Special Report: Multiple Molecular Testing of Cancers to Identify Targeted Therapies
BackgroundMeasurement of genetic or other molecular markers in cancer tissue is established in the diagnosis, staging, and treatment of cancer. Currently, there is interest in the utility of measuring a large number of molecular markers at a single time in order to identify a treatment which targets the biological pathway involving that molecular marker. The available methods, or assays, may include molecular markers that individually might be indicated for a specific cancer, but are not indicated for most cancers. This may result in a different treatment than usually selected for a patient based on the type of cancer and its stage.
ObjectiveThe objective of this Special Report is to review important concepts underlying the use of multiple molecular testing to guide cancer treatment. Issues and problems underlying the assessment of efficacy of such strategies will be discussed. Published studies that use multiple molecular testing methods to select targeted therapies and that report patient outcomes will be presented.
Search StrategyA series of targeted electronic searches was performed on MEDLINE (via PubMed) through March 2013 to identify recent (i.e., published within the last 5 years) review articles or systematic reviews, initially using terms “molecular profiling” and “cancer” and “treatment.” In addition, we used the “Related Articles” function of PubMed to search for additional citations, including original works. We also hand-searched the bibliographies of relevant reviews to identify citations to original studies that reported clinical outcomes, for example, survival.
Selection CriteriaFor the evidence review of this Special Report, we selected published studies that used multiple molecular testing to select treatment and reported patient outcomes. We excluded case reports and studies that anecdotally reported the outcomes of a few patients.
Main ResultsUnderlying Concept. The use of multiple molecular testing to select targeted therapy is based on a shift in thinking about cancer behavior and treatment. Rather than thinking about cancer based on site and histology, molecular markers represent biological pathways that may be common across cancers. Choosing treatment based on these biological pathways is hypothesized to be a better method of selecting treatment.
Use of Multiple Molecular Testing. Use of multiple molecular markers to select treatment can generally be categorized in two ways. Performing a large number of tests might increase the probability of a positive test, which indicates possible susceptibility of the cancer to a targeted therapy usually not indicated for that particular cancer. Alternatively, the results of large numbers of tests might be integrated in some manner to construct an interlinked biologic pathway for that particular cancer, thereby providing insight into a potentially more effective targeted therapy for that particular patient.
Evaluation of Strategy. This strategy of selecting treatment for cancer patients raises important issues in the design of clinical trials to demonstrate effectiveness. Molecular marker-stratified studies will be difficult to perform because there will be small numbers of patients in each molecular marker subgroup. Studies using molecular marker strategy designs also have critical problems. Patients with different cancers, different molecular markers, and different treatments may be combined in clinical trial arms, raising issues regarding the uniformity of treatment effects.
Review of Evidence. Three published studies report health outcomes for patients whose treatments were selected using multiple molecular marker panels. Two of the studies compare the time to progression on the targeted treatment to the time to progression on the most recently failed treatment. This is not an established measure of efficacy or treatment response. One study compares patients who had targeted treatment to another group of patients who did not have targeted treatment. This study was not randomized and thus may be subject to confounding. In two of the studies, subjects were given targeted treatments in Phase I trials. Outcomes of these patients could be dependent on the experimental treatment rather than the selection strategy. Also discussed in the review of evidence is a planned clinical trial using a molecular marker strategy design to select treatment for patients. The planned trial highlights important issues in the conduct and analysis of trials attempting to evaluate multiple molecular testing.
Use of multiple molecular testing to assist in making treatment decisions for cancer patients is rapidly evolving. Strong evidence of clinical effectiveness of this approach is not available, and a number of issues remain to be solved, particularly patient selection. Different approaches may be taken to the interpretation of multiple molecular marker panels. Clinical trials to determine the effectiveness of this approach will be challenging to complete.