The cell adhesion glycoprotein E-cadherin (CDH1) is commonly inactivated in breast tumors. Precision medicine approaches that exploit this characteristic are not available. Using perturbation screens in breast tumor cells with CRISPR/Cas9-engineered CDH1 mutations, we identified synthetic lethality between E-cadherin deficiency and inhibition of the tyrosine kinase ROS1. Data from large-scale genetic screens in molecularly diverse breast tumor cell lines established that the E-cadherin/ROS1 synthetic lethality was not only robust in the face of considerable molecular heterogeneity but was also elicited with clinical ROS1 inhibitors, including foretinib and crizotinib. ROS1 inhibitors induced mitotic abnormalities and multinucleation in E-cadherin–defective cells, phenotypes associated with a defect in cytokinesis and aberrant p120 catenin phosphorylation and localization. In vivo, ROS1 inhibitors produced profound antitumor effects in multiple models of E-cadherin–defective breast cancer. These data therefore provide the preclinical rationale for assessing ROS1 inhibitors, such as the licensed drug crizotinib, in appropriately stratified patients. Significance: E-cadherin defects are common in breast cancer but are currently not targeted with a precision medicine approach. Our preclinical data indicate that licensed ROS1 inhibitors, including crizotinib, should be repurposed to target E-cadherin–defective breast cancers, thus providing the rationale for the assessment of these agents in molecularly stratified phase II clinical trials. Cancer Discov; 8(4); 498–515. ©2018 AACR. This article is highlighted in the In This Issue feature, p. 371

Purpose: The Cancer Esophagus Gefitinib trial demonstrated improved progression-free survival with the epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor gefitinib relative to placebo in patients with advanced esophageal cancer who had disease progression after chemotherapy. Rapid and durable responses were observed in a minority of patients. We hypothesized that genetic aberration of the EGFR pathway would identify patients benefitting from gefitinib. Methods: A prespecified, blinded molecular analysis of Cancer Esophagus Gefitinib trial tumors was conducted to compare efficacy of gefitinib with that of placebo according to EGFR copy number gain (CNG) and EGFR, KRAS, BRAF, and PIK3CA mutation status. EGFR CNG was determined by fluorescent in situ hybridization (FISH) using prespecified criteria and EGFR FISH-positive status was defined as high polysomy or amplification. Results: Biomarker data were available for 340 patients. In EGFR FISH-positive tumors (20.2%), overall survival was improved with gefitinib compared with placebo (hazard ratio [HR] for death, 0.59; 95% CI, 0.35 to 1.00; P = .05). In EGFR FISH-negative tumors, there was no difference in overall survival with gefitinib compared with placebo (HR for death, 0.90; 95% CI, 0.69 to 1.18; P = .46). Patients with EGFR amplification (7.2%) gained greatest benefit from gefitinib (HR for death, 0.21; 95% CI, 0.07 to 0.64; P = .006). There was no difference in overall survival for gefitinib versus placebo for patients with EGFR, KRAS, BRAF, and PIK3CA mutations, or for any mutation versus none. Conclusion: EGFR CNG assessed by FISH appears to identify a subgroup of patients with esophageal cancer who may benefit from gefitinib as a second-line treatment. Results of this study suggest that anti-EGFR therapies should be investigated in prospective clinical trials in different settings in EGFR FISH-positive and, in particular, EGFR-amplified esophageal cancer.

One approach to identifying cancer-specific vulnerabilities and therapeutic targets is to profile genetic dependencies in cancer cell lines. Here, we describe data from a series of siRNA screens that identify the kinase genetic dependencies in 117 cancer cell lines from ten cancer types. By integrating the siRNA screen data with molecular profiling data, including exome sequencing data, we show how vulnerabilities/genetic dependencies that are associated with mutations in specific cancer driver genes can be identified. By integrating additional data sets into this analysis, including protein-protein interaction data, we also demonstrate that the genetic dependencies associated with many cancer driver genes form dense connections on functional interaction networks. We demonstrate the utility of this resource by using it to predict the drug sensitivity of genetically or histologically defined subsets of tumor cell lines, including an increased sensitivity of osteosarcoma cell lines to FGFR inhibitors and SMAD4 mutant tumor cells to mitotic inhibitors.

Pre-operative chemoradiotherapy (CRT) for MRI-defined, locally advanced rectal cancer is primarily intended to reduce local recurrence rates by downstaging tumours, enabling an improved likelihood of curative resection. However, in a subset of patients complete tumour regression occurs implying that no viable tumour is present within the surgical specimen. This raises the possibility that surgery may have been avoided. It is also recognised that response to CRT is a key determinant of prognosis. Recent radiological advances enable this response to be assessed pre-operatively using the MRI tumour regression grade (mrTRG). Potentially, this allows modification of the baseline MRI-derived treatment strategy. Hence, in a ‘good’ mrTRG responder, with little or no evidence of tumour, surgery may be deferred. Conversely, a ‘poor response’ identifies an adverse prognostic group which may benefit from additional pre-operative therapy. TRIGGER is a multicentre, open, interventional, randomised control feasibility study with an embedded phase III design. Patients with MRI-defined, locally advanced rectal adenocarcinoma deemed to require CRT will be eligible for recruitment. During CRT, patients will be randomised (1:2) between conventional management, according to baseline MRI, versus mrTRG-directed management. The primary endpoint of the feasibility phase is to assess the rate of patient recruitment and randomisation. Secondary endpoints include the rate of unit recruitment, acute drug toxicity, reproducibility of mrTRG reporting, surgical morbidity, pathological circumferential resection margin involvement, pathology regression grade, residual tumour cell density and surgical/specimen quality rates. The phase III trial will focus on long-term safety, regrowth rates, oncological survival analysis, quality of life and health economics analysis. The TRIGGER trial aims to determine whether patients with locally advanced rectal cancer can be recruited and subsequently randomised into a control trial that offers MRI-directed patient management according to radiological response to CRT (mrTRG). The feasibility study will inform a phase III trial design investigating stratified treatment of good and poor responders according to 3-year disease-free survival, colostomy-free survival as well as an increase in cases managed without a major resection. ClinicalTrials.gov, ID: NCT02704520 . Registered on 5 February 2016.