Abstract
BACKGROUNDS: The patient-generated subjective global assessment (PG-SGA) is one of the screening criteria for malnutrition, the skeletal muscle radiodensity (SMD) and skeletal muscle mass index (SMI) are associated with survival in colorectal cancer patients. Body composition parameters can be easily assessed; however, few studies have examined the association between total muscle wasting scores in PG-SGA and body composition parameters and two muscle abnormalities. METHODS: This cohort study included 1,637 stage I-III CRC patients from 2 clinical centers in China, who were enrolled in the training cohort (n = 1,005) and validation cohort (n = 632). Baseline data were collected prospectively from patients including age, BMI, staging, gait speed, hand grip strength (HGS), peak expiratory flow (PEF), neutrophil-lymphocyte ratio (NLR), intermuscular adipose tissue (IMAT), visceral fat area (VFA) and total muscle wasting score in PG-SGA. Relevant risk factors were subjected to logistic regression analysis and Cox regression analysis to identify characteristics associated with muscle abnormalities and survival. Based on the logistic model results, normograms were established to predict muscle abnormalities, and its discrimination and calibration were assessed using the receiver operating characteristic (ROC) curve and calibration curve. The Kaplan-Meier curves were used to assess the survival of colorectal cancer patients with malnutrition or sarcopenia in an inflammatory state (assessed by NLR). RESULTS: The mean age of all participants was 57.7 ± 10.6 years (56.9% males) and the prevalence of low SMD and low SMI was 32.2 and 39.5%, respectively. Low SMD rate was significantly associated with age, TNM stage, BMI, IMAT, walking speed, total muscle wasting score and NRS2002 score by logistic regression analysis (p < 0.05). Low SMI rate was significantly correlated with age, NLR, BMI, PEF, handgrip strength, calf circumference, walking speed, total muscle wasting score and NRS2002 score (p < 0.05). The AUCs of the diagnostic nomograms were 0.859 (95% CI, 0.831-0.886) for low SMD and 0.843 (95% CI, 0.813-0.871) for low SMI in the validation cohort. We also found that patients with colorectal cancer with malnutrition or sarcopenia had a worse prognosis when NLR ≥3.5. CONCLUSION: Muscle abnormalities and malnutrition are strongly associated with mortality in patients with non-metastatic colorectal cancer. Early identification and intervention of the associated risk factors may offer new ways to improve patient prognosis.