Abstract
Medical devices represent a complex category of medicinal products with varying definitions depending on the regional jurisdiction of regulatory agencies. A common aspect of these definitions is that a medical device is intended to be used for specific medicinal purpose where the primary intended action of the device is not achieved through pharmacologic (or other chemical) means. While regional regulatory frameworks for medical devices are different than for pharmaceutical or biological products, medical device manufacturers are required to evaluate the safety and performance of these products in the context of their intended use. In biological safety evaluation, histopathology plays a relevant role in assessing medical device biocompatibility. This manuscript provides a broad overview of biocompatibility assessment with a deeper look at the role of the toxicologic pathologist in assessing innovative and emerging bone therapies. The content of this manuscript is based on individual presentations delivered at the 2023 International Academy of Toxicologic Pathology (IATP) Satellite Symposium held in conjunction with the Annual Congress of the European Society of Toxicologic Pathology (ESTP) on 26 September, in Basel, Switzerland.
Keywords
Introduction
The International Academy of Toxicologic Pathology (IATP) Satellite Symposium is a 1/2-day scientific session traditionally held with the annual congress of the European Society of Toxicology Pathology (ESTP) meeting. In 2023, the theme of the IATP Satellite Symposium, held in Basel, Switzerland, was “Medical Device Safety Assessment: Toxicology and Pathology Perspectives.” Given that the medical device industry has undergone major changes in recent years, with respect to biological evaluation, the intent of this symposium was to highlight and discuss some of these changes.
The session included a broad overview on medical device safety assessment starting with an historical look at the origin of “biocompatibility,” through the transition from a “checklist” to a risk-based approach and, finally pointing out the increased interest in the development of alternative methods to address biological endpoints in compliance with the 3Rs principle (refine, reduce, replace). In this respect, chemical characterization and toxicological risk assessment of medical devices represents the first step for medical device safety assessment. Key changes in the ISO 10993-guidelines series dedicated to chemical characterization and toxicological risk assessment have been described in one presentation, with focus on providing information for the design, execution, and risk assessment of extractable/leachable studies. Challenges might arise during medical device safety assessment, either from chemical testing/toxicological assessment or from in vitro/in vivo biological endpoints, especially in the pathology field. In this symposium, a comprehensive analysis was presented on the history, challenges, lessons learned, and next steps in one specific area of medical device pathology, that is, bone implant. Finally, the regulatory framework for medical device assessment was presented, with particular focus on Europe, that is, the Medical Device Regulation (MDR).
This article provides synopses of all presentations, including representative figures, selected images, and major points of discussion.
Medical Device Toxicology: Modern Day Safety Evaluation for the Potential Biological Risks Associated With Medical Devices
The safety of a medical device is established through a rigorous process of balancing the benefits of the device for its intended use against the potential risks of such use. Dr. Nicole Soucy (Boston Scientific) used a generic fuzzy logic model associated with the sorites paradox (paradox of the heap) to introduce the risk management framework that has evolved for the evaluation of medical device safety. Within this risk management framework, the key question being answered is: when is a device ‘biocompatible’?
As medical device regulations began proliferating in the 1970s, there was a need to establish a harmonized means for determining if these devices, and the materials they are manufactured from, were safe. The Tripartite Biocompatibility Guidance, an agreement between the United States, the United Kingdom, and Canada, which borrowed heavily from methods established by the U.S. Pharmacopeia for the classification of plastics used in pharmaceutical container closure systems, was the first attempt at harmonization. 21 This framework formed the skeleton of the first edition of ISO 10993-1 published in 1992, which is the internationally harmonized basis on which medical device safety is established today. While this series of standards has evolved over time, the actual test methods themselves largely remain unchanged from these origins, which Dr. Soucy highlighted as causing challenges in modern times as alternative test methods are being developed to replace these older, and often unvalidated, models.
Historically, an evaluation of “biocompatibility” has been a key aspect of medical device safety evaluation and is defined in ISO 10993-1 as “ability of a medical device or material to perform with an appropriate host response in a specific application.” 23 The limitation of this definition is that it is interpreted to mean that a material should have the ability to reside in the body, often for a long time, with only low levels of foreign body response. 51 Modern day medical device safety evaluation involves a comprehensive evaluation of the potential adverse effects of the device, inclusive of its materials of construction and manufacture and an assessment of the overall patient safety risk related to the intended use of the device. This assessment is more appropriately aligned with the definition of an adverse effect in the field of toxicology which evaluates “a change in response to a stimulus, which adversely affects the performance of the whole organism or reduces the organism’s ability to respond to an additional environmental challenge.” 38 Therefore, Dr. Soucy posited that “Medical Device Toxicology” is a more appropriate description of the overall biological safety evaluation process for these products.
Given that medical device regulatory frameworks are based on principles of risk management, the biological evaluation of a medical device follows a tiered approach based on where and for how long the device is used. This is called device categorization and is driven by the nature and duration of contact. The nature of contact is categorized based on where the device interfaces with the body, such as intact skin, breached or compromised surfaces, externally communicating, or implantation in tissue, bone, or in contact with blood. For each of these categories, there is further subcategorization based on the duration of device contact across three durations: <24 hours, <30 days, or >30 days. As the nature of contact increases from lower to higher risk, and as the duration of contact increases from shorter to longer periods, the biological endpoints requiring evaluation also increase. As an example, the biological endpoints relevant for a device implanted in the body, having contact with blood for greater than 30 days would include cytotoxicity, irritation, sensitization, material-mediated pyrogenicity, implantation effects, hemocompatibility, an evaluation of systemic toxicity (acute, subchronic, and chronic), genotoxicity, and carcinogenicity. 23 Historically, the list of endpoints relevant for biological evaluation represented a “checklist of tests” to be conducted. Since 2009, however, these represent endpoints which must be evaluated, often through a combination of information and testing.
The information relevant to this overall evaluation begins with characterization of the physical, morphological, and chemical characteristics of the device. It is also important to understand where in the anatomy the device will be placed, how many devices would be used in a single patient, and if there is anything specific about the patient population that makes them particularly sensitive (e.g. neonate or child). With this information in hand, the assessor identifies the potential biological hazards associated with the substances, estimates the potential biological risks as assessed through scientific rationale and/or testing, and finally determines the overall acceptability of these risks as they relate to the intended clinical use of the device. As objective evidence, this evaluation is documented with overall biological safety conclusions for each medical device. Within the risk management framework, this conclusion must consider the risk profile of the device balanced against its intended clinical benefit.
In recent years, there has been a strong interest in refining, reducing, and replacing animal tests for the evaluation of medical devices. Ex vivo and in vitro models have been or are actively being validated for the direct replacement of certain in vivo tests, existing preclinical studies in large animal models are being refined to address additional biological endpoints, and analytical extractables methods are being used to allow for evaluation of toxicological risks related to potential device leachables. Each of these are essential to modern medical device safety evaluation. Dr. Soucy discussed the development and validation of in vitro irritation methods as an example of the challenges that arise when alternative test methods are developed to replace “gold-standard” in vivo methods that were never formally validated to begin with. Here, the in vivo rabbit method is known to be subjective with poor predictive accuracy. 52 An in vitro alternative test method, using reconstructed human epidermis (RhE) models has been formally validated through an international collaborative effort 10 and has been adopted as the preferred biological method in ISO 10993-23. 27 Despite wide-spread global adoption, some regulatory authorities still expect to see the testing conducted in accordance with the in vivo methods, including the U.S. Food and Drug Administration (FDA). 55
In her concluding remarks, Dr. Soucy highlighted that while the ISO 10993 series of standards are globally harmonized, individual regulatory agencies across the globe often have unique interpretations of these standards which can complicate the ability of multinational corporations in ensuring that the overall assessment of a given device is suitable for every region of the world.
Extractable and Leachable Studies for Medical Devices: Study Design and Toxicological Assessment
To ensure patient safety, knowledge about potential leachable substances associated with medical device materials is critical. For this reason, robust physical and chemical characterization of the medical device is the mandatory first step of the biological evaluation process as defined by ISO 10993-1:2018. 23 During her presentation, Dr. Flora Wegener focused on the regulatory framework which guides medical device manufacturers on toxicological risk assessment within a risk management process.
The extent of this analysis is dependent on the available knowledge of the device including material characterization, nonclinical and clinical safety data, toxicological data as well as information on possible residual process aids or additive used in the manufacturing process. With increasing complexity of medical devices and combination products, a more defined framework to allow identification of potential biological hazard, estimation, and control of biological risk potentially deriving from the constituents of the medical device is required. The recent update of ISO 10993-18:2020 “Biological Evaluation of Medical Devices—Part 18: Chemical Characterization of Medical Devices within a Risk Management Process” provides this framework 26 and describes how to conduct chemical characterization and how to use the obtained information to address a variety of questions which might arise during the development and/or maintenance of a medical device (e.g., evaluation of biological endpoints, screening of potential new materials, targeting specific chemical substances, and establishing material or device equivalence). Importantly, chemical characterization is not limited to extractable or leachable testing alone. As defined in ISO 10993-18:2020 section 3.6 chemical characterization is a “process of obtaining chemical information, accomplished either by information gathering or by information generation, for example, by literature review or chemical testing.” 26
Essentially, chemical characterization is a stepwise approach which starts with qualitative and quantitative information gathering (e.g., from manufacturing site, supplier material) on material composition, followed by a gap analysis and, when necessary, generation of the missing information via chemical testing including extractable/leachable testing (Figure 1). Dr Wegener emphasized the importance of clearly defining the “why” before study execution. The purpose of the study should be defined (e.g. development of new devices, authority-specific request, change in manufacturing), to properly select the type of study (i.e. target analysis, extractable study, and leachable study), the design of the study (e.g. beginning or end of shelf-life, final finished device or single component, extraction condition and analytical methods to be used) and, ultimately to interpret the results (e.g. which biological endpoints will be addressed with the study results).
Stepwise approach to conduct chemical characterization according to ISO 10993-18:2020.
The standard gives guidance on how extractable/leachable studies should be designed to generate a comprehensive understanding of the related risks. Dr Wegener presented an “Extractable/Leachable Roadmap,” and while discussing each step, highlighted the most common issues that might arise in the conduct of such studies (Figure 2).
Representative “extractable/leachable roadmap.”
The analytical evaluation threshold (AET), introduced in ISO 10993-18:2020, 26 converts a dose-based threshold (TTC or Threshold of Toxicological Concern) to a concentration-based threshold (AET); determining the AET is the first step. Extractable/leachables which are below the AET do not need to be identified, quantified, or reported because they are considered to have an acceptable safety risk without further toxicological assessment. Extractable/leachables above the AET require toxicological assessment according to ISO 10993-17 requirements. 25 The formula used to derive the AET takes into consideration both the clinical exposure to the device used (maximal number/day of devices used under normal clinical condition) and the TTC value related to the device categorization, as well as the number of devices used to generate the extract, the volume of the extract and an uncertainty factor (UF) applied to the analytical method. Therefore, a close interaction between analytical chemist—or laboratory conducting the analytical study—and medical device toxicologist is required to adequately set this value.
Following steps include the selection of extraction condition (i.e., simulated, exaggerated, and exhaustive), extraction vehicles (i.e., polar, mid-polar, and not polar), and adequate analytical methods (i.e., gas chromatography/mass spectrometry, liquid chromatography/mass spectrometry, inductive coupled plasma—optical emission spectrometry) to detect both inorganic (mostly metal ions), and organic compounds. The organic compounds include volatile organic (VOC), semi-volatile organic (SVOC), and non-volatile organic (NVOC) compounds. Selection of these various parameters/conditions will depend on the risk associated with the medical device category based on type, duration of contact, and clinical use. Once available, analytical data are provided to the toxicologist for toxicological risk assessment of compounds detected above the AET and is conducted according to the requirements of ISO 10993-17. 25 This guideline was significantly revised in 2023 (from the original 2002 version) to align toxicological risk assessment of medical devices with the current best practices used in other fields of toxicology. Although general principles of toxicological risk assessment remain the same in both versions, new concepts and definitions have been introduced. From a regulatory perspective, a transition period to comply with the new requirements varies based on the regulatory agency reviewing the data.
Dr. Wegener highlighted the importance, for the conduct of the toxicological risk assessment, to establish what potential extractables the patient is exposed to, based on the identification and quantification of the compounds detected in the analytical studies. For identified compounds with sufficient hazard information in the literature, a point of departure (PoD) is selected, and a tolerable intake (TI) is calculated. To characterize the risk, the Margin of Safety (MoS), given by the ratio of the maximal daily exposure to the extractable/leachable and the TI, is provided. When a given extractable/leachable has an MoS above 1, the estimated exposure to such compounds is considered unlikely to present a risk for the patient when the device is used as intended. If data are unavailable, or of limited quality, toxicological evaluation of a compound is performed using the concept of TTC in accordance with ISO 10993-1:2018 23 and ISO/TS 21726:2019 “Biological evaluation of medical devices—Application of the threshold of toxicological concern (TTC) for assessing biocompatibility of medical device constituents.” 29
In this framework, compounds with structural alerts for mutagenic and/or carcinogenic potential are evaluated against a TTC value based on the duration of exposure to the device. These TTC values are defined in ISO/TS 21726:2019 29 and derived from ICH M7. 13 In silico analysis, using both statistical and rules-based models, (e.g. OECD QSAR Toolbox, VEGA QSAR and/or Toxtree5,45,54) are used to detect structural alerts for potential mutagenic or carcinogenic potential. If compounds do not show positive structural alerts, TTC values based on Cramer classification can be applied to evaluate the potential for noncancer toxicity according to ISO/TS 21726:2019. 29 In this framework, high-order toxicity compounds (Cramer Class III) have a TTC of 90 µg/day, intermediate toxicity (Cramer Class II) have a TTC of 540 µg/day and low toxicity (Cramer Class I) have a TTC of 1800 µg/day. 41 Adjustment of Cramer Class TTC values for less-than-lifetime exposure have not been harmonized or universally accepted in the scientific community. Under the current requirements, an overestimation of the risk is possible due to conservatism across the entire process (i.e. conservative AET selection, rigorous extraction condition, worst case exposure assessment). Therefore, it is legitimate to consider that the hypothetical risk estimated may not represent the actual risk of adverse effects in a patient population.
A current challenge in toxicological risk assessment of medical devices is related to analytical detection of “unknown” or “partially identified” compounds. No definitive definition exists for levels of identification, nor is there scientific consensus on what level of identification is appropriate for use in toxicological risk assessment, both are active areas of discussion within the scientific community.
Although guidance exists for design of extractable/leachable studies and subsequent toxicological assessment, it is important to remember that medical devices are complex test articles and difficulties regularly arise (i.e., compatibility of extraction solvent with device material, mismatch between AET calculated and effective limit of detection of the analytical method, high numbers of unidentified compounds, etc.). The appropriate weight should be given to these results within the overall biological evaluation of the medical device. Therefore, the “how” and “why,” when planning an extractable/leachable study, should always be clearly defined and a close collaboration between the analytical chemist and qualified toxicologist should be established to develop meaningful data.
Challenges in Biocompatibility and Regenerative Medicine: Focus on Bone Implant Pathology—a Workup, Shortfalls, and Forecasts
After having outlined a brief history of 70 years of the evolution of biomaterials pathology, starting in the 50s and covering four generations of implant biomaterials (bioinert, bioactive, bioresorbable, biomimetic), Dr. Antoine Alves focused on some challenges in biocompatibility and regenerative medicine and specifically on bone implant pathology.
The increase of an aging population and associated risks such as weakened bones, trauma, and degenerative bone disorders, among other conditions, drive high demand in bone implants. In Europe and the United States, diminished quality of life and survival have been reported following hip fracture with almost one out of three patients over 50 years of age, dying in the first year following hip fracture. 33 Nearly 30% of patients in orthopedic medicine are young, with an increase of 35% in the last decade; risk of revision surgeries is significantly higher with younger age groups. 49 Even though joint replacements are among the most performed surgeries worldwide, implant failure rates can still exceed 10%. 47 To improve the clinical outcomes related to implant failure, active research and innovation are ongoing to optimize the bone implant clinical success.
In 2022, the size of the global bone implants market reached approximately USD 35.3 billion. The annual growth rate of the global bone implants market is prospected to be 6.2% from 2023 to 2030. 46 For context of this posited growth, bone grafting is the second most frequently performed tissue transplantation in human medical practice after blood transfusion. 58
Regulatory authorities and clinical practitioners require a burden of proof supported by preclinical evidence. Pathologists play an essential role in assessing innovative and emerging bone therapies in preventing clinical implant failure, and associated negative outcomes but, in doing so, they face numerous challenges. Understanding the various material properties, the mechanical and biological laws which interplay at the material interface, and other technical challenges such as tissue preparation and imaging can help in better evaluating the implant safety and performance, in accordance with international standards. While there is no international consensus on the terminologies employed in medical device pathology, 57 this does not constitute a major obstacle in medical device pathology practice when evaluating the local effects of the device.
A brief workup drawn from Dr Alves experience in bone implant pathology, shared at the IATP satellite symposium held in Basel, 26 September 2023, is recapitulated here.
Bone Implant Safety Challenges
Local and systemic effects assessment of bone implants and regenerative devices have a central position in the regulatory testing programs. Commonly, macroscopic and microscopic evaluation of safety (defined as the absence of unreasonable risk of injury with the use of the device) are mostly regulated by the standard ISO 10993-Part 6: 2016 “Biological Evaluation of Medical Devices—Part 6: Test for local effects after implantation,” 28 whereas the systemic effects (acute to chronic toxicity studies) are covered by the ISO 10993-Part 11: 2017 “Biological Evaluation of Medical Devices—Part 11: Test for systemic toxicity.” 24 The test device effects (in particular referring to inflammation) are compared to that of a control/predicate device. A predefined microscopic scoring system (ISO 10993-Part 6: 2016 28 ) suggests semiquantitation of the local tissue effects through a Reactivity Ranking Score (RRS). In practice, the test or control Reactivity is calculated by using the following formulae: scores of (polymorphonuclear cells + lymphocytes + plasma cells + macrophages + multinucleated giant cells + necrosis) weighted with a factor 2 + (neovascularization + fatty infiltrate + fibrosis). The RRS is calculated as being Test Reactivity (average score)—Control Reactivity (average score). The RRS value falls then into either a null to minimal reaction (0.0–2.9), a slight (3.0–8.9), moderate (9.0–15.0), or severe reaction (>15.0).
Although the ISO guidance is mostly set forth as recommendations, the study pathologists may face some challenges in the evaluation of the local osseous effects as per ISO 10993-Part 6: 2016. 28 For instance, adipocyte formation (fatty infiltrate) may not be a negative finding, particularly in cancellous bone repair. This is because, in healthy research animals, bone marrow can contain a significant amount of adipocytes. Hematopoietic/stromal cells and adipocytes may also be noted in adult research animals especially during mid to late stages of healing, reflecting suitable osseous restoration process. As such, the fatty infiltrate criteria should be considered with caution among the RRS parameters to avoid disfavoring a therapy promoting faster and better bone (and bone marrow) regeneration. Then the bone marrow formation should be considered as a performance parameter. Instead, lesions such as osteolysis could be added in the ISO semiquantitative table for the RRS calculation. For instance, with biodegradable implants, inadequate design of 3D-printed implants or mechanical factors resulting in stress-shielding, certain xenogeneic materials, inorganic coatings, improper biocorrosion, particulate debris, and their byproducts may all elicit chronic inflammatory reaction and/or active local bone resorption resulting in bone loss (Figures 3 and 4). Adverse tissue reaction with extensive osteolytic lesions associated with bioabsorbable fixation devices was reported in patients following 11 weeks or 4.3 years postoperatively. 6 Clinically, revision rates related to stress shielding-induced bone loss can exceed 50% in total hip replacement. 19 Hence, osteolysis (and not physiologic bone resorption) as a pathologic descriptor, is an important histopathological finding which can be attributed to local tissue responses in the assessment of innovative bone implants.
Alveolar bone reconstruction after tooth extraction in dog. Eight weeks after implantation of an experimental resorbable biomaterial (B), a moderate to marked infiltration of well-defined mononuclear cells including plasma cells characterized the interfacial reactive tissue. No bone growth observed around the biomaterial (B). Suspicion of byproducts release. Un-decalcified resin section, 10-µm thick, obtained with the latest technology for microscopy using a contact-freezer cutting equipment. McNeal tetrachrome stain. Original objective 60X.
Rabbit distal femur, 4 weeks postimplantation showing active signs of osteolysis (red arrows) at the interface of an orthopedic metal alloy implant bearing a specific surface texture. Un-decalcified ground resin section, approx. 50-µm thick section, Modified Paragon surface stain, Original objective 20X.
Another area of needed discussion relates to the importance and inclusion of neovascularization as a key histopathological parameter in the RRS evaluation. Ingrowth of neovessels is generally viewed as a positive factor in bone repair (performance criteria 61 ). High neovascularization scores in a treatment group may unfortunately skew results and generate an unfavorable RRS score, while actually being a component of favorable healing. On the contrary, a low level of revascularization has been reported as being coupled to delayed bone healing. 39 This would translate to lower RSS scoring, although the absence or decreased neovascularization would actually represent a negative outcome. Neovascularization should be removed as a component of the RRS parameters used for scoring and could instead be replaced with pathological lesion that consistently represent detrimental findings such as chronic fibrotic reaction or encapsulation that may result in aseptic loosening.
As with neovascularization, fibrosis is a temporal and important component of the healing process and can progress from granulation tissue which plays a crucial role in early-stage stabilization of damaged tissue, but can progress to scar formation which may be indicative of chronic trauma in the healing process (healing impairment, Figure 5). These findings may suggest a different use of the fibrosis parameter in the ISO 10993-Part 6: 2016 28 scoring system, either among the RRS parameters or among the performance criteria depending on the clinical significance.
Implant failure. After 3 months of implantation in the sheep proximal humerus, the rate of biodegradation of the grafted bone substitute was higher than the rate of the bone regeneration. This event resulted in a poorly repaired and functional bone as shown by the predominant fibrotic scar tissue (blue arrows) and absence of edge-to-edge bone bridging. The yellow line represents the osseous defect margin. Un-decalcified ground resin sections, approx. 50-µm thick section, Modified Paragon surface stain, Original objective 2X.
Another point of investigation is the local toxic effects of bone implants on bone mineralization. Abrogation of the mineralization process is not easily noted in conventional histopathology. In Dr Alves’ test facility (i.e. NAMSA), it has been demonstrated how analysis of bone dynamics following in-life injection of bone fluorochrome markers can improve detection of inhibitory effects of implants on the mineralization that had otherwise been very difficult to detect morphologically. 4 While mineralization is not currently an assessed criteria in the international standards, Dr Alves and colleagues would advocate using degree of mineralization (through bone markers) as a descriptor to confirm the presence or absence of toxic effects on bone mineralization. Mineralization defect may also affect the peri-implant bone function as argued below.
Bone Implant Performance Challenges
Orthopedic implants (primarily referring to fixators), dental implants, and bone graft materials (scaffolds) have a mechanical and/or a biologic function. The fixators used for fracture healing (bone plates, screws, nails, rods, pins, and wires) are mostly comprised of metals or degradable polymers/metals. The scaffolds are generally comprised of autologous, allogenic and xenogeneic bone, or alloplastic materials employed as bone graft substitutes. They present in different forms, including granules, porous blocks, cements, moldable, injectable fluids, and so on. Degradable ceramic materials are the most common biomaterials used as bone graft substitutes for bone defect repair 36 although composite materials and metals are used as well. Depending on the clinical outcome targeted, one category of implants may be more important than the other. For instance, fixators used for fracture healing will serve a predominantly mechanical support function, during the immobilization and bone alignment healing period. Bone graft materials for posterolateral spine fusions provide biologic stimulus for new bone formation, with little mechanical function. Still, bone graft materials may require sufficient mechanical properties for osseous reconstruction in critical sized defects (considered defects that do not spontaneously heal without intervention), and in load-bearing applications, such as in limb salvage procedures or segmental defects.35,36
A complex pathobiological interaction exists at the host-bone implant interface. To evaluate the clinical/biological performance (specific properties of a medical device that are expected to contribute to the treatment of an injury), the study pathologist should be aware of the material properties and bone implant healing mechanisms taking place at the recipient site so as to properly document the study endpoints. From Dr Alves experience, so far, the main material properties, as listed below, have both qualitative and quantitative influences on the bone healing, and are in general found to be interrelated with safety aspects, in the histopathological evaluation of bone implants. A way to histologically characterize the selected material properties is described below as being biological readouts.
Osteoinduction
Osteoinduction is defined as the promotion of undifferentiated cells to form osteoprogenitor cells with production of bone at heterotopic sites such as the skeletal muscle.3,59 It should not be used to qualify the property of a product as being osteoconductive, after implantation in bone, due to the presence of osteocompetent cells. However, it is important for osteoinductive products to be tested as well in specific bone indications for the appraisal of the local and systemic response. For information, BMP-7 a powerful osteoinductive bone morphogenetic protein mixed with a collagen biomaterial and approved in 2001, has been withdrawn off the market for safety reasons.14,20 BMP-2 showed clinical side effects as well. 31
Biological readout
Microscopic extent of the ectopic ossicles formation (Guidance in ASTM 2529-13 2015). X-rays, Micro-CT may help in identifying the radiopaque formation of ectopic ossicles for technical preparations.
Osteogenesis
Osteogenesis refers to the formation of bone either from the recipient bone or from the graft material (containing osteogenic cells or not) as demonstrated by J Davies, 2003, 9 pointing out two distinct patterns of de novo bone formation.
Distance osteogenesis: bone formation originated from the bone defect margins.
Contact osteogenesis: bone formation starting at the implant surface.
Biological readout
Scoring following sequential injection of bone fluorochrome markers (Figure 6) has proved to be reliable in distinguishing and qualifying contact or distance osteogenesis in our experience. 4 Other markers such as alkaline phosphatase and Alizarin red may be used as indicators of the mechanistic event, distance, or contact osteogenic differentiation.
A ceramic 3D printed porous scaffold (PS), inspired from natural osseous trabecules, has been implanted in the sheep distal femur for 4 weeks. Injection of bone fluorochrome markers at day 7 (Calcein Green), day 14 (Xylenol Orange), and day 21 (Oxytetracycline) pointed out no bone-material apposition at day 7. Still an active and dynamic sequential bone ingrowth driven by osteoconduction at day 14 (osteogenesis in orange) and day 21 (osteogenesis in yellow) have been identified. Only distant osteogenesis was observed here. Undecalcified ground resin sections, Epifluorescence analysis, approx. 50-µm thick section, Original objective 4X. Courtesy from Cerhum SA, Belgium.
Osseointegration
Osseointegration is defined by the formation of a direct contact between the living bone and the implant surface (at the microscopic level), without signs of intervening soft tissue. 3
Biological readout
Semiquantitative scoring and/or quantitative histomorphometric measurement of the bone-to-implant contact (BIC%). Mechanical tests such as pushout and detachment tests will measure the fixation strength of the direct bone-implant bonding, to complement the microscopic findings and clinical significance.
Some pitfalls in the evaluation or interpretation may occur when, for instance, a primary osseointegration is observed at early-stage postimplantation, where the direct BIC is a physical contact without having a functional connection. Having a new bone formation visible at the intimate material surface warrants a direct BIC measurement, as well as assessment of implant stability and clinical success. Poor signs of osseointegration, even in the absence of significant local inflammation, would be indicative of an undesirable reaction (representing safety concerns) and should follow with interpretation as to whether or not integration of the bone implant is expected. In different indications, direct BIC may not be desirable. For instance, in osteosynthesis procedure, a bone plate intended to be surgically removed after achievement of a functional fracture repair, should not promote extensive BIC.
Osteoconduction
Osteoconduction is the process by which ongrowth and ingrowth of bone is directed to conform to a scaffold surface so as to bridge a defect (revisited definition inspired from Albrektsson & Johansson, 2001 and Weber 20193,59)
Biological readout
Semiquantitative scoring and/or histomorphometric measurement of the bone ingrowth rate into a scaffolding material (bone area density %42,59) are means to evaluate the osteoconduction property of a medical device.
Osteoconductive materials have the potential of shortening the healing time by bypassing the fibrocartilaginous phase, similar to what occurs in the primary fracture healing process. Bioactive materials (ensuring bone-implant bonding by chemical bonding), such as calcium phosphate ceramics are known to possess higher osteoconductivity properties when compared to bioinert materials (stainless steel material) or low biocompatible materials. The anatomical location, blood supply, implant bed conditions, and device architectural design (surface chemistry, topography, and pores design) have a strong influence on the process of osteoconduction. 42
Biodegradation
Biodegradation refers to the breakdown of a material mediated within a biological system. 62
Biological readout
Semiquantitative scoring and/or histomorphometric measurement of the material area density. The rate of degradation % is measured relative to the initial state of the device. Polarized light enables the detection of birefringent biomaterials particularly at their late stage of breakdown where only a few remnants may be hard to localize.
Regulatory guidelines recommend evaluation of the material degradation at least at 3 time-periods (onset of degradation, peak of degradation, and end-stage degradation) including examination of systemic effects. 57 One of the most critical histopathological challenges is to evaluate material degradation (various patterns of material breakdown) and to determine the differences between the inflammatory response related to the degradation process (normal phagocytic and pro-repair inflammation), and those responses which are reflective of irritancy and pro-inflammatory reactions. These differences may be subject to confounding factors that can challenge the analysis and interpretation of the performance and safety endpoints, in light of the ISO 10993-6.28,53 The profile of the inflammatory reaction (monomorphic or pleomorphic/polytypic) may guide the interpretation and clinical significance of the degradation taking place in the targeted indications. 53 The lack of correlation between the rate of implant degradation and the rate of bone healing (Figure 5) may be the result of adverse event such as implant failure (disappearance of the osteoconductive scaffold, improper degradation/loss of mechanical function), notable inflammatory reaction, osteolysis and fibrosis, or compromise of bone repair or of guided bone regeneration.4,44 Additional investigations may be of interest in determining the nature of the byproducts and material debris observed at the local or remote organs by using Fourier-transform infrared (FTIR) microscopy or energy dispersive X-ray (EDAX) methods for instances.
Biomechanical Factors
Bone is a tissue that changes its density and microarchitecture through adaption to loading conditions; a phenomenon ruled by the Wolff’s law. Bone implants, besides their biological functions, provide a mechanical support particularly at weight-bearing location. Stimulation of osteogenesis and bone remodeling or activation of unwanted bone resorption (osteopenia) depends on the strain/stress transfer between an implant and the recipient bone. When stress transfer is not homogeneous between an implant and the recipient bone (difference in Young modulus or stiffness), this induces a reduction of the bone density in the disused bone by stress-shielding, which may lead to implant failure. 36 Stiffer implants relative to the recipient bone result in frequent issues such as bone atrophy, implant failure, and aseptic loosening.
As such, it is relevant to distinguish bone resorption provoked by the nature of biomaterial (chemistry) from that caused by mechanical factors in a nonconfounding assessment of the safety and performance of the device. Influence of the mechanical factors has often been overlooked in bone implant pathology. In orthopedic therapy, since 2007, the so called Diamond Concept of Giannoudis considered (1) the osteogenic and angiogenic potential of the sites, (2) the osteoconductive scaffold properties, (3) the growth factors and osteoinductive mediators, in addition to (4) the mechanical factors, all working together to favor/enhance the bone healing performance. 2 In the quest for optimal bone devices, new materials with adjusted mechanical properties (new composite materials, 3D additive printing materials, nanomaterials, smart materials, biomimetic materials, regenerative medicine solution) are extensively being developed to solve common issues in bone healing in orthopedics. 36
In bone implant pathology, it is recommended that biomechanical factors be considered in the microscopic evaluation of bone regeneration, through collection of mechanical data of both the tested devices and the recipient bone in the animal model selected, to improve and refine the appraisal of implant effects.
Biological Laws
As recently commented by Vallet-Regi, 56 development of biocompatible substitutes and repair materials has been a leading goal for many years. After years of observations however, evidence has mounted that the physiologic environment receiving the implants must also be taken in consideration to improve treatment outcomes. Choi et al 8 reported that dental implants have a slower rate of integration and a statistically higher rate of long-term failure when placed in the maxillary anterior and molar positions as opposed to the mandible. Following the well-known primary and secondary bone healing mechanisms acting around the biomaterials at various degree, some additional nonexhaustive biological mechanisms governing the bone-implant healing are outlined below:
-
-
-
-
-
1 In the presence of a slow resorbing (or permanent) graft, the amount of the newly formed bone should be inversely proportional to the size of the graft material. Using an osteoconductive bone substitute with two different granule sizes in cancellous bone could demonstrate that.
2 Using one of the bone substitutes indicated above, in cortical bone, the bone growth should fill the free space available at nearly 100%.
The law of the bone capping would result in an equation as readout support:
This is to demonstrate that with an osteoconductive, permanent or slow resorbing scaffold, bone regrowth may translate to an adaptive response governed by a biologic law, limiting the bone growth after having reached its homeostatic constant. The newly formed bone, including the incorporated bone implant, should be regarded as a functional unit, mechanically competent. The use of those osteoconductive permanent or slow resorbing materials may contribute to valid functional bone repair.
Illustration of the two bone waves theory. Left panel: after 4 weeks of implantation in the sheep distal femur, the three-dimensional printed porous metal scaffolds (3DS) proved to be highly osteoconductive, as demonstrated by the massive invasion of a dense young trabecular woven bone. Right panel: following 26 weeks of implantation, bone reshaping by virtue of a remodeling and maturation process occurred, resulting in a thicker trabecular bone and lower bone density compared to 4 weeks. Un-decalcified ground resin sections, approx. 50-µm thick section, Modified Paragon surface stain, original objective 4X. Courtesy of Waldemar GmbH, Germany.
Technical Challenges
Owing to the diversity of medical devices, sites of implantation, necropsy techniques and variability in slide preparation, a custom-based approach should be considered for the management of the explants. Bone implant pathology requires advanced instrumentation. In vivo and ex-vivo imaging to monitor the implant and surrounding tissues, nondecalcified resin histology, various microcutting systems preserving the implant in situ, sophisticated microscopy and computer-assisted image analyzer system are mostly used. An algorithm of technical preparation options is presented in Figure 8. Additional technical challenges are described in the literature.4,30
Suggested algorithm scheme for routine selection of embedding media, cutting process, and stains. 4
The FDA requests comparative quantitative histomorphometric measurements in bone implant performance evaluation. As discussed above, bone and implant surface areas and BIC parameters are historically provided. With the advent of artificial intelligence and machine learning and for research purposes or to highlight specific competitive performance variables, it is now possible to extract much more data from the slides than in the past. For instance, bone anisotropy, erosion parameters, bone architecture parameters (trabecular thickness, number, and separation), osteoid parameters, granulometry, intergranular distance, scaffold porosity, and dynamic parameters using bone fluorochrome markers are among many other new parameters which could be exploited in bone implant pathology.
In regenerative medicine, one of the present and future challenges for pathologists is to histomorphometrically constitute a set of the normal osseous constants with respect to each animal model used in preclinical testing. The new regenerative treatments should be compared with predicates, the sham-operated group and with normal osseous constants.
Conclusion and Forecast
To evaluate the safety and performance of a bone implant, different factors should be considered together. There are a range of new disruptive technologies which combined with deeper scientific understanding in osteopathology (notably biological laws, role of mechanical factors, AI, regulatory evolution among others) are influencing the shift in interpretation of bone pathology. A diagram based on Dr Alves and colleagues’ experience in bone implant pathology contributes to our understanding of how to approach bone implant evaluations (Figure 9).
The four facets model a toolbox example for bone implant evaluation in pathology.
In conclusion, an important question has been discussed with the community of practice: “What is needed in bone pathology for the upcoming years?.” Some of the needs have been highlighted by Dr Alves as follows: Terminology: Consensus for unified definitions and terminology among the stakeholders. Safety: (a) Refining the ISO 10993-6 Reactivity Ranking score in bone implant pathology,
28
(b) Implementation of bone fluorochrome markers to detect toxic effects not visible in static conditions. Performance: (a) Constituting normal values of the bone tissues for each relevant animal model, (b) Discovering more biological laws to continue building the foundations of bone implant pathology.
11
Technique: Medical device pathology needs optical advancements in microscopy with integrated equipment offering polarization light, reflexion mode, epifluoresent UV light, and spectroscopy mode altogether.
Finally, bone implant pathology, as for other medical device pathologies, is still in early stages of development. Collaboration across multidisciplinary fields is necessary for further maturation of bone-implant pathology to advance.
Regulatory Requirements for Medical Devices Regarding Toxicological Properties
The ISO 10993-1 series discussed by previous presenters focuses specifically on the requirement for the biocompatibility safety assessment of medical devices. The ISO 10993 series provides guidance on how to specifically address various aspects of biological safety still leaving manufacturers a certain freedom on how to conduct testing. The presentation of Dr. Bussmeyer aimed to explain basic concepts and frameworks for regulatory approval of medical devices in the European Union and other geographic areas and highlighted recent developments with relevance to the toxicologist and pathologist. The respective requirements are less detailed but strictly binding.
EU Requirements
Historically, the Council Resolution 85/C 136/01 of 7 May 1985 settled on the so-called “new approach” to technical harmonization for medical devices. Technical harmonization was approached by only harmonizing a limited number of requirements called the “essential requirements” (ER) and by applying reference to standards to demonstrate fulfillment of the ER. Mutual recognition of this approach was intended to allow free movement of goods within the EU for medical devices carrying a CE-mark. To this end, three directives were published: Directive 90/385/EEC 1 regarding active implantable medical devices, Directive 93/42/EEC 17 regarding medical devices, and Directive 98/79/EC 15 regarding in vitro diagnostic medical devices.
In September 2012, the European Commission proposed a new legislation to increase safety, traceability, and transparency. Two regulations replace the aforementioned three directives: The Medical Devices Regulation, MDR (EU) 2017/745, 18 and the In Vitro Diagnostic Medical Device Regulation, IVDR (EU) 2017/746. 16 The main legal act laying down the requirements for medical devices is the European MDR/Regulation (EU) 2017/745 (Figure 10).
The regulatory framework: from former MDD/AIMDD and IVDD to MDR and IVDR.
To obtain regulatory approval, a conformity assessment is conducted, which is defined as “the process demonstrating whether the requirements of this Regulation [MDR] relating to a device have been fulfilled” (MDR, article 2, Definitions, [40]). 18 After successful conformity assessment, manufacturers may affix the CE mark on their devices. The conformity assessment route depends on the risk class of the device, which is determined by application of the 22 classification rules laid down in Annex VIII of MDR. Devices are divided into classes I (low risk), IIa, IIb (medium risk), and III (high risk), taking into account the intended purpose of the devices and their inherent risks. High-risk devices are subject to the most stringent conformity assessment procedures. The conformity assessment generally involves an audit of the manufacturer’s quality system. Moreover, depending on the type of device, a review of the technical documentation on the safety and performance of the device is conducted.
The principles of classification according to MDR overlap with those laid down in ISO 10993-1 in the sense that high invasiveness generally corresponds to a higher risk class. Medical device regulation classification rules can be divided into four groups depicted in Figure 11.
Classification rules under MDR.
The “General Safety and Performance Requirements” (GSPR) laid out in Annex I of the MDR define the prerequisites, which every medical product must fulfill before obtaining certification and in turn being allowed to be placed on the EU market. The GSPR concern safety as well as technical and medical performance of the device. They replace the “Essential Requirements” of MDD.
To demonstrate compliance with the GSPR, manufacturers must (1) check each product type or model against each requirement, (2) determine whether the requirement is applicable, (3) apply harmonized standards to demonstrate compliance, (4) acquire documented evidence of compliance, and (5) and follow one of the conformity assessment procedures suitable for the determined type of device and its risk class.
The GSPR are divided into three chapters: Chapter I—general requirements (safety and performance requirements [SPR] 1-9), Chapter 2—requirements regarding design and manufacture (SPR 10-22), and Chapter 3—requirements regarding the information supplied with the device (SPR 23).
Chapter I lists general requirements. Products must work in accordance with their design and intended functionality. They must not endanger the health or safety of the patient, user, or any other person. The risks associated with a medical device must be reduced to the maximum extent possible. Manufacturers must maintain a risk management system that continues to be updated throughout the product lifecycle. Moreover, manufacturers must provide patients and users with information about any potential risks.
Chapter 2 states requirements regarding design and manufacture comprising topics such as chemical, physical, and biological properties of a device, toxicological aspects, medicinal substances used in conjunction with a device, microbial contamination, and biological materials (animal/human origin). The MDR puts more emphasis on toxicological aspects than MDD and AIMDD. Although medical devices do not have a pharmacological function, there are various toxicological aspects to be considered when seeking regulatory approval for medical devices. The GSPR require manufacturers to take toxicological aspects of medical devices into consideration from the early development phase when settling on device design and specifically when choosing materials and substances to be used (Figure 12).
GSPR and toxicology.
SPR 10 specifies the requirements related to the chemical, physical, and biological properties of a device. SPR 10.1 outlines general considerations for materials regarding the device function, toxicity, and biological safety. It requires manufacturers to consider physiological processes such as absorption, distribution, metabolism, and excretion (ADME). It points to the importance of physical and chemical characterization as part of the design process and biological safety evaluation.
SPR 10.2 considers risks from contaminants and residues, and SPR 10.3 relates to compatibility of a device with materials and substances.
SPR 10.4 lists the requirements regarding substances contained in and released from the device.
Devices, which are invasive and come into direct contact with the human body or (re)administer medicines, body liquids or other substances, or transport the latter have specific requirements regarding CMR (carcinogenic, mutagenic or toxic to reproduction) substances or agents with endocrine-disrupting properties. Concentration of substances of concern must stay below 0.1% by weight. Moreover, MDR demands that the European Commission shall provide its scientific committee with a mandate to prepare guidelines including a risk-benefit assessment of phthalates as well as other CMR and endocrine-disrupting substances. It is hence advisable to examine regularly if new guidelines on substances of concern have been published when seeking approval for devices containing such agents.
SPR 10.6 regards risks related to particle size. Special care must be taken when designing and manufacturing devices incorporating nanomaterials, which have high or medium potential for internal exposure. Risks linked to the size and properties of particles shall be reduced as far as possible unless these come into contact with intact skin. Of note, the risk class of devices incorporating a nanomaterial is dependent on the level of potential for internal exposure as stated in rule 19, Annex VIII of the MDR. The use of nanoparticles can place a device in a higher risk class that may significantly complicate regulatory approval.
SPR 12 deals with the requirements for medicinal products and substances which are absorbed or locally dispersed in the body. The most significant change in comparison to the MDD/AIMDD consists in the inclusion of substances or combinations of substances that are absorbed by or locally dispersed in the human body, along with the medicinal considerations. Devices, which are absorbed or locally dispersed in the human body and devices that contain a medicinal substance having an action ancillary to that of the device must fulfill requirements beyond biocompatibility as per ISO 10993.
SPR 12.1 refers to Article 1(8) of the MDR determining which combination products are in scope of this Regulation. Only medicinal substances having an action “ancillary to that of the device” are included. Combination products require consultation with a competent authority as outlined in Annex IX, Chapter II, Section 5.2 of the Regulation.
SPR 12.2 includes entirely new specific provisions for devices composed of substances which are introduced into the body and absorbed or locally dispersed in the body. Such devices shall comply with the relevant requirements in Directive 2001/83/EC for medicinal products for the evaluation of certain toxicological data including ADME, possible interactions of those substances, their metabolism in the human body, their local tolerance and toxicity. Data on single-dose toxicity, repeat-dose toxicity, genotoxicity, carcinogenicity and reproductive and developmental toxicity may have to be provided to obtain a product certificate. In addition, the Notified Body conducting the medical device assessment must seek the scientific opinion of a competent authority as defined in Annex IX, Chapter II, Section 5.4
Devices without a medical purpose were excluded from the scope of the MDD and AIMDD but are included in the MDR. SPR 9 clarifies how to apply the “risk-benefit” and “performance” requirements (SPRs 1 and 8) for those devices without a medical purpose or medical benefit. Annex XVI of the Regulation is devoted to a listing of these devices. In addition, the so-called common specifications (CS), that is, the Commission Implementing Regulation (EU) 2022/2346 of 1 December 2022 apply. The CS detail requirements related among other things to risk management and safety information.
Annex II of the MDR describes the basic requirements regarding the contents of the product Technical Documentation (Figure 13). Chapter 6 of Annex II is of interest for toxicologists as it specifies the documentation requirements regarding product verification and validation as well as additional information required in specific cases. For absorbable products, data on ADME, local tolerance and toxicity must be included. A justification must be provided in the case of devices containing CMR or endocrine-disrupting substances.
MDR documentation requirements.
Some parts of the MDR are certainly important to consider when conducting a toxicological assessment within the frame of a medical device approval process (Figure 14). In summary, introduction of MDR increased the importance of toxicology for medical device certification.
Sections of MDR related to toxicology.
U.S. Requirements for medical devices are governed by the FDA in the United States. The Center for Devices and Radiological Health (CDRH) is the section responsible for performance and safety assessment and premarket approval of medical devices including the related quality management audits. Its assessment is based on the applicable sections of Title 21 of the Code of Federal Regulations (CFR), which governs food and drugs within the United States; 21 CFR part 820 Quality System Regulation corresponds to ISO 13485 While 21 CFR part 814 applies to class III devices requiring premarket approval. 12 Furthermore, ICHQ10 applies, which describes a model for quality management system 22 and applies to the development and manufacture of pharmaceutical drug substances and drug products, including biotechnology and biological products, throughout the product lifecycle.
In the regulatory context, non-US and non-EU countries are often referred to as “Rest-of-World” (ROW) countries. Most ROW countries have established an independent registration process. Approvals within recognized reference agencies (e.g., US FDA, EC certificate) can be a prerequisite for starting registration applications in certain countries and may help to expedite the registration timeline in certain countries. The ISO 10993 series is recognized by many of the ROW countries for evaluation of biological safety of medical devices or have formed a basis for local standard development, for example, China GB/T 16886.1, Japan JIS T 0993-1.32,43 Periodic updates may be published by local authorities to align with ISO parent standard(s). Continuous compliance to local standards is mandatory for device registration and maintenance. In conclusion, toxicological aspects are increasingly important to consider for global regulatory approval of medical devices. Regulations and guidelines beyond ISO 10993 must be considered.
Footnotes
Acknowledgements
The authors thank their respective employers and the IATP for supporting their travel and lodging to deliver their lectures in-person at the IATP Satellite Symposium; Dr. Deepa Rao for chairing and moderating the session in-person; the ESTP’s Scientific Organizing Committee and the IATP’s Education Committee in developing, organizing, and supporting the 2023 IATP Satellite Symposium held on 26 September, in Basel, Switzerland.
Author Contributions
FW, AA, UB, and NVS led analysis, writing, and interpretation for their respective sections based on their in-person lectures at the IATP Satelline Symposium. The overall drafting of the manuscript was led by FW and UB. All authors shared responsibility for revision and final approval for publication.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: FW and UB are employee shareholder of Merz Aesthetics, AA is an employee of NAMSA, and NVS is an employee shareholder of Boston Scientific. The manuscript’s content was developed solely by the authors. Statements in this manuscript are those of the authors and not of their employer.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
