Biomarker reporting has increasingly become a key component of pathology reporting, providing diagnostic, prognostic, and actionable therapeutic data for patient care.
To expand and improve the College of American Pathologists (CAP) biomarker protocols.
We surveyed CAP members to better understand the limitations they experienced when reporting cancer biomarker results. A Biomarker Workgroup reviewed the survey results and developed a strategy to improve and standardize biomarker reporting. Drafts of new and revised biomarker protocols were reviewed in both print and electronic template formats during interactive webinars presented to the CAP House of Delegates. Feedback was collected, and appropriate revisions were made to finalize the protocols.
The first phase of the CAP Biomarker Workgroup saw the development of (1) a new stand-alone general Immunohistochemistry Biomarker Protocol that includes reporting for ER (estrogen receptor), PR (progesterone receptor), Ki-67, HER2 (human epidermal growth factor receptor 2), PD-L1 (programmed death ligand-1), and mismatch repair; (2) a new Head and Neck Biomarker Protocol that updates the prior 2017 paper-only version into an electronic template, adding new diagnostic and theranostic markers; (3) a major revision to the Lung Biomarker Protocol to streamline it and add in pan-cancer markers; and (4) a revision to the Colon and Rectum Biomarker Protocol to add HER2 reporting.
We have taken a multipronged approach to improving biomarker reporting in the CAP cancer protocols. We continue to review current biomarker reporting protocols to reduce and eliminate unnecessary methodologic details and update with new markers as needed. The biomarker templates will serve as standardized modular units that can be inserted into cancer-reporting protocols.
Genomic biomarker testing has grown dramatically in recent years, from single-gene testing to larger panels with hundreds of genes or even the entire exome and genome.1 Beyond individual mutations, aggregate measures such as characterizing microsatellite instability and overall mutational burden (tumor mutational burden) have also become more common and clinically relevant.2 However, reporting approaches vary widely across academic medical centers and independent laboratory vendors.3 Other difficulties arise as well, such as when testing is performed at multiple laboratories, with results in different places in the electronic medical record or not there at all.
Use of immunohistochemical (IHC) biomarker testing to evaluate biomarkers has also grown recently. In particular, PD-L1 (programmed death ligand-1) IHC testing has increased significantly, given the reported associations with immunotherapy response, and represents a complex, dynamic, and heterogeneous reporting landscape.4 A trend that PD-L1 IHC testing highlights is that a given IHC biomarker tends to be applied across multiple tumor types (as exemplified by mismatch repair testing by immunohistochemistry), albeit with tumor type–specific reporting aspects in certain contexts.
The College of American Pathologists (CAP) cancer protocols have been successful in standardizing pathologist reporting of critical features of cancer resections, particularly in the context of staging information.5 For laboratories that are accredited by CAP or for cancer centers accredited by the American College of Surgeons Commission on Cancer, reports must include the required data elements identified in the applicable cancer protocol (ANP.123506 ). While biomarker protocols have existed since 2014,7 these have been underused and not completely standardized.
METHODS
The CAP Biomarker Workgroup was formed with members from the CAP Cancer Committee and Pathology Electronic Reporting Committee. The group met regularly and contributed their own ideas on potential issues with the biomarker protocols. CAP members were also surveyed for their opinions on the protocols and possible improvements.
With these data on problems and limitations with the protocols, we created a new biomarker protocol for IHC, a new electronic version of an outdated Head and Neck Biomarker Protocol, and a major revision of the Lung Biomarker Protocol. These were developed within CAP’s new single-source protocol system to simplify creation of electronic and paper checklists.
The IHC biomarker reporting protocol attempts to create a structure of basic reporting elements for reporting the results of various IHC tests. We used a method called entity attribute value modeling in which the specimen and IHC biomarker are considered the main components (entity) and the results and methods are considered the attributes (Figure 1, A and B). Within the results and methods, we created a set of attributes (eg, interpretation, score, antibody clone) and associated value sets (eg, Positive/Negative/Equivocal/Indeterminate, 0/1+/2+/3+/Other, HercepTest/4B5/SP3/Other) that could be used in the context of a specific IHC biomarker. There was a focus on flexibility, in that very little is mandatory.
A, Overview of the data-modeling approach for immunohistochemistry biomarker reporting, associating specimen information to results and methods of the biomarker tested. B, Templates for programmed death ligand-1, mismatch repair, HER2 (human epidermal growth factor receptor 2), estrogen receptor, progesterone receptor, and Ki-67 were developed. The implementation for HER2 is shown as formatted in the print version of the IHC biomarker protocol. Abbreviations: IHC, immunohistochemistry; LDT, laboratory-developed test.
A, Overview of the data-modeling approach for immunohistochemistry biomarker reporting, associating specimen information to results and methods of the biomarker tested. B, Templates for programmed death ligand-1, mismatch repair, HER2 (human epidermal growth factor receptor 2), estrogen receptor, progesterone receptor, and Ki-67 were developed. The implementation for HER2 is shown as formatted in the print version of the IHC biomarker protocol. Abbreviations: IHC, immunohistochemistry; LDT, laboratory-developed test.
The Head and Neck Biomarker Protocol was converted for the first time from paper format with the goal of simplification and updating of key molecular alterations to include theranostic markers and newly available methodology. Similar to other protocols, methods sections were removed as they were considered redundant to the official molecular report. Human papillomavirus reporting was updated to include E6/E7 mRNA (messenger RNA) in situ hybridization. Newer markers including NTRK were added; SWI/SNF (SWItch/Sucrose Non-Fermentable) surrogate IHC stains and PAX (paired box) rearrangements were included in the sinonasal section. Salivary duct carcinoma HER2 (human epidermal growth factor receptor 2) reporting was modeled after that of the IHC biomarker reporting protocol.
The Lung Biomarker Protocol underwent a major revision with a goal of simplification. Realizing the protocol would be a supplement to the official molecular reports, we removed the methods sections to reduce the amount of copying required for the pathologist and considered the treating physician unlikely to use this information. The results were standardized to use single-letter amino acid codes, again as these were considered to be most likely what the oncologists would recognize (eg, “BRAF p.V600E”). The most common mutations were provided as options for selection, as well as an “other” option with free-text entry. New biomarkers since the last update were added, including NTRK, tumor mutational burden, mismatch repair/microsatellite instability status, and PD-L1. PD-L1 was transferred from the new IHC protocol for consistency.
The new lung, head and neck, and IHC protocols were presented at an open virtual meeting for CAP members to collect feedback. Comments were positive, though interpretations were requested for the Lung Biomarker Protocol and were added. It was noted that protocols like these may become unwieldy over time, and less necessary as large panels become the norm.
RESULTS
CAP member survey results regarding the original, older protocols are shown in Figure 2. Of respondents, 48% (65 of 135) reported no “pain points” in using the protocols; 33% (45 of 135) said reports were inconsistent; 17% (23 of 135) had issues related to testing being performed in other laboratories; 16% (22 of 135) said vendor systems could not handle biomarker results; and 13% (18 of 135) said current biomarker reports are difficult to use.
“Pain points” for old biomarker reports. About half of users reported issues with the old biomarker templates.
“Pain points” for old biomarker reports. About half of users reported issues with the old biomarker templates.
Nonusers reported various issues described in Figure 3. Of nonusers, 37% (30 of 81) prefer internal templates; 33% (27 of 81) think the CAP biomarker protocols collect too many data elements; 32% (26 of 81) noted that biomarker results are not available at the same time; and 30% (24 of 81) said the templates do not match their workflow.
Reasons why College of American Pathologists biomarker protocol templates are not used.
Reasons why College of American Pathologists biomarker protocol templates are not used.
We received numerous suggestions for improvement, shown in the Table. These included multiple requests to make the biomarker protocols briefer. Some had issues with integration into the electronic health record, such as not being able to use a synoptic in an addendum. Other specific comments included the need to include National Comprehensive Cancer Network–recommended biomarkers, making them easier to use, making them modular, adding PD-L1 reporting, and a desire for automated result extraction. Some, such as standardizing PD-L1 across organs, are beyond the scope of our group.
The group considered these comments and developed ideas for improvements. We removed the methods sections, suspecting it would be difficult for general pathologists to find and transcribe these. Since the template is a supplementary report, this required information would still be present in the original reports. We also removed specific drug names and treatment information, knowing this would change rapidly and become inaccurate. However, after later feedback on the new templates, we added brief comments on treatment such as “This mutation is targetable with specific tyrosine kinase inhibitors.”
DISCUSSION
We were able to create updated, simplified lung8 and head and neck9 biomarker protocols as well as a generic, flexible, and standardized approach to IHC biomarker reporting,10 spanning biomarkers such as PD-L1, ER, Ki-67, and the mismatch repair panel. This framework represents a starting point to consider how the CAP could begin harmonization of the IHC biomarkers reporting across various tumor types that are currently embedded in the organ-specific protocols. The challenge with this will be that reporting of a given biomarker may or may not be tumor-type specific, and allowing for that is complex, although technically possible. Arguably, the more provocative question revolves around the decision in this first iteration to avoid associating prognostic and/or therapeutic “explanatory notes.” The reasoning for this was to focus on modeling the discrete data capture around IHC testing, in which the pathologist interpretation is the primary source of the data. Certainly, as has been the case with genomic biomarkers, a knowledge store of associations could be paired with the objective findings from any given IHC biomarker(s).
Still, the work on these new and modified protocols made clearer the issues with the manually entered checklist approach to biomarkers. The entering pathologist must find results from multiple methodologies, potentially from different laboratories, and transcribe them without error. While we may have improved this by eliminating some sections, the bulk of the problems remain.
A unified report could happen 2 ways. First, if all testing is performed centrally, potentially as part of a large next-generation sequencing panel, it can be reported in an automated and reproducible fashion by the laboratory. The disadvantages of this approach include the difficulty of implementing all important tests centrally, and the fact that this would favor large commercial laboratories. The second possibility, if testing is performed at multiple laboratories or at separate time points and one unified report is not initially provided, is to piece together the final unified report by health communications standards, that is, HL7 (Health Level Seven). Though not designed for complex results such as large genomic data, larger datasets can be embedded in HL7 messages such as with an embedded.json (JavaScript Object Notation). The Biomarker Workgroup has started discussions with large laboratories on options to transmit data automatically and populate the synoptic reports.