First-Line Treatment for Advanced Non–Small-Cell Lung Cancer

With best supportive care alone, patients with metastatic non-small-cell lung cancer (NSCLC) have a median survival of 4 to 5 months and 1-year survival rates of approximately 10%.[1] Metastatic NSCLC: Chemotherapy Trials carried out in the 1980s and 1990s comparing chemotherapy to best supportive care reported variable efficacy results; however, a pivotal meta-analysis of these data indicated that cisplatin(Drug information on cisplatin)-based chemotherapy provided a survival benefit in advanced NSCLC.[2] The meta-analysis included eight trials involving 778 patients who received cisplatin-based chemotherapy vs best supportive care. The hazard ratio for death was 0.73 in favor of chemotherapy (P < .0001). This equated to a 1.5-month increase in median survival and a 10% increase in survival at 1 year. Chemotherapy for patients with advanced disease has also demonstrated that it can improve a patient's quality of life and be costeffective.[ 3,4] In the past decade newer agents such as gemcitabine(Drug information on gemcitabine) (Gemzar), vinorelbine, paclitaxel(Drug information on paclitaxel), and docetaxel(Drug information on docetaxel)(Taxotere) have all demonstrated activity in NSCLC as single agents; consequently, these agents have been combined with cisplatin or carboplatin(Drug information on carboplatin). Randomized phase III trials have endeavored to identify an optimal platinum- based doublet therapy regimen. Phase III Trials
The largest of these trials are summarized in Table 1. For example, Southwest Oncology Group (SWOG) investigators compared the reference regimen of vinorelbine/cisplatin to paclitaxel/carboplatin in SWOG9509.[5] With over 200 patients per arm, this trial showed no difference in quality of life, objective response rates, median survival, or 1-year survival between the arms. However, grade 3/4 hematologic toxicities were significantly higher on the vinorelbine/ cisplatin arm and more of these patients discontinued therapy because of toxicity (P =.001). The Eastern Cooperative Oncology Group (ECOG) undertook a similar trial (E1594) in which 1,207 chemotherapy-naive patients with advanced NSCLC were randomized to receive cisplatin/paclitaxel as the reference standard (based on earlier work from E5592) or one of three experimental arms: cisplatin/gemcitabine, cisplatin/docetaxel, or carboplatin/paclitaxel.[ 6,7] The gemcitabine/cisplatin arm was associated with a statisticallysignificant longer time to progression but a higher rate of grade 3/4 thrombocytopenia, although this was not clinically significant. Overall response rates and survival did not differ significantly between the reference regimen of cisplatin/paclitaxel and the three investigational arms. Based on toxicity differences, both the SWOG and ECOG chose carboplatin/paclitaxel as their reference regimen for future studies. The European Organisation for Research and Treatment of Cancer (EORTC) compared cisplatin/paclitaxel to gemcitabine/cisplatin or gemcitabine/paclitaxel. Although not statistically significant, the non- platinum-containing regimen gemcitabine/ paclitaxel appeared to be somewhat less efficacious. Overall, there were no significant differences in responserates, time to progression, or survival between the three regimens.[ 8] Similarly, the Italian Lung Cancer Project compared two new agent- platinum regimens (cisplatin/gemcitabine and carboplatin/paclitaxel) with a reference regimen of cisplatin/vinorelbine in 612 chemotherapy-naive patients.[9] Patients on the cisplatin/ vinorelbine arm had the highest percentages of dose reductions and omissions, mainly because of hematologic toxicities, while peripheral neuropathy and alopecia were significantly more common with carboplatin/paclitaxel. The results again revealed no significant difference in response rates, time to progression, or overall survival. The largest trial to date is TAX- 326.[10] In this multicenter randomized phase III study involving 1,220 patients, cisplatin/docetaxel and carboplatin/ docetaxel were compared with cisplatin/vinorelbine in chemotherapy- naive patients with advanced and metastatic NSCLC. The incidence of grade 3/4 neutropenia was substantial (74% to 79%), but the incidence of febrile neutropenia was low (4%) across all arms of the study. The cisplatin/docetaxel arm had a significantly higher response rate (32% vs 25%) and median survival (11.3 vs 10.1 months, P = .044) compared with cisplatin/vinorelbine. The 1- and 2-year survival rates were significantly different at 47% and 21% for docetaxel/cisplatin compared to 41% and 14% for cisplatin/vinorelbine. The carboplatin/docetaxel arm performed similarly to the control arm of cisplatin/vinorelbine, suggesting that cisplatin may provide a slight response and survival advantage over carboplatin in this disease; however, this trial was not designed to compare the two docetaxel arms. On the basis of these data, the combination of cisplatin/ docetaxel has recently been added to the list of platinum-based doublets approved by the US Food and Drug Administration for first-line treatment of advanced NSCLC. Although there are minor differences in response rates and survival times between these five trials, it should be noted that the ratio of stageIII to stage IV disease differed in many of them. For example, in the TAX- 326 study only about two-thirds (67%) of the patients in TAX-326 had stage IV disease, compared with 80% to 90% in the other large trials. As patients with stage IIIB disease tend to live longer, this difference potentially introduces some bias and makes it difficult to compare across trials. In summary, numerous studies have been conducted comparing the platinumbased doublets amongst themselves and no clinically significant differences have been identified. Other Aspects of the Chemotherapy Approach
Other approaches to improving chemotherapy in metastatic NSCLC have included the manipulation of chemotherapy duration and the addition of a third chemotherapy agent. In general, regimens containing three chemotherapy agents have failed to demonstrate any superiority over two-drug regimens in advanced NSCLC.[11,12] Typically three-drug combinations cause increased toxicity and financial cost without an improvement in survival. Triplet therapy cannot be recommended outside the confines of a clinical trial.[13] The duration of first-line chemotherapy in patients with advanced disease has been somewhat controversial and recommendations have ranged from 3 to 12 months of chemotherapy. Prospective trials investigating this issue include a phase III trial in which 230 patients with stage IIIB/IV NSCLC were randomized to receive carboplatin and paclitaxel every 21 days for four cycles or continuous treatment until progression.[14] At progression, patients on both arms were to receive weekly paclitaxel. The majority of patients (57%) were able to receive the four planned cycles ofchemotherapy, and on the continuous treatment arm the median number of cycles delivered was also 4 (42% received ≥ 5 cycles and 18% received ≥ 8 cycles). Objective response rates did not differ between arms (22% vs 24%, P = .80). Importantly, all responses in both arms occurred within the first four cycles of treatment. This trial and others like it, failed to show a benefit for continuing therapy in advanced NSCLC beyond four cycles.[ 14,15] In addition to response and survival rates, the palliative benefits of chemotherapy appear to plateau after the first few cycles.[16,17] For example, Ellis and colleagues administered mitomycin(Drug information on mitomycin)/ vinblastine(Drug information on vinblastine)/cisplatin to 110 symptomatic patients with advanced or metastatic NSCLC.[17] A total of 69% of patients had complete or significant resolution of symptoms, with 61% of those achieving symptomatic improvement after one course of therapy and 96% experiencing this benefit after two courses. In conclusion, in the past decade it has become clear that doublet platinum- based regimens are the standard of care for most patients with metastatic NSCLC. Doublet regimens are superior to single agents with respect to efficacy as well as economic and quality-of-life data.[18-20] Randomized phase II studies have demonstrated that all of the new agent-platinum doublet combinations appear roughly equivalent with regard to objective responses rate and survival. Subtle differences between them allow a clinician some flexibility to choose among toxicity profiles, convenience, and cost. Metastatic NSCLC: Molecularly Targeted Therapy With the use of chemotherapy, median survivals of 8 to 10 months and 1-year survival rates of 30% to 40% are routinely reported for large cooperative group studies (Table 1). Though it is clear that chemotherapy is an appropriate treatment for many patients with lung cancer, there is a sense that the use of traditional chemotherapeutic agents has reached a therapeutic plateau. Increased understandingof cancer biology has revealed numerous potential therapeutic strategies. EGFR Inhibitors
One such approach is to interfere with the epidermal growth factor receptor (EGFR) on the cell surface in order to disrupt intracellular signaling. EGFR is overexpressed in a variety of solid tumors and is particularly common in NSCLC, with 40% to 80% of tumors expressing EGFR.[21-24] When stimulated, the transmembrane receptors trigger a cascade of intracellular signaling that affects cellular proliferation and apoptosis. Strategies to target the EGFR receptor include the use of tyrosine kinase inhibitors, monoclonal antibodies, antisense oligonucleotides, and ligand-toxins or immunoconjugates. There are many compounds at various stages of clinical testing; however, gefitinib(Drug information on gefitinib) (Iressa) and erlotinib (Tarceva) are furthest along in clinical trials. These agents will be discussed in greater detail in this issue in the paperby Dr. Buter and Dr. Giaccone, but suffice it to say that both of these agents have shown activity in certain subsets of patients with previously treated NSCLC, with erlotinib currently the only EGFR-TKI that has been shown to prolong survival when compared to placebo in patients with one or two prior chemotherapies.[25] Largely because of the tolerability of these drugs in phase I combination trials and the promise demonstrated in the single-agent phase II studies, randomized phase III trials of chemotherapy with or without EGFR tyrosine kinase inhibitors (TKIs) were initiated. The INTACT (Iressa in NSCLC Trial Assessing CombinationTherapy) 1 and 2 trials were large randomized phase III studies in chemotherapy- naive patients with advanced stage III or IV NSCLC.[26,27] The trials were similar in design and both randomized patients to receive chemotherapy with gefitinib or placebo (INTACT 1 used cisplatin/gemcitabine and INTACT 2 used carboplatin/ paclitaxel). Together, INTACT 1 and 2 enrolled over 2,000 patients, twothirds of whom received gefitinib in addition to standard chemotherapy. Unfortunately, the results of these trials did not confirm that the addition of gefitinib resulted in an improvement in overall survival. The results of a parallel set of trials conducted using the small molecule EGFR tyrosine kinase inhibitor erlotinib instead of gefitinib demonstrated similar results. The TRIBUTE (carboplatin/paclitaxel with or without erlotinib) and TALENT (cisplatin/ gemcitabine with or without erlotinib) studies together enrolledover 2,200 patients and again failed to demonstrate any difference in survival associated with the addition of erlotinib to chemotherapy.[28,29] The reason for this apparent lack of benefit remains undefined. Some experts have proposed that despite sound preclinical data, the concurrent use of chemotherapy and the EGFR TKIs might have been antagonistic, and have suggested that a sequential use of these drugs might be more successful. Others feel that it is related to our inability to discriminate between subgroups of patients who are biologically more likely to respond to TKIs, and that the true benefit was diluted in an unselected population.[30] Prediction of Response
There is an accumulation of evidence suggesting that there may be a subgroup of patients with NSCLC who do respond preferentially to the EGFR TKIs. It was noted from the phase I trials and anecdotal evidence that women, nonsmokers, and patients with adenocarcinoma and bronchioloalveolar carcinoma (BAC) seem to respond better to the TKIs. Important evidence to support this observation has come from single-institution retrospective analyses of patients treated with single-agent TKIs. For example, a multivariate analysis identified the presence of any BAC features and being a never-smoker was significantly correlated with response to gefitinib in a retrospective review of 139 patients treated at Memorial-Sloan- Kettering Cancer Center.[31] Because of these observations, two prospective phase II trials were undertaken specifically in patients with BAC or adenocarcinoma with BAC features. In a SWOG study 138 such patients received single-agent gefitinib (102 chemo-naive, 36 previously treated).[32] Using standard criteria to assess tumor response the chemonaive group had a response rate of 21% whereas the previously treated group had a response rate of 10%. The rates of stable disease were 30% and 36%, and the median survival was 12 and 10 months for the previously untreated and treated groups, respectively. In subgroup analyses female gender, never-smokers, and thedevelopment of a rash from the gefitinib all predicted for improved survival. A similar study of erlotinib in 69 patients with BAC treated with erlotinib also found that smoking status and the presence of a rash predicted for response.[33] Such trials have raised the question of whether patients who fit these specific criteria should be considered for first-line therapy with an EGFR TKI, but this issue is highly controversial. Another focus of intensive research is the use of molecular biology to predict responders to targeted therapy such as the EGFR TKIs. Exciting reports from two institutions have identified mutations in the tyrosine binding domain of the EGFR that appear to correlate with patients who have dramatic responses to gefitinib.[34,35] These studies had a limited number of patients but the results are intriguing and demand serious attention. A massive worldwide effort is under way to replicate and expand on these results and to better define EGFR gene mutations. Other Promising Agents
Despite some controversial results molecularly targeted therapy is an important paradigm shift and represents the future of lung cancer treatment. There are many other agents that have demonstrated great potential activity in phase II studies that are currently being tested in the phase III setting. The rexinoid bexarotene (Targretin) selectively activates RXR receptors which modulate functions associated with differentiation, inhibition of cell growth, apoptosis, and metastasis. Two randomized phase III trials were performed to determine the survival benefit of adding bexarotene to either carboplatin/paclitaxel or cisplatin/ vinorelbine chemotherapy in patients with chemotherapy-naive advanced or metastatic NSCLC. In each of these trials over 600 patients were randomized to chemotherapy consisting of either carboplatin at an AUC of 6 and paclitaxel 200 mg/m² IV every 3 weeks or cisplatin (100 mg/m² every 4 weeks) and vinorelbine (25 mg/m² every week) IV with or without bexarotene (400 mg/m²/d). There wereno differences with respect to the primary end point, overall survival in either study (P = .19). An initial trend analysis suggests a relationship between bexarotene dose intensity and biomarker response (triglyceride elevation) with survival is being further evaluated in parallel with other risk factor analyses to better identify determinants of benefit and risk to bexarotene in a firstline setting.[36,37] Bevacizumab(Drug information on bevacizumab) (Avastin) is a recombinant humanized monoclonal antibody to the vascular endothelial growth factor (VEGF) and has shown great promise in phase II NSCLC trials and in phase III studies in colorectal cancer.[38,39] One concern in the phase II NSCLC trial was a high incidence of life-threatening pulmonary hemorrhage in patients with squamous cell histology. The ECOG reported on the results of their large randomized phase III study of carboplatin/paclitaxel with or without bevacizumab in patients with advanced nonsquamous NSCLC.[40] The primary end point of this randomized, multicenter trial was to compare the effects of the addition of bevacizumab to paclitaxel/carboplatin (PC) on overall survival in patients with previously untreated nonsquamous NSCLC. Secondary end points included response rate, time to progression, and tolerability. Eligibility criteria included ECOG performance status 0 or 1 and adequate hematologic, renal, and hepatic function. Patients with brain metastases and squamous cell histology were excluded. Patients were randomized to receive paclitaxel at 200 mg/m² plus carboplatin at an area under the concentration- time curve of 6 on day 1 every 3 weeks or PC plus bevacizumab at 15 mg/kg day 1 (PCB) every 3 weeks (Figure 1). Patients on PCB continued single-agent bevacizumab after six cycles until progressive disease or intolerable toxicity. The study opened in July 2001, was suspended for a planned toxicity evaluation from February 2002 through August 2002, and accrual was completed in April 2004 with 444 patients assigned to PC and 434 to PCB. The results presented were based on a second planned interim analysisthat was conducted with 469 of a planned 650 deaths (72.2%) for full analysis. The response rate (10% vs 27%, P < .0001, Fisher's exact test), progression-free survival (4.5 vs 6.4 months, P < .0001 [two-sided]), and median survival (10.2 vs 12.5 months, P = .0075, two-sided Wald test) all favored the PCB arm. Both regimens were well tolerated with selected toxicities as follows (PC vs PCB): grade 4/5 neutropenia (16.4% vs 24%) and thrombocytopenia (0% vs 1.6%), grade 3/4 thrombosis/embolism (3% vs 3.8%), grade 3-5 neuropathy (11.9% vs 9.5%) and hemorrhage (1.0% vs 4.1%). There were 10 treatment- related deaths (PC arm: 2; PCB arm: 8); 5 were due to hemoptysis, all on the PCB arm .[40] Summary Advances in chemotherapy have improved progression-free and median survival for patients with advanced NSCLC from 4 to 6 months with best supportive care to 8 to 10 months. Biologically targeted therapies have demonstrated activity and excellent tolerability in early trials, and the optimal way to integrate these exciting new agents is an area of active investigation.