Change in tumor cellularity of breast carcinoma after neoadjuvant chemotherapy as a variable in the pathologic assessment of response

BACKGROUND. Complete pathologic response of breast carcinoma to neoadjuvant chemotherapy is a well defined outcome that correlates with prolonged survival. Categorization of incomplete response depends on accurate measurement of residual tumor size but is complicated by the variable histopathologic changes that occur within the tumor bed. In the current study, the authors investigated the contribution of assessing tumor cellularity in the pathologic evaluation of response to chemotherapy. METHODS. The slides from diagnostic core needle biopsy and the subsequent matched resection specimens were examined in 240 patients with breast carcinoma: 120 "treated" patients who received neoadjuvant chemotherapy and 120 "control" patients who received primary surgical management within a few weeks of diagnosis. Clinical response and residual tumor size were evaluated in 108 treated patients who completed a clinical trial with paclitaxel and then received combined 5-fluorouracil, doxorubicin, and cyclophosphamide chemotherapy. Tumor cellularity was assessed from hematoxylin and eosin-stained tissue sections as the percentage of tumor area that contained invasive carcinoma. RESULTS. After neoadjuvant chemotherapy, tumor cellularity decreased from a median of 40% in core needle biopsy to 10% in resection specimens (P < 0.01; Wilcoxon signed rank test). The cellularity of core needle biopsy (median, 30%) tended to underestimate the cellularity of resection specimens (median, 40%) in the control group (P < 0.01). Changes in cellularity varied within each clinical response category, particularly partial response and minor response. The greatest reduction was observed in the cellularity of residual primary tumors that measured less than or equal to 1 cm (pathologic T1a [pT1a] and pT1b tumors), but changes in cellularity varied in the pT1, pT2, and pT3 residual tumor categories. The shape of the distribution of tumor size, expressed as the greatest dimension in cm, was similar in the control group and the treatment group (excluding complete pathologic response); however, when residual tumor size and cellularity were combined, the distribution of pathologic response shifted left (toward complete response) with a steep decline, suggesting that many tumors had a large reduction in cellularity but little change in the tumor size. CONCLUSIONS. Cellularity of the tumor mass was reduced significantly by neoadjuvant chemotherapy, and the change varied widely in different categories of clinical response. Although residual tumors measuring less than or equal to 1 cm in greatest dimension had the most reduction in tumor cellularity, there was broad variability for all residual tumor groups (pT1-pT3). The frequency distribution of residual tumor size was altered markedly by the inclusion of tumor cellularity, indicating that the product of pathologic size and tumor cellularity may provide more accurate pathologic response information than tumor size alone. (C) 2004 American Cancer Society.