Table 2.
Secondary Efficacy Outcomes.*
| RECIST 1·1 per central review | RECIST 1·1 per investigator | |||
|---|---|---|---|---|
| Atezolizumab plus cabozantinib (N=259) | Cabozantinib (N=254) | Atezolizumab plus cabozantinib (N=263) | Cabozantinib (N=259) | |
| Confirmed objective response, n (% [95% CI]) | 105 (41% [35%–47%])† |
104 (41% [35%–47%])† |
100 (38% [32%–44%])‡ |
108 (42% [36%–48%])‡ |
| Complete response, n (%) | 0 | 2 (1%) | 4 (2%) | 2 (1%) |
| Partial response, n (%) | 105 (41%) | 102 (40%) | 96 (37%) | 106 (41%) |
| Stable disease, n (%) | 131 (51%) | 121 (48%) | 127 (48%) | 120 (46%) |
| Progressive disease, n (%) | 11 (4%) | 13 (5%) | 24 (9%) | 17 (7%) |
| Not evaluable or missing, n (%) | 12 (5%) | 16 (6%) | 12 (5%) | 14 (5%) |
| Ongoing objective response at data cutoff, n/N (%)§ | 53/105 (50%) | 55/104 (53%) | 58/100 (58%) | 48/108 (44%) |
| Median duration of response (95% CI), mo | 12·7 (10·5–17·4) | 14·8 (11·3–20·0) | NE (10·4–NE) | 12·2 (9·7–14·5) |
| Median duration of response range, mo | 2·1+ to 22·9+ | 2·3+ to 25·6+ | 2·1+ to 23·2+ | 2·1+ to 25·6+ |
Included are patients who presented with measurable disease according to RECIST 1·1, as assessed by either a central review facility or by investigators.
The estimated difference in objective response rate per central review between the atezolizumab-cabozantinib group and the cabozantinib group was −0·4 (95% CI, −9·3 to 8·5).
The estimated difference in objective response rate per investigator assessment between the atezolizumab-cabozantinib group and the cabozantinib group was −3·7(−12·5 to 5·1).
Included are patients with complete or partial response who did not experience disease progression or death.
CI denotes confidence interval, NE not evaluable, RECIST Response Evaluation Criteria in Solid Tumours, version 1·1. The plus sign indicates a censored value.