Table 3.
a Concordance matrix for the agreement of the two expert pathologists for PD-L1 status in the MA31 cohort, at patient level. b: Re-classification of a second pathologist for PD-L1 and PD-1 status in the BCCA and MA31 cohorts
| a | ||||
|---|---|---|---|---|
| Patients classified by pathologist #1\#2 as | Negative | Positive | Total | Agreement |
| Negative | 246 | 6 | 252 | 97.60% |
| Positive | 4 | 19 | 23 | 82.60% |
| Total | 250 | 25 | 275 | |
| Agreement | 98.40% | 76% | ||
| b | ||||||
|---|---|---|---|---|---|---|
| Cohort | Classified positive by pathologist #1 | Re-classification by pathologist #2 as | Pathologists’ agreement | |||
| Negative | Positive | No tumor | ||||
| Classified negative by the system | BCCA-test (PD-L1) | 8 | 3 | 4 | 1 | 50% |
| MA31 (PD-L1) | 2 | 2 | 0 | 0 | 0% | |
| BCCA-test (PD-1) | 2 | 2 | 0 | 0 | 0% | |
Re-classification was done only for cases classified as negative by the system but positive by the first pathologist.