Adverse surgical outcomes in screen-detected ductal carcinoma in situ of the breast

Background The Sloane Project is the largest prospective audit of ductal carcinoma in situ (DCIS) worldwide, with over 12,000 patients registered between 2003 and 2012, accounting for 50% of screen-detected DCIS diagnosed in the United Kingdom (UK) over the period of accrual. Methods Complete multidisciplinary data from 8313 patients with screen-detected DCIS were analysed for surgical outcome in relation to key radiological and pathological parameters for the cohort and also by hospital of treatment. Adverse surgical outcomes were defined as either failed breast conservation surgery (BCS) or mastectomy for small lesions (<20 mm) (MFSL). Inter-hospital variation was analysed by grouping hospitals into high, medium and low frequency subgroups for these two adverse outcomes. Results Patients with failed BCS or MFSL together accounted for 49% of all mastectomies. Of 6633 patients embarking on BCS, 799 (12.0%) required mastectomy. MFSL accounted for 510 (21%) of 2479 mastectomy patients. Failed BCS was associated with significant radiological under-estimation of disease extent and MFSL significant radiological over-estimation of disease extent. There was considerable and significant inter-hospital variation in failed BCS (range 3-32%) and MFSL (0-60%) of a hospital's BCS/mastectomy workload respectively. Conversely, there were no differences between the key radiological and pathological parameters in high, medium and low frequency adverse-outcome hospitals. Conclusions This evidence suggests significant practice variation, not patient factors, is responsible for these adverse surgical outcomes in screen-detected DCIS. The Sloane Project provides an evidence base for future practice benchmarking.