Similarly, of the 1,010 patients enrolled in the EORTC-10853 trial, mammography detected lesions in 71% of the women. At a median follow-up of 10.5 years, the overall rate of in-breast tumor recurrence was reduced from 26% to 15% (P < .001) with a similarly effective reduction of invasive (13% to 8%, P = .065) and noninvasive (14% to 7%, P = .001) recurrence rates.[Level of evidence: 1iiDii] In this analysis, parameters associated with an increased risk of in-breast recurrence included age 40 years or younger, palpable disease, intermediate or poorly differentiated DCIS, cribriform or solid growth pattern, and indeterminate margins. Elsewhere, margins of less than 1 mm have been associated with an unacceptable local recurrence rate, even with radiation therapy. In both of the studies reported here, the effect of radiation therapy was consistent across all assessed risk factors.
Given that lumpectomy and radiation therapy are generally applicable for most patients with DCIS, can a subset of patients be identified with such a low risk of local recurrence that postoperative radiation therapy can be omitted? To identify such a favorable group of patients, several pathologic staging systems have been developed and tested retrospectively, but consensus recommendations have not been achieved.[9,10,11,12]
The Van Nuys Prognostic Index, which combines three predictors of local recurrence (i.e., tumor size, margin width, and pathologic classification), was used to retrospectively analyze 333 patients treated with either excision alone or excision and radiation therapy. Using this prognostic index, patients with favorable lesions, who received surgical excision alone, had a low recurrence rate (i.e., 2% with a median follow-up of 79 months). A subsequent analysis of these data was performed to determine the influence of margin width on local control. Patients whose excised lesions had margin widths 10 mm or larger in every direction had an extremely low probability of local recurrence with surgery alone (4% with a mean follow-up of 8 years). These reviews are retrospective, noncontrolled, and are subject to substantial selection bias. By contrast, no subset of patients was identified in the prospective NSABP trial that did not benefit from the addition of radiation therapy to lumpectomy in the management of DCIS.[2,4,14,15]