Background
Prospective, single-arm cohort study (ECOG 5194). 670 patients with DCIS treated with breast-conserving surgery (BCS) alone (no radiation therapy). Enrolled in two cohorts based on pathological features: Cohort 1 (n=565): Grade 1 or 2 DCIS, ≤2.5 cm, margins ≥3 mm; Cohort 2 (n=105): Grade 3 DCIS, ≤1 cm, margins ≥3 mm. Approximately 31% received tamoxifen (optional). This is the 12-year outcomes update of the original ECOG 5194 study.
Interventions and follow up
Arm A: BCS + observation (no RT)
Arm B: None — single-arm study
Primary endpoint: Rate of ipsilateral breast events (IBE = local recurrence, invasive or in situ)
mFollow up: 12 years (median)
Arm B: None — single-arm study
Primary endpoint: Rate of ipsilateral breast events (IBE = local recurrence, invasive or in situ)
mFollow up: 12 years (median)
Results
12-year IBE (Cohort 1): 14.4% (7.5% invasive, 6.9% DCIS)
12-year IBE (Cohort 2): 24.6% (15.1% invasive, 9.5% DCIS)
Tamoxifen use: No statistically significant reduction in IBE from tamoxifen in this study
12-year IBE (Cohort 2): 24.6% (15.1% invasive, 9.5% DCIS)
Tamoxifen use: No statistically significant reduction in IBE from tamoxifen in this study
Adverse events
Main adverse events: Not applicable — no treatment administered beyond BCS. IBE rates represent natural history outcomes with observation alone.
Conclusions
In highly selected DCIS patients — grade 1-2 with small lesions and adequate margins — the 12-year IBE with observation alone was approximately 14%, roughly half of which was invasive cancer. The higher-risk cohort (grade 3, small lesion) had a 24.6% IBE at 12 years, with 15% being invasive, supporting the need for RT in grade 3 DCIS regardless of size.
Key Limitations
Key Limitations: Non-randomized single-arm design — cannot directly compare to RT-treated patients within the same trial. Patient selection is highly restrictive (margins ≥3 mm is not standard practice today). Tamoxifen was not mandated, limiting generalizability. The 14–25% IBE rates may not be acceptable to all patients or clinicians. Long follow-up shows recurrence risk continues to accumulate beyond 10 years.
Clinical Context
ECOG 5194 is frequently cited in discussions of DCIS omission trials alongside RTOG 98-04. The ongoing COMET trial is randomizing low-risk DCIS patients to active surveillance vs standard treatment to provide prospective RCT-level evidence. Currently, omission of RT after BCS for DCIS remains non-standard outside shared decision-making discussions in low-risk patients (grade 1-2, small lesion, wide margins, ER+, older age).
References