The epidermal growth factor receptor (EGFR) is a central regulator of proliferation and progression in human cancers. Five EGFR inhibitors, two monoclonal antibodies and three TKIs, have recently gained FDA approval in oncology (cetuximab, panitumumab, erlotinib, gefitinib and lapatinib). These strategies of EGFR inhibition demonstrate major tumor regressions in approximately 10-20% of advanced cancer patients. However, many tumors eventually manifest acquired resistance to treatment. In this study we established and characterized a model to study molecular mechanisms of acquired resistance to the EGFR monoclonal antibody cetuximab. Using high-throughput screening we examined the activity of 42 receptor tyrosine kinases in resistant tumor cells following chronic exposure to cetuximab. Cells developing acquired resistance to cetuximab exhibited increased steady-state EGFR expression secondary to alterations in trafficking and degradation. In addition, cetuximabresistant cells manifested strong activation of HER2, HER3 and cMET. EGFR upregulation promoted increased dimerization with HER2 and HER3 leading to their transactivation. Blockade of EGFR and HER2 led to loss of HER3 and PI(3)K/Akt activity. These data suggest that acquired resistance to cetuximab is accompanied by dysregulation of EGFR internalization/degradation and subsequent EGFR-dependent activation of HER3. Taken together these findings suggest a rationale for the clinical evaluation of combinatorial anti-HER targeting approaches in tumors manifesting acquired resistance to cetuximab.

The epidermal growth factor receptor (EGFR) is a receptor tyrosine kinase that plays a major role in oncogenesis. Cetuximab is an EGFR-blocking antibody that is FDA approved for use in patients with metastatic colorectal cancer (mCRC) and head and neck squamous cell carcinoma (HNSCC). Although cetuximab has shown strong clinical benefit for a subset of cancer patients, most become refractory to cetuximab therapy. We reported that cetuximab-resistant NSCLC line NCI-H226 cells have increased steady-state expression and activity of EGFR secondary to altered trafficking/degradation and this increase in EGFR expression and activity lead to hyper-activation of HER3 and down stream signals to survival. We now present data that Src family kinases (SFKs) are highly activated in cetuximab-resistant cells and enhance EGFR activation of HER3 and PI(3)K/Akt. Studies using the Src kinase inhibitor dasatinib decreased HER3 and PI(3)K/Akt activity. In addition, cetuximab-resistant cells were resensitized to cetuximab when treated with dasatinib. These results indicate that SFKs and EGFR cooperate in acquired resistance to cetuximab and suggest a rationale for clinical strategies that investigate combinatorial therapy directed at both the EGFR and SFKs in patients with acquired resistance to cetuximab.

163 Background: We initiated a nurse practitioner (NP)-based model of palliative care delivery embedded within an oncology clinic guided by a shared mental model (SMM) between palliative care, oncology, and psychosocial oncology clinicians in order to foster best practice collaboration and closed-loop communication between teams. These data represent processes and outcomes three years after initial implementation of the program. Methods: We evaluated program growth as well as advance care planning, hospice use, and utilization in patients with advanced cancer seen by the palliative care NP compared to patients receiving usual care from March 2014 to March 2017 at University-based oncology clinics. We developed a palliative care quality improvement tool integrating administrative and clinical data from multiple sources, including the electronic health record (EHR) and external hospices, using progressive methods of pulling data, such as natural language processing, in order to identify patients with advanced cancer and key process and end of life utilization measures. We used chi square tests to compare care received by the two groups. Results: There was good adoption of the intervention. The number of participating oncologists increased from 2 to 5 and the palliative care NP shifted from part-time to full-time after the first 1.5 years of implementation. Patients enrolled in the NP-based model of palliative care delivery were more likely to have a documented goals of care conversation (74.6% v. 9%, p < 0.01), to be referred for additional psychosocial support (52.5% v. 30.9%, p < 0.01), and to complete physician orders for life sustaining treatment (POLST) (20.3% v. 4.5%, p < 0.01). There was no statistically significant difference in advance directive completion (28.8% v. 23.5%). Among decedents, patients enrolled in the NP-based model were more likely to be enrolled in hospice (50.5% v. 29.1%, p < 0.01). There were non-statistically significant trends toward less hospital (4.6 days v. 5.6 days) and ICU use (1.0 day vs. 1.3 days) in the last 30 days of life. Conclusions: An NP-model of palliative care delivery within an oncology clinic led to important improvements in key palliative care processes and outcomes.

105 Background: We aimed to engage oncologists to disseminate a successful pilot-tested shared mental model (SMM) for the integration of early advance care planning (ACP) and identification of palliative care (PC) needs across a health system’s oncologic practice. Methods: Our Oncology Communication Collaborative Team (OCCT) had oncology leadership support and included a multidisciplinary team representing leaders in oncology, ACP, PC, psycho-social oncology and quality. To communicate the SMM developed by our team, the OCCT developed an interactive Saturday session (1-hr didactic, 3-hr small group role-play) that focused on early ACP and the cognitive and emotional aspects of communication. Before and after the training, we asked participants to rate their ability to communicate with patients as well as their readiness, self-efficacy, and need for help to improve communication regarding prognosis, ACP, end of life care and symptom management using a previously validated survey. We computed means and compared matched pairs of pre and post surveys using a paired t-test. We also surveyed participants about whether they would recommend the course to others and planned changes to practice. Results: All but one oncologist (52/53), 3/4 invited fellows, and 12/14 oncology nurse practitioners participated and 90% of attendees completed pre and post surveys. Participants rated their communication ability higher (6.7 v. 7.6, p < 0.01) on a 10-point scale after the training. Readiness to improve communication in this domain (9.1 v. 9.2, p = 0.35) was similar before and after the training. Self-efficacy (1.5 v. 1.5, p = 0.70) and needing help to improve (1.6 v. 1.7, p = 0.37) were rated highly (1 = A lot and 4 = Not at all) but did not change with training. All but one participant reported they would recommend the course to others and free text responses about changes they planned to make to their practice based on the training included: having earlier ACP discussions, focusing on patient goals/priorities and asking open-ended questions. Conclusions: Conducting a training to disseminate a SMM of oncology and PC is feasible, valuable, and can be the first step for partnered continuous quality improvement.

No abstract