Table 2.
Composition of the patient population in subgroups stratified by underlying PD-L1 expression and TMB level, and first-line treatment for each strategy
| PD-L1 | TMB | 1L Treatment | PFS-HRa | Percentage of each subgroup constituting the population per strategyb | |||||
|---|---|---|---|---|---|---|---|---|---|
| A: PD-L1 alone | B: TMB alone | C: PD-L1 and TMB | D: PD-L1 + DB | E: TMB + DB | F: PD-L1 and TMB + DB | ||||
| PD-L1 high | TMB high | Pembro | 0.375 | 8.75 | 8.75 | 8.75 | 6.39 | 6.39 | 6.39 |
| TMB high | Pembro+ChT | 0.270 | 0.00 | 0.00 | 0.00 | 2.36 | 2.36 | 2.36 | |
| TMB low | Pembro | 0.635 | 16.25 | 0.00 | 0.00 | 11.86 | 0.00 | 0.00 | |
| TMB low | Pembro+ChT | 0.457 | 0.00 | 16.25 | 16.25 | 4.39 | 16.25 | 16.25 | |
| PD-L1 intermediate | TMB high | Pembro | 0.750 | 0.00 | 10.15 | 0.00 | 0.00 | 7.41 | 0.00 |
| TMB high | Pembro+ChT | 0.413 | 10.15 | 0.00 | 10.15 | 10.15 | 2.74 | 10.15 | |
| TMB low | Pembro | 1.270 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | |
| TMB low | Pembro+ChT | 0.699 | 18.85 | 18.85 | 18.85 | 18.85 | 18.85 | 18.85 | |
| PD-L1 low | TMB high | Pembro | 1.266 | 0.00 | 16.10 | 0.00 | 0.00 | 11.75 | 0.00 |
| TMB high | Pembro+ChT | 0.563 | 16.10 | 0.00 | 16.10 | 16.10 | 4.35 | 16.10 | |
| TMB low | Pembro | 1.786 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | |
| TMB low | Pembro+ChT | 0.770 | 29.99 | 29.99 | 29.99 | 29.99 | 29.99 | 29.99 | |
| Total percentage of Pembro | 25.00 | 35.00 | 8.75 | 18.25 | 25.55 | 6.39 | |||
| Total percentage of Pembro+ChT | 75.00 | 65.00 | 91.25 | 81.75 | 74.45 | 93.61 | |||
1L first-line, ChT chemotherapy, DB disease burden, Pembro pembrolizumab, Pembro+ChT pembrolizumab with chemotherapy (here assumed ChT as cisplatin-pemetrexed), PD-L1 programmed death-ligand 1, PFS-HR progression-free hazard ratio of 1L treatment compared to ChT, TMB tumor mutational burden
aPFS-HR values were computed assuming independent predictive values of TMB and PD-L1 on treatment benefit [8, 18, 20] (details are given in Sect. 1.2 of the ESM)
bPercentage depends on the prevalence of the biomarker used in the treatment choice (i.e., PD-L1 and TMB) and the percentage of patients with high tumor/disease burden