Assessment of Nutritional and Imaging Biomarkers to Predict PEG Tube Placement in Non-HPV-Associated Head and Neck Patients Undergoing Concurrent Chemoradiotherapy

Purpose/Objective(s)

Patients with head and neck cancers undergoing chemoradiotherapy may require percutaneous endoscopic gastrostomy (PEG) tube placement because of dehydration or significant weight loss from their locally-advanced disease or treatment. Modifiable risk factors, such as sarcopenia, may identify high-risk populations who could benefit from intervention. Objective measures of sarcopenia include skeletal muscle index (SMI), skeletal muscle density (SMD), and skeletal muscle gauge (SMG), a novel measure which incorporates both SMI and SMD. We reviewed our institutional experience to assess pretreatment risk factors for nutritional deficiency with respect to serum and skeletal muscle biomarkers.

Materials/Methods

Definitive and post-operative patients undergoing concurrent chemoradiotherapy for locally advanced (AJCC 7^th edition Stage III-IVB) non-human papilloma virus (HPV)-associated head and neck cancer at an academic medical center between 2011 and 2016 were reviewed. Imaging analysis was assessed with three different skeletal muscle measures: SMI, SMD, and SMG. Measurements were obtained from radiation planning computed tomography scans by contouring muscles with an axial slice at the C3 vertebral body. Association between patient factors and PEG tube placement were calculated in univariate and multivariate analyses.

Results

A total 129 patients were included consisting of mostly males (91.5%) with primary tumors located in the larynx/hypopharynx (45.7%) and oropharynx (29.5%). Mean age was 62 years old (range 19-87) with a majority former or current smokers (80.6%) with ECOG performance status 0-1 (83.0%). 70% were treated with definitive chemoradiation and 30% were treated post-operatively. PEG tube was placed in 87/129 (67.4%) patients. Weight loss ≥10% was less common in patients with versus without PEG (30.0% vs 45.2%, p=0.086). Predictive factors on univariate analysis for PEG tube included older age (p=0.01), Caucasian race (p<0.01), former and current smoking status (p<0.01), low SMI (p=0.01), low SMD (p=0.04), deficiency of all 3 skeletal muscle measures (p<0.01), and low pretreatment albumin (p<0.01). On multivariate analysis, factors that persisted included Caucasian race (OR 5.13, p=0.01), current smoking status (OR 6.06, p<0.01), and low pretreatment albumin (OR 4.88, p<0.01). A receiver operating characteristic curve to estimate multivariate model sensitivity and specificity provided an area under the curve of 0.814 (95% CI, 0.738-0.890).

Conclusion

Pretreatment albumin level may serve as a predictive and modifiable pretreatment characteristic of head and neck patients receiving PEG tube placement for treatment. This evaluation identifies patients needing aggressive nutritional guidance, targeted supplementation, and early symptom management as a component of their care.