Clinical breast examination and its relevance to diagnosis of palpable breast lesion.

To evaluate if clinical breast examination (CBE) is a reliable tool for diagnosis of palpable breast lesions, from July 2002 to October 2003, 371 patients (445 palpable breast lesions) were examined by single experienced clinician and compared with fine needle aspiration (FNA) results. CBE had sensitivity of 57.14%, specificity 97.11%, positive predictive value (PPV) 76.60%, negative predictive value (NPV) 93.20%, false negative (FN) 0.06%, false positive (FP) 0.02%, and overall accuracy 91.44%. There was concordance in 91.44%. Of a total of 397 benign clinical suspections, 199 episodes were found to be cysts (50.13%). CBE alone even in an experienced clinician is not a reliable tool for diagnosis of palpable breast lesion whether it is malignant or not, but it can be used as a primary and simple tool for benign suspicious palpable breast lesion, particularly cystic breast lesion.

Sentinel node localization in breast cancer using intradermal dye injection: results, influencing factors and learning curve.

To evaluate the identification rate, false negative rate, concordance, negative predictive value of sentinel node localization in breast cancer using intradermal isosulfan blue injection whether this is accurate enough for surgical approach in breast cancer surgery and whether there is a significant learning curve for this technique. Factors affecting the outcomes of the procedure are also determined. From August 2002 to September 2003, 66 cases of stage 0-IIIB operable breast cancer patients underwent sentinel lymph node biopsy before standard breast cancer operation. Overall, identification rate was 80.3%, false negative rate was 10.6%, concordance was 86.8%, negative predictive value was 83.3%, sentinel node was the only node that was positive in 45.5%, and mean operative time was 55.1 minutes. Factors found to lower sentinel node identification rate are neoadjuvant chemotherapy and large tumor (T3-4) while previous excision was not found to affect the identification rate. There is significant learning curve in this technique and this should be performed at least 40-45 cases in the learning phase to accomplish a high identification rate and lower false negative rate before implicating into clinical practice.