Addressing Overtreatment in DCIS: What Should Physicians Do Now?

JNCI: Journal of the National Cancer Institute, Dec 2015

Morrow, Monica, Katz, Steven J.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://academic.oup.com/jnci/article-pdf/107/12/djv290/17986206/djv290.pdf

Addressing Overtreatment in DCIS: What Should Physicians Do Now?

Ductal carcinoma in situ (DCIS) is unique in the growing debate about overtreatment in breast cancer because the marked increase in rates of diagnosis has not reduced the incidence of invasive breast cancer (1) and breast cancer-specific mortality after treatment is extremely low (2). This has fueled concerns that many patients with DCIS receive treatment that is too extensive and that some may not require treatment at all. In this issue of the Journal, Worni et al. (3) examine trends in treatment patterns and mortality using data from the Surveillance, Epidemiology, and End Results (SEER) program on over 121 000 women diagnosed with DCIS between 1991 and 2010. Much of the news here is good. Ten-year disease-specific survival across all locoregional treatment subgroups exceeds 98%. In general, patients have received less extensive surgery over time. Over the 20-year study period, rates of unilateral mastectomy declined from 44.9% of patients to 19.3%, while use of lumpectomy and radiotherapy increased from 24.2% to 46.8%. However, the rate of bilateral mastectomy also increased from 0% to 8.5%. Axillary dissection was largely replaced by sentinel node biopsy, although 15.3% and 2.8% of patients undergoing mastectomy and breast-conserving surgery, respectively, had an axillary dissection in 2010. SEER data are ideal to track treatment trends over time and identify areas where practice does not match evidence. But using SEER data for comparative effectiveness studies is much more challenging. Indeed, it is misleading to draw conclusions from the Worni study (3) about tiny absolute differences in mortality outcomes between treatment groups. Attempts in the study to address baseline differences between treatment groups through propensity score stratification are flawed because of missing data. Even the most basic clinical factors such as histology, tumor size, and receptor status were missing for one-quarter to one-half of the patients and varied markedly across treatment groups. Furthermore, other potentially relevant clinical factors, such as medical comorbidity, extent of disease of DCIS in the breast, and margin status after lumpectomy, were not measured. What do the Worni et al. (3) results suggest about the comparative effectiveness of lumpectomy alone vs lumpectomy with radiation? They suggest that patients who undergo lumpectomy alone have worse survival (Worni et al., Table 4 and Figure 3) (3). This paradox reinforces the nettlesome problem of accounting for treatment selection effects in an observational study design using SEER data: The finding that all-cause mortality is higher in patients who receive lumpectomy alone supports the idea that surgeons are selecting the least extensive treatment for patients at highest risk of death from other causes. More valid approaches have demonstrated the major benefit of radiation in reducing recurrence in DCIS patients (4). The Early Breast Cancer Trialists’ Collaborative Group meta-analysis of randomized trials demonstrates a 50% reduction in the risk of local recurrence with the use of radiation, with 10-year rates of local recurrence in the excision-alone groups ranging from 24% to 30% (5). In a randomized trial examining the efficacy of radiotherapy in low-risk DCIS (low or intermediate grade, <2.5cm in size, clear margins), the use of radiotherapy reduced local recurrence rates at seven years from 6.7% to 0.9% (6). Prospective studies of well-characterized patient populations thought to be at low risk for recurrence based on small tumor size and low to intermediate grade provide additional information to inform discussions with patients. The Eastern Cooperative Oncology Group (ECOG)–American College of Radiology Imaging Network (ACRIN) E5194 study reported a 12-year rate of local recurrence of 14.4% (7.5% invasive) for widely excised low to intermediate DCIS lesions with a median size of 0.6cm and an annual recurrence risk of 1.2% per year, which had not plateaued at 12 years (7). Even for patients selected as low risk with a multigene expression score, 10-year rates of local recurrence were 10.6% (8). Thus, although cause-specific survival rates remain high across all surgical treatments for DCIS, the suggestion that outcomes are equivalent is not supported in randomized trials. Of greatest concern is the misleading assertion by the authors that the excellent outcomes across all treatment options “support implementation of less invasive treatment options including active surveillance in thoughtfully selected patients.” No patients were explicitly selected for active surveillance. Less than 1% of patients in the sample were coded as not receiving surgery or radiation, but many of the patients in this tiny group may have received excisional biopsy with clear margins. We know virtually nothing about the impact of active surveillance on survival. The studies that are used to inform us about the natural history of DCIS all include lesions that were init (...truncated)


This is a preview of a remote PDF: https://academic.oup.com/jnci/article-pdf/107/12/djv290/17986206/djv290.pdf

Morrow, Monica, Katz, Steven J.. Addressing Overtreatment in DCIS: What Should Physicians Do Now?, JNCI: Journal of the National Cancer Institute, 2015, Volume 107, Issue 12, DOI: 10.1093/jnci/djv290