2/neu overexpression were not associated with pCR. The presence of ALI in the core biopsy, post-therapy excision, or both was associated with axillary lymph node metastases (P = 0.0062, P = 0.0249, and P = 0.0021, respectively). Although preliminary, our study suggests that the presence of tumor necrosis and ALI in the core biopsy may be important features to be included in the standard report. The pCR of a primary breast can- cer to chemotherapy, though an impor- tant prognostic factor for survival, is observed only in a small proportion of patients. The ability to identify histo- logic and molecular characteristics of breast cancer that predict a pCR to primary chemotherapy would be im- portant in designing specific therapy regimens in clinical settings. Many studies, including this one by Pu and colleagues, have investigated histopathologic and molecular features of breast cancer in samples of the ini- tial tumor, collected by core biopsy or fine-needle aspiration, before chemo- therapy. Many of these studies have shown that hormone receptor status, histologic type and grade, proliferation index, and apoptosis are useful for pre- dicting response to therapy. However, most such studies, including this one, had only small numbers of cases and evaluated only a few markers. Pu and colleagues found tumor necrosis in the biopsy sample to be a significant pre- dictive factor. Two main issues must be considered in attempts to use a single pathologic parameter to make decisions about ther- apy in clinical settings. The first issue is tumor heterogeneity, which is a well rec- ognized characteristic of many solid tu- mors, especially breast cancer. A single histopathologic finding, such as tumor necrosis identified in a prechemotherapy sample, may or may not be representa- tive of the entire tumor. The second issue is the need to standardize evaluations of histopathologic features in small sam- ples; no such criterion has been estab- lished for tumor necrosis. In this study, tumor necrosis was defined as the pres- ence of groups of cells with pyknosis, karyorrhexis, and eosinophilic debris, a definition that is not very specific. The reproducibility of this feature as as- sessed by different pathologists would require large multicenter studies. Many of the histologic features of breast can- cer such as proliferation, grade, and tu- mor necrosis are related to each other, and they most likely reflect tumor differ- entiation. Molecular profiling is perhaps the most promising technology for evalu- ating multiple markers related to tumor differentiation. A. A. Sahin, MD Breast Diseases: A Year Book ® Quarterly 57 Vol 17 No 1 2006 57 ANNOTATED BIBLIOGRAPHY Accuracy of Intraoperative Frozen-Section Analysis of Breast Cancer Lumpectomy-Bed Margins Cendán JC, Coco D, Copeland EM III J Am Coll Surg 201:194-198, 2005 At present across the United States, roughly 40% of patients who have undergone lumpectomy for breast cancer have to return to the operating room for positive or close surgical margins, as would have been the case in this study if intraoperative frozen-section evaluation of margins had not been used. Cendan and colleagues found that the need for a second operation was avoided in nearly half of the patients who had positive margins on final pathologic analysis. Careful evalu- ation of surgical margins, whether by palpation, frozen-section analy- sis, touch preparation cytology, 1 or intraoperative sonography, 2 is im- portant for optimal results after conservative breast surgery. 3 V. S. Klimberg, MD References 1. Klimberg VS, Westbrook KC, Korourian S: Use of touch preps for diagnosis and evaluation of surgical margins in breast can- cer. Ann Surg Oncol 5:220-226, 1998. 2. Henry-Tillman R, Johnson AT, Smith LR, et al: Intraoperative ultrasound and other techniques to achieve negative margins. Semin Surg Oncol 20:206-213, 2001. 3. Pinotti JA, Carvalho FM: Intraoperative pathological monitor- ization of surgical margins: A method to reduce recurrences af- ter conservative treatment for breast cancer. Eur J Gynaecol Oncol 23:11-16, 2002. Fibroepithelial Lesions With Cellular Stroma on Breast Core Needle Biopsy: Are There Predictors of Outcome on Surgical Excision? Jacobs TW, Chen Y-Y, Guinee DG Jr, et al Am J Clin Pathol 124:342-354, 2005 Fibroepithelial lesions in core biopsy specimens are the most prob- lematic diagnostically when a moderate degree of stromal hypercellu- larity is present. Jacobs and colleagues retrospectively evaluated the pathologic features of fibroepithelial lesions on core biopsy and cor- related the results with histologic findings in the corresponding surgi- cal excision specimens, with particular emphasis on those lesions with moderate stromal hypercellularity in the core biopsy specimen.