The management of patients with advanced non-small-cell lung cancer (NSCLC) remains a daunting task. The majority of patients have inoperable disease that is locally advanced (stage IIIA-B) or metastatic to distant sites (stage IV). Until the past several years, the primary therapy available to patients was radiation treatment and supportive care. Because of the predominantly older population with significant comorbidity as well as the morbidity of the disease itself, advances in treatment have been limited. However, during the past decade, several randomized trials have confirmed a survival advantage with chemotherapy when applied to patients with good performance status who have either stage III or IV non-small-cell lung cancer.
Although these effects have been modest, they have been consistent. For supportive care alone, the median survival time across eight trials was 9 to 21 weeks, with a median of 16 weeks.[2-9] For patients receiving combination chemotherapy, generally including cisplatin(Drug information on cisplatin) (Platinol), the median survival time ranged from 20 to 35 weeks, with a median of 28 weeks. Although the difference in median survival time was only a few months, it represented nearly double the survival time associated with supportive care alone. Furthermore, this increase in median survival time also has been associated with increases in 1- and 2-year survival rates for individuals with advanced disease and increases in 5-year survival rates for those with stage III disease.
All of these increases in survival rates were achieved using chemotherapy regimens commonly available in the 1980s and early 1990s. During the past several years, many newer agents have been identified that hold promise for further improvement in outcomes with chemotherapy for non-small-cell lung cancer. A number of promising phase II regimens have been reported using vinorelbine (Navelbine), paclitaxel(Drug information on paclitaxel) (Taxol), docetaxel (Taxotere), gemcitabine(Drug information on gemcitabine) (Gemzar), topotecan(Drug information on topotecan) (Hycamtin), and irinotecan(Drug information on irinotecan) (Camptosar).
Many of these trials have provided encouraging response rates and apparent improvements in toxicity profiles. However, comparative trials are essential to document overall clinical benefit. Because chemotherapy has been established as being superior to supportive care alone in patients with advanced non-small-cell lung cancer, appropriate comparisons must be made between regimens containing the newer agents with those containing the older agents, and ultimately among the newer agents themselves.
Vinorelbine has become the first new agent to be approved for the treatment of non-small-cell lung cancer in more than 20 years. Data supporting vinorelbine as a superior agent continue to accumulate and now include results from three large, randomized, prospective, multicenter trials. These three trials include the US trial of single-agent vinorelbine vs fluorouracil(Drug information on fluorouracil) (5-FU), the European trial of single-agent vinorelbine vs vindesine(Drug information on vindesine)/cisplatin vs vinorelbine/cisplatin, and the Southwest Oncology Group (SWOG) trial of single-agent cisplatin vs vinorelbine/cisplatin. Each one of these trials is discussed elsewhere in this supplement.
It is important to look at each of these studies individually with respect to the end points of response rate, toxicity profile, symptom benefit, and quality of life. It is also useful to evaluate these three trials collectively for the relative impact of these various regimens on survival. There are some differences in the patient populations across these studies. However, in all three trials, patients predominantly had stage IV disease and survival was superior with the vinorelbine-containing regimens compared with the control arms.
As outlined in Table 1, single-agent cisplatin or single-agent 5-FU yielded a 1-year survival rate of approximately 16%, whereas single-agent vinorelbine was associated with a 1-year survival rate of 25% in the US trial (in which all patients had stage IV disease) and 30% in the European trial (in which a subgroup of patients had stage III disease). These 1-year survival rates of 25% and 30% with vinorelbine are comparable with the 27% 1-year survival rate associated with the combination of vindesine and cisplatin in the European trial. However, these rates were inferior to the 35% 1-year survival rate associated with vinorelbine/cisplatin demonstrated in both the European and SWOG studies. These survival results with vinorelbine form the benchmark for survival in advanced non-small-cell lung cancer, either as a single agent or in combination therapy.
Although cisplatin and vinorelbine have been widely used in the treatment of advanced non-small-cell lung cancer, there is a lack of consensus concerning the best regimen. This lack of consensus reflects the fact that improvement in outcome with vinorelbine alone or in combination, although clinically significant, remains modest. This issue is further complicated by the perceived and real toxic effects of cisplatin per se, as well as the hope that newer combinations may produce even better outcomes. Of these newer regimens, the one used most widely is the combination of paclitaxel and carboplatin(Drug information on carboplatin). Use of this combination evolved from the broad activity of paclitaxel evident in a number of disease settings and the apparent improvement demonstrated when this agent was used in combination with cisplatin. However, the combination regimen of cisplatin and paclitaxel was limited by neurotoxicity, which led to the substitution of carboplatin for cisplatin.
Although several phase II trials have been performed, it is instructive to examine the studies of Langer et al and Johnson et al. The overall response rate was 62% and the 1-year survival rate was 54% in the study by Langer et al, whereas the response rate was 27% and the 1-year survival rate was 32% in the study by Johnson et al. Both trials used similar doses of carboplatin and paclitaxel by 24-hour infusion; however, Langer et al used intraindividual dose escalation of paclitaxel to a maximum of 215 mg/m2 with granulocyte colony-stimulating factor (G-CSF) support, whereas Johnson et al used a starting dose of 115 mg/m2 with dose reduction (as necessary) as a strategy for managing neutropenia. In addition, the patients in the trial by Langer et al had more favorable performance status.
This range of outcomes emphasizes the pitfalls of interpreting phase II results and underscores the need for phase III, randomized trials involving direct comparisons among vinorelbine/cisplatin, paclitaxel/carboplatin, and the other newer combination regimens.
We initiated a phase I/II trial of vinorelbine and carboplatin in patients with advanced non-small-cell lung cancer. Preliminary results of the phase I portion of the study have been reported, and the final results will be reported elsewhere.
The rationale for this combination was based on the favorable toxicity profile observed with vinorelbine as a single agent in the US trial, as well as the documented improvement in survival with the addition of cisplatin,[11,12] but also the excessive toxicity associated with the addition of cisplatin. The substitution of carboplatin for cisplatin was attractive in an attempt to reduce the incidence of nausea and vomiting, fatigue, neurotoxicity, ototoxicity, nephrotoxicity, and alopecia, which are more often associated with cisplatin. However, the potential limitation of this change was the enhanced potential for myelotoxicity with carboplatin vs cisplatin. There was also the perception in the lung cancer community that carboplatin was inferior to cisplatin. This perception was never based on comparative trials; in fact, before the development of paclitaxel, carboplatin was the single most active agent tested by the Eastern Cooperative Oncology Group, yielding a 1-year survival rate of 20% when used alone. This would appear to be at least comparable, if not superior, to the results obtained with cisplatin in the SWOG trial noted in Table 1.
Phase I Trial
In the phase I trial conducted at both Duke University Medical Center and the Greenville Cancer Center, 22 patients received monthly carboplatin at an area under the concentration vs time curve (AUC) of 7 (according to the Calvert formula), plus weekly vinorelbine in doses ranging from 0 to 30 mg/m2 in various patient cohorts. The trial was designed to incorporate the use of G-CSF in attempt to maintain the intended dose of vinorelbine.
In the cohort of patients receiving carboplatin alone, not one patient reached dose-limiting toxicity and one of five patients had a partial response to treatment. No responses were achieved in the cohort of patients receiving vinorelbine 15 mg/m2/week, and two of the five patients required G-CSF support to maintain the dose intensity of the regimen. In two cohorts combined for a total 12 patients, including one cohort receiving vinorelbine 22.5 mg/m2/week, and a second cohort receiving vinorelbine 30 mg/m2/week, 5 of 12 patients achieved a partial response, but 9 of 12 patients required G-CSF support. In a subsequent cohort treated with vinorelbine 37.5 mg/m2/week, all patients required G-CSF support; despite the use of G-CSF, the intended dose intensity could not be maintained because of an increase in nonhematologic toxicity (particularly constipation and ileus).
Based on the results of this phase I experience, carboplatin at an AUC of 7 by the Calvert formula on a monthly basis, plus vinorelbine 30 mg/m2/week with G-CSF support, was chosen for the phase II study. The incidence of grade 4 neutropenia was high at this dose, but only one of seven patients experienced febrile neutropenia. Also, although the majority of patients developed anemia requiring red blood cell support over the course of several cycles of treatment, only one of seven patients required a platelet transfusion.