A Prospective Observational Study of Comprehensive Genomic Profiling in Previously Untreated Metastatic Non-small Cell Lung Cancer

The understanding of NSCLC is undergoing a rapid evolution from one disease treated with empirically chosen, costly cytotoxic chemotherapies, laden with adverse effects and benefits lasting only weeks or a few months, to a rapidly growing set of genomically defined diseases that can be matched to targeted therapies with markedly improved outcomes, including response rates of 70-80% and survivals often measured in years rather than months. The need for comprehensive genomic profiling tests able to accommodate the expanding list of genomic markers without necessitating new biopsies or rising cost is not just a convenience, but a medical necessity.

In October 2014, the National Comprehensive Cancer Network (NCCN) specifically recommended that patients with metastatic NSCLC undergo "broad molecular profiling" with the goal of identifying rare driver mutations for which effective drugs may already be available, or to appropriately counsel patients regarding the availability of clinical trials, including not only EGFR and ALK, but also BRAF, ERBB2 (HER2), MET, RET and ROS1. For example, while ERBB2 alterations are certainly more common in breast cancer, these same alterations have also been found in NSCLC and thus would predict response to targeted therapy with anti-HER2 agents. Specifically, each of these additional targetable alterations indicate a greatly increased likelihood of response to a respective targeted treatment, even though some of the treatments would be considered 'off-label' for lung cancer (e.g. trastuzumab or afatinib targeting patients with ERBB2 alterations; crizotinib for MET amplification and ROS1 rearrangements), or on-label drugs approved for lung cancer (e.g. crizotinib targeting patients with ALK rearrangements).

Problems in Genomic Testing: The substantial increase in the number of recommended targetable alterations in NSCLC presents multiple clinical and logistic challenges to the current model of gene-by-gene testing. These include non-validated testing, missed alterations that could have been matched to targeted therapy, insufficient tissue due to sequential or parallel testing, unacceptably long turnaround times, and physician frustration and confusion related to delays in obtaining all needed results and complex interpretation of multiple reports. Based on recent publications, it is reasonable to anticipate the list of targetable alterations that should be evaluated as 'standard of care' will only continue to expand. This reality makes the current model of individual tests in a gene-by-gene model unsustainable. For example traditional testing would require multiple modalities, including polymerase chain reaction (PCR)-based tests, capillary sequencing, fluorescence in situ hybridization (FISH)-based testing, and mass spectrometry-based sizing assays. Insufficient tissue to supply all of these tests is a particular problem. Traditional pathologic evaluation, including testing for the seven targetable alterations in the NCCN guidelines, may require more than 26 slides -- a tissue requirement that is usually beyond the amount obtained from minimally invasive biopsies. Competing demands for scarce tissue may place clinicians in an uncomfortable position of having to guess which tests to run when slides are running out, or alternatively, to consider conducting additional biopsy procedures. These challenges will only grow in complexity as biopsies become less invasive, the number of molecular markers and targeted therapeutic options increase, and evidence for newer markers becomes stronger.

FoundationOne is a comprehensive genomic profile that applies next generation sequencing in a unique manner to identify all 4 types of genomic alterations across all genes known to be unambiguous drivers of solid tumors with high accuracy. The test simultaneously sequences the coding region of 315 cancer-related genes plus introns from 28 genes often rearranged or altered in cancer to a typical median depth of coverage of greater than 500X. Each covered read represents a unique DNA fragment to enable the highly sensitive and specific detection of genomic alterations that occur at low frequencies due to tumor heterogeneity, low tumor purity and small tissue samples. FoundationOne detects all classes of genomic alterations, including base substitutions, insertions and deletions (indels), copy number alterations (CNAs) and rearrangements using a small, routine FFPE sample (including core or fine needle biopsies).

Foundation Medicine recently collaborated with Memorial Sloan-Kettering Cancer Center (MSKCC) to evaluate the clinical utility of this approach. After appropriate institutional approvals, a group of patients with advanced NSCLC who had completed in-house clinical testing in the CLIA-accredited MSKCC lab and who tested negative for all using conventional diagnostics, and who were alive and receiving systemic cytotoxic chemotherapy were identified. Before they were tested using FoundationOne, 24 of the first 34 such patients had required additional biopsies to complete the initial requisite testing. In 9 of the 34 clinical cases, FoundationOne could not be performed due to lack of available tissue after the initial rounds of MSKCC testing. FoundationOne was successfully performed in 25 patients uncovering one or more genomic alterations linked to a targeted agent based on NCCN guidelines in 36% (9/25 cases) and a targeted agent available on an open clinical trial at MSKCC in 32% (8/25) of patients. This approach simultaneously analyzes the multiple genes and classes of genomic alteration required for clinical care and are readily adapted to efficiently incorporate a rapidly expanding landscape of targetable alterations.