Lung cancer remains the single largest cause of cancer-related deaths, and the burden of the disease in the elderly population will only grow as life expectancy increases.
Rogerio C. Lilenbaum, MD
Lung cancer remains the single largest cause of cancer-related deaths, and the burden of the disease in the elderly population will only grow as life expectancy increases.1,2An estimated 50% of new lung cancers are diagnosed in patients 65 years of age and over, and up to 40% are diagnosed in patients more than 70 years of age.2,3
Targeted agents that inhibit the tyrosine kinase activity of the epidermal growth factor receptor (EGFR) and those that target the vascular endothelial growth factor (VEGF) pathway have been valuable additions to the treatment of non-small cell lung cancer (NSCLC), resulting in improved survival; however, in clinical trials designed to evaluate safety and efficacy of these agents, elderly patients are generally a minority.4,5
Chandra P. Belani, MD, on Treating Older Patients with NSCLC
Belani is the deputy director at Penn State Hershey Cancer Institute.
In a recent single-center, retrospective, cohort study of elderly Chinese patients (N = 56; >70 years) with advanced/metastatic, nonresectable NSCLC, the median age at diagnosis was 73 years, and most patients (60.7%) had received only a single line of treatment, with less than one quarter (23.3%) receiving more than two successive lines of therapy.2In this study, paclitaxel/platinum doublets were the most commonly used first-line chemotherapy.2The median progression-free survival (PFS) and overall survival (OS) for this population of elderly patients was 10 and 19 months, respectively; in univariate analysis.2Only better baseline performance status (PS) was correlated with longer PFS, while better OS was predicted by better baseline PS, AD cell type, more lines of treatment, and better Charleston Comorbidity Index (CCI) scores.2Remarkably, the tolerability of chemotherapy or tyrosine kinase inhibitor (TKI) treatment was similar in the elderly population, and survival was similar to that seen in younger patients, suggesting that PS rather than age should be at least one clinical parameter to consider when deciding on chemotherapy versus TKI-based therapy in the elderly.2
Because of the under-representation of elderly patients in clinical trials, subgroup analyses, both preplanned and unplanned, have been used as an alternative to clinical studies that are specifically targeted to elderly patient populations. These analyses have generally indicated that, while the elderly may be subject to increased toxicity, they are able to tolerate both combination therapy and the use of targeted agents.4-6Indeed, it was recognized as early as 2003 that elderly patients with advanced NSCLC could be safely administered both single-agent and some combination-agent chemotherapies.3For example, in 2001, a study of NSCLC patients 70 and over (The ELVIS Trial) found single-agent vinorelbine with best supportive care (BSC) to be superior to BSC alone, with median survival of 28 and 21 weeks in the respective groups. Importantly, patients on vinorelbine + BSC experienced improved quality-of-life measures, including global health status and measures of cognitive, social, and physical functioning, although drug-related toxicities did affect scores for nausea/vomiting and constipation, which were lower in the vinorelbine group.3,7Similarly, in the CALGB study of single-agent paclitaxel versus paclitaxel with carboplatin, both response rate (RR) and survival were superior in the combination-therapy arm, and, in the elderly subgroup of patients, results were similar to those of the overall population, although the benefit in survival (median 8.0 months vs 5.8 months) did not reach significance.3,8Although neutropenia rates were higher in the combination arm, there were no neutropenia-related deaths and only 3% or less of patients in either group experienced grade 4 infections.3,8
More recent studies have demonstrated the efficacy and safety of gemcitabine/carboplatin and pemetrexed/carboplatin combinations for the treatment of elderly patients with NSCLC, with response rates comparable to those seen in general NSCLC populations.9Costs associated with these different strategies are also important to consider in the elderly, who may be more susceptible to chemotherapy-related adverse events (AEs).10For example, a recent cost-efficacy analysis comparing the use of the TKI erlotinib versus chemotherapy as a first-line option for fit elderly patients with NSCLC found that, while overall patient outcomes did not differ significantly, erlotinib followed by chemotherapy compared favorably from a cost perspective as compared with the reverse strategy.10
Rogerio C. Lilenbaum, MD, from the Department of Medical Oncology at Smilow Cancer Hospital and the Yale Cancer Center, identified some recent data that effectively summarize the current status of treatment for elderly patients with NSCLC. The first study, published in 2011 inLancet Oncologyand authored by Elisabeth Quoix, MD, University Hospitals of Strasbourg, France, and coworkers, examined the use of combination or single-agent therapy in elderly patients with NSCLC.11Elderly patients (70-89 years of age) in this multicenter trial were randomly assigned (1:1) to doublet chemotherapy with carboplatin and paclitaxel, or to single-agent therapy with gemcitabine or vinorelbine; the primary endpoint was OS.11The trial enrolled 451 patients with a mean age of 77 years across treatment groups; at baseline, 118 patients (26.1%) had a PS of 2, and the median follow-up was 30.3 months. An independent data-monitoring committee recommended that the recruitment be stopped based on favorable results seen at the second interim analysis. The results showed doublet chemotherapy was superior for OS (multivariate analysis hazard ratio [HR] = 0.64; 95% confidence interval [CI] = 0.52-0.78;P<.0001) and for PFS (HR = 0.51; 95% CI= 0.42-0.62;P<.0001). For grade 3 and 4 AEs, neutropenia, febrile neutropenia, thrombopenia, sensory neuropathy, and anemia were more common in the double group as compared with the monotherapy group; a total of 10 deaths (4.4%) attributable to treatment occurred in the doublet arm, as compared with 3 deaths (1.3%) in the monotherapy arm.11Thus, despite increased toxicity, there was a significant benefit for elderly patients assigned to doublet, platinum-based chemotherapy, as compared with single-agent gemcitabine or vinorelbine.11Lilenbaum noted that the Quoix study was a benchmark trial, in that it “answered a decade-long question of whether or not elderly patients should be treated with a combination or a single agent,” with the answer to this question being “a resounding ‘yes’ for combination therapy.” He cautioned, however, that toxicity in the trial was not trivial, and there were “probably more treatment-related deaths than one would consider acceptable in practice,” which may possibly have been due to some of the dosages that were used for the combination regimen or due to other supportive care issues. Nonetheless, he noted that the results “set the tone for any discussion of chemotherapy treatment of elderly patients.”
Clearly, however, some elderly patients are not appropriate candidates for combination therapy because of the presence of significant comorbidities and/or borderline or compromised PS. Therefore, the issue for most clinicians becomes how best to stratify elderly patients for combination or single-agent therapy. In this regard, Lilenbaum pointed to some data emerging from a French trial, Elderly Selection on Geriatric Index Assessment (ESOGIA), which was presented at the 15th World Conference on Lung Cancer in October 2013, which essentially took the question a step further.12In this trial, the experimental arm (arm B) used the Comprehensive Geriatric Assessment (CGA) to evaluate patients, and depending on CGA score, the patients (N = 493, randomized) were categorized and stratified into different treatment groups; for fit patients a carboplatin-based doublet was used, for vulnerable patients monotherapy was used, and for frail patients BSC was used. The control arm in the study (arm A) was a group in which the decision to use combination or single-agent therapy was driven by a standard treatment algorithm, based on investigator assessment alone (ie, essentially dictated by PS and age). The main outcome was time to treatment failure (TTF).12“This was the first time ever in more than 2 decades of research of geriatric assessments in which the assessment guides the type of treatment,” Lilenbaum said. Interestingly, but perhaps disappointingly, however, he noted that the results showed that the assessment-guided arm was not better than the physician-guided arm (median TTF, 99 days vs 98 days in arms A and B, respectively;P= .7149); meaning that, essentially, physicians without a formal instrument were still doing an effective job of assessing patients for combination versus single-agent therapy. Evaluating the results from this study, and also noting the extensive research that has been done to create reliable and consistent geriatric assessments for NSCLC, Lilenbaum noted that while “it would be desirable to have a reproducible instrument that we can all use instead of just our clinical judgment… in the end, it looks like clinical judgment still does a good job. I think the elderly patient, in general, should be treated with a combination regimen,” Lilenbaum said, noting that the Quoix trial included a population of patients that was “pretty representative of that seen in our clinical practice,” ranging in age from 70 to 89, and with roughly a quarter of the patients having a PS of 2. Despite these findings, he cautioned that there are patients with high-risk features, based on comorbidities and other issues that are captured on the geriatric assessment, who are likely not appropriate for more aggressive (ie, combination) therapy. “How we identify these patientsâ”€upfrontâ”€still remains an area of investigation.”