Lung cancer is the most lethal cancer known to man, claiming the lives of over 160,000 people each year in the United States alone.[1] The mainstay treatment for advanced stage non–small-cell lung cancer (NSCLC) has been the use of platinum-based doublet therapy, however this treatment has led only to a modest improvement in survival and symptomatic relief.[2] Current research focuses on studying and identifying molecular targets on cancer cells in order to generate personalized treatments that are based upon an individual’s molecular profile. Two of the most studied and validated targets are epidermal growth factor receptor (EGFR) and vascular endothelial growth factor (VEGF). EGFR, important for cell proliferation and over-expressed in a large portion of NSCLC patients, can be inhibited either by tyrosine kinase inhibitors (TKI), such as erlotinib or gefitinib(Drug information on gefitinib), or by the monoclonal antibody cetuximab(Drug information on cetuximab). Cetuximab has shown promise in increasing survival benefit and response rates in a large phase III trial when added to platinum-based chemotherapy[3] though, in other studies it failed to increase PFS.[4] VEGF is an important angiogenesis-stimulating factor and is inhibited by the monoclonal antibody bevacizumab(Drug information on bevacizumab). A randomized phase II trial showed an increase in survival benefit in patients with advanced NSCLC who were given bevacizumab in addition to standard chemotherapy (versus standard chemotherapy alone), but these studies also showed a moderate increase in the risk of treatment-related deaths.[5]
The varying success of cetuximab and bevacizumab in treatment for lung cancer leaves much room for further research toward the goal of successful clinical application. In this issue of ONCOLOGY, Cesare Gridelli and Antoni Rossi call attention to current studies involved in identifying clinical and biological markers that may predict treatment outcomes and be used for selecting appropriate patients for monoclonal antibody treatment.
The authors examined large-scale studies such as the Eastern Cooperative Oncology Group (ECOG – E4599) study and the AVAiL trail for bevacizumab outcomes.[5, 6] In these studies, bevacizumab was reported to cause more clinically significant bleeding, specifically in the presence of tumor cavitations instead of squamous histology or central tumor localization. Gender, hypertension, VEGF, and ICAM were all considered as clinical and biological markers predictive for the beneficial effect of bevacizumab therapy, but no markers were sufficiently reliable for implementation in clinical practice.[7]
The authors described the large phase II First-Line ErbituX in lung cancer (FLEX) trial, the multicenter phase II BMS099 study and the Cetuximab in Advanced Lung Cancer (CALC-1) study in order to illustrate the data on the effects of cetuximab. These studies showed improvement in objective response rate (ORR), overall survival (OS), and progression-free survival (PFS) (with the exception of PFS in the smaller BMS099 study) in all histology groups when cetuximab was added to platinum-doublet therapy (cisplatin/vinorelbine or carboplatin(Drug information on carboplatin)/paclitaxel).[3, 4, 8, 9] Acne-like rash indicated predictive positive responses to cetuximab treatment and increased OS. A retrospective analysis of the FLEX trial also showed cetuximab to be effective independent of mutations in K-ras or the EGFR kinase domain.[10]
This article constructs a comprehensive review of the varying factors found in studies that influence the results of monoclonal antibody treatment in advanced NSCLC. Although including cetuximab or bevacizumab in treatment regimens for advanced NSCLC clearly benefits certain patients, clinicians still lack specific biomarkers that can accurately predict patient response to therapy or toxicity. The use of biomarkers could potentially exclude patients who will not benefit or even be harmed by antibody treatment. As increasing biomarker research has brought about new and innovative cancer-specific molecular targets, the heterogeneous nature of tumors has also become significantly apparent, thereby diminishing our hopes for the effectiveness of generalized, single-treatment regimens for all patients, even among those with the same histology.[11] Here we point to some recent developments in research that should further our goal of providing personalizing treatment for patients.
One biomarker that is currently under investigation for the selection of patients to cetuximab therapy is EGFR amplification, measured by fluorescent in-situ hybridization (FISH). A phase II study evaluating EGFR FISH as a predictive marker in NSCLC patients treated with cetuximab plus chemotherapy showed promising predictive results. FISH-positive patients, with a high level of polysomy (> four copies of the gene in > 40% of cells), were compared to FISH- negative patients, with no or low genomic gain (< four copies of the gene in > 40% of the cells)[12] and showed a significantly improved progression-free survival advantage: Six months compared to three months, and an overall survival benefit of 15 months versus seven months, respectively.[13] The results of this study, which is being further validated by the randomized phase III SWOG 0819 study (which compares chemotherapy with and without cetuximab) demonstrates not only the importance of receptor presence but also the quantity of receptor expression in predicting patient response. It is known, however, that the growth of tumor cells primarily involves multiple changes and pathways of resistance that extend beyond single genetic alterations such as amplifications or mutations.[14] Currently, new technologies involving protein and gene arrays have allowed entire genomes and proteomes to be assayed for new biomarkers and have identified useful profiles or “molecular signatures.”[15] These signatures could eventually become important tools to quantify EGFR pathway activation and better predict response to cetuximab therapy.
Personalized medicine research looks beyond the molecular changes taking place within tumor cells, to individual inherited genetic variability. Over the last decade, single nucleotide polymorphisms (SNPs) within VEGF and VEGF receptor 2 (VEGFR-2) have emerged as potential markers for predicting cancer prognosis and treatment outcomes.[16,17] SNPs found in VEGF and VEGFR-2 are appealing to study because of their germline origins and their effects on rare-mutating endothelial cells, separate from tumor cells prone to common mutations.[18] This allows for the convenient study of SNPs at any point, regardless of tumor status. A recent study in advanced breast cancer suggested that polymorphisms in VEGF could serve as predictors for the outcome of patients treated with bevacizumab and paclitaxel(Drug information on paclitaxel) versus paclitaxel alone. Results from SNP analysis of patients enrolled in the ECOG 2100 trial showed that VEGF-2578 AA had a superior OS (HR = 0.58;
P = .023) when treated with bevacizumab versus control. VEGF 1154-AA also had a similar OS improvement (HR = 0.62; P = .001), however neither SNP predicted a superior PFS or RR for either arm.[16] Another study analyzed serum samples from E4599 patients for selected SNPs involved in the angiogenesis pathway. Patients were treated with bevacizumab plus carboplatin and paclitaxel (BPC) or chemotherapy alone. Results showed increased OS in patients with VEGF G-634C and intercellular adhesion molecule (ICAM)-1 T469C polymorphisms when treated with BPC. SNPs in ICAM1, EGF, and CXCR2 were also associated with greater PFS in BPC-treated patients.[19]
The specific uses of cetuximab and bevacizumab continue to be an exciting venue of research in cancer therapy. However, despite the discovery of new, promising molecular targets and techniques, there are still many hurdles to clear before truly personalizing treatment for lung cancer. Biomarkers need to be validated for clinical implementation using prospective randomized clinical trials. These trials need to be appropriately designed, perhaps using adaptive randomization, where accrued data from patient outcomes direct subsequent patient treatment assignments, and suitable endpoints are needed to display the beneficial mechanisms of treatment.[20,21] New trial designs such as the BATTLE (Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination) trials developed at MD Anderson have already been successfully implemented, using adaptive randomization while focusing on comprehensive molecular profiling of patient tumor and blood samples.[22] With the new direction in research and clinical trials, treating patients who most benefit from treatment with minimal toxicity will continue to be the mainstay of personalized cancer therapy.
Financial Disclosure: The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
