ATYPICAL RESPONDERS LANDSCAPE REVIEW ∙ OCTOBER, 2017 26 Several biological determinants may predict chemosensitivity and chemoresistance. Results from theTAILORxtrialrecentlyshowedthatthe21-geneprofileassay,OncotypeDX® (GenomicHealth,Inc.), accurately predicts which patients with hormone receptor–positive, HER2-negative, axillary node– negative invasive stage I or II breast cancer can be treated solely with hormone therapy and avoid adjuvant chemotherapy, which would otherwise be recommended based on clinicopathological features alone [74]. Another example is MammaPrint® (Agendia), which uses a score derived from the relative expression of 70 genes [75]. Although these evidence-based decision support tools are useful in early-stage breast cancer patients, similar tools are yet to be developed for MBC. Non-invasive technologies to detect circulating tumor DNA may be useful for detecting molecular aberrations [76, 77], some of which may confer resistance or sensitivity and suggest new therapeutic targets. Although promising, some of these tools remain preliminary. Key drawbacks are that tumors are heterogeneous, the complete tumor (and other tumors within the patient) is not assessed, the tumor’s microenvironment is excluded, and the biological and social/lifestyle determinants that may influence the response are not elucidated. Nevertheless, an atypical response may become more predictable with these assays, allowing optimal selection and avoidance of therapies. An effort launched in the 1950s to predict an individual patient’s response to treatment by placing their tumor cells in a test tube and exposing them to various chemotherapy agents was not as successful as anticipated because, among other reasons, the tumor’s microenvironment was excluded [78]. This is important because the microenvironment is different in vivo compared to in vitro. In vitro conditions lack important factors such as macrophages and lymphocytes, which are immune system cells [79]. Recent innovations aim to overcome these obstacles by conducting real-time testing of ex vivo cancer cells and including microenvironment attributes. Predicting a clinical response to therapy with these techniques remains controversial [78], and robust studies are warranted. The Weisenthal Cancer Group has established a chemosensitivity assay test called “Cytometric Profiling” in which live, whole cancer cells from an individual patient are exposed to a library of chemotherapeutic agents or combinations of agents to determine which drugs/drug combinations will work best for that tumor and which will not [80]. A study of 30 women with ovarian cancer that compared therapies based on chemosensitivity testing with therapies aimed at genetic aberrations revealed that chemosensitivity testing is a superior predictor of treatment An atypical response may become more predictable with chemosensitivity/ chemoresistance assays, allowing selection of optimal therapies and avoidance of therapies that are unlikely to work or that are needlessly toxic.