Cancer outcomes in women without upfront surgery for ductal carcinoma in situ: observational cohort study

Ryser MD, Thomas SM, Li Y, et al. BMJ. 2025 Jul 8;390

This US observational cohort study quantified the risk of invasive cancer in women with ductal carcinoma in situ (DCIS) managed without upfront surgery. Women with needle biopsy–proven, pure DCIS diagnosed in 2008–2015 who were alive and free of invasive cancer at 6 months and had received no breast surgery or radiotherapy in that period were included (n=1780, 751 sites). Median age was 63; 86% of lesions were screen-detected, 59% low/intermediate grade, and 75% hormone receptor–positive. Thirty-eight per cent ultimately had ipsilateral surgery and 25% received endocrine therapy.

The primary endpoint was ipsilateral invasive breast cancer (iIBC), defined to include ipsilateral nodal or distant metastasis or breast cancer death without prior invasive event. Secondary endpoints were breast cancer–specific survival and other-cause mortality. Time-to-event analyses from the 6-month landmark used Kaplan–Meier methods; competing risks methods estimated cumulative incidence of other-cause death, and univariable Cox models examined associations with clinicopathological factors and endocrine therapy. A “low-risk” group mirrored active-surveillance trials (age ≥40, screen-detected, grade I/II, HR-positive); all others were “high-risk”.

Over median 53.3 months’ follow-up, 115 women (6.5%) developed iIBC and 29 (1.6%) died of breast cancer. Cumulative iIBC incidence was 2.6%, 7.2% and 10.7% at 2, 5 and 8 years, respectively; corresponding 8-year risks were 8.5% in low-risk versus 13.9% in high-risk DCIS (P<0.001). Eight-year breast cancer–specific survival was 96.4% overall (98.1% low-risk; 94.4% high-risk). Other-cause mortality reached 27.8%, exceeding age-matched expectations (19%). Endocrine therapy showed no consistent association with iIBC in time-dependent sensitivity analyses.

The authors note limitations related to registry-based data abstraction, potential misclassification and confounding by indication in this highly selected, older, more comorbid cohort. Nonetheless, they argue that the observed invasive risk approximates that of atypical hyperplasia and supports considering surveillance, particularly for low-risk DCIS and in older or comorbid women, with decisions guided by risk stratification and shared decision-making while randomised trial data mature.

Summary author: Alona Courtney, Clinical Research Fellow Princess Grace Hospital, Specialist Registrar in General surgery Health Education North West London