Standard curative schedules of radiotherapy to the breast deliver 25 fractions of 2·0 Gy over 5 weeks. In a randomised trial, we tested whether fewer, larger fractions were at least as safe and as effective as standard regimens. In this analysis, we assessed the long-term results of tumour control in the same population.

In 1986–98, we randomly assigned 1410 women with invasive breast cancer (tumour stage 1–3 with a maximum of one positive node and no metastasis) who had had local tumour excision of early stage breast cancer to receive 50 Gy radiotherapy given in 25 fractions, 39 Gy given in 13 fractions, or 42·9 Gy given in 13 fractions, all given over 5 weeks. The primary endpoint was late change in breast appearance, which has been reported elsewhere. Here, we report ipsilateral tumour relapse, one of the secondary endpoints. Relapse was defined as any appearance of cancer in the irradiated breast. Analysis was by intention to treat.

After a median follow-up of 9·7 years (IQR 7·8–11·8) for the 838 (95%) patients who survived, the risk of ipsilateral tumour relapse after 10 years was 12·1% (95% CI 8·8–15·5) in the 50 Gy group, 14·8% (11·2–18·3) in the 39 Gy group, and 9·6% (6·7–12·6) in the 42·9 Gy group (difference between 39 Gy and 42·9 Gy groups, χ² test, p=0·027). The sensitivity of breast cancer to dose per fraction was estimated to be 4·0 Gy (95% CI 1·0–7·8), similar to that estimated for the late adverse effects in healthy tissue from breast radiotherapy.

Breast cancer tissue is probably just as sensitive to fraction size as dose-limiting healthy tissues. If this finding is confirmed, radiotherapy schedules can be greatly simplified by the delivery of fewer, larger fractions without compromising effectiveness or safety, and possibly improving both.