- Design
- Retrospective comparative clinicopathological study
- Population
- 58 borderline or malignant phyllodes tumours of the breast
- Primary outcome
- Recurrence and metastasis by classification system
- Effect
- Mitoses 30 or more per 10 HPF (P = 0.005) and malignant heterologous elements (P < 0.001) associated with metastasis
Researchers re-reviewed 62 phyllodes tumours originally diagnosed as borderline or malignant between 2015 and 2025, reclassifying four as benign and analysing 58 (42 borderline, 16 malignant). Each was classified under WHO 2019 criteria and two alternative systems. Median follow-up was about 32 months.
Six borderline tumours recurred locally without metastasis. Seven malignant tumours metastasised, with two deaths. A stromal mitotic count of 30 or more per 10 high-power fields and malignant heterologous elements were independently associated with metastasis. One alternative system agreed with WHO 2019 in 97% of cases; the other reclassified 52% of borderline tumours as malignant, but none of those went on to metastasise.
The series is small and follow-up short, so it cannot settle classification. It supports staying with WHO criteria for now, and reinforces that the two features carrying most weight, mitotic activity and heterologous elements, are both sampling-dependent.
- Sample phyllodes tumours generously; mitotic hotspots and heterologous elements can be missed with limited blocks.
- Report the stromal mitotic count per 10 high-power fields explicitly, noting the field area used.
- Highlight malignant heterologous elements prominently in the report.
- Be cautious about adopting classification systems that upgrade many borderline tumours to malignant without outcome evidence.
Why it matters
Upgrading half of borderline tumours to malignant would change surgery and follow-up for many women without, on this evidence, predicting who metastasises.
Don't overread it
With 58 tumours and about 32 months of follow-up, absence of metastasis among reclassified tumours is weak evidence.
The statistics, in plain English
P values of 0.005 and below 0.001 indicate the associations with metastasis are unlikely to be chance, but with only seven metastases the estimates of how strongly each feature predicts spread are imprecise.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for pathology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free