Are companion diagnostics useful?

We define a companion diagnostics (CDx) test as any diagnostic tool that guides the selection of patient treatment. In the US, these tests include diagnostics cleared by the Food and Drug Administration (FDA), as well as laboratory-developed tests (LDTs) run in CLIA-accredited laboratories. The answer to the title’s question is obviously in the affirmative. We discuss the growing need to improve, accelerate, and standardize oncology CDx that benefit patients. Recent advances in our understanding of the mutational landscape of cancers have led to the rapid development of targeted therapies for pathogenomic targets. At the same time, advances in next-generation sequencing (NGS) have revealed the complexity of heterogeneous cancers. With the fast pace of change occurring in clinical oncology, flexible approaches to CDx development are needed while test accuracy is maintained for delivering precision medicine to patients. The remarkable efficacies of new targeted therapies have recently changed the paradigm for oncology to one of matching the right patient with the right therapy (1 ). This approach requires robust laboratory testing of patient samples. FDA-cleared tests exist for a limited number of markers (Table 1). In contrast, LDTs developed in CLIA-certified laboratories are used for 2000 genetic tests (2 ). LDTs are currently used in oncology for patient care and in clinical trials, which are now evaluating 500 agents targeting 100 genomic alterations. The recent success in the development of 2 targeted therapies, crizotinib and vemurafenib, exemplifies the CDx process in which the FDA has co-cleared a diagnostic test (3, 4 ). The development of the CDx included analytical validation and elements of quality systems (good manufacturing practice, design control, personnel, software, instrumentation, and other parameters) as critical components of the regulatory package. The tests were used to recruit patients to pivotal phase II trials for the presence of the mutation/ gene fusion. The clear clinical response observed in patients treated with the novel therapies provided clinical validation of the drugs and the diagnostic tests. The drug clearance included the CDx package of analytical validation, adherence to quality systems regulation, and the clinical-validation data package. On the other hand, the FDA’s refusal to clear omacetaxine for targeted T315I mutation–positive chronic myelogenous leukemia was based on a failure to show interlaboratory concordance, which can lead to misclassification of patients and potential risk. In contrast to these examples, CDx development for anti– epidermal growth factor receptor (EGFR) antibody therapies in colorectal cancer highlights the evolving landscape of precision medicine. Tumor production of EGFR, as measured by immunohistochemistry with an FDA-cleared CDx test, was used to select patients for pivotal phase III trials of cetuximab and panitumumab; however, the detection of EGFR protein expression was later discovered to be variable and not to demonstrate clinical utility (5 ). This result quickly led to the use of panitumumab and cetuximab without testing for EGFR protein expression. Meanwhile, cetuximab and panitumumab were discovered to be effective only when the wild-type KRAS (v-Kiras2 Kirsten rat sarcoma viral oncogene homolog) gene is in the tumor (6, 7 ). After these data emerged, European and American Society for Clinical Oncology guidelines recommended testing for KRAS mutation with LDT-based tests to exclude patients harboring KRAS mutant tumors before initiating therapy. Despite the almost universal clinical practice of KRAS mutation testing for colorectal cancer over the past 4 years, an FDA-cleared CDx for KRAS became available only this year for use in EGFR antibody therapies. Notably, the need to test for EGFR protein expression before anti-EGFR antibody therapy remains on the drug label. Complexity like that observed for colorectal and lung cancer is likely to be seen again with the increasing