A 45-year-old Caucasian man presented to the hospital with a 3-month history of fatigue, bilateral upper and lower limb paresthesias and gradually worsening ascending paralysis. A few weeks later, he developed acute renal failure requiring haemodialysis. Investigations revealed presence of myeloperoxidase (MPO) perinuclear antineutrophil cytoplasmic antibodies (ANCA). Renal biopsy was conclusive for rapidly progressive glomerulonephritis with crescents. Treatment for ANCA positive vasculitis was initiated with pulsed steroids, cyclophosphamide and plasmapheresis. The hospital course took an unexpected turn when the patient developed acute chest pain with an EKG consistent with inferior ST elevation myocardial infarction (STEMI). Urgent left heart catheterisation revealed distal occlusions in multivessel coronary distribution. Coronary involvement is rare in ANCA vasculitis and STEMI has not been reported in MPO-ANCA positive vasculitis, to the best of our knowledge.

e15058 Background: Recent retrospective and subgroup analyses have evaluated the effects of capecitabine with acid-reducing agents both in the adjuvant setting for colon cancer and in the palliative setting for stage IV gastroesophageal malignancies. Efficacy outcomes in these studies are contradictory. Acid-reducing agents are thought to reduce the absorption of capecitabine by increasing gastric pH leading to a concern of decreased efficacy. Studies thus far have not included stage IV colorectal cancers, a common population who receives capecitabine-based regimens. Methods: We conducted a retrospective chart review of patients diagnosed with stage IV Gastro Intestinal (GI) cancers treated with capecitabine-based therapies between 2012 and 2018. Electronic health records were used for data collection. Descriptive statistics and COX regression model were utilized for data analysis. Results: Of the 249 charts that were screened for inclusion, 58 were included in this review. Thirty-eight (65.5%) had a diagnosis of colon cancer, 11 (19%) had rectal cancer, and 9 (15.5%) had other GI malignancies. The most common capecitabine-based regimens included XELOX alone (29%) or with bevacizumab (36%). During capecitabine treatment, 20 (34.5%) patients received concomitant acid-reducing therapy. This included 19 patients taking a PPI and 1 patient taking a H2 Inhibitor. Patients who did not receive acid-reducing therapy had a 1-year PFS of 23.7% compared to 25% in patients who did not, (HR 0.55, p = 0.46). The 1-year OS was 68.4% in the non-acid-reducing agent group and 60% in the patients who did receive acid-reducing agents during capecitabine treatment, respectively (HR 0.17, p = 0.68). Conclusions: Receiving acid-reducing agents concomitantly with capecitabine-based regimens appears to not have an effect on 1-year PFS or OS in patients with Stage IV GI cancers. These outcomes may have been affected by the sample size, incomplete home medication lists, and compliance issues with capecitabine. Larger population studies are needed to validate these findings and identify the significance of utilization of acid-reducing agents in patients with Stage IV GI cancers receiving capecitabine.

187 Background: Difficult conversations require an immense amount of patience. Yet on average doctors allow patients 11 seconds to speak before interrupting, which makes determining the patient’s agenda and goals of care incredibly difficult. Most patients expect their doctor to initiate prognosis and end of life care conversations. An inexpensive, effective way to increase these conversations is with a Question Prompt List (QPL). Our study examines QPL use with patients who are at least three months into treatment for metastatic cancer and includes a pre and post-visit interview to provide a within-group analysis of the QPL’s effect on their understanding of disease state, prognosis, and goals of treatment. Our patient population includes patients of lower socioeconomic background and a higher percentage of African Americans. Methods: 35 out of 155 eligible patients identified by chart review were enrolled in the study. They were interviewed about their diagnosis and prognosis before their oncologists visit. At the end of the interview, patients were given a QPL and encouraged to read it. Once the oncologists completed the visit, the researcher conducted a final interview about their disease state. Results: Before seeing their oncologist, 26 patients (74%) completely read the QPL. Most patients would recommend more physicians use this tool (77%). Of note, patients thought of new questions or concerns (51%), felt more comfortable asking questions (46%), and reported asking more questions compared to prior visits (34%). A staggering majority (68%) of patients reported never having a prognosis discussion prior to today’s visit. Previously, only 15 patients (43%) understood their treatment was palliative compared to 19 patients (54%) after this visit. Conclusions: The QPL was viewed as beneficial by most patients and increased question asking, prognosis conversations, and accuracy of goals of treatment knowledge. The QPL tool is inexpensive and requires minimal work by office staff to implement. We plan to expand our study and if results remain unchanged, we will look at studies to investigate a shortened QPL given at an initial consultation and used at subsequent visits.

31 Background: CPIs have been established as essential components of cancer immunotherapy across multiple cancer types with proven clinical benefit, improved outcomes, and less toxicity. Studies in lung and head and neck cancers found that low ALC, a marker of immune exhaustion, was associated with poor response to CPIs and worse progression-free survival. We explored the effect of pALC and pNLR on CPI response rates in patients with RCC and BC. Methods: We retrospectively reviewed every RCC and BC patient that received CPIs at Overton Brooks VA Medical Center and LSUHSC-S between 2015 and 2019. Patients’ pALC and pNLR were calculated. The patients were divided according to pALC into 2 groups: Group A with pALC > 1000 and Group B with pALC < 1000. Similarly, using NLR’s established upper normal limit of 3, 2 groups were created: Group 1 with pNLR < 3 and Group 2 with pNLR > 3. Our primary outcome of interest was defined as the presence or absence of CPI response. Patients who attained stable disease, partial response, and complete response were categorized as responders. Those who progressed on CPIs were labeled as non-responders. The significance of the association between pALC and pNLR groups and the occurrence of any response was analyzed statistically. Results: Twenty patients (13 RCC, 7 BC) were treated with CPIs and had documented responses. Twelve patients had pALC > 1000 (Group A) whereas 8 patients had pALC < 1000 (Group B). Both groups were comparable with respect to age, sex, race, and types of CPIs. Group A had a significantly higher response rate (75% vs 25%, p = 0.027). As to pNLR, 10 patients had pNLR < 3 (Group 1) and another 10 patients had pNLR > 3 (Group 2). Patients with pNLR > 3 had worse response rates to CPIs compared to those with pNLR < 3 (30% vs 80%, p = 0.024). Conclusions: This is the first report from a real-world clinical setting to show a detrimental association between pALC < 1000 and pNLR > 3 and CPI response rates in a retrospective cohort of consecutive non-selected kidney and bladder cancer patients. This association and its clinical utility require further confirmation in a prospective larger cohort.