Current treatment approaches rely on pathologic stage as the major determinant in the decision to recommend adjuvant chemotherapy for patients with early-stage non–small-cell lung cancer (NSCLC). A number of randomized clinical trials have established a definitive role for adjuvant chemotherapy in patients with stage II or IIIA NSCLC.[1-3] Although many of these studies have included stage I patients, subset analyses have demonstrated a lack of survival benefit in this subgroup. The article by Calhoun et al published in this issue of ONCOLOGY addresses the heterogeneity in patient subsets within stage IB disease and how that might have affected outcomes in clinical trials of adjuvant chemotherapy.
The authors correctly point out the differences among patients with stage IB disease. For example, they note that the impact of tumor size on patient outcome with chemotherapy might be different from that of visceral pleural invasion, even if both situations are considered stage IB disease according to the current staging system. The new staging system proposed by the International Association for the Study of Lung Cancer (IASLC) will lead to more accurate prognostic information based on tumor size, extent, and pattern of involvement.[4] This will also facilitate proper stratification of patients in future studies of early-stage NSCLC.
Clinical Trials
Only two randomized clinical trials have been conducted exclusively in patients with stage I NSCLC.[5,6] In the Japanese study, the use of uracil/tegafur (UFT) was superior to placebo in patients with an adenocarcinoma histology.[5] Even among patients with stage I NSCLC, a survival benefit was seen only for those with stage IB disease. In contrast, the Cancer and Leukemia Group B (CALGB) study conducted to evaluate the role of the carboplatin(Drug information on carboplatin)/paclitaxel combination in stage IB disease failed to demonstrate a survival advantage, although the median disease-free survival was improved with chemotherapy.[6]
One of the notable differences between these two studies is in sample size, which was much larger for the Japanese study (N = 999) than the CALGB trial (N = 344). The CALGB study was closed early after a planned interim analysis demonstrated a P value for overall survival that was less than a prespecified stopping boundary. Could the lack of survival difference in CALGB be merely a result of inadequate sample size? Or is this proof of a true lack of benefit with chemotherapy in this setting?
The risk of recurrence increases with stage, and it is evident from the Lung Adjuvant Cisplatin(Drug information on cisplatin) Evaluation (LACE) meta-analysis that the magnitude of benefit with adjuvant chemotherapy is higher in patients with greater risk of recurrence following surgery for early-stage NSCLC.[7] Therefore, the benefit of adjuvant chemotherapy, if any, would be smallest for stage I among all subsets of early-stage NSCLC. In order to demonstrate a modest survival benefit in this subgroup, a study with a much larger sample size would be required. The CALGB study was powered to detect a hazard ratio (HR) of 0.67 with adjuvant chemotherapy, but only noted a nonsignificant HR of 0.80, with a wide confidence interval. Similar findings were noted in the other phase III studies, which showed a nonsignificant trend toward improved disease-free and overall survival with chemotherapy in stage IB disease, although the differences did not reach statistical significance.[1,8]
All of these findings support the concept that adjuvant therapy is beneficial to patients with stage IB disease, although much less so than in those with stage II or IIIA disease. That being the case, is it necessary to conduct an adequately powered large clinical trial exclusively for stage I NSCLC to establish the role of adjuvant chemotherapy? Given that it would take a few years to complete such a study, would the results be clinically relevant by the time it is done?
As Calhoun and colleagues report, a number of novel risk-prediction methods are under evaluation.[9] Utilization of genomic arrays to identify patients at high risk of recurrence and, thereby, to select adjuvant chemotherapy, or individualization of chemotherapy based on the molecular characterization of the tumor[10] are more likely to yield influential information and will be a major step forward in improving outcomes for patients with early-stage NSCLC. In addition, it is hoped that the incorporation of molecularly targeted agents, currently under investigation in clinical trials, will improve the cure rate in this setting.
Recommendations
What should current practice be for patients with resected stage IB disease outside the setting of a clinical trial? The subset analyses from the CALGB and National Cancer Insitute of Canada’s JBR.10 studies support the use of adjuvant chemotherapy in patients with a primary tumor larger than 4 cm. This could form the basis for an informed discussion with individual patients regarding the risks associated with chemotherapy as well as the potential benefits.
Furthermore, the use of adjuvant chemotherapy in stage IB disease should be restricted to patients with minimal comorbid illness and a good performance status (Eastern Cooperative Oncology Group [ECOG] 0 or 1). Patients should also be eligible to receive cisplatin-based regimens, since these are preferred over carboplatin-based regimens in the curative setting for resected NSCLC. Patients with resected stage IA disease should be considered for enrollment in the ongoing chemoprevention clinical trial of selenium(Drug information on selenium) (ECOG 5597) or other appropriate studies. In conclusion, recent progress in the treatment of early-stage NSCLC provides hope for further improvements in the near future.
—Suresh S. Ramalingam, MD
—Seth Force, MD
—Fadlo R. Khuri, MD
