Management of Recurrent Non–Small-Cell Lung Cancer | 9 |
INTRODUCTION
The management of recurrent non–small-cell lung cancer (NSCLC) has become more challenging over the past 10 years. Previously, few chemotherapy options were available and all patients were viewed as having the same disease with no discrimination based on tumor histology or genetic biomarkers. Front-line therapeutic options were often toxic and many patients were too ill to consider second-line treatment at the time of cancer progression. With the advent of less toxic therapy and improved symptom management, many patients are now able to receive therapy beyond the front-line regimen.
It is important to be mindful of the goals of therapy when treatment is considered. The majority of patients with recurrent disease will be treated with palliative intent to improve symptoms and maintain quality of life. Rarely, recurrent NSCLC may present in a limited manner with the potential for cure, and these situations should not be missed.
PRESENTATION OF RECURRENCE
NSCLC has a high rate of recurrence regardless of the initial stage and treatment. Stage I–II patients with completely resected disease have a 30% to 60% chance of metastatic spread that often manifests within the first 2 years postresection. Similarly, up to 80% of patients with stage III disease treated with definitive local-regional therapy will recur. Stage IV NSCLC is almost uniformly incurable despite aggressive front-line, platinum-based therapy, or targeted therapy for driver mutation-positive disease. These numbers suggest that 80% of patients with NSCLC will ultimately relapse and may be eligible for subsequent therapy.
Upon evidence of recurrent disease, repeat staging scans and a biopsy to confirm relapse are strongly recommended, especially if oligometastatic disease is identified on imaging and aggressive treatment is considered. For patients with early stage disease, relapse may occur locally at the site of surgical resection or radiation therapy (RT). In addition, regional spread to hilar and/or mediastinal lymph nodes is not uncommon after primary resection. Rarely, patients may present with a metachronous, isolated, metastatic site in the brain, contralateral lung, or adrenal gland. However, the majority of patients develop diffusely metastatic disease. Therapeutic options at the time of relapse are based on the initial treatment received, the location of relapse, tumor burden, tumor histology, presence of actionable genetic biomarkers, and patient performance status (PS).
INTRATHORACIC LOCAL-REGIONAL RECURRENCE
Patients may develop a local recurrence after definitive therapy, such as surgical resection for stage I–II NSCLC or chemoradiotherapy for stage III disease. If the repeat staging work-up fails to identify extrathoracic metastases, further local therapy should be considered since some patients may still be cured. A new, second primary lung malignancy can occur in up to 5% of such patients annually and should be treated with surgical resection or stereotactic body radiation (SBRT) if detected at an early stage. Similarly, patients may present with a regional (e.g., hilar or mediastinal) nodal recurrence after surgical resection for stage I or II disease, and definitive chemoradiotherapy should be strongly considered with intent to cure. Progression of a tumor nodule previously treated with SBRT may be treated with radiofrequency ablation (RFA) with good long-term outcome. It is important to consider these potentially curative treatment options for recurrent disease isolated to the thorax, since systemic therapy alone will only delay disease progression.
OLIGOMETASTATIC RECURRENCE
Adrenal Gland Metastasis
Aggressive local treatment should be considered for NSCLC that metastasizes to an adrenal gland after definitive therapy for early-stage disease. An adrenal gland metastasis identified after surgical resection may also be surgically resected with the possibility for improved long-term outcome (1). A pooled analysis of the published literature reported a 5-year survival rate of 25% after adrenalectomy, although the benefit of this approach is typically limited to patients with previously resected stage I disease who develop an adrenal metastasis months after the primary tumor was removed. The benefit of resection of an isolated adrenal metastasis in resected stage II disease remains unproven, and adrenal metastectomy should not generally be considered in stage III NSCLC given the extremely high likelihood of further metastatic disease.
Brain Metastasis
Metachronous brain metastasis after definitive treatment for localized NSCLC is a relatively common occurrence. Although treatment of brain metastases is beyond the scope of this chapter, surgical resection should be considered for a solitary brain metastasis, especially if the patient has significant neurologic symptoms. One to three metastases may be treated with stereotactic radiosurgery with or without whole brain radiation (WBRT). More than three brain metastases will likely require WBRT, given the high chance of subclinical brain metastases not identified on imaging. A careful discussion between the patient, medical oncologist, and radiation oncologist is required to personalize the treatment approach based on symptoms, comorbidities, treatment preferences, and the extent of extracranial disease. Systemic therapy following definitive local treatment of grossly visible brain metastases has not been shown to improve patient outcome and should be considered on a case-by-case basis.
Other Metastatic Sites
Rarely, NSCLC may present as an isolated metastasis to the liver, bone, or a distant lymph node, or other site. Limited case reports suggest an improved long-term outcome with definitive local therapy, although no high-quality data support this approach. In the absence of other metastatic sites, a definitive treatment approach with RT or resection is an option, especially if the patient has a good PS and has had a prolonged disease-free interval after treatment of early stage disease. This should be discussed in a multidisciplinary forum with input from surgery, radiation oncology, and medical oncology.
SYSTEMIC RECURRENCE
Non–Small-Cell Lung Cancer Without Driver Mutations
Immunotherapy
PD1 inhibitors: nivolumab and pembrolizumab
Agents that enhance the innate antitumor immune response have garnered much interest in oncology, especially in the treatment of lung cancer. Several immune checkpoint-inhibitor antibodies have been developed that target either the programmed death 1 (PD-1) receptor on activated T-cells or its ligand (PD-L1), which is expressed on tumor cells and infiltrating immune cells. The binding of PD-L1 to PD-1 results in the inactivation of T-cells and thus, the protection of tumors from cytotoxic immune surveillance. Monoclonal antibodies to either PD-1 or PD-L1 can interfere with ligand-receptor binding, allowing activated T-cells to attack the tumor. Two PD-1-directed monoclonal antibodies, nivolumab and pembrolizumab, are currently approved for second-line therapy after the failure of initial platinum-based therapy in NSCLC regardless of tumor histology.
Nivolumab (Odivo®) is a fully human IgG4 antibody that targets PD-1. Brahmer et al. randomized 272 patients with relapsed squamous cell NSCLC to nivolumab 3 mg/kg every 2 weeks or docetaxel 75 mg/m2 every 3 weeks until disease progression (2). The primary endpoint of overall survival favored nivolumab over docetaxel (median, 9.2 vs. 7.3 months, 1 year 42% vs. 24%; HR, 0.59, 95% CI 0.44–0.79; P < .001). Response rate (20% vs. 9%) and median progression-free survival (PFS; 3.5 vs. 2.8 months) were also improved with nivolumab compared to docetaxel. In addition, 20% of patients receiving nivolumab remained progression-free 1 year from initiation of treatment, suggesting a prolonged duration of response to therapy. Fewer grades 3 and 4 toxicities, mainly fatigue, decreased appetite, and leukopenia, were identified with nivolumab than with docetaxel (7% vs. 55%), and nivolumab did not cause any treatment-related deaths. The most common grades 1 and 2 toxicities with nivolumab were rash, nausea, diarrhea, and fatigue. Tumor cell PD-L1 expression by immunohistochemistry did not correlate with either prognosis or response to treatment.
Similarly, Borghaei et al. randomized 582 patients with relapsed non-squamous NSCLC to nivolumab 3 mg/kg every 2 weeks or docetaxel 75 mg/m2 every 3 weeks until disease progression (3). Overall survival, the primary endpoint, again favored nivolumab over docetaxel (median, 12.2 vs. 9.4 months; HR 0.73, 95% CI 0.59–0.89; P = .002). Response rate also favored nivolumab (19% vs. 12%), but median PFS favored docetaxel (2.3 vs. 4.2 months). Despite this unexpected finding, 19% of patients receiving nivolumab remained progression-free 1 year from initiation of treatment compared to 8% with docetaxel. Fewer grades 3 and 4 toxicities, mainly fatigue, nausea, and diarrhea, were identified with nivolumab than with docetaxel (10% vs. 54%). Common grades 1 and 2 toxicities with nivolumab included rash (9%), pruritus (8%), diarrhea (8%), and hypothyroidism (7%). One treatment-related death from encephalitis occurred with nivolumab. Contrary to the squamous trial, PD-L1 expression levels had a significant predictive association with all efficacy endpoints. However, this biomarker is not currently indicated to guide the use of nivolumab in routine practice. Nivolumab is now approved as second-line therapy in NSCLC regardless of histology or PD-L1 tumor expression.
Pembrolizumab (Keytruda®) is another monoclonal antibody that binds PD-1 and inhibits the binding of PD-L1 and PD-L2. KEYNOTE-001 was a phase I study of pembrolizumab in several tumor types, including NSCLC, which enrolled 61 patients with previously treated NSCLC with at least 50% of tumor cells with membranous PD-L1 expression (4). Pembrolizumab demonstrated a response rate of 44% and a median PFS of 6.1 months (95% CI 2.9–12.5). The median duration of response was 12.5 months in this selected patient population. Interestingly, pembrolizumab demonstrated a better response rate in current and former smokers than in never smokers (23% vs. 10%), and it is postulated that this is related to a higher tumor mutational burden resulting in greater neoantigen expression in smokers. As with nivolumab, the common side effects of pembrolizumab included fatigue (19%), pruritus (11%), decreased appetite (10%), and rash (9.7%). Pneumonitis was identified in 3.6% of patients, with one death (0.2%) from this complication. Based on an analysis of 1,143 NSCLC patients, 23% of previously treated patients had membranous PD-L1 expression in at least 50% of their tumor cells. According to the Keytruda (pembrolizumab) package insert (October 2015), the current indication for pembrolizumab is, “for the treatment of patients with metastatic non-small cell lung cancer (NSCLC) whose tumors express programmed death ligand 1 (PD-L1) as determined by an FDA-approved test, with disease progression on or after platinum-containing chemotherapy.” The need to confirm PD-L1 tumor expression will likely limit the use of pembrolizumab, since nivolumab can be prescribed as second-line therapy without the need for PD-L1 analysis.
PD-L1 inhibitors: atezolizumab and Durvalumab
A second checkpoint-inhibitor strategy targets PD-L1, thereby inhibiting interaction with PD-1. The two anti-PD-L1 monoclonal antibodies that are farthest along in clinical development are atezolizumab (MPDL3280A) and Durvalumab. Over 80 patients \3 weeks until disease progression (5). Although survival data are immature, the response rate was 17% for all evaluable patients and 27% for those with high PD-L1 expression. Fatigue, nausea, and decreased appetite were the most commonly reported toxicities.
A phase I study of Durvalumab included an expansion cohort of 178 patients with relapsed NSCLC treated with 10 mg/kg every 2 weeks (6). The response rate for patients treated with only one prior line of therapy was 19%, and was higher in patients with squamous cell carcinoma than with non-squamous NSCLC (21% vs. 13%). PD-L1-positive patients and current or former smokers had a higher chance of response. Preliminary results also suggest that patients with PD-L1-positive tumors have improved overall survival compared to those with PD-L1-negative tumors. Common side effects of Durvalumab included diarrhea (7%), rash (8%), hyperthyroidism (4%), and hypothyroidism (4%).
Pseudo-progression
Pseudo-progression has complicated the response assessment of immune checkpoint-inhibitors. Early studies, particularly in melanoma, identified a subset of patients whose tumors grew or who developed a new lesion after the initial few weeks of immunotherapy. Although treatment was discontinued, subsequent imaging revealed significant disease regression. In this select group of patients, biopsies of enlarging lesions demonstrated inflammatory cell infiltrates or tumor necrosis. However, this is a relatively rare occurrence in patients with solid tumors, including NSCLC, and reviews of limited trial data suggest that pseudoprogression occurs in less than 5% of patients (7). If early imaging suggests pseudoprogression, but symptoms related to the cancer are improved or controlled, it is reasonable to continue treatment and repeat imaging in 4 to 6 weeks to assess for delayed response. However, progressive disease on imaging accompanied by worsening symptoms, such as weight loss, fatigue, dyspnea, cough, or pain, suggests true progression and immune checkpoint therapy should be discontinued.
Practical use of immunotherapy
Given the improved survival and favorable toxicity profile of immune checkpoint-inhibitors when compared to standard chemotherapy, these agents are commonly the preferred second-line therapy for patients with relapsed NSCLC. Patients with a documented history of autoimmune disease (except vitiligo, Grave’s disease, or psoriasis not currently on therapy) or inflammatory bowel disease (i.e., Crohn’s disease or ulcerative colitis), or who are on immunosuppressive therapy (i.e., >10 mg of prednisone daily or its equivalent), should not be treated with immune checkpoint therapy.
It is encouraging that approximately 20% of unselected patients with advanced squamous and non-squamous NSCLC will achieve a durable response with anti-PD-1 or anti-PD-L1 antibodies. Unfortunately, clinical benefit is not seen in the majority of patients who receive these agents and rare immune-mediated side effects, such as pneumonitis, hepatitis, nephritis, colitis, and hypophysitis, can be life threatening. Patients must be carefully monitored and prescribers must have a low threshold to discontinue the agent and administer corticosteroid therapy to avoid serious complications. Future studies will focus on the toxicity and efficacy of combining immune checkpoint-inhibitors with RT, chemotherapy, and other immune modulators to enhance the therapeutic benefit of these agents.
Single-Agent Chemotherapy
The vast majority of patients with advanced NSCLC will not have a targetable driver mutation identified and will be treated with front-line, platinum-based chemotherapy. Of note, in patients who received adjuvant chemotherapy for resected, early-stage disease, or chemoradiotherapy for locally advanced disease, retreatment with a platinum-doublet regimen is reasonable if recurrence is identified greater than 6 months from completion of prior therapy. Patients who recur within 6 months of completing adjuvant chemotherapy or definitive chemoradiotherapy should be treated with a second-line agent. With few exceptions, clinical trials have demonstrated that multidrug regimens in the relapsed setting result in increased response rates and modest improvements in PFS, but are uniformly more toxic and have not improved overall survival (8). Patients with non-squamous NSCLC not tested for driver mutations at initial presentation should have the tumor analyzed for EGFR mutation and ALK and ROS1 gene rearrangement given the significant impact that therapy targeting these driver mutations can have on outcomes. Figure 9.1 provides an algorithm for managing patients with relapsed NSCLC without a targetable driver mutation.
Docetaxel
Docetaxel is the most extensively studied agent in patients with relapsed or refractory NSCLC. Two randomized, phase III trials have evaluated the efficacy of docetaxel in patients with advanced NSCLC who had progressed after prior chemotherapy (9,10). In the study by Shepherd et al., 204 patients who had been previously treated with a platinum-based regimen were randomized to receive docetaxel 100 mg/m2 or best supportive care (9). All patients had a PS of 0 to 2 and had not had any prior taxane therapy. Due to excessive toxicity in the first 49 patients assigned to receive docetaxel 100 mg/m2, including three treatment-related deaths, the dose was reduced to 75 mg/m2 for the subsequent 55 patients randomized to the treatment arm and no further treatment-related mortality was observed. Patients who received docetaxel 75 mg/m2 had a response rate of only 5.5%, but demonstrated a significant improvement in median (7.5 vs. 4.6 months) and 1-year overall survival (37% vs. 19%) compared to those assigned to best supportive care.
Fossella et al. conducted a three-arm randomized trial comparing two doses of docetaxel (75 mg/m2 and 100 mg/m2) to a control arm of either single-agent vinorelbine or ifosfamide in 373 patients who had progressed after one or more platinum-based regimens (10). Response rates were 6.7% for docetaxel 75 mg/m2, 10.8% for docetaxel 100 mg/m2, and 0.8% for control chemotherapy. One-year survival rates for patients on docetaxel 75 mg/m2, docetaxel 100 mg/m2, and control chemotherapy were 32%, 21%, and 19%, respectively. Neither response rate nor survival was significantly impacted by prior taxane exposure, suggesting that docetaxel and paclitaxel are not completely cross-resistant.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
