Maurie Markman, MD
Editor-in-Chief of OncologyLive
Senior vice president for Clinical Affairs and National Director for Medical Oncology
Cancer Treatment Centers of America, Eastern Regional Medical Center
It is increasingly recognized that the realm of molecular diagnostic testing in oncology is undergoing as much change, if not more, as the world of cancer therapeutics. With the widely anticipated introduction during the next several years of whole genome sequencing strategies acceptably priced under $5000 for patients with and without cancer, it is highly likely that the pace of revolutionary changes in this area will only quicken.
Further, as exciting associations between new molecular targets and novel antineoplastic agents are reported, there is an understandable need to develop reliable diagnostic tests to determine the optimal patient populations to receive the specific therapy.
However, for the present, the reporting of individual molecular tests remains the standard of care for documenting the appropriateness of delivery of novel targeted therapy to an individual patient with cancer. The existing scenario raises the issue of the costeffectiveness associated with the decision to obtain a single or possibly multiple molecular tests when an oncologist attempts to design an optimal management program for a particular patient.
Consider, for example, the question of whether all patients with advanced-stage non-small cell lung cancer should have molecular testing performed at diagnosis to determine the presence of a sensitizing epidermal growth factor receptor (EGFR
) gene mutation, an ALK rearrangement, and a BRAF
mutation. Similarly, while it is currently routine to examine a tumor for the presence of estrogen and progesterone receptors and for HER2 overexpression in an initial tumor specimen in women with breast cancer, would it be appropriate to add 1 or more additional molecular tests (eg, BRCA1, BRCA2
) for all patients if therapy might conceivably be impacted by knowing this information (eg, treatment with a PARP inhibitor)? Would this be a cost-effective approach to disease management?
Although it is now possible to order a series of tests for multiple molecular targets for a given patient, is that a cost-effective strategy? Such questions loom over current options.
In a provocative report, investigators in Switzerland have attempted to examine this question in individuals with metastatic colorectal cancer.1
The specific issue addressed in this analysis was whether it was costeffective to obtain an examination of KRAS
mutation status which, if found to be “wild type,” would be followed immediately with an attempt to evaluate the specific cancer’s BRAF
mutational status, prior to administering cetuximab (a monoclonal antibody directed at the EGFR receptor). The alternative scenarios examined in this exercise were KRAS
testing alone (without performing BRAF
mutational analysis) and the delivery of cetuximab in the absence of any molecular testing.
Utilizing data from previously published studies and assuming that the presence of a mutation (in either KRAS
) would prevent treatment with this agent, the researchers found that the approach of KRAS
testing followed by BRAF
analysis in individuals with wildtype KRAS
to be the most cost-effective strategy when compared with simply not administering cetuximab to any patient, and therefore denying any group of individuals the opportunity to experience clinical benefit from the drug.
All such economic modeling studies in oncology can be challenged, considering the many assumptions required to complete the analysis, and the uncertain relationship between trial data and the “real world” patients undergoing therapy in routine oncologic practice (eg, age, performance status, presence of relevant comorbidities).