Abstract
Pathology peer review verifies and improves the accuracy and quality of pathology diagnoses and interpretations. Pathology peer review is recommended when important risk assessment or business decisions are based on nonclinical studies. For pathology peer review conducted before study completion, the peer-review pathologist reviews sufficient slides and pathology data to assist the study pathologist in refining pathology diagnoses and interpretations. Materials to be reviewed are selected by the peer-review pathologist. Consultations with additional experts or a formal (documented) pathology working group may be used to resolve discrepancies. The study pathologist is solely responsible for the content of the final pathology data and report, makes changes resulting from peer-review discussions, initiates the audit trail for microscopic observations after all changes resulting from peer-review have been made, and signs the final pathologist’s report. The peer-review pathologist creates a signed peer-review memo describing the peer-review process and confirming that the study pathologist’s report accurately and appropriately reflects the pathology data. The study pathologist also may sign a statement of consensus. It is not necessary to archive working notes created during the peer-review process.
Summary of Recommendations for Pathology Peer Review
Pathology peer review is most commonly performed before a study is completed. Recommendations for pathology peer review conducted before the final pathology report has been signed are presented below and may be modified to achieve the peer-review objectives for a specific study. Pathology peer review is recommended prior to study completion when important risk assessment or business decisions will be based on nonclinical studies. The decision to peer review a study should be made on a case-by-case basis, considering the importance of decisions based on these studies, the experience and skill of the study pathologist, and regulatory requirements. The pathology peer review should be conducted by a pathologist with appropriate training and experience. The peer-review pathologist may work for a contract research organization (CRO), the sponsor, or a third party. The sponsor’s peer-review pathologist may bring extensive knowledge to the peer-review process that can improve the quality of the pathology data and interpretations. If pathology peer review will be performed before the completion of the study, the protocol or a protocol amendment should state that a peer review will be conducted. Detailed methods for the pathology peer review are not required in the study protocol. The peer-review pathologist generally is responsible for reviewing the relevant pathology-related data, often including but not limited to clinical pathology results, organ weight data, macroscopic observations, microscopic findings, ancillary pathology information (electron microscopy, immunohistochemistry, in situ hybridization, etc.), data interpretations, and the draft narrative report prepared by the study pathologist. The peer-review pathologist must have broad flexibility to select materials for review within wide, general guidelines (standard operating procedures and the objectives of the peer review) and must have latitude to examine additional organs and other study data necessary to thoroughly evaluate all potential pathology findings. The peer-review pathologist may choose to consult with the study pathologist in selecting the materials to be evaluated but is not obligated to do so. Peer review of target organs varies with the study design: In rodent toxicity studies, target organs from affected sexes should be examined from all animals in the control group, all animals in the highest dose group lacking the finding, and sufficient animals (50% or more) in affected groups to characterize the finding. In nonrodent toxicity studies, target organs from affected sexes should be reviewed in all control animals, all animals in affected dose groups, and all animals in the highest dose group that lacks the finding. In recovery groups in rodent and nonrodent toxicity studies, all target organs should be examined in all control animals and all treated animals in the dose groups and sexes in which the finding was found at the end of the treatment period. In rodent toxicity studies, the peer-review pathologist generally should examine all protocol organs in at least 30% of high-dose animals of each sex in the treatment phase. In nonrodent toxicity studies, the peer-review pathologist should examine all protocol organs in at least half of the high-dose animals in the treatment phase (minimum of two/sex). Examination of all protocol organs from a subset of control animals in rodent or nonrodent toxicity or carcinogenicity studies may be performed at the discretion of the peer-review pathologist. In rodent carcinogenicity studies, the peer-review pathologist should examine the following: All organs in at least 10% of rodents per sex in the high-dose animals for 2-year or lifetime rodent carcinogenicity studies. All organs from five high-dose animals per sex in 6-month alternative (genetically engineered mouse) carcinogenicity studies. All neoplasms in all animals. All target organs in all animals in all dose groups in which neoplastic treatment-related findings are observed or suspected in any dose group. Each sex should be considered separately. Target organs with nonneoplastic, treatment-related findings in all control animals, all animals of the highest dose group lacking the finding (to establish the no observed effect level [NOEL] for the finding), and sufficient animals in affected groups and sexes to verify the finding. At least 30% of animals in affected groups should be reviewed, and the review should concentrate on animals with target organ findings noted by the study pathologist. Differences of opinion between the peer-review pathologist and the study pathologist regarding the overall study interpretation may be resolved through consultation with other pathologists and/or subject matter experts or by convening a formal (documented) pathology working group (PWG). If the peer review is performed before the final study report is signed, the audit trail for microscopic findings should not be initiated until all changes resulting from the peer-review process have been made to the pathology data set and interpretations. The pathology peer-review memo should be prepared using these general guidelines: The peer-review pathologist prepares, signs, and dates a peer-review memo documenting materials, methods, and conduct of the review process and the peer-review pathologist’s general agreement with the study pathologist’s pathology report. An optional signed statement by the study pathologist documenting consensus with the reviewed pathology data and study interpretations may be included in the peer-review memo. The peer-review memo may be signed before the quality assurance review of the pathology data, after the quality assurance review, or after the study pathologist’s contributing scientist’s report (or an integrated study report signed by the study pathologist) has been signed because peer review is a voluntary and flexible process. The peer-review memo should be archived and/or included in the study pathologist’s report or the final study report but is not considered raw data. Preliminary diagnoses, interim notes and worksheets of the study pathologist and peer-review pathologist, and the draft versions of the study pathologist’s reports are not raw data. It is not necessary to retain or archive these documents after the final study report is signed. The study pathologist is solely responsible for making any changes resulting from the peer review process, initiating the audit trail for microscopic observations, and signing the study pathologist’s final report. As a result, the peer-review pathologist is not considered a principal investigator for multisite studies and is not required to sign a compliance statement.
Pathology peer reviews conducted after studies have been completed and the study pathologist’s report has been signed often are performed to address specific concerns. These peer reviews should be designed and conducted to achieve a particular goal, and the materials examined and documentation will vary based on the peer review objectives. All changes to the study data and interpretations and all discrepancies with the original pathology data and study pathologist’s report should be documented. If the original data are modified, the changes and the reasons for each change should be recorded in the audit trail. The peer-review objective and process should be documented and signed by the peer review pathologist. In some cases, a second independent report issued by the peer-review pathologist is appropriate. Changes to study data or interpretations resulting from peer review conducted after a good laboratory practices (GLP) study report has been finalized are considered raw data and should be documented in a study report amendment.
Introduction
Pathology evaluations in nonclinical toxicology and carcinogenicity studies are composed of two critical steps: (1) the accurate diagnosis and recording of all pathology findings and (2) the integrated interpretation of all pathology information within the context of the entire study to identify and characterize treatment-related findings. Unlike body weights and organ weights, food consumption, pharmacokinetics, and most clinical pathology data, gross and microscopic pathology interpretations are subjective and qualitative. Microscopic interpretations in particular are developed, refined, and harmonized throughout the entire microscopic assessment and construction of the pathology report. Consultation commonly is used to further refine diagnostic terminology and criteria. This reiterative evaluation and revision to define and explain the entire spectrum of treatment-related microscopic findings justifies delay of audit trail initiation and data locking for microscopic observations until the study pathologist’s report is completed.
During pathology peer review conducted before a study is completed, the peer-review pathologist assists the study pathologist with the assessment and refinement of microscopic observations, appropriate terminology, all relevant study data, and the draft pathology report using published literature, historical understanding of the test article and related compounds, and knowledge of additional experts as needed. Ultimately, these interactions verify and improve the quality of the final recorded pathology observations as well as the interpretation and integration of pathology findings within the study pathologist’s final report. This article primarily addresses peer-review before the study report has been finalized and signed, with limited discussion of pathology peer review after a study has been finalized.
Technological advances in pathology assessment and communications together with evolving expectations within regulatory agencies have stimulated recent global discussions of best practices for pathology peer review (Barale-Thomas and Bradley 2009; Organisation for Economic Co-operation and Development [OECD] 2009; McKay et al. 2010). The Society of Toxicologic Pathology (STP) and many individual toxicologic pathologists have published recommendations and commentary on the topic of pathology peer review over the past two decades (Eighmy 1996; Frantz 1997; Mann 1996; McCullough et al. 1997; Morton et al. 2006; Sahota 1997; STP 1991, 1997; Tuomari et al. 2004; Ward et al. 1995). Collectively, these publications outline the evolving perspectives regarding the purpose of the pathology peer review, the role of the peer-review pathologist, and the most appropriate conduct and documentation of the peer-review process. In the present document, prior recommendations are consolidated and expanded to define current preferred pathology peer-review practices for nonclinical toxicity and carcinogenicity studies.
Historical Perspectives on Pathology Peer Review
An early objective of pathology peer review was “to assure the development of accurate pathology data which clearly and cogently support the scientific conclusions” (STP 1991). In general, pathology peer reviews occur before finalization of the study report (Isaacs 2007; Ward et al. 1995) because the peer-review process is integral to refining and improving the accuracy of the final pathology diagnoses and interpretations. Completing the peer review before the study pathologist locks the database not only reduces the documentation required to obtain a high-quality final product (Crissman et al. 2004) but also promotes incorporation of changes that more accurately reflect the findings in a study. The interim notes and worksheets of the study pathologist and the peer-review pathologist are not essential for the reconstruction and evaluation of the pathology section of the final study report and thus are not considered raw data. Therefore, it is not necessary to retain these interim notes and worksheets once the consensus pathology data are incorporated into a signed final report (STP 1997; United States Federal Register 1987; Ward et al. 1995). The single overriding principle of the pathology peer-review process is that it should facilitate “direct interaction between the original and the reviewing pathologists” to produce agreement regarding the nature and interpretation of the pathology data and the final pathology report (STP 1991).
Current Perspectives on Pathology Peer Review
The Purpose and Extent of Pathology Peer Review
The primary objective of a complete pathology peer review performed before the study is completed is to assist the study pathologist to refine, verify, and improve the accuracy and quality of the final pathology data and interpretations. In executing this function, the peer-review pathologist should examine all microscopic slides from a sufficient number of high-dose animals to generate confidence that all significant microscopic treatment-related findings have been identified. (Throughout this article, treatment-related [or test article–related] findings are defined as direct or indirect [secondary], adverse or nonadverse changes caused by the test article, device, or other intentional manipulation studied in the animal model.) confirm that microscopic diagnoses are consistently recorded, terminology and grading criteria are consistently and appropriately applied across the entire pathology data set, microscopic findings are appropriately correlated to necropsy findings, and pathology findings are appropriately correlated to causes of clinical signs, morbidity, or death. examine all target organs from all control animals and all treated animals down to and including all target organs from all animals at the NOEL for each treatment-related microscopic finding. In rodent studies with high incidences of treatment-related findings within specific organs, it may be sufficient for the peer-review pathologist to evaluate a subset of animals in affected groups to confirm the target organ finding, nomenclature, and consistency of severity modifiers. ensure that the study pathologist’s draft report accurately reflects and is supported by the individual animal data tables and summary tables, identifies treatment-related microscopic observations and the dose groups in which these findings occur, and integrates these findings with treatment-related in-life findings, clinical pathology results, organ weight findings, gross necropsy observations, and available ancillary pathology findings (e.g., immunohistochemistry, EM) within the context of the entire study.
Selection of Studies for Pathology Peer Review
Few regulatory agencies have issued guidelines mentioning pathology peer review. The only clear regulatory requirement for pathology peer review of nonclinical studies supporting pharmaceutical registration is found in the European Medicines Agency (EMEA) guidance on the conduct of carcinogenicity studies (EMEA 2002). Draft guidance under consideration by the OECD emphasizes the role and value of pathology peer review but does not provide detailed recommendations on methodology (OECD 2009). The U.S. Environmental Protection Agency (EPA) requires peer review of pathology data and confirmation of changes by a PWG for all submissions requesting reconsideration of carcinogenicity decisions for chemicals based on changes in the pathology diagnoses (EPA 1994). Despite the paucity of mandated requirements for pathology peer review, sponsors and regulatory agencies generally acknowledge that pathology peer review can increase confidence in the pathology data and pathology interpretations.
In the absence of a clear requirement by regulatory agencies for pathology peer review, the decision to peer review each study should be made by the sponsor on a case-by-case basis, considering the importance of decisions that will be based on these studies, the experience and skill of the study pathologist, and regulatory requirements. Pathology peer review is recommended when important risk assessment or business decisions may be based on pathology interpretations in nonclinical studies. Pathology peer review usually is appropriate for GLP toxicity and carcinogenicity studies and biomarker qualification studies and may add value for mechanistic and investigative studies with pathology endpoints and exploratory (non-GLP) toxicity studies that guide compound development decisions and dose selection.
Who Should Perform the Peer Review?
The individual(s) primarily responsible for performing pathology peer review must have the training and experience to competently evaluate all pathology data within the context of the entire study (Bolon et al. in press). Peer-review pathologists should be experienced with studies using the same species and of similar duration and design as the study to be peer reviewed. Experts engaged to consult on specific topics are not required to have the qualifications of a peer-review pathologist who evaluates the entire study, but their contributions should be limited to their areas of expertise.
Pathology peer review may be performed by a pathologist within the same organization or company, the sponsor’s pathologist, or a third party. When appropriate, it may be of value to have several individuals with specialized expertise (clinical pathology, special knowledge of a given organ system, ultrastructure, etc.) participate in the peer-review process. Regardless of their affiliation(s), peer-review pathologists assist the study pathologist to verify and refine the diagnoses and interpretations to be included in the study pathologist’s report.
Peer Review of Studies at CROs by the Sponsor’s Pathologist (Sponsor Peer Review)
The sponsor’s peer-review pathologist frequently has access to information regarding target biology, pharmacology and pharmacokinetic data, mechanisms of target-mediated and off-target toxicity, previous findings with the same compound, supporting scientific literature, and class effects with similar compounds that is not available to the CRO’s study pathologist or an independent peer-review pathologist. This additional knowledge significantly improves the quality of the pathology peer-review process and the final pathology interpretations. Regulatory agencies may have concerns that the sponsor’s peer-review pathologists and others within the sponsor’s organization may inappropriately influence the study pathologist’s interpretations, especially when the study pathologist is employed by a CRO (McKay et al. 2010). Sponsors and their pathologists have little to gain and much to lose by misrepresenting pathology findings. The sponsor’s long-term interests are best served by ensuring that data interpretations in all studies are of the highest possible quality and that the peer-review process is rigorous and objective so that the most advantageous business decisions can be made.
Responsibilities of the Peer-Review Pathologist and the Study Pathologist
The peer-review pathologist is responsible for the design and conduct of the peer review within broad guidelines defined by standard operating procedures. “It is the responsibility of the peer review pathologist to ensure that the method of review employed is sufficient to verify the accuracy of the histopathologic findings” (STP 1997). The peer-review pathologist designs the process, often in consultation with the study pathologist. The approach to the pathology peer review and the materials selected for review may vary based on institutional procedures, the objective of the peer review, the study design, and/or the study pathologist’s findings. The peer-review pathologist should examine sufficient data to ensure identification and interpretation of treatment-related findings in each dose group. This applies to the interpretation of all pathology data and is not simply limited to evaluation of the microscopic slides. The peer-review pathologist may benefit from discussions with the study pathologist prior to peer review and should have access to the study protocol and protocol amendments, mortality information, in-life observations and interpretations, clinical pathology and organ weight data (with means and results of statistical analyses, if performed), all microscopic slides, the study pathologist’s microscopic diagnoses for each animal, appropriate summary data tables, the draft pathology report narrative, toxicokinetic data (if available), and other relevant data. For routine peer review of all pathology findings, the peer-review pathologist usually evaluates the individual and summary data tables and interpretations for clinical pathology, organ weights, macroscopic observations, microscopic findings, and ancillary pathology information (results of electron microscopy, special histochemical and immunohistochemical stains, in situ hybridization or in situ polymerase chain reaction information, etc.).
The peer-review pathologist should review the pathology terminology used and the draft pathology narrative report. In-life observations and toxicokinetic data may be reviewed if appropriate. When the peer review of clinical pathology data is conducted by an individual other than the pathologist responsible for the peer review of gross and microscopic findings, the peer-review pathologist(s) should review enough clinical pathology data to ensure that clinical pathology findings are appropriately integrated with other study data in the narrative of the study pathologist’s report. Correlation of pathology findings with in-life data and causes of deaths should be reviewed. If a focused or targeted pathology peer review has been requested to address a specific concern or issue, sufficient data must be reviewed to fully evaluate the concern or issue.
The peer-review pathologist and study pathologist should discuss all aspects of the data and their interpretation and then come to general agreement that the target organs have been appropriately identified; nomenclature for treatment-related findings is clear, well defined, and consistently applied; the affected dose groups for each treatment-related finding are clearly specified; and related findings have been appropriately integrated in the study pathologist’s narrative. Ultimately, the study pathologist is responsible for all pathology data and interpretations in the final pathologist’s report. Any changes to study data and the pathologist’s narrative report based on the peer-review consultation are made only by the study pathologist. After all changes to the pathology data and draft pathology report resulting from pathology peer review have been made, the study pathologist initiates the audit trail (“locks” the data) and signs the separate final pathologist’s report or an integrated study report containing pathology findings. In Japan, it is common for the pathology data and the pathology narrative to be incorporated within the study report without a separate pathologist’s report, and the study pathologist may not sign the study report. In this situation, the study pathologist should retain responsibility for the pathology data and interpretations and should initiate the audit trail of the microscopic data after changes resulting from peer-review have been made. In the past, Japanese regulators have expected the audit trail to be initiated before the peer-review process began. Japanese authorities are considering revisions to expectations for the pathology peer-review process.
The study pathologist is not obligated to agree with all suggestions of the peer-review pathologist. It is also not necessary to agree on every diagnosis or every severity classification if these discrepancies do not significantly alter the overall interpretation of the pathology data. Discrepancies in terminology or classification for microscopic findings that are unrelated to treatment and minor differences in incidence and severity grading of treatment-related findings are to be expected and are acceptable.
Review of All Organs in a Subset of High-Dose Animals in Rodent Toxicity Studies
In rodent toxicity studies, all organs from at least 30% of high-dose animals per sex from the treatment phase should be reviewed to verify treatment-related findings and to generate confidence that no treatment-related findings have been overlooked. Thus, in rodent studies with 10 animals per sex per group necropsied at the end of treatment, all organs should be examined in at least 3 animals per sex from the high-dose group. In chronic rodent studies with 15 animals per sex in the treatment phase, examination of all organs in 5 high-dose animals per sex is recommended. In complicated studies with multiple routes of delivery or administration of combinations of multiple compounds, there may be more than one high-dose group. Review of all organs in a subset of recovery animals should not be required and is not routinely performed.
Some companies recommend that the peer-review pathologist evaluate at least a certain percentage of the total number of animals from the treatment phase of the study (e.g., 10%). In this case, the number of animals reviewed in the high-dose group is not specified. This strategy is acceptable only if the peer-review pathologist evaluates sufficient high-dose animals to provide reasonable assurance that all target organs have been identified. If this approach is used, the peer-review pathologist often will need to examine all organs in at least 30% of high-dose animals to confirm all treatment-related findings with reasonable confidence. In some cases, three animals per sex are insufficient to confidently identify all target organs. Ultimately, the peer-review pathologist should examine all organs in enough animals to be confident that all target organs have been identified.
Review of All Protocol Organs in a Subset of High-Dose Animals in Nonrodent Toxicity Studies
Nonrodent toxicity studies often use only 3 or 4 animals per sex in each dose group, so peer-review examination of all organs in at least 50% of animals (at least two per sex) of the high-dose group in the treatment phase is recommended for adequate peer review. It is common in nonrodent studies to review all organs from all high-dose animals sacrificed before or at the end of the treatment period.
Should All Protocol Organs Be Examined in a Subset of Control Rodents or Nonrodents?
The decision to examine all protocol organs from a subset of control animals and the number of animals to be examined should be determined by the peer-review pathologist based on review of the pathology data tables, microscopic findings in treated groups, the potential value of this examination, and applicable standard operating procedures. Review of all organs from a subset of control animals should not be required for all studies.
Confirming No-Effect Levels by Review of Target Organs
In both rodent and nonrodent toxicity studies, the peer-review pathologist should examine each target organ in the affected sexes from all control animals and from all animals in the highest dose group lacking the treatment-related finding in order to confirm the NOEL for each finding. In addition, the peer-review pathologist usually examines all target organs in all animals of the affected sexes in all affected dose groups. In some rodent studies, examination of target organs in a subset of animals (50% or more) in affected groups may be sufficient to adequately confirm the findings. If target organs are reviewed in a subset of rodents in affected groups, the review should concentrate on animals with treatment-related findings noted by the study pathologist. If only a few animals are affected, all affected animals identified by the study pathologist should be reviewed.
If the study includes recovery groups, target organs should be reviewed in all recovery animals of the affected sexes in the control group and in groups with treatment-related microscopic findings identified in the treatment phase. Review of target organs in recovery groups receiving the test article that did not have treatment-related microscopic findings at the end of the treatment phase should not be required.
Review of All Organs in Additional Animals when Mortality Is High
When more than 50% of the animals in the high-dose group die or are terminated prior to the end of the treatment period, the highest surviving dose group with at least 50% survival usually is reviewed as described above for the high-dose group. The peer-review strategy for the group(s) with high mortality should be determined by the peer-review pathologist, study pathologist, study director, and sponsor with consideration of the duration of survival, the length of the study, and other study-specific factors; however, within the broad objectives of the peer review, the ultimate responsibility for selection of the materials to be reviewed rests with the peer-review pathologist. Peer review of additional organs in dead or moribund animals in any dose group may be helpful in confirming the cause(s) of death or moribundity and evaluating treatment-related findings at lethal doses.
Pathology Peer Review in Rodent Carcinogenicity Studies
Peer-review requirements outlined in European guidance for 2-year rodent carcinogenicity studies involving pharmaceuticals include examination of slides for all target organs, review of all organs in 10% of animals in all treated and control groups, and review of 10% of all neoplasms (EMEA 2002). These requirements do not address peer review for alternative 6-month carcinogenicity models using genetically modified mice.
This article recommends a different strategy for pathology peer review of rodent 2-year or lifetime carcinogenicity studies and 6-month studies designed to assess carcinogenic potential in genetically modified mice. The rationale for this recommendation is to ensure that an adequate amount of material has been examined to verify treatment effects and affected dose groups. Examine all protocol organs in at least 10% of high-dose animals in each sex in 2-year studies and at least 5 high-dose animals/sex in 6-month alternative mouse model studies. Peer review of all organs from a subset of animals in carcinogenicity studies does not contribute meaningfully to the evaluation of carcinogenic potential but may assist in the identification of nonneoplastic, treatment-related findings. Review of all organs in some control animals may be performed at the discretion of the peer-review pathologist to accurately distinguish nonneoplastic, treatment-related findings from incidental background observations and to comply with regulatory expectations. Routine review of all organs in selected animals in the low- and intermediate-dose groups is not recommended. Examine all neoplasms identified by the study pathologist. Some neoplasms may be difficult to classify, and multiple possible diagnoses may exist for similar lesions. For all neoplastic findings, standardized terminology must be used correctly and consistently, and “Fatal” and “Nonfatal” designations must be assigned (taking into account all findings in each animal). Misclassification or reclassification of one or a few neoplasms in a study may change the pathology interpretation by altering the statistical and qualitative identification of treatment-related neoplastic findings or affected dose groups. Therefore, peer-review evaluation of every neoplasm in every animal in all groups is considered best practice. Examine all organs in all animals of all dose groups of affected sexes with observed or suspected treatment-related neoplastic findings in any dose group. Examine target organs with nonneoplastic, treatment-related findings in all control animals, all animals of the highest dose group lacking the finding (to establish the NOEL for the finding), and sufficient animals in affected groups and sexes to verify the finding. If target organs are reviewed only in a subset of animals, at least 30% of animals in the affected groups should be reviewed. The review should concentrate on animals with target organ findings noted by the study pathologist. If only a few animals have a specific treatment-related microscopic finding, all affected animals with this finding identified by the study pathologist should be reviewed. Examination of apparently incidental, nonneoplastic proliferative findings should be made at the discretion of the peer-review pathologist based on knowledge of the test article’s toxicity profile and mechanism of action, review of pathology data tables, and consultation with the study pathologist. It may be appropriate to examine nonneoplastic proliferative lesions considered precursors of neoplasia (e.g., atypical hyperplasia, some forms of focal hyperplasia, and foci of cellular alteration) to confirm that these lesions are not neoplasms. However, examination of very common proliferative lesions without evidence of a treatment effect (e.g., adrenal gland spindle cell hyperplasia and endometrial hyperplasia in mice, pituitary gland adenohypophyseal hyperplasia, and adrenal gland medullary hyperplasia in rats) adds little value to the peer-review process. In 6-month mouse carcinogenicity studies with positive control groups, review at a minimum all neoplasms in expected target organs in the positive control group.
Resolving Differences of Opinion
In most studies that include a pathology peer review, the study pathologist and peer-review pathologist should reach consensus on target organs, terminology for treatment-related findings, dose groups with treatment-related findings, and other major conclusions within the study pathologist’s report. When differences of opinion regarding pathology findings or interpretations exist that could affect risk assessment, informal consultations and/or formal (documented) consultations with additional experienced pathologists and/or subject matter experts often lead to resolution of discrepancies. If consultation results in agreement by the study pathologist and peer-review pathologist on the major pathology findings and dose groups affected, documentation of the informal consultations with other experts should not be required because consensus has been reached and is documented in the peer-review memo.
If the study pathologist and peer-review pathologist cannot reach agreement through informal discussion and consultation during a peer review conducted before a study is completed, a formal (documented) PWG consisting of the study pathologist, the peer-review pathologist, and at least one other pathologist may be created to resolve differences of opinion (Ward et al. 1995). A PWG may also be convened to review the overall pathology interpretations within a study or if the study pathologist and peer-review pathologist seek additional expertise to resolve and document a complicated issue. Additional pathologists and/or subject matter experts may be included in a PWG. If the study has been conducted at a CRO and the peer-review pathologist is employed by the sponsor, membership of the PWG should be balanced, and at least one PWG member (in addition to the study pathologist) should be a neutral party not employed directly by the sponsor. It is appropriate for the sponsor’s pathologists to participate in the PWG because the sponsor’s representatives often possess the most complete knowledge of the history, pharmacology, and toxicity of the test article. The method the PWG will use to resolve differences of opinion should be determined in advance, and documentation of the conclusions should be archived with the study data. The PWG reviews the pertinent data and specimens to resolve interpretations of pathology findings. If consensus agreement cannot be reached within the PWG, discrepancies may be resolved by majority vote.
When Should the Audit Trail of Microscopic Findings Begin when Peer Review Is Conducted before a Study Is Completed?
It is recommended that the audit trail be initiated (i.e., microscopic findings “locked”) after changes resulting from pathology peer review have been completed but before the quality assurance review of the pathology data and draft pathology report. Regulatory agencies and quality assurance units often expect that quality assurance review of the microscopic findings and study pathologist’s draft narrative should occur after the audit trail for microscopic findings has been initiated. Initiating the audit trail just prior to beginning the quality assurance review will clearly demonstrate any changes to the microscopic observations made during or after the quality assurance review while facilitating implementation of changes identified during the pathology peer-review process.
The definition of raw data for microscopic findings varies in different regulatory jurisdictions. The U.S. Food and Drug Administration has declared that only the study pathologist’s signed final report constitutes raw data for microscopic observations. The 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 report of the pathologist comprises the raw data respecting the histopathological evaluation of tissue specimens. (United States Federal Register 1987)
Recommended Documentation when the Pathology Peer Review Occurs before the Study Completion
If pathology peer review is planned prior to completion of a study, the pathology peer review should be mentioned in the study protocol or acknowledged in a protocol amendment. It is sufficient to state that a pathology peer-review will be performed. Detailed protocol methods describing the pathology peer-review process are not recommended, as these should be determined by the peer-review pathologist within broad limits defined by the peer review objectives and standard operating procedures.
The peer-review pathologist should create and sign a peer-review memo (peer-review statement). The peer-review memo should identify the study, specimens, and data reviewed and include a clear declaration that the peer-review pathologist(s) agrees with the overall interpretation of the pathology data for the study. If a standard operating procedure clearly describes the specimens and data to be examined, it may not be necessary to reiterate the materials examined in the peer-review memo. If more than one formal pathology peer review is conducted, peer-review memos should be created for each peer-review process. If a formal PWG is used, a memo documenting the issue or issues addressed, the members of the PWG, the specimens and data examined, and the conclusions of the PWG (signed by all members of the PWG) should be created. An optional signed statement by the study pathologist confirming agreement regarding the study interpretations may be added to a peer-review memo to strengthen the documentation of consensus. The peer-review memo should be archived with other pathology and study records. A copy of the peer-review memo may be included in the study pathologist’s report or the study report, but this should not be required because documentation of data verification and review processes generally are not included in study reports. If the peer review and the study pathology assessment are not performed at the same site, appropriate chain-of-custody records documenting the transfer and inventory of specimens (slides) shipped between sites should be maintained and archived.
The peer-review memo should be signed after all changes resulting from the peer-review process have been made to the microscopic findings and draft pathology report. The exact timing of the signature on the peer-review memo should be selected by the sponsor because peer review is an optional, unregulated process (with the exception of 2-year rodent carcinogenicity studies in Europe) and the timing of the signature does not affect the value of the interactions between the peer-review pathologist and the study pathologist. It is acceptable to sign the peer-review memo at the conclusion of the peer-review process prior to initiation of the audit trail for the microscopic findings and quality assurance review (Isaacs 2007), after quality assurance review of the pathology data and draft report are completed, or after the pathology report or an integrated study report containing the pathology findings is finalized and signed. This timing documents the peer-review pathologist’s agreement with the study pathologist’s final report.
Completing the pathology peer review and signing the peer-review memo prior to quality assurance review (option 1) is consistent with the execution of data verification procedures before data sets and draft reports are submitted for quality assurance review and enables the quality assurance unit to verify that the peer-review process was conducted as described in the study protocol and standard operating procedures. This timing complies with all current published regulatory guidance. In this scenario, the peer-review memo documents the peer-review pathologist’s agreement with the study pathologist’s diagnoses and draft report at the time the peer-review memo is signed. The study pathologist would finalize the microscopic observations and initiate the audit trail for microscopic data soon after the peer-review memo is signed and before the initiation of the quality assurance review. The study pathologist is responsible for ensuring that the peer-review pathologist is consulted regarding proposed changes before the final pathologist’s report is signed. If major changes are proposed after the original peer-review memo is signed, the peer-review pathologist must be given the opportunity to consider their validity and if necessary to reevaluate the pathology data and slides. Consensus following this reappraisal should be documented by the peer-review pathologist via an annotation on the original peer-review memo or by preparing a second peer review memo.
As stated previously, the peer-review memo is a record of data verification and is not raw data. The notes and worksheets created or annotated by the peer-review pathologist or study pathologist as well as records of changes made to the microscopic findings and study report as a result of peer review are also not raw data and do not need to be retained after the study report is signed (STP 1997; McKay et al. 2010).
The peer-review pathologist should not be considered a principal investigator and should not be required to sign (and should not sign) the final study pathologist’s report. In addition, the peer-review pathologist should not be expected to sign a compliance statement because the study pathologist assumes the responsibility for evaluation of pathology data and interpretations in compliance with regulatory guidance (OECD 2002).
Documentation when the Peer Review Occurs after Study Completion
If the peer-review process is conducted after the study pathologist’s report and the final study report are finalized and signed, it is preferable to include the original study pathologist in this review process, but this is not always possible. In some cases, the peer-review pathologist must perform a complete, independent review of all pathology data and prepare a separate signed pathology report. If changes to pathology data or the study pathologist’s narrative are required in a GLP-compliant study after the final study report has been signed, a formal study report amendment is required that documents all changes to the findings and interpretations described in the original study pathologist’s report. A peer-review memo, a PWG memo, or a pathology report documenting the purpose and scope of the peer review, the materials and information examined, and any modifications or disagreements with the original pathology data and report should be signed by all reviewers. If the original microscopic data files are changed, the audit trail should document each change.
Remote Peer Review and Peer Review Using Digital Images
Direct and detailed communication between the study pathologist and the peer-review pathologist is essential to successful peer review. While face-to-face interaction over a multiheaded microscope is ideal, pathology peer review often can be performed successfully at a site remote from the study pathologist. Discussion and reconciliation of findings and wording in the draft report usually can be managed by telephone, e-mail, and/or sharing of written communications and images. When findings are unusually complicated, face-to-face discussion and concurrent review of selected glass slides should be considered to resolve details of the pathology findings.
Direct light microscopic examination of tissue sections on glass slides is the gold standard for histopathologic assessment. Recent advances in the quality of digital imaging technology permit entire microscopic sections to be scanned and reviewed from any location with an Internet connection. Sections may be examined at high resolution over the full range of magnifications typically available to an anatomic pathologist using a light microscope. In principle, therefore, pathology peer review could be conducted using digital images if the peer-review pathologist can effectively evaluate tissue changes using images of each entire specimen represented on the glass slides. The study pathologist and peer-review pathologist(s) at distant locations then might discuss and interpret individual microscopic findings from the same image, often working by teleconference in real time. If the digital image cannot be suitably interpreted, the glass slide(s) should be used to resolve any concerns (Tuomari et al. 2007). Microscopic examination of histologic sections using digital images should become more common over time as appropriate processes are defined and accepted.
Conclusion
Pathology peer review has evolved over several decades of collective experience by thousands of toxicologic pathologists. Despite multiple publications about the peer review process, many details of pathology peer review have not been widely described or universally accepted. This article presents current views on pathology peer review in an attempt to further harmonize and communicate effective pathology peer-review strategies and practices.
It is anticipated that pathology peer-review expectations within companies, research institutions, and regulatory agencies worldwide will continue to evolve and that the recommendations proposed in this article will be further refined in the future. However, the key features of pathology peer review conducted prior to study completion, that is, flexibility to meet changing needs and unique study requirements, the ability of the peer review pathologist to fully explore all possible evidence of toxicity within nonclinical toxicity and carcinogenicity studies, direct interaction of two or more pathologists to ensure the quality and consistency of pathology data and reports, and increased confidence in the data and interpretations, should not change over time. Any future changes in the process should facilitate and enable these current strengths.