Rosen's Breast Pathology, 4e

xvi

Introduction

induce carcinomas at a later date, but this proved to be of less concern because in subsequent years, advances in mam- mography technology led to substantially reduced radiation exposure. Many later studies documented the feasibility of mammography screening and also confirmed that it reduced breast carcinoma deaths in the screened populations. 25–30 Encouraged by the success of breast screening and with the availability of suitable tests, screening for the detection of occult tumors in other organs such as the prostate gland and lungs was introduced, again prompting criticisms of “overdiagnosis” and “overtreatment.” 31,32 In this regard, a study that reported a reduction in deaths due to carcinoma of the lung after screening with low-dose CT scans also noted that 96.4% of “positive” screening findings did not prove to be carcinoma, resulting in a large number of diag- nostic procedures that did not benefit these individuals. 33 In March 2012, the NCI sponsored a meeting to once again address concerns that screening results in the “over- diagnosis” of cancer. A report summarizing the conclusions of the participants was published in July, 2013 34 in an article that attracted wide public attention. The authors of the re- port defined “overdiagnosis” as a diagnosis “…which occurs when tumors are detected that, if left unattended, would not become clinically apparent or cause death. Overdiagnosis, if not recognized, generally leads to overtreatment.” It was concluded that “overdiagnosis” most often occurs as a re- sult of screening when clinically asymptomatic, “indolent” cancers are likely to be detected. In this context, the authors defined “cancer” as a disease “…with a reasonable (my ­Italics) likelihood of lethal progression if left untreated.” The word “reasonable” was not defined by the authors, but pre- sumably referred to an unspecified risk that a patient would experience a fatal outcome. The authors also recommended that the word “cancer” should be dropped from what were referred to as “premalig- nant conditions” such as DCIS that should be renamed “in- dolent lesions of epithelial origin” under the acronym IDLE conditions. They suggested that this change would remove the frightening connotation associated with “cancer” and re- duce “overtreatment” by making it easier to recommend less aggressive therapy for “­indolent” lesions. In support of this position, it was argued that not all in situ carcinomas in vari- ous organs progress to an invasive stage if left untreated, as evidenced by the clinical biology of prostatic (PIN) and cer- vical (CIN) neoplasia, and that in some instances (e.g., the prostate gland) the invasive neoplasms that ultimately arise are so indolent that they pose little danger in the lifetime of the patient. A corollary of latter argument was that the cur- rent tendency to manage all in situ carcinomas with equally aggressive treatments results in the overtreatment of some patients who might not have needed the recommended therapy, and may have been harmed by it. Finally, being able to replace costly treatment with observation would reduce healthcare costs. In summary, it was suggested that simply changing the name of a disease would result in important improvements in patient well-being and save money.

Before addressing the foregoing proposal itself, it is nec- essary to comment on a matter of semantics relating to the words “cancer” and “carcinoma” as they were used by the authors of the aforementioned proposal, as well as many other authors cited among the references in this book. The online Merriam Webster Dictionary defines cancer as “a malignant tumor of potentially unlimited growth that ex- pands locally by invasion and systemically by metastasis.” Carcinoma is defined as “a malignant tumor of epithelial origin.” Thus, carcinoma refers to the subset of malignant tumors arising from epithelium, whereas cancer refers to the entire spectrum of malignant tumors, including carci- nomas, sarcomas, lymphomas, leukemias, and malignant neoplasms of the central nervous system. Regrettably, Drs. Esserman, Thompson, and Reid confused these terms throughout their paper. Although they were mainly con- cerned with “overtreatment” relating to screening-detected carcinomas arising at various sites, and pigmented skin le- sions, they repeatedly used the words cancer and carcinoma interchangeably. In the second paragraph of the article, they refer to “breast cancer and prostate cancer” when they ap- pear to mean carcinoma. The authors’ misuse of these terms is best appreciated in the following quotation from one of the summary recommendations: “First, premalignant con- ditions (eg. ductal carcinoma in situ or high-grade prostatic intraepithelial neoplasia) should not be labeled as cancers or neoplasia, nor should the word cancer be in the name.” This recommendation is meaningless because the word “cancer” already does not appear in the names of the cited lesions. It is to be hoped that Drs. Esserman, Thompson, and Reid are not seeking to deny the concept of in situ carci- noma generally, and in the breast specifically, by referring to it as a “premalignant” condition. All invasive carcinomas arise from a preinvasive stage of the disease that develops in the epithelium from which the carcinoma originates. The duration of the preinvasive stage is variable, depending on factors that are largely not known. At the histologic level, the cytologic appearance of in situ carcinoma cells is often indis- tinguishable from that of the invasive carcinoma it has given rise to. Molecular studies have shown a high level of concor- dance in the genetic alterations between these components in a given tumors that consists of DCIS and invasive ductal carcinoma, as discussed in Chapters 11 and 12, as well as Chapters 31 and 32 in the context of lobular carcinoma. Rather than denying the existence of preinvasive carcinoma, what is needed is further study to identify the molecular alterations that endow DCIS (and lobular carcinoma in situ [LCIS]) with the ability to invade and metastasize, as well as changes in the patient’s “resistance” that might enable these events to occur. Turning to the flawed proposal, which, if adopted in its current form, would probably be more harmful than beneficial, it is self-evident that changing the name of a disease would not change the disease itself. Despite some general principles that invasive carcinomas appear to have in common, such as epithelial origin and a preinvasive,