Scientific and Regulatory Policy Committee Points to Consider for Medical Device Implant Site Evaluation in Nonclinical Studies

Abstract

Nonclinical implantation studies are a common and often critical step for medical device safety assessment in the bench-to-market pathway. Nonclinical implanted medical devices or drug-device combination products require complex macroscopic and microscopic pathology evaluations due to the physical presence of the device itself and unique tissue responses to device materials. The Medical Device Implant Site Evaluation working group of the Society of Toxicologic Pathology’s (STP) Scientific and Regulatory Policy Committee (SRPC) was tasked with reviewing scientific, technical, and regulatory considerations for these studies. Implant site evaluations require highly specialized methods and analytical schemes that should be designed on a case-by-case basis to address specific study objectives. Existing STP best practice recommendations can serve as a framework when performing nonclinical studies under Good Laboratory Practices and help mitigate limitations in standards and guidances for implant evaluations (e.g., those from the International Organization for Standardization [ISO], ASTM International). This article integrates standards referenced by sponsors and regulatory bodies with practical pathology evaluation methods for implantable medical devices and combination products. The goal is to ensure the maximum accuracy and scientific relevance of pathology data acquired during a medical device or combination drug-device implantation study.

This Points to Consider article is a product of a Society of Toxicologic Pathology (STP) Working Group commissioned by the Scientific and Regulatory Policy Committee (SRPC) of the STP. It has been reviewed and approved by the SRPC and Executive Committee of the STP but does not represent a formal best practice recommendation of the Society; rather, it is intended to provide key “points to consider” in designing studies or interpreting data from toxicity and safety studies intended to support regulatory submissions. The views expressed in this article are those of the authors and do not represent the policies, positions, or opinions of their respective agencies and organizations. Readers of Toxicologic Pathology are encouraged to send their thoughts on these articles or ideas for new topics to the Editor.

Introduction

Medical devices are defined in Section 201(h) of the United States Food, Drug, and Cosmetic Act to include products

intended to affect the structure or any function of the body of man or other animals, and which does not achieve its primary intended purposes through chemical action within or on the body of man or other animals or which is not dependent upon being metabolized for the achievement of its primary intended purposes. ¹

This broad definition includes multiple classes of medical devices from relatively low-risk and low exposure Class I products, to intermediate-level Class II products, to high-risk life-sustaining invasive Class III products produced by complex or emerging technologies (e.g., active implanted medical devices). Importantly, medical devices are tested for both safety (e.g., appropriate tissue and host response [i.e., not adverse] using the final finished form [FFF] introduced in a clinically relevant location) and efficacy (e.g., functionally appropriate response in the tissue and host). In addition, small molecules or biologics may be fully integrated into medical devices, resulting in a FFF as a combination product (drug or biologic and/or a medical device). ² The pathways for testing safety and efficacy of new biomedical products generally include multiple in vitro and in vivo studies, with the in vivo studies ranging from acute to chronic tissue tolerability of a simple biomaterial (e.g., intramuscular implantation) to multi-year implantation of the FFF in an animal model of the intended clinical indication (e.g., cardiac assist device in an animal model of congestive heart failure).

The wide range of medical device products and target organs/tissues has triggered the development of an equally broad range of tissue/organ and device-specific standards and guidance documents. ³ Assessment is further complicated by the ever-evolving range of biomaterials, which may include novel formulations, physiochemical modifications that alter tissue and immunologic responses, self-forming gels, and electronics with integrated software. In addition, the regulatory landscape for medical devices can be complex, with differing worldwide approaches for the classification of, standards for, and approval pathway of medical devices and drug-device combination products.

Pathologists encountering these documents for the first time, particularly the International Organization for Standardization (www.ISO.org) testing standards, can find the number and breadth of published standards to be overwhelming and access to the standards restrictive due to their for-purchase nature. Pathologists may be further challenged by the expectation of rigid application of pre-defined scoring systems for identifying histopathological findings and assigning microscopic severity grades which are described in the currently published standards; these histology grading schemes/rubrics are widely but not judiciously applied within the industry. In most cases, these pre-defined tissue response scores cannot be accurately applied to many medical devices or implant locations, and thus are insufficient to fully communicate the microscopic pathology findings in a study. The purpose of the pathology evaluation is to identify macroscopic and microscopic pathology findings within a study and to interpret them in context of the study design, and it is not possible to define or predict ahead of time all of the possible microscopic findings that can occur at an implant site. Medical device pathologists (often toxicologic pathologists experienced in nonclinical studies) also must have a working knowledge of commonly used terms and definitions that apply to regulation, testing, and evaluation of medical devices. Some relevant terms and definitions can be found in regulatory documents, but there is limited consensus on the definition of most terms.

To provide clarity for its members who work in this field, the Society of Toxicologic Pathology (STP) tasked its Scientific and Regulatory Policy Committee (SRPC) to empanel a working group to review the scientific, technical, and regulatory considerations for nonclinical studies designed to evaluate the safety of medical devices and combination drug-device products. The objective of the Medical Device Implant Site Evaluation working group (MDISE) was to compile a list of major elements to consider when performing the pathology evaluation for medical devices or combination products. As the subject material in this field is broad, the MDISE focused this article on points to consider when conducting macroscopic (gross) and microscopic analyses of target tissues, as a means of evaluating local biological responses to implantable medical devices or implantable drug-device combination products. Particular emphasis was placed on demonstrating that well-established principles of microscopic evaluation as applied routinely in toxicologic pathology are relevant to medical devices and combination products. A second critical goal was to provide key references that detail specific evaluation parameters for individual organ systems and organ/tissue-specific bioabsorbable devices, which have been provided as Supplemental Material. This points-to-consider document highlights approaches for conducting the macroscopic and microscopic pathology portions needed to assess the local safety of implanted materials during a nonclinical safety study. The specific objectives undertaken by the MDISE were as follows:

• To provide background information and practical considerations for medical device implantation studies to pathologists, engineers, sponsors, and regulatory professionals;

• To provide a summary of regulatory documents and considerations for study planning;

• To recognize deficiencies in the pathology methodology and interpretations presented in currently accepted standards; and

• To suggest suitable methods for evaluating, documenting, and reporting the pathology of implantation studies, and considerations for applying these approaches to the ISO-10993-6:2016 ⁴ (the most commonly used standard for implant studies).

Background

Safety assessment has been applied to medical device materials only recently, and thus has evolved along a different path than that of small molecules and biologics. Historically, biomaterials were considered inert and biological/toxicological safety concerns did not bring about regulatory standards until the 1980s. Testing standards/recommendations from different organizations were based on historical protocols for chemical and materials testing protocols, and safety testing procedures were often driven by the considerations and needs of engineers rather than chemists, biologists, or pathologists. Thus, scientific approaches to implant biocompatibility assessment evolved largely as a separate discipline with only recent safety contribution from toxicologic pathologists. As a result of the diverse worldwide regulatory groups responsible for assessments of medical devices and combination products, nonclinical testing procedures for medical devices are generally based on standard testing of physical materials rather than the biological testing guidelines developed globally by national regulatory agencies for drug and biological nonclinical safety assessment. These differences in safety assessment of biomedical materials are amplified when drug and/or biologics are combined with medical devices as combination products.

While the nonclinical testing program for medical devices has many similarities to those for small molecules and biologics, there are notable differences. ⁵ The goal of nonclinical testing for pharmacologically active products is to develop understanding of target organ toxicity and safety margins prior to clinical testing. Uniformity of nonclinical testing and evaluation of pharmacological products has been refined over many decades through publication of guidance documents offered by multinational efforts such as The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) and the Organisation for Economic Cooperation and Development (www.OECD.org). There are also efforts at standardizing nomenclature through the development of globally recognized microscopic terminology such as the Global Open Registry Nomenclature Information System (goRENI [www.goRENI.org]), and the multiple organ- and species-focused toxicologic pathology lexicons prepared under the International Harmonization of Nomenclature and Diagnostic Criteria (INHAND) initiative (https://www.toxpath.org/inhand.asp#pubg). However, accepted microscopic nomenclature has not been produced for medical device nonclinical implant site evaluations, largely due to the diversity and complexity of tissue changes identified across the range of ever-evolving implantable medical devices and drug-device combination products. Consequently, the lack of a standardized nomenclature specific to biomaterials and variable application of the INHAND terminology fosters ambiguity in the microscopic criteria used in pathology evaluations for medical device and combination product nonclinical studies. Ultimately, striving for establishment of standardized nomenclature for medical device studies will require flexibility to accommodate novel biomaterials, device innovation, and the inevitable novel microscopic findings.

Except for the term biocompatibility (the ability of a material to perform with an appropriate host response in a specific situation.),^1,6 consensus definitions for basic terms that apply to medical devices are not readily available. The choice of preferred terms and their denotations and connotations are frequently based on material sciences, engineering, and country-specific laws rather than a pathobiological foundation of tissue responses. Meetings to define consensus terminology by the biomaterial societies, and inclusion of definitions in ISO standards, US Food and Drug Administration (FDA) documents (webpages and legal definitions), and dictionaries are the current available sources for lexical items (Table 1). Where there is no available preferred terminology or clear definition, the MDISE’s collective experience indicates that any vested parties on an individual development team (e.g., engineers, pathologists, toxicologists, regulators, industry sponsors, and contract research organizations [CROs]) should agree on usage and, as appropriate, include definitions for preferred terms within the pathology and/or study reports.

Table 1.

Reference sources for general and selected specific definitions for medical device–related terms.

Source	Terms	Reference number or website
FDA	General terms	www.FDA.gov 7
	Biocompatibility	https://www.FDA.gov 7
	Degradation	www.FDA.gov 7
	Combination product	www.FDA.gov 7
	Final finished form (FFF)	www.FDA.gov 7
	Extractables	www.FDA.gov 7
	Leachables	www.FDA.gov 7
ISO	General terms	All ISO 10993 series standards at www.ISO.org 8
	Searchable website for ISO terms and definitions	ISO Online Browsing Platform at www.ISO.org/obp/ui/
	Biocompatibility	10993-6, 10993-9 a
	Biodegradation	10993-9 a
	Biomaterial	10993-9 a
	Degradation	10993-9, 10993-13, 10993-14, 10993-15, 10993-16, 10993-17 a
	Final finished product	10993-9 a
	Extractables/leachables	10993-16, 10993-17, 10993-18 a
International Union of Societies for Biomaterials Science and Engineering	General terms	9
	Biocompatibility
	Biodegradation
	Biomaterial
	Degradation
Dictionaries	Biomaterial general terms	Szycher 10
		Williams 11

Abbreviations: FDA, US Food and Drug Administration; FFF, final finished form; ISO, International Organization for Standardization.

a

For each ISO 10993 document, please visit www.ISO.org. ⁸

Generally, the currently accepted algorithm for medical device evaluation begins with physical/chemical information (such as material composition and extractables and/or leachables) and often will include assays to assess cytotoxicity, sensitization, irritation, hemocompatibility, and pyrogenicity (bacterial and/or material-mediated endotoxins). ⁶ Dependent on duration and nature of contact, some assessment of genotoxicity, local tissue impacts, and acute through chronic systemic toxicity can be required, although some basic medical devices or materials may be de-risked by physical and chemical properties alone. In addition, considerations for manufacturing/packaging sterilization, particulates and contaminants, and degradation are also included in the risk assessment. When necessary, additional testing for safety pharmacology, reproductive and developmental toxicity, immunotoxicity, and carcinogenicity may be considered. These criteria are broadly addressed in the medical device nonclinical biocompatibility testing standards published by the ISO within the 10993 series. ⁸ The standards most relevant to toxicologic pathologists will be discussed in more detail below.

The timing of medical device safety studies may differ substantially compared with safety studies for most bio/pharmaceutical nonclinical programs. Nonclinical safety evaluations of medical devices are routinely conducted prior to any clinical evaluations, whereas pharmaceutically active products are filed as Investigational New Drug (IND) products that normally use a staged approach of nonclinical testing in a series of progressively longer studies which overlap with the early phases of clinical testing. Consequently, there may be many years before the completion of the required nonclinical testing for pharmaceuticals and biological products. Combination products also may encounter a staged approach whereby a certain amount of nonclinical device testing (in conjunction with testing for the drug or biologic) allows for clinical trials early in development. Another fundamental difference is that nonclinical efficacy studies are generally required for the final clinically packaged, and fully functioning form of a medical device (e.g., use of the FFF such as a wound dressing in a wounded animal model) prior to any initiation of clinical trials, whereas typically, safety is the primary focus for nonclinical assessment of small molecules and biologics. The relevance of identifying dose-dependent effects in device studies is highly study-contingent depending on the device material or presence of pharmaceutical activity. The FFF is frequently considered to be the maximum implantable device (i.e., the device is often not scalable to change the “dose” of the device, without affecting efficacy and/or function). For many functioning medical devices (e.g., pacemaker), a typical dose response is irrelevant, while in certain circumstances, a dose response may be relevant (e.g., implanted biomaterials where various volumes, weights, or numbers of implanted biomaterials may be evaluated).

Typical animal models utilized in safety studies for small molecules and biologics include rodents (mice and rats), beagle dogs, minipigs, and non-human primates (e.g., cynomolgus macaque or rhesus macaques).^12,13 In addition to use of these species, large FFF devices that require efficacy testing in anatomically site-appropriate animal models may require more suitable test animals such as small ruminants (usually sheep), large hounds rather than beagle dogs, mini and full-size pigs, and other less common laboratory animal species. Compared with testing of small molecules and biologics, this wider range of test species and strains has greater genetic and environmental variability compared with conventional toxicology species. This increases the complexity of histopathologic evaluation and interpretation due to the decreased availability of standardized terminology, reduced documentation of normal variability (i.e., incidental background findings), reduced numbers of animals on study, and increased breadth of spontaneous and induced pathology that might be encountered in unusual species that are seldom used to test small molecules and biologics.

The primary objective of the pathology evaluation in nonclinical safety studies for pharmaceuticals/biologics is most often to identify and characterize the systemic dose-dependent morphological responses of both the intended target organ(s) and any off-target organ(s). In these cases, an appropriate assessment may result in a morphological diagnosis of “not remarkable” for many organs and tissues in such studies. Very detailed evaluations of the local impact (the site of primary administration or absorption) are often less critical for small molecules and biologics, unless it is relevant to the route of administration and/or mechanism of action. However, the local implantation site in most device implant studies, including the control groups implanted with reference materials, will merit characterization of microscopic findings using detailed histopathologic descriptions and potentially, selective photographic documentation. Possible systemic effects of implanted medical devices are evaluated, if warranted, by risk level, duration of contact, and site of implantation (such as devices with a higher degree of risk or longer duration of implantation). ⁶ In addition, a balance of local and/or systemic assessment may be more critical for combination drug-device products, as longer term systemic effects of a medical device may be de-risked by methods of extraction (methods which provide potential materials/chemicals from a device under more extreme conditions which do not degrade the device) and leachate (methods which provide potential materials/chemicals from a device by replicating normal physiologic conditions of use) profile analyses, which may replace longer in vivo studies. Virtually all implantable medical devices or implanted combination products result in local tissue changes related to alteration of regional topography due to the physical disruption imparted by the implant, the procedure required for the implant, and any response related to material composition. The cellular/tissue response to an implant reflects alteration of the local biological milieu; this pathophysiological response may vary based on the composition, location, intended purpose of the device. Observed changes do not always indicate that the tissue responses are abnormal or even undesirable. Some medical device–related tissue responses can be actively utilized to support and extend desirable tissue remodeling outcomes (e.g., osteointegration ensuring long-term stability of a bone implant). Just as the flexibility should exist for terminology and grading of systemic morphological findings in drug and biological studies, basic suggested descriptors limited to specific cell types and fixed numerical grades (as noted in ISO 10993-6:2016, Annex E) ⁴ are often not suited to the robust descriptions needed to fully characterize the variety of local tissue responses in medical device implant studies.

Considerations for Medical Device Implant Site Evaluation

Successful evaluation of medical device implant sites requires appropriate planning and standard operating procedures (SOPs). The following points to consider have been devised to make the generation, analysis, interpretation, and communication of accurate and useful pathology data a straightforward endeavor for toxicologic pathologists who participate in safety assessment for implantable medical devices or drug-device products. The points for medical devices are typically also relevant for drug-device combination products; thus, the references below are typical to “medical device.” Where applicable, variations are noted.

• 1. General Approaches for Implant Study Design

• 2. Guidance Regarding Study Pathologist Qualifications

• 3. Organization of the Sponsor and Implant Study Team

• 4. Considerations for the Surgeon/Interventionalist

• 5. Necropsy Procedures for Nonclinical Implant Studies

• 6. Trimming and Histology Processing Procedures for Nonclinical Implant Studies

• 7. Ancillary Imaging/Non-Routine Analyses Relevant to Pathology Evaluation

• 8. Generation, Interpretation, and Communication of Histopathologic Data

• 9. Pathology Peer Review of Pathology Reports for Medical Device Studies

• 10. Regulatory Aspects Relevant to Implant Studies

• 11. Utilization of Early Regulatory Feedback on Device Studies

• 12. Post-Marketing Surveillance of Medical Devices

• 13. How to Work with International Standards for Implant Studies

• 14. Navigating Differences between ISO 10993-6:2016 ⁴ and Standard Toxicologic Pathology Practices for Pharmaceuticals and Biologics

Table 2.

MDISE practical applications, limitations, and advisable modifications of the ISO 10993-6:2016 standard for implant study pathology assessment.

ISO 10993-6:2016 standard sections	MDISE notes and useful modifications to ISO 10993-6:2016 for implant studies based on current nonclinical scientific principles for histopathologic evaluation
4.2 Preparation of samples: Physical features of the control implant should be “matched as closely as reasonably possible,” should represent the final form of the product, and should be used in a manner that represents actual clinical use	• There are different types of controls, and it should be understood if a control is intended to assess tissue reactions to surgical implantation alone vs the tissue reaction to an article similar to the test article.• There is no universal control material, and use of a control material or inclusion of additional control groups (e.g., sham surgery group) should be determined on a case-by-case basis in consultation with the study pathologist.• It may be challenging or impossible to find an appropriate control with similar structure/material composition, but thoughtful selection of one or more controls optimizes interpretation of the tissue responses for the test article. Controls with a similar structure/material composition (e.g., predicate control, competitor control, marketed control) are used to assess the tissue response to a marketed product compared with the novel test article. High-Density Polyethylene or HDPE is a common “negative” or more appropriately “reference” control for surgical implantation of solid materials, and saline is sometimes a “negative” or “reference” control for injected or combination medical devices.• Use of inappropriate tissue response controls (e.g., a non-bioabsorbable implant control vs a bioabsorbable test implant) should be avoided.• An excessive/unexpected tissue reaction to a negative/reference control may indicate a procedural issue (e.g., surgical trauma or contamination problems) unrelated to the specific articles implanted. Conversely, increased tissue responses around reference controls relative to test devices with surface modifications or that elute anti-inflammatory agents may need to be evaluated by inclusion of an additional control group with the reference material plus an injected depot anti-inflammatory.• The study pathologist should be consulted if there is any question regarding appropriateness of control material or the need for other possible control groups, including surgical controls.• Interpretation and communication of the study results should always include the relevance of the control material, note any substantive physiochemical differences between it and the test material, and state whether the control material serves a specific purpose (e.g., biomaterial vs surgical control).
5.1.1 Note: Use of vertical standards to replace ISO 10993-6:2016 when relevant	• The study pathologist should reference the use of an appropriate ISO vertical standard (i.e., outside of ISO 10993-6:2016) 4 when one exists.• Principles of good histopathologic evaluation (i.e., describing and interpreting the results visible in the tissue sections for the current study) should be utilized for microscopic scoring instead of trying to warp current findings to fit pre-specified histopathology scoring rubrics detailed in vertical standards.• Consult the study pathologist regarding study design when needed.
5.3.5 Combination of local toxicity studies under ISO 10993-6:2016 with systemic toxicity studies under ISO 10993-11:2017	• This section of the standard is not clear, and specific details for combining an implantation study (ISO 10993-6:2016) 4 with a systemic toxicity study (ISO 10993-11:2017) 17 are not provided.• Consultation with the study pathologist and toxicologist are strongly encouraged when both local implantation and systemic toxicity studies are to be combined.• The FFF may require use of an animal model of disease or use of non-traditional toxicology species/strains of animals (i.e., non-rodent or large animals) which can make the toxicology and histopathology evaluations more challenging. In addition, the ISO 10993-6:2016 4 recommendation of having either a reference standard control or marketed material control in the same animal along with the FFF to evaluate local tolerability cannot be performed with pharmacologically active combination products or where a potential material-related systemic impact is anticipated. Some challenges with using large animal models for device studies can be the need for frequent body weights, organ weights (with no historical data for comparison), and collection of all ISO 10993-11:2017 17 Tier I and II tissues in large animal models. Frequently, the Tier I tissues are evaluated histologically, and Tier II tissues are collected in case of future evaluation needs.• ISO 10993-11:2017 17 Tier I and II tissue lists do not accurately reflect all appropriate organ and tissue designations that are typical of comprehensive organ and tissue sampling lists used for GLP studies. In practice, tissue sampling as for a GLP study will avoid mistakes in or insufficient sampling of organs and tissues in systemic evaluation of medical devices.• When feasible, combining local and systemic toxicity assessments can reduce animal use, which is in line with animal welfare standards for both ISO and ICH, and may allow reduction of the “parallel study requirements” that are often encountered in combination product development. Combining local and systemic evaluations in a single study is most successful when relevant agencies are involved in the discussion of study design prior to initiation.
5.5.2 Macroscopic assessment	• The methods described in the standard are generally appropriate.• In general, descriptive terminology should be used for recording gross observations and providing incidence tables of gross findings by implant groups. Discrete interpretive terms that have a cellular basis (i.e., “hemorrhage”) should be avoided in favor of descriptive terms (e.g., “Focus, red”) as many gross findings have a common appearance but the actual cellular/tissue change can only be determined microscopically.• Macroscopic photography should be taken of the implantation site when relevant (and as dictated in the study protocol and/or by identification of gross findings at necropsy).• Collection of gross lesions should follow the study protocol and relevant SOPs of the test facility, and when possible, gross lesions should be submitted for microscopic evaluation whether associated with the implant site or not.• As per best practices for toxicological pathology generally, a macroscopic-microscopic correlation should be provided in the pathology data tables.• Discuss device-related gross findings in the report narrative if relevant and indicate their microscopic correlate.
5.5.3 Implant retrieval and tissue sample collection	• The size of tissue samples and number of tissue samples should be determined on a case-by-case basis as the sample size of 2-5 mm recommended in the standard may be insufficient for some implants and/or implant sites.• If microscopic findings routinely extend to tissue margins on the section or the mechanism of action is likely to result in a wide radius of tissue changes, additional tissue evaluation may be necessary.• The pathologist should evaluate tissue-device interface integrity and provide a comment in the pathology report for studies for which the medical device was removed prior to embedding (i.e., is sufficient interface tissue present for assessment of cellular response, or is insufficient interface tissue available due to attrition during device removal). If retention of tissue-device interface information is unknown for a given device, a small pilot study should be done to determine if conventional paraffin embedding following device removal will provide sufficient representation of the tissue-device interface to allow for proper evaluation, or if other methods (e.g., plastic resin embedding) may be helpful to ensure retention of interface tissue.
5.5.3 Implant retrieval and tissue sample collection—lymph nodes	• Responses in regional (draining) lymph nodes and lymphoid tissues are frequently supportive of the local tissue response to an implanted device, so collection should not be restricted to biodegradable implants.• Migration of biodegradable materials is often not demonstrable in lymph nodes as soluble ions and micro to nano particles will not be microscopically visible and/or will pass through the lymph nodes while large particles will not gain entry to circulation.• In addition to draining lymph nodes near the device, collection of contralateral lymph nodes is strongly encouraged (either as a control response or as backup lymph nodes), along with distant superficial or internal lymph nodes and/or systemic lymphoid tissues (spleen, thymus) to permit evaluation of systemic (“control”) lymphoid tissue responses. Implants in or near mucosal tissues should include collection of mucosa-associated lymphoid tissue (MALT). Microscopic evaluation of contralateral lymph nodes, distant lymph nodes, MALT, and/or systemic lymphoid tissues can provide additional information for more accurate interpretation of changes within any draining lymph nodes.• Consultation with the study pathologist is recommended to determine the extent of lymph node evaluation (basic or enhanced) that may be required. Note 1 in ISO 10993-6:2016 4 suggests that draining lymph nodes cannot always be located, but an experienced pathologist can provide the anatomical information on localization of appropriate lymphoid tissues for collection (see Section 5.h of the text for additional information on lymph node and lymphoid tissue evaluations in device studies).
5.5.4 Microscopic assessment	• “The biological response parameters” that are assessed and recorded should at a minimum include an evaluation of cellular responses, and where appropriate epithelial, stromal, vascular and neural responses, and the characteristics of the device (or absence of the device) at the implantation site.• The cellular, stromal, and other tissue responses listed in ISO 10993-6:2016 4 Annex E are factors that can be directly scored (if present) or noted as absent in a foot note or narrative if so desired. Alternatively, the convention within toxicologic (drug) pathology dictates that negative (absent) findings are not recorded if not present, so this approach may also be utilized.• Semi-quantitative evaluation for microscopic findings should be conducted in accordance with current best practices for toxicologic pathology evaluations for assigning microscopic diagnoses and establishing severity grades within an experimental study.44,49,50 • Suggested “biological response parameters” as listed in part 5.5.4 may be utilized if relevant, but this list is not comprehensive or representative of all possible descriptive/diagnostic terminology that may occur at an implant site. The pathologist must use good judgment and training to employ the most appropriate terminology to capture the relevant/appropriate microscopic findings qualitatively and/or semi-quantitatively at an implant site as they occur within a study.• Quantitative values (e.g., micrometers) should not be routinely used without a specific goal, proper methodology, and documentation for quantitative evaluation (e.g., planned measurements of capsular thickness). The pathologist should be consulted for the use of discrete quantitative values and their applicability (or not) to the histopathological evaluation.• Photomicrographs should be taken of the tissue-device interface and other key microscopic findings as warranted (and as dictated in the study protocol) to help explicate key points in the pathology narrative.
Test Methods for Implantation: Annex A (subcutaneous tissue), Annex B (muscle), Annex C (bone), Annex D (brain)	• General concepts are appropriate for biomaterial evaluation at these sites.• These scoring rubrics should not be used as the sole evaluation criteria for large or complicated (i.e., multiple materials, complex technology) medical devices; these require study-specific study design in consultation with an experienced pathologist.
Annex E: Examples of evaluation of local biological effects after implantation, Tables E.1 and E.2	• Major modifications are essential in adapting these rubrics.• “The biological response parameters” that are assessed and recorded should be limited to the cellular/tissue changes that are actually present within the tissues under study, and the report should not be restricted to the parameters listed in Tables E.1 and E.2 as it is not sufficient to capture all possible biological responses at an implant site. The pathology evaluation should report on the relevant changes that are present at the implant site within the study.• A more detailed evaluation and reporting of individual cell populations (e.g., separate neutrophils and eosinophil populations rather than just combined polymorphonuclear cells) can provide insight into the tissue responses when relevant. 34 Giant cells should generally be amended to specify multinucleated giant cells or foreign body giant cells, as a giant cell implies an enlarged cell but not necessarily the syncytial cells with multiple nuclei that are typically found as part of implant reactions.• The term “fatty infiltration” is discouraged due to the lack of consensus on the definition of this term. The general descriptive terms of “Increased cellularity, adipocyte” or “Increased adipocytes” are encouraged to avoid dissent over mechanistic implications associated with “fatty infiltration.” Particularly in skeletal muscle, adipocytes occurring during reparative processes are not infiltrative but are associated with a fibro/adipogenic progenitor cell that differentiates to adipocytes, so “metaplasia, adipocytes” may be the most appropriate term. 51 In addition, identifying an increase in adipocytes may not be relevant in anatomic sites where adipocytes are commonly present (e.g., intermuscular fat pads, subcutis).44,45 • The use of mathematical manipulation proposed in paragraph 4 of part E.1 may not follow suggested statistical methods for ordinal scoring. Trend interpretation by the study pathologist of microscopic ordinal scoring data via incidence/severity tables offers the greatest likelihood of identifying useful cellular/tissue response information (i.e., biocompatibility) in a study. The role of the pathologist is to use their training and experience to identify the actual cellular tissue/changes within the implant site, determine the appropriate microscopic terminology, and set relevant/appropriate grading criteria that are appropriate for each study. The few templated scoring systems in this Annex do not provide sufficient latitude to achieve optimal qualitative and/or semi-quantitative evaluations across a wide range of tissues and medical devices.
Annex E: Table E.3—Example of semi-quantitative scoring scheme	• Major modification is essential for the sample methods detailed in this Annex. The ISO 10993-6:2016 4 standard clearly indicates Table E.3 as an example, but in practice this scoring scheme is routinely expected by sponsors and regulators for a “biocompatibility assessment.”• For simple implant or biomaterials studies, parts F.1 and F.2 in this table may be retained or modified at the discretion of the study pathologist.• Mathematical manipulation (addition and multiplication) of ordinal scoring data is not scientifically appropriate and should never be used due to the context-specific, non-linear nature of these data.44,45 Rather than use of mathematical manipulation, interpretation of trends in the incidence/severity tables by the study pathologist allows for the most accurate and biologically/clinically relevant interpretation of biocompatibility.• Note for non-pathologists: microscopic evaluation of tissue changes in a drug or implant site toxicity study typically is performed at a low objective (2X, 4X, and/or 10X microscope objectives) by a trained and experienced pathologist, although higher magnifications may be used as needed to discern subtle effects.
Annex E.2: Conclusion—“Reaction/Reactivity score”	• Major modification is essential. • Generally, a composite reaction/reactivity score (formerly an irritation score) should not be assigned, particularly for larger implants or FFF devices. The reactivity score does not provide an actual or complete understanding of the biological/clinical relevance of microscopic changes; this biological reaction interpretation is accomplished by a trained pathologist interpreting incidence/magnitude (severity) trends in the ordinal scoring data generated by their microscopic evaluation.• Study conclusions should generally indicate the following: (1) Whether or not microscopic findings are similar between control and test medical devices and discussion of important differences if present; (2) Characteristics of the major tissue responses; (3) Recognition of major findings of biological/clinical significance (i.e., inflammation if deemed important, or extensive necrosis, or temporal features of tissue responses); (4) Comment on biocompatibility under the conditions for that particular study, when appropriate; (5) Other items as needed/relevant for a given implant type at the discretion of the study pathologist.• It is important to recognize that similar microscopic findings between a control article and test article do not equal biocompatibility, particularly if a poorly matched control article is used and serious microscopic findings are present in both groups (e.g., necrosis). The determination of biocompatibility by the study pathologist takes into account the entire study context and provides an indication of relative safety between the test and the control materials.
Alternative to the Annex E.2 “Reaction/Reactivity score”—how to efficiently communicate findings in the study conclusion	• In a straightforward simple biomaterial implant study with incidence/severity trends that are similar between control and test articles, a biologically/clinically relevant conclusion can easily be provided in a few sentences: “Microscopic findings at the implant site are similar between control and test implants at all time points. The implanted material exhibits biocompatibility under the conditions of this study based on (1) microscopic features of a minimal acute inflammation and resolution of inflammation by late time points, and (2) orderly evolution of tissue encapsulation progressing from a thick disorganized field at early time points to a thin, mature fibrous connective tissue capsule by 180d.” This text is only an example of wording; actual study wording should be selected as appropriate within the study by the study pathologist. Strive for brevity whenever possible.• More complex study designs will need more text to clearly describe all relevant test article–related findings and may benefit by inclusion of specific report sub-sections to highlight key points.
Annex E: Table E.4—An example of a histological system—neural tissue response	• Major modifications are essential. Follow recommendations as described above for navigating Annex E: Table E.3—Example of semi-quantitative scoring scheme.
6.6 Macroscopic and microscopic evaluation—use of statistics	• Modification is strongly suggested • Statistical evaluation generally should not be used to analyze ordinal microscopic data.• Instead, biologically relevant trends should be identified by examining incidence/severity tables.• Descriptive statistics (e.g., the group average [mean]) should not be used as the sole factor in interpreting ordinal microscopic scoring data, but such values may be used at the discretion of the study pathologist when relevant (this is uncommon). Incidence/severity tables should always be evaluated in addition to descriptive statistics when possible.• Inferential statistics (even nonparametric) should not be used to assess the relevance of ordinal microscopic data.44,45

Abbreviations: FFF, Finished Final Form (clinically relevant form of a medical device under development); GLP, Good Laboratory Practice; HDPE, High Density Polyethylene; ICH, International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use; ISO, International Organization for Standardization; MALT, mucosa-associated lymphoid tissue; MDISE, Medical Device Implant Site Evaluation working group of the Society of Toxicologic Pathology; SOP, standard operating procedure.

Conclusions

Complex biological responses are induced by implanted medical devices through a combination of topographical, physiochemical, and sometimes pharmacologic (i.e., drug-containing medical devices) influences. These influences paired with the constantly evolving nature and often highly technical engineering of biomaterials makes implantable medical devices some of the most challenging products that a pathologist may encounter. Historically, methods for macroscopic and microscopic pathology evaluation of devices have emphasized inflexible sampling and analytical approaches. The more flexible principles of macroscopic and microscopic evaluation established by toxicologic pathologists for assessing other biomedical products (e.g., pharmaceuticals and biologics) can be used to greatly improve the scientific quality and clinical applicability of device pathology evaluations in medical implant studies. Attention to these practical points to consider when designing and performing implant studies will improve the consistency and standards for evaluation of implanted medical devices.

Supplemental Material

sj-docx-1-tpx-10.1177_01926233221103202 – Supplemental material for Scientific and Regulatory Policy Committee Points to Consider for Medical Device Implant Site Evaluation in Nonclinical Studies

Supplemental material, sj-docx-1-tpx-10.1177_01926233221103202 for Scientific and Regulatory Policy Committee Points to Consider for Medical Device Implant Site Evaluation in Nonclinical Studies by Maureen T. O’Brien, JoAnn C. L. Schuh, Lyn M. Wancket, Sarah D. Cramer, Kathleen A. Funk, Nicolette D. Jackson, Kamala Kannan, Kevin Keane, Abraham Nyska, Serge D. Rousselle, Adrienne Schucker, Valerie S. Thomas and Stefan Tunev in Toxicologic Pathology