The Certainties and the Uncertainties of Ductal Carcinoma In Situ

Ductal carcinoma in situ (DCIS) was an uncommon problem before the use of screening mammography, but it is estimated that 55 700 new cases of DCIS were diagnosed in the United States in 2003 (1). Our lack of understanding of the natural history of DCIS, coupled with treatment options ranging from excision alone to mastectomy plus tamoxifen therapy, has resulted in confusion among both women and their physicians regarding the selection of an appropriate management strategy. This is reflected in the study by Baxter et al. (2), in this issue of the Journal, who documented that between 1992 and 1999 the use of both mastectomy and axillary surgery for women with DCIS decreased, while treatment with excision alone remained common throughout the study period. The authors concluded that their study documents both overtreatment (mastectomy) and undertreatment (failure to use radiation) and that care could be improved with the use of standards. What information exists to develop standards for treating DCIS? First, we know that, even with the availability of large-core biopsy needles and vacuum-assisted biopsy devices, 11%–20% of mammographic lesions diagnosed as DCIS will be found to contain invasive cancer when completely excised (3,4). Complete excision is clearly a management requirement to be certain that only DCIS is present, but it may necessitate mastectomy. Although DCIS is thought of as early-stage cancer, DCIS lesions are often quite large. In a study of mastectomy specimens containing DCIS (5), 46% of the lesions were larger than 3 cm in size, even though 86% were mammographically detected. At present, no data are available from population-based studies to indicate how often mastectomy is medically necessary in DCIS, but a report from a single institution (6) suggests that mastectomy is medically indicated statistically significantly more frequently in DCIS than in stage I invasive breast cancer, primarily because of the large size of many of the lesions. Information from a large population-based dataset on the medical necessity of mastectomy would provide valuable insight into the issue of overtreatment. Another certainty is that pure DCIS does not metastasize to the axillary lymph nodes, or elsewhere, so that axillary dissection should not be a part of surgical therapy. This fact appears to be widely recognized, as shown by the 8.5% incidence of dissection reported by Baxter et al. (2) in women treated with breast-conserving surgery. The higher rate of axillary surgery (although it is debatable whether the removal of six lymph nodes during a mastectomy constitutes planned axillary surgery) seen in women undergoing mastectomy reflects recognition of the substantial risk of undiagnosed invasive carcinoma in women with large DCIS lesions (3,4) and is consistent with practice standards during the period studied (7). The increasing use of sentinel node biopsy for axillary staging in women with DCIS undergoing mastectomy should make this a moot point in the future. The other certainty in DCIS is that the cause-specific survival is extremely high. In the National Surgical Adjuvant Breast and Bowel Project (NSABP) Protocol B-17 study (8) of breast-conserving surgery with and without radiation, only 14 deaths (1.6%) attributable to breast cancer had occurred in 814 women after a mean follow-up of 8 years. A retrospective multi-institutional study (9) of women with DCIS treated with excision and irradiation reported a 15-year actuarial cause-specific survival of 96%. The extremely favorable survival in DCIS, regardless of the type of local therapy, has led some to advocate a minimalist approach to treatment in the majority of cases. But, although survival rates are high, the risk of recurrence varies considerably for mastectomy, excision and radiation, and excision alone, ranging from 1% to 2% at 10 years after mastectomy to 32% at 12 years after excision alone in NSABP B-17 (10,11). This wide variation in recurrence rates is the crux of the uncertainty in DCIS. There are two areas, at opposite ends of the research spectrum, where our limited understanding prevents us from developing meaningful therapeutic guidelines. The first of these is our inability to identify which DCIS lesions will progress to invasive carcinoma, and in what time interval. Conventional prognostic factors, such as patient age and tumor grade, subtype, and size, provide information on the time course of local recurrence and the magnitude of risk reduction achieved with radiotherapy, but these factors do not identify those women who will have a disease recurrence with potentially life-threatening invasive cancer. Efforts to identify a molecular signature for DCIS lesions that will recur as invasive carcinoma are of enormous interest, but such efforts have been limited by the lack of tissue for study and the need for a very large population with long-term follow-up to generate an adequate number of invasive recurrences (12,13). In the absence of data that allow us to identify which women with DCIS are at risk for invasive recurrence, how should therapy be selected? Treatment in DCIS is more properly considered the prevention of invasive carcinoma. As with any prevention intervention, the individual’s values, desires, and perceptions of what constitutes an acceptable level of risk should be the primary determinants of the prevention strategy used. Evidence suggests that the critical difference between the prevention goal of DCIS treatment and the therapeutic goal of invasive cancer management is not well appreciated by women with the disease. Rakovitch et al. (14) compared the perceptions of women diagnosed with DCIS with those of women diagnosed with stage T1 or T2 N0 breast cancer. Both groups had similar levels of anxiety and depression, and women with DCIS estimated their risk of dying of breast cancer to be 27%, identical to the estimate of the women with invasive cancers. De Morgan et al. (15) reported a high level of confusion among women diagnosed with DCIS regarding the nature of the disease and dissatisfaction with the information that they received during the decision-making process. In the absence of this understanding, rational decision making regarding therapy is difficult. Katz et al. (16) surveyed 183 Detroit-area women treated for DCIS or invasive cancer. No correlation between the patient’s preferences and attitudes and the surgeon’s treatment recommendation was observed. However, women concerned about recurrence reported that they chose to undergo mastectomy, challenging the traditional paradigm that high rates of mastectomy are attributable to surgeon recommendation. Until the genetic changes that govern the progression from in situ to invasive disease are identified, a better understanding of how to communicate what is and is not known about the nature of DCIS and a better understanding of patient preferences and factors that influence the decision-making process are critical to helping women make decisions that meet their needs. Patients who perceive that they have a choice of therapy and the ability to participate in the decision-making process are more satisfied with the outcome of care, regardless of the treatment received (16). Baxter et al. (2) have documented various patterns of care for DCIS that they termed over- and undertreatment. Because variations in care were observed on the basis of age, race, and geographic location, the authors conclude that they most likely reflect physician practice patterns. They may equally reflect variations in attitudes toward risk among different groups of women. The use of breast implants for augmentation varies on the basis of age and geographic region, from 1.64 per 1000 women in the East to 16.4 per 1000 women in Texas (17). Is this undertreatment in the East and overtreatment in Texas, or is it a reflection of the diversity in preferences among American women? The ability to predict which women with DCIS will develop invasive cancer will ultimately solve the dilemma of DCIS. For the present, understanding who makes treatment decisions and why is likely to do more to improve the care of women with DCIS than any treatment guidelines.