Regulatory Forum Commentary *

Abstract

During 2011, International Harmonization of Nomenclature and Diagnostic Criteria for Lesions in Rats and Mice (INHAND) Global Editorial Steering Committee representatives had discussions with representatives of the Food and Drug Administration (FDA) Center for Drug Evaluation and Research (CDER), Clinical Data Interchange Standards Consortium (CDISC), and the National Cancer Institute (NCI) Enterprise Vocabulary Services (EVS) to examine the potential use of INHAND terminology for SEND (Standard for Exchange of Nonclinical Data) submissions to the FDA. The interest in utilizing the INHAND nomenclature, based on input from industry and government toxicologists as well as information technology specialists, suggests that there will be wide acceptance of INHAND nomenclature. The purpose of this article is 2-fold: (1) to provide a brief historical background on the development of SEND and how it is structured and (2) to discuss the impact of SEND on toxicologic pathology and the role of INHAND.

Introduction

Harmonization of nomenclature and diagnostic criteria, especially for rats and mice, has been a long-term goal of pathologists working in the profession of toxicologic pathology. In 2005, the Society of Toxicologic Pathology (STP) and European Society of Toxicologic Pathology (ESTP), in conjunction with Registry of Industrial Toxicology Animal data (RITA), developed a collaborative process to review, update, and harmonize existing nomenclature documents and databases. In 2006, the British Society of Toxicologic Pathology (BSTP) and the Japanese Society of Toxicologic Pathology (JSTP) joined the initiative, making this a global effort. The result of these discussions was the INHAND proposal (International Harmonization of Nomenclature and Diagnostic Criteria for Lesions in Rats and Mice). A Global Editorial Steering Committee (GESC) oversees the activities of the project and the development of harmonized terminology for each organ system resides within Organ Working Groups (OWG), drawing upon experts from North America, Europe, and Japan (Vahle et al. 2009).

Considerable progress has been made with four major systems published to date—respiratory, hepatobiliary, urinary, and central/peripheral nervous systems. These documents are available in Toxicologic Pathology as supplements as well as on a website—www.goReni.org (member access for scientists working in the field of toxicologic pathology, members of any society of toxicologic pathology or of a regulatory agency). INHAND guides not only give toxicologic pathologists diagnostic criteria and guidelines for recording lesions observed in rodent toxicity and carcinogenicity studies, they include representative photomicrographs of lesions and information regarding pathogenesis, along with pertinent references.

During 2011, INHAND GESC representatives had discussions with representatives of the Food and Drug Administration (FDA) Center for Drug Evaluation and Research (CDER), Clinical Data Interchange Standards Consortium (CDISC), and the National Cancer Institute (NCI) Enterprise Vocabulary Services (EVS) to examine the potential use of INHAND terminology for SEND (Standard for Exchange of Nonclinical Data) submissions to the FDA. The interest in utilizing the INHAND nomenclature, based on input from industry and government toxicologists as well as information technology specialists, suggests that there will be wide acceptance of this nomenclature.

The purpose of this article is 2-fold: (1) to provide a brief historical background on the development of SEND and how it is structured and (2) to discuss the impact of SEND on toxicologic pathology and the role of INHAND.

Background

At its core, SEND is a formal mechanism for submitting data from nonclinical studies to the FDA electronically and in a standardized format. To better understand SEND, it is useful to have some understanding of how it developed and its basic structure.

History

From fiscal years 2002 through 2007, the FDA developed regulations and published guidance to improve the consistency of electronic submission of regulatory documents and data. During this time frame, a significant increase in the number of electronic submissions of documents sent to FDA was attributed to the Prescription Drug User Fee Act (PDUFA) III strategy to implement the Electronic Common Technical Document (eCTD) submission format, the implementation of the FDA Electronic Submissions Gateway (ESG), and the implementation of the Electronic Labeling Rule (ELR) and the Physicians Labeling Rule (PLR). The development and publishing of guidance to industry and regulation changes were critical to the success of these initiatives (FDA 2011). During PDUFA IV, the FDA has continued to work with pharmaceutical and chemical industry on increasing electronic submissions. Regulations and guidance to improve consistency of data organization, submission processing, access to documents and data, and evaluation of submission information are in development. Format and data standards are fundamental to the receipt of electronic submissions of data. The FDA wants to work with stakeholders to coordinate the implementation of standards through public meetings, pilot testing, external training, and tutorial sessions.

In order to adopt a standard, FDA first identifies an existing standard that will meet the business need or identifies and works with a well-recognized Standards Development Organization (SDO). For new health information exchange standards, the FDA works within Health Level Seven (HL7), a not-for-profit SDO, dedicated to providing a comprehensive framework and related standards for the exchange, integration, sharing, and retrieval of electronic health information that supports clinical practice and the management, delivery, and evaluation of health services (HL7 2013).

FDA works with its end user community and appropriate SDO or terminology standards maintenance organizations to update standards as needed. The FDA CDER is collaborating with CDISC to bring harmonization to the clinical and nonclinical data submission process. CDISC is a global, open, nonprofit, multidisciplinary organization that has established data standards to support the acquisition, exchange, submission, and archive of clinical research data and metadata. The foundation for the standardized clinical content is the CDISC Study Data Tabulation Model (SDTM). NCI EVS codes, maintains, and publishes all CDISC controlled terminologies, including SDTM, and EVS also supports many FDA terminology standards (NCI EVS 2012).

The SDTM includes nonclinical requirements based on the SEND models that are being harmonized with the SDTM. Since its inception, the SEND team has been involved in FDA pilot efforts. SEND is currently supported by CDER and EVS and will ultimately be a requirement for U.S. FDA submissions. More information on the SEND guidelines is available on CDISC’s website (CDISC.org/SEND).

The FDA has outlined their goals for PDUFA V (FDA 2012a). Section XII of this document deals with standardization of electronic drug application data and electronic submissions and their implementation. It states that FDA, after consultation with stakeholders, will issue draft guidance on these standards and on the format of electronic submissions by December 31, 2012. With publication of version 3.0 of the SEND implementation guide in 2011, development is well underway. FDA will issue the final guidance no later than 12 months after the public comment period on the draft guidelines. Electronic submissions will be phased in from 24 to 36 months from when the final guidance is issued. At the end of this period, virtually all submissions will likely be electronic.

FDA is committed to move quickly toward the goal of fully electronic submission. While work on the nonclinical side had been slow initially, the pace has picked up in recent years. SEND will be a reality in the near future and will impact toxicologic pathologists. We need to be prepared.

Structure of the SDTM and SEND

The majority of subject (animal)-level observations collected during a toxicity study should be represented according to one of the three SDTM general observation classes: interventions, events, or findings. Interventions class captures investigational, therapeutic, and other treatments that are administered to the subject (with some actual or expected physiological effect) as specified by the study protocol (e.g., exposure to study drug). Events class captures planned protocol milestones such as randomization and study completion and occurrences, conditions, or incidents independent of planned evaluations occurring during the study or prior to the study (e.g., room temperature excursions, water system malfunction, etc.). Findings class captures the observations resulting from tests or evaluations such as laboratory tests, body weights, food and water consumption, clinical observations, and so on. Usually the majority of data, which typically consists of measurements or responses to tests usually at specific days or time points, will fit the Findings General Observation Class.

Under the Findings General Observation Class, domains or major categories of data such as body weights, clinical observations, gross findings, microscopic findings, and so on, have been adapted for SEND from the SDTM and include only those categories relevant to nonclinical studies.

A logical extension of the “domain” discussion leads us to the microscopic finding domain. This domain captures the microscopic evaluations/histopathology of the study. This domain should contain at least one record for every scheduled tissue for all subjects (animals) in the study (i.e., if an organ was examined and found not remarkable, it should have a record indicating “UNREMARKABLE”). In addition, animals with nonprotocol required tissues that were examined should also have a record. Subjects (animals) that were not scheduled for examination should not have records unless they were examined. Having a record for each animal in the microscopic domain supports creating incidence tables and statistical analysis on histopathology data. There are a number of associated fields for this domain, including such things as study identifier, unique subject (animal) ID, group identifier, and others along with items related to the actual diagnosis.

SEND and INHAND and Toxicologic Pathology

Within the microscopic finding domain of the SEND software, there will be a field for the original diagnosis as the pathologist enters it into his or her company’s data entry system. The original diagnosis includes the base process term along with any modifiers (e.g., necrosis, acute, centrilobular, moderate). The original diagnosis is always available to the FDA reviewer. The SEND software will also require a microscopic standardized result. This is where INHAND comes in. Based on the discussions INHAND GESC representatives had with FDA, CDISC members, and EVS, the controlled terminology (CT) for this variable will be derived from INHAND. This aligns well with FDA’s interest to adopt or adapt standards that are already in place.

All microscopic findings recorded by the pathologist in a study will be automatically mapped (bucketed) into base processes (e.g., degeneration, necrosis, inflammation, and hyperplasia) and modifiers by SEND-compliant software. Computer system vendors will be offering SEND-compliant modules or sponsors may elect to migrate data using internally developed proprietary software.

The published, individual INHAND microscopic diagnoses are being used as the standardized terminology and are being mapped into appropriate base processes and modifiers, which will be used to populate the CDISC CT lists for the software that will convert pathology data into the standardized format for FDA submission. All standardized terminology and codelists will be coded and maintained by EVS in the NCI Thesaurus (NCIt) and published as a subset of CDISC terminology. The additional information associated with INHAND diagnoses, such as descriptions, pathogenesis, and photomicrographs, will be available to FDA reviewers. If diagnoses other than the INHAND terminology are used by study pathologists, these can still be submitted. However, supplemental information similar to that available with INHAND will not be available to the FDA reviewer. Nor will they become part of the CT lists. They will only be available on a given study. Requests for terminology to be added to codelists can be submitted.

The mapping of the INHAND terminology will be phased in over the next few years:

Phase I (2012–2014)

The INHAND GESC will assist the SEND CT Committee in setting up appropriate categories to segment (bucket) microscopic terminology and help provide appropriate definitions for each category. “Buckets” include the diagnosis or “basic process” (e.g., necrosis) and various types of modifiers such as severity or distribution.

The GESC will map the current published terminology to the SEND codelist (e.g., map the INHAND terminology to the appropriate “buckets” for SEND).

Phase II (Ongoing)

The GESC will become a permanent standing committee of the various STPs with a defined appointment and term of members. It will act as a “clearing house” for comments, requests for updates, and so on, to the INHAND terminology, be responsible for manuscript updates/creation, and coordinate the manuscript updates by collaborating with subject matter experts. The GESC will continue to liaise with global open REgistry Nomenclature Information system (goRENI) and various journal contacts to ensure the basic structure for future endeavors is maintained and act as a “scientific advisory board” to the SEND terminology committee (including EVS) charged with maintaining the SEND terminology.

Phase III (2013–2015)

The GESC will address nomenclature of other species. This will be done on a species-by-species basis rather than an organ system basis. The GESC will establish species-specific committees which will likely include dog, monkey, rabbit, and minipig. Committees would be tasked with providing a manuscript on terminology that would be unique for each species as well as identifying rodent terminology not appropriate for use in the nonrodent species under consideration. This information would be posted on goRENI. Publication as an article or supplement would be determined on a case-by-case basis. New terminology would then be mapped to SEND CT codelists.

Benefits

Standardizing submitted data will result in submission efficiencies such as providing one standard used by sponsors and vendors, reduce review time, and increase reviewer efficiency.

Electronically submitted data for a drug development program, when submitted in a standardized format, can be searched within a study, across studies within a program, or across different programs. It will allow reviewers to communicate their questions more precisely to the sponsor. There will be specialized software for FDA to replicate study tables and graphs (as both line listing and summary tables) submitted by sponsors and view and subset any data (e.g., by dose, time, test, gender, etc.). The FDA, in close collaboration with the NCI Center for Biomedical Informatics and Information Technology (CBIIT), is developing a standards-based repository of subject-level clinical trial data to support regulatory review and patient-centered outcomes research (PCOR). The FDA and NCI CBIIT are committed to open-source, open-access software and tools with the goal of promoting the maximum amount of sharing with industry (FDA 2012b). A similar tool will be available for the nonclinical data.

Pathology remains a descriptive science and there will always be “shades of gray” in the morphologic changes we capture in words. The move toward standardization will be challenging. It is the intent that INHAND will provide standardized nomenclature for the common histopathologic findings observed in toxicology studies but not all findings will be addressed in the current versions of INHAND. With this in mind, SEND, EVS, and the GESC have developed mechanisms so that new terminology can be submitted and incorporated into both the INHAND terminology and the SEND CT lists. In addition, there will always be unusual or unique findings that require new terminology. The current SEND Implementation Guide and CDISC governance provide for these situations.

This collaboration between the INHAND GESC, CDER, CDISC, and EVS heralds a renewed effort toward harmonization of nonclinical data for regulatory submission, enhancing communication between toxicologists, pathologists, and their regulatory counterparts worldwide.