2 Clarke Drive
Suite 100
Cranbury, NJ 08512
© 2024 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
Adjuvant treatment with the chemotherapy doublet of pemetrexed and cisplatin did not lead to a significant improvement in overall survival compared with vinorelbine plus cisplatin in patients with stage II to IIIA nonsquamous non–small cell lung cancer.
Adjuvant treatment with the chemotherapy doublet of pemetrexed and cisplatin did not lead to a significant improvement in overall survival (OS) compared with vinorelbine plus cisplatin in patients with stage II to IIIA nonsquamous non–small cell lung cancer (NSCLC), according to findings from the phase 3 JIPANG study (UMIN000006737) published in the Journal of Clinical Oncology.1
At a median follow-up of 77.3 months (95% CI, 74.2-79.2), the median OS in both the pemetrexed plus cisplatin arm (n = 389) and the vinorelbine plus cisplatin arm (n = 394) was not yet reached (NR; HR, 1.03; 95% CI, 0.80-1.32; 1-sided log-rank P = .589). The 3-year OS rates were 87.0% (95% CI, 83.2%-90.0%) vs 84.1% (95% CI, 80.0%-87.3%), respectively, and the 5-year OS rates were 75.0% (95% CI, 70.3%-79.0%) vs 75.6% (95% CI, 71.0%-79.6%), respectively. A total 123 patients in the pemetrexed plus cisplatin arm died compared with 119 patients in the vinorelbine plus cisplatin arm.
Additional results demonstrated that recurrence or death occurred in 57% of patients who received pemetrexed and 60% of patients treated with vinorelbine in the modified intent-to-treat population. The median updated recurrence-free survival (RFS) was 43.4 months vs 37.5 months, respectively (HR, 0.95; 95% CI, 0.79-1.14). The 3-year RFS rates did not differ significantly between the 2 arms, at 51.6% (95% CI, 46.5%-56.4%) vs 51.0% (95% CI, 49.4%-74.9%), respectively; the 5-year rates were 44.9% (95% CI, 39.8%-49.8%) vs 42.6% (95% CI, 37.7%-47.5%), respectively.
The restricted mean survival time (RMST) for RFS in the pemetrexed arm was comparable to that of the vinorelbine arm at 57.8 months vs 55.9 months, respectively (P = .56). The RMST for OS was 86.7 months vs 86.9 months, respectively (P = .92).
“To our knowledge, this JIPANG study is the first phase 3 study to compare cisplatin-based chemotherapy regimens as a postoperative adjuvant treatment among patients with completely resected nonsquamous NSCLC,” study authors wrote. “The JIPANG study found no significant difference in OS between the pemetrexed and cisplatin [arm] and the vinorelbine plus cisplatin [arm], even with longer follow-up. The prognosis of patients with completely resected NSCLC receiving cisplatin-based chemotherapy has improved over the past 2 decades, and the study's results provide important historical data for future research.”
JIPANG was an open-label Japanese study that enrolled patients who were 20 to 75 years old with nonsquamous NSCLC who received pathologically confirmed complete surgical resection where the level of surgical resection was higher than lobectomy.1 Eligible patients had stage IIA or IIIA disease, an ECOG performance status of 1 or less, and adequate hematologic and organ function.1 Patients with severe postoperative complications, interstitial pneumonia, and those who received neoadjuvant chemotherapy or planned to receive an EGFR TKI were excluded from the study.2
Eligible patients were randomly assigned 1:1 to be treated with 500 mg/m2 of pemetrexed and 75 mg/m2 of cisplatin via intravenous (IV) infusion on day 1 of each treatment cycle, or IV vinorelbine at a dose of 25 mg/m2 on days 1 and 8 and 80 mg/m2 of cisplatin on day 1. Each treatment cycle was repeated every 3 weeks until the completion of 4 cycles. Subsequent chemotherapy or postoperative radiotherapy was not allowed for patients who did not experience disease recurrence.
The study’s initial primary end point was OS. However, this was amended to RFS due to a lack of sufficient death events needed to evaluate survival at the end of patient accrual. Secondary end points included OS, rate of treatment completion, and toxicity.
The baseline patient characteristics were well balanced between the 2 arms; the median age was 64 years (range, 28-75) in the pemetrexed arm compared with 65 years (range, 33-75) in the vinorelbine arm. Most patients in both arms were male (58.4% vs 59.4%) and had an ECOG performance status of 0 (75.8% vs 77.4%), stage IIIA disease (52.4% vs 52.0%), adenocarcinoma (95.9% vs 95.9%), and EGFR wild-type disease (75.1% vs 75.9%). Nearly all patients in both arms underwent lobectomy (99.2% vs 97.7%) and the median time from surgery was 43 days (range, 22-58) in the pemetrexed arm vs 42 days (range, 22-58) in the vinorelbine arm.1
A subgroup analysis revealed that OS significantly favored the pemetrexed arm among patients with stage II disease (HR, 0.63; 95% CI, 0.41-0.95; P = .002) and those aged at least 70 years (HR, 0.66; 95% CI, 0.39-1.09; P = .055). OS in all other subgroups examined either favored the vinorelbine arm or did not significantly favor either arm.
Previously published safety findings showed that patients in the pemetrexed arm (n = 392) experienced any grade adverse effects (AEs) at a rate of 99.7% compared with 100% in the vinorelbine arm (n = 395); grade 3 to 5 AEs occurred at rates of 47.4% vs 89.4%, respectively. The most common any grade AEs in the pemetrexed arm included anorexia (82.1%), nausea (78.8%), and decreased neutrophil count (78.3%). In the vinorelbine arm, common any grade AEs included decreased neutrophil count (95.5%), decreased white blood cell count (94.7%), and anorexia (80.3%).2
After completing study treatment, 366 patients received chemotherapy, 229 received radiotherapy, and 42 underwent surgery.1
“One limitation of this Japanese study is that standard subsequent treatment for patients with recurrent/metastatic NSCLC was changed during this study, including ICIs,” the investigators concluded. “Therefore, various chemotherapy regimens were used for subsequent treatment. Despite this limitation, the JIPANG study provides important historical data and long-term follow-up data remain essential in adjuvant settings for NSCLC.”