Table 2.
The concordance and discordance rates between the TNM stage and MTVwb risk stratum in both the modeling and validation cohorts.
| Modeling cohort | MTVwb risk stratum | ||||
|---|---|---|---|---|---|
|
| |||||
| TNM stage (Substages) | I | II | III | IV | Total |
| I (IA/IB) | 116 (96/20) | 29 (4/25) | 0 (0/0) | 0 (0/0) | 145 (100/45) |
| II (IIA/IIB) | 11 (0/11) | 52 (29/23) | 28 (1/27) | 0 (0/0) | 91 (30/61) |
| III (IIIA/IIIB/IIIC) | 13 (11/2/0) | 39 (25/11/3) | 68 (20/35/13) | 56 (22/17/17) | 176 (78/65/33) |
| IV (IVA/IVB) | 10 (9/1) | 30 (22/8) | 54 (24/30) | 93 (37/56) | 187 (92/95) |
|
| |||||
| Total | 150 | 150 | 150 | 149 | 599 |
|
| |||||
| Validation cohort | MTVwb risk stratum | ||||
|
| |||||
| TNM stage (Substages) | I | II | III | IV | Total |
|
| |||||
| I (IA/IB) | 67 (59/8) | 15 (1/14) | 0 (0/0) | 0 (0/0) | 82 (60/22) |
| II (IIA/IIB) | 4 (0/4) | 24 (14/10) | 5 (0/5) | 0(0/0) | 33 (14/19) |
| III (IIIA/IIIB/IIIC) | 14 (11/3/0) | 40 (28/10/2) | 36 (12/15/9) | 25 (6/11/8) | 115 (57/39/19) |
| IV (IVA/IVB) | 6 (4/2) | 30 (18/12) | 30 (10/20) | 40 (13/27) | 106 (45/61) |
|
| |||||
| Total | 91 | 109 | 71 | 65 | 336 |
Note: The data is expressed as the number of patients in both of the TNM stage (rows) and MTVwb risk (columns) stratum, and in the parentheses are the number of patients in the TNM substage. Concordance between the TNM stage and the MTVwb risk stratum means the TNM stage and the MTVwb risk stratum are the same in the patients. For example, in TNM stage I of the modeling cohort, 116 patients (96 patients with stage IA and 20 patients with stage IB) had MTVwb risk stratum I, and 29 patients (4 patients with stage IA and 25 patients with stage IB) had MTVwb risk stratum II. Therefore in the TNM stage I, the TNM stage and MTVwb risk stratum were concordant in 116/145 patients (80.0%) and discordant in 29/145 patients (20%), which are equal to the concordance and discordance rates between the TNM substage and the MTVwb risk stratum as the TNM substages are included in their respective TNM stage. Of the 599 patients of modeling cohort, 270 patients (45.1 %) were discordant between TNM stage and MTVwb risk stratum, with a majority of these being within +/− one MTVwb risk stratum (36.2%). The discordance rate for deviations greater than +/− one MTVwb risk stratum in the modeling cohort was 8.8%. Of the 336 patients in the validation cohort, 167 patients (50.3%) were discordant with a majority of these being +/− one MTVwb risk stratum (35.4%). In the validation cohort, the discordance rate for deviations greater than +/− one MTVwb risk stratum was 14.9%.