However, the significance of the associations of positive CA IX staining with overall and disease-specific mortality disappeared in multivariate analysis (P = 0.2), after correction for positive lymph nodes, tumour size, grade of malignancy, hormone receptor and HER2 receptor status, menopausal status/systemic treatment, and randomization, which were all independent prognostic markers.
Overall size of the malignancy measured at both mammography and ultrasound correlated with histological tumour size (r = 0.54 and r = 0.56, respectively).
However, a preliminary study by Demartini et al. [ 31] with 15 patients demonstrated CAD tumor sizes are less accurate in predicting the size of residual malignancy than those measured by the radiologist.
Nodule size is one of the most important characteristics, as good evidence exists for a strong correlation between size and risk of malignancy.
For this study, size and type of malignancy were gathered in the database.
Most investigators revealed that HER-3 protein expression on breast carcinomas was associated with poor prognostic factors such as large tumor size, high grade of malignancy and lymph node metastasis[ 22, 23].
Estrogen receptor (ER) status, progesterone receptor (PR) status, nodal status, tumor size, and grade of malignancy are the classical parameters used to date by clinicians to narrow down prognoses and weight treatment options [ 2].
Lumen roundness (GAVG) is the only one ranked high by maximal relevance, yet all three features are consistently chosen by the classifier, since they provide different, complementary information on a tissue: greater circularity of lumens and increase in the number of lumens and the size of nuclei indicate malignancy of a tissue.
Evidence has shown that nodular size is not predictive of malignancy risk, and that nodular morphology, rather, is the more important criterion [4, 7, 10, 11, 12, 13].
