Stromal necrosis as the dominant prognostic histologic feature in malignant phyllodes tumors of the breast: a single-center retrospective study with observations on stromal overgrowth

Malignant phyllodes tumor (MPT) of the breast is a rare fibroepithelial neoplasm with aggressive clinical behavior, for which well‑validated prognostic markers remain limited. Stromal overgrowth has been widely recognized as a hallmark of malignancy; by contrast, the independent prognostic role of stromal necrosis is less well‑characterized. This retrospective single‑center study aimed to compare the prognostic value of stromal necrosis and stromal overgrowth in MPT, and to explore whether stromal necrosis alone, instead of a combined‑feature model, predominantly drives the association with adverse patient outcomes. A retrospective review of 23 patients with histologically confirmed MPT was performed. Clinicopathological and histopathological data were collected for analysis. Survival analyses were performed using Kaplan-Meier curves with the log‑rank test and univariate Cox proportional‑hazards regression. Patients were stratified based on the presence or absence of stromal necrosis and stromal overgrowth. The prognostic value of stromal necrosis was primarily evaluated by comparing necrosis‑positive versus necrosis‑negative subgroups. The prognostic significance of stromal overgrowth was assessed as a secondary exploratory comparison. Stromal overgrowth was present in 11 patients (47.8%) and stromal necrosis in 13 (56.5%); all overgrowth‑positive cases also exhibited stromal necrosis. Patients with stromal necrosis had significantly worse disease‑free survival (DFS) and overall survival (OS) compared with necrosis‑negative patients (log‑rank P = 0.004 and P = 0.005, respectively). In univariate Cox regression, stromal necrosis was associated with significantly elevated risks of recurrence (HR = 7.19, 95% CI 1.52–34.06, P = 0.013) and death (HR = 11.59, 95% CI 1.39–96.68, P = 0.024). In our cohort, stromal overgrowth showed a weaker prognostic association: it reached statistical significance for OS (HR = 5.98, P = 0.030), while the association with DFS was only borderline (HR = 3.26, P = 0.062). Of note, although tumor size (> 8 cm vs. ≤ 8 cm) showed differences in log‑rank testing for both DFS ( P = 0.022) and OS ( P = 0.009), this association did not reach statistical significance in univariate Cox regression analysis. Within this single-center retrospective cohort, stromal necrosis appears to be the dominant histologic prognostic marker for malignant phyllodes tumors. Complete overlap between stromal necrosis and stromal overgrowth in our cohort precluded reliable assessment of the independent prognostic effect of stromal overgrowth. Given the limited sample size of 23 cases, these findings should be interpreted with caution and require validation in larger multicenter cohorts prior to clinical application.