A Zero-Cost Bedside Nomogram Integrating TSF and Pain for Predicting Severe Malnutrition in Nasopharyngeal Carcinoma: A Prospective Cohort Study

Shasha Shen¹Kai Zhou²Guoping He³Hui Xu¹Jinzhi Wu¹Xin Tian¹Bi Feng¹Yan Lou¹Hu Ma^3*Xiaoxia Gou^1*

• ¹Head and Neck Oncology, The Second Affiliated Hospital of Zunyi Medical University, Zunyi, China
• ²Abdominal Oncology, The Second Affiliated Hospital of Zunyi Medical University, Zunyi, China
• ³Thoracic Oncology, The Second Affiliated Hospital of Zunyi Medical University, Zunyi, China

Objective: To examine the relationship between nutritional status and quality of life (QoL) in hospitalized patients with nasopharyngeal carcinoma (NPC), identify independent predictors of severe malnutrition, and establish a practical "zero-cost" predictive nomogram to guide early clinical intervention. Methods: This study is a secondary analysis of a prospective cohort from the Investigation on Nutrition Status and Clinical Outcome of Common Cancers (INSCOC) project. We included 216 NPC patients prospectively enrolled at a participating center between January 2014 and December 2019. Clinical data, anthropometrics (including Triceps Skinfold Thickness, TSF), and patient-reported outcomes (NRS 2002 and pain scores) were collected upon admission. Patients were randomly split into a training cohort (n = 152) and an internal validation cohort (n = 64) at a 7:3 ratio. Nutritional status was assessed using the Patient-Generated Subjective Global Assessment (PG-SGA), with severe malnutrition defined as a PG-SGA score ≥ 9. Logistic regression was performed to develop the nomogram. Results: Among the 216 patients (77.8% male; mean age 50.5 ± 10.6 years), 26.9% were at nutritional risk (NRS 2002 ≥ 3), and 23.2% suffered from severe malnutrition. Worsening nutritional status was significantly associated with lower functional domain scores and higher symptom burden in QoL (p < 0.05). Multivariate logistic regression identified NRS 2002 score (OR = 28.52, p < 0.001), Pain score (OR = 2.81, p = 0.014), and TSF (OR = 0.83, p = 0.004) as independent predictors. The nomogram demonstrated excellent discrimination, with an AUC of 0.903 in the training set and 0.825 in the internal validation set. Calibration curves and decision curve analysis (DCA) confirmed the model's clinical utility. Conclusion: Based on the prospective INSCOC dataset, we developed and validated a novel, zero-cost bedside nomogram integrating NRS 2002, TSF, and Pain scores. By bypassing the need for blood tests, this tool enables immediate risk stratification for NPC patients, particularly in resource-limited settings, facilitating timely and personalized nutritional interventions.