Ductal Carcinoma in Situ: Dilemma or Denouement

Abstract

TO THE EDITOR: On the basis of multiple randomized trials for ductal carcinoma in situ (DCIS), it has long been claimed that no low-risk subset defined by favorable clinical or pathologic features could be identified and spared irradiation. The report by Hughes et al successfully identified for the first time a low-risk subset in a prospective trial that was spared radiation therapy. Their results were based on careful mammographic-pathologic correlation and complete sequential tissue processing—methods that are essential to define the size and margin status and to exclude invasion. This type of comprehensive approach has been advocated for some time and was the consensus recommendation of three consecutive image-detected breast cancer consensus conferences and the recently published College of American Pathologists’ guidelines for DCIS. As the accompanying Editorial noted, such methods were not employed by any of the randomized trials. As a result, there were much higher local recurrence rates with or without irradiation and a sizeable number of metastatic first events in B17 and B24, a reflection of unrecognized invasive disease compared with prospective nonrandomized studies. Moreover, this approach was decidedly not part of the prospective study of Wong et al, on which we have previously commented. In our experience, the two therapeutic interventions for which patients have the greatest dread are radiation therapy and axillary lymph node dissection. Relatively few of our patients seek a bilateral mastectomy for DCIS, and most who initially discuss this will choose alternative breast-conserving treatments when they are properly informed. For the lowto intermediate-grade lesions, 25 mm or smaller with 3 mm or more margins, the 7-year rate of IBE was 10.5%, half of which were invasive. In our experience, 10% of patients with invasive recurrences after an initial diagnosis of DCIS die of metastatic breast cancer. That means if 400 patients are not irradiated, we can expect 42 recurrences, 21 of which will be invasive, two of whom will die. If these 400 patients are irradiated, the numbers will be halved to 21 recurrences, 10 of which are invasive, one of whom will die. In other words, we must irradiate 400 patients to save one life. Of the 400 patients, 358 do not benefit from radiation because they would not have recurred. All of them, however, are exposed to the morbidities of radiation therapy. Are we certain that no lives will be shortened because of irradiation? We are not. Long-term data with current techniques do not exist; however, available long-term outcome data of breast irradiation note small increases in mortality related to cardiac disease and smaller risks of lung carcinoma. It hardly seems worth irradiating 400 women to prevent 21 recurrences and one death. Does it not make more sense to carefully follow these patients, re-excise the 42 who recur, and irradiate them at that time, thus sparing 358 the time, expense, and morbidity of breast radiation therapy? Michael D. Lagios The Breast Cancer Consultation Service, Tiburon, CA