Bronchoesophageal and tracheoesophageal fistulas are an uncommon but well-documented complication of Hodgkin lymphoma (HL). To our knowledge, a fistula directly connecting the esophagus with the lung (esophageal-pulmonary fistula) and resulting from HL has not been reported in the literature. We present a case of HL complicated with an esophageal-pulmonary fistula. The early recognition and treatment of an esophageal-pulmonary fistula in patients with HL are imperative since, with appropriate treatment, HL complicated with an aero-esophageal fistula has the same prognosis as those without one, unlike the dismal prognosis in esophageal and lung cancer. Endoscopic esophageal stenting followed by chemotherapy is the preferred treatment approach. This leads to the healing of the fistula and the prolongation of patient survival.

e12096 Background: Chemotherapy-radiotherapy time intervals (CRTI) in breast cancer vary based on recovery from chemotherapy side effects and preferences of physicians and patients. Our study aimed to determine the association of the time-interval between chemotherapy and radiation with prognosis (recurrence and survival duration) in non-metastatic breast cancer patients (stages I-III). Methods: We included female patients from Karmanos Cancer Center (KCC) database > 21 years, diagnosed with non-metastatic breast cancer from 2005-2015, who underwent surgery, chemotherapy and radiation. CRTI was divided into <24, 24-30 and >30 weeks. Cox regression analysis using age, race, stage, grade, ER/PR status and CRTI variables was done to determine the independent predictors of prognosis. Kaplan-Meier survival curves were used to demonstrate differences in survival time of CRTI groups. Results: We included 553 patients, majority were Caucasian (460 [83.5%]) with a mean age of 57.0 years (SD 11.9). Patients in the >30 weeks CRTI group were more likely to have stage III breast carcinoma when compared to <24 and 24-30 weeks groups (45.0% vs.12.4% vs. 27.8%, p-value <0.001). There were no significant differences in the local and distant recurrence rates among the 3 groups. Patients in <24 weeks CRTI group had a greater mean survival (118.7 vs. 109.2 vs.100.9 months; p-value 0.016) when compared to 24-30 and >30 weeks groups. Only clinical stage (stage 3 vs. 1 HR 3.60; 95% CI 1.89-7.02) and ER/PR status (negative vs. positive HR 1.74; 95% CI 1.02-2.94) were independent predictors of overall survival in multivariate analysis. CRTI was not significantly associated with survival in multivariate analysis (<24 weeks ref, 24-30 weeks HR 1.09; 95% CI 0.59-1.93, >30 weeks HR 1.48; 95% CI 0.82-2.59). However, patients in >30 weeks group were at greater risk of dying when compared to <24 weeks, even after controlling for stage. Conclusions: Our study showed a trend towards a worse outcome in patients with longer chemo-radiation interval, despite controlling for other variables in the multivariate analysis. The conclusions of this single center study need to be verified by further studies with larger sample size.