A new analysis from the NRG Oncology/NSABP B-35 randomized clinical trial has reopened an important question in the surgical management of ductal carcinoma in situ (DCIS): should close lumpectomy margins automatically lead to reexcision?
The study, published in JAMA Surgery, examined ipsilateral breast tumor recurrence among postmenopausal women with hormone receptor–positive DCIS who underwent lumpectomy, whole-breast irradiation, and adjuvant endocrine therapy. Although narrower margins were associated with slightly higher unadjusted recurrence rates, margin width was not a statistically significant independent predictor after adjustment for other patient and tumor characteristics.
In correspondence addressing the study, Kefah Mokbel, MB BS, MS, FRCS, Professor of Breast Cancer Surgery and Chair of the Multidisciplinary Breast Cancer Program at The London Breast Institute, cautioned that the findings should not yet be interpreted as evidence for broadly relaxing current margin recommendations.
The study, authored by Irene L. Wapnir, MD; Reena S. Cecchini, PhD; James J. Dignam, PhD; and colleagues, was published in JAMA Surgery and examined ipsilateral breast tumor recurrence among postmenopausal women with hormone receptor–positive DCIS who underwent lumpectomy, whole-breast irradiation, and adjuvant endocrine therapy.
What Did NSABP B-35 Show?
Wapnir and colleagues analyzed prospectively collected margin-width data from the NSABP B-35 trial, which enrolled postmenopausal women with hormone receptor–positive DCIS treated with breast-conserving surgery, whole-breast irradiation, and either tamoxifen or anastrozole.
At 10 years, the unadjusted cumulative incidence of ipsilateral breast tumor recurrence was:
- 5.6% among patients with margins smaller than 1 mm, compared with 4.0% among those with margins of at least 1 mm.
- 5.3% among patients with margins smaller than 2 mm, compared with 3.8% among those with margins of at least 2 mm.
However, after adjustment for age, treatment, and tumor-related factors, margin width was not a statistically significant predictor of recurrence. At the 2-mm threshold, the adjusted hazard ratio was 1.33 (95% CI, 0.86–2.06).
The investigators concluded that the absolute differences in recurrence risk were small and suggested that reexcision for margins below 1 mm or 2 mm could be reconsidered in appropriate patients.
Absence of Statistical Significance Is Not Proof of Equivalence
Professor Mokbel’s central concern is that the absence of a statistically significant association should not be interpreted as evidence that narrow margins are clinically equivalent to wider margins.
The adjusted hazard ratios remained above 1 at both thresholds:
- 1.38 for margins below 1 mm
- 1.33 for margins below 2 mm
The confidence intervals were wide and included the possibility of a clinically relevant increase in local recurrence risk. According to Professor Mokbel, this raises the possibility that the analysis may have lacked sufficient precision to detect a meaningful difference.
Only approximately 88 to 90 ipsilateral breast tumor recurrence events informed the multivariable analyses. This limited event number increases the possibility of a type II error, meaning that a potentially important association may not have been detected.
Importantly, NSABP B-35 was not designed as a non-inferiority trial. A non-inferiority analysis would be needed to determine whether accepting a narrower margin produces a recurrence risk that remains within a clinically acceptable limit compared with current standards.
Tumor Size May Still Complicate the Interpretation
Pathological tumor size was the strongest predictor of recurrence in both adjusted models, with hazard ratios of approximately 2.1.
However, Professor Mokbel noted that roughly one-third of tumor sizes were estimated rather than directly measured. Tumor size also differed significantly between the margin groups, with a reported P value below .001.
This matters because an inaccurately measured confounder may not be fully corrected through statistical adjustment. When tumor size is unevenly distributed between patients with narrower and wider margins, residual confounding may influence the apparent association between margin width and recurrence.
Sensitivity analyses that account for uncertainty in tumor-size measurement could help determine how robust the study’s findings are.
Can These Results Be Applied to Every Patient With DCIS?
The NSABP B-35 population was highly specific: postmenopausal women with hormone receptor–positive DCIS who received both whole-breast irradiation and endocrine therapy.
The results should therefore not automatically be extended to:
- Younger women
- Patients with hormone receptor–negative DCIS
- Patients not receiving radiotherapy
- Patients who do not initiate endocrine therapy
- Other treatment settings with different baseline recurrence risks
Professor Mokbel emphasized that the findings are reassuring for selected patients, particularly those receiving comprehensive adjuvant treatment. However, they may not support a universal change in surgical practice.
Reexcision Decisions Remain Individualized
The study adds prospective evidence to a long-standing clinical discussion and may help reduce unnecessary surgery in carefully selected patients. Still, the findings do not establish that margins below 1 mm or 2 mm are non-inferior to wider margins.
For now, decisions about reexcision should continue to consider the full clinical picture, including DCIS extent, imaging-pathology concordance, patient age, tumor biology, planned radiotherapy, endocrine treatment, cosmetic implications, and patient preference.
The key message is not that margin width no longer matters. Rather, in postmenopausal women with hormone receptor–positive DCIS treated with lumpectomy, whole-breast irradiation, and endocrine therapy, the impact of a close negative margin may be smaller than previously assumed, but the evidence may not yet justify relaxing consensus recommendations across all patients.
Written by Nare Hovhannisyan, MD
Read more stories shaping the future of cancer care on OncoDaily