 1.410 | < 0.001 | 1.8 (0.7) | 1.7 (0.6) | 0.263 | < 0.01 |
| ECOG PS | ||||||||
| 0 or missing | 333 (90.2%) | 378 (83.8%) | 0.192 | < 0.01 | 87.0% | 86.0% | 0.028 | 0.768 |
| 1 | 36 (9.8%) | 73 (16.2%) | 13.0% | 14.0% | ||||
| Comorbidities | ||||||||
| Diabetes | 21 (5.7%) | 33 (7.3%) | 0.066 | 0.428 | 6.0% | 6.5% | 0.020 | 0.813 |
| Chronic pulmonary disease | 19 (5.1%) | 17 (3.8%) | 0.067 | 0.431 | 4.6% | 4.0% | 0.025 | 0.760 |
| Atrial fibrillation | 4 (1.1%) | 16 (3.5%) | 0.164 | < 0.05 | 1.7% | 2.6% | 0.065 | 0.454 |
aThe mean of stabilized truncated weights calculated from the propensity scores among patients in the CheckMate 238 and real-world cohorts was 0.96 (SD, 0.84) and 0.92 (SD, 0.62), respectively
b451 out of 452 patients in the real-world cohort were included because one patient with missing comorbidity profiles was excluded
cA standardized difference of < 0.1 was considered an inconsequential imbalance between the two cohorts
dThe index date was defined as the date of randomization to adjuvant nivolumab treatment in the CheckMate 238 cohort and the initiation date of the adjuvant nivolumab treatment in the real-world cohort
AJCC-8 American Joint Committee on Cancer, Cancer Staging Manual, eighth edition, ECOG PS Eastern Cooperative Oncology Group performance status, IPTW Inverse probability of treatment weighting, SD Standard deviation