Search

3548 Background: The factors associated with the rise of young-onset colorectal cancer (YOCRC) remain unclear. However, environmental exposures are believed to be associated with YOCRC in addition to hereditary factors. Therefore, we evaluated the association between the national level Environmental Quality Index (EQI) and YOCRC in the US. Methods: We queried the SEER database to identify adult colorectal cancer patients diagnosed between 2010-2016. YOCRC was defined as age at diagnosis < 50 years. EQI (2005-2010) is a measure of county-level cumulative environmental exposures that includes 5 domains: sociodemographic, built, air, land, and water. A higher value represents a lower environmental quality. We distributed the total EQI and each EQI domain into five quintiles. Multivariable logistic regression analysis was used to assess the relationship between YOCRC and quintiles (upper-most vs. lowest) of EQI after adjusting by race (White, Black, and Others), gender, and stage at diagnosis. The age-adjusted incidence rate was also calculated using the SEER*Stat, and correlation efficiency was estimated between EQI domains and incidence rate. Results: A total of 261,417 CRC patients were included; 11% were YOCRC. In the adjusted multivariable analysis, poor built EQI (OR 1.15 [1.11-1.20]) and water EQI (OR 1.08 [1.03-1.12]) were more likely to be associated with YOCRC. Poor built EQI was more strongly associated with Black YOCRC (OR 1.21 [1.09-1.35]) as compared to White YOCRC (OR 1.14 [1.09-1.19]). Poor sociodemographic EQI was more strongly associated with Others (OR 1.47 [1.25-1.72]) compared to Black YOCRC (OR 1.14 [1.03-1.25]). In addition, poor built EQI (OR 1.19 [1.12-1.27]) and water EQI (OR 1.12 [1.05-1.19]) were more strongly associated with the metastatic disease among YOCRC patients. However, the total poor EQI was not associated with YOCRC (OR 0.99 [0.95-1.03]). On incidence analysis, there was a positive correlation between the incidence rate of YOCRC and sociodemographic EQI (rho = 0.49, p < 0.001), air EQI (rho = 0.30, p < 0.001), and land EQI (rho = 0.18, p < 0.001). Conclusions: This study evaluated a population-based ecological approach and showed that YOCRC was associated with lower environmental quality, including built and water domains. EQI domains were also associated with different racial groups among YOCRC. Instead of studying single environmental exposures, the impact of the cumulative environmental index should be recognized and studied further to understand the mechanism of cancer development.

Background: We aimed to study the effect of socioeconomic differences and molecular characteristics on survival in patients with young-onset pancreatic neuroendocrine tumors (YOPNET) and typical-onset PNET (TOPNET). Methods: We identified the patients with YOPNET (&lt, 50 years) and TOPNET (&ge, 50 years) who underwent definitive surgery diagnosed between 2004 and 2016 using the National Cancer Database. We evaluated overall survival (OS) using the Kaplan&ndash, Meier and Cox regression methods before and after propensity score matching. A publicly available genomic dataset was used to compare mutation frequencies among the two groups. Results: A total of 6259 patients with PNET were included, of which 27% were YOPNET. Patients with YOPNET were more likely to be Black, Hispanic, female, and have private insurance versus patients with TOPNET (all p &lt, 0.001). Patients with YOPNET had a lower comorbidity score, but higher stage and tumor size (all p &lt, 0.001). YOPNET was associated with a greater improved OS than TOPNET before and after propensity score matching (p &lt, 0.001). On multivariable analysis, this survival difference persisted for YOPNET as an independent prognostic factor (unmatched p = 0.008, matched p = 0.01). For genomic analysis, patients with YOPNET had a lower rate of multiple endocrine neoplasia type-1 (MEN-1) mutation than patients with TOPNET (26% vs. 56%, p &lt, 0.001). Conclusions: YOPNET represents a disease with distinct clinical features. Patients with YOPNET who underwent definitive surgery had better OS than patients with TOPNET despite having higher stage and tumor size. YOPNET also had lower rate of MEN-1 mutation.

123 Background: The effect of hospital volume on the outcome of stage 4 pancreatic cancer is not well known. We evaluated the effect of hospital volume on time to treatment and survival in patients with stage 4 pancreatic adenocarcinoma (PDAC). Methods: We used the National Cancer Database, including 1,319 hospitals to identify the study population. Adult (≥ 18 years) patients with stage 4 PDAC were included. We excluded the patients who had other histologies, unknown follow-up, and survival of less than 30 days. We categorized the hospital volume as three groups: low- (75th centile). Time from diagnosis to chemotherapy initiation was classified as early (≤ 6 weeks) or late (> 6 weeks). Kaplan-Meier and Cox regression methods were used to evaluate the overall survival (OS) between HVH and low-volume hospital (LVH) groups. Results: Among 72,531 patients with stage 4 PDAC; 65% received chemotherapy. Patients treated at HVH had higher rates of chemotherapy ( 73% vs. 60%, p

e16716 Background: Gallbladder cancer (GBC) is an uncommon but highly fatal malignancy. Surgery remains the only potentially curative treatment for GBC. The role of neoadjuvant chemotherapy in patients with locally advanced GBC undergoing surgery is unknown. We studied the association of neoadjuvant chemotherapy on survival in locally advanced GBC patients who underwent resection. Methods: We identified adult patients with locally advanced (stage III-IV) GBC who underwent definitive surgery between 2004 and 2016 using the National Cancer Database. Treatment was categorized as neoadjuvant chemotherapy plus surgery (NAT), surgery plus adjuvant therapy (AT), and surgery alone (SA). Categorical variables were compared using the chi-square test with Bonferroni correction. Kaplan-Meier and Cox regression were used for survival analyses. We used 1:3 nearest neighbor propensity score matching based on NAT for each group. Results: Out of a total of 5,962 patients, 122 (2.2%) received NAT, 2934 (53.6%) AT, and 2421 (44.2%) SA. NAT was associated with private insurance and treatment at an academic/research facility (all p < .001) while SA patients were older, Hispanic, had government insurance, and higher comorbidities (all p < .001). Although all groups had similar lymph node assessment (NAT: 45%, AT: 46%, SA: 37%, p < .001), NAT was associated with lymph node negative disease (NAT 23%, AT 13.2%, SA 13.2%, p < .001). Median overall survival was higher in NAT compared to AT or SA (21 vs. 14 vs. 6 months, p < .001) which persisted after propensity score matching (21 vs. 15 vs. 9 months, p < .001) and multivariable regression analysis (Table). In node positive GBC, NAT was associated with improved median overall survival (NAT 24, AT 18, SA 8 months, p < .001). Conclusions: NAT is infrequently used in patients with locally advanced GBC. NAT is associated with improved median overall survival compared to AT and SA, and appears to be most beneficial in node positive disease. Prospective studies are needed to evaluate the role of neoadjuvant chemotherapy in locally advanced GB. [Table: see text]

e16710 Background: Pancreatic neuroendocrine tumor (PNET) is a rare disease. There is a lack of data on the association of age and sociodemographic characteristics on survival in patients with NET. Our study aimed to evaluate clinical, molecular differences, and survival outcomes in young-onset PNET (YOPNET) and late-onset PNET (LOPNET). Methods: We used the National Cancer Database to identify patients with YOPNET ( < 50 years) and LOPNET (≥50 years) who underwent definitive surgery diagnosed between 2004 and 2015. We calculated the survival outcomes using Kaplan Meier and Cox regression methods before and after 1:1 nearest neighbor propensity score matching. We used the AACR Genie cohort to compare mutation frequency with the Benjamini-Hochberg method between YOPNET and LOPNET. Results: A total of 6,259 patients with PNET were selected, of which 27% were YOPNET. Patients with YOPNET were more likely to be female, black, and Hispanic, have private insurance versus LOPNET (all p < .001). Patients with YOPNET had less comorbidity score, but greater tumor size ( > 4 cm) and higher stage (all p < .001). Patients with LOPNET were more likely to be treated in an academic/research facility ( p < .001). Patients with YOPNET were more likely to be located in the head of the pancreas ( p = .004). Patients with YOPNET had better overall survival (OS) as compared to patients with LOPNET which persisted after propensity score matching ( p < .001). This difference continued after adjusting for clinically significant variables (unmatched p = .008; matched p = .01) (Table). For genomic analysis, we identified 177 patients; 27% was YOPNET, and patients with YOPNET had a significantly lower rate of MEN1 mutation than patients with LOPNET (26% vs. 56%, p < .001). Rate of other mutations was not different. Conclusions: Patients with YOPNET who underwent surgery had better OS than LOPNET and were more likely to have higher stage and greater tumor size. YOPNET represents a disease with distinct clinical and molecular features. [Table: see text]

TPS599 Background: Phosphatidylserine (PS) is an immunosuppressive lipid segregated to the inner leaflet of the plasma membrane of normal cells. PS is externalized to the outer leaflet of the plasma membrane on cells that line tumor blood vessels, tumor cells and exosomes in the tumor microenvironment creating a specific target for anticancer treatments. Bavituximab is a PS targeting antibody that binds to PS via β2 glycoprotein-1 (β2GP1), an abundant serum glycoprotein. Bavituximab can modulate the tumor immune microenvironment to promote anti-tumor immune activity. This clinical trial is evaluating the effect of combining bavituxumab and the checkpoint inhibitor pembrolizumab in patients with advanced hepatocellular carcinoma (HCC). Methods: This is a single arm phase 2 clinical trial for patients with histologically confirmed HCC which is not eligible for curative and/or loco-regional therapy. No prior systemic therapy for HCC is allowed. Patients should be > 18 years, have Child-Pugh Score A, ECOG 0-1 and meet the following laboratory criteria: total bilirubin ≤ 2.0, INR ≤ 1.7; Hgb ≥ 8.5 g/dl; AST, ALT ≤5 times ULN; platelet ≥ 50,000/mm3; and albumin ≥ 2.5 g/dl. Major exclusion criteria include thromboembolic event within the last 6 months, clinically evident ascites, GI bleeding within 2 months, uncontrolled hypertension, HIV and autoimmune disease (except DM1, childhood allergies and thyroid abnormalities if adequately controlled. Dosing is as follows: Bavituximab: 3 mg/kg IV weekly, pembrolizumab: 200 mg IV every 3 weeks. The primary objective is to determine the overall response rate (ORR) of therapy. Secondary objectives are to determine the safety, tolerability and other measures of clinical efficacy (overall survival, 6 month PFS, and duration of response). Tumor and blood samples are being collected to assess for pharmacodynamic biomarkers. A minimax two-stage method will be used to analyze ORR after the first 15 patients are accrued. If 3 or more of the first 15 patients have either a complete or partial response, 13 additional patients will be enrolled. ORR will be analyzed by RECIST 1.1 criteria. Clinical trial information: NCT03519997.

674 Background: Young-onset pancreatic adenocarcinoma (YOPC) is uncommon but there are limited studies for these patients. We used a population based registry to compare the characteristics and outcomes of young-onset vs. older patients with pancreatic adenocarcinoma. Methods: We selected the patients with pancreatic adenocarcinoma from the SEER registry diagnosed between 2004 and 2015. Cases with age of diagnosis less than 50 were termed young-onset pancreatic cancer. Stage 4 patients were excluded. We compared baseline characteristics of YOPC vs. older using Chi-square. Kaplan Meier and Cox regression were used for survival analysis of these patients. Results: Of 28,904 patients, 1,415 (4.9%) had YOPC while 27,489 (95.1%) were older. YOPC were more likely to have stage 3 compared to older patients with PC (31.6% vs. 25.3%). YOPC had a higher rate of surgery than older patients (40% vs. 29.1%, p < 0.001), were more likely to be male, black and of Hispanic ethnicity. The primary tumor location was not different between the two groups. Overall survival (OS) was higher in YOPC versus older patients (12 vs. 9 months, p < 0.001). The analysis of multivariable Cox regression confirmed that there is a significant association between survival and YOPC group after adjusting for stage, grade, gender, ethnicity, surgery and race (HR 1.23, 95% CI: 1.13-1.33, p < 0.001). Conclusions: Patients with non-metastatic YOPC represent a group of patients with distinct clinical characteristics. YOPC have a higher rate of surgery and better overall survival compared to older patients. [Table: see text]

Cutaneous involvement as the presenting sign of internal carcinoma is rare and is associated with poor prognosis. Here we present the case of a 66-year-old woman with anorectal adenocarcinoma who presented with tumor involvement of the skin.

Background ARQ761 (β-lapachone) is a hydroquinone analog that exploits the unique elevation of NQO1 found in 90% of pancreatic cancer cells and causes tumor-specific cell death by eliciting a futile redox cycle generating high levels of reactive oxygen species and ultimately PARP1 hyperactivation-dependent cell death. ARQ761 induces cell death which is: (i) not dependent on p53 status; (ii) not dependent on cell cycle status; (iii) not affected by known oncogenic driver or carrier mutations; and (iv) not affected by loss of caspases. We demonstrated that ARQ761 has synergistic antitumor activity with chemotherapy in pancreatic cancer preclinical models. Methods This was a single arm phase-1b clinical trial of ARQ-761 plus gemcitabine and nab-paclitaxel for patients with pancreatic cancer. The primary objective was to determine the maximum tolerated dose of ARQ 761 in combination with chemotherapy. Secondary objectives were to assess clinical activity, safety, tolerability and pharmacodynamic effects. Patients with metastatic pancreatic adenocarcinoma not treated with prior gemcitabine with adequate hematologic, renal and liver function and good performance status were enrolled. Dose levels are described in the table. Immunohistochemistry was used to assess for NQO1 expression. Results Fifteen patients were evaluable for safety, 67% were male, 53% had received prior chemotherapy and 96% tested positive for NQO1 expression. The most common treatment related toxicities were anemia and fatigue. The following dose limiting toxicities were observed (dose level 1: neutropenia, anemia, dose level -1 fatigue). The maximum tolerated dose of ARQ761/gemcitabine/nab-paclitaxel is 156/800/100mg/m2. Tumour response was stable disease in 53% Pharmacodynamic analysis was performed on post treatment biopsies. Conclusions We report results of a phase 1b clinical trial of the novel NQO1 bioactivatable compound, ARQ761, in combination with gemcitabine/nab-paclitaxel. Treatment related toxicities were observed which were noted in preclinical models and managed with routine dose adjustment. Clinical trial identification NCT02514031. Legal entity responsible for the study UT Southwestern Medical Center. Funding Pancreatic Cancer Action Network, Rising Tide, Gateway, Arqule Pharmaceuticals. Disclosure M.S. Beg: Research grant / Funding (institution): Arqule; Advisory / Consultancy: Array; Advisory / Consultancy, Speaker Bureau / Expert testimony: BMS; Advisory / Consultancy, Speaker Bureau / Expert testimony: Ipsen; Advisory / Consultancy: Boston Biomedical; Advisory / Consultancy: Guardant; Advisory / Consultancy: Genentech. D. Boothman: Research grant / Funding (institution): Arqule; Advisory / Consultancy: Toray global; Speaker Bureau / Expert testimony: systems oncology. Y. Arriaga: Full / Part-time employment: IBM. A. Sanjeeviaiah: Advisory / Consultancy, 1500$: Guardant health. B. Schwartz: Full / Part-time employment: Arqule. All other authors have declared no conflicts of interest.

143 Background: Patients with GI malignancies suffer from weight loss, sarcopenia and malnutrition contributing to poor outcomes and reduced survival. Early nutritional intervention in at-risk patients by a clinical dietitian can prevent or delay the onset of malnutrition. In the GI Oncology clinic at the UTSW SCCC, the rate of documented nutritional plan by a clinical dietitian within first 90 days of new patient encounter was low. Methods: We aimed to increase the rate of a documented nutrition assessment by a clinical dietitian to 65% within 90 days of a new patient encounter. Baseline data from the electronic medical record were obtained from Sep 2017-Oct 2018. Group sessions were arranged to apply QI methodologies to determine steps to a documented nutritional plan by a clinical dietitian. Patient advocates were interviewed to assess patient perspective. Sequential PDSA cycles were planned to improve rates of nutritional plan documentation and data were obtained every 2 weeks. Results: At baseline, 41.1% of new patients in the two-month data collection period had documented nutrition plans within 90 days of their first appointment. Most of these patients were on intravenous chemotherapy. EMR-based nutrition assessments identified 17% of all new patients at risk of malnutrition. Multiple causes for low baseline rates of nutrition plans were discovered, including patient or family characteristics and needs, clinical dietitian resources, physician limitations, process flaws, as well as difficulty with the EMR. Patient-centered PDSA cycles directed toward patients, and clinical staff to increase the rate of documented nutritional plan are ongoing. After the first PDSA cycle, early 2-week assessment shows documented rate of nutritional plan of 28%. Authors expect it to increase with longer follow-up and subsequent PDSA cycles. Conclusions: Malnutrition in GI cancer is prevalent and under-recognized in routine clinical encounters. Addressing malnutrition is important aspect from patient perspective. We are continuing ongoing efforts to increase the rate of nutritional assessments in these patients.

10027Background: Previous studies have suggested that racial disparities exist for DNR orders in cancer patients. We examined racial differences in DNR orders for pancreatic cancer patients in an u...

489 Background: There has been an increase in Palliative care utilization in cancer patients. We examined trends of palliative care and intensive care utilization in pancreatic cancer patients in an urban setting safety net hospital. Methods: This is a retrospective analysis of pancreatic cancer patients seen at the Parkland Health and Hospital System between January 1999 and September 2016. Cancer cases and receipt of palliative care were identified from prospectively maintained registries. Health care utilization including intensive care unit (ICU) was reviewed. All statistical analysis was done using IBM SPSS version 24. Results: We identified 455 new diagnoses of pancreatic cancer, mean age 61 years, 227 (50%) female and 228 (50%) white. Of these, 277 (61%) received palliative care ever. Patient who received palliative care were more likely to be younger (mean age, 59.3+-12 vs 62.8 +- 12 years) and have stage 4 disease vs stage 1-3 disease (p 0.006, and p 0.003 respectively). There was no statistically significant difference in palliative care utilization between gender and ethnicity groups. 140 patients had a DNR order and 29 required ICU admission at any point. A first contact with palliative care consult was obtained < = 7 days before death for 29 (10%) patients, < = 30 days before death for 86 (31%) patients, 30-60 days before death for 50 (18%) and more than 60 days before death for 141 (51%) patients. Patients receiving palliative care were more likely to have a DNR status (p < 0.001) but had no difference in ICU use within the last 30 days of life (p 0.285). Conclusions: The rate of palliative care in patients with pancreatic cancer in this cohort from a safety net hospital is higher than nationally reported studies. Most patients received palliative care > 30 days before death. While patients received early palliative care, it did not result in reduced ICU care. Factors influencing ICU care utilization near the end of life need further study.

463 Background: FOLFIRINOX therapy is associated with improved outcome in patients with gastrointestinal cancers. The regimen can be associated with significant toxicity and empiric dose modifications are often used. We analyzed 1) real-world prescribing patterns of FOLFIRINOX and 2) toxicity of therapy. Methods: Patients undergoing FOLFIRINOX chemotherapy at an academic, NCI-Designated Comprehensive Cancer Center were identified and electronic medical records reviewed. Patients who received at least one dose of FOLFIRINOX were included. Chemotherapy dose, growth factor use and toxicity data was abstracted for the first 8 weeks. ‘Standard FOLFIRNOX’ was defined as the regimen utilized by Conroy et al (NEJM 2011). Any empiric reduction/withholding of drug dose for cycle 1 was classified as ‘modified FOLFIRINOX’. Bivariate analysis was performed on the data. Results: There were 111 patients seen between 5/2011-3/2017 and 94% had pancreatic cancer. Age range was 29-87 years and 52% were female. 59% received ‘modified FOLFIRINOX’ and 20% received empiric growth factors. Line of therapy for standard vs modified respectively was 71.1% vs 45.5% for 1st, 17.8% vs 36.4% for 2nd, and 11.1% vs 18.2% for beyond 2nd (p = 0.03). Patients with ‘modified FOLFIRINOX’ were more likely to have metastatic disease (p = 0.01), have received second line or beyond, and higher ECOG score (p = 0.03). Patients with ‘modified FOLFIRINOX’ had a trend toward fewer treatment-related ED visits or hospitalization vs ‘standard FOLFIRINOX’ (27.2% vs 42.2% p = 0.10) and fewer treatment delays (25.8% vs 42.2% p = 0.07). Conclusions: In the real world setting, a majority of patients on FOLFIRINOX receive empiric dose modifications. Although modified dose did not translate to a significant difference in ED visits, hospitalizations or treatment delays, there was a trend toward fewer events.

e15752 Background: Insulin resistance (IR) in pancreatic cancer (PC) patients is associated with cachexia and poor outcome. Pioglitazone (PIO) improves insulin sensitivity via activation of peroxisome proliferator-activated receptor (PPAR gamma). Effects of PIO on insulin sensitivity, glucose homeostasis, and circulating adipokine levels in PC have not been examined. Methods: Patients with metastatic PC were administered PIO 45mg/day orally for 8 weeks concurrent with chemotherapy. Patients with known DM at enrollment were identified Fasting plasma was collected at baseline, weeks 2, 4, 6, 8 of pioglitazone treatment and at 2 weeks-post treatment. The primary objective was to describe changes in indicators of IR, including glucoregulatory hormone levels, glucose tolerance, and inflammatory cytokines. Results: Fourteen patients (age 64y), with a mean BMI of 28 were enrolled. Mean adiponectin increased from baseline to week 8 of treatment (14.2± 3.3 and 46.9±11.4µg/ml, respectively, P ≤ 0.01), and returned to baseline levels at 2 weeks post-treatment (15.1±1.9 µg/ml). Markers of IR (serum glucose, insulin, glucagon, and response to an oral glucose bolus) did not correlate with tumor size, inflammatory cytokine levels, GM-CSF, or CA19-9. A dichotomous response to PIO treatment was observed between non-diabetic and T2DM patients. Fasting insulin increased 70.6±31.3% from baseline to treatment week 8 in patients with T2DM. In contrast, treatment of non-diabetic patients decreased fasting insulin by 40.2±6.3%, demonstrating a significant, P ≤ 0.01, difference in treatment response between PC patients with or without T2DM. Conclusions: We have demonstrated that PC patients respond to 8 weeks of PIO treatment with a significant rise in the insulin sensitizing adipokine, Adiponectin, with a complete wash-out after 2 weeks of cessation. PIO results in opposing fasting insulin responses in non-diabetic and DM patients suggests that diabetes status plays a significant role in the glucoregulatory effects of PIO treatment in PC patients. Clinical trial information Clinical trial information: NCT01838317.

651 Background: Regorafenib is an oral multi kinase inhibitor with clinical activity in colorectal cancer, gastrointestinal stromal tumors and hepatocellular cancer. Drug- drug interactions are commonly encountered in clinical practice and warfarin continues to be the most commonly prescribed anticoagulant among cancer patients in the clinic. The interaction of regorafenib with warfarin has not been studied. Methods: Patients at a single institution being prescribed regorafenib were identified and charts reviewed for concurrent warfarin therapy. Baseline characteristics, indications for regorafenib and warfarin therapy was determined. Patients were followed up to the first 8 weeks of combination therapy. Results: We identified 6 patients on concurrent regorafenib and warfarin. All patients had refractory colon cancer. Median age was 74, 4 were male and 5 white. Indication for warfarin was venous thromboembolism in 5 and atrial fibrillation in 1. Baseline INR ranged from 1.1 to 2.4. An increase in INR was seen in all six patients (peak INR range 4.5 to > 12) which improved with dose modification. Time to peak INR was less than 6 weeks (INR trends in all patients shown in figure 1). There were no bleeding complications noted in any patient. Conclusions: We report a clinically significant interaction between warfarin and regorafenib with patients demonstrating an increase in INR. Patients who are anticoagulated with warfarin and start regorafenib should be closely monitored and upfront warfarin dose reduction should be considered. Further drug-drug interaction studies are needed to determine the exact mechanism of drug-drug interaction between these medications.