Systemic Options for Second-Line Therapy and Beyond

Summary of Key Points

There are a growing proportion of patients who are candidates for second-line therapy and beyond.
Treatment choices are dictated by tumor histology, molecular phenotype (e.g., EGFR , ALK , ROS1 , etc.), and components of frontline chemotherapy including also the use of maintenance and bevacizumab.
In patients with no actionable molecular targets, several options are available that include chemotherapy (docetaxel, pemetrexed), epidermal growth factor receptor (EGFR)-targeted therapies (erlotinib and afatinib for squamous cell carcinoma), and ramacirumab (an anti-vascular endothelial growth factor receptor [VEGFR] 2 monoclonal antibody) in combination with docetaxel and immune checkpoint inhibitors.
Recent studies comparing EGFR tyrosine kinase inhibitors (TKIs) with single-agent chemotherapy for patients with known wild-type EGFR tumors have cast doubts on the clinical efficacy of second-line EGFR TKIs.
Patients with actionable molecular targets such as EGFR sensitizing mutations, ALK, and ROS1 translocation, who did not receive the appropriate targeted therapy in front line, must receive it in a second line strategies.
Third-generation EGFR TKI such as osimertinib for patients with tumors positive for T790 resistant mutation and second-generation anaplastic lymphoma kinase (ALK) inhibitors (ceritinib, alectinib and brigatinib) showed relevant clinical activity when tested in the second-line setting.
In the second-line setting immune checkpoint inhibitors (nivolumab, pembrolizumab, and atezolizumab) showed an overall response rate of 5% to 40% in patients with durable responses, but at the present time there are no reliable biomarkers predicting the subgroup of patients who would most benefit. Candidate biomarkers include PDL-1 expression, mutational load, and neoantigen expression.
Although erlotinib is approved for third-line therapy, the strength of the evidence is limited and a new wave of clinical trials is needed.
Studies comparing EGFR TKIs with single-agent chemotherapy for patients with known wild-type EGFR tumors have cast doubts on the clinical efficacy of second-line EGFR TKIs.

Acknowledgment

Dr. Goss thanks his research associate Johanna Spaans and his administrative assistant Valerie Smaglinskie for their support in drafting this chapter.

In order to discuss the systemic options for the management of patients in whom first-line chemotherapy has failed, it is necessary to revisit certain aspects of the disease that are covered elsewhere in this text in greater detail. Lung cancer is the most common cancer worldwide and it accounts for approximately 28% of all cancer deaths. For therapeutic purposes, nonsmall cell lung cancer (NSCLC), which includes 85% of all lung cancers, is divided into squamous and nonsquamous histology, for which there are differing approaches to treatment. Substantial advances in our understanding of the biology and molecular pathology of NSCLC have allowed us to identify oncogenic drivers and molecular biomarkers predictive of efficacy with targeted therapies, which further divides NSCLC into smaller therapeutic subgroups. However, even in the case of the most intensively studied of these oncogenic drivers, the EGFR gene, there are still many unanswered questions. For example, at a histologic level, the mechanism of the evolution of NSCLC to small cell lung cancer is unknown and our knowledge of the genetic evolution of NSCLC between the first- and second-line settings is scanty. Large studies evaluating the genetics of lung cancer in primary and metastatic sites sequentially across first, second, and subsequent lines of therapy have not been reported in the literature. In the setting of first-line treatment, controversy exists about treating patients with maintenance chemotherapy or waiting until disease progression. In the second-line setting, the appropriate treatment is unresolved and the number of patients eligible for second-line therapy following maintenance chemotherapy is uncertain, as there are conflicting data in the literature. A study by Fidias et al. comparing early maintenance chemotherapy with treatment at the time of progression demonstrated that only 37% of patients received second-line treatment at the time of progression compared with 95% who were treated with immediate maintenance. Overall survival favored the maintenance arm (median overall survival, 12.3 vs. 9.7 months; p = 0.853), suggesting that immediate maintenance chemotherapy was the treatment of choice. However, the authors noted that the median overall survival for patients in the deferred-treatment arm who received second-line therapy (37%) was 12.5 months, equivalent to that for patients in the maintenance arm. In contrast, Bylicki et al., in a three-arm randomized study that enrolled 464 patients, found that with careful observation, 95% of patients in the observation arm were eligible for second-line chemotherapy at the time of disease progression (84% received study-defined second-line treatment) and that there was no difference in survival between the maintenance arm and the observation arm. The close observation required to establish eligibility for second-line treatment, however, may not be feasible outside of a clinical trial. In an older retrospective review by Murillo and Koeller , published in 2006, among patients with stage IIIb and IV NSCLC treated at 10 community centers in the United States, 84% received first-line, 56% received second-line, 26% received third-line, 10% received fourth-line, and 5% received fifth-line chemotherapy. Further questions have recently been generated by the addition of the immune checkpoint inhibitors to our second-line armamentarium. These agents are by no means the universal panacea with overall only 5% to 40% of patients responding to treatment and our ability to identify the benefiting group mediocre at best. With preliminary data suggesting intriguing signals of efficacy of immune checkpoint inhibitor monotherapy in certain treatment-naive subgroups and additional activity in combination with chemotherapy in the upfront management of unselected patients with advanced NSCLC, the role of immune checkpoint inhibitors as second-line agents is uncertain.

History

It was not until 1995 that a large meta-analysis of eight randomized studies comparing cisplatin-based combination chemotherapy with best supportive care for advanced NSCLC demonstrated with certainty that chemotherapy had a modest impact on survival, with a median survival improvement from 4 to 7 months and a 1-year survival rate of 5% to 15%. The result of the meta-analysis was subsequently confirmed in a four-arm randomized phase III study evaluating response and survival rates between third-generation chemotherapies (e.g., paclitaxel, docetaxel, and gemcitabine) when combined with a platinum agent (either cisplatin or carboplatin). This study demonstrated a modest survival improvement, with a median survival of 7.9 months and a 1-year survival rate of 33%. However, it should be noted that platinum-doublet chemotherapy control arms of recent trials have been associated with better median overall survival than chemotherapy arms in earlier trials, most likely because of better performance status of the study population and some stage migration.

Until the two publications by Shepherd et al. and Fossella et al. in 2000, the role of second-line chemotherapy was uncertain. The literature consisted of phase I and phase II trials, most of which were small and consisted of fewer than 30 patients; in addition, details about prior treatment and patient performance status were frequently not included in the publication. Furthermore, although response rates were reported, very few studies provided median survival or 1-year survival rates. A review of the literature demonstrated disappointing results for clinical trials in a second-line setting, with most studies showing response rates of less than 10% and median survival times of 4 months or less. The agents most frequently evaluated in phase II studies included the vinca alkaloids vindesine and vinorelbine, the taxanes (paclitaxel and docetaxel), and gemcitabine. Variable and conflicting results were reported with second-line vinorelbine and in two trials in which vinorelbine, 25 mg/m ² or 20 mg/m ² , was administered weekly, no responses were seen. However, in a small trial of vinorelbine 30 mg/m ² that included only 10 patients a 20% response rate was reported by Sandora et al. Several studies that tested paclitaxel also produced conflicting results, perhaps in part because of the variability in both the dose and administration time. No responses were seen in a small study in which paclitaxel 140 mg/m ² was administered over 96 hours. In another trial, paclitaxel 200 to 250 mg/m ² was administered over 24 hours and 2 (14%) confirmed partial responses were noted, with two additional responses that lasted less than 4 weeks. In two trials in which varying doses of paclitaxel were given over 1 hour, 1 (2.5%) of 13 patients in the first study had a response, whereas 26 responses (25%) occurred in the second study. Gemcitabine was also investigated, and Gridelli et al. noted partial responses in 6 (20%) of 30 patients treated with gemcitabine 1000 mg/m ² weekly for 3 weeks of a 4-week cycle. However, the most extensively studied agent was docetaxel. In phase II trials, docetaxel was administered at 100 mg/m ² every 3 weeks, with objective response rates ranging from 15% to 22%. These promising results led to two randomized studies of second-line docetaxel for patients previously treated with cisplatin-based chemotherapy. (These studies will be discussed in more detail later in the chapter.)

Second-Line Chemotherapy

Only one randomized phase III trial has compared chemotherapy plus best supportive care to best supportive care alone for patients with advanced NSCLC previously treated with platinum-based chemotherapy. Patients with a performance status of 0 to 2, stage IIIb or IV disease with either measurable or evaluable tumor who had received one or more platinum-based chemotherapy regimens were randomly assigned to docetaxel 100 mg/m ² or 75 mg/m ² plus best supportive care every 3 weeks or best supportive care only. The primary outcome of the study was overall survival, and the secondary end points included objective tumor response, duration of response, and change in quality of life. All patients in the docetaxel arms were assessed every 3 weeks. Of the 204 patients randomized, 104 patients were assigned to the docetaxel arm, 84 had measurable lesions, and 6 (7.1%) of the 84 had a partial response. Time to progression was longer for patients treated with docetaxel than for patients who received best supportive care only (10.6 vs. 6.7 weeks; p < 0.001), as was the median survival (7.0 vs. 4.6 months; log-rank test, p = 0.047). The difference was more significant for the 75-mg/m ² dose of docetaxel compared with the best supportive care arm (median survival, 7.5 vs. 4.6 months; log-rank test, p = 0.010; and 1-year survival rate, 37% vs.11%; p = 0.003). Adverse events included febrile neutropenia, which occurred in 11 patients treated with docetaxel at 100 mg/m ² , 3 of whom died, and in 1 patient treated with docetaxel, 75 mg/m ² . Grade 3 or grade 4 nonhematologic toxicity with the exception of diarrhea occurred at a similar rate in both the docetaxel and best supportive care arms. In this study, the 100-mg/m ² dose was associated with five reported toxicity-related deaths. Three of the deaths were docetaxel-related, and an association with docetaxel treatment could not be ruled out for the other two deaths. At this dose, the median number of cycles delivered was only two and this, combined with a 10% early death rate, probably accounted for the lack of improved survival in the 100-mg/m ² treatment arm. When the docetaxel dose was reduced to 75 mg/m ² in the second half of the trial, dose delivery improved, with a median of four cycles, and the rate of febrile neutropenia decreased from 22% to 2%, with no toxicity-related deaths. This high rate of toxicity-related death had not been seen in other phase II studies involving a dose of 100 mg/m ² , but led the authors to conclude that only docetaxel at a dose of 75 mg/m ² is associated with significant prolongation of survival. Of note, clinical benefit in this study could be demonstrated by end points other than response and survival. A significant positive effect of docetaxel was evident in the analysis of both the usage of narcotics and nonnarcotics for pain and in the need for radiotherapy. In summary, this was the first trial to document that in patients with advanced NSCLC and good performance status, second-line chemotherapy with docetaxel 75 mg/m ² after platinum-based chemotherapy is justified, with a significant prolongation of survival and reduced pain.

The above findings were supported by a three-arm multicenter, open-label randomized phase III trial of patients with stage IIIb or IV NSCLC who progressed on platinum-based therapy. The trial was designed to compare docetaxel 100 mg/m ² every 3 weeks and docetaxel 75 mg/m ² every 3 weeks to a control arm of vinorelbine 30 mg/m ² administered intravenously on days 1, 8, and 15 of the 3-week cycle or ifosfamide 2 g/m ² on days 1–3 of a 3-week cycle (with the choice of drug left to the investigator’s discretion). Patients had to have either measurable or assessable lesions and an Eastern Cooperative Oncology Group (ECOG) performance status of 0 to 2. No restriction was based on the number of prior regimens or the amount of prior chemotherapy. A total of 373 patients were randomly assigned, and the three treatment groups were well-balanced for key patient characteristics. The overall response rates were 10.8%, 6.7%, and 0.8% for the docetaxel 100 mg/m ² , docetaxel 75 mg/m ² , and the vinorelbine or ifosfamide arms, respectively. Patients who received docetaxel had a longer time to disease progression ( p = 0.046) and a greater progression-free survival (PFS) at 26 weeks ( p = 0.05). Although the overall survival was not significantly different between the three arms, the 1-year survival was significantly greater for the docetaxel 75 mg/m ² arm when compared with the control arm (32% vs. 19%; p = 0.025). Prior exposure to paclitaxel did not decrease the likelihood of response to docetaxel nor did it impact survival. The authors concluded that clinical benefit as determined by objective response, PFS, and 1-year survival favored patients who received docetaxel. Grade 4 neutropenia and fever were higher in the two docetaxel arms than in the control arm; however, other treatment-related adverse events were similar across the three arms.

These two studies, supported by data from multiple phase II studies, resulted in docetaxel being registered with the US Food and Drug Administration (FDA) and the European Medicines Agency as an approved chemotherapy agent for second-line treatment for advanced NSCLC. However, despite the prolongation in 1-year survival by 10% to 20% and improved quality of life when compared with ifosfamide, vinorelbine, or best supportive care alone, these gains were modest, which led to the evaluation of pemetrexed, a novel multitargeted, antifolate in the second-line setting. This compound inhibits the enzyme thymidylate synthase, resulting in decreased thymidine necessary for pyrimidine synthesis. As a drug that also inhibits dihydrofolate reductase and glycinamide ribonucleotide formyl transferase, vitamin supplementation with folate and vitamin B12 is required to limit hematologic and nonhematologic toxicity associated with pemetrexed, including neutropenic fever. Therefore, supplementation with folic acid at 0.35 mg to 1 mg orally daily and vitamin B12 at 1000 μg intramuscularly every 9 weeks is essential to control the toxicity of this drug and has been used in most trials investigating this agent. Phase II studies of pemetrexed in previously untreated patients with NSCLC demonstrated single-agent response rates of 17% to 23%. In a phase II study of pemetrexed for patients with advanced NSCLC who had disease progression within 3 months after completing first-line chemotherapy, the response rate was 8.9% and the median survival was 5.7 months. Based on the similar overall survival found for pemetrexed and docetaxel and the expected lower toxicity with pemetrexed, a multinational phase III study comparing these two agents in the second-line treatment of NSCLC was undertaken. The primary objective of this noninferiority study was to compare overall survival between the two treatment groups on an intent-to-treat basis. Secondary objectives were to compare toxicities, response rate, PFS, time to progression, time to treatment failure, time to response, duration of response, and quality of life between the two treatment groups. Eligible patients had to have a performance status of 0 to 2 and have received previous treatment with one chemotherapy regimen for advanced NSCLC. The study included 571 patients who were randomly assigned to receive pemetrexed 500 mg/m ² intravenously on day 1 plus vitamin B12, folic acid, and dexamethasone every 21 days or to receive docetaxel 75 mg/m ² intravenously on day 1 plus dexamethasone every 21 days. The overall response rate was 9.1% and 8.8% for pemetrexed and docetaxel, respectively ( p = 0.105). The PFS was 2.9 months in both arms, and the median survival was 8.3 months and 7.9 months, respectively. The 1-year survival rate in each arm was 29.7%. Patients receiving docetaxel were more likely to have grade 3 or 4 neutropenia ( p < 0.001), febrile neutropenia ( p < 0.001), and neutropenia with infection ( p = 0.004); hospitalization for neutropenic fever was more frequent in the docetaxel arm (13.4% vs. 1.5%; p < 0.001). Use of granulocyte colony-stimulating factor support was also greater in the docetaxel arm (19.2% vs. 2.6%; p = 0.001) than pemetrexed. The authors concluded that, in patients with advanced NSCLC in whom one line of previous chemotherapy had failed, pemetrexed was equivalent to docetaxel in terms of clinical efficacy but with fewer side effects and should therefore be considered a standard treatment option in the second-line NSCLC setting.

Despite good overall clinical efficacy in these trials, not all patients benefit from pemetrexed. In a retrospective analysis of phase III pemetrexed studies Scagliotti et al. found significant treatment-by-histology interaction for overall survival and PFS. Specifically, patients with nonsquamous tumors treated with pemetrexed had a significantly longer overall survival (hazard ratio [HR]: 0.78; 95% confidence interval [CI], 0.61–1.00; p = 0.48) and PFS (HR: 0.82; 95% CI, 0.66–1.02; p = 0.076) than patients treated with docetaxel. Conversely, patients with squamous tumors treated with pemetrexed appeared to have a worse overall survival and PFS (overall survival: HR: 1.56; 95% CI, 1.08–2.26; p = 0.018 and PFS: HR: 1.40; 95% CI, 1.01–1.96; p = 0.004) compared with docetaxel. The treatment-by-histology interaction test for overall survival and PFS was p = 0.001 and p = 0.004, respectively. This finding, confirming the benefit of pemetrexed for nonsquamous histology, has been supported by the findings of studies with pemetrexed in the first-line and maintenance settings.

While a number of questions have been answered in trials, reviews, or meta-analyses of the literature, a number of questions remain. Is a combination of two or more drugs superior to single-agent chemotherapy, and is a weekly schedule better than an every 3-week schedule?

Choice of Chemotherapy Agent

A review of multiple randomized phase II trials comparing docetaxel to single-agent paclitaxel, gemcitabine, ifosfamide, vinorelbine, and pemetrexed demonstrated that none of these agents was superior to docetaxel in the second-line setting. In this review, two-drug combinations, both platinum and nonplatinum, were also compared with docetaxel in multiple phase II randomized studies. Among platinum-based doublets, none was found to be superior to docetaxel in the second-line setting. Four randomized studies compared a single-agent with a two-drug nonplatinum-based regimen, and three trials compared docetaxel to a combination of docetaxel plus gemcitabine or docetaxel plus irinotecan. Of note, in all trials reviewed, none of the two-drug regimens was shown to improve survival. Furthermore, toxicities were more common among combination regimens, sometimes leading to toxicity-related deaths or negative outcomes related to symptom relief, prolongation of survival, and improved quality of life for patients with incurable disease, which are the primary aims of second-line treatment. Similarly, doublet therapy that includes pemetrexed does not appear to improve survival compared with single-agent pemetrexed in the second-line setting, based on the findings of a meta-analysis. The comparable efficacy of single-agent and doublet chemotherapy in the second-line setting has been supported by four meta-analyses in the literature.

Scheduling of Chemotherapy

Three randomized trials (one phase II and two phase III studies) compared weekly docetaxel delivery to the classic schedule of every 3 weeks. In the phase II study, response, median survival, and 1-year survival were not significantly different but favored the every 3 weeks regimen. Similarly, the two phase III studies did not show a difference in overall survival or quality of life.

The number of cycles of chemotherapy that a patient should receive in the second-line setting is a matter of debate. This question has not been answered, as it has not been formally addressed in a randomized trial. In the trials of both Shepherd et al. and Hanna et al., patients were treated until disease progression and the mean number of cycles was four. The reason leading to treatment discontinuation has been inconsistently reported in the literature, but it is most likely due to drug-related toxicity and disease progression. Given that time to progression in randomized phase III trials ranges from 2 to 3 months, corresponding with three or four cycles of chemotherapy, progression may be considered the main reason for discontinuation of second-line treatment. In conclusion, following reviews of four large meta-analyses of second-line trials in the literature, single-agent docetaxel or single-agent pemetrexed administered every 3 weeks remains the gold standard for good performance status patients (without a known treatable oncogenic driver) eligible for chemotherapy. This is detailed in guidelines from both the National Comprehensive Cancer Network and the American Society of Clinical Oncology.

Third- and Subsequent-Line of Chemotherapy

If patients treated with a targeted agent are excluded, there are scanty data regarding the outcomes of patients treated with chemotherapy following first- and second-line chemotherapy for advanced NSCLC. In a retrospective analysis, Massarelli et al. reviewed 700 patient records and identified patients who had received at least two chemotherapy regimens, including at least one course of platinum-based chemotherapy and one course of docetaxel. In this review, the response rate to first-line chemotherapy for all 700 patients was 20.9%; the rates were 16.3%, 2.3%, and 0% for second-line, third-line, and fourth-line chemotherapy, respectively. The disease-control rate also decreased dramatically from first- to fourth-line treatment, although it was higher for second-line treatment (74.4%) than for first-line treatment (62.8%). The median overall survival time from the start of the last chemotherapy, either first- or fourth-line treatment, was 4 months. Patients with stage III disease at initial diagnosis had a longer overall survival from diagnosis than patients with stage IV disease ( p = 0.02). These data suggest that treating patients with currently available chemotherapy regimens following two lines of chemotherapy should not be standard of care and that further chemotherapy should be explored in the context of a clinical trial.

Second-Line Treatment with Molecularly Targeted Agents

Many of the molecularly targeted therapies were first investigated as second- or third-line therapy, especially at the time when a reliable biomarker was not available. Here, we review the role of a number of molecularly targeted therapies and their comparative data with single-agent chemotherapy.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here