NATURE REVIEWS | CLINICAL ONCOLOGY ADVANCE ONLINE PUBLICATION | 1 NEWS & VIEWS BIOMARKERS Exceptional responders— discovering predictive biomarkers Naoko Takebe, Lisa McShane and Barbara Conley Modern genomics technologies enable the identification of genetic alterations, even those present at a low frequency, and can contribute to unveiling the mechanistic rationale behind the unexpected clinical response of ‘exceptional responders’. This approach will drive the identification of molecular biomarkers that can be integrated into clinical trials and predict response to a specific therapy. Takebe, N. et al. Nat. Rev. Clin. Oncol. advance online publication 17 February 2015; doi:10.1038/nrclinonc.2015.19 Most oncologists during their professional careers have had the experience of a patient who had a surprising response to a treatment, in instances when the majority of patients did not respond. Such ‘exceptional responders’, in the context of clinical trials, raise intriguing questions with regard to understanding why these patients showed long-lasting responses to either single-agent or combination thera- pies that are otherwise deemed ‘failures’ based on the overall study outcomes. 1 Until recently, the analysis of the molecular characteristics of the exceptional responders was difficult owing to limitations of the technologies and to the unavailability of sufficient biopsy samples from these patients, especially considering that the majority of available patient archival material is represented by tissues embedded in formalin-fixed, paraffin embedded blocks, in which nucleic acid are highly degraded. Recent advances in molecu- lar and computational analysis techniques have enabled the identification of genetic alterations, even those existing at very low frequencies in tumour tissues—providing a valuable tool to identify the alterations that might specify the molecular aetiology behind these exceptional responses. The advent of next-generation sequencing coupled with large-scale sequencing efforts to identify somatic gene aberrations in tumour tissues has resulted in publicly available data- bases, such as The Cancer Genome Atlas (TCGA) and International Cancer Genome Consortium (ICGC), which have helped to combine the biological knowledge of specific pathways inhibited in exceptional respond- ers with the identification of mutations that could then be used as predictive molecu- lar biomarkers of clinical response. The frequency of the mutations identified by screening patient tumours and that can serve as molecular biomarkers might be ≤5%. 2 Nonetheless, ongoing global efforts are aimed at the identification of such low frequency actionable mutations, which can then be incorporated as biomarkers into master protocol clinical trials. 2 For this strat- egy to be successful, a nationwide effort to search for exceptional responders, delineate the mechanisms of action of the particular agents, and make the results available in a public database, is an important research project to undertake. In this context, the National Cancer Insti- tute (NCI) has launched several precision medicine initiatives so that recently acquired knowledge of molecular variants active in cancers could enable, in a more rational and efficient way, further therapeutic develop- ment. 3 Specifically, the ‘NCI Exceptional Responders Pilot Study’ was launched in October 2014 (Figure 1). 4,5 This initiative is based on a NCI Cancer Therapy Evaluation Program review of 10 years of unpublished data from phase II clinical trials, which identified the occurrence of exceptional responses in up to 10% of patients in trials in which the agent did not receive FDA approval in that setting. Under this proto- col, clinicians (within or outside the NCI), who are interested in proposing a case for the Exceptional Responders Pilot Study and have access to the patient’s tumour tissue (and normal tissue if available) prior to the exceptional response will be able to make an initial submission via e-mail. Of note, exceptional responders will not be limited to cases treated only by targeted therapies or single-agent therapy, or in a clinical trial. The submission should include a short descrip- tion of the case (without patient identifiers) including diagnosis (and time of diagnosis); treatment; type and duration of exceptional response; informed-consent status; partici- pation in a clinical trial or not; availability of tumour tissue; and the nature of the sample. The protocol initially defines exceptional Propose exceptional responder nciexceptionalresponders@mail.nih.gov (tissue available) Case review @NCI IRB approval Execute MTA Submit clinical data (CTSU) Submit tissue and path report Execute contract, complete data submission, submit consent form QA/QC isolation of nucleic acids DNA sequencing, whole exome and deep targeted panel if possible: RNAseq, WGS etc. Data coordinating centre Review of data and correlation with drug’s mechanism of action Controlled access database Figure 1 | Algorithm of the NCI Exceptional Responders Initiative. This scheme describes the different steps from evaluation to sample analysis of the Exceptional Responders Initiative. For more detailed information or to submit a case, please see the Clinicaltrials.gov and NCI websites, 4,5 or contact NCI Exceptional Responders Study coordinator Erin Souhan through the Exceptional Responder Mailbox (nciexceptionalresponders@mail.nih.gov). Abbreviations: CTSU, Clinical Trials Support Unit; IRB, Institutional Review Board; MTA, material transfer agreement; NCI, National Cancer Institute; QA, quality assurance; QC, quality control; WGS, whole genome sequencing. © 2015 Macmillan Publishers Limited. All rights reserved