W K de Roos1, P Kaye, D M Dent. 1. Department of Surgery, Groote Schuur Hospital and University of Cape Town, South Africa.
Abstract
BACKGROUND: Local recurrence and death from metastases are occasional, but consistent, themes in reports of patients with phyllodes tumours. Factors that might contribute to these outcomes were sought. METHODS: Data from 38 patients with a phyllodes breast tumour were reviewed retrospectively, reclassifying the pathological material using the Pietruszka and Barnes criteria. RESULTS: At a median of 12 months, nine patients had developed a local recurrence and four had died from metastases. Following local excision in 24 patients (for diagnosis in 13, for 'fibroadenoma' in nine and for phyllodes tumour in two patients), 13 had no further surgery and five had local recurrence (three of eight benign tumours, two of two malignant tumours). Wide local excision or mastectomy in 18 patients was followed by four recurrences (one of eight borderline tumours, three of ten malignant tumours). All patients with recurrence had margin involvement on histological examination, but not all patients with margin involvement developed recurrence. Lack of statistical correlation between local recurrence and age, delay, size, grade or type of surgery was confounded by selection bias for more extensive surgery for malignant tumours. Death correlated with size (P = 0.05) and grade (P = 0.03) of tumour. CONCLUSION: Inadequate preoperative diagnosis ('fibroadenoma' or failure of triple assessment) frequently led to local excision with positive margins. Without revision this often resulted in local recurrence. Local recurrence of any grade was usually followed by further recurrence. Death was related to tumour size and histological grade, confirming these as prognostic factors.
BACKGROUND: Local recurrence and death from metastases are occasional, but consistent, themes in reports of patients with phyllodes tumours. Factors that might contribute to these outcomes were sought. METHODS: Data from 38 patients with a phyllodesbreast tumour were reviewed retrospectively, reclassifying the pathological material using the Pietruszka and Barnes criteria. RESULTS: At a median of 12 months, nine patients had developed a local recurrence and four had died from metastases. Following local excision in 24 patients (for diagnosis in 13, for 'fibroadenoma' in nine and for phyllodestumour in two patients), 13 had no further surgery and five had local recurrence (three of eight benign tumours, two of two malignant tumours). Wide local excision or mastectomy in 18 patients was followed by four recurrences (one of eight borderline tumours, three of ten malignant tumours). All patients with recurrence had margin involvement on histological examination, but not all patients with margin involvement developed recurrence. Lack of statistical correlation between local recurrence and age, delay, size, grade or type of surgery was confounded by selection bias for more extensive surgery for malignant tumours. Death correlated with size (P = 0.05) and grade (P = 0.03) of tumour. CONCLUSION: Inadequate preoperative diagnosis ('fibroadenoma' or failure of triple assessment) frequently led to local excision with positive margins. Without revision this often resulted in local recurrence. Local recurrence of any grade was usually followed by further recurrence. Death was related to tumour size and histological grade, confirming these as prognostic factors.