Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Pathology Peer Review
1
Introduction
In vivo regulatory-type animal toxicity studies that include pathology endpoints are generally not hypothesis-driven experiments. Instead, these studies are meant to gather information using multiple, medical diagnostic endpoints, assessed both antemortem (i.e., in life) and postmortem, to discover or characterize the toxicity profile that results from the administration of relatively high doses of a test article to animals. Characterizing the toxicity profile necessitates integration of these diagnostic endpoints that form the basis of the conclusions of the study. Typically, the study pathologist (SP), as the principal medical diagnostician (veterinarian or physician) on the study team, is responsible not only for describing the macroscopic (gross) observations, organ weights, and microscopic pathology findings but integrating those finding with the antemortem endpoints to fully describe the toxicity profile in the pathology report; in some instances, the SP also may be asked to address clinical pathology parameters (e.g., hematologic and serum chemistry values) along with anatomic pathology endpoints. It follows that, from the perspective of the SP, a toxicity study is, in fact, a medical diagnostic exercise.
Diagnosis in a toxicity study goes beyond the postmortem microscopic anatomic evaluation of tissue sections from a comprehensive set of samples from bodily organs, although the SP is the contributing scientist who generates these diagnoses. Proper diagnosis requires integration of clinical endpoints; such as body weight and food consumption; clinical signs and symptoms and their time of onset (including evaluating the cause of death, if early death or unscheduled euthanasia occurs); and clinical pathology parameters (e.g., serum chemistry, hematology, urinalysis) in addition to the postmortem endpoints of organ weights and the gross and microscopic evaluation of the organs. For practical purposes, the necropsy essentially equates to a thorough, and final, physical examination.
The SP's evaluation that leads to the final diagnoses and their interpretation is perhaps the most essential component of identifying and characterizing potential hazards to human or animal health associated with exposure to the test article in most studies. Therefore, it is crucial that the SP's evaluation is thorough, and the diagnoses are accurate, for this becomes the foundation upon which the risk assessment is built. Procedures or practices should ordinarily be in place at a testing facility to ensure diagnostic accuracy and minimize misdiagnosis or inadvertent diagnostic error. The principal practice to ensure accuracy, and one that is required by the global Good Laboratory Practice (GLP) regulations and guidance documents ( ; ), is assuring that SPs have sufficient pathology education, training, and experience to evaluate toxicity studies ( Pathology and GLPs, Quality Control and Quality Assurance ; Vol 1, Chap 27 ; ; ). However, achieving an accurate pathology diagnosis is an iterative process of progressive diagnostic refinement that is, at its best, a collaborative activity. While the SP is ultimately responsible as the contributing scientist for the generating the pathology raw data and preparing the final pathology report, most pathologists do not reach their conclusions in isolation.
Although generally not required in the global GLP regulations, the practice of pathology peer review serves as an important quality control (QC) procedure to increase confidence in the accuracy of the pathology diagnoses and their interpretations. Pathology peer review entails having a second pathologist (known as a peer review pathologist [PRP]), or occasionally a group of three or more additional pathologists (known as a pathology working group [PWG]), evaluate a subset of the pathology postmortem findings originally identified by the SP (microscopic diagnoses as well as diagnostic terminology and severity grades) in addition to the overall interpretation of these findings in the report. This chapter will review current pathology peer review practices and their applications for the various types of animal toxicity studies conducted by industry and government research institutes as well as the current state of government regulation of peer review in industry, including appropriate documentation and reporting practices.
2
Peer Review Timing and Pathology Raw Data
Pathology peer review is not required by regulatory agencies, but studies in which the pathology data have undergone peer review are particularly welcome by regulatory reviewers due to the perception that the QC afforded by the additional level of review has improved the quality of the pathology raw data. Recommendations for performing an effective pathology peer review have been prepared by the Society of Toxicologic Pathology (STP); these are available both in print ( ) and online ( https://doi.org/10.1177/0192623310383991 ). These peer review best practices have been endorsed by nine other societies of toxicologic pathology from around the world as well as the American College of Veterinary Pathologists (ACVP).
There are two types of pathology peer review, and they differ in respect to timing ( Figure 26.1 ) A contemporaneous (or prospective ) peer review is conducted after the SP has generated draft or preliminary pathology diagnoses and prepared a draft narrative report but prior to signing and dating the final pathology report (i.e., prior to the creation of the pathology raw data, as discussed below). This is the most common type of pathology peer review conducted in industry and in government research institutes. During a contemporaneous peer review, the PRP works with the SP to refine the pathology diagnoses and terminology. In contrast, a retrospective peer review is performed after the pathology report has been finalized, signed, and dated—often weeks, months, or years later—in order to review a specific question(s) or render a second opinion. In other words, a retrospective pathology peer review is performed on a study for which pathology raw data already exist, so the outcome of the peer review may lead to changes in existing data and the need to prepare an amended pathology report.
2.1
Pathology Raw Data and Peer Review
Understanding the difference between a contemporaneous and retrospective peer review requires an understanding of the definition of “raw data” as it pertains to histopathology diagnosis. The definition of “pathology raw data” influences why certain documentation actions are, or are not, required with contemporaneous or retrospective peer review. Pathology raw data has long been considered by the U.S. Food and Drug Administration (FDA) as the signed and dated final report of the SP (to include the finalized histopathology diagnoses). Importantly,
[t]he pathologist's interim notes, therefore, which are subject to frequent changes as the pathologist refines the diagnosis, are not raw data because they do not contribute to study reconstruction. Accordingly, only the signed and dated final report of the pathologist comprises raw data respecting the histopathological evaluation of tissue specimens.
In a contemporaneous peer review, the PRP is reviewing the SP's preliminary diagnoses and interpretation as recorded in the draft report, thereby serving a QC function. The PRP is not generating raw data in this peer review but merely is producing a list of suggested adjustments to hone the diagnoses, interpretation, and/or terminology—if any seem to be needed. In general, the PRP does not recommend any alteration to the preliminary diagnoses and severity grades if the impressions of the SP and PRP use diagnostic terminology from a widely recognized lexicon (e.g., International Harmonization of Nomenclature and Diagnostic Criteria [ , https://www.goreni.org/gr3_download_nom.php ]) and are only one severity grade apart. Discussion between the SP and PRP regarding the PRP's suggested adjustments may be used to further refine the preliminary diagnoses. The SP may accept all, some, or none of the PRP's recommendations since the SP is ultimately the sole scientist responsible for generating the final diagnoses and interpretation of the pathology portion of the study. Thus, in a contemporaneous peer review, pathology raw data are not yet established at the initiation of the peer review because the SP has only recorded preliminary pathology findings and written the draft narrative/interpretation. From the (above), it is important to note that the SP's preliminary (or “interim”) notes regarding draft diagnoses are not pathology raw data; the raw data are established only when the final pathology report is signed by the SP. Logically, notes or annotations made by the PRP to the SP's preliminary histopathology diagnoses and/or draft report and then shared with the SP are also not pathology raw data. The pathology raw data are listed in the final signed pathology report, so neither the SP nor the PRP is required to retain their interim notes. The basis for this regulatory perspective is that the pathology results for the study may be recapitulated, if necessary, by the original SP or another pathologist by reviewing the tissue sections and report, without the need to review the SP's (and/or PRP's) interim notes. Thus, documentation of the peer review is limited to a statement (often called a certificate, statement, or memorandum) by the PRP of the peer review procedure. See the section below on “Documentation of Contemporaneous Peer Review“ for additional detail on appropriate documentation of contemporaneous peer review.
In contrast, a retrospective peer review is conducted after the pathology report has already been finalized so that the pathology raw data already exists for the study. Therefore, any changes made as a result of the retrospective peer review will alter the original pathology raw data. Such changes will be subject to documentation in an audit trail and will require preparation of an amendment to the original pathology report. Such amended reports will not only reflect the updated pathology data and, if warranted, revised interpretation but also will incorporate an additional report subsection noting what portions of the original report were amended, how they read before and after the new data were inserted, and the reason for writing the amended report.
In certain situations, a second pathology report may be written by the retrospective PRP to address the new pathology raw data and its interpretation. This circumstance is discussed in Section 5 in the discussion of Retrospective Peer Review by an Expert. There may also be instances when the original SP is not available possibly necessitating a second report by the retrospective PRP or a report from a PWG.
3
Peer Review Process
3.1
Consultation
Informal consultation with colleagues is frequently utilized by the SP as part of the process of refining histopathologic diagnoses. Seeking such spontaneous opinions from colleagues is not a peer review and need not be documented in the pathology report or study file. In this regard, consultation includes seeking opinions through casual discussions with colleagues either individually or in group settings (e.g., institutional pathology rounds) and/or through mentoring by more experienced pathologists. Consultation is used routinely during the primary evaluation of the study to get opinions on complex or unusual findings. The consultation can be as simple as sharing a slide or digital image with a fellow pathologist to obtain their opinion. Slightly more elaborate input may be sought by asking a colleague to review a small set of slides to see if they can sort the slides into those with the change and those without, often using criteria suggested by the SP but sometimes using criteria developed by the person being consulted. The SP will have already gone through this same sorting exercise and made their initial commitment. Consultation serves as an excellent “reality check” for the SP since it is quick and helps reinforce or refine the SP's initial diagnosis. Regardless of how extensive the consultation may be, it can contribute to SP's confidence in their histopathologic diagnosis, attribution of changes to treatment, and assignment of adversity in the draft pathology report.
3.2
Peer Review: Contemporaneous Peer Review
General Process
Although there is generally no regulatory requirement for peer review, organizations conducting toxicity studies routinely conduct contemporaneous peer reviews to optimize the quality of the pathology data. Additional considerations for whether a peer review may be appropriate include whether a nonclinical study may be submitted to a regulatory agency or support substantial financial and/or operational decisions. This motivation may extend to include peer review of investigative toxicology studies, efficacy studies, non-GLP dose range-finding toxicity studies, and due diligence evaluations of toxicity studies of compounds that have been acquired or are being considered for acquisition.
The intent to perform a peer review (either contemporaneous or retrospective) must be documented for GLP studies (and is highly recommended for all nonclinical studies with pathology endpoints) in the protocol or in a protocol amendment. The protocol or amendment should list the reviewing pathologist's name, credentials, and affiliation. The PRP is not listed as a principal investigator since they are not creating raw data but rather as a contributing scientist. Details on how the peer review will be planned, conducted, and documented are generally included in a peer review SOP and the peer review statement rather than the protocol.
A contemporaneous peer review generally starts with a review of the relevant peer review SOP. The peer review SOP to be followed can be that of the sponsor (if the PRP is from a sponsoring organization), the testing facility, or third-party entity if the peer review is subcontracted. The SOP should provide guidelines regarding what study related material to review, the groups and number of animals for examination of full tissues, the review process for target tissues, how to proceed if there is disagreement between the PRP and SP, and how to document the peer review. Detailed considerations for the initial selection for review of animals for full tissue, target tissues, and neoplasms and proliferative changes are included in the SOP and discussed later in the section below on “Core Planning Considerations.”
If a study has recovery groups or necropsies at multiple time points, conducting a single peer review at the end of a study typically is easiest administratively and logistically. This arrangement also optimizes the review across all phases of the study, thereby minimizing drift (i.e., shifts in diagnostic terminology and/or grading schemes over time) within diagnoses and severity scoring. However, development program needs may require a peer review to be conducted with each phase of the study to keep the entire project on the fastest possible track. One of the most common times the peer review is split is for studies with a recovery cohort, when the initial peer review is conducted at the end of the dosing phase and the later peer review at the end of recovery phase. The peer review concepts are still the same for both phases regardless of whether the study is reviewed in a single defined time frame or reviewed in separate phases, and the approach to the planning, conduct, and documentation of the peer reviews is similar for all portions of a study regardless of how many phases are reviewed.
Based on the study design, peer review SOP, draft pathology report, and preliminary pathology data, the PRP designs the peer review plan. Considerations in organizing the peer review process include the initial dose groups (control and treated) to assess, the number of animals (per group) to evaluate, target tissues to review, and if relevant how extensive the review will be for systemic findings that reflect a singular mechanism of action (e.g., collection of vacuolated macrophages in numerous tissues associated with systemic antisense oligonucleotide exposure). The PRP should have access to the study protocol and protocol amendments as well as the full range of study data such as mortality information, in-life observations, clinical pathology and organ weight data, macroscopic observations and the SP's preliminary microscopic diagnoses, and other relevant information (if available) such as toxicokinetic, biodistribution, and transduction data. The PRP also may consult with the SP for input in advance of the review, but the PRP has the authority and bears the sole responsibility for designing the peer review.
The PRP then proceeds with reviewing the slides from the selected subsets of animals and target tissues and compares their diagnoses and severity scores to those generated by the SP. The PRP is tasked with checking that substantive treatment-related findings are identified. This endorsement is accomplished by confirming that microscopic diagnoses, terminology, and grading criteria are consistently and appropriately applied, recorded, and interpreted across the pathology data set that was selected for review. At this point, the PRP shares their comments with the SP. The SP considers the PRP's suggestions and then decides through discussions with the PRP whether the suggested changes will be incorporated (fully or with modifications) in the revised report draft or ignored. Failure to reach consensus between the SP and PRP may require additional levels of peer review by other pathologists as covered in the section below on “Resolution of Disagreements.”
Key Study Elements to Review
The PRP carries several responsibilities during the review. This section will highlight key elements the PRP must focus on to maximize their contribution to quality of the pathology data. The PRP will ultimately be working toward consensus on these items with the SP.
Completeness of the Pathology Review (Target Tissues Identified)
The scope of the PRP's review will vary by the study type and may be adjusted as deemed necessary by the PRP within the guidelines of the appropriate peer review SOP.
In general, peer reviews examine all tissues from 30% of high-dose animals, tissues from control animals as deemed necessary by the PRP, all tissues from early deaths, target tissues from all animals of all dose groups, and all gross findings as well as preneoplastic and neoplastic changes ( ). For this purpose, the “high-dose group” should be the highest-dose group with a sufficient number of surviving animals. The baseline review for short-term general toxicity studies may be more streamlined since few neoplasms typically will need to be reviewed. Specific options for peer review of different study types for different species are detailed below under “Core Planning Considerations.”
For the reviewed tissues, the PRP must ensure that the SP's review was complete and that all treatment-related findings in target tissues were identified, diagnosed, and graded appropriately and consistently. The review includes assessment of whether key threshold levels (e.g., the no observed effect level [NOEL] or no observed adverse effect level [NOAEL]) were also properly assigned.
Consistent Use of Diagnostic Terminology and Assignment of Severity Grades
Consistent and proper use of diagnostic terminology is foundational for broad acceptance and confidence in histopathology diagnoses and their interpretation. During the primary evaluation of a large and/or complex study, the SP may drift inadvertently over time in their use of a histopathologic diagnosis and/or assignment of severity scores. The PRP should also be sensitive to potential shifts in diagnostic thresholds by the SP. In addition, the SP will routinely encounter changes that are variations in tissue morphology, tissue artifacts, and spontaneous background lesions. Therefore, an SP may choose to not record an observation (sometimes referred to as “thresholding”) as a means to avoid creating an overly complex data set with the appearance of differences when none exists ( ). The review of control animals during a peer review helps the PRP understand the spectrum of changes that may be “thresholded out” by the SP. The PRP is perfectly situated during their focused and time-limited review to identify possible shifts in diagnostic nomenclature, severity scores, and/or thresholding.
Appropriate Interpretation and Correlation in the Pathology Report
Once the SP generates the preliminary histopathology data set, the findings are categorized with respect to their relationship to the test article—usually as treatment related, procedure related, or incidental. These decisions should be clearly articulated in the draft pathology report. The SP typically should comment in the pathology report regarding whether or not a specific microscopic finding is adverse or nonadverse. It is essential in the peer review that the PRP ensures the appropriateness of the SP's interpretation of the findings because treatment-related histopathologic changes are the basis of the dose level assignments for setting dose/exposure levels (e.g., NOEL, NOAEL) used by regulators as a snapshot of a test article's potential for toxicity.
Once the slide review is complete, the PRPs then review the draft report to ensure it accurately describes the treatment-related microscopic observations and the dose groups affected, and that it adequately correlates the microscopic findings to other study parameters such as gross necropsy findings, organ weights, clinical signs, clinical pathology changes, morbidity, or early death. This report review is critical to ensure that ultimately, in the integrated study report, anatomic pathology changes are accurately conveyed and that the NOEL or NOAEL are assigned to the proper dose level if histopathologic changes contribute to that assignment.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here