Chemotherapy in Metastatic Non–Small-Cell Lung Cancer

Introduction

Lung cancer is the leading cause of cancer death. Approximately 85% of patients diagnosed with this neoplasm die from their disease. Complete surgical resection and cure can be considered in only 25% of cases.[1] Of the estimated 164,100 people in the United States who will be diagnosed with lung cancer in 2000,[2] 75% will have non–small-cell lung cancer (NSCLC) and the majority will die either of micrometastatic disease from tumors that initially appeared to be surgically resectable or from surgically unresectable locally advanced or metastatic disease.

Despite this grim picture, recent therapeutic advances may justify cautious optimism about future gains in the battle against lung cancer. Over the last decade, a group of new agents with differing mechanisms of action have shown great promise in early clinical studies in NSCLC. These include the taxanes docetaxel (Taxotere) and paclitaxel (Taxol), the nucleoside analog gemcitabine (Gemzar), the vinca alkaloid vinorelbine (Navelbine), the topoisomerase-I inhibitor irinotecan (Camptosar, CPT-11), and the bioreductive agent tirapazamine.

New Chemotherapeutic Agents in Combination Therapy

Having demonstrated activity as single agents, these new agents are now being evaluated in combination, mostly with cisplatin (Platinol, CDDP) (Tables 1, 7). Cisplatin has been the “backbone” of combination chemotherapy in patients with NSCLC because of its proven single-agent activity. In particular, therapy with cisplatin is associated with a relative lack of myelosuppression, a unique cytocidal mechanism of action, and the resultant possibility of synergistic cytotoxicity when combined with other agents. The latter has been confirmed in in vitro studies when cisplatin was used in combination with etoposide (VePesid, VP-16).[3,4]

Paclitaxel and Cisplatin-Based Therapy

In preclinical studies of the combination of cisplatin and paclitaxel, it was found that paclitaxel inhibited the repair of DNA adducts formed by platinum agents.[5] Preclinical studies have proven these agents to be synergistic when used in combination in in vitro models.[6] Phase I and II trials determined that the combination of cisplatin and paclitaxel was feasible. Rowinsky et al showed that paclitaxel at doses ranging from 135 mg/m² to 170 mg/m² administered over 24 hours with cisplatin 75 mg/m² was well tolerated.[7] When paclitaxel was administered following cisplatin, there was an increased incidence of grade 4 neutropenia. However, this same response was not observed when paclitaxel was administered prior to cisplatin and was attributed to a possible decrease in paclitaxel clearance due to the effect of cisplatin on the cytochrome P-450 system.[8] Subsequent trials found that paclitaxel at 200 to 250 mg/m² was tolerated and could be administered to patients with cisplatin when given with G-CSF granulocyte colony-stimulating factor (G-CSF, filgrastim, [Neupogen]) support (Table 2). Phase I and II studies demonstrated that when cisplatin is combined with paclitaxel, response rates of 31% to 47% are achieved.[9-12]

The Eastern Cooperative Oncology Group (ECOG) attempted to corroborate these results in a randomized phase III study of the combination of paclitaxel and cisplatin (Table 2).[13] The ECOG study was also designed to evaluate the dose-intensity for paclitaxel by including two doses of paclitaxel in combination with cisplatin. For the trial, 560 patients were randomized to one of three arms: (1) cisplatin 75 mg/m² plus etoposide 100 mg/m² (PE) administered intravenously on days 1-3 with cycles repeated every 3 weeks; (2) cisplatin 75 mg/m² plus paclitaxel 135 mg/m² administered over 24 hours (PT); or (3) cisplatin 75 mg/m² plus paclitaxel 250 mg/m² given over 24 hours with G-CSF support (PT-G).

The response rates were 12% for the PE control arm, 26% for the PT arm, and 32% for the PT-G arm (P < .001). Median survivals were 9.56 and 9.99 months for the PT and PT-G arms, respectively. This compared favorably with median survival in the PE control arm, which was 7.69 months (P = .091). There was no significant difference when the low- and higher-dose paclitaxel arms were compared. Moreover, a stronger trend for longer survival (P = .051) was observed when the paclitaxel-treated cohort was compared with patients treated with PE. Treatment-related mortality rates for arms PE, PT, and PT-G were similar: 2%, 4.4%, and 5.3%, respectively. The results of this study culminated in the combination of cisplatin and 24-hour infused paclitaxel 135 mg/m² being designated as the control arm for future ECOG trials in metastatic NSCLC.

Paclitaxel and Carboplatin in Combination

Several investigators have combined paclitaxel with carboplatin (Paraplatin), a congener of cisplatin. These trials were designed to reduce the significant nonhematologic toxicity seen with cisplatin. Carboplatin has minimal renal toxicity, reduced neurologic toxicity, and is less emetogenic. The principle toxicities of carboplatin relate to its effects on bone marrow suppression, particularly neutropenia and thrombocytopenia. The results of studies combining carboplatin with paclitaxel as a 3- or 24-hour infusion are summarized in Table 1. This combination has revealed interesting activity with reported response rates of approximately 30% to 50% in phase I or phase II trials in small numbers of patients (n = 10 to 53).

The primary toxicity was neutropenia of short duration. There were relatively few episodes of neutropenic fever. Patients receiving the shorter infusion of paclitaxel experienced peripheral neuropathy, particularly with repeated cycles of treatment. An interesting observation with this combination is the unexpectedly low incidence of significant thrombocytopenia.[14] Investigators have theorized that this is attributable to the protective effect mediated by paclitaxel against carboplatin-induced thrombocytopenia.[15] These studies have also demonstrated that the dosing of carboplatin with paclitaxel delivered over 24 hours is limited by myelosuppression, whereas neurotoxicity is the dose-limiting effect when the same combination is given over 1 or 3 hours.[16]

In most studies, the regimens were scheduled every 3 weeks, except for a study by Johnson in which a 4-week cycle was evaluated.[17] There appears to be an appreciable difference in response rates between the two schedules: the 4-week schedule had a response rate of 27%, whereas the majority of the studies utilizing the 3-week schedule had response rates of 50% or greater.[15,16,18-20] Until the results of ongoing phase III trials are complete, however, the optimal paclitaxel and platinum regimen cannot be defined. Nonetheless, the conclusions reached from studies completed to date have been interpreted as demonstrating that paclitaxel and the platin compounds are active combinations that are well tolerated and can be administered on an outpatient basis every 3 weeks.

The results of three recent phase III trials of platin compounds and paclitaxel vs cisplatin and etoposide or teniposide (Vumon) have recently been presented.[21-23] Belani et al randomized 179 patients to 75 mg/m² cisplatin on day 1, 100 mg/m² etoposide on days 1 to 3 of a 21-day cycle, and 190 patients to carboplatin with an area-under-curve (AUC) of 6 and 3-hour–infused paclitaxel at 225 mg/m².[21] All patients in this study had a performance status greater than 70% and stage III or IV disease (Table 7). Although there was no difference in the incidence of grade 3 or 4 neutropenia between the treatment arms, the incidence of grade 1 to 3 neuropathy was 4% in the taxane-containing arm and 0.6% in the etoposide arm. The response rate was 14% for the control arm compared to 21.6% for the carboplatin/paclitaxel arm. There was no difference in overall survival.

In another study of 332 patients, Giaccone et al on behalf of the European Organization for Research and Treatment of Cancer (EORTC) compared paclitaxel (175 mg/m² infused over 3 hours) and cisplatin (80 mg/m²) administered on day 1 of a 21-day cycle with the same dose of cisplatin plus teniposide (100 mg/m² administered on days 1 to 3) of a 21-day cycle.[22] The teniposide-containing arm had a higher incidence of grade 3/4 neutropenia (83%) and thrombocytopenia (36%) than the taxane arm (55% and 2%, respectively), whereas the taxane arm had more grade 2/3 neurotoxicity (29% vs 7%). Response rates favored the taxane arm significantly (41% vs 28%; P = 0.18), however, this did not translate into a survival advantage as the

1-year overall survival rate was 41% in the teniposide arm and 43% in the taxane-containing arm. Gatzemeier compared high-dose cisplatin (100 mg/m²) administered every 3 weeks with cisplatin (80 mg/m²) plus paclitaxel (175 mg/m²) administered every 3 weeks (Table 7).[23] In this study of 414 patients with NSCLC, 70% had stage IV disease and 30% had stage IIIB. Patients in the single-agent arm received a median of three courses of treatment compared with five courses in the combination arm. The response rate in the cisplatin alone arm was 17% vs 26% in the combination arm (P = .0002) and the time-to-disease progression was 2.7 months vs 4.1 months respectively (P = .028). The median survival in the cisplatin arm was 8.6 months vs 8.1 months in the cisplatin plus paclitaxel arm (P = .826). Peripheral neuropathy, neutropenia, and alopecia were worse in the combination treatment arm, whereas nausea/vomiting and constipation were more problematic in the cisplatin arm.

Docetaxel and Cisplatin

This combination has also been evaluated in patients with advanced or metastatic NSCLC. The response rates (25% to 53%) were similar to those seen when paclitaxel was given in combination with cisplatin, with acceptable toxicity. Docetaxel was administered in doses of 75 mg/m² with 75 to 100 mg/m² cisplatin given on an every-3-week schedule.[24-26] The principle toxicity was grade 4 neutropenia. The incidence of febrile neutropenia was approximately 10% among patients who received this combination. The accumulation of peripheral edema appears to be ameliorated by the use of short courses of dexamethasone administered shortly before, and for a few days after, docetaxel.

Vinorelbine and Platinum-Based Chemotherapy

Phase I/II trials established the activity and acceptable toxicities of this combination (Table 1). Gebbia et al reported the results of a phase I/II study of cisplatin 80 mg/m² on day 1 and vinorelbine 25 to 30 mg/m² on days 1 and 8 of a 3-week cycle.[27] An overall response rate of 46% and a median survival of 34 weeks was observed in 30 patients, all of whom had good performance status and either stage III or IV disease. The toxicities were World Health Organization (WHO) grade 3 leukopenia (33%) and injection vein phlebitis (16%). Similar findings were noted by Berthaud et al when vinorelbine 30 mg/m2 weekly was administered with cisplatin 120 mg/m² every 4 to 6 weeks.[28] The median age of the 32 patients in that study was 55 years. Most had an ambulatory performance status with locally advanced metastatic disease. The response rate was 33% and the median survival was 44 weeks.

Three phase III trials have demonstrated the efficacy of the vinorelbine and cisplatin combination. In a study conducted by LeChevalier et al, 612 patients were prospectively randomized to one of three arms: (1) cisplatin 120 mg/m² on days 1 and 29 and then every 6 weeks plus vinorelbine 30 mg/m² weekly; (2) cisplatin 120 mg/m² alone on days 1 and 29 and every 6 weeks plus vindesine (Eldisine) 3 mg/m² weekly for 6 doses and then every other week vs (3) single-agent vinorelbine 30 mg/m² weekly.[29] Approximately 65% of the patients in this study had a performance status of 0 or 1 and stage IV disease. The median age of patients was 60 years. The authors reported response rates favoring the cisplatin plus vinorelbine arm, 30% vs 19% in the cisplatin plus vindesine arm (P = .02).

A modest survival advantage was also found in favor of the cisplatin/vinorelbine combination, with a median survival of 40 weeks compared with 32 weeks for the cisplatin plus vindesine combination (P = .01). Survival in the single-agent vinorelbine arm was similar to that in the vindesine/cisplatin arm, 31 weeks vs 32 weeks, respectively. There was increased myelosuppression with the cisplatin/vinorelbine combination, but less neurotoxicity than with the vindesine/cisplatin combination.

The Southwest Oncology Group (SWOG) conducted a second randomized trial evaluating vinorelbine in combination with cisplatin.[30] In this study, 432 patients with advanced or metastatic NSCLC and good performance status were randomized to receive either cisplatin 100 mg/m² once every 4 weeks or cisplatin 100 mg/m² once every 4 weeks plus vinorelbine 25 mg/m² weekly. The combination of cisplatin plus vinorelbine elicited a superior response rate compared to single-agent cisplatin, 26% vs 12%, respectively (P = .0002). Overall median survival and 1-year survival favored the combination arm with a median survival of 8 months and 1-year survival of 36% vs 6 months and 20%, respectively (P = .0018), in the single-agent arm. As expected, the combination arm had a much higher incidence of WHO grade 3/4 neutropenia (81%) compared with the single-agent cisplatin arm (5%).

Based on the results of their previous randomized trial, the SWOG conducted a second trial in 408 eligible patients with stage IIIB (12%) and stage IV (88%) NSCLC. Patients were randomized to receive either vinorelbine plus cisplatin (VC) or paclitaxel plus carboplatin (PC).[31] There was no difference in response rate (27% for each arm), median survival (8 months for each arm), or 1-year survival (36% for VC and 33% for PC). Economic and quality-of-life parameters were compared. The preliminary analysis revealed no differences in quality of life, but a significant difference in pharmacoeconomic data favoring the VC arm.

Santammagio et al have combined carboplatin with vinorelbine in a study in 77 patients with NSCLC. The study had standard inclusion criteria. A response rate of 31% and median survival of 9 months were reported.[32] Toxicities included seven cases of grade 3 nausea/vomiting, two cases of grade 3 granulocytopenia, and six cases of grade 3 anemia.