Objectives/HypothesisTo assess for potential urban and rural disparities in head and neck cancer (HNC) outcomes within a single‐payer healthcare system.Study DesignA large retrospective population‐based cohort analysis of consecutive HNC patients treated in British Columbia, Canada between 2001 and 2010 was conducted.MethodsAll patients diagnosed with HNC from 2001 to 2010 and referred to any one of five British Columbia Cancer Agency centers for management were reviewed. Based on census data, patients were classified into: 1) rural, 2) small urban, 3) moderate urban, and 4) large urban areas. Kaplan‐Meier methods and Cox regression models were used to correlate site of residence with overall survival (OS), controlling for prognostic factors that included sociodemographic and other tumor and treatment‐related characteristics.ResultsWe identified 3,036 patients; the median age was 64 years, 26% were women, and 32% had Eastern Cooperative Oncology Group (ECOG) 0 or 1. The majority resided in large urban areas (55%) followed by rural (22%), moderate urban (13%), and small urban (10%). In regression analyses, smoking (hazard ratio [HR]: 2.10, 95% confidence interval [CI]: 1.28‐3.45, P < .001), ECOG 2 + (HR: 3.44, 95% CI: 2.26‐5.22, P < .001), oral cavity (HR: 1.54, 95% CI: 1.03‐2.32, P = .04) and hypopharyngeal tumors (HR: 2.31, 95% CI: 1.42‐3.77, P = .00), and large tumor size (HR: 1.69, 95% CI: 1.08‐2.64, P = .02) were correlated with inferior OS, but site of residence was not. When stratified by type of treatment, OS remained similar irrespective of urban or rural residence.ConclusionsUrban–rural differences in HNC survival outcomes were not observed.Level of Evidence2c. Laryngoscope, 128:852–858, 2018

OBJECTIVES:Optimal treatment of rectal cancer (RC) requires multidisciplinary care. We examined whether distance to treatment center or community size impacts access to multimodality care and population-based outcomes in RC.METHODS:Patients diagnosed with stage II/III RC from 1999 to 2009 and treated at 1 of 6 regional cancer centers in British Columbia were reviewed. Distance to treatment center was determined for each patient. Communities were classified as rural, small, medium, and large population centers. Logistic and Cox regression models assessed associations of distance and community size with treatment received as well as cancer-specific (CSS) and overall survival (OS).RESULTS:Of 3,158 patients, 93.6% underwent surgery, 86.3% received radiotherapy, and 51.3% were treated with adjuvant chemotherapy (AC). Median time from diagnosis to oncologic consultation was longer for those >100 km from a treatment center or residing in medium/rural communities. Logistic regression demonstrated no correlation between distance or community size and receipt of treatment modality. Univariate analysis showed similar CSS (P = .18, .88) and OS (P = .36, .47) based on community size and distance, respectively. In multivariate analysis, distance >100 km had inferior CSS (Hazard Ratio [HR] 1.39, 95% CI: 1.03-1.88; P = .031). There was no consistent trend between decreasing community size and outcomes; however, living in a small center was associated with improved OS (HR 0.58, 95% CI: 0.38-0.88; P = .011) and CSS (HR 0.42, 95% CI: 0.25-0.70; P = .001).CONCLUSIONS:In this population-based study, there were no urban-rural differences in access to multidisciplinary care, but increased distance may be associated with worse cancer-specific outcomes.

Adjuvant folinic acid, fluorouracil, and oxaliplatin (FOLFOX) chemotherapy for resected high-risk colon cancer is associated with a low risk of febrile neutropenia (FN). Neutropenia, however, is a common cause of dose modification or delay with unknown consequences on outcomes. We examined the effect of neutropenia-related and other dose-limiting toxicities and relative dose intensity of oxaliplatin and 5-FU, on relapse-free and overall survival in patients treated with FOLFOX chemotherapy for resected high-risk colon cancer.

423 Background: Adjuvant FOLFOX chemotherapy for resected high-risk colon cancer is associated with a low risk of febrile neutropenia (FN). Nonetheless, neutropenia is a common cause of dose modification or delay with unknown consequences on outcomes. We examined the effect of neutropenia-related and other dose limiting toxicities (DLT), relative dose intensity (RDI) of oxaliplatin and 5-FU, on relapse-free (RFS) and overall survival (OS) in patients treated with FOLFOX chemotherapy for resected high-risk colon cancer. Methods: A chart review was conducted on patients treated at the British Columbia Cancer Agency with ≥ 1 cycle of FOLFOX chemotherapy for resected stage II or III colon cancer between January 1, 2006 and December 31, 2007. RFS and OS were analyzed by the Kaplan-Meier method and multivariate Cox proportional-hazards models. Results: 114 patients (median age 59.2 years, 43.8% male, 98% stage III, median follow-up 5.2 years) were included. 90% of patients experienced any DLT, while 58% of patients had a neutropenia related DLT. There were no documented episodes of FN. G-CSF was used in 9.6% of patients. Median RDI was 81% and 85% for oxaliplatin and 5-FU, respectively. Oxaliplatin and 5-FU RDI were not associated with RFS or OS when analyzed as continuous variable or categorically (Table). Conclusions: DLTs affect the majority of patients on adjuvant FOLFOX for high-risk colon cancer, but RFS and OS do not appear to be affected by the associated lower RDI of oxaliplatin and 5-FU. [Table: see text]

6570 Background: Management of HNC is becoming more specialized where effective treatments frequently require multidisciplinary and multimodality care. Concerns exist that access to such complex care may be suboptimal for marginalized subsets of the population. Our aim was to examine for potential urban and rural disparities in HNC outcomes within a population-based single payer healthcare system. Methods: All patients diagnosed with HNC from 2001 to 2010 and referred to any 1 of 5 regional comprehensive cancer centers in British Columbia, Canada were reviewed. Based on census data, patients were classified into 4 categories: 1) rural 2) small urban 3) moderate urban and 4) large urban areas. Kaplan Meier methods and Cox regression were used to correlate site of residence with overall survival (OS), controlling for prognostic factors that included socio-demographics and other tumor and treatment-related characteristics. Results: A total of 3,036 patients were included: median age was 64 years, 74% were men, and 32% were ECOG 0/1. The majority resided in large urban areas (55%) followed by rural (22%), moderate urban (13%), and small urban (10%). There were no clinically significant differences in baseline characteristics across the 4 groups. In multivariate-adjusted models, advanced age >/= 65 years (HR 1.58, 95%CI 1.21-2.06, p<0.001), ECOG 2+ (HR 4.20, 95%CI 2.41-4.93, p<0.001), and lack of multimodality treatment (HR 2.88, 95%CI 1.72-4.81, p<0.001) correlated with inferior OS, but site of residence did not (Table). In subgroup analyses that stratified by type of treatment (radiation, chemotherapy, and/or surgery) and anatomic location of HNC (oral cavity, oropharynx, larynx, hypopharynx, nasopharynx), OS remained similar irrespective of urban or rural residence. Conclusions: Urban-rural differences in outcomes were not observed. The centralization of HNC management in this large population-based cohort represents an appropriate model of care for cancers in which multimodality treatments are increasingly complex and where disparities in access may be prevalent. Residence HR for death 95%CI P-value Rural 1.0 -- -- Small urban 1.27 0.77-2.10 0.35 Moderate urban 0.84 0.52-1.37 0.49 Large urban 1.19 0.80-1.56 0.51

e14603 Background: While urban-rural differences in cancer care are well described, the etiology of these disparities is unclear. Our aims were to 1) characterize differences in AC use based on community size and 2) determine if such disparities are mediated through variations in driving distance (DD) and travel time (TT) to closest cancer center. Methods: Patients diagnosed with stage 2 and 3 RC from 1999 to 2009 and referred to any 1 of 5 regional cancer centers in British Columbia were reviewed. Communities were classified as rural, small urban, moderate urban and large urban based on census data. Using zip codes and a distance matrix application interface, DD and TT to the closest cancer center were determined and categorized into quartiles. Stepwise logistic regression models were constructed to explore AC use based on urban vs rural communities, adjusting for DD and TT. Results: A total of 3,017 patients were identified: median age was 67 years (IQR 58-75), 64% were men and 58% received AC. Patients were distributed across various communities: rural 36%; small urban 12%; moderate urban 13%; and large urban 39%. There were no differences in baseline patient and disease characteristics based on community size (all p>0.05). Compared to patients in large urban centers, those living in rural, small urban and moderate urban areas were less likely to be treated with AC (62 vs 49 vs 54 vs 58%, respectively, p<0.001). Likewise, DD and TT were shortest for large urban and longest for rural residents (both p<0.001). In multivariate analyses that controlled for confounders, urban-rural disparities in receipt of AC persisted, but these differences significantly diminished after adjusting for DD, TT, or both (Table). Conclusions: Urban-rural disparities in AC use is partly mediated by commute. Strategic distribution of cancer services that reduce DD and TT to cancer centers may improve access to AC for a number of RC patients who are living in smaller communities. [Table: see text]

486 Background: Trials have strict inclusion and exclusion criteria to maintain internal validity. However, study findings are often applied to pts in clinical practice who do not satisfy all of the CTEC. Whether these pts benefit from treatment is unclear. Our objectives were 1) to characterize cancer-specific (CSS) and overall survival (OS) in a population-based cohort of trial-eligible (TE) and trial-ineligible (TI) pts receiving adjuvant chemotherapy (CT) and 2) to compare their outcomes with those not treated with CT. Methods: Pts diagnosed with stage III CC between 2006 and 2008, referred to 1 of 5 regional cancer centers in British Columbia, and evaluated for possible adjuvant CT within 12 weeks of curative surgery were reviewed. Pts were defined as TE if aged 18 to 79 years, ECOG 0 to 1, CEA <10 ng/ml, had not received prior CT or radiation, and had adequate blood counts, cardiac, liver, and kidney function. All other pts were considered TI. Using Kaplan-Meier and Cox regression analyses, we compared outcomes between TE and TI pts who received CT vs. those who did not. Results: A total of 821 pts were identified: median age was 68 years, 52% were men, 85% were ECOG 0 to1, and 71% received adjuvant CT. Among pts treated with CT, 405 (70%) were TE and 177 (30%) were TI. Compared to TI pts, those who were TE were younger (p<0.01) and more likely to receive FOLFOX than capecitabine (65 vs. 46%, p<0.01). CSS and OS were significantly different among pts who were TE, TI, and those who did not receive CT (p<0.01) (Table). In multivariate analyses that adjusted for confounders, both TI pts and those not treated with CT had worse prognoses than TE pts (HR for CC deaths 1.32, 95%CI 0.86-2.02 and 2.77, 95%CI 1.92-3.99, respectively, p trend <0.01; HR for all deaths 1.24, 95%CI 0.85-1.80 and 2.95, 95%CI 2.17-4.00, respectively, p trend <0.01). Conclusions: A considerable number of stage III CC pts who did not fit CTEC were treated with adjuvant CT. While outcomes in the TI group were worse than those in the TE group, CSS and OS were still better than the subset that did not receive any CT. [Table: see text]