Table 1.
Cohort characteristics at study entry
| PRIMM–UK (n = 54) | PRIMM–NL (n = 74) | Manchester (n = 17) | Leeds (n = 19) | Barcelona (n = 11) | All cohorts (n = 175) | P value | |
|---|---|---|---|---|---|---|---|
| Age (years), median (range) | 64 (19–94) | 60 (21–85) | 66 (38–87) | 57 (35–88) | 64 (37–88) | 63 (19–94) | 0.127 |
| Sex (female), n (%) | 19 (35) | 37 (50) | 7 (41) | 7 (37) | 5 (45) | 75 (43) | 0.530 |
| BMI (kg m−2), mean (range) | 28.6 (18.83–47.66) | 27.02 (15.43–40.74) | 26.92 (18.99–40.40) | 28.46 (20.90–38.57) | 26.27 (20.96–36.08) | 27.63 (15.43–47.66) | 0.075 |
| Metastatic stage, n (%) | 0.006 | ||||||
| Stage 3 unresectable | 5 (9) | 2 (3) | 0 (0) | 1 (5) | 0 (0) | 8 (5) | |
| Stage 4 M1a | 12 (22) | 7 (9) | 5 (29) | 3 (16) | 4 (36) | 31 (18) | |
| Stage 4 M1b | 12 (22) | 14 (19) | 2 (12) | 5 (26) | 5 (45) | 38 (22) | |
| Stage 4 M1c | 20 (37) | 23 (31) | 8 (47) | 7 (37) | 2 (18) | 60 (34) | |
| Stage 4 M1d | 5 (9) | 28 (38) | 2 (12) | 3 (16) | 0 (0) | 38 (22) | |
| BRAF mutant, n (%) | 18 (33) | 42 (57) | 2 (12) | 9 (47) | 3 (27) | 74 (42) | 0.004 |
| ECOG performance status ≥1, n (%) | 36 (67) | 17 (23) | 8 (47) | 2 (11) | 1 (9) | 64 (37) | 1.729 × 10−6* |
| Outcomes following ICB | |||||||
| PFS ≥12 months, n (%) | 27 (50) | 32 (43) | 8 (47) | 11 (58) | 5 (45) | 83 (47) | 0.824 |
| irAEs, n (%) | 38 (70) | 44 (59) | 9 (53) | 9 (47) | 7 (64) | 107 (61) | 0.399 |
| Colitis, n (%) | 13 (24) | 10 (14) | 3 (18) | 4 (21) | 3 (27) | 33 (19) | 0.570 |
| Treatment details | |||||||
| ICB combination therapy (anti-CTLA-4/anti-PD-1), n (%) | 29 (54) | 15 (20) | 2 (12) | 11 (58) | 1 (9) | 58 (33) | 1.60 × 10−5* |
| Previous BRAF or MEK inhibition, n (%) | 10 (19) | 28 (38) | 2 (12) | 0 (0) | 1 (9) | 41 (23) | 0.001* |
| PPI use at baseline, n (%) | 13 (24) | 24 (32) | 4 (24) | 6 (32) | 1 (9) | 48 (27) | 0.495 |
| Antibiotics use at baseline, n (%) | 9 (17) | 11 (15) | 2 (12) | 3 (16) | 0 (0) | 25 (14) | 0.694 |
Baseline characteristics are presented as mean and s.d. or median (range) for continuous variables and as counts and percentages for categorical variables. χ2 tests for categorical variables and two-sided Wilcoxon tests for continuous data were performed to calculate differences between cohorts. P values written in bold indicate nominally significant differences between cohorts (P < 0.05). *Statistical significance under a false discovery rate of 5%. UK, United Kingdom; NL, the Netherlands.