Breast conservation or mastectomy

Four randomised trials settled this and all four reported at twenty years or more. NSABP B-06 randomised 1,851 women three ways and found no significant difference in disease-free, distant-disease-free or overall survival between total mastectomy, lumpectomy alone and lumpectomy with irradiation; the hazard ratio for death after lumpectomy with irradiation against mastectomy was 0.97 (95 percent confidence interval 0.83 to 1.14). Milan I randomised 701 women to Halsted radical mastectomy or quadrantectomy with radiotherapy and found death from any cause at twenty years of 41.2 and 41.7 percent (p=1.0). EORTC 10801 randomised 868 patients including tumours up to 5 cm and found time to death at a median 22.1 years no different (hazard ratio 1.11, 0.94 to 1.33). NSABP B-04, at twenty-five years, found no advantage to radical mastectomy over lesser operations at all.

What does differ is the chance of the cancer returning in the breast that was kept: 14.3 percent at twenty years in NSABP B-06 with radiotherapy, against 39.2 percent without it, and 8.8 against 2.3 percent between quadrantectomy and radical mastectomy in Milan I. A recurrence in a conserved breast is usually salvageable by mastectomy, which is part of why survival is unaffected: in the EORTC boost trial, mastectomy was the first salvage treatment for 279 of 354 such recurrences (79 percent) in one arm and 178 of 237 (75 percent) in the other.

Mastectomy is still the right operation for some: a tumour too large for the breast to give a good result, several tumours in different quadrants, inflammatory breast cancer, a margin that cannot be cleared by re-excision, a woman who cannot have radiotherapy, and a woman who has weighed the options and wants the breast off. Risk-reducing mastectomy in a BRCA1 or BRCA2 carrier is a different operation for a different reason.

In England NICE NG101 recommendation 1.4.3 offers further surgery when tumour is at the inked margin, and recommendation 1.4.5 considers it when tumour cells lie within 1 mm of the radial margin, a threshold the guideline lowered from 2 mm in 2024 to reduce re-operations.