e17540 Background: Prophylactic percutaneous gastrostomy tube (PEG) insertion reduces toxicity from chemoradiation to the head and neck but is thought to increase long term feeding tube dependence. This study retrospectively examines incidence and risk factors for treatment related complications of oropharynx cancer patients with and without prophylactic PEGs. Methods: Rush University Medical Center oropharynx cancer patients who received definitive chemoradiation treatment between 2007-2018 were included. Classifications were: “therapeutic” PEG (pretreatment for immediate use due to inability to swallow), prophylactic PEG, reactive PEG (patient/physician preference or 10% weight loss from baseline), and no PEG inserted on treatment. We compared patients with reactive or no PEG to (1) patients with prophylactic PEG, and (2) patients with prophylactic PEG and therapeutic PEG. Multivariate linear and logistical regression models were used to test PEG effect on weight loss, hospital admission, and incidence of acute kidney injury (AKI). Models were adjusted for covariates (age, gender, race, HTN, CAD, DM, other comorbidity). Acute kidney injury (AKI) was creatinine 1.5-2x above baseline. Results: In all, 104 patients were included with mean age 60.1 (SD = 8.65) and baseline BMI 29.6 (SD = 5.62). 53.4% (N = 55) had a prophylactic PEG, 38.8% (N = 40) had reactive or no PEG, 7.8% (N = 8) had a therapeutic PEG. 80 (76.9%) were treated with cisplatin. For all patients, analyses showed that reactive PEG or no PEG patients were more likely to develop AKI during treatment compared to patients with a prophylactic PEG (OR:3.2, p = 0.03), and to patients with prophylactic PEG and therapeutic PEG combined (OR:3.5, p = 0.02). There were no statistically significant differences between PEG groups for weight loss and hospital admission rate. In cisplatin treated patients, reactive PEG or no PEG patients were more likely to be admitted to the hospital compared to prophylactic PEG patients (OR:3.8, p = 0.04). Compared to patients with prophylactic and therapeutic PEG combined, however, there was no statistically significant difference. Patients with reactive or no PEG were more likely to have AKI than prophylactic PEG (OR:5.2, p < 0.01), and as compared to patients with therapeutic or prophylactic PEG (OR:4.4, p = 0.02). Conclusions: Reactive PEGs were associated with increased AKI and hospitalizations compared with prophylactic PEG. With a reactive PEG model, patients may need to have routine lab work and monitoring adjusted to reduce treatment complications.