Strategies for maintenance therapy in advanced non-small cell lung cancer: Current status, unanswered questions and future directions

Critical Reviews in Oncology/Hematology 82 (2012) 338–360

Strategies for maintenance therapy in advanced non-small cell lung cancer: Current status, unanswered questions and future directions Ana Custodio ∗ , Javier de Castro Medical Oncology Department, Hospital Universitario La Paz, Paseo de la Castellana, 261, 28046 Madrid, Spain Accepted 16 August 2011

Contents 1. 2.

3. 4.

5. 6.

7.

Introduction. What is maintenance therapy? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Maintenance therapy with the same chemotherapeutic agents used in the first-line regimen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Continuation of platinum-based doublet first-line chemotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Continuation of the non-platinum component of the first-line CT: “continuation maintenance” strategy . . . . . . . . . . . . . . . . . Maintenance treatment with different agents from those used in the first-line regimen: switch-maintenance or early second-line therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Maintenance therapy with molecularly targeted agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Maintenance treatment with targeted agents already present in the induction phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Sequential treatment with different targeted agents from those used in the first-line regimen . . . . . . . . . . . . . . . . . . . . . . . . . . . Meta-analysis of the maintenance therapy studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Unanswered questions about maintenance therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. What is the most appropriate study end point?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Maintenance treatment and quality of life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3. Who should receive maintenance treatment? In patients for whom a drug holiday may be appropriate, how long should the holiday be? How best should we monitor patients during a break from treatment? . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4. How should we select the right patient population who would benefit from maintenance treatment? . . . . . . . . . . . . . . . . . . . . 6.5. Implications for drug development and clinical trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Biographies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

339 340 340 341 344 346 346 349 352 354 354 354 355 355 356 357 357 357 360

Abstract Systemic chemotherapy (CT) with platinum-based doublets result in modest improvements in both overall survival (OS) and quality of life in good performance status patients with advanced non-small cell lung cancer (NSCLC). However, although substantial progress has been made in the therapeutic options currently available for these patients, the overall outcome remains poor. Maintenance therapy for patients who achieved at least stable disease after first-line treatment has been an area of intense investigation in recent years as a way of improving outcomes in metastatic NSCLC. Several alternative strategies for prolongation of initial treatment have been evaluated. These include the prolongation of the initial combination CT regimen until disease progression, unacceptable toxicity or a predefined greater number of cycles, continuation with a lower intensity version of the first-line CT regimen or administration of a new active agent immediately after completion of the first-line therapy (switch-maintenance or early second-line therapy). Treatments that have

∗

Corresponding author. Tel.: +34 91 2071138; fax: +34 91 7277118. E-mail address: [email protected] (A. Custodio).

1040-8428/$ – see front matter © 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.critrevonc.2011.08.003

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

339

been studied in randomized trials to date include CT, molecularly targeted agents, and immunotherapy approaches. Phase III trials have not revealed a survival benefit for extended first-line CT with combination regimens for more than 4–6 cycles. Nevertheless, early second-line therapy with pemetrexed in nonsquamous tumours and erlotinib have demonstrated to improve OS results, especially in select patient groups characterized by histology and/or molecular profile. This article reviews recent data with maintenance therapy in advanced NSCLC and discusses the implications for routine patient care and future drug development. © 2011 Elsevier Ireland Ltd. All rights reserved. Keywords: Maintenance therapy; Pemetrexed; Erlotinib; Bevacizumab; Epidermal growth factor receptor; Vascular endothelial growth factor receptor; Clinical trial

1. Introduction. What is maintenance therapy? Non-small cell lung cancer (NSCLC) is the leading cause of cancer mortality in the industrialized world, responsible for more than a million deaths worldwide each year. Most patients are diagnosed with locally advanced or metastatic disease (≈70–80%), and in this palliative setting balancing efficacy with toxicity is of the utmost importance [1]. The current standard of care for treatment of advanced stage NSCLC patients with a preserved performance status (PS) is a platinum-based regimen, which results in modest prolongation of survival, improvement in cancer-related symptoms and quality of life (QoL) [2–4]. Nevertheless, only approximately 60% of patients will experience disease control at 8 weeks, and the median overall survival (OS) observed in recent studies of platinum-based doublets was 10–13 months, with <5% survival at 5 years [5,6]. Recently, epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) have emerged as a first-line treatment option for patients whose tumours harbour an activating mutation of the receptor tyrosine kinase [7–10]. Although it is generally accepted that the EGFR inhibitor therapy should be continued indefinitely until disease progression, the optimal duration of first-line chemotherapy (CT) is unclear. Several studies investigating the optimal treatment duration have demonstrated that continuation of combination CT beyond four to six cycles only results in added toxicity without a meaningful improvement in progression-free survival (PFS) or OS, suggesting that the maximum benefit of CT is yielded by the first few cycles [11–15]. So, current guidelines from the American Society of Clinical Oncology (ASCO) [2], the National Comprehensive Cancer Network (NCCN) [16] and the European Society of Medical Oncology (ESMO) [3] all recommend up to a maximum of 6 and minimum of 4 cycles of first-line platinum-based doublet CT for responding patients or those with stable disease (SD). The current practice of adopting a “watch and wait” approach after achieving maximal response provides the opportunity for patients to experience a “drug holiday”, but it is often associated with anxiety about disease progression, with particular concern for clinical deterioration and the inability to receive secondline treatment. Before discussing the options, it is worth considering what proportion of patients actually receives second-line therapy. Evidence from recent major clinical trials, such as the Eastern Cooperative Oncology Group (ECOG) 4599 study [17], First-Line Erbitux in Lung Cancer

(FLEX) [18] or the trial comparing cisplatin/pemetrexed with cisplatin/gemitabine [19] suggest that this figure is approximately 50–60% of patients treated with front-line therapy. The addition of molecularly targeted agents such as cetuximab or bevacizumab to first-line combination CT is associated with modest improvement in survival [18–20]. With this approach the targeted agent is usually continued beyond the initial induction phase with the combination therapy. Although this is a biologically rational approach, there is no clinical evidence that patient outcomes are improved with such prolonged treatment. Furthermore, the targeted therapy combinations may be suited for selected subsets of patients with advanced stage NSCLC based on clinical characteristics, anatomy and tumour histology [21]. The relative brief duration of disease control even after a major response to front-line treatment has prompted investigators to pursue other novel strategies to delay progression and improve survival for advanced stage NSCLC [22,23]. Maintenance therapy (MT) is the continued administration of therapy after a specified number of treatment cycles once maximum tumour response or disease stabilization have been achieved. The two treatment strategies to extend the duration of treatment in advanced NSCLC that have been more intensively investigated in last years include “continuation maintenance” and “switch maintenance”. Continuation maintenance describes the strategy of continuing a CT or targeted agent that was part of the first-line induction platinum-doublet regimen after a defined number of cycles of combination therapy. If a non-crossresistant agent is used as MT before disease progression after first-line platinum-based CT, this approach can be defined as “switch maintenance”, early second-line or sequential therapy. Although a number of active chemotherapeutic and targeted agents are now available for the treatment of advanced NSCLC, it is clear that not all of them are suited for administration for prolonged number of cycles. The optimal MT agent should be associated with proven efficacy, a favourable toxicity profile and the ability to prolonged administration without significant risk of serious cumulative toxicity. The challenges that lie in interpreting the literature come from the heterogeneity of studies and the lack of consensus with respect to what constitutes MT. This review summarizes the rationale, current data and perspectives of maintenance and early-second line treatment in patients with advanced NSCLC.

340

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

Table 1 Phase III trials of prolonged platinum-based doublet CT. Trial/year

Treatment arms

Median PFS

Median OS

Grades 3–4 toxicities

Smith et al. [11]

Mitomycin, vinblastine and cisplatin × 3 cycles (n = 155) vs. Mitomycin, vinblastine and cisplatin × 6 cycles (n = 153) Carboplatin and paclitaxel × 4 cycles (n = 114) vs. Carboplatin and paclitaxel until progression (n = 116) Carboplatin and vinorelbine × 3 cycles (n = 150) vs. Carboplatin and vinorelbine × 6 cycles (n = 147) Third-generation platinum-doublet × 4 cycles (n = 156) vs. Third-generation platinum-doublet × 6 cycles (n = 158)

5 months

6 months

3 cycles: 12% leucopenia, 10% infection, 6% emesis

5 months (p = 0.4)

7 months (p = 0.2)

Socinski et al. [12]

Von Plessen et al. [13]

Park et al. [14]

6 cycles: 22% leucopenia, 5% infection, 5% emesis

NR

6.6 months

NR

8.5 months (p = 0.63)

39% neutropenia, 14% neuropathy 42% neutropenia, 27% neuropathy

16 weeks

28 weeks

21 weeks (p = 0.21)

32 weeks (p = 0.75)

52% leucopenia, 20% infection, 3% anemia 46% leucopenia, 16% infection, 9% anemia

4.6 months

15.9 months

6.2 months (p = 0.001)

14.9 months (p = 0.461)

9.6% neutropenia, 3.2% febrile neutropenia, 0.6% anemia 12.7% neutropenia, 8.2% febrile neutropenia, 9.5% anemia

PFS: progression-free survival; OS: overall survival; NR: not reported.

2. Maintenance therapy with the same chemotherapeutic agents used in the first-line regimen 2.1. Continuation of platinum-based doublet first-line chemotherapy Continuation in the context of strategies for prolonging therapy in advanced NSCLC involves continuing the platinum-based first-line regimen until evidence of progressive disease in the absence of significant toxicity or administration of a predefined greater number of treatment cycles. The largest randomized trials that have examined the value of extending the duration of the initial cytotoxic CT for metastatic NSCLC are summarized in Table 1 [11–14]. These studies varied with respect to the number of cycles in the “standard” arm (three gave 3 cycles and one gave 4 cycles) and in the experimental arm (three trials administered six cycles and only one trial employed a true “maintenance” approach with continuation of CT until progression). There are also a number of small studies, or studies reported only in abstract form that use this trial design outline. Smith et al. were the first to address the question of optimal duration of treatment by random assignment of 308 patients with stage IIIB or IV NSCLC who had achieved an objective or symptomatic response to three cycles of mitomycin, vinblastine and cisplatin (MVP) to stop or to continue for a further three cycles [11]. Authors failed to demonstrate

any advantage from six cycles: median time to disease progression (TTP) was 5 months for both arms (p = 0.4) and median OS was 6 vs. 7 months for three vs. six courses, respectively (p = 0.2). However, it should be noted that while more than 70% of the patients completed 3 cycles of therapy, less than one-third (31%) in the experimental arm were able to receive 6 cycles. There were no differences in QoL parameters between the two groups for the first 9 weeks of treatment, but during weeks 9–18, patients continuing on CT reported significantly more fatigue (p = 0.03) and a trend toward increased nausea and vomiting (p = 0.06). Socinski et al. performed a second phase III study to compare a regimen of paclitaxel plus carboplatin every 3 weeks for either 4 cycles (arm A) or continuously until there was radiographic evidence of progression (arm B) [12]. All patients received second-line single agent weekly paclitaxel after the scheduled 4 cycles in arm A or at progression in arm B. The coprimary endpoints were survival and QoL. Interestingly, the median number of treatment cycles on both arms was four, although in the investigational arm, 42% patients received greater than four cycles and 18% received eight or more cycles. Only 45% of patients received second-line therapy (42% in arm A vs. 47% in arm B). The study showed no overall benefit in terms of response or survival to continuing treatment beyond four cycles. Overall response rates (ORR) were 22% and 24% for arms A and B, respectively (p = 0.080). Median OS time and 1-year survival rates were 6.6 months and 28% for arm A and 8.5 months and 34%

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

for arm B, respectively (log-rank p = 0.63). The lack of a significant difference in survival between the study arm was evident both in the overall population and when comparing subgroups of patients who had received at least 4 treatment cycles. Rates of hematologic and nonhematologic toxicity were similar between the two arms, except for neuropathy. As it might be expected given the recognized side effect profile of paclitaxel, the rate of grades 2–4 neuropathy increased from 19.9% at cycle 4 to 43% at cycle 8. There were no differences in QoL. The study conducted by von Plessen et al. evaluated a carboplatin and vinorelbine regimen every 3 weeks given for either three (C3) or six (C6) courses [13]. Key endpoints were QoL at 18 weeks and OS. As in the study by Smith [11], 117 (78%) patients across the study completed 3 cycles of therapy; however, only 79 (54%) patients assigned to the longer regimen completed 6 cycles. Although median PFS and OS favoured the six-cycle arm, the differences were not significant (p = 0.21 for PFS and p = 0.75 for OS). In addition, there were no significant group differences in global QoL, pain or fatigue up to 26 weeks. The dyspnea palliation rate was lower in the C3 arm at 18 and 26 weeks (p < 0.05), but this finding was inconsistent across different methods of analysis. Park et al., representing the Korean Cancer Study Group, conducted a phase III trial of slightly different design to define the optimal duration of CT in patients with advanced NSCLC whose disease had demonstrated sensitivity to platinumbased CT [14]. A total of 452 Korean patients with stage IIIB with malignant effusion or stage IV NSCLC were randomized to receive either four (arm A, six-cycle group; n = 158) or two (arm B, four cycle group; n = 144) additional cycles of third-generation, platinum-based CT, after response to only two induction cycles of the same CT regimen. This study was designed to demonstrate noninferiority in 1-year survival rate with only two additional cycles. Again, there was a disparity in terms of completed cycles, with 68.4% of patients assigned to arm A and 92.3% of those assigned to arm B completed four and six planned cycles of treatment, respectively. The difference in the 1-year survival rate between the two groups was 3.4% (59% in arm A and 62.4% in arm B) and met the predefined criteria of noninferiority. Median TTP was significantly longer in arm A (6.2 months compared with 4.6 months in arm B; p = 0.001), although there was no difference in OS (median survivals of 14.9 and 15.9 months for arms A and B, respectively; p = 0.461). The authors postulated that the main reason why the TTP benefit did not translate into a survival benefit probably involved the dilution effect of the secondline treatment, because 62.7% of the six-cycle group and 74.4% of the four-cycle group received the second-line CT, respectively (p = 0.026). Toxicity rates were similar in both arms. However, patients assigned to four cycles experienced less nausea/vomiting, sore mouth and dyspnea (p < 0.05) and more favourable QoL than those receiving six cycles. It is important to note that the difference in this trial, compared with previous studies, is that only patients who responded after two cycles were randomized to continue. This fact

341

probably explains the overall higher proportion of patients who went on to receive their full complement of planned CT, and survived longer compared with previous studies. Based on these data we can conclude that continuation of combination CT beyond a defined number of cycles did not provide additional benefit, but it was associated with higher toxicity, leading to a discontinuation of further clinical investigation. However, as it has been previously described, a substantial proportion of patients enrolled in the longer duration treatment of these trials experienced either disease progression or unacceptable toxicity and received fewer than the planned number of cycles. This may have contributed to the similar OS in the two treatment arms. 2.2. Continuation of the non-platinum component of the first-line CT: “continuation maintenance” strategy It is widely accepted that toxicity from a platinum-based regimen is derived mainly from the platinum component, and that most patients cannot tolerate prolonged administration of platinum-based doublets. Nevertheless, it does not occur with third-generation single-agent CT, as demonstrated in second-line therapy trials [24,25]. Using a drug with known single-agent activity for MT is theoretically possible to slow disease progression and improve disease-related symptoms, with minimal side effects. In view of the favourable tolerability of third-generation chemotherapeutic agents, a recent strategy has investigated the value of continued treatment of responding and stable patients with the non-platinum component of the first-line induction regimen. Table 2 shows trials with this design. Two studies have evaluated paclitaxel as MT after first-line CT with carboplatin and paclitaxel [26,27]. In the first study conducted by Belani et al. [26], 401 patients were initially treated with one of three paclitaxel/carboplatin induction regimens for 16 weeks. Patients who responded (n = 130) were randomly assigned to either weekly paclitaxel therapy (n = 65) or observation (n = 65). The primary end point was TTP. Although the study was not powered to compare outcomes during the maintenance phase, median TTP (38 vs. 29 weeks; p = 0.124), median OS (75 vs. 60 weeks) and 1year survival rates (72% vs. 60%) were longer in patients who received maintenance paclitaxel. During MT, 86% of patients in the paclitaxel group reported at least one adverse experience, and 45% reported at least one grade 3/4 adverse event (AE). In a second study by the same authors, patients with previously untreated stage IIIB/IV NSCLC were randomly assigned to either arm 1 (carboplatin every 4 weeks with weekly paclitaxel) or arm 2 (carboplatin and paclitaxel, both administered every 3 weeks) [27]. Following 4 cycles of therapy, patients with response or SD were eligible to receive MT with weekly paclitaxel until progression. For patients included in the MT phase (n = 141), the median number of paclitaxel doses was 10 (1–51) for arm 1 and 9 (1–80) for arm 2. Median TTP was 33 weeks and 29 weeks, respectively for arms 1 and 2 for those patients receiving MT compared with

342

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

Table 2 Phase III trials of continuation maintenance therapy. Trial/year Belani et al. [26]

Belani et al. [27]

Brodowicz et al. [28]

Belani et al. [29]

Perol et al.c [30,31]

Paz-Ares et al. [32]

Induction therapy

Maintenance therapy 70 mg/m2

Paclitaxel/carboplatin: random assignment to one of three regimens × 16 weeks (n = 401) Paclitaxel/carboplatin: random assignment to weekly (arm 1) or every 3 weeks paclitaxel (arm 2) × 4 cycles (n = 444)

Paclitaxel weekly for 3 of 4 weeks vs. Observation (n = 130) Paclitaxel 70 mg/m2 weekly for 3 of 4 weeks vs. Observation (n = 141)

Cisplatin 80 mg/m2 on day 1 + gemcitabine 1250 mg/m2 on days 1 and 8 every 3 weeks × 4 cycles (n = 352) Carboplatin AUC5 on day 1 + gemcitabine 1000 mg/m2 on days 1 and 8 every 3 weeks × 4 cycles (n = 519)

Gemcitabine 1250 mg/m2 on days 1 and 8 every 21 days + BSC vs. BSC (n = 206) Gemcitabine 1000 mg/m2 on days 1 and 8 every 21 weeks + BSC vs. BSC (n = 255)

Cisplatin 80 mg/m2 on day 1 + gemcitabine 1250 mg/m2 on days 1 and 8 every 3 weeks × 4 cycles (n = 834) Cisplatin 75 mg/m2 on day 1 + pemetrexed 500 mg/m2 on day 1 every 3 weeks × 4 cycles (n = 979)

Gemcitabine 1250 mg/m2 days 1 and 8 every 3 weeks (n = 154) vs. Observation (n = 155) Pemetrexed 500 mg/m2 on day 1 every 3 weeks + BSC (n = 359) vs. Placebo + BSC (n = 180)

Median PFS 38

weeksa

Median OS

Grades 3–4 toxicities

75 weeks

All grades 3–4 toxicities: 45% for paclitaxel maintenance

29 weeks (p = 0.124)

60 weeks (p = NR)

Median TTP 33 weeks (arm 1) vs. 29 weeks (arm 2)

76 weeks 29 weeks (p = NR)

Median TTP 12 weeks (arm 1) vs. 11 weeks (arm 2) (p = NR) 6.6 months

13 months

5 months (p < 0.001)

11 months (p = 0.195)

3.9 monthsb

8 monthsb

3.8 months (HR = 1.09; p = 0.575)

9.3 months (HR = 1.97; p = 0.838)

3.8 months

NR

1.9 months (HR = 0.55; p < 0.001)

NR (HR = 0.86; 95% CI: 0.66–1.12)

3.9 months

NR

2.6 months (HR = 0.64; p = 0.00025)

NR

Maintenance paclitaxel: neutropenia 2.1%, neurophaty 2.1%, arthralgia 2.1%, fatigue 2.8%, dyspnea 2.1% Maintenance gemcitabine: neutropenia 14.9%, anemia 2.6%, thrombocytopenia 1.7%, emesis 0.8% Neutropenia 13.3%, anemia 9.4%, thrombocytopenia 9.4%, fatigue 3.9% Neutropenia 1.6%, anemia 2.4%, thrombocytopenia 1.4%, fatigue 1.6% At least 1 grade 3/4 AE: gemcitabine arm 27.9%, observation 2.6%

Neutropenia 3.6%, anemia 4.5%, fatigue 4.2% Neutropenia 0%, anemia 0.6%, fatigue 0.6%

PFS: progression-free survival; OS: overall survival; TTP: time to progression; NR: not reported; BSC: best supportive care; HR: hazard ratio; CI: confidence interval. a TTP and OS were adjusted for 16 weeks of initial treatment for the study of Belani et al. [26]. b PFS was calculated from first-line therapy in the study by Belani et al. [29]. In other studies, TTP and PFS were reported from time to randomization. c Three-arm trial, of 834 patients enrolled 465 randomized. Results of erlotinib arm included in Table 4.

12 weeks (arm 1) and 11 weeks (arm 2) for patients not eligible for MT (n = 261) or who chose not to receive it (n =n = 42). Median OS was 76 weeks (MT) vs. 29 weeks (no MT). One and 2-year survival rates were 69% and 33% (MT) vs. 25% and 9% (no MT), respectively. Improvements in survival were associated with minimal toxicity. Grade 4 neutropenia was noted in 2.1% of the patients during the maintenance phase, as well as grade 3 neuropathy. Despite these interesting results, no definitive conclusions can be drawn regarding the role of low-dose weekly maintenance paclitaxel because neither of the two studies was designed with this primary endpoint. So, the role of this drug as MT is yet to be established in a phase 3 study.

The clinical utility of maintenance gemcitabine has also been investigated in patients with advanced NSCLC [28–31]. The Central European Cooperative Oncology Group (CECOG) conducted a phase III study to evaluate maintenance gemcitabine vs. best supportive care (BSC) for patients who did not experience disease progression after first-line treatment with 4 cycles of gemcitabine and cisplatin [28]. The primary objective was TTP for the whole study period and OS was a secondary endpoint. Of the 352 patients enrolled, 206 (59%) were randomized to gemcitabine (n = 138) or BSC (n = 68). For the maintenance period, patients received a median of three gemcitabine cycles (range, 0–38) or two cycles of BSC. TTP was significantly longer for patients who

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

received MT both throughout the entire study (6.6 months vs. 5 months) and during the maintenance period (3.6 vs. 2 month) (p < 0.001 for both). There was also a non statistically significant trend toward longer OS throughout the entire study in the maintenance gemcitabine group. A subset analysis of good and poor PS patients was performed for OS. Patients with good PS (PS > 80, n = 99) experienced a statistically significant worse survival in the BSC arm compared with the gemcitabine arm (median OS, 8.3 vs. 22.9 months; HR = 2.1; 95% CI: 1.2–3.8), while patients with a poor PS (PS ≤ 80, n = 107) experienced similar OS from time of randomization in the BSC arm compared with the gemcitabine arm (median OS, 7.7 vs. 7.0 months; HR = 0.8; 95% CI: 0.5–1.3). Overall, toxicity was mild, with neutropenia representing the most frequent treatment-related complication. Maintenance CT did not result in QoL deterioration compared with BSC. Two additional phase III studies with maintenance gemcitabine have been reported at 2010 ASCO Annual Meeting [29,30]. In the first study [29], advanced NSCLC patients achieving response or SD after first-line therapy with 4 cycles of carboplatin–gemcitabine were randomized 1:1 to receive maintenance gemcitabine + BSC or BSC alone until disease progression. The primary endpoint was the comparison of OS between the two arms and the secondary endpoint was PFS. Patients were required to have an Eastern Cooperative Oncology Group (ECOG)-PS of 0–2 at the time of enrollment. A total of 519 patients were enrolled and 255 (50%) not-progressing after induction therapy were included in the maintenance phase. The study closed after 6 years due to slow accrual, at which point 179 of the 238 planned events had occurred. There were no statistically significant differences in PFS or OS between the two arms. In a subgroup analysis, variables associated with worsened survival were PS ≥2 compared with PS = 1 and male gender. In the gemcitabine and BSC arms, a high percentage of patients had a PS of ≥2 (56% and 58%, respectively), and, not surprisingly, only a minority of patients received poststudy therapy (16% and 17%, respectively). MT was well tolerated despite a higher incidence of grades 3–4 toxicity (anemia 9.4% vs. 2.4%, neutropenia 13.3% vs. 1.6%, thrombocytopenia 9.4% vs.1.4%, and fatigue 3.9% vs. 1.6% in maintenance and BSC arms, respectively). These trials suggest that poor PS patients may not be good candidates for continuation maintenance with gemcitabine, but the prolonged survival among patients with a good PS in the trial by Brodowicz et al. [28] suggests that this may be a viable strategy for patients with a good PS. The second three-arm phase III study (IFCT-GFPC 0502) [30,31], however, showed a strong PFS benefit for patients who continued gemcitabine after first-line cisplatin and gemcitabine first-line therapy. Stage IIIB/IV patients with an ECOG PS of 0–1 whose tumours did not progress following four cycles of cisplatin–gemcitabine were randomized to observation or to receive either gemcitabine or erlotinib as MT until disease progression. The trial was not designed to compare the efficacy of maintenance gemcitabine to erlotinib. Of the 834 patients who received induction CT, 464 (56%) were

343

included in the MT phase. Moreover, patients in all treatment arms were offered pemetrexed at the time of disease progression. Since the study was designed before the interaction of pemetrexed with histology was known, 19% of patients in the observation group had squamous cell histology. The data related to the erlotinib arm will be discussed later in the EGFR TKI maintenance section. PFS by independent review, the primary endpoint of the study, was significantly prolonged by gemcitabine (3.8 vs. 1.9 months; HR = 0.55; p < 0.001) vs. observation. The OS data are not mature, but no significant difference in OS has been observed with 69.6% of events having occurred (HR = 0.86; 95% CI: 0.66–1.12). Impressively, 60% and 76% of patients randomized to gemcitabine or observation received second-line pemetrexed therapy, respectively, which is the highest salvage therapy rate reported in the maintenance studies and balances much better the two arms in terms of overall therapies delivered. The median number of cycles was three in both arms and the ORR to pemetrexed observed in the gemcitabine and observation arms was 8.1% and 15.2%, respectively. MT was well tolerated, although grades 3–4 treatment-related AEs were most common with gemcitabine (27%) than with observation (2%). Finally, pemetrexed continuation maintenance following pemetrexed–platinum first-line therapy has recently been evaluated in the PARAMOUNT phase III trial [32]. In this double-blind, placebo-controlled trial, 979 patients with advanced nonsquamous NSCLC and an ECOG-PS of 0/1 participated in the induction phase, specified as four cycles of cisplatin 75 mg/m2 and pemetrexed 500 mg/m2 on day 1 of a 21-day cycle. Patients who had not progressed during cisplatin–pemetrexed first-line therapy were randomized (2:1, stratified for disease stage, PS and response to induction treatment) to maintenance pemetrexed 500 mg/m2 on day 1 of a 21-day cycle +BSC (n = 359) or placebo+ BSC (n = 180) until disease progression. The primary endpoint was PFS. Patient characteristics were balanced between arms. The trial met its primary endpoint by showing a 36% reduction in the risk of progression (HR = 0.64; 95% CI: 0.51–0.81; p = 0.00025). The median independently reviewed PFS, measured from randomization, was 3.9 months (95% CI: 3–4.2) on the pemetrexed arm, and 2.6 months (95% CI: 2.2–2.9) on the placebo arm. PFS benefit was seen across all subgroups. The disease control rate (DCR) was 71.8% on the pemetrexed arm, and 59.6% on the placebo arm (p = 0.009). The study was fully powered for OS, but OS data were not yet mature at data cut-off. Pemetrexed MT was well tolerated. The drug-related serious AE rate was 8.9% on the pemetrexed arm, and 9.2% of patients had grade 3/4 laboratory AEs. On the placebo arm, the rates were 2.8% and 0.6%, respectively. The main differences were in terms of grade 3–4 fatigue (4.2% vs. 0.6%), anemia (4.5% vs. 0.6%) and neutropenia (3.6% vs. 0%). Discontinuation due to AEs was 5.3% on the pemetrexed arm and 3.3% on the placebo arm. In conclusion, none of the previously described phase III studies that have investigated continuation MT with CT agents have demonstrated an improvement in OS in the

344

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

Table 3 Maintenance treatment with sequential noncross-resistant chemotherapeutic agents. Trial/year

Induction therapy

Maintenance therapy

Median PFS

Median OS

Grades 3–4 toxicities

Westeel et al. [35]

Mitomycin + ifosfamide + cisplatin (MIC) × 2–4 cycles (n = 573)

Vinorelbine 25 mg/m2 weekly for 6 months vs. Observation (n = 181)

5 months

12.3 months

3 months (HR = 0.77; p = 0.11)

12.3 months (HR = 1.08; p = 0.65)

GIP vs. Paclitaxel 225 mg/m2 every 3 weeks until disease progression (n = 281)

4.4 months

11.9 months

4 months (p = 0.56)

9.7 months (p = 0.17)

Maintenance vinorelbine: 45.9% leucopenia, 12.6% infection, 9.1% anemia, 7.4% peripheral neuropathy Leucopenia 77.5%, thrombopenia 42.8%, infection 7.9%

Immediate docetaxel 75 mg/m2 every 3 weeks (maximum 6 cycles) Delayed docetaxel 75 mg/m2 every 3 weeks (maximum 6 cycles) (n = 309)

5.7 months

12.3 months

2.7 months (p < 0.001)

9.7 months (p = 0.0853)

Pemetrexed 500 mg/m2 every 3 weeks + BSC vs. BSC

4.0 months

13.4 months

2.0 months (HR = 0.60; p < 0.001)

10.6 months (HR = 0.79; p = 0.012)

Sculier et al. [60]

Fidias et al. [37]

Ciuleanu et al. [38]

Gemcitabine + ifosfamide + cisplatin (GIP) × 3 cycles (n = 485)

Cisplatin + gemcitabine × 4 cycles (n = 563)

Platinum-based doublet × 4 cycles (n = 745)

Leucopenia 71%, thrombopenia 15.8%, infection 3.6% Total 47.4%, fatigue 9.7%, neutropenia 27.6% Total 47.9%, fatigue 4.1%, neutropenia 28.6% Total 16%, fatigue 5%, anemia 3%, neutropenia 3% Total 4%, fatigue < 1%, anemia < 1%, neutropenia < 1%

PFS: progression-free survival; OS: overall survival; HR: hazard ratio; BSC: best supportive care.

intention-to-treat population. The clinical trial design not focusing to assess the feasibility of MT [26,27], the inclusion of a high percentage of patients with poor PS [29] or the use of PFS as the primary endpoint [28,30–32] may explain the lack of a survival benefit.

3. Maintenance treatment with different agents from those used in the first-line regimen: switch-maintenance or early second-line therapy Switch-maintenance treatment involves the administration of an agent with established activity in advanced NSCLC, immediately after completion of initial CT. In a sequential regimen, the switch from one treatment to another does not require documented progression, in contrast to the switch from the first to second-line treatment. Drugs are given as planned, even in the presence of a response with no signs of resistance at the end of the first part of the sequence. Delivering different CT regimens for induction and maintenance treatment is based on the non-cross-resistant regimens

theory, supported by the Goldie and Coldman hypothesis of drug resistance and mathematical models derived from laboratory experiments [33,34]. According to these theories, even the smallest detectable cancers contain at least one drug-resistant clone and the appearance of resistant clones depends on spontaneous mutations and emerges as tumours grow and progress. Therefore, the early use of non-cross resistant antineoplastic agent might increase the probability of destroying more cancer cells before chemoresistance arises. This strategy has been used in the design of recent maintenance trials in advanced NSCLC, in which patients are treated with one schedule of induction CT and then switched to an alternative, non-cross resistant agent if they respond or remain stable on initial treatment. Results from clinical trials that evaluated early second-line treatment with chemotherapeutic agents are summarized in Table 3. At least two large trials that used older regimens for the first-line CT failed to demonstrate a survival benefit when a third generation agent was used for sequential, noncrossresistant maintenance [35,36].

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

In a French phase III trial that included 573 patients with advanced NSCLC, initial treatment consisted of two monthly mitomycin, ifosfamide and cisplatin (MIC) cycles followed by thoracic radiotherapy for patients with stage IIIB NSCLC and four monthly MIC cycles for those with “wet” stage IIIB or stage IV NSCLC [35]. Of 227 patients who responded or stabilized on induction treatment, 181 (32%) were randomized to receive maintenance intravenous weekly vinorelbine for 6 months or BSC. The primary endpoint was to compare OS between the two groups. The median duration of vinorelbine CT was 13.8 weeks, and only 23% of patients completed the full 6 months of treatment. The most common reasons for stopping CT prematurely were progressive disease (38%) and treatment toxicity (21%). Maintenance vinorelbine did not improve the OS or PFS times. Median PFS was 5 months in the vinorelbine group and 3 months in the observation group (HR = 0.77; p = 0.11) and the median survival from the date of randomization was 12.3 months in both the vinorelbine and observation groups (HR = 1.08; p = 0.65). As typically reported with single agent vinorelbine, the main toxicity in the maintenance group was hematologic. The relatively small number of patients in each of the treatment arms and the modest activity of single-agent vinorelbine in the second-line setting may have contributed to these negative results. Sequential CT consisting of a cisplatin-based regimen followed by paclitaxel also failed to demonstrate a survival benefit in patients with advanced NSCLC compared with cisplatin-based CT reserving paclitaxel as salvage treatment [36]. A total of 485 CT-naïve patients were initially treated with three courses of GIP (gemcitabine + ifosfamide + cisplatin) and those with nonprogressing tumours (n = 281) were then randomised to further similar courses of GIP (n = 140) or courses of paclitaxel every 3 weeks until disease progression (n = 141). Patients with progressive disease on GIP were crossed over to paclitaxel. The primary objective of the trial was to compare OS between the two groups. The median number of post-randomisation courses was three in both arms. There was no significant difference (p = 0.56) in PFS with a median duration of 4 and 4.4 months in the paclitaxel and GIP arms, respectively. In terms of post-randomisation survival, there was also no statistically significant difference (p = 0.17) with median survival times of 9.7 (95% CI: 7.8–11.6) and 11.9 (95% CI: 9.4–14.3) months for paclitaxel and GIP arm, respectively. Multivariate analysis demonstrated that sex and haemoglobin were independent prognostic factors. After adjustment for these factors, the observed HR was 0.81 (95% CI: 0.63–1.04) in favour of GIP (p = 0.10). Toxicity was tolerable; there was a significantly higher rate of grade III/IV thrombocytopenia with the use of GIP but more alopecia with paclitaxel. In contrast, two contemporary phase III trials which have evaluated approved second-line agents as MT suggest that some patients do benefit from sequential CT with a noncrossresistant agent [37,38].

345

Preclinical and clinical data have shown that taxanes are active in platinum-resistant NSCLC [39,40], indicating that the delivery of these agents using a switch-maintenance strategy after platinum-based induction CT may be a viable option. A phase III study by Fidias et al. compared the relative efficacy and toxicity profile of docetaxel given either immediately after first-line therapy or at disease progression in patients who had stable or responding disease after four cycles of carboplatin and gemcitabine (CG) [37]. This is the only study to date that has examined directly the issue of CT timing, since both groups were scheduled to receive the same regimen. Of the original cohort (n = 563), 398 patients completed the 4 cycles of CG, and 309 (55%) were randomly assigned to immediate (n = 153) or delayed (n = 156) docetaxel (patients given BSC after randomization and the same docetaxel regimen after first evidence of progressive disease) for a maximum of 6 cycles. The trial was designed to detect an improvement in OS of 4 months. Secondary end points were ORR, PFS, toxicity and QoL. The ORR was 31.1% for the CG phase, 11.7% for the immediate docetaxel arm (n = 145) and 11.2% for the delayed docetaxel arm (n = 98). A PFS analysis (from randomization to first evidence of progressive disease or death) showed statistically significant (p < 0.001) longer median PFS in the immediate docetaxel arm (5.7 months) compared with the delayed docetaxel arm (2.7 months). However, although OS trended in favour of immediate docetaxel therapy, results did not reach statistical significance (median 12.3 vs. 9.7 months with delayed CT; p = 0.0853). The toxicity profile of both docetaxel arms was comparable and no significant difference in patients QoL was observed between the treatment arms (improvement: 15.6% for both arms; worsening: immediate 11.0% and delayed 18.4%; overall p = 0.76). Of importance, whereas 145 (94.8%) of the 153 patients randomized to immediate docetaxel received at least one treatment cycle, only 98 patients (62.8%) of 156 patients randomly assigned to the delayed arm actually went on to receive docetaxel at progression. The main reason for not receiving CT in the delayed arm was significant symptomatic deterioration and declining PS at the time of progression that precluded further treatment. Thus, it can be argued that the true benefit with “immediate” docetaxel in this study could be attributed to the higher proportion of patients receiving active therapy in the maintenance setting. This result was supported by a post hoc analysis that documented identical OS of 12.5 months for the 245 patients in the immediate arm and for the 98 patients in the delayed arm who actually receive docetaxel. The frequency and the type of imaging studies used in the control arm (computed tomography evaluations every 3 months) have also been cited as unfavourably delaying detection of disease progression, compared with contemporary practice in the industrialized world. Finally, it has to be noted that although this study was negative, the HR for PFS and OS was fairly consistent with the other switch-maintenance trials [38], suggesting that the lack of statistically significance may relate more to the power of the study than to the lack of effectiveness of the strategy.

346

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

A slightly different trial design was used for the phase III study of pemetrexed as early second-line therapy for advanced NSCLC (JMEN trial) [38]. Pemetrexed has been shown to be noninferior to docetaxel in the second-line treatment of NSCLC and better tolerated [25,41]. The drug has also been shown to be active and well tolerated in the firstline treatment of nonsquamous NSCLC in combination with a platinum analogue [19]. Based on these data, Ciuleanu et al. randomly assigned (2:1) a total of 663 patients who had not progressed after 4 cycles of platinum-based first-line CT (cisplatin or carboplatin plus gemcitabine, docetaxel or paclitaxel) to receive pemetrexed plus BSC or placebo plus BSC until disease progression. The initial therapy did not include pemetrexed. The primary endpoint was PFS and a prespecified analysis by histology was incorporated in the protocol. The median number of maintenance cycles delivered was 5 (range 1–55) in the pemetrexed group and 3.5 (range 1–46) in the placebo group. Of the patients randomized to pemetrexed, 48% received ≥6 cycles and 23% received ≥10 cycles. The ORR (3.4% vs. 0.5%; p = 0.042) and DCR (49.1 vs. 28.9%; p < 0.0001) were significantly higher for pemetrexed than for placebo according to an independent review. MT significantly improved the independently assessed PFS (4 vs. 2 months; HR = 0.60; p < 0.001) and OS (13.4 vs. 10.6 months; HR = 0.79; p = 0.012) compared with placebo. The clinical benefit of pemetrexed was only evident in patients with nonsquamous histology (PFS: 4.4 vs. 1.8 months; HR = 0.47; p < 0.001; OS: 15.5 vs. 10.3 months; HR = 0.70; p = 0.002), whereas the analysis in patients with squamous cell histology showed no improvement in any survival endpoint (PFS: 2.4 vs. 2.5 months; HR = 1.03; OS: 9.9 vs. 10.8 months; HR = 1.07). This differential activity of pemetrexed based upon histology is consistent with observations in trials where pemetrexed was used in both first-line [19] and second-line [25] treatment of advanced NSCLC and may be a result of a higher expression of thymidylate synthase, the primary target of pemetrexed, in squamous NSCLC and a lower expression level in adenocarcinomas, leading to greater sensitivity to pemetrexed in the second group [42,43]. Treatment with pemetrexed was globally well tolerated, with no drug-related deaths. Grades 3–4 toxicities were higher with pemetrexed (16% vs. 4%; p < 0.001), specifically fatigue (5% vs. 0.5%) and neutropenia (2.9% vs. 0%). However, time to symptom worsening analysis favoured pemetrexed maintenance for control of pain (6.1 vs. 4.6 months; p = 0.041) and haemoptysis. Poststudy treatment was given to 51.5% of patients assigned to maintenance pemetrexed (third line) and 67% of those who had received placebo. Pemetrexed was used as poststudy therapy in only 19% of patients on the placebo arm because the study did not mandate that patients on the control arm be crossed over to receive pemetrexed at the time of disease progression. Other second-line therapies included docetaxel (29%), erlotinib (21%) and gefitinib (10%). The discretion granted to investigators in the choice of secondline therapy has been cited as a major limitation of the study because it failed to provide any insight into the possibility

that the benefit of maintenance pemetrexed treatment may be attained at the time of disease progression. Nonetheless, the majority of the patients on the placebo arm received salvage therapy, and given the therapeutic equivalence of the currently available second-line treatments, it is unlikely that this factor unduly influenced the impressive overall results of the study. Based upon the results of this trial, two major regulatory bodies, the European Medicines Agency (EMA) and the United Stated Food and Drug Administration (FDA), have approved pemetrexed in July 2009 as MT for patients with locally advanced or metastatic nonsquamous histology, for whom disease has not progressed after four cycles of a platinum-based first-line CT. Of interest, the EMA specifically noted that first-line treatment should be a platinum doublet with gemcitabine, paclitaxel or docetaxel.

4. Maintenance therapy with molecularly targeted agents Several molecularly targeted agents are approved for the treatment of NSCLC, most notably, agents targeting the epidermal growth factor (EGF) [7,44,45] and the vascular endothelial growth factor (VEGF) signalling pathways [17,20]. In the first line setting, bevacizumab, a monoclonal antibody that targets VEGFR [17,20], and cetuximab [18,46], an antibody that targets the EGFR, have been studied in large randomized trials. Modest improvements in ORR, PFS and OS have been demonstrated in most studies with the addition of these agents to platinum-based CT doublets. All studies were designed to continue maintenance antibody therapy in responding and stable patients after completion of first-line CT. Whether this is necessary, or whether it simply adds the cost and the potential for toxicity, is unknown at this time, because there have been no trials in which responding and stable patients were randomized either to stop or to continue treatment. The EGFR TKIs gefitinib and erlotinib have been shown to prolong survival when administered as first-line [7] or salvage therapy in advanced NSCLC [44,45]. On the basis of the established activity of these agents in NSCLC and the well-tolerated toxicity profile, they are potentially ideal candidates for MT. The cytostatic properties of targeted therapies could work effectively in patients with minimal disease and maintain tumour regression after an initial response to CT. Two different strategies have been investigated: MT with targeted agents already present in the induction phase and sequential treatment with different targeted agent from those used in the induction regimen. 4.1. Maintenance treatment with targeted agents already present in the induction phase The inhibition of angiogenesis is one of the targeted approaches most widely studied in the treatment of advanced NSCLC [47]. Among the angiogenesis inhibitors, the antiVEGF monoclonal antibody bevacizumab represents the

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

most successful targeted therapy. In fact, the addition of bevacizumab to platinum-based CT as first line treatment has demonstrated to offer tangible clinical benefit for bevacizumab-eligible patients with advanced nonsquamous NSCLC in two large phase III trials [17,20]. Both studies met their primary endpoints. ECOG 4599 compared carboplatin and paclitaxel with and without bevacizumab and demonstrated an improvement in PFS (6.2 months vs. 4.5 months in the bevacizumab and placebo groups, respectively; HR = 0.66; 95% CI: 0.57–0.77; p < 0.001) and OS (12.3 months vs. 10.3 months in the bevacizumab and placebo groups, respectively; HR = 0.79; 95% CI: 0.67–0.92; p = 0.003) with the addition of bevacizumab [17]. The Avastin in Lung Cancer (AVAIL) study compared cisplatin, gemcitabine and placebo to the combination with bevacizumab 7.5 mg/kg or 15 mg/kg [20]. PFS was significantly longer in both bevacizumab treatment arms compared with the placebo arm, with HRs for PFS of 0.75 (median PFS, 6.7 vs. 6.1 months for placebo; p = 0.003) in the low-dose group and 0.82 (median PFS, 6.5 vs. 6.1 months for placebo; p = 0.03) in the high-dose group compared with placebo. Median OS, a secondary endpoint, in both bevacizumab study arms was similar to that in the CT alone arm. In both studies, bevacizumab continued in responding and stable patients as MT after six cycles of first-line CT until disease progression or unacceptable toxicity. Given that VEGF is a key proangiogenic mediator, continued inhibition of VEGF is essential to prevent tumour revascularization and/or neovascularization, which are required for further growth. In the ECOG 4599 trial [17], 215 (53%) of the 407 patients starting treatment with CT and bevacizumab continued with bevacizumab monotherapy, and of these patients, 107 (50%) received more than five cycles in the maintenance phase. In the AVAIL trial [20], the percentages of patients completing six cycles of CT and bevacizumab were 49% in the low-dose (7.5 mg/kg) bevacizumab (n = 345) and 45% in the high-dose (15 mg/kg) bevacizumab group (n = 351). Of these, 42% and 41% of patients continued single-agent bevacizumab as MT after completion of induction therapy, respectively. In fact, 94% of patients eligible to receive bevacizumab were still receiving MT at cycle 7. These two trials represent the first evidence of an improvement in outcomes using CT combined with targeted therapies as first-line therapy of advanced NSCLC, and also represent a successful strategy of MT. However, it is not possible to determine from these data whether or what proportion of the survival benefit in the bevacizumab arms might have come from the maintenance phase of the treatment as neither trial was designed to answer this question. Indeed, the entire PFS and OS benefits of bevacizumab may have come from the higher ORR in the bevacizumab arms reported in the studies and may not have been due to the prolonged administration of this agent. Considering the potential for toxicity and the cost of this drug, the issue of MT with bevacizumab is one that deserves further attention. The role of bevacizumab plus pemetrexed as MT in patients with CT-naive, advanced, nonsquamous NSCLC has

347

also been assessed in a phase II study [48]. Patients received pemetrexed, carboplatin and bevacizumab 15 mg/kg every 21 days for up to 6 cycles, and for those with non-progressive disease, pemetrexed and bevacizumab were continued until disease progression or unacceptable toxicity. Fifty patients were enrolled: 30 patients (60%) and 7 (14%) patients completed ≥6 and 18 cycles of therapy, respectively. Forty-nine patients were evaluable for response: 1 complete response (CR) and 26 PR were observed, for an ORR of 55% (95% CI: 41–69%). At a median follow-up of 9.5 months, the median PFS was 7.8 months and the median survival time was 14.1 months. The regimen has a favourable toxicity profile; there were no grade ≥3 hemorrhagic events nor did any patients experience grade 3/4 hypertension. Most serious AEs were observed during the initial six cycles of therapy, with no increase in the relative incidence of AEs typically associated with CT or bevacizumab, even after extended treatment. In view of its promising efficacy and modest toxicity, the regimen of carboplatin, pemetrexed and bevacizumab appears ideal for further development in previously untreated patients with advanced, nonsquamous NSCLC. The EGFR pathway is another promising therapeutic target in NSCLC. Cetuximab, an anti-EGFR immunoglobulin G1 monoclonal antibody, in combination with first-line platinum-based CT and its role as MT has also been evaluated in large phase III trials [18,46,49]. The FLEX (First-Line Erbitux in lung cancer) study compared cisplatin and vinorelbine with and without cetuximab [18]. OS, the primary end-point, was significantly prolonged in the CTplus-cetuximab group compared with the CT-alone group (median 11.3 vs. 10.1 months, respectively; HR = 0.871; 95% CI: 0.762–0.996; p = 0.044). The combination of CT and cetuximab was also better than CT alone in terms of ORR, but PFS time was not different (median 4.8 months in both groups; HR = 0.943; 95% CI: 0.825–1.077; p = 0.39). In the BMS-099 study [46], ORR was significantly higher with the addition of cetuximab to CT (25.7% vs. 17.2%; p = 0.007), while PFS, the primary endpoint, was not significantly different (median 4.4 vs. 4.2 months with CT alone; HR = 0.902; 95% CI: 0.761–1.069; p = 0.236). Median OS was neither significantly prolonged (9.69 months with cetuximab-CT vs. 8.38 months with CT alone; HR = 0.89; 95% CI: 0.754–1.051; p = 0.169), but the difference was similar to that observed in the FLEX trial. In both trials, cetuximab continued after the end of first-line CT in non-progressing patients until disease progression or unacceptable toxicity occurred. In the FLEX study [18], 241 of the 548 patients assigned to the cetuximab arm did not experience disease progression or unacceptable toxicity and were eligible for maintenance cetuximab. The median number of CT cycles was four (range 0–6 for CT plus cetuximab and 1–7 for CT alone) and cetuximab was given for a median duration of 18 weeks (range 1–135). In the BMS-099 study [46], the median number of CT cycles was four in both arms and the median number of cetuximab infusions in the cetuximab/CT arm was 12 (range, 1–125 infusions). As in the bevacizumab trials, it

348

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

is not possible to extrapolate the extent of benefit from the addition to cetuximab vs. the role of MT with this agent. However, it is of interest in the FLEX trial [18] that the survival curves start to separate at the end of first-line CT when cetuximab maintenance begins. This seems to underline a potential survival effect for cetuximab MT. In conclusion, these trials reveal that the continuation of the maintenance monoclonal antibody is feasible and a substantial minority of patients are likely to receive therapy for a extended period of time. To assess the risk and benefits of this strategy accurately, a phase III trial specifically investigating maintenance cetuximab or bevacizumab compared with placebo after first-line platinum-based CT would be required. In addition to monoclonal antibodies, there are a number of oral angiogenesis inhibitors directed against VEGFR TK, which have been investigated in combination with CT [50,51]. All trials performed to date continued the oral angiogenesis inhibitor after treatment with CT in responding and stable patients. No study has demonstrated a significant PFS or OS benefit to date, and so, at this time, no conclusions can be drawn concerning the potential for benefit from MT with this strategy. The ECOG conducted a randomised discontinuation design phase II study (E2501) in third line and beyond disease to assess the role of sorafenib as MT [52]. Three hundred and forty-two patients received sorafenib twice daily for 2 cycles (2 months) and ninety-seven responding and stable patients then were randomized to receive sorafenib or placebo MT. Patients randomized to placebo were allowed to cross over to sorafenib on progression. Two months after randomization, 22% of patients in the placebo group and 35% in the sorafenib group continued to have responding or SD (p = 0.01), with an overall benefit in PFS favouring sorafenib MT (median PFS 3.6 vs. 1.9 months; HR = 0.50; p = 0.01). The HR for OS also favoured sorafenib, even though 61% of patients on the placebo arm crossed over to sorafenib at the time of progression (HR = 0.68; p = 0.15). This is the first study to suggest that maintenance of response with a VEGF TKI may prolong not only PFS but also OS. These results require confirmation in a larger well-powered study. Concurrent administration of small molecules EGFR TKIs, gefitinib and erlotinib, with CT has been evaluated as part of their drug development programs in advanced NSCLC patients, but their addition to platinum-containing CT regimens in the first-line setting have failed to show survival benefit even though EGFR TKIs were continued as maintenance after CT in all trials [53–56]. In two phase III studies, named Iressa NSCLC Trial Assesing Combination Therapy (INTACT)-1 [53] and INTACT-2 [54], patients with unresectable stage III or IV NSCLC received platinum-based CT (cisplatin–gemcitabine in INTACT-1 and carboplatin–paclitaxel in INTACT-2) plus either gefitinib 500 mg/day, gefitinib 250 mg/day or placebo. After a maximum of six cycles, patients without non-progressive disease were maintained on gefitinib or placebo until disease progression or drug intolerance. There was no difference in RR,

TTP or OS between the treatment groups. However, in the INTACT-2 trial, a retrospective subgroup analysis showed a trend toward improved survival for patients with adenocarcinoma who had received CT for ≥90 days (patients would have received at least the median number of CT cycles) in the gefitinib 250 mg/day arm (p = 0.05), suggesting a possible effect of single-agent gefitinib as MT. Similarly, the combination of erlotinib with platinum-based polychemotherapy has been evaluated in two large phase III randomized trials [55,56]. In both studies, patients received erlotinib 150 mg/day or placebo in combination with a maximum of six 21-day cycles of platinum-based CT (cisplatin–gemcitabine in the TALENT trial [55], and carboplatin–paclitaxel in the TRIBUTE [56]), followed by maintenance monotherapy with erlotinib/placebo in the absence of disease-progression. There were no differences in outcome between treatment arms, although patients who reported never smoking experienced improved survival in the erlotinib group in both studies. However, retrospective subgroup analysis of the TRIBUTE trial showed a median OS of 13.6 months with erlotinib vs. 12.2 months with placebo (p = 0.04) for the 861 patients who survived beyond 4 months (408 patients randomized to erlotinib and 453 patients randomized to placebo). Of the 740 patients who survived beyond 6 months (erlotinib 348; placebo 392), median survival with erlotinib was 15.4 vs. 13.8 months for placebo. Likewise, in TALENT, patients treated with erlotinib for more than 150 days showed increased response duration in the erlotinib arm compared with placebo. These results are suggestive of TKI efficacy as maintenance monotherapy after platinum-based CT for selected patients. Preclinical evidence suggests that a possible reason for the lack of success in these studies was the potential antagonism between the constituents of the combinatory therapy. EGFR TKIs induce G1-phase cell cycle arrest, which protects cells from the cytotoxic effects of cell cycle phase-dependent chemotherapeutic agents [57]. In contrast, sequential administration of EGFR TKIs following CT achieves pharmacodynamic separation of these two therapeutic approaches and has been shown to provide greater efficacy than concurrent administration. Thus, the effects of a CT and EGFR TKI combination may be sequencedependent [58,59]. This hypothesis is supported by the results of the FAST-ACT trial (First-line Asian Sequential Tarceva plus Chemotherapy Trial) [59], a placebo-control randomized phase II study in which erlotinib was administered not concurrently with CT, but on days 15–28 of each CT cycle. A total of 154 unselected patients from Asia (94%) and Australia with chemonaïve advanced NSCLC were randomized to receive either erlotinib 150 mg/day or placebo orally on days 15–28 of 4-weekly CT cycles. CT consisted of gemcitabine on days 1 and 8 plus cisplatin or carboplatin on day 1 for a maximum of 6 cycles (GC). Responding patients continued to receive erlotinib or placebo as MT. The ORR was higher in the erlotinib arm (35.5% vs. 24.4% in the placebo arm; p = 0.12). Although the primary endpoint of nonprogression rate (NPR) at 8 weeks was not met (80.3%

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

349

Table 4 Phase III trials of maintenance treatment with sequential noncross-resistant molecularly targeted agents. Trial/year

Induction therapy

Maintenance therapy

Median PFS

Median OS

Grades 3–4 toxicities

Cappuzo et al. [61]

Platinum-based doublet × 4 cycles (n = 1949)

Erlotinib 150 mg/day vs. Placebo (n = 889)

12.3 weeks

12 months

Rash 9%, diarrhea 2%

11 weeks (HR = 0.71; p < 0.001)

Rash 0%, diarrhea 0%

Cisplatin 80 mg/m2 on day 1 + gemcitabine 1250 mg/m2 on days 1 and 8 every 3 weeks × 4 cycles (n = 834) Platinum-based doublet + bevacizumab × 4 cycles (n = 1160)

Erlotinib 150 mg/day (n = 155) vs. Observation (n = 155)

2.9 months

11 months (HR = 0.81; p = 0.0088) 11.8 months

1.9 months (HR = 0.82; p = 0.002)

10.7 months (HR = 0.91; p > 0.05)

Bevacizumab 15 mg/kg/3 weeks + erlotinib 150 mg/day vs. Bevacizumab 15 mg/kg/3 weeks + placebo (n = 768) Observation (n = 298)

4.76 months

15.9 months

3.75 months (HR = 0.72; p = 0.0012)

13.9 months (HR = 0.91; p = 0.26)

4.6 months

13.7 months

Gefitinib 250 mg/day (n = 300)

4.3 months (HR = 0.68; p < 0.001)

12.9 months (HR = 0.86; p = 0.11)

Perol et al.a [30,31]

Miller et al. [62] and Kabbinavar et al. [63]

Takeda et al. [71]

Zhang et al. [72]

Gaafar et al. [73] and Surmont et al. [74]

Platinum-based doublet × 6 cycles vs. Platinum-based doublet × 4 cycles (n = 604) Platinum-based CT × 4 cycles (n = 298)

Platinum-based doublet × 4 cycles (n = 173)

Gefitinib 250 mg/day (n = 148) vs. Placebo (n = 148) Gefitinib 250 mg/day (n = 86) vs. Placebo (n = 87)

4.8 months

18.7 months

2.6 months (HR = 0.42; p < 0.0001) 4.1 months

16.9 months (HR = 0.84; p = 0.26)

2.9 months (HR = 0.61; p = 0.0015)

9.4 months (HR = 0.81; p = 0.204)

10.9 months

Total 15.5%, rash 9%, diarrhea 0.6% Total 2.6%, rash 0%, diarrhea 0% Total 44.1%, rash 10.4%, diarrhea 9.3% Total 30.4%, rash 0.5%, diarrhea 0.8%

Leucopenia 39.9%, anemia 21.8%, rash 0.7%, diarrhea 2% Leucopenia 37%, anemia 13.4%, rash 0.3%, diarrhea 1.7% Total 10%, 0% rash, 0% diarrhea, 2% ALT increase Total 0% Total 19%, rash 5.1%, diarrhea 2.1% Total 0%, rash 0%, diarrhea 0%

PFS: progression-free survical; OS: overall survival; HR: hazard ratio. a Three-arm trial, of 834 patients enrolled 465 randomized. Results of gemcitabine arm included in Table 2.

in the GC-erlotinib arm and 76.9% in the GC-placebo arm; p = 0.51), patients in the erlotinib plus CT arm had a higher NPR at 16 weeks (64.5% vs. 53.4%; p = 0.14) and obtained a significant 53% improvement in PFS compared with those in the placebo plus CT arm (median 29.4 vs. 23.4 weeks; p = 0.0002). Planned subgroup analysis showed that the PFS benefit was consistent across all clinical subgroups. OS was similar between the two arms (median 74.1 weeks for GCerlotinib vs. 75.7 weeks for GC-placebo; p = 0.42), as well as overall safety profiles. The high tumour ORR in the erlotinib arm suggested the benefit of sequential administration. The separation of the Kaplan–Meier curves for PFS that began during the sequential combination phase of treatment was sustained during weeks 24–48, suggesting that MT with erlotinib also contributes to the overall PFS observed. This study allows for a novel strategy to treatment using intercalated doses of EGFR TKI with CT for advanced-NSCLC patients.

4.2. Sequential treatment with different targeted agents from those used in the first-line regimen The strategy of early second-line treatment has also been tested for targeted therapies. The EGFR TKIs, gefitinib and erlotinib, represents the most studied agents in this setting. Phase III trials of MT with sequential non-crossresistant molecularly targeted agents are summarized in Table 4. Based on its efficacy in advanced NSCLC, the EGFR TKI erlotinib has been evaluated as switch-maintenance therapy in three randomized phase III studies: Sequential Tarceva in Unresectable Lung Cancer (SATURN) [61], ATLAS [62,63] and IFCT-GFCP 0502 [30,31]. The SATURN study evaluated the efficacy and safety profile of erlotinib MT in patients with non-progressive disease after treatment with at least 4 cycles of standard first-line platinum-based doublet CT [61]. The coprimary end-points were PFS in all patients and EGFR immunohistochemistry (IHC)-positive

350

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

(IHC+) patients. A total of 1949 patients initiated therapy with platinum-based CT and 889 (45%) of them did not experience progressive disease and were randomized to erlotinib (n = 438) or placebo (n = 451). Best tumour ORR was 11.9% with erlotinib vs. 5.4% with placebo (p = 0.0006) and DCR (CR + PR + SD > 12 weeks) was 40.8% with erlotinib vs. 27.4% with placebo (p < 0.0001). PFS was significantly prolonged with erlotinib vs. placebo in all analysable patients, irrespective of EGFR status (median PFS 12.3 vs. 11.1 weeks; HR = 0.71; 95% CI: 0.62–0.82; p = 0.000003) and in EGFR IHC+ patients (median PFS 12.3 vs. 11.1 weeks; HR = 0.69; 95% CI: 0.58–0.82; p = 0.00002). More importantly, OS was significantly prolonged with erlotinib vs. placebo in the intention-to-treat (ITT) population (median 12 vs. 11 months; HR = 0.81; 95% CI: 0.70–0.95; p = 0.0088) and the EGFR IHC-positive population (HR = 0.77; 95% CI: 0.64–0.93; p = 0.0063). Patients who had SD after firstline CT seemed to have more pronounced OS benefit with maintenance erlotinib (median 11.9 vs. 9.6 months with placebo; HR = 0.72; 95%; CI: 0.59–0.89; p = 0.0019) than those who had a previous CR or PR (median 12.5 vs. 12 months with placebo; HR = 0.94; 95% CI: 0.74–1.20; p = 0.618) [64]. Interestingly, pre-planned subgroups analysis of PFS and OS by clinical characteristics suggested improved outcome with erlotinib compared with placebo regardless of gender, ethnic origin, PS or smoking habit. Among patients with squamous histology (n = 360), patients in the erlotinib arm experienced a statistically significantly longer PFS (HR = 0.76; 95% CI: 0.60–0.95) compared with placebo, but did not experience a statistically significant difference in OS (HR = 0.86; 95% CI: 0.68–1.10). Patients with adenocarcinoma histology (n = 401) treated with maintenance erlotinib experienced a statistically significant improvement in both PFS (HR = 0.60; 95% CI: 0.48–0.75) and OS (HR = 0.77; 95% CI: 0.61–0.97). Subgroup analysis according to EGFR mutation status showed a PFS benefit both in patients with EGFR-activating mutations (HR = 0.10; 95% CI: 0.04–0.25; p < 0.001) and in those with wild-type EGFR (HR = 0.78; 95% CI: 0.63–0.96; p = 0.0185). OS was significantly prolonged in patients whose tumours did not harbour activating EGFR mutations (HR = 0.77; 95% CI: 0.61–0.97; p = 0.0243). However, the PFS benefit seen for patients with EGFR mutation-positive tumours did not translate into an equally impressive OS benefit, probably due to the high degree of censoring (most patients have not yet experienced an event) and the 67% cross-over rate to second-line EGFR TKI therapy in the placebo group. Moreover, an important caveat to the survival data is the low rate of subsequent EGFR TKI use in the placebo arm (16%). Because EFGR TKIs have known therapeutic benefit, the lack of crossover in this trial reinforces the data showing that erlotinib is an effective agent in NSCLC, but does not definitively prove that the modest improvements in PFS and OS are from the switch maintenance strategy. A prospective biomarker analysis of the SATURN trial has also evaluated the influence on outcome of EGFR protein expression using IHC (positive in 84%),

EGFR gene copy number by fluorescent in situ hybridization (FISH) (positive in 48%), K-ras mutations (positive in 18%) and EGFR exon 19 deletions and/or L858R mutations (positive in 11% of samples) [65]. A PFS benefit was observed with erlotinib in all biomarker subgroups analyzed. Only patients with EGFR sensitizing mutations in exons 19 or 21 derived significantly greater PFS benefit with maintenance erlotinib compared with those patients with EGFR wild-type tumours, with a treatment by mutation interaction test highly significant (interaction p < 0.001). In contrast, patients with K-ras mutations obtained a similar clinical benefit from erlotinib to those with wild-type K-ras (HR = 0.77 in K-ras mutation positive and 0.70 in K-ras wild-type; p = 0.9480). Erlotinib was well tolerated in the maintenance setting, with a tolerability profile similar to that seen in the phase 3 BR.21 study [44]. The most common reported treatment-related AEs were rash (60% with erlotinib vs. 9% with placebo) and diarrhea (20% with erlotinib vs. 5% with placebo), with most reported to be of grade 1/2. Although there was no significant difference for time to deterioration in QoL in the two arms (HR = 0.96; 95% CI: 0.79–1.16), there was an improvement in time to pain and time to analgesic use in favour of the erlotinib arm (HR = 0.61 and 0.66, respectively) [66]. Authors conclude that the SATURN study is the first trial to show that MT with a targeted agent following conventional CT can significantly prolong PFS and OS in advanced NSCLC across patients subgroups and regardless of the status of the most commonly studied biomarkers. Based on this trial, maintenance erlotinib was recently approved by the FDA in all patients who do not experience disease progression after platinum-based therapy. The EMA approved maintenance erlotinib only for patients with SD after 4 cycles of standard platinum-based first-line therapy. Another relevant phase III study, named ATLAS, was designed to build on the use of bevacizumab as MT for patients who receive this agent in combination with CT for first-line treatment [62,63]. On the basis of the preclinical and clinical evidence in support of a favourable interaction between EGFR and VEGFR signalling pathways in tumorigenesis [67–69], the ATLAS study sought to determine whether the combination of bevacizumab and erlotinib is more effective than bevacizumab alone when used for MT. A total of 1160 patients were enrolled for first-line therapy with a platinum-based doublet (choice of CT at investigators’ discretion) and bevacizumab 15 mg/kg every 3 weeks. Patients who had not experienced disease progression (n = 768, 66%) at the end of four cycles and had no significant toxicity were continued on bevacizumab and randomly assigned to either erlotinib (n = 373) or placebo (n = 370). ATLAS is therefore unique among these trials, since its design incorporates standard bevacizumab maintenance in both arms. Although bevacizumab should generally be used with caution in patients with squamous cell carcinomas, central nervous system metastases, and/or those with an increased risk of bleeding, such patients were eligible for study participation if they had peripheral and/or extra-thoracic squamous tumours,

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

if brain metastases had been treated, or they were undergoing anticoagulation with a low-molecular-weight heparin. The safety profile for bevacizumab + erlotinib was consistent with known profiles for the two agents separately, although patients in the bevacizumab and erlotinib arm compared with bevacizumab and placebo experienced a higher rate of rash (10.4% and 0.5%, respectively) and diarrhea (9.3% and 0.8%, respectively). PFS, the primary endpoint, was significantly longer in the bevacizumab + erlotinib arm (4.76 vs. 3.75; HR = 0.722; 95% CI: 0.592–0.881; p = 0.0012). The 3-and 6-months PFS rates were 67.7% and 40.3% for the combination maintenance group, compared with 53.4% and 28.4% for the placebo arm. Pre-planned subgroups analysis showed that the PFS benefit with bevacizumab + erlotinib MT was seen regardless of age, gender, ethnicity, smoking habit or histology. Biomarker status was available in 367 patients and, similar to SATURN, evaluated the role of EGFR IHC (positive in 52% of available samples), EGFR FISH (positive in 23.7%), K-ras mutation (positive in 25%) and EGFR mutation (positive in 14%) on PFS [70]. Patients who were EGFR FISH positive, EGFR mutated and K-ras wild-type enjoyed the greatest benefit from therapy with erlotinib and bevacizumab in the maintenance setting (HR = 0.66, 0.44 and 0.67, respectively). In contrast to the PFS benefit, OS showed a nonsignificant 2-month difference in favour of the experimental group (15.9 vs. 13.9 months; HR = 0.91; 95% CI: 0.80–1.04; p = 0.2686). The high baseline OS in the control arm, the highest reported in maintenance studies could be explained for patient selection or effect of bevacizumab in the maintenance setting. Potentially, the addition of another maintenance targeted agent does not confer any additional benefit. Alternatively, based on the number of patients randomly assigned and the high survival attained by the bevacizumab only group, the study might not have the power to detect OS differences. It has to be noted that the study closed after a second planned interim efficacy analysis confirmed the PFS superiority in the erlotinib arm, which limits the interpretation of the post hoc analysis. Moreover, this study did not address the fundamental question of how much additional benefit is derived from the use of bevacizumab as MT in patients with nonsquamous NSCLC treated with bevacizumab-containing induction regimen. Although recognizing the perils of cross-trial comparison, the 12-month median OS with single-agent erlotinib maintenance in the SATURN trial and 15.9-month median OS with the combination of erlotinib and bevacizumab in the ATLAS trial, highlight the importance of establishing the relative contribution of each agent when a combination therapy strategy is being evaluated in the maintenance setting. Finally, the previously described French multicenter trial IFCT-GFPC 0502 compared erlotinib to observation in patients with non progressive disease following four cycles of cisplatin–gemcitabine [30,31]. As mentioned above, patients were mandated to receive pemetrexed on progression, and indeed a large portion of these patients was treated (pemetrexed, 76.1%; erlotinib, 3.1%; and docetaxel, 1.9%). Of

351

the 834 patients who were enrolled, 464 were randomized and 155 were assigned to the erlotinib and observation arms. Baseline characteristics were balanced. Similar to the previous studies, PFS was significantly prolonged with maintenance erlotinib (2.9 vs. 1.9 months; HR = 0.82; p = 0.002) and benefit was seen irrespective of histology, smoking status or gender. Preliminary OS data showed a non-significant improvement with erlotinib vs. observation (11.8 vs. 10.7; HR = 0.91; 95% CI: 0.80–1.04) [30,31]; however, OS was improved with erlotinib vs. observation in patients receiving second-line pemetrexed (HR = 0.82; 95% CI: 0.69–0.96). Results based on EGFR mutation status are not currently available. All these data confirm that MT with erlotinib achieves promising results, in contrast to the data previously observed for EGFR TKIs when administered concurrently with CT. The role of maintenance gefitinib has also been assessed in three randomized phase III trials in advanced NSCLC [71–74]. The West Japan Thoracic Oncology Group conducted a randomized phase III study (WJTOG0203) to evaluate whether gefitinib improves survival as sequential therapy after platinum-doublet CT [71]. A total of 604 Asian CT-naïve NSCLC patients were randomly assigned to receive either platinum-doublet CT up to six cycles (arm A) or three cycles of platinum doublet followed by gefitinib 250 mg/day until disease progression (arm B). The median number of cycles was three in both arms, and 57.3% (n = 172) of patients in the sequential arm were treated with gefitinib. The median treatment duration of gefitinib was 69.5 days, and the maximum treatment duration was 1324 days. Toxicity results were consistent with previous Japanese studies of advanced NSCLC patients who received a platinum-doublet regimen. Furthermore, no significant severe AEs were seen that were not predictable from the safety profiles of gefitinib in sequential therapy after CT. The ORR was 29.3% for CT alone and 34.2% for CT followed by gefitinib. There was no significant difference between treatment arms (p = 0.20). The overall DCR were 71.0% and 75.5% in arm A and in arm B, respectively. There was a statistically significant improvement in PFS with gefitinib (4.6 vs. 4.3 months with observation; HR = 0.68; 95% CI: 0.57–0.80; p < 0.001). However, OS results did not reach statistical significance (13.7 vs. 12.9 months; HR = 0.86; 95% CI: 0.72–1.03; p = 0.11). Of the patients in the CT alone and sequential arms, 54.5% and 75.2%, respectively, received EGFR TKI therapy at some point. In an exploratory subset analysis of OS by histologic group, patients in arm B with adenocarcinoma did significantly better than patients in arm A (HR = 0.79; 95% CI: 0.65–0.98; p = 0.03). Another subset of smokers had a survival advantage with the sequential strategy (HR = 0.79; 95% CI: 0.64–0.98), whereas there was no difference between the two treatment groups in the subset of never smokers (HR = 0.94; 95% CI: 0.66–1.33). Patients were not selected on the basis of the target EGFR mutation status, because when this study was planned, the EGFR mutation as a predictive factor of response to gefitinib had not been recognized.

352

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

A second Asian phase III trial (INFORM, C-TONG 0804) to investigate the efficacy and safety of gefitinib as MT in patients with advanced NSCLC has been reported at the 2011 ASCO Annual Meeting [72]. Patients from 27 centers across China with an ECOG-PS of 0–2 were treated with standard first-line platinum based CT and those without progression/unacceptable toxicity after four cycles (n = 296) were randomized 1:1 to gefitinib 250 mg/day or placebo until disease progression. PFS was the primary endpoint. Secondary endpoints included OS, ORR, DCR, symptom improvement and tolerability. Patient characteristics were balanced between arms; overall, 54.1% patients were never-smokers, 70.6% had adenocarcinoma and 40.9% were female. Tissue samples of 79 patients (27%) were evaluable for EGFR mutations. The overall EGFR mutation-positive rate was 38%. A total of 8.1% of patients in the gefitinib arm and 31.8% in the placebo arm received subsequent therapies with targeted agents. The ORR was 23.6% in the gefitinib arm and 0.7% in the placebo arm (p = 0.0001). The overall DCR were 71.6% and 50.7% in the gefitinib and placebo arms, respectively. The study met its primary endpoint and demonstrated a statistically significant increase in PFS with gefitinib vs. placebo (median PFS 4.8 vs. 2.6 months; HR = 0.42; 95% CI: 0.32–0.54; p < 0.0001). The greatest magnitude of effect on PFS was observed in patients with EGFR mutationpositive tumours (median PFS 16.6 months in the gefitinib vs. 2.8 months in the placebo arm; HR = 0.17; 95% CI: 0.07–0.42), but no significant PFS benefit was found for those patients with EGFR mutation-negative tumours (median PFS, 2.7 vs. 1.5 months; HR = 0.86; 95% CI: 0.48–1.51). The difference in OS between treatments was not statistically significant (median OS, 18.7 vs. 16.9 months; HR = 0.84; 95% CI: 0.62–1.14; p = 0.2608). Gefitinib was well tolerated. Most common AEs (any grade) were rash (49.7%), diarrhea (25.2%), and alanine aminotransferase increase (21.1%) which were generally mild/moderate. Overall incidence of serious AEs was 6.8% in the gefitinib arm and 3.4% in the placebo arm. The European Organization for the Research and Treatment of Cancer (EORTC) 08021-ILCP trial was similar in design to the SATURN study [73,74]. Initially, all patients with advanced NSCLC not-progressing after 4 cycles of platinum-based CT were randomized to receive either gefitinib 250 mg/day or matched placebo until progression or unacceptable toxicity. The study was designed to detect a 3month increase in OS. The protocol was amended to require evidence of EGFR protein expression by IHQ. This resulted in a slowing of recruitment, which ultimately led to study closure before it reached its target accrual. The continuous administration of gefitinib following platinum-based CT is well tolerated. After inclusion of 173 patients and based on 149 of the required 514 deaths, PFS was significantly different in favour of gefitinib (4.1 vs. 2.9 months; HR = 0.61; 95% CI: 0.45–0.83; p = 0.0015), but no difference in OS could be detected (median OS 10.9 vs. 9.4 months; HR = 0.83; 95% CI: 0.60–1.15; p = 0.204). Exploratory subgroup analysis of

PFS showed benefit for gefitinib regardless of stage, EGFR IHQ, smoking status and best CT response. Gefitinib showed benefit in PS 0–1, but yet not in those with PS 2. The small number of accrued patients in the EORTC study, the reduced number of patients who received gefitinib immediately after platinum-based CT and the high rate of subsequent use of this agent in the WJTOG0203 and INFORM trials make an assessment of sequential gefitinib difficult. Data to date do not support the use of this agent in the maintenance setting. In addition to the EGFR TKIs, sequential MT with multitargeted TKIs not administered during the induction CT phase is currently being investigated. Vandetanib is a oncedaily oral inhibitor of VEGFR, EGFR and Rearranged during Transfection (RET) signalling. Several phase III trials have shown the efficacy and tolerability of vandetanib in patients with advanced NSCLC, both as a single agent and in combination with CT [75–77]. In addition to these trials, a double-blind, placebo-controlled phase II study has assessed the role of vandetanib MT in NSCLC [78]. Patients who had not progressed after 4 cycles of platinum-based CT were randomized to vandetanib 300 mg/day or placebo. The primary endpoint was PFS at 3 months. The ITT analysis comprised 117 patients (75 vandetanib and 42 placebo), excluding one patient randomized to vandetanib but not treated. The median duration of follow-up was 5.9 months. Baseline characteristics were balanced in both arms (64% male, 74% PS 1, 65% smokers, 74% adenocarcinoma histology and 63% partial responders to CT). A preplanned interim analysis showed that PFS at 3 months was 29% with placebo, leading this arm to be stopped; PFS at 3 months with vandetanib was 37.5%, allowing this arm to proceed to the second stage. The ITT analysis showed 37.3% of patients in the vandetanib arm achieved PFS at 3 months. Overall, the most common AEs in patients receiving vandetanib were rash (12% grade 3/4), diarrhea, pruritus and anorexia. Based on these results, vandetanib MT should be considered for further study. The Cancer and Leukemia Group B (CALGB) 30607 trial (NCT00693992) is designed to address the potential benefit from MT with sunitinib, an oral, selective multi-targeted TKI. In this randomized phase III study, patients receive four cycles of platinum-doublet CT (bevacizumab is allowed in eligible patients), and responding and stable patients are randomized to receive sequential sunitinib 37.5 mg daily or placebo daily. The primary end point is PFS, and cross-over to sunitinib is allowed at the time of progression (Table 5). 5. Meta-analysis of the maintenance therapy studies Although most randomised trials assessing the role of MT fail to prove gain in OS, some researchers advocated that new trials with larger sample size would be necessary to prove the surrogacy between PFS and OS. Based on this hypothesis, the optimal duration of CT in advanced NSCLC has been the subject of recent meta-analysis [79–81]. Lima and colleagues targeted seven randomised controlled

ECOG: Eastern Cooperative Oncology Group; SWOG: Southwest Oncology Group; CALGB: Cancer and Leukemia Group B; CT: chemotherapy; PFS: progression-free survival; OS: overall survival; BSC: best supportive care. a Patients are randomized at the start of treatment. b Patients are stratified based on eligibility for bevacizumab: patients ineligible will receive carboplatin + palitaxel with or without cetuximab.

PFS 244 Bevacizumab + cetuximab Sunitinib vs. placebo NCT00693992 (CALGB 30607)

Carboplatin + paclitaxel + bevacizumab × 6 cyclesb vs. Carboplatin + pemetrexed + bevacizumab + cetuximab × 6 cyclesb Platinum-based CT × 4 cycles NCT00946712 (SWOG 0819)a

362 Cisplatin + pemetrexed + bevacizumab × 4 cycles

NCT01107626 (ECOG 5508)

1282

NCT00961415 (AVAPERL1)

1546

All histologies

OS

PFS

Nonsquamous histology All histologies

OS Nonsquamous histology

Bevacizumab + pemetrexed Bevacizumab vs. Pemetrexed vs. Bevacizumab + pemetrexed Bevacizumab vs. Bevacizumab + pemetrexed Bevacizumab

900 Bevacizumab Bevacizumab NCT00762034 (Point Break)a

Nonsquamous histology

PFS without grade 4 toxicity OS Pemetrexed

Carboplatin + pemetrexed × 4 cycles vs. Carboplatin + pemetrexed + bevacizumab × 4 cycles Carboplatin + paclitaxel + bevacizumab × 4 cycles vs. Carboplatin + pemetrexed + bevacizumab × 4 cycles Carboplatin + paclitaxel + bevacizumab × 4 cycles NCT00948675a

Nonsquamous histology 360

Maintenance therapy Initial therapy Trial

Table 5 Ongoing phase III trial of maintenance therapy.

Sample size

Histology

Primary endpoint

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

353

trials (1559 patients) comparing different durations of firstline treatment of advanced NSCLC [79]. All trials tested platinum-based regimens [11–14,27,28] except one [82] and studies that include target drugs were excluded. The metaanalysis showed that treatment with more than 4 cycles was associated with an improved PFS (HR = 0.75; 95% CI: 0.65–0.85; p < 0.0001) in the absence of an OS benefit (HR = 0.97; 95% CI: 0.84–1.11; p = 0.65). Performing the analysis with only third-generation platinum-based treatments, the result was similar, with no differences between groups (HR = 1.08; 95% CI: 0.90–1.28; p = 0.41). Longer treatment was associated with more severe leukopenia but with no significant increase in nonhematological toxicities. The same conclusion was reached in a second metaanalysis performed by Soon et al. [80], in which thirteen trials including 3027 patients were analyzed [11–14,26,28,35,37,38,82–85]. This study included some of the most recently published trials not previously addressed in other reviews with the following designs: (1) a defined number of cycles vs. continuation of the same CT until disease progression [12,82]; (2) a fixed number of cycles vs. a larger defined number of cycles of identical CT [11,13,14,83,84]; (3) a defined number of cycles of initial CT vs. identical initial CT followed by additional cycles of an alternative CT [26,28,35,37,38,85]. The primary outcome was OS as reported in each study; secondary outcomes included PFS, health-related QoL and toxicity. The number of cycles of CT in the standard duration groups ranged from 2 to 8 while the number of additional cycles by which CT was to be extended ranged from 6 cycles to continuing until progression or prohibitive toxicity. Extending CT was associated with a clinically substantial and statistically significant 25% decrease in the hazard for progression as compared with the standard duration of CT (HR = 0.75; 95% CI: 0.69–0.81; p < 0.001), as well as in a clinical modest, but statistically significant reduction in the hazard for death (HR = 0.92; 95% CI: 0.85–0.99; p = 0.03). However, the survival benefit only became significant when the phase III of Ciuleanu et al. [38] evaluating the maintenance role of pemetrexed was added to the meta-analysis. This review did not include the SATURN [61] and ATLAS [62,63] studies and it would be safe to assume that its overall results would only be strengthened by the addition of these two large studies. A subgroup analysis revealed the effect on PFS was significantly greater for trials extending CT with a third generation agent than an older agent (HR = 0.70 vs. 0.90; p = 0.003), for trials extending CT with a non-platinum agent rather than a platinum (HR = 0.69 vs. 0.82; p = 0.03), and for trials of design 3 rather than design 2 (HR = 0.67 vs. 0.82; p = 0.01). There were no statistically significant differences in effects on OS between subgroups defined by trial design, or use of platinum agents or third generation CT agents. Longer duration of CT was also associated with more frequent AEs in all trials where it was reported. However, of the seven trials included in the metaanalysis that reported on QoL [11–14,28,37,82], five noted no major differences between treatment and control groups.

354

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

Authors conclude that extending CT beyond a standard number of cycles, particularly with a third-generation regimen, improved PFS substantially, but has only modest effects on OS in patients with advanced NSCLC and is associated with higher rates of AEs and not significant impact on QoL. A third meta-analysis of randomized studies that evaluated single agent MT has recently been reported [81]. Ten randomized trials with a total of 3451 patients have been included. The primary outcome of interest was OS and the secondary endpoint was PFS. The OS (HR = 0.87; 95% CI: 0.82–0.94; p = 0.0003) and PFS (HR = 0.84; 95% CI: 0.80–0.88; p < 0.0001) were superior with MT, and improvements in PFS and OS were observed with both cytotoxic agents (OS: HR = 0.88; 95% CI: 0.80–0.97; p = 0.018. PFS: HR = 0.87; 95% CI: 0.82–0.91; p < 0.0001) and EGFR-targeted agents (OS: HR = 0.86; 95% CI: 0.78–0.95; p = 0.006. PFS: HR = 0.76; 95% CI: 0.70–0.83; p < 0.0001). “Switch” MT was associated with significant OS (HR = 0.86; 95% CI: 0.80–0.93; p = 0.00046) and PFS (HR = 0.71; 95% CI: 0.66–0.77; p < 0.0001) However, “continuation” MT was not associated with a survival benefit (HR = 0.92; 95% CI: 0.77–1.08; p = 0.33) despite a modest improvement in PFS (HR = 0.92; 95% CI: 0.87–0.98; p = 0.007).

6. Unanswered questions about maintenance therapy Optimizing treatment of advanced NSCLC involves consideration of delivering the most effective therapies, in the right combination, at the right time, while minimizing toxic adverse effects and QoL deterioration. Approximately 40–60% of advanced NSCLC patients complete 4 cycles of platinum-based CT without progression or unacceptable toxicity. For these patients, continuation of CT is a feasible option. However, a number of important questions of how to deliver the additional therapy remains to be answered and deserves careful evaluation in the setting of appropriately designed prospective clinical trials. 6.1. What is the most appropriate study end point? Although most published trials showed a benefit of MT in terms of PFS, only the most recently reported prospective trials have shown a clear OS benefit today [38,61]. The benefits of using PFS as the basis for the adoption of a new therapy have the advantage of eliminating the confounding factor of poststudy treatments, offering a more rapid assessment of efficacy, and reducing the number of patients required for efficacy analysis compared with an endpoint of OS. Nevertheless, this endpoint is very dependent on the interobserver variability in the assessment of disease progression or the frequency of response evaluation across studies. To ensure accurate PFS measurement, a strict schedule of efficacy assessment and patient follow-up should be implemented and an independent review of radiographic images is frequently performed in large clinical trials. Furthermore,

small absolute improvements in PFS may not translate into clinical benefit or improvement in OS. OS is often considered a more definitive endpoint, but the availability of multiple agents and practice patterns that are used in the poststudy setting could potentially confound the effect of a maintenance regimen on it. It is not possible to control for poststudy therapy in the maintenance and non-maintenance study arms, and outcomes with subsequent treatments may be influenced by many clinical and biological factors. Because phase III trials have documented an improvement in OS with erlotinib and pemetrexed [38,61], this endpoint should be the preferred one for trials seeking to change the standard of care or become a standard maintenance agent. PFS may be the preferred endpoint when investigating a novel agent in the maintenance setting compared with placebo or an established agent. Perhaps, rather than discuss the utility of PFS as a useful efficacy endpoint, it may be more important to look into ways to standardize PFS assessment in maintenance phase III trials and to use available data to guide an agreement on what degree of PFS benefit is acceptable as a valid evidence of efficacy in the absence of OS improvement. 6.2. Maintenance treatment and quality of life In addition to PFS and OS benefit, important points to consider for MT are the symptomatic benefit, as well as the acute and cumulative toxicity of prolonged treatment and their impact on patients’ QoL. PFS without any symptomatic benefit should not be considered an appropriate endpoint for MT. So, supplementary information including symptom improvement may add value to the prolongation of PFS. In the maintenance setting the clinicians’ willingness to risk patient exposure to AEs may be less, particularly if toxicity may last for months or even years. Although the rate of grade ≥3 toxicities observed with MT has been low, a prolonged exposure to grades 1 and 2 toxicities may adversely impact patients’ QoL. Unfortunately, not all studies in the maintenance setting provided carefully collected prospective data about QoL. Earlier studies occasionally showed worsening in QoL with prolongation of CT, although such studies often included cisplatin regimens. More recent studies with newer generation drugs, especially targeted biological agents, have shown no detriment in QoL when comparing different durations of CT or when comparing delayed vs. immediate second-line therapy [13,32,37,38,61]. In the JMEN study [38], time to worsening of patient-reported symptoms was measured from the date of randomisation to the first date of worsening for each of six symptoms and three summary items of the Lung Cancer Symptom Scale (LCSS). Worsening was defined by a 15-mm increase from the baseline lung cancer symptom score on a 100-mm scale. Authors recorded significant delays in symptom worsening in favour of pemetrexed for pain (median 6.1 months [95% CI: 4.6–9.6] vs. 4.6 months [95% CI: 3.3-6.0]; HR = 0.76; 95% CI: 0.56-0.99; p = 0.041) and haemoptysis (HR = 0.58; 95% CI: 0.34–0.97; p = 0.038). None of the other time-to-worsening comparison differed significantly

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

between groups. In the PARAMOUNT study [32], authors used the EuroQoL dimensional scale (EQ-5D) to evaluate QoL. Questionnaires were administered at baseline, on day 1 of each cycle of induction or maintenance treatment and 30 days after completion of the study. No clinically relevant differences in QoL scales were observed between treatment arms. Fidias and colleagues observed a statistically significant improvement in PFS when docetaxel was administered immediately after front-line CT, without decreasing QoL [37]. Using the LCSS questionnaire, the same number of patients (15.6%) showed improved average symptom burden index (ASBI) in each docetaxel arm. The majority of patients in both docetaxel arms had stable ASBI; stable ASBI was slightly more prevalent in the immediate docetaxel arm (58.7%) compared with the delayed docetaxel arm (53.2%). Conversely, there were more patients in the delayed docetaxel arm (18.4%) with worsened ASBI compared with the immediate docetaxel arm (11%). However, overall ASBI results were not statistically different (p = 0.76) between the two docetaxel arms. In the SATURN trial [61], QoL was assessed using the Functional Assessment of Cancer Therapy-Lung (FACT-L) questionnaire at 6-weeks intervals until week 48, and every 12 weeks thereafter. There was not a statistically significant difference in QoL for patients receiving erlotinib compared with those receiving placebo (HR = 0.96; 95% CI: 0.79–1.16 for time to deterioration in QoL). A post hoc analysis showed that time to pain (HR = 0.61; 95% CI: 0.42–0.88; p = 0.008) and time to analgesic use (HR = 0.66; 95% CI: 0.46–0.94; p = 0.02) were both significantly improved with erlotinib vs. placebo, although time to cough (HR = 0.77; 95% CI: 0.49–1.21; p = 0.2546) and time to dyspnea (HR = 0.75; 95% CI: 0.48–1.17; p = 0.2054) were not significantly affected [66]. Finally, Soon and colleagues were unable to statistically pool results about health-related QoL in their meta-analysis because of differences among studies in the scales and methods used [80]. Three trials used the EORTC Core QoL questionnaire and Lung Cancer Specific Module; two trials used LCSS; one trial used the FACT-L; and one trial used an investigator-designed questionnaire. Only two of seven trials showed trends toward better health-related QoL associated with a standard duration of CT, rather than a extended CT. Five trials did not demonstrate major differences in global QoL between their two treatment groups despite differences in some subscales. To conclude, the issues of QoL and symptomatic benefit need to be addressed prospectively in future clinical trials, particularly for agents that have a narrow therapeutic index. 6.3. Who should receive maintenance treatment? In patients for whom a drug holiday may be appropriate, how long should the holiday be? How best should we monitor patients during a break from treatment? Experience from clinical practice suggests that only approximately 60–70% of patients with advanced NSCLC are able to receive second-line therapy. Interestingly, in the

355

trial by Fidias et al. [37], patients in the “delayed” docetaxel arm had the same OS as those receiving CT up front. This suggests that the trend toward improved outcome in the immediate arm was a result of more patients being able to receive an active drug. So, we can conclude that MT should improve survival, not because it does a better job of killing cancer cells when given earlier, but simply because this approach enables us to give patients access to more lines of effective treatment and treat a greater number of patients before other complications of the lung cancer render them unable to receive additional lines of therapy [86]. Moreover, the fact that only 51.5% of patients who were administered the placebo in the study of Ciuleanu et al. received any form of salvage CT probably also influenced the superior OS in the pemetrexed arm [38]. At present, there are no definitive markers of risk that allow us to identify which patients may safely receive a treatment holiday after successful first-line therapy, and which patients will progress quickly after discontinuation of treatment and never receive second-line therapy. Currently, data are only available for the docetaxel study [37], where characteristics such as PS, sex, or stage IIIB/IV were not found to predict for eventual second-line therapy in the control group. Furthermore, none of the studies has identified whether degree of response to first-line therapy predicts for the likelihood of receiving second-line treatment at the time of progression. A consensus should be reached as to what constitutes the surveillance regimen for those patients who would like a therapeutic break and to develop clinical and other biomarkers to predict those patients who fall into a high-risk category for early progression. These two strategies might go some way toward addressing these therapeutic dilemmas. The converse is also true, because it is difficult to defend exposing patients to treatment toxicity from MT if they have relatively indolent tumours. 6.4. How should we select the right patient population who would benefit from maintenance treatment? Closely related with the use of appropriate therapeutic agent for MT is the selection of the right patient population. In reality, when the survival benefit derived by unselected patients in maintenance CT and maintenance molecular therapy trials is examined critically, the benefits must be considered modest, at best, particularly when offset against the cost, toxicity, and inconvenience for patients. However, preplanned and unplanned subset analysis in these trials showed that select patients characterized by histologic or molecular profile derive a higher benefit from the MT in terms of significant reduction in the risk of disease progression and death. In the JMEN study [38], preplanned analysis for histology showed survival benefit only for nonsquamous tumours. Although the HR in the SATURN study also favoured patients with adenocarcinoma tumours [61], both in terms of PFS (HR = 0.60 vs. 0.76 for squamous cell cancers) and OS (HR = 0.77 vs. 0.86), the benefit for

356

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

maintenance erlotinib was seen irrespective of histology. Similarly, there was no treatment-by-histology interaction in the study by Fidias et al. [37] or the IFCT-GFPC-052 trial for either maintenance erlotinib or gemcitabine [30,31]. By design, all of the recently reported trials enrolled patients who derived some benefit from outline induction CT. Subgroups analysis in the larger phase III trial have assessed if a differential survival benefit with MT could be expected based on response to front-line CT; however, data are inconsistent. Results from the JMEN study showed benefit only for patients with SD at the time of random assignment. The HR for PFS was 0.53 for responders and 0.48 for SD patients, whereas the HR for OS was 0.90 and 0.68, respectively [38]. Similarly, the HR OS in the SATURN study favoured the SD patients (HR = 0.72 vs. 0.94 for responsive patients) [61]. Nevertheless, subgroup analysis of the docetaxel trial suggested that responsive patients might benefit more than those with SD. The HR for PFS was 0.47 and 0.81, whereas for OS the HR was 0.61 and 1.02 for response and SD patients, respectively [37]. Data from the continuation gemcitabine maintenance in the IFCT-GFPC-052 study were concordant with the docetaxel trial (PFS HR = 0.44 vs. 0.68 for responders and SD patients, respectively), whereas there was no difference in the switch-maintenance approach with erlotinib between these two groups (HR = 0.80 vs.0.85, respectively) [30,31]. Biomarker analysis has also suggested possible patient subgroups which may derive greater benefit with EGFRTKIs in the maintenance setting. It is clear that PFS, but not OS, with maintenance erlotinib is significantly better in patients with EGFR sensitizing mutations in exons 19 or 21 in both the SATURN and ATLAS trials [61,62,65,70]. Patients with K-ras mutation, however, seem to have less benefit. PFS was statistically significant only for patients with wild-type tumours in the SATURN trial (HR = 0.70; 95% CI: 0.57–0.87; p = 0.0009 vs. HR = 0.77; 95% CI: 0.50–1.19; p = 0.2246 for mutant tumours), although on further analysis there was no interaction with treatment effect for this variable (p = 0.95) [65]. Similarly, patients with K-ras wild-type status enjoyed the greatest PFS benefit from maintenance erlotinib and bevacizumab therapy in the ATLAS study (HR = 0.67; 95% CI: 0.49–0.91 for K-ras wild-type vs. HR = 0.93; 95% CI: 0.55–1.56 for K-ras mutant NSCLC) [70]. In addition to these phase III trial in the maintenance setting, several studies show questionable benefit with EGFR-TKIs in K-ras mutant NSCLC [87,88], and in one study, significantly worse survival was documented with erlotinib in such patients [89]. Although these data cannot be considered conclusive, a practice favouring treatment with EGFR TKIs for K-ras wild type tumours is currently the most advisable approach. In conclusion, consistent with the evolving practice in the first-line induction treatment setting, using such a well characterized tumour profile to select patients for MT may further improve the therapeutic index leading to a greater survival benefit. Moreover, the issue of selection of patients who benefit from MT is interrelated with its economic cost because a

better strategy of selecting patients may improve the benefit observed with the therapy and reduce the number of patients who receive an ineffective treatment. 6.5. Implications for drug development and clinical trials Several phase III studies are investigating the role of CT or targeted therapy in the maintenance setting. Ongoing or planned clinical trials are summarized in Table 5. It is to be noted that the study of Ciuleanu et al. [38] did not include pemetrexed induction regimens, mainly because it was started before the completion of the pivotal firstline study of pemetrexed–cisplatin vs. gemcitabine–cisplatin [19]. So, it is not possible to draw definitive conclusions from the study about the safety and efficacy of pemetrexed maintenance after initial treatment with pemetrexed-containing regimens. However, the PARAMOUNT phase III trial [32] has investigated this question, showing that pemetrexed continuation MT compared with placebo is an effective and well tolerated treatment in patients with advanced nonsquamous NSCLC with response or SD after four cycles of induction treatment with cisplatin–pemetrexed. Furthermore, translational research will address whether thymidylate synthase expression correlates with efficacy outcomes in patients treated with pemetrexed. A second phase III trial with similar design is comparing four cycles of carboplatin and pemetrexed followed by pemetrexed MT with four cycles of carboplatin, paclitaxel and bevacizumab followed by bevacizumab maintenance (NCT00948675). The primary endpoint is PFS without grade 4 toxicity. Apart from the ATLAS trial [62,63], results of recent maintenance studies do not take into account the use of bevacizumab. It is unclear whether patients who receive a platinum-based regimen with bevacizumab should continue bevacizumab, switch to another agent, or continue bevacizumab in combination with a second agent for MT. Several phase III trials are underway to examine this issue. In the Point-Break study, designed to detect an improvement in median OS of almost 3 months, patients are being randomly assigned to four cycles of the Patel regimen (carboplatin, pemetrexed and bevacizumab followed by maintenance pemetrexed and bevacizumab) or four cycles of the ECOG 4599 regimen (carboplatin, paclitaxel and bevacizumab with maintenance bevacizumab alone) (NCT00762034) [90]. The ECOG has proposed a three-arm phase III trial with similar design (ECOG 5508) in patients with nonsquamous NSCLC who are eligible for bevacizumab. Patients who do not experience disease progression after four cycles of carboplatin, paclitaxel and bevacizumab will be randomized to MT with pemetrexed alone, bevacizumab alone, or the combination of bevacizumab and pemetrexed. A complementary phase III study of Avastin (Bevacizumab) with or without pemetrexed as maintenance therapy after avastin in first line in patients with non-squamous non-small cell lung cancer (AVAPERL1) is investigating whether adding pemetrexed to bevacizumab

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

MT improves clinical outcomes over bevacizumab alone following 4 cycles of induction therapy with cisplatin, pemetrexed and bevacizumab (NCT00961415). Primary endpoint is PFS. Safety data have been presented at the 2011 ASCO Annual Meeting with 373 patients who received >1 dose of any study drug included in the analysis [91]. At a median follow-up time of 4.4 months, 43% of patients are on treatment, 34% in follow-up and 23% off-study. A total of 227 patients (61%) completed first-line therapy and initiated MT. First-line cisplatin–pemetrexed–bevacizumab was well tolerated with a safety profile comparable to that of regimens indicated for use with bevacizumab. Most AEs were grades 1–2. The most common grade >3 AEs were neutropenia (8%), embolism (3%), hypertension (2), anemia (2%), diarrhea (2%), nausea and vomiting (2%) and pneumonia (2%). The incidence of grade >3 AEs was similar in both maintenance arms (35% vs. 38% in the bevacizumab and bevacizumab–pemetrexed arms, respectively), but more serious AEs were reported in patients treated with pemetrexed–bevacizumab (20% vs. 31% in the bevacizumab and bevacizumab–pemetrexed arms, respectively). Finally, the Southwest Oncology Group has initiated a phase III trial, SWOG 0819, that will investigate the combination of carboplatin and paclitaxel (with bevacizumab in eligible patients) compared with the carboplatin, paclitaxel and cetuximab (with bevacizumab in eligible patients) (NCT00946712). Patients in both arms will receive six cycles of platinum-based CT and then receive maintenance CT with bevacizumab alone or in combination with cetuximab.

7. Conclusion Based on the currently available data, extended first-line CT with combination regimens for more than 4–6 cycles is not recommended because of cumulative toxicities and the lack of proven advantages in survival with the increased duration of therapy [2,3]. Nevertheless, the early use of an effective and well-tolerated non-crossresistant anticancer agent as MT appears beneficial in advanced disease. Phase III trials have demonstrated an improvement in PFS and OS with pemetrexed among patients with nonsquamous histology [38] and erlotinib MT after induction platinum-based CT [61]. The trials supporting sequential therapy all permitted second-line therapy, but this was not uniformly applied. Thus, the question of whether using an agent as first-line MT is better than using the same agent at progression remains unanswered. Patients with advanced NSCLC not-progressing after completion of their first-line CT regimen should be fully informed about both the potential risks and benefits of MT. They account for 50–60% of patients who initiated platinumbased CT. Phase III trials in the maintenance setting have shown that there was no difference in survival in patients actually receiving the prescribed second-line therapy [37,38]. Thus a treatment break after initial therapy followed by second-line treatment at the earliest sign of progression

357

appears to be an acceptable management option; however, patients must be closely monitored for both subjective and objective evidence of progression because the median PFS is approximately 2–3 months. Alternatively, sequential use of a non-cross-resistant therapy is appropriate for patients who place a higher value on delaying any disease progression or who are felt to be at high-risk of relapse. Our ability to select patients who are most likely to have a limited duration of response or SD after platinum-based first-line CT and/or benefit from MT is limited. Ultimately, physician and patients will have to decide on the timing of additional therapy based on prior treatment response and tolerance, the expected acute and cumulative toxicities of the proposed regimen, and potential benefits in terms of survival and symptom control. We believe that patients with high-volume or symptomatic disease, a preserved PS and no significant toxicity with previous treatments should be strongly considered for MT in routine practice. For patients who choose the maintenance approach, either a cytotoxic agent, such as pemetrexed in nonsquamous histology, docetaxel, or an EGFR TKI may be considered. What constitutes the optimal strategy is yet to be determined because there have been no comparative trials of maintenance chemotherapeutic agents against targeted agents or other strategies. Similarly, the duration of monoclonal antibody therapy has yet to be studied before maintenance of these agents has become standard of care after completion of first-line CT. With regard to future developments in the maintenance setting, special attention will be paid to the selection of patients based on their clinical and molecular profiles to maximize benefit and to reduce the risk of unnecessary toxicity and cost. Finally, patients should be encouraged to participate in clinical trials whenever possible in order to provide clarity about the optimal use of the maintenance approach in advanced NSCLC.

Conflict of interest The authors declare that they have no conflict of interest relating to the publication of this manuscript.

References [1] Jemal A, Siegel R, Xu J, et al. Global statistics, 2010. CA Cancer J Clin 2010;60:277–300. [2] Azzoli CG, Baker Jr S, Temin S, et al. American society of clinical oncology clinical practice guideline update on chemotherapy for stage IV non-small-cell lung cancer. J Clin Oncol 2009;27:6251–66. [3] D’Addario G, Felip E. Non-small-cell lung cancer: ESMO clinical recommendations for diagnosis, treatment and follow-up. Ann Oncol 2009;20(Suppl. 4):68–70. [4] Ngeow J, Leong SS, Gao F, et al. Impact of comorbidities on clinical outcomes in non-small cell lung cancer patients who are elderly and/or have poor performance status. Crit Rev Oncol Hematol 2010;76:53–60.

358

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

[5] Schiller JH, Harrington D, Belani CP, et al. Comparison of four chemotherapy regimens for advanced non-small cell lung cancer. N Engl J Med 2002;346:92–8. [6] Klastersky J, Awada A. Milestones in the use of chemotherapy for the management of non-small cell lung cancer (NSCLC). Crit Rev Oncol Hematol 2011, doi:10.1016/j.critrevonc.2011.02.002. [7] Mok TS, Wu LY, Thongprasert S, et al. Gefitinib or carboplatin–paclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009;361:947–57. [8] Zhou C, Wu Y, Chen G, et al. Efficacy results from the randomise phase III OPTIMAL (CTONG 0802) study comparing first-line erlotinib versus carboplatin plus gemcitabine, in Chinese advanced non-small cell lung cancer patients with EGFR activating mutations (abstract LBA13). Ann Oncol 2010;21(Suppl. 8):viii6. [9] Rosell R, Moran T, Queralt C, et al. Screening for epidermal growth factor receptor mutations in lung cancer. N Engl J Med 2009;361: 958–67. [10] Pirker R, Filipits M. Monoclonal antibodies against EGFR in non-small cell lung cancer. Crit Rev Oncol Hematol 2010, doi:10.1016/j.critrevonco.2010.10.008. [11] Smith IE, O’Brien ME, Talbot DC, et al. Duration of chemotherapy in advanced non-small-cell lung cancer: a randomized trial of three versus six courses of mitomycin, vinblastin, and cisplatin. J Clin Oncol 2001;19:1336–43. [12] Socinski MA, Schell MJ, Peterman A, et al. Phase III trial comparing a defined duration of therapy versus continuous therapy followed by second-line therapy in advanced-stage IIIB/IV non-small-cell lung cancer. J Clin Oncol 2002;20:1335–43. [13] Von Plessen C, Bergman B, Andresen O, et al. Palliative chemotherapy beyond three courses conveys no survival or consistent quality-oflife benefits in advanced non-small-cell lung cancer. Br J Cancer 2006;95:966–73. [14] Park JO, Kim S-W, Ahn JS, et al. Phase III trial of two versus four additional cycles in patients who are nonprogressive after two cycles of platinum-based chemotherapy in non-small-cell lung cancer. J Clin Oncol 2007;27:5233–9. [15] Socinski MA, Stinchcombe TE. Duration of first-line chemotherapy in advanced non-small-cell lung cancer: less is more in the era of subsequent therapies. J Clin Oncol 2007;25:5155–7. [16] National comprehensive cancer network: NCCN clinical practice guidelines in oncology: non small cell lung cancer; 2011. [17] Sandler A, Gray R, Perry MC, et al. Paclitaxel–carboplatin alone or with bevacizumab in non-small-cell lung cancer. N Engl J Med 2006;355:2542–50. [18] Pirker R, Pereira JR, Szczesna A, et al. Cetuximab plus chemotherapy in patients with advanced non-small-cell lung cancer (FLEX): an openlabel randomised phase III trial. Lancet 2009;373:1525–31. [19] Scagliotti GV, Parikh P, von Pawel J, et al. Phase III study comparing cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapy-naive patients with advanced-stage non-small-cell lung cancer. J Clin Oncol 2008;26:3543–51. [20] Reck M, von Pawel, Zatloukal P, et al. Overall survival with cisplatin–gemcitabine and bevacizumab or placebo as first-line therapy for nonsquamous non-small-cell lung cancer: results from a randomised phase III trial (AVAIL). Ann Oncol 2010;21:1804–9. [21] Mountzios G, Dimopoulos MA, Soria JC, Sanoudou D, Papadimitriou CA. Histophatologic and genetic alterations as predictors of response to treatment and survival in lung cancer. Crit Rev Oncol Hematol 2010;75:94–109. [22] Stewart DJ. Tumor and host factors that may limit efficacy of chemotherapy in non-small cell and small cell lung cancer. Crit Rev Oncol Hematol 2010;75:173–234. [23] Wotjalla A, Arcaro A. Targeting phosphoinositide 3-kinase signaling in lung cancer. Crit Rev Oncol Hematol 2011, doi:10.1016/j.critrevonc.2011.01.007. [24] Shepherd FA, Dancey J, Ramlau R, et al. Prospective randomized trial of docetaxel versus best supportive care in patients with non-small-cell

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

lung cancer previously treated with platinum-based chemotherapy. J Clin Oncol 2000;18:2095–103. Hanna S, Shepherd FA, Fosella FV, et al. Randomized phase III trial of pemetrexed versus docetaxel in patients with non-small-cell lung cancer previously treated with chemotherapy. J Clin Oncol 2004;22: 1589–97. Belani CP, Barstis J, Perry MC, et al. Multicenter, randomized trial for stage IIIB or IV non-small-cell lung cancer using weekly paclitaxel and carboplatin followed by maintenance weekly paclitaxel or observation. J Clin Oncol 2003;21:2933–9. Belani CP, Ramalingam S, Perry MC, et al. Randomized, phase III study of weekly paclitaxel in combination with carboplatin standard every-3-weeks administration of carboplatin and paclitaxel for patients with previously untreated advanced non-small-cell lung cancer. J Clin Oncol 2008;26:468–73. Brodowicz T, Krakowski M, Zwitter M, et al. Cisplatin and gemcitabine first-line chemotherapy followed by maintenance gemcitabine or best supportive care in advanced non-small cell lung cancer: a phase II trial. Lung Cancer 2006;52:155–63. Belani CP, Waterhouse DM, Ghazal H, et al. Phase III study of maintenance gemcitabine (G) and best supportive care (BSC) versus BSC, following standard combination therapy with gemcitabine-carboplatin (G-Cb) for patients with advanced non-small cell lung cancer (NSCLC). J Clin Oncol 2010;28(Suppl.):540s (abstr 7506). Perol M, Chouaid C, Milleron MJ, et al. Maintenance with gemcitabine or erlotinib versus observation with predefined second-line treatment after cisplatin–gemcitabine induction chemotherapy in advanced NSCLC: IFCT-GFPC 0502 phase III study. J Clin Oncol 2010;28(Suppl.):540s (abstr 7507). Perol M, Zalcman G, Monnet I, et al. Final results from the IFCTGFCT 0502 phase III study: maintenance therapy in advanced NSCLC with either gemcitabine (G) or erlotinib (E) versus observation (O) after cisplatin–gemcitabine induction chemotherapy (CT), with a predefined second-line treatment. Ann Oncol 2010;21(Suppl.):8s (abstr 370PD). Paz-Ares LG, de Marinis F, Dediu M, et al. PARAMOUNT: phase III study of maintenance pemetrexed (pem) plus best supportive care (BSC) versus placebo plus BSC immediately following induction treatment with pemetrexed plus cisplatin for advanced nonsquamous non-small cell lung cancer (NSCLC). J Clin Oncol 2011;29(Suppl.) (abstr CRA 7510). Goldie D, Coldman A. A mathematic model for relating the drug sensitivity of tumors to their spontaneous mutation rate. Cancer Treat Rep 1979;63:1727–33. Day RS. Treatment sequencing, asymmetry, and uncertainty: protocol strategies for combination chemotherapy. Cancer Res 1986;46:3876–85. Westeel V, Quoix E, Moro-Sibilot D, et al. Randomized study of maintenance vinorelbine in responders with advanced non-small-cell lung cancer. J Natl Cancer Inst 2005;97:499–506. Sculier JP, Lafitte JJ, Lecomte J, et al. A phase III randomised trial comparing sequential chemotherapy using cisplatin-based regimen and paclitaxel to cisplatin-based chemotherapy alone in advanced nonsmall-cell lung cancer. Ann Oncol 2007;18:1037–42. Fidias PM, Dakhil SR, Lyss AP, et al. Phase III study of immediate compared with delayed docetaxel after front-line therapy with gemcitabine plus carboplatin in advanced non-small-cell lung cancer. J Clin Oncol 2009;27:591–8. Ciuleanu T, Brodowicz T, Zielinski C, et al. Maintenance pemetrexed plus best supportive care versus placebo plus best supportive care for non-small-cell lung cancer: a randomised, double-blind, phase 3 study. Lancet 2009;374:1432–40. Gandara DR, Vokes E, Green M, et al. Activity of docetaxel in platinumtreated non-small-cell lung cancer: results of a phase II multicenter trial. J Clin Oncol 2000;18:131–5. De Marinis F, Grossi F. Clinical evidence for second-and third-line treatment options in advanced non-small cell lung cancer. Oncologist 2008;13:14–20.

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360 [41] Favaretto AG, Pasello G, Magro C. Second and third line treatment in advanced non-small cell lung cancer. Discov Med 2009;8:204–9. [42] Scagliotti G, Hanna N, Fosella F, et al. The differential efficacy of pemetrexed according to NSCLC histology: a review of two phase III studies. Oncologist 2009;14:253–63. [43] Scagliotti G, Monica V, Ceppi P, et al. Baseline thymidilate synthase expression according to histological subtypes of non-small cell lung cancer. J Clin Oncol 2009;27(Suppl.):15s (abstr 7521). [44] Shepherd FA, Rodrigues Pereira J, Ciuleanu T, et al. Erlotinib in previously treated non-small-cell lung cancer. N Engl J Med 2005;353:123–32. [45] Maemodo M, Nou A, Kobayashi K, et al. Gefitinib or chemotherapy for non-small-cell lung cancer. N Engl J Med 2010;362:2380–8. [46] Lynch TJ, Patel T, Dreisbach L, et al. Cetuximab and first-line taxane/carboplatin chemotherapy in advanced non-small-cell lung cancer: results of the randomized multicenter phase III trial BMS099. J Clin Oncol 2010;28:911–7. [47] Wozniak A. Challenges in the current antiangiogenic treatment paradigm for patients with non-small cell lung cancer. Crit Rev Oncol Hematol 2011, doi:10.1016/j.critrevonc.2011.05.003. [48] Patel JD, Hensing TA, Rademarker A, et al. Phase II study of pemetrexed and carboplatin plus bevacizumab with maintenance pemetrexed and bevacizumab as first line therapy for nonsquamous non-small-cell lung cancer. J Clin Oncol 2009;27:3284–9. [49] Liu L, Cao Y, Tan A, Liao C, et Gao F. Cetuximab-based therapy versus non-cetuximab therapy for advanced cancer: a meta-analysis of 17 randomized controlled trials. Cancer Chemother Pharmacol 2010;65:849–61. [50] Scagliotti G, Novello S, von Pawel J, et al. Phase III study of carboplatin and paclitaxel alone or with sorafenib in advanced non-small cell lung cancer. J Clin Oncol 2010;28:1835–42. [51] Goss GD, Arnold A, Shepherd FA, et al. Randomized, double blind trial of carboplatin and paclitaxel with either daily oral cediranib or placebo in advanced non-small-cell lung cancer: NCIC clinical trials group BR24 Study. J Clin Oncol 2009;28:49–55. [52] Schiller JH, Lee JW, Hanna NH, et al. A randomized discontinuation phase II study of sorafenib vs placebo in patients with non-small cell lung cancer (NSCLC) who failed at least two prior chemotherapy regimens: E2501. J Thorac Oncol 2009:C1.6. [53] Giaccone G, Herbst RS, Manegold C, et al. Gefitinib in combination with gemcitabine and cisplatin in advanced non-small-cell lung cancer: a phase III trial-INTACT 1. J Clin Oncol 2004;27:777–84. [54] Herbst RS, Giaccone G, Schiller JH, et al. Gefitinib in combination with paclitaxel and carboplatin in advanced non-small-cell lung cancer: a phase III trial-INTACT 2. J Clin Oncol 2004;22:785–94. [55] Gatzemeier U, Pluzanska A, Szczesna A, et al. Phase III study of erlotinib in combination with cisplatin and gemcitabine in advanced non-small-cell lung cancer: the Tarceva lung cancer investigation trial. J Clin Oncol 2007;25:1545–52. [56] Herbst RS, Prager D, Hermann R, et al. TRIBUTE: a phase III trial of erlotinib hydrochloride (OSI-774) combined with carboplatin and paclitaxel chemotherapy in advanced non-small-cell lung cancer. J Clin Oncol 2005;23:5892–9. [57] Gandara DR, Gumerlock PH. Epidermal growth factor receptor tyrosine kinase inhibitors plus chemotherapy: case closed or is the jury still out? J Clin Oncol 2005;23:5856–8. [58] Xu JM, Azzariti A, Colucci G, Paradiso A. The effect of gefitinib (Iressa, ZD1839) in combination with oxaliplatin is scheduledependent in colon cancer cell lines. Cancer Chemother Pharmacol 2003;52:442–8. [59] Davies AM, Ho C, Lara Jr PN, Mack P, Gumerlock PH, Gandara DR. Pharmacodynamic separation of epidermal growth factor receptor tyrosine kinase inhibitors and chemotherapy in non-small-cell lung cancer. Clin Lung Cancer 2006;7:385–8. [60] Mok T, Wu YL, Yu CJ, et al. Randomized, placebo-controlled, phase II study of sequential erlotinib and chemotherapy as first-line treatment for advanced non-small-cell lung cancer. J Clin Oncol 2009;27:5080–7.

359

[61] Cappuzo F, Ciuleanu T, Stelmakh L, et al. Erlotinib as maintenance treatment in advanced non-small-cell lung cancer: a multicentre, randomised, placebo-controlled phase 3 study. Lancet Oncol 2010;11:521–9. [62] Miller VA, O’Connor P, Soh C, et al. A randomized, double-blind, placebo-controlled, phase IIIb trial (ATLAS) comparing bevacizumab (B) therapy with or without erlotinib (E) after completion of chemotherapy with B for first-line treatment of locally advanced, recurrent, or metastatic non-small cell lung cancer. J Clin Oncol 2009;27(Suppl.):18s (abstr LBA8002). [63] Kabbinavar FF, Miller VA, Johnson BE, et al. Overall survival (OS) in ATLAS, a phase IIIB trial comparing bevacizumab (B) therapy with or without erlotinib (E) after completion of chemotherapy (chemo) with B for first-line treatment of locally advanced, recurrent, or metastatic non-small cell lung cancer (NSCLC). J Clin Oncol 2010;28(Suppl.):15s (abstr 7526). [64] Cappuzzo F, Ciuleanu T, Park K, Wu Y, Giaccone G. Erlotinib maintenance therapy in patients with stable disease after first-line platinum-doublet chemotherapy for advanced NSCLC. Ann Oncol 2010;21(Suppl.):8s (abstr 369PD). [65] Brugger W, Triller N, Blasinska-Morawiec S, et al. Biomarker analyses from the phase III placebo-controlled SATURN study of maintenance erlotinib following first-line chemotherapy for advanced NSCLC. J Clin Oncol 2009;27(Suppl.):15s (abstr 8020). [66] Juhász E, Kim JH, Stelmakh L, et al. Maintenance therapy of non-small cell lung cancer with erlotinib preserves quality of life. Ann Oncol 2010;21(Suppl.):8s (abstr 439P). [67] Herbst RS, Johnson DH, Mininberg E, et al. Phase I/II trial evaluating the anti-vascular endothelial growth factor monoclonal antibody bevacizumab in combination with the HER-1/epidermal growth factor receptor tyrosine kinase inhibitor erlotinib for patients with recurrent non-small-cell lung cancer. J Clin Oncol 2005;23:2544–55. [68] Herbst RS, O’Neill VJ, Fehrenbacher L, et al. Phase II study of efficacy and safety of bevacizumab in combination with chemotherapy or erlotinib compared with chemotherapy alone for treatment of recurrent or refractory non small-cell lung cancer. J Clin Oncol 2007;25: 4743–50. [69] Herbst RS, Ansari R, Bustin F, et al. Efficacy of bevacizumab plus erlotinib vs erlotinib alone in advanced non-small-cell lung cancer after failure of standard first-line CT (BeTa): a double-blind, placebocontrolled, phase 3 trial. Lancet 2011;377:1845–54. [70] Johnson B, Miller V, Amler L, et al. Biomarker evaluation in the randomized, double-blind, placebo-controlled, phase IIIb ATLAS trial, comparing bevacizumab (B) therapy with or without erlotinib (E), after completion of chemotherapy with bevacizumab for the treatment of locally-advanced, recurrent, or metastatic non-small cell lung cancer (NSCLC). Eur J Cancer 2009;7:5 (abstr 8LBA). [71] Takeda K, Hida T, Sato T, et al. Randomized phase III trial of platinumdoublet chemotherapy followed by gefitinib compared with continued platinum-doublet chemotherapy in Japanese patients with advanced non-small-cell lung cancer: results of a West Japan Thoracic Oncology Group Trial (WJTOG0203). J Clin Oncol 2010;28:753–60. [72] Zhang L, Shenglin M, Song X, et al. Efficacy, tolerability and biomarker analysis from a phase III, randomized, placebo-controlled, parallel group study of gefitinib as maintenance therapy in patients with locally advanced or metastatic non-small cell lung cancer (NSCLC; INFORM; C-TONG 0804). J Clin Oncol 2011;29(Suppl.) (abstr LBA 7511). [73] Gaafar RM, Surmont V, Scagliotti G, et al. A double-blind, randomized, placebo-controlled phase III intergroup study of gefitinib (G) in patients (pts) with advanced NSCLC, non-progressing after first-line platinum-based chemotherapy (EORTC 08021-ILCP 01/03). J Clin Oncol 2010;28(Suppl.):15s (abstr 7518). [74] Surmont VF, Gaafar RM, Scagliotti GV, et al. A double-blind, randomized, placebo-controlled phase III study of gefitinib (G) versus placebo (P) in patients (PTS) with advanced NSCLC, non-progressing after first-line platinum-based chemotherapy (EORTC 08021-ILCP). Ann Oncol 2010;21(Suppl.):8s (abstr 368PD).

360

A. Custodio, J. de Castro / Critical Reviews in Oncology/Hematology 82 (2012) 338–360

[75] Herbst R, Sun Y, Eberhardt WE, et al. Vandetanib plus docetaxel versus docetaxel as second line treatment for patients with advanced nonsmall-cell lung cancer (ZODIAC): a double-blind, randomised, phase 3 trial. Lancet Oncol 2010;11:619–26. [76] De Boer R, Arrieta O, Yang CH, et al. Vandetanib plus pemetrexed versus pemetrexed for the second-line treatment of advanced nonsmall-cell lung cancer: a randomized, double-blind phase III trial. J Clin Oncol 2011;29:1067–74. [77] Natale RB, Thongprasert S, Greco FA, et al. Phase III trial of vandetanib compared with erlotinib in patients with previously treated advanced non-small-cell lung cancer. J Clin Oncol 2011;29:1059–66. [78] Ahn M, Park K, Ahn JS, et al. Randomized, double-blind, placebocontrolled phase II study of vandetanib maintenance for advanced or metastatic non-small-cell lung cancer (NSCLC) following 1st-line platinum-doublet chemotherapy. Ann Oncol 2010;21(Suppl.):8s (abstr 418P). [79] Lima J, Santos L, Sasse E, Sasse A. Optimal duration of first-line chemotherapy for advanced non-small cell lung cancer: a systematic review with meta-analysis. Eur J cancer 2009;45:601–7. [80] Soon YY, Stockler MR, Askie LM, Boyer MJ. Duration of chemotherapy for advanced non-small-cell lung cancer: a systematic review and meta-analysis of randomized trials. J Clin Oncol 2009;27:3277–83. [81] Behera M, Owonikoko TK, Chen Z, et al. Single-agent maintenance therapy for advanced-stage non-small cell lung cancer: a meta-analysis. J Clin Oncol 2011;29(Suppl.) (abstr 7553). [82] Buccheri GF, Ferrigno D, Curcio A, Vola F, Rosso A. Continuation of chemotherapy versus supportive care alone in patients with inoperable non-small cell lung cancer and stable disease after two or three cycles of MACC. Results of a randomized prospective study. Cancer 1989;63:428–32. [83] Tourani JM, Penaud D, Caubarrere I, et al. Cisplatin and vindesine for disseminated non small cell lung cancer: results of a prospective trial with 81 patients. Bull Cancer 1990;77:1107–13. [84] Barata FJ, Parente B, Texeira E, et al. Optimal duration of chemotherapy in non-small-cell lung cancer: multicenter, randomized, prospective clinical trial comparing 4 vs 6 cycles of carboplatin and gemcitabine. J Thorac Oncol 2007;2(Suppl.):S666 (abstr P2-235). [85] Zarogoulidis K, Ziogas E, Dermitzakis G, et al. The survival benefit of maintenance therapy of NSCLC patients responding to initial treatment: a randomized trial. Eur J Cancer 1995;31(Suppl.):S231–2 (abstr 1112). [86] Schiller JH. Duration of chemotherapy for metastatic non-small-cell lung cancer: more may be better after all. J Clin Oncol 2009;27:3265–7. [87] Jackman DM, Yeap BY, Lindeman NI, et al. Phase II clinical trial of chemotherapy-naïve patients > or = 70 years of age treated with erlotinib for advanced non-small-cell lung cancer. J Clin Oncol 2007;25:760–6. [88] Zhu CQ, da Cunha Santos G, Ding K, et al. Role of KRAS and EGFR as biomarkers of response to erlotinib in National Cancer Institute of Canada Clinical Trials Group study BR.21. J Clin Oncol 2007;25:4268–75. [89] Eberhard DA, Johnson BE, Amler LC, et al. Mutations in the epidermal growth factor receptor and in KRAS are predictive and prognostic

indicators in patients with non-small-cell lung cancer treated with chemotherapy alone and in combination with erlotinib. J Clin Oncol 2005;23:5900–9. [90] Patel JD, Bonomi P, Socinski MA, et al. Treatment rationale and study design for the PointBreak study: a randomized, openlabel phase III study of pemetrexed/carboplatin/bevacizumab followed by maintenance pemetrexed/bevacizumab versus paclitaxel/arboplatin/bevacizumab followed by maintenance bevacizumab in patients with stage IIIB or IV nonsquamous non-small-cell lung cancer. Clin Lung Cancer 2009;10: 252–6. [91] Barlesi F, Scherpereel A, Gervais R, et al. AVAPERL1 (MO22089): maintenance (mtc) bevacizumab (bev) with or without pemetrexed (pem) in patients (pts) with advanced nonsquamous non-small cell lung cancer (NSCLC) treated with first-line (1L) bevacizumab-cisplatin (cis)-pem: interim safety data. J Clin Oncol 2011:29 (Suppl.; abstr 7562).

Biographies Ana Custodio M.D. is currently a staff physician of the Medical Oncology Department at La Paz University Hospital in Madrid, Spain. Dr. Custodió interests include research in lung cancer, gastrointestinal cancer and neuroendocrine tumours, focusing on molecular biology and translational oncology. She is author and co-author of several papers and book chapters in these fields. Dr. Custodio is a member of several scientific societies including the Spanish Society of Medical Oncology (SEOM), the Spanish Lung Cancer Group (GECP) and the Spanish Group of Neuroendocrine Tumours (GETNE). Javier de Castro M.D., Ph.D. is currently Associate Professor of Clinical Oncology and Palliative Care at La Paz University Hospital and at the School of Medicine, in the Autonoma University in Madrid, Spain. Dr. de Castró main research interests include basic an clinical research in lung and gastrointestinal cancer, focusing on molecular-targeted therapies and molecular biology. He is co-author of more than 100 papers, books, and book chapters in this field. Dr. de Castro is also a member of several scientific societies including the American Society of Clinical Oncology (ASCO), the European Society of Medical Oncology (ESMO), the Spanish Society of Medical Oncology (SEOM) and the Spanish Society for the Study and Investigation in Cancer (ASEICA).