Proposing New Criteria for Classification of Benign Fibroepithelial Lesions Based on Clinicopathologic Evaluations of 507 Cases with Clinical Outcome.

Cellular fibroadenoma (cFA), benign phyllodes tumor (BePT), and borderline phyllodes tumor (BoPT) demonstrate overlapping clinicopathologic features. We evaluated 507 cases including 91 cFAs, 230 BePTs, and 186 BoPTs. The World Health Organization-recommended morphologic criteria were used to classify cFA and BePT. BoPT was diagnosed when a tumor presented at least 1 of the 5 morphologic features (permeative border, markedly increased stromal cellularity, marked stromal atypia, ≥5 mitoses/10 high-power fields, and the presence of stromal overgrowth) but did not meet our recently proposed refined diagnostic criteria for malignant phyllodes tumor. The median follow-up time was 79 months. Overall, 23 (4.5%) of the 507 cases developed recurrence. None had distant metastasis or recurred as malignant phyllodes tumor. The recurrence rates were 7.7%, 2.2%, and 5.9% in cFA, BePT, and BoPT, respectively. Of the 507 cases, younger age, larger tumor size, the presence of myxoid stroma, and stromal overgrowth were significantly associated with higher recurrence rate in univariate analysis (all P < .05); in multivariate analysis, younger age (years) and larger tumor size were significantly associated with recurrence, whereas the other 2 variables were marginally associated with recurrence. Tumors > 6.5 cm had a recurrence rate of 13.5% vs 3.5% in tumors ≤ 6.5 cm; tumors with stromal overgrowth had a recurrence rate of 15% vs 4.1% in tumors without stromal overgrowth. In cFA, being of Hispanic race and the presence of myxoid stroma were significantly associated with recurrence; in BePT, increased number of leaf-like structure and the presence of myxoid stroma were significantly associated with recurrence; and in BoPT, younger age (years) and large tumor size were significantly associated with recurrence. Surgical margin status (positive vs negative) was not associated with recurrence in cFAs, BePTs, or BoPTs. The recurrence rate in benign fibroepithelial lesions was low. No metastasis or recurrence as malignant phyllodes tumor was observed. Surgical margin status was not associated with recurrence rate. These results indicate that the current classification does not correlate with clinical outcomes. We propose to use tumor size of 6.5 cm and/or the presence of stromal overgrowth as the diagnostic criteria to classify these lesions as benign and borderline fibroepithelial lesions because they best correlate with clinical outcome. The significance of myxoid stroma warrants further investigation.