Maintenance therapy is a new treatment strategy that aims to sustain a reduced tumor size and relieve tumor-related symptoms, in contrast to conventional chemotherapy that aims to maximize tumor cell death.

Based on research data, the National Comprehensive Cancer Network (NCCN) guidelines recommend maintenance therapy in non-small-cell lung cancer (NSCLC) after first-line therapy, and many clinical studies of multiple regimens and modalities are currently underway.

Maintenance therapy can be classified into two types: switch maintenance therapy and continuous maintenance therapy. Switch maintenance therapy involves switching to a different non-cross-resistant regimen in an alternating or sequential fashion if the response is complete remission (CR), partial remission (PR), or stable disease (SD) after four to six cycles of first-line platinum-based combination chemotherapy. Continuous maintenance therapy is to continue one or all component of the current regimen after four to six cycles of chemotherapy if there is a SD or better response. Both maintenance therapies can be continued until disease progression1.

Much clinical and biological evidence regarding maintenance therapy exists2. First, about 20-80% of NSCLC patients cannot receive second-line chemotherapy for multiple reasons, including poor compliance. Maintenance therapy can suppress disease progression and provide the opportunity to receive additional treatment. Second, according to the Goldie-Coldman theory3, resistant and slowly growing cancer cells remain after first-line chemotherapy which has killed the sensitive and rapidly proliferating cells. Use of different non-cross-resistant chemotherapy regimens is effective in eradicating the remaining resistant cancer cells. Third, continuing low-dose cytotoxic chemotherapies can damage endothelial cells, ultimately decreasing angiogenesis in the tumor microenvironment. Fourth, maintenance chemotherapy can activate the immune system against tumor cell. Despite number of studies showing an effectiveness of maintenance therapy, many concerns remain regarding the financial burden and uncertainty of improving quality of life (QOL). Additionally, absence of the parameters used to predict the effectiveness of maintenance therapy and debates over whether progression-free survival is the optimal primary endpoint in clinical trials to elucidate benefit of maintenance therapy should be considered in practical application of it.

NCCN version 2.2013 recommends continuous maintenance therapy using bevacizumab (category 1), cetuximab (category 1), pemetrexed (category 1), bevacizumab plus pemetrexed, and gemcitabine and switch maintenance therapy using pemetrexed, erlotinib, and docetaxel (category 2B). It also recommends that close surveillance of patients without therapy is a reasonable alternative to maintenance.

More specifically, in anaplastic lymphoma kinase-negative non-squamous NSCLC patients, continuation of bevacizumab (category 1), cetuximab (category 1), pemetrexed (category 1), bevacizumab plus pemetrexed, or gemcitabine and switch therapy to pemetrexed or erlotinib improve progression-free survival and overall survival (OS). In patients with squamous cell carcinoma, continuation of cetuximab (category 1) or gemcitabine and switch to erlotinib or doxetaxel (category 2B) are recommended.

Continuation of first-line platinum doublet chemotherapy can be understood in the same context with clinical studies evaluating the optimal chemotherapy duration. More than four to six cycles of platinum-based chemotherapy showed no benefit in increasing the overall response rate (ORR) or survival rate; rather, this chemotherapy regimen augments toxicity. Socinski et al.4 compared the OS and QOL in two groups of stage IIIb/IV NSCLC patients: arm A (four cycles of carboplatin at an area under the curve of 6 and paclitaxel 200 mg/m² every 21 days) or arm B (continuous treatment with carboplatin/paclitaxel until progression). That study showed no overall benefit in OS, ORR, or QOL to continuing treatment with carboplatin/paclitaxel beyond four cycles in advanced NSCLC. Additionally, other studies comparing six cycles and three to four cycles of chemotherapy also demonstrated that progression-free survival (PFS), OS, and the proportion of patients undergoing second-line chemotherapy were not significantly different5,6.

Paclitaxel, gemcitabine, pemetrexed, vinorelbine, and doxetaxel can be combined with platinum as first-line chemotherapy for NSCLC patients. Among them, paclitaxel, gemcitabine, and pemetrexed were investigated as maintenance chemotherapy. Belani et al.7 showed that weekly paclitaxel and carboplatin followed by weekly maintenance paclitaxel (70 mg/m², 3 of 4 weeks) was beneficial compared with observation alone after first-line chemotherapy. In this study, the median PFS and OS were 38 and 76 weeks in maintenance therapy group, respectively, and 29 and 60 weeks for the control group, respectively. However, this study did not analyzed statistical significance according to maintenance therapy; therefore, a need for further validation of the efficacy of weekly paclitaxel maintenance therapy exists.

The Central European Cooperative Oncology Group (CECOG) compared gemcitabine maintenance therapy and best supportive care (BSC) after combination chemotherapy using cisplatin (80 mg/m²) and gemcitabine (1,250 mg/m², days 1 and 8)8. Patients showing CR/PR or SD after four cycles of first-line chemotherapy were randomized and maintenance arm received gemcitabine (1,250 mg/m², days 1 and 8) every 3 weeks until tumor progression. The differences in OS between the two arms were not statistically significant. With statistical significance, the PFS was 6.6 months in the maintenance group compared with 3.6 months in the BSC group. However, subgroup analysis showed that good-performance patients (Karnofsky scale>80) had a shorter OS in BSC arm than in maintenance arm (8.3 months vs. 22.9 months, respectively; hazard ratio (HR), 2.1; n=107). This means that good performance status (PS) can be valuable predictor in maintenance therapy.

Belani et al.9 compared gemcitabine maintenance therapy and BSC after carboplatin (area under the curve [AUC], 5) and gemcitabine (1,000 mg/m², days 1 and 8) combination chemotherapy. No significant difference was found in the overall OS and PFS. And greater than grade 3-4 toxicities were noted in patients treated with gemcitabine maintenance therapy.

Intergroupe Francophone de Cancerologie Thoracique-Groupe Francais de Pneumo-Cancerologie (IFCT-GFPC 0502) is a well-designed controlled study to demonstrate the effects of gemcitabine maintenance therapy. In that study, good-performance patients (European Cooperative Oncology Group [ECOG] grade 1-2) with CR/PR or SD after four cycles of first-line chemotherapy using gemcitabine (1,250 mg/m², days 1 and 8) plus cisplatin (80 mg/m², day 1) were randomized to gemcitabine or erlotinib maintenance therapy or BSC until tumor progression. All patients were treated with pemetrexed at progression. PFS was improved in the gemcitabine maintenance therapy group compared with BSC (3.8 months vs. 1.9 months; HR, 0.55; p<0.001); however, no OS benefit for maintenance therapy was observed. The study did not compare the benefits of gemcitabine and erlotinib maintenance therapy.

Continuation maintenance with pemetrexed was validated by the PARAMOUNT study10. Patients who did not progress after completion of four cycles of cisplatin plus pemetrexed, had an ECOG PS of 0 or 1, and achieved an induction response (CR, PR, or SD) were randomly assigned to receive maintenance therapy with either pemetrexed (500 mg/m² every 21 days) plus BSC or placebo plus BSC until disease progression. Only non-squamous cancer patients with good PS (ECOG 0 or 1) were enrolled in that study. Among the 359 patients randomized to continuation maintenance with pemetrexed, a significant reduction was noted in the risk of disease progression over the placebo group (HR, 0.62; 95% confidence interval [CI], 0.49-0.79; p<0.0001). The median PFS, measured from randomization, was 4.1 months (95% CI, 3.2-4.6) for pemetrexed and 2.8 months (2.6-3.1) for the placebo. A significant benefit in OS was observed in the pemetrexed maintenance therapy group compared with the placebo group (16.9 months vs. 14.0 months, respectively).

Vinorelbine, paclitaxel, docetaxel, and pemetrexed has been studied as maintenance therapy in NSCLC.

Westeel et al.11 conducted a randomized trial to compare maintenance vinorelbine therapy with observation. Patients with stage IIIB NSCLC having two cycles of monthly mitomycin-ifosfamide-cisplatin (MIC) followed by radiotherapy and those with "wet" stage IIIB (pleural or pericardial involvement), stage IIIB with supraclavicular node involvement, or stage IV (i.e., metastatic) NSCLC received 4 cycles of monthly MIC were enrolled11. Patients who responded to MIC treatment were randomly assigned to receive intravenous vinorelbine at a dose of 25 mg/m² for 6 months or no further treatment. No difference in PFS was found between these arms (log-rank p=0.32).

Switch maintenance therapy using paclitaxel also did not show a benefit regarding OS and PFS12. In that study, patients were treated with three cycles of GIP (gemcitabine [days 1 and 8]+ifosfamide [day 1]+cisplatin [day 1]) chemotherapy. Patients with CR/PR or SD were randomly assigned to receive GIP maintenance or paclitaxel (225 mg/m² per 3 weeks) switch maintenance therapy every 3 weeks. However, no significant difference in OS was observed. Regarding toxicity, the incidence of thrombocytopenia was increased in the GIP group, and that of peripheral neuropathy was increased in the paclitaxel-treated group, but the difference was not statistically significant.

Fidias et al.13 conducted a phase III study of immediate compared with delayed docetaxel after front-line therapy with gemcitabine plus carboplatin in advanced NSCLC. The enrolled chemotherapy-naïve patients had either stage IIIB NSCLC with pleural effusion or stage IV NSCLC. Gemcitabine (1,000 mg/m²) was administered on days 1 and 8 with carboplatin (AUC, 5) on day 1. After 4 cycles of gemcitabine plus carboplatin per 3 weeks, patients who did not show progression were randomly assigned either to an immediate docetaxel group (docetaxel 75 mg/m²) on day 1 every 21 days (with a maximum of six cycles) or to a delayed docetaxel group using docetaxel after observation and progression of tumor. The median PFS for immediate docetaxel (5.7 months) was significantly greater (p=0.0001) than for delayed docetaxel (2.7 months). The median OS for immediate docetaxel (12.3 months) was greater than for delayed docetaxel (9.7 months), but the difference was not statistically significant (p=0.0853). The QOL results were not significantly different (p=0.76) between two docetaxel groups. Thus, patients seem to have benefited from docetaxel therapy, and the patients in the immediate docetaxel arm trended toward improved OS because a greater number of patients could receive treatment. Treatment was discontinued in 37.2% of patients in the delayed docetaxel group, and the major reason was disease progression. However, in the immediate docetaxel group, only 5% of patients interrupted the treatment, suggesting that NSCLC patients may be healthier and prone to receive additional therapy if maintenance therapy is offered immediately after front-line chemotherapy.

Ciuleanu et al.14 conducted a randomized, double-blind, phase III study to validate the benefit of pemetrexed maintenance therapy. Patients who had not progressed on four cycles of platinum-based chemotherapy were randomly assigned (2:1 ratio) to receive pemetrexed (500 mg/m², day 1) plus BSC or placebo plus BSC in 21-day cycles until disease progression (JMEN study). Pemetrexed maintenance therapy compared with placebo, significantly improved PFS (4.3 months vs. 2.6 months, respectively; p<0.0001) and OS (13.4 months vs. 10.6 months, respectively; p=0.012). With these results, pemetrexed was approved as switch maintenance therapy for NSCLC in the United States and Europe. This study included some squamous NSCLC cases (26% in the pemetrexed group and 30% in the placebo group). Sub-analysis revealed that pemetrexed maintenance is not beneficial toward PFS and OS in squamous NSCLC, a finding that is consistent with the results of a previous study of the effectiveness of pemetrexed used exclusively for non-squamous NSCLC.

Targeted therapies, including bevacizumab, cetuximab, erlotinib, and gefitinib, can be used as maintenance therapy for NSCLC. Continuous maintenance therapy with bevacizumab or cetuximab shows a benefit in PFS and OS. The results of clinical trials of erlotinib and gefitinib as continuous maintenance therapy were disappointing, but switch maintenance therapy with erlotinib and gefitinib produced a significant benefit compared with placebo or observation.

In most studies of the effect of maintenance therapy, PFS was the primary endpoint. Additionally, few studies report a significant difference in OS, indicating that maintenance therapy can increase PFS but have little impact on OS. Furthermore, in contrast to the definite standard of OS, that of PFS might differ among researchers. PFS cannot guarantee objectivity without an independent review. Thus, some issues arise concerning the real significance of PFS improvement in patients treated with maintenance therapy. This problem becomes more obvious when evaluating PFS after six cycles of chemotherapy rather than four cycles. In the SATURN study, OS was significantly increased only in the patients who showed SD after four cycles of chemotherapy. These data suggest that the SD in those patients might actually be progressive SD according to the Response Evaluation Criteria In Solid Tumors (RECIST), and so erlotinib maintenance therapy is practically early second-line chemotherapy25.

Insufficient data exist concerning the proportion of patients treated with second-line chemotherapy and the adequacy of regimens used as second-line treatment. In the PARAMOUNT study, only 4% of patients in the control group received pemetrexed as second-line chemotherapy at disease progression. The increase in PFS and OS with pemetrexed maintenance therapy might be due mainly to whether the patients were treated with pemetrexed, rather than the impact of maintenance therapy.

A clinician's expectation of the outcome of maintenance therapy is not merely an extension of PFS or OS. One of the major goals of maintenance therapy is increasing QOL. However, most clinical trials, except large-scale studies, did not evaluate QOL. Some studies, such as the PARAMOUNT and SATURN studies, reported that maintenance therapy did not affect QOL significantly. The JMEN study evaluated QOL objectively by assessing the time to symptomatic adverse events. The results indicated that pemetrexed maintenance therapy delays significantly serious adverse events such as severe pain or hemoptysis14. However, the value of maintenance therapy in terms of improvement of QOL remains to be clarified. When deciding to administer maintenance therapy, the pros and cons of this therapy in terms of multiple aspects should be considered.

Generally, only a proportion of NSCLC patients, ~20-80%, receive second-line chemotherapy. Many studies have reported consistent data that maintenance therapy markedly increases the rate of receiving second-line therapy, and might be a major benefit of maintenance therapy. However, Fidias et al.13 reported no OS improvement in maintenance therapy using docetaxel. Two possible interpretations of these results can be considered: 1) using docetaxel is an important issue or 2) sufficient PS to receive second-line therapy is more important issue than maintenance therapy.

According to the results of a study in Korea26, the proportion of patients receiving second-line therapy in Korea is high (86%) compared with that in other countries. The benefit of maintenance therapy in Korea, where most NSCLC patients undergo second-line chemotherapy, should thus be considered carefully.

More than four to six cycles of platinum-based combination chemotherapy increases toxicity and shows no survival benefit although those results in modest improvements in OS and QOL in NSCLC patients. However, recently updated evidence for maintenance therapy is encouraging. Patients are likely to cause tumor related symptom can get benefit from the maintenance treatment, who have a large tumor causing symptoms or no severe adverse event during previous therapy27. Thus far, no randomized studies have compared the effect of regimens in maintenance therapy, and the optimal drug for maintenance therapy remains to be identified. Despite much debate, continuous maintenance therapy with pemetrexed and switch maintenance treatment using erlotinib or pemetrexed has been revealed to be beneficial, and these treatments are currently recommended. More comprehensive studies are needed to validate the value of maintenance therapy.