Table 3.
Association of pemetrexed treatment and efficacy by ASBI and PS subgroups
| LSB (ASBI < 25) |
HSB (ASBI ≥ 25) |
|||
|---|---|---|---|---|
| Pemetrexed + BSC (N = 222)a | Placebo + BSC (N = 111)a | Pemetrexed + BSC (N = 92)a | Placebo + BSC (N = 39)a | |
| PFS (95% CI, months) | ||||
| Median | 5.1 (4.4, 6.2) | 2.4 (1.5, 2.9) | 3.7 (2.8, 5.4) | 2.8 (1.5, 3.8) |
| HR (pemetrexed versus placebo) | 0.50 (0.39, 0.63), P < 0.0001 | 0.54 (0.35, 0.82), P = 0.0036 | ||
| Adjusted HR (pemetrexed versus placebo)b | 0.49 (0.38, 0.62), P < 0.0001 | 0.50 (0.32, 0.80), P = 0.0033 | ||
| Interaction effect between treatment and ASBI subgroup | P = 0.6699 | |||
| OS (95% CI, months) | ||||
| Median | 17.5 (14.0, 20.6) | 11.0 (8.1, 14.3) | 11.8 (8.6, 14.1) | 10.6 (6.8, 13.4) |
| HR (pemetrexed versus placebo) | 0.64 (0.49, 0.85), P = 0.0019 | 0.91 (0.60, 1.39), P = 0.6729 | ||
| Adjusted HR (pemetrexed versus placebo)b | 0.63 (0.47, 0.83), P = 0.0012 | 1.02 (0.65, 1.60), P = 0.9188 | ||
| Interaction effect between treatment and ASBI subgroup | P = 0.1499 | |||
| PS 0 (n = 133)a | PS 0 (n = 60)a | PS 1 (n = 190)a | PS 1 (n = 96)a | |
| PFS (95% CI, months) | ||||
| Median | 5.5 (4.1, 6.9) | 1.7 (1.4, 2.9) | 4.4 (4.2, 5.4) | 2.8 (1.7, 2.9) |
| HR (pemetrexed versus placebo) | 0.40 (0.29, 0.56), P < 0.0001 | 0.58 (0.45, 0.76), P < 0.0001 | ||
| Adjusted HR (pemetrexed versus placebo)c | 0.36 (0.25, 0.52), P < 0.0001 | 0.60 (0.45, 0.78), P = 0.0002 | ||
| Interaction effect between treatment and PS subgroup | P = 0.0900 | |||
| OS (95% CI, months) | ||||
| Median | 17.7 (12.5, 22.4) | 10.3 (8.5, 14.4) | 14.1 (12.1, 17.3) | 10.6 (7.0, 13.2) |
| HR (pemetrexed versus placebo) | 0.57 (0.39, 0.82), P = 0.0027 | 0.80 (0.60, 1.06), P = 0.1200 | ||
| Adjusted HR (pemetrexed versus placebo)c | 0.54 (0.37, 0.80), P = 0.0019 | 0.78 (0.58, 1.05), P = 0.1045 | ||
| Interaction effect between treatment and PS subgroup | P = 0.2154 | |||
Bold P-values are significant. Treatment effect within each subgroup was significant at P < 0.05; the interaction effect between treatment and subgroup was significant at P < 0.2.
aOf the 481 non-squamous patients, 17 were missing LCSS assessments and two were missing ECOG PS.
bAdjusted for age group (aged <65 versus ≥65 years), gender (female versus male), race (Caucasian versus other), smoking status (no versus yes), stage (IIIB versus IV), previously treated brain metastases (yes versus no), platinum component of induction (cisplatin versus carboplatin), best response to prior chemotherapy (CR/PR versus SD), and ECOG performance status (0 versus 1).
cAdjusted for age group (aged <65 versus ≥65 years), gender (female versus male), race (Caucasian versus other), smoking status (no versus yes), stage (IIIB versus IV), previously treated brain metastases (yes versus no), platinum component of induction (cisplatin versus carboplatin), and best response to prior chemotherapy (CR/PR versus SD).
ASBI, average symptom burden index; PS, performance status; BSC, best supportive care; LSB, low symptom burden; HSB, high symptom burden; PFS, progression-free survival; CI, confidence interval; HR, hazard ratio; OS, overall survival.