Search

59 Background: 5-Fluorouracil (5-FU) is a component of first-line treatment regimens for metastatic colorectal cancer (mCRC). Historically, 5-FU is administered as a bolus followed by an infusion. However, the bolus dose adds substantial toxicity and is often withheld in patients with limited functional status or high-risk comorbidities, but its impact on treatment outcomes remains unclear. Small studies suggest that it may be omitted. The aim of this study was to determine whether omission of the 5-FU bolus is associated with a difference in overall survival (OS). Methods: An electronic health record-derived national multicenter oncology database from Flatiron Health was queried to select patients with mCRC who received a first-line 5-FU-containing regimen. Demographics, relevant labs, treatment details, and survival outcomes were collected. Propensity score (PrS) matching and OS analysis were performed incorporating age, race, sex, ECOG score, combination drug regimen, and baseline creatinine and bilirubin. Variables with p < 0.10 in univariable Cox proportional hazards models were included in the multivariable analysis. Results: We included 9741 patients with mCRC who received 5-FU-based regimens. All individuals received a 5-FU infusion, and 7901 (81%) also received a 5-FU bolus. Among our entire cohort, 43% were female, 23% were > 70 years, 66% were white, and 89% had ECOG ≤ 1. Over a median follow-up time of 19 months, 5847 patients (60%) died. In the unmatched univariable (HR 0.94, 95% CI 0.88-1.00, p = 0.06) and multivariable (aHR 0.85, 95% CI 0.93-1.06, p = 0.85) analyses, there was no association between the use of bolus 5-FU and OS. A number of factors were associated with an increased risk of death, including older age, high ECOG scores, and elevated bilirubin or creatinine levels. Similarly, in our PrS-matched dataset (n = 6126), the use of a 5-FU bolus was not associated with OS (HR 0.98; 95% CI 0.91-1.06; p = 0.64). Conclusions: The findings of our PrS-matched multicenter cohort study indicate that, after adjusting for host and treatment factors, 5-FU bolus dosing was not associated with an overall survival benefit among patients with mCRC. These results suggest that, in mCRC, the addition of a 5-FU bolus does not appear to add efficacy to regimens utilizing infusional 5-FU. Future work is necessary to determine the role of bolus dosing in the adjuvant setting and its impact on other clinically-relevant outcomes.

11059 Background: As the largest integrated healthcare system in the US, the Veterans Health Administration (VHA) is chartered “to serve as the primary backup for any health care services needed...in the event of war or national emergency” according to a 1982 Congressional Act. This mission was invoked during the COVID 19 pandemic to divert clinical and research resources, resulting in dramatic changes in the activities of investigators and research staff. This study uses established behavioral change frameworks to examine the experience of VHA cancer researchers during the pandemic. Methods: We conducted a mixed methods study on the effects of the pandemic on VHA researchers. We used an electronic questionnaire composed of Likert scale statements and open-ended questions, which was constructed using the 14 domains of the Theoretical Domains Framework (TDF) and the Capability, Opportunity, and Motivation (COM-B) model for behavior change. Likert scale statements were quantified and content analysis was performed on open ended responses. Results: The questionnaire was distributed electronically to 118 researchers participating in national VHA collaborations, with 42 responses (36%). Several significant themes and findings emerged. Only 36% of respondents did not feel that their research focus changed during the pandemic. (TDF: professional role and identity maps to COM-B Motivation). And 81% of respondents reported a diverse range of research interests (COM-B: M). Only 11 of 42 respondents (26%) reported prior experience with infectious disease research (TDF: skills maps to COM-B: C), and 31 of 42 (74%) agreed that they gained new research skills during the pandemic (COM-B: C). When asked to describe support structures helpful during the pandemic, 29% mentioned local supervisors, mentors and research staff (COM-B: C, M), 15% cited larger VHA organizations (COM-B: C) and 18% mentioned remote work abilities or telehealth (COM-B: O). Lack of timely communication from supervisors and remote work, particularly for individuals with caregiving responsibilities, were seen as limiting factors for research during the pandemic (COM-B: O). Although 83% of respondents agreed that they will be prepared to respond to future crises after COVID (COM-B: C), fewer than half (40%) of respondents felt professionally rewarded for pursuing research related to COVID, with 50% feeling neutral (COM-B: M). Conclusions: This study demonstrated the tremendous effects of the COVID 19 pandemic on research activities of VHA investigators. We identified perceptions of insufficient recognition and lack of professional advancement related to pandemic era research. Improving the structure of remote work and team communication represent high yield areas for improvement. It is essential to understand barriers and facilitators of crisis redeployment in the VHA, to make the transition effective and fulfilling for those carrying out the mission.

e18617 Background: Precision oncology, targeting therapy to the molecular specifics of one patient’s cancer, improves outcomes and decreases toxicity. Previous studies have documented underuse of next generation sequencing (NGS) in both the Veterans Health Administration (VHA) and private sector cancer care. Our group’s work has found that requiring providers to order and review tests in multiple electronic medical record systems for a single patient is a significant barrier to precision oncology. Following the incorporation of electronic medical record (EMR) NGS ordering into the VA EMR (CPRS) at a New York VA Medical Center, we examined the effect of the ordering change on precision oncology care for veterans. Methods: We identified patients eligible for NGS testing who had medical oncology clinic visits at the Manhattan VA in 2021, and divided patients into those whose NGS required external ordering (EO) (i.e. direct to Foundation Medicine (FM)) and those whose NGS used internal ordering (IO) (i.e. integrated CPRS ordering). We collected demographic and clinical information, including intention to order NGS, date of NGS order, specimen receipt, and testing results, on all patients from CPRS and FM web-based portal. Descriptive statistics were calculated and variables were compared with chi-square, T-tests, and Mood’s median test. Results: A total of 138 patients were identified, 99 in the EO group and 38 in the IO group. The groups were vastly male, (100% EO vs 95% IO), had similar ages (average age 69 and 75, respectively, p = 0.24), approximately half were self-reported Black race (48% EO vs 46% IO; p = 0.89), and the preponderance of prostate cancer was higher in the EO group vs the IO group (81% EO vs 62% IO, p = 0.02). For the 99 EO patients, 74 (75%) had a documented intention to order, and 47 (64%) of those had a specimen received by FM. Of the 38 patients in the IO group, 38 (100%) had intention to order and 25 had results received by FM (66% vs 64% IO vs EO p = 0.91). Compared to the EO group, the IO group had shorter time from intention to order to sample received (among those with positive values): mean 172 days (d) vs 18 d and median 40 d vs 6 d (p = 0.04). Among those with metastatic disease and documented intention to order NGS (EO n = 72, IO n = 32), the time from diagnosis of metastatic disease to intention to order NGS was significantly shorter after implementation of internal ordering: mean 764 d vs 698 d, median 226 d vs 15 d, (p = 0.004). The time from diagnosis of metastatic disease to NGS result in FM was also significantly shorter for IO patients (mean 980 d vs 251 d; median 46 d vs 428 d, p = 0.0003). Conclusions: We found that incorporating NGS testing into CPRS decreased delay from diagnosis of metastatic disease to intention to order NGS testing, and from intention to order NGS testing to results received, ultimately enabling faster precision oncology care for patients.

e17659 Background: Development of the EGFR tyrosine kinase inhibitors (TKI’s) and FDA approval of erlotinib in 2004 began a decade of advances in targeted therapy for advanced lung cancer. Lung ade...

TPS8132 Background: Metformin (MTF), an oral biguanide, has recently demonstrated preventative and therapeutic effects in both preclinical and observational clinical studies in non-small cell lung ...

8558 Background: The optimal surveillance imaging in the management of patients (pts) with non-Hodgkins lymphoma (NHL) remains undefined. Unnecessary scans increase health care costs and radiation ...

6571 Background: The optimal use and sequence of radiation therapy (RT), chemotherapy, and surgery for patients with stage III non-small cell lung cancer (NSCLC) is unknown. We conducted a cross-sectional study using the Surveillance Epidemiology and End Results (SEER) population-based cancer registry to examine the relationship between year of diagnosis and receipt of neoadjuvant RT for patients with stage III NSCLC diagnosed in the United States from 2000 to 2008, with additional focus on the role of race and treatment location. Methods: Thirteen SEER registries were used for data analysis. Patients diagnosed with stage III NSCLC treated with both surgery and RT between the years 2000 and 2008 were included. We used multiple logistic regression to describe the relationship between type of therapy for lung cancer and year of diagnosis. Covariates identified for analysis included age, grade, sex, race, ethnicity, and SEER registry. Results: The total sample size was 393,711 of which 5,949 patients met our inclusion criteria. The cohort had a mean age of 64, was predominantly white (85.8%) and had predominantly high grade tumors (50.9%). Use of neoadjuvant RT for stage III NSCLC was lowest in 2000 and higher in every year from 2002 to 2008 (OR 1.325 in 2002, p= 0.035; 1.921 in 2007, p

e17566 Background: Shortages of intravenous leucovorin were initially reported in late 2008, and availability plummeted nationally and at Continuum Cancer Centers of New York (CCCNY) in late 2011. NCCN guidelines recommend either the use of levoleucovorin or low dose leucovorin during the shortage. The impact of the leucovorin shortage on patient therapy remains unknown. Methods: We reviewed patient charts for all outpatients treated with leucovorin at CCCNY between April and September of 2010, 2011, and 2012. We recorded patient characteristics, and leucovorin use (dose, number of treatments); and calculated descriptive statistics. We classified each dose as either low (20-40 mg/m2) or high (200-500mg/m2) and compared leucovorin use between years with Chi Square and ANOVA tests. We also reviewed pharmacy purchasing data to evaluate the economic effect of the leucovorin shortage. Results: We identified 55 patients treated with 313 doses of leucovorin in 2010, 99 patients treated with 582 doses in 2011 and 118 patients treated with 742 doses in 2012. No patients received levoleucovorin. Patient characteristics, disease and stage were similar between years with colorectal cancer accounting for 78%, 69%, and 70% of patients in 2010, 2011, and 2012 (p=0.87). Low dose leucovorin was used in 30.0% of doses in 2010, 30.4% in 2011, and 99.1% in 2012 (p

e17589 Background: Recent evidence suggests that HIV may be a risk factor of lung cancer, independent of smoking . Due to the expanding population of HIV patients with lung cancer, there is a need to define the clinical course and tumor biology of these patients. Our analysis seeks to characterize the clinical and molecular features of HIV associated non small cell lung cancer (HIV-NSCLC) and evaluate outcomes in a New York City Cancer center that serves a racially and economically diverse population. Methods: We searched the Continuum Cancer Center Registry for cases of HIV-NSCLC diagnosed from 2002 to 2012. Charts were reviewed to determine patient and tumor characteristics, treatment and outcomes. Kaplan-Meier curves were constructed for survival and compared by means of the log rank test. Patient characteristics were compared to national data from the SEER database. Mutational analysis of archival tissue was performed by OnkoMatch Tumor genotyping. Results: HIV-NSCLC was idenfied in 74 patients. Median age (MA) was 55 compared to MA of 70 for non HIV-NSCLC. Patients were predominantly male (72%). Histology distribution was reflective of the non-HIV population (SEER): adenocarcinoma [31 (42%)], squamous [18 (24%)], NSCLC NOS [8 (11%)], poorly differentiated [6 (8%)], and other [9 (15%)]. Distribution of stage was similar to SEER with 39 (53%) patients presenting with stage IV. Lowest recorded CD4 count was or < 200), or by receipt of chemo or radiation. Mutational analysis on 7 patients demonstrated 3 with cMet overexpression, 2 KRAS mutations and 1 BRAF mutation. Conclusions: Among our HIV-NSCLC cohort, patients were diagnosed at younger age, but had similar stage and histology distributions as SEER database averages. HIV associated lung cancer appears to have a poor prognosis similar to that of the general population. Rates of treatment were low in our cohort and the potential for undertreatment warrants further study. Observed increased rate of cMET overexpression should prompt further molecular profiling in this population.