Purpose: Percutaneous endoscopic gastrostomy (PEG) placement and dependency is a major morbidity associated with radiotherapy for treatment of head and neck cancer. There is much controversy surrounding the use of prophylactic PEG placement. There are many publications evaluating risk factors for PEG dependency and the role of prophylactic PEG in the general population, but such information in the veteran population is lacking. The purpose of this study is to identify risk factors for long-term PEG dependency (defined as placement with no removal of PEG during follow-up) and assess the need for prophylactic PEG in veterans.
Methods: Between 2011 and 2013, 40 patients underwent radiotherapy to the primary site and regional lymph nodes for squamous cell carcinoma of the head and neck. Medical records were reviewed to collect data, including primary tumor site, stage, age, radiation dose, use of chemotherapy, weight, and body mass index (BMI) at radiation start. Data were collected on timing of PEG placement to evaluate the use of prophylactic PEG and its impact on weight loss during radiotherapy, treatment time, and long-term PEG dependency. Data were also collected on whether patients were seen by Speech Pathology and Nutrition before and during radiotherapy to determine the impact of these visits on reduction of PEG dependency. Data analysis was performed to identify which risk factors are most strongly associated with long-term PEG dependency. The effects of prophylactic PEG placement on long-term PEG dependency, weight loss during treatment, and treatment time were also evaluated.
Results: Patients had primary tumors of the nasopharynx, hypopharynx, larynx, oropharynx, or oral cavity. Thirty patients underwent prophylactic PEG placement, and 10 did not. All patients coded as PEG dependent long-term have not had the PEG removed during follow-up and either died with the PEG in place or have been PEG dependent > 1 year. There was no significant relationship between primary tumor site, stage, age, radiation dose, use of chemotherapy, smoking at consultation, or living alone with long-term PEG dependency. Visits with Speech Pathology and Nutrition prior to radiotherapy and during radiotherapy were not associated with reduced long-term PEG dependency. Factors significantly associated with long-term PEG dependency included prophylactic PEG placement vs no prophylactic PEG placement (63% vs 10%, P = .0084), BMI 22 (75% vs 26%, P = .0038), and weight 170 lb (63% vs 23%, P = .041). Treatment times were not significantly different between patients with prophylactic PEG and those without prophylactic PEG. Patients with prophylactic PEG experienced less weight loss during radiotherapy than those with no prophylactic PEG (median weight loss was 22 lb for no prophylactic PEG vs 8 lb for patients with prophylactic PEG, P = .0067). Of the 10 patients who did not undergo prophylactic PEG, 2 underwent reactive PEG placement.
Conclusions: Within the veteran population, factors significantly associated with long-term PEG dependency include use of prophylactic PEG, low pretreatment BMI, and low pretreatment weight. The use of prophylactic PEG did not improve treatment time but did reduce weight loss during treatment. Though this sample size is small, the differences noted were pronounced and statistically significant. Therefore, prophylactic PEG placement should not be recommended routinely but considered in patients with poor nutritional status at baseline. Veterans with risk factors such as low weight/BMI and prophylactic PEG may require increased monitoring of swallow function.