SP5.7 Prognostic Value of ‘Cachexia Index’ in Patients with Colorectal Cancer Undergoing Surgical Resection

Aim Numerous phenotypical features of cachexia have been associated with poorer outcomes in patients with cancer. However, at present, very few of these markers are routinely considered during the shared decision-making process. This study aims to assess the prognostic value of a novel composite marker of cachexia, termed ‘Cachexia Index’, in patients with colorectal cancer. Method Consecutive patients who underwent resection of a newly diagnosed colorectal cancer, between Jan 2019 and December 2020, were identified from prospectively maintained databases across two sites. Cachexia index was calculated as (L3 skeletal muscle index) x (serum albumin) / (neutrophil-lymphocyte-ratio). The sex-stratified, cut-off values for cachexia index were determined using area under receiver operating characteristic curves (AUC) and Youden’s Index. The primary outcome of interest was overall survival. Results Overall 273 patients (53% male, median 67 years) underwent resection of a colon (n=178) or rectal (n=95) cancer across the study period. Optimum cut-off values for a low cachexia index were defined as <50 for males (AUC=0.792, sensitivity = 0.821, specificity = 0.727) or <35 for females (AUC=0.744, sensitivity = 0.700, specificity = 0.756). Patients with a low cachexia index were older (median 71 vs. 66 years, p=0.029) and had a higher proportion of cT4 (p=0.035) or cN1/2 tumours (p=0.044). Tumour site and ASA grade were comparable between groups. Following adjustment for confounders, using cox-multivariable regression, low cachexia index was associated with decreased overall survival (HR: 3.77, 95% CI: 1.93-7.37, p<0.001). Conclusion Cachexia index may be a clinically useful marker for poor prognosis in patients undergoing surgical resection of a colorectal cancer.