Regulatory Forum Opinion Piece*

It seems that every few years concerns are raised about the practice of pathology peer review, even though peer review has been part of toxicologic pathology since the late 1970s. While the individual procedures have varied slightly, the overarching intent of pathology peer review has been to provide diagnostic consistency and accuracy of interpretation of the pathology data to the global regulatory authorities to aid in their decision making regarding the safety of biopharmaceuticals and chemicals. The intent of peer review is not what raises periodic cause for concern among regulators, but rather other technical aspects of the process that surface periodically. That is why we believe that worksheets should be retained even though this step is not “required” or espoused by Society of Toxicological Pathology (STP; Morton et al. 2010).

Kerlin (2012) briefly touched on some of the perceived reasons behind issues with peer review in his Regulatory Forum Opinion Piece on pathology raw data. Often times the “documentation” of the peer review is regarded as pathologist’s “interim notes” and are discarded after the report is completed, as these “notes” are not deemed to be necessary for the reconstruction of the pathology evaluation. While this practice may be expeditious in completing the peer review and reduces retained documents, it can be perceived as a “black box” around the peer-review process and opens the study for criticisms relative to data integrity and manipulation of data. While pathologists have long stated that peer review is intended to ensure the quality of the pathology evaluation in safety studies (Ward et al. 1995; Boorman et al. 2010), it does no good if the process used creates doubt in the skeptical mind of compliance and regulatory authorities relying on the final evaluation and interpretation of the data. It has previously been suggested that the peer review worksheets could be included as an appendix to the pathology report or maintained in the study file (Ward et al. 1995). Long (1996) and Mann (1996) continued to stress the need for a well-documented peer review process to give regulatory reviewers greater confidence in the pathology evaluation. However, a recent commentary by McKay et al. (2010), based upon current practices in the United States, European Union, and Japan, restated the position that peer review worksheets are not raw data and do not need to be retained or submitted. The issue is not what constitutes raw data and what should be retained, but instead the issue is one of transparency in the process.

The latest controversy has focused not on the raw data but on the integrity of the data generated by the study pathologist. This is not a new issue, as Peters (1996) has already highlighted the need to maintain peer review worksheets to add credibility to the process and to ensure that the recommended data changes as a result of the peer review were made. More recently, a series of Food and Drug Administration (FDA) compliance inspections and interactions with agency representatives both in the United States and in Europe have questioned the timing of when the database should be “fixed or locked.” The “influence” by the sponsor and/or the peer review pathologist on the study pathologist’s findings has raised concerns on the part of regulators regarding how to maintain the integrity of the database. This has led to a recommendation from compliance inspectors that pathology data be locked prior to the peer review in order to track changes made during or after the peer review.

The concerns raised by the regulators distill to transparency of the peer-review process. Although STP has provided recommendations on pathology peer review (Morton et al. 2010), these do not directly address the issue of transparency. Changes in diagnoses during the peer review, particularly of neoplasms, raises concerns in the eyes of the regulator relative to the role of “influence” by the peer-review pathologist and/or sponsor. We believe that being able to trace what was done and when is one way to alleviate this concern. Locking of the data before the review and tracking subsequent change as a result of peer review is another way to achieve greater transparency. This is, though, not the most efficient way to meet the desired level of transparency.

Attempts have been made recently by regulatory authorities both in the United States and European Union to draft guidance documents for pathology peer review. Both of these efforts illustrated a misunderstanding of the science of toxicologic pathology and the purpose of peer review. It was clear from these draft guidance documents and discussions at the Third Global Quality Assurance Conference in 2011 held in Kyoto, Japan, that peer review is where science and compliance collide and that the desire of the regulators is to be able to follow changes in diagnoses. At the simplest level, the “locked” data prior to peer review would provide a mechanism for this. This procedure does, however, create additional burdens on both the study and peer-review pathologists. It has been simpler to conduct a peer review prior to “locking” the data to harmonize nomenclature. How then does one provide the desired transparency in peer review? We propose that the necessary transparency comes through treating the “documentation” of the peer review as nothing more than quality control (QC) interim pathologist notes and retaining these peer-review documents or notes with the pathology files. These notes are not considered to be “raw data” as they are not necessary for the reconstruction of the pathology report; however, these files would create a level of transparency that has not always been present and harkens back to the suggestions made by Ward et al. (1995). Retaining these notes not only facilitates review of the pathology report by the peer review pathologist to ensure the necessary changes to diagnoses and interpretation have been made by the study pathologist but also provides transparency to the recreation of the pathology evaluation at a later date.

So the question remains, should pathology data be “locked” before or after the peer review? In our view, “locked” versus “unlocked” really does not matter. As long as events are recorded and retained to maintain transparency in the process, peer review can continue as it has with achieving consensus prior to signing of the pathology report. While the STP position paper advocates maintaining the status quo (Morton et al. 2010), we believe that a course correction is warranted. There are likely many STP members who will disagree with this view, but when properly performed this method is efficient and does not add time or expense to the reporting process. As with previous pieces written for this column, this article brings attention to a controversial topic and it is now incumbent on the global societies of toxicologic pathology to come together and provide a workable recommendation for tracking the pathology peer review that will satisfy a need for transparency while maintaining the independence and integrity of the process in an efficient manner.