Trying to Understand Differing Results of FIRE-3 and 80405: Does the First Treatment Matter More Than Others?

It was not that long ago that sequencing the lines of therapy for patients with advanced colorectal cancer (CRC) meant using bolus fluorouracil followed by infusional fluorouracil or vice versa. Although there were many factors responsible for the dismal 10-month life expectancy of these patients in the 1980s, the absence of different and effective treatment options was an obvious impediment. The development of irinotecan in the 1990s heralded a new attitude among researchers and pharmaceutical companies alike and ushered in two decades of active exploration of new agents in colorectal cancer. Just this year, two more agents—ramicurumab and TAS102—will be added, which will bring the compendium of drugs available in the United States for the treatment of CRC to 11 from the existing nine (ie, fluorouracil, irinotecan, oxaliplatin, capecitabine, cetuximab, bevacizumab, panitumumab, regorafenib, and zivaflibercept). Each agent was studied in and gained initial approval on the basis of specific study criteria. However, in the management of metastatic cancer, oncologists typically make decisions about the best way to sequence therapies, and so it has been for metastatic CRC. In some instances, the decision is based on preclinical or actual clinical data (eg, bevacizumab beyond progression); in other instances, on toxicity avoidance (eg, strategies such as OPTIMOX (Optimized Leucovorin-Fluorouricil-Oxaliplatin) or the use of treatment holidays); in still others, on market or other external factors (eg, pathways or guidelines). However these agents are integrated, their use has almost certainly contributed to the major improvements in median survival that we see in studies today. Because patients with metastatic CRC usually enjoy a preserved performance status for much of their course, most will receive many, if not all, of the available classes of agents. However, there are little data to govern the sequence that yields the best outcomes. Given the number of permutations one could consider, a rigorous study of the optimal order of these systemic agents is implausible even without including the ablative or surgical interventions that are also frequently employed. The classic study by Tournigand et al, which assessed the activity of FOLFOX (infusional fluorouracil, leucovorin, and oxaliplatin) and FOLFIRI (fluorouracil, leucovorin, and irinotecan) in either order, was possible when just two treatment options existed and showed equivalence in that era. Today, though, asking the sequencing question would be next to impossible and would require a study design that is hopelessly complex. Biomarkers have brought some organization to the treatment paradigm: The 50% to 60% of patients whose tumors harbor any mutation in RAS are not suited for the anti–epidermal growth factor receptor (EGFR) antibodies, whereas the 5% to 8% of patients whose RAS wild-type tumor has a V600E mutation in BRAF will generally experience a poor response and are often considered for clinical trials that target the RAF mutation. However, even with some biomarker refinement, the current National Comprehensive Cancer Network guidelines take nine pages to enumerate the treatment options for patients with CRC, and the recent European Society for Medical Oncology guidelines for CRC are equally complex. In fact, the term continuum of care has replaced the classic lines of therapy terminology; this change acknowledges the great variation in how oncologists navigate through the various options available to treat patients with CRC. In this issue of the Journal of Clinical Oncology, Modest et al explore the possible impact of subsequent therapies on patients with advanced CRC treated on the FIRE-3 trial. FIRE-3, conducted in Germany and Austria, sought to establish which of two monoclonal antibodies—the EGFR-targeting antibody cetuximab or the vascular endothelial growth factor (VEGF) –targeting antibody bevacizumab—was the best to use in combination with FOLFIRI in the first-line management of metastatic CRC. The decision to use FOLFIRI as the backbone is an example of regional variation in chemotherapy preferences. Initiated a decade ago, thestudybeganwithoutbiomarkerpatientenrichmentbutwasamended to reflect the understanding of the importance of RAS mutations for the identification of patients who could benefit from cetuximab. FIRE-3 found that patients with KRAS wild-type (at codons 12 and 13) tumors who began treatment with FOLFIRI and cetuximab had an overall survival (OS) of 28.7 months compared with 25.0 months in the group treated with FOLFIRI plus bevacizumab. (Subgroup analysis showed a larger difference between cetuximab and bevacizumab treatment arms of 33.1 months versus 25.6 months, respectively, when all KRAS and NRAS mutations were excluded.) This result should speak for itself, because OS is an uncontroversial end point. However, FIRE-3 failed to meet its primary end point of overall response, as judged by the investigator, as well as its secondary end point of progression-free survival (PFS). For FIRE-3 to then demonstrate an OS advantage stands in stark contrast with the analysis JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 33 NUMBER 32 NOVEMBER 1