Abstract
PURPOSE: The aim of the study was to assess sentinel lymph node biopsy (SLNB) results in colon cancer (CC) regarding basic anthropometric features of the studied population and their derivatives calculated using mathematical formulas. METHODS: One hundred three SLNBs in CC have been analysed. Various indicators were calculated for every patient using mathematical formulas: BMI, Roher's index, lean body weight, body fat percentage and body weight/ideal body weight for a given height ratios using the following formulas: Broca's, Broca's ideal weight, Broca-Brugsch, Lorenz's, Potton's, Devine's, Robinson's, Miller's and Hamwi. The results were compared with accuracy, sensitivity and false negative results percentage by means of ROC curves and the test for structure indicators (for determined cut-off points). RESULTS: No statistically significant relationship between the results and patients' sex or age were found. ROC curve analysis did not reveal statistically significant relationships between the obtained results and indicators calculated on the basis of growth and weigh (all p > 0.05). The analyses of sensitivity and accuracy with determined cut-off point, in spite of differences amounting to 19 % (analysis of lean body weight/weight ratio), showed no statistical significance for any of the relationships (all p > 0.05). CONCLUSIONS: No indicator with high diagnostic and prognostic value has been found. The problem of qualifying patients for SLNB in CC in regard of the anthropometric features of the population and body composition assessment formulas remains open and requires further analysis on larger populations.