Rosen's Breast Pathology, 4e

xviii

Introduction

invasive carcinoma and 3 (8%) had DCIS. 42 The other re- port described the detection of invasive carcinoma in 1 of 28 (3.8%) women and DCIS in 4 (14.3%) others during a median follow-up of 38 months. 43 The consequences of short follow-up and clinical moni- toring in the foregoing prospective studies are significant. In every instance, subsequent DCIS was detected more frequently than invasive carcinoma. Had the patients not been monitored closely it is highly likely that many of those subsequently found to have DCIS would have developed invasive carcinoma in later years. In fact, it is troubling that the frequency of invasive carcinoma was so high (3.6% to 6.8%) after a median follow-up of only 5 years or less in pa- tients known to have DCIS. This suggests that at least some of these women may have harbored invasive carcinoma when DCIS was first detected, and/or that progression from DCIS to invasion can occur rapidly. In any case, none of the preceding prospective studies can be cited as evidence in fa- vor of the proposal to eliminate the word “carcinoma” from DCIS, or in support of a “treatment” plan based on clinical follow-up alone that underestimates the potential of DCIS to give rise to invasive carcinoma. In view of the foregoing evidence, and other data pre- sented in Chapter 11, there is an urgent need to better understand the factors involved in the biology of untreated DCIS, especially the low-grade variants of the disease. This will certainly involve intensive molecular studies, including gene expression profiling of DCIS and invasive ductal car- cinoma lesions, but it is important not to overlook the role that the “host” environment might play in modulating the evolution of DCIS. As noted above, the same issue pertains to LCIS, and despite more than a decade of investigation, there is still no reliable guide to predicting the likelihood that invasive carcinoma will appear in a particular patient with classical LCIS. Deleting the word “carcinoma” from DCIS will not solve the fundamental problem. Understating the risk associated with failing to adequately treat low-grade DCIS is likely to create a false sense of security, and might cause significant harm . The 2012 World Health Organization (WHO) publication on the classification of mammary carcinoma replaces the term “invasive ductal carcinoma, NOS,” with “invasive car- cinoma of no special type” because the word “ductal” in this context “…perpetuates the traditional but incorrect concept that these tumors are derived exclusively from mammary ductal epithelium in distinction to lobular carcinomas, which were deemed to have arisen from within lobules, for which there is also no evidence.” 44 The muddled thinking of the authors of the WHO book who made these asser- tions is manifested by the fact that they retained terms such as DCIS, atypical ductal hyperplasia, and invasive lobular Does Invasive Ductal Carcinoma, NOS, Have a Future?

diagnostic biopsy sample in some cases. In others, however, it is possible that persistent DCIS remained dormant, either failing to develop the capacity to invade and metastasize or actually regressing, possibly to extinction. Data from two studies involving a total of 38 women with low-grade micropapillary, papillary, and cribriform DCIS are particularly relevant to this issue. 35,36 Previously overlooked DCIS in these cases was found in the course of reviewing breast biopsies that had been diagnosed originally as benign, a circumstance that attests to the low-grade nature of most of the lesions. With follow-up averaging 21.6 years 35 and 30 years, 36 respectively, invasive carcinoma was found in 16 (48%) of the combined total of 38 patients. In the latter study, 36 the frequency of subsequent carcinoma was 9.1 times expected, with a 95% CI of 4.73 to 17.5. A third, more recent review of biopsies classified as benign found 13 instances of previously unrecognized DCIS (4 low grade, 6 intermediate grade, and 3 high grade). 37 In the course of follow-up, subsequent ipsilateral invasive carcinoma was diagnosed in 6 of the 13 (46%) patients after intervals of 4 to 18 years. In this study, the odds ratio of developing invasive carcinoma in women with DCIS when compared to women with nonproliferative breast disease was 13.5, with a 95% CI of 3.7 to 49.7. A number of prospective studies of patients with DCIS who had no treatment after a biopsy are also available. The studies differ from the foregoing retrospective studies in two important respects: (1) because the patients were known to have DCIS and were monitored clinically after the initial biopsy, new abnormalities were promptly investigated; and (2) reported clinical follow-up rarely exceeded 5 years. Two studies described selected patients with clinically occult DCIS that was detected by mammography. In one, 70 patients were followed for a median of 47 months after biopsy-proven DCIS was diagnosed. 38 Among the DCIS le- sions, 51% had a comedo (high grade) component and 29% were predominantly comedo-DCIS. During the course of follow-up, invasive carcinomas were detected in 3 (4.3%) patients, and 8 (11.4%) were found to have further evidence of DCIS. Another report describing a selected group of untreated DCIS patients included 59 women. 39 During a median follow-up of 37 months, invasive carcinoma was detected in 4 (6.8%) and additional DCIS in 6 (10.9%). There are two noteworthy population-based studies of women with DCIS who had no treatment after a diagnostic biopsy. One series consisted of 112 patients with a median follow-up of 53 months during which 5 (4.4%) invasive car- cinomas and 19 (17%) instances of DCIS were detected. 40 A smaller study of 21 patients described the finding of invasive carcinoma in 3 (14.3%) patients during a median follow-up of 7 years. 41 Finally, two prospective studies of women with untreated DCIS detected by mammography screening can be cited. In one report, 38 women with predominantly low-grade cribriform DCIS were followed for a median period of 60 months, during which time 2 (5%) were found to have