Background: Memorial Sloan Kettering Cancer Center (MSKCC) and MD Anderson Cancer Center (MDACC) have established nomograms to predict sentinel node positivity. We propose the addition of two novel variables-distance of tumor from the nipple and from the skin-can improve their performance.
Methods: Ultrasounds of clinical T1/T2 tumors were reviewed. Distances of the tumor from the skin and from the nipple were measured. MSKCC and MDACC nomogram predictions and the AUC-ROC for each model were calculated. The added utility of the two variables was then examined using multiple logistic regression.
Results: Of 401 cancers studied, 79 (19.7 %) were node positive. The mean distance of tumors from the nipple in node-positive patients was 4.9 cm compared with 6.0 cm in node-negative patients (p = 0.0007). The mean distance of tumors from the skin was closer in node-positive cases (0.8 cm) versus node-negative cases (1.0 cm, p = 0.0007). The MSKCC and MDACC nomograms AUC-ROC values were 0.71 (95 % CI 0.64-0.77) and 0.74 (95 % CI 0.68-0.81). When adjusted for the MSKCC predicted probability, addition of both distance from nipple (p = 0.008) and distance from skin (p = 0.02) contributed significantly to prediction of nodal positivity and improved the AUC-ROC to 0.75 (95 % CI 0.70-0.81). Similarly, distance from nipple (p = 0.002), but not distance from skin (p = 0.09), added modestly to the MDACC nomogram performance (AUC 0.77; 95 % CI 0.71-0.83).
Conclusions: Distance of tumor from the nipple and from the skin are important variables associated with nodal positivity. Adding these to established nomograms improves prediction of nodal positivity.