Novel Agents in the
Treatment of Lung Cancer
New Approaches to First-Line Therapies
A Continuing Medical Education Activity sponsored by InforMEDical Communications, Inc.
InforMEDical Communications, Inc. •
83 Acton Street • Carlisle, MA 01741
tel: 978-318-9582 • fax: 978-318-9583 • email: info@informedicalcme.com
Defining the Optimal Duration of Chemotherapy in Advanced NSCLC
In the setting of curable solid tumors, a short course of chemotherapy—typically 3 or 4 cycles—is sufficient to achieve its goal. Similarly, a total of 4 cycles has also been used in successful trials of adjuvant chemotherapy in NSCLC. However, when dealing with incurable lung cancer the duration of treatment is not well defined. Clinicians are aware that the benefit of a certain regimen, whether radiographic or clinical, will plateau at some point, while the toxicity of the treatment continues to accumulate. Assuming that this is the case, it becomes clear that the benefit of tumor control may be outweighed by the risks of ongoing treatment.
Prolonged Chemotherapy
Several studies have evaluated the exact duration of a single regimen in advanced NSCLC, and these data have served as the basis for the guidelines published by the American Society of Clinical Oncology.67 Smith et al. randomized 308 patients with stage IIIB or IV NSCLC to receive either 3 or 6 cycles of MVP (mitomycin, vinblastine, cisplatin) (Table 2).68 Randomization occurred at the time of enrollment because the authors anticipated that patients would not accept random assignment after chemotherapy was initiated. The study was negative, both in terms of progression-free survival (PFS) (5 vs 5 months) and overall survival (OS) (6 vs 7 months). A second study of similar design was performed by Socinski et al., who randomized patients to either a defined course of 4 cycles of carboplatin and paclitaxel, or to the same chemotherapy until progression.69 All patients were to receive weekly paclitaxel upon relapse. Overall survival was almost 2 months longer in the prolonged chemotherapy arm (8.5 vs 6.6 months), but this difference was not statistically significant. Of note, 42% of patients in the shorter duration treatment arm received second-line therapy.
While the studies mentioned previously used OS as the primary endpoint, von Plessen et al. examined differences in the quality of life (QoL) of patients treated with either 3 or 6 cycles of carboplatin and vinorelbine.70 Accordingly, the doses used were attenuated to limit toxicity (carboplatin AUC 4 and vinorelbine 25 mg/m2 on days 1 and 8 every 21 days). Interestingly, there was no detriment in the QoL of patients assigned to the longer duration of chemotherapy, but there were also no statistically significant differences detected in PFS (3.6 vs 4.8 months) or OS (6.4 vs 7.3 months).
There are 2 major limitations in these studies: the inclusion of patients destined to progress early on treatment, which dilutes any possible effect of prolonged therapy, and the inability to deliver therapy according to protocol. It is worth mentioning that the percentage of patients actually completing the "prolonged" therapy is fairly small. Only 31% and 34% of patients received 6 cycles of MVP or carboplatin and paclitaxel, respectively. A larger proportion (54%) completed 6 cycles of carboplatin and vinorelbine, likely reflecting the milder nature of this regimen. In all of these studies, about 50% to 70% of patients were able to receive 3 to 4 cycles of therapy on either arm, suggesting that the initial few cycles are more important for the treatment effect. This is also illustrated by the fact that the response rates in the prolonged treatment group did not increase with more chemotherapy: 32% at 3 cycles of MVP versus 38% at 6 cycles; and 24% both at 4 or more than 4 cycles of carboplatin and paclitaxel.
Maintenance Chemotherapy
The lack of additional benefit with more cycles of the same treatment, coupled with the high rate of dropout due to toxicity, deterioration in the absence of progression, and patient refusal indicated that a different strategy was needed. Subsequent studies utilized a limited number of upfront chemotherapy cycles and relied on a lower intensity, prolonged therapy as the experimental approach.
Two studies used the agent initially coupled with platinum as the preferred regimen, giving a "maintenance" dose of the "induction" chemotherapy (See Table 3). Belani et al. randomized patients with stable or responsive disease after 16 weeks of carboplatin and paclitaxel to either observation until progression or weekly paclitaxel.71 This was primarily a randomized phase II study of different regimens of carboplatin and paclitaxel, with a secondary randomization following the initial regimen, and therefore had limited statistical power to detect differences in the prolonged treatment arm. The second study, by Brodowicz et al., randomized patients with at least stable disease after 4 cycles of cisplatin and gemcitabine to observation or single-agent gemcitabine, and used PFS as the primary end point.72 Both studies showed an improvement in PFS favoring the experimental arm, which was statistically significant in the Brodowicz study (3.6 vs 2 months; P<0.001). However, OS showed no significant differences (Belani: 13.6 vs 10.1 months; Brodowicz: 10.2 vs 8.1 months). Despite the absence of a survival benefit, there are some important conclusions that can be drawn from these efforts. It is evident that patients are generally willing to accept randomization, even those with responsive disease, after their initial chemotherapy. Additionally, there was a definite PFS benefit, with numerical superiority in OS for the maintenance arm in these underpowered studies, suggesting that further exploration of such regimens may be reasonable. Lastly, it is important to note that only 30% to 60% of patients undergoing front-line therapy were considered for maintenance therapy.
Early Second-Line Chemotherapy
More recent trials used a similar basic design, but instead of continuing with the same single agent they utilized a non-cross-resistant agent, typically one already approved for the second-line therapy of NSCLC.
Fidias et al. enrolled patients with stage IIIB/IV NSCLC, and randomized those with stable or responsive disease after 4 cycles of carboplatin and gemcitabine to either immediate treatment with docetaxel for 6 additional cycles or delayed use of the same chemotherapy at the time of relapse.73 Belani et al. recently reported the updated results of an international study that randomized 663 patients following 4 cycles of a platinum doublet to either pemetrexed or placebo.5 Although the primary endpoint of the study was PFS, there was adequate power to detect meaningful differences in OS. Two additional studies explored the value of early use of erlotinib. SATURN was an international study that evaluated erlotinib or placebo given immediately after initial treatment with at least 4 cycles of platinum-based chemotherapy.8 ATLAS was also a large global trial that utilized 4 cycles of bevacizumab in combination with the investigators' choice of platinum-based doublet as front-line therapy.9 Patients without evidence of progressive disease (PD) were subsequently randomized to receive either bevacizumab and erlotinib or bevacizumab and placebo. The details of these trials are discussed by Dr. Hanna, but the hazard ratios both for PFS and OS are very similar among these trials, suggesting a similar general effect is obtained with any of these agents in this setting (Table 4).
Maintenance Bevacizumab
Over the last several years, bevacizumab has become the agent most frequently used in the maintenance treatment of advanced NSCLC. Theoretically, such an ongoing suppression of angiogenesis may be important, since the target of bevacizumab, VEGF, is genetically stable and expressed throughout the entire tumor life cycle. However, at this point there is an absence of randomized phase III data demonstrating the superiority of maintenance bevacizumab over observation.
Data from ATLAS may be informative, although the inclusion criteria were restrictive due to the use of bevacizumab. The median PFS for patients randomized to bevacizumab and placebo in this study was 3.75 months, which is much longer than that reported in the observation arm of several recent studies (Fidias: 2.7 months; Belani: 2 months; SATURN: 2.8 months).5,8,9,73 Whether this difference is due to the use of maintenance bevacizumab or simply patient selection will need to be explored in a randomized trial.
Such a trial has recently been completed in ovarian cancer (GOG 218), in which patients were randomized either to carboplatin and paclitaxel; carboplatin, paclitaxel and concurrent bevacizumab; or the same triplet with 1 year of maintenance bevacizumab. Similarly, ECOG is planning a phase III trial for advanced NSCLC that will randomize non-progressing patients after 4 cycles of carboplatin, paclitaxel, and bevacizumab to either bevacizumab alone, pemetrexed alone, or the combination of pemetrexed and bevacizumab.
Conclusion
Multiple studies have demonstrated that 4 cycles of initial chemotherapy is sufficient treatment for patients with advanced NSCLC. For those without progression, there is emerging evidence that early second-line therapy, especially with pemetrexed in patients with nonsquamous histology, is associated with an OS benefit compared to observation. Whether such benefit applies to patients treated with maintenance bevacizumab or patients initially treated with the same agent in the upfront setting remains to be determined. Although we need more information about the symptom control afforded and the QoL associated with maintenance therapy, this approach appears to be here to stay.
(click table to view larger version)