Duration of First-Line Chemotherapy in Advanced Non–Small-Cell Lung Cancer: Less Is More in the Era of Effective Subsequent Therapies

Abstract

The treatment of advanced, metastatic stage IIIB/IV non– small-cell lung cancer (NSCLC) has come a long way. Fifteen to 20 years ago, debate existed as to whether these stages should be treated at all. In 1995, the NSCLC Collaborative Group published a meta-analysis based on individual patient data, exploring the survival benefit of cisplatin-based chemotherapy (CBC) in this disease. The trials evaluated in this meta-analysis used agents typically referred to as firstor second-generation agents, and included drugs such as mitomycin, etoposide, and the vinca alkaloids; most of these are not commonly used in the modern day treatment of advanced NSCLC. In advanced disease, CBC improved survival compared with best supportive care (BSC). The absolute survival benefit was modest (2-month improvement in median survival time; 4 months for BSC and 6 months for CBC) and a 10% improvement in the 1-year survival rate (10% for BSC and 20% for CBC). This benefit of CBC was associated with a hazard ratio of 0.73, translating to a 27% risk reduction in death over the course of the disease. In 1997, the American Society of Clinical Oncology (ASCO) endorsed treatment of patients with stages IIIB/IV NSCLC who had a good performance status (PS) at the time of diagnosis. Around the turn of the century, a plethora of phase III trials evaluated the new third-generation agents (paclitaxel, docetaxel, vinorelbine, irinotecan, and gemcitabine) in combination with either a platinum, or in novel nonplatinum-based doublets. These new drugs modestly improved the therapeutic index of therapy, but no combination seemed superior. Two-drug combinations seemed to optimize the balance between the survival and palliative benefits of treatment and the risk of toxicity commonly seen with platinum-based regimens. In 1999, the first demonstration that secondor third-line treatment could improve survival and palliate symptoms lead to the approval of docetaxel for second-line treatment. Since then, three additional agents (gefitinib, pemetrexed, and erlotinib) have been approved for the treatment of refractory disease. Recently, two important biologic pathways (the epidermal growth factor receptor [EGFR] and vascular endothelial growth factor [VEGF] pathways) have recently been validated as therapeutic targets in advanced NSCLC. Lastly, the recent discovery of EGFR activating mutations has created optimism that continued insights into the biology of lung cancer will lead to further therapeutic gains. Although we have much work ahead, it is important to recognize that substantial progress has been made in the therapeutic options currently available for patients with advanced NSCLC. Despite these advances, the overall survival profile remains poor for most patients, and platinum-based doublets remain the cornerstone of treatment. Although these doublets improve survival, they are not curative options and are associated with substantial toxicity, much of which is cumulative. The goal of therapy in this setting is to maximize the impact platinum-based therapy has on survival, and the palliation of disease-related symptoms without undue toxicity. The original ASCO guidelines addressed the appropriate duration of therapy in this setting and cited a single trial that randomly assigned 74 patients to two to three cycles of a non–platinum-based combination regimen compared with continuous treatment. This trial did not show a significant difference in survival. Based on this trial, ASCO recommended that no more than eight cycles be delivered to patients with stages IIIB/IV NSCLC. However, in most clinical trials evaluating various strategies in this population, the median number of cycles delivered was typically three to four, suggesting that either the disease was refractory to treatment or the patient could not tolerate the toxicities of treatment beyond three to four cycles. Likewise, when reported, objective responses and palliation of symptoms typically occurred within the first two to three cycles. With this in mind, Smith and colleagues were the first to address the question of the duration of therapy in advanced NSCLC by random assignment of 308 patients to three, compared with six cycles of mitomycin, vinblastine, and cisplatin. They could not demonstrate an advantage from six cycles, as the response rate, time to disease progression, and overall survival were identical between the two arms. A second trial randomly assigned 230 patients with stage IIIB/IV disease to four cycles of carboplatin plus paclitaxel (CbP), compared with treatment until progression. No benefit was seen for those patients treated beyond four cycles. In a subset analysis, those patients who could have received more cycles but stopped at four fared as well as those patients continuing on CbP until progression. A third trial randomly assigned 297 patients with advanced NSCLC to either three or six cycles of carboplatin plus vinorelbine; no response, survival, or quality-of-life differences were demonstrated. A consistent theme in all these trials was the increasing risk of cumulative JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 25 NUMBER 33 NOVEMBER 2