Patient disposition is summarized in Figure 1. A total of 108 patients were screened for the study, of whom 1 did not meet the eligibility criteria and 1 was lost to follow-up. Therefore, 106 patients were enrolled and received at least 1 dose of induction therapy. Following the induction phase, a total of 80 (75.4%) patients had no disease progression documented, although 55 patients of them were considered protocol-qualified and were randomized to maintenance treatment with pemetrexed/BSC (n = 28) or BSC alone (n = 27). All of the patients in the pemetrexed/BSC arm received at least 1 dose of maintenance therapy. The most common reason for premature study discontinuation in both arms was death (pemetrexed/BSC, 60.7%; BSC, 55.6%), which was predominantly due to non-squamous NSCLC. Five (17.9%) patients in the pemetrexed/BSC arm and 10 (37.0%) patients in the BSC arm were alive and progression-free at the end of the study. The main reasons for treatment discontinuation in the pemetrexed/BSC and BSC arms, respectively were: progressive disease (64.3% vs 77.8%), patient decision (17.9% vs 7.4%), and death (14.3% vs 7.4%).
Baseline patient characteristics
Table 1 presents some key baseline characteristics for the randomized patients, before the start of induction therapy, and the patients’ best tumor response during the induction therapy phase. Baseline characteristics were generally similar between the two treatment arms except that there were slightly higher proportions of patients who were never smokers or had stage IV disease in the pemetrexed/BSC arm (pemetrexed/BSC arm, 42.9% and 67.9% vs BSC arm, 37.0% and 63.0%, respectively), and there was a higher proportion of females in the BSC arm (pemetrexed/BSC arm, 37.0% vs BSC arm, 28.6%). The best overall tumor responses to induction therapy were generally similar between the arms in terms of the numbers of responding and stable patients. The numbers of patients with ECOG PS scores of 0, 1 or 2 at the start of the maintenance phase were generally similar between the two maintenance arms (2 patients in each arm had an ECOG PS of 2; other data not reported here).
Study treatment and additional therapy
Of the 106 patients that initiated induction therapy, 79 (74.5%) patients completed 3 cycles of induction therapy and 67 (63.2%) patients completed 4 cycles of induction therapy. During induction therapy, there were very few dose reductions or delays, and the median dose intensities (defined as the actual dose/planned dose) were 94.0% for pemetrexed and 94.6% for cisplatin. During the induction phase, all 106 patients were compliant with the required vitamin B12 supplementation, and compliance with the required folic acid supplementation ranged from 89.3-100.0%.
Of the 28 patients in the pemetrexed/BSC maintenance arm, 18 (64.3%) received ≥4 cycles, 7 (25.0%) received ≥6 cycles, 4 (14.3%) received ≥12 cycles, 3 (10.7%) received 14 cycles, and 1 (3.6%) received 15 cycles. During the maintenance phase, the median number of cycles of pemetrexed given was 4.0 and the median dose intensity of pemetrexed was 95.3% (range: 75%-102%). There were no pemetrexed dose reductions required during the maintenance phase, but 3 (10.7%) patients required a pemetrexed dose delay for 1 cycle and 1 (3.6%) patient had 2 delayed cycles. During the maintenance phase, compliance was 85.7% for vitamin B12 supplementation and an average of 78.7% for folic acid supplementation. After the discontinuation of study therapy, a similar proportion of patients in each arm received at least one type of additional anticancer therapy (pemetrexed/BSC, 46.4%; BSC, 44.4%). The additional therapies were: chemotherapy (39.3% vs 37.0%) and/or radiotherapy (14.3% vs 14.8%).
Progression-free survival and overall survival
The median PFS time for the maintenance phase for both treatment arms was 3.2 months (pemetrexed/BSC arm, 95% CI: 2.9 to 6.1; BSC arm, 95% CI: 2.2 to 4.3). The PFS HR stratified by the best tumor response for induction therapy was 0.76 (two-sided 95% CI: 0.42 to 1.37; one-sided p-value = 0.1815). When the analysis model was also adjusted for the 3 prespecified patient characteristics, PFS remained longer for the pemetrexed/BSC arm than the BSC arm (maintenance phase: HR = 0.65, 95% CI: 0.35 to 1.20, one-sided p-value = 0.0846; overall study period: HR = 0.60, 95% CI: 0.33 to 1.09, one-sided p-value = 0.0461). The median OS time in the pemetrexed/BSC arm was 12.2 months (95% CI: 5.6 to 20.6) and 11.8 months (95% CI: 6.3 to 25.6) in the BSC arm (HR: 1.13, two-sided 95% CI: 0.56 to 2.28). Summarized PFS and OS data are shown in Table 2. The one-year survival probabilities were similar for the two treatment arms for the maintenance phase (pemetrexed/BSC, 54.4%; BSC, 49.5%) and the overall study period (pemetrexed/BSC, 58.6%; BSC, 57.8%). The unadjusted Kaplan-Meier curves for PFS and OS for each arm are shown in Figures 2 and 3, respectively.
Of the 106 patients who commenced induction therapy, 27 patients had a best overall response of CR (n = 1, 0.9%) or PR (n = 26, 24.5%), 49 (46.2%) patients had a best response of SD, and 4 (3.8%) patients were defined per protocol as unknown. Therefore, the disease control rate for the induction phase was 71.7% (76/106).
For the maintenance phase, there were no cases of a best response of CR or PR, thus the overall response rate was 0.0% for both arms. Sixteen (57.1%) patients in the pemetrexed/BSC arm and 12 (44.4%) patients in the BSC arm had a best response of SD.The disease control rate for the maintenance phase was not significantly different between the pemetrexed/BSC arm and the BSC arm (57.1% vs 44.4%, p = 0.3463). Nine (32.1%) patients in the pemetrexed/BSC arm and 10 (37.0%) patients in the BSC arm showed progressive disease. For the overall study period, the disease control rate was similar for both treatment arms (pemetrexed/BSC, 96.4%; BSC, 96.3%; p = 0.9791).
During the induction phase, 73 (68.9%) patients experienced at least 1 treatment-emergent adverse event (TEAE) and 65 (61.3%) patients experienced at least 1 drug-related TEAE. Approximately one-third (31.1%) of the patients experienced at least 1 grade 3/4 TEAE and 12.3% of the patients experienced at least 1 SAE. Twelve (11.3%) patients died during induction therapy. The most common TEAEs (≥10%) during the induction phase were: vomiting (34.0%), decreased appetite (24.5%), fatigue (23.6%), nausea (14.2%), anemia (12.3%), and neutropenia (10.4%). The observed toxicities were consistent with the known safety profile of the induction therapy regimen.
During the maintenance phase, 12 (42.9%) patients in the pemetrexed/BSC arm and 9 (33.3%) patients in the BSC arm experienced at least 1 TEAE. Five patients in each arm experienced at least 1 grade 3/4 TEAE (pemetrexed/BSC, 17.9%; BSC, 18.5%) and 1 patient in each arm experienced at least 1 SAE (pemetrexed/BSC, 3.6%; BSC, 3.7%). There were in total 5 patients (9.1%) with at least 1 grade 3/4 TEAE that were considered possibly drug-related, including 4 patients in pemetrexed/BSC arm (neutropenia, n = 2, 7.1%; anemia, n = 1, 3.6%; hypokalemia, n = 1, 3.6%) and 1 patient in BSC arm (anemia, n = 1, 3.7%). None of the SAEs were drug-related. The most common (≥5% in at least 1 arm) TEAEs during the maintenance phase are summarized in Table 3. The proportions of patients that experienced TEAEs during the maintenance phase were relatively low, and the observed toxicities for the pemetrexed/BSC arm were mostly mild to moderate (grade 1 or 2) and not unexpected for pemetrexed monotherapy. The occurrence of grade 3/4 TEAEs was relatively low in both arms and there were no significant between-arm differences in the proportion of patients experiencing any grade 3/4 TEAE. Three (10.7%) patients required red blood cell transfusions and 1 (3.6%) patient was hospitalized due to NSCLC. Four (14.3%) patients in the pemetrexed/BSC arm and 2 (7.4%) patients in the BSC arm died during the maintenance phase. In the pemetrexed/BSC arm, 3 (10.7%) deaths were due to non-squamous NSCLC and 1 (3.6%) death was due to an AE not related to the study drug (sudden death). In the BSC arm, 1 (3.7%) death was attributed to non-squamous NSCLC and 1 (3.7%) death was due to an AE (cardiac arrest).
For the overall study period, 75/106 (70.8%) patients who received at least 1 dose of study drug experienced at least 1 TEAE, including 20 (71.4%) patients in the pemetrexed/BSC arm and 18 (66.7%) patients in the BSC arm. The occurrence of drug-related grade 3/4 TEAEs was generally low. Eight (28.6%) patients in the pemetrexed/BSC arm, 12 (44.4%) patients in the BSC arm, and 38/106 (35.8%) patients from the entire study population experienced at least 1 grade 3/4 TEAE during the overall study period. As shown in Table 4, the most common TEAEs (≥15%) in the entire treated patient population (n = 106) for the overall study period were: vomiting (35.8%), fatigue (25.5%), decreased appetite (24.5%), and anemia (16.0%), and the most common grade 3/4 drug-related TEAEs in the entire treated patient population for the entire study period were: neutropenia (8.5%), anemia (5.7%), vomiting (1.9%), and fatigue (1.9%).
There were a total of 17 (60.7%) deaths in the pemetrexed/BSC arm and 15 (55.6%) deaths in the BSC arm for the overall study period, mostly due to non-squamous NSCLC (pemetrexed/BSC, n = 14 [50.0%]; BSC, n = 12 [44.4%]). One death in each arm was due to an AE not study-drug related and no death was reported due to an AE related to study drug. Overall, 5 (17.9%) patients in the pemetrexed/BSC arm and 3 (11.1%) patients in the BSC arm required red blood cell transfusions. During the overall study period, 1 (3.6%) patient in the pemetrexed/BSC arm was hospitalized due to non-squamous NSCLC and 1 (3.7%) patient in the BSC arm was hospitalized due to an AE.