College of American Pathologists
Printable Version

  Hows and whys behind updated cancer protocols


CAP Today




March 2012
Feature Story

Susan C. Lester, MD, PhD

The surgical pathology report is the most important medical document for the majority of cancer patients. The pathologic information is used to predict the most likely outcome for the patient and to determine the most effective local and systemic treatments. Therefore, it is generally agreed that standardized reporting of cancers is a worthy goal. The CAP Cancer Committee in the 1980s took on the difficult task of developing protocols for reporting major cancer types. The first invasive breast cancer protocol was released in 1986.

A project to revise and update the protocols began in 2005. As part of this process, each Cancer Committee member identified a panel of interested physicians that consisted of pathologists from private and academic practices, located throughout the United States, as well as physicians in related specialties such as surgery, radiation oncology, and medical oncology. Prior breast cancer protocols listed three authors. The current breast review panel was expanded to include 14 physicians in three specialties.

Developing a consensus guideline for pathology reporting is complex because of the central role of pathology reports in patient care, epidemiology, quality assurance, clinical trials, and other forms of research. In addition, in the transition to electronic reporting, many questions of the best means of “paper reporting” versus “computer reporting” arose. The ideal pathology report had to acknowledge, balance, and prioritize the many demands on pathology reporting for differing purposes.

The breast cancer review panel carefully considered the elements of reporting that are necessary to describe the extent of the carcinoma for T and N classification and local control, for prognostication, and for correlation with clinical findings. Elements were included only if there was a consensus on their importance and scientific support for their use in patient management. Unlike in the preceding checklist, all the elements required for AJCC/UICC (American Joint Committee on Cancer/Union for International Cancer Control) classification were included as separate elements, though some of the findings are unusual. The reporting of ductal carcinoma in situ was expanded to accommodate cases in which DCIS is the primary component with only microinvasion present. Because estrogen receptor (ER), progesterone receptor (PgR), and HER2 testing are now recommended for all invasive carcinomas, a section for reporting these results was also included.

The checklist had to be lengthened to accommodate many types of specimens containing a wide variety of carcinomas. There had to be sufficient scope to report a mastectomy with extensive DCIS and microinvasion, as well as a local excision containing a solitary focus of invasive carcinoma. The alternative, creating multiple checklists for multiple types of specimens and cancers, had disadvantages and was not pursued.

After the updated protocol was released in 2009, it became clear that, as hard as the panel had tried, the optimal checklist had not been created. Thus, the protocol underwent an open comment period in April 2011. Seventy-nine individuals responded. Although not all of the suggestions could be incorporated into the protocol revision, all were carefully considered.

The 2009 protocol can continue to be used if the reporting of ER and PgR are updated to the guidelines of the CAP and American Society of Clinical Oncology. The changes that have been made to the revised protocol build in flexibility. Pathologists can choose whether or not to adopt them.

The comment we heard most often was about the order of the elements. In the original protocols the Cancer Committee developed, all elements were in the same order across all protocols. Uniformity of the checklists across tumor sites made it easier, in the committee’s view, to ensure all important elements were present and information could be located easily in the same place in each report. However, this resulted in elements being reported for some cancers in an order not typical for those cancers. A specific order for the elements was not required for synoptic reporting, but it was never made clear that users were able to modify the order. Thus it has been clarified that the elements in any protocol can be rearranged into any order the pathologist believes will communicate the information best. The order of the elements of the invasive breast protocol has been modified to better reflect the importance and relationship of the elements to each other and to reflect more traditional reporting practices. However, some pathologists may choose to organize the elements in other ways.

Requiring uniformity across tumor sites meant, too, that some elements that were most applicable to only some tumor sites were in protocols where they were less applicable. The elements of “specimen” and “specimen integrity” have been removed from the breast cancer protocol.

The length of the protocol was another source of major concern. Achieving the optimal length requires balancing the need for a checklist that can encompass a wide variety of specimens and carcinomas with the need to avoid creating a checklist containing too many elements that may not pertain to the current case and may obscure diagnostic information. Because not all elements are relevant to all specimens, many elements become “pertinent negatives.” The value of reporting a pertinent negative finding is that it is definite information provided by the pathologist. If an element is not reported, a variety of possible assumptions can be made, including that the finding is not present, the finding cannot be evaluated (for example, the associated structure is not present), the finding is present but the pathologist chose not to report it, or the finding is present and the pathologist chose to report it but the element is not present in the report (for example, due to transcriptional error). For checklist elements not relevant to a specific specimen, the review panel decided it was often preferable to leave decisions about reporting up to the pathologist. Therefore, the reporting of elements related to skin, nipple, and chest wall involvement, multiple foci of invasive carcinoma, lymph nodes, and distant metastasis were made optional when not relevant to the case.

A third major concern was about the inability to use the traditional method of including procedure(s), laterality, and location on a single line as a report header rather than as separate data elements. This issue concerns some of the conflicts between the current state of electronic reporting versus more traditional narrative reporting. Separate elements are the most compatible form for electronic reporting. However, headings summarize information and are traditional and easily understood. Therefore, these data elements can now be on a single line and still fulfill the requirement for synoptic reporting. AJCC/UICC T, N, and M categories can also be reported on a single line. In addition, there is no need to include the definitions of each category.

The inclusion of reporting guidelines for estrogen receptor, progesterone receptor, and HER2 was also reconsidered. These results are generally not available at the time of sign-out of the main report. Therefore, the elements present are optional and not required in the main report if there will be a supplemental report. The ASCO/CAP guidelines for reporting ER/PgR and HER2 provide additional guidelines for the primary reports of these studies.

Other minor changes were made to help customize reports (for example, the reporting of lobular carcinoma in situ and the size of the largest metastasis was made optional), to simplify reporting (for example, the style of reporting margins was modified), or to provide more information (for example, the number of mitoses per mm2 for grading was provided for each group).

A major limitation of standardized reporting is that it creates the impression that all cancers can be described using a finite set of features, which all pathologists know is not true. An identification card carries standard data such as the person’s name, height, and eye color, but these elements do not replace the information contained in the person’s photograph. In a similar fashion, the checklist should be considered to be only the start of providing important pathologic information. “Additional pathologic findings” and notes should be used to capture the rich and complex nature of cancers that all have a unique biologic origin and arise in unique individuals.

The ultimate goal of standardized reports is to provide better and higher quality care to patients and more accessible information for epidemiologic and other studies, and to be an adjunct to research that will lead to new ways to improve patient outcomes. The Cancer Committee and breast review panel hope that the revised protocol will provide pathologists with a better balance between the competing demands on this important document.

Dr. Lester served on the CAP Cancer Committee from 2006 through 2011, during which time she was head of the breast review panel. She is chief of breast pathology services in the Department of Pathology, Brigham and Women’s Hospital, Boston. The revised breast carcinoma protocol will be posted May 25 at

Related Links: CAP Learning Related Links: CAP Learning