Table 1 Clinicopathological features of all cohorts
From: Swarm learning for decentralized artificial intelligence in cancer histopathology
Variable | TCGA | DACHS | Epi700 | YCR BCIP | QUASAR |
|---|---|---|---|---|---|
Use in this study | Train | Train | Train | Test | Test |
Cohort type | Population | Population | Population | Population | Clinical trial |
No. of patients | 632 | 2,448 | 661 | 889 | 2,190 |
Median age (years) | 68 | 69 | 72.7 | 71 | 63 |
IQR for age (years) | 18 | 14 | 14.5 | 15 | 12 |
Male | 322 (50.9%) | 1,436 (58.7%) | 358 (54.2%) | 494 (55.6%) | 1,334 (60.9%) |
Female | 292 (46.2%) | 1,012 (41.3%) | 303 (45.8%) | 395 (44.4%) | 848 (38.7%) |
Unknown sex | 18 (2.85%) | 0 | 0 | 0 | 8 (0.4%) |
MSS/pMMR | 392 (62%) | 1,836 (75%) | 471 (71.3%) | 760 (85.5%) | 1,529 (69.8%) |
MSI/dMMR | 65 (10.3%) | 210 (8.6%) | 136 (20.6%) | 129 (14.5%) | 246 (11.2%) |
Unknown MSI status | 175 (27.7%) | 402 (16.4%) | 54 (8.1%) | 0 | 415 (19%) |
MSI/MMR method | PCR 5-plex | PCR 3-plex | PCR 5-plex | IHC | IHC |
Wild-type BRAF | 471 (74.5%) | 1,930 (78.8%) | 553 (83.7%) | 32 (3.6%a) | 1,358 (62%) |
Mutated BRAF | 63 (10%) | 151 (6.2%) | 92 (13.9%) | 75 (8.4%a) | 129 (5.9%) |
Unknown BRAF status | 98 (15.5%) | 367 (15%) | 16 (2.4%) | 782 (88%) | 916 (41.8%) |
BRAF detection method | Sequencing37 | ColoCarta[b,40 | NA | Pyrosequencing41 | |
Stage I | 76 (12%) | 485 (19.8%) | 0 | 169 (19%) | 5 (0.2%) |
Stage II | 166 (26.3%) | 801 (32.7%) | 394 (59.6%) | 317 (35.7%) | 53 (2.4%) |
Stage III | 140 (22.2%) | 822 (33.6%) | 267 (40.4%) | 370 (41.6%) | 1,653 (75.5%) |
Stage IV | 63 (10%) | 337 (13.8%) | 0 (0%) | 33 (3.7%) | 268 (12.2%) |
Stage unknown | 187 (29.5) | 3 (0.1%) | 0 | 0 | 211 (9.7%) |
Left-sided CRC | 248 (39.2%) | 1,607 (65.6%) | 280 (42.3%) | 487 (54.8%) | 1,158 (52.9%) |
Right-sided CRC | 176 (27.8%) | 819 (33.5%) | 375 (56.7%) | 332 (37.3%) | 754 (34.4%) |
Unknown side | 209 (33%) | 22 (0.9%) | 6 (1%) | 70 (7.9%) | 278 (12.7%) |
