Abstract
Preparing and processing medical device implants for evaluation is a relatively high-risk and high-dollar process in which studies get made and endpoints can be lost with no second chance. It is important to customize every aspect of the preparation process to the type of device and the study endpoints. Some standard and proven approaches for a few types of implants are discussed, addressing fixation, special dissection and extraction techniques, preprocessing imaging, trimming techniques, sample processing, embedding media, and cutting and staining options.
Introduction
The choice of approach to device handling during the preparation and processing phase of a device study plays a vital role in the quality of the data that one harvests from a study. In a field in which every device presents a new set of conformational circumstances, it is imperative to precisely conceive and plan for implant-site dissection, fixation, imaging, trimming, and cutting to capture all necessary endpoints. Experience and common sense are key factors in this approach, as is knowledge of the device one is dealing with. Any error can result in the loss of valuable data. The field of implant pathology is very unforgiving, technically challenging, and labor and equipment intensive. The pathologist should be involved throughout the whole study, from protocol design through trimming and from imaging to histopathological evaluation.
Every Device is Unique
Certain types of devices are more commonly encountered than others (e.g., cardiovascular stents and orthopedic implants). However, the types of devices are as varied as the applications and imagination of the engineers designing them. The device–tissue interface and its preservation present an additional challenge for pathology. Three major factors come into play when approaching a new device: three-dimensional structure, chemical composition of the implant (metals, polymers, coatings, biodegradable materials), and surface feature for tissue interaction. The evaluation steps described below have to be tailored to a particular device to capture and document critical information.
Imaging of Implantation Sites
Evaluation of medical devices often requires more than just microscopic evaluation, and various imaging techniques may be applied before extraction and dissection of the sample. Furthermore, to obtain optimal samples for evaluation, various imaging modalities may help in the extraction and dissection process.
Gross Photography:
Generally, a record of the gross appearance of the sample is required. Standard calibration and labeling procedures and appropriate stage lighting are used to demonstrate critical features (e.g., patency and surface features). Unlike toxicology studies, the images produced are often a standard part of the pathology report.
Micro Computed Tomography (Micro CT) Scanning:
Micro CT offers excellent resolution of in situ structural detail. This modality can be performed with or without contrast medium injection (Microfil) to outline a hollow space or vascular lumen. Hollow metal devices may pose a technical challenge because of radiographic shadowing.
Radiography:
Radiography with a Faxitron X-ray machine using high-resolution film is an important imaging modality for medical devices. A Faxitron is a small, cabinet sized X-ray machine that can easily fit into most laboratory settings. It has a shielded, radiation-safe enclosure. High-resolution radiography can be critical to demonstrate conformational mismatch, device-fatigue fractures, or calcification. They are also essential in assisting during trimming.
Microradiography:
Microradiographs are defined as high-resolution radiographs of microscopic slides (e.g., thick bone and implant sections). This type of radiography requires very long exposure times at low voltage. They can be used for bone density and trabecular bone conformation measurements.
Gross Observations
There are certain implants for which gross observations cannot be obtained without disrupting the interface between the tissue and the device. For other implants, gross observations would not be a useful endpoint, and therefore, are not worth pursuing. Examples of these circumstances include intracranial implants, bone implants, some types of articular or intervertebral disc implants, intraocular implants, and pacemaker leads.
In contrast, there are certain medical devices for which gross observations are critical and are the principal endpoint: for example, complex vascular connectors, anastomotic clips, bypass grafts, cardiac valves, vein filters, and patent foramen ovale (PFO) occluders.
Finally, in some cases, gross observations might be useful, but by exposing the device, one risks compromising the site, and potentially, the acquisition of other endpoints. Stents, experimental aneurysms, linear/axial vascular shunts, and self-expanding stent grafts or any self-expanding device under spring-loaded tension would all pose a challenge for acquisition of gross observations.
Extraction and Dissection of the Implant Site
Handling of the site and of the excised specimen requires careful consideration to avoid alterations in device conformation and relationships to adjacent tissues. For example, in cardiovascular stent sections, poor handling can result in dislodging the stent struts from their previous location in vivo (Figure 1) and/or alteration of planimetric endpoints (lumen collapse). There are often a lot of adhesions and fibrous scarring at implant sites, resulting from healing after the surgical procedure. It is generally not wise to try to dissect through this adhesion formation and scarring. With that said, in rare instances, it can be critical to remove extraneous scar tissue that forms after hematoma resolution to access the site and capture essential gross observations. Examples include any anastomotic site (anastomotic clips and some complex vascular grafts) in which critical and focal patency may not be reliable if captured or sampled microscopically only.
Fixation
General rules for fixation that are used for tissues in general apply for fixation of a medical-device implantation site. Fixatives can be applied by immersion or perfusion. Immersion works best when tissues are thin and permeable. Often, pretrimming will be necessary. Access holes may need to be drilled into bones and enclosed cavities (e.g., the skull or joints). Ideally, large samples should be fixed by perfusion of fixative via the supplying vasculature, and the blood should be rinsed out when vascular patency is a morphological endpoint.
Because the tissue–device interface is important to preserve, it is best to minimize relative shrinkage of tissue by not using fixatives associated with considerable shrinkage. Because bone is not subject to shrinkage, this factor becomes less crucial when dealing with devices implanted into bone. In the case of large samples, in which pretrimming or perfusion fixation may not be possible, it is important not to use fixatives with slow or limited penetration abilities (e.g., Bouin’s).
Finally, if one wants to consider doing immunohistochemistry, the choice of fixative must be considered based on the sensitivity of the antigen being tested to different fixatives.
Embedding
Paraffin offers the best medium for embedding in terms of cost, ease of handling, and flexibility of staining procedures. Anything that can be sectioned on a standard microtome (e.g., most polymers and biological implants) can be handled in paraffin, barring any biomaterial and paraffin incompatibilities.
However, paraffin is not hard enough to support metals or undecalcified bone and is not hard enough to support production of semithin sections. Table 1 lists resins that are commonly used in medical-device work.
Because many medical devices contain metals and other hard components, they require a hard embedding medium to support the specimens during sectioning (e.g., methylmethacrylate acetate [MMA] or glycomethacrylate acetate [GMA] and Spurr’s resin). Handling of these substances requires certain safety precautions because of toxicity. Because of its relative softness, GMA can be cut on a microtome but is not suitable for grinding. Also, during polymerization, significant shrinkage may occur, depending on the resin, and shrinkage will affect hard tissue, soft tissue, and implants to different degrees. Infiltration times can be very long for large specimens (days to weeks). Furthermore, there are compatibility issues that may arise. For example, any methacrylate-based implant or coating will generally not be compatible with MMA processing. A good review of MMA-based techniques for cardiovascular stents is provided in Rippstein et al. (2006). Spurr’s, LR White, LR Gold, and Lowicryl are useful for electron microscopy and are also variably used for light microscopy and immunohistochemistry.
Microtomy
Thin sections (2 to 5 μm) may be taken of paraffin-embedded sections or undeplasticized GMA or Epons, as in the case of peripheral nerves or some vascular grafts. Thin sections may also be deplasticized (MMA or Spurr’s) for stents, undecalcified bone, and polymers. These sections should be taken with motorized microtomes for speed control. Regularly sharpened tungsten carbide blades are also needed. In our experience, solid knives that can be regularly sharpened yield better results than those seen with disposable knives.
Thick sections (15 to 50 μm) are required for large polymer or metallic implants. These thick sections require a precision diamond band, disc, or wire saw to produce. Precision sectioning is followed by leveling down of the section with a grinder or polisher. These processes are expensive, requiring specialized equipment and a large number of hours by skilled technologists.
Staining
Paraffin-embedded tissue, the medium most pathologists are used to dealing with, obviously can be stained with a broad range of stains. Similarly, deplasticized MMA or Spurr sections, if completely deplasticized, are also amenable to numerous stains. However, for undeplasticized MMA sections (thin or thick sections), the available stains are more limited (Table 2). Furthermore, staining of thick sections is limited to the surface. GMA, which cannot be deplasticized, is water miscible and can be stained with most water-soluble stains (e.g., HE or PAS [hematoxylin and eocin, or periodic acid - Schiff]).
Immunohistochemistry
The success of immunohistochemistry in paraffin or plastic-embedded tissues is antigen-dependent and, to a lesser extent, antibody-dependent. In general, the choice of method to be used for immunohistochemistry, such as devices, must be approached on a case-by-case basis depending on the markers being investigated, and a systematic approach to determining the best method (e.g., fixation, resin handling, antigen retrieval) is required. Thin sections (5 μm or less) are required, and if the cellular reaction at the tissue-device interface is being investigated, then that must also be intact.
As mentioned previously, fixation plays a role, and many antigens that have undergone formalin or paraformaldehyde fixation will likely require antigen retrieval. The same considerations for fixatives apply to immunohistochemical staining with a particular antibody in any tissue.
Choice of the embedding resin is based on factors such as the hardness of the resin required, the amount of heat generated by the polymerization reaction, and the need for plasticization. Getting a 5-μm section is easier with some resins than with others because of hardness of the resin. Whereas GMA and MMA can support thin sections of specimens containing bone or metal, the highly exothermic polymerization reaction may damage antigenicity. Some control of specimen temperature may be accomplished by polymerizing the specimen in a controlled, low-temperature environment; if the specimen is large, controlling the internal temperature of the tissue may prove difficult. Technovit 8100 is a GMA-embedding resin that polymerizes at a lower temperature and has been shown to give good immunohistochemistry results in stented arterial tissue (Malik et al., 1998). Technovit 9100 also polymerizes at a lower temperature, but is based on polymethyl methacrylate and not GMA. Technovit 9100 New is a recently released MMA-based resin designed specifically for immunohistochemistry, enzyme histochemistry, and in situ hybridization (Yang et al., 2003a).
Some resins—MMA, Technovit 9100, Technovit 9100 New—need to be deplasticized before immunohistochemical staining. GMA-based resins, including JB4 and Technovit 8100, cannot be deplasticized or deacrylated, and staining must be done with the resin in place. Similarly, LR White and LR Gold are not deplasticized before immunohistochemical staining. There is some evidence to suggest that reduced sensitivity and penetration of the immunostaining reagents may result from lack of resin removal (Malik et al., 1998; Saito et al., 1998; Yang et al., 2003b), which may affect detection of some antigens.
Conclusion
In a field in which every device presents a new set of conformational and material circumstances, it is imperative to precisely conceive and describe implant-site dissection, fixation, imaging, trimming, and cutting to capture all necessary endpoints. Experience and common sense are required. Any error results in the irreparable loss of experimental units. The field of implant pathology is very unforgiving, technically challenging, and labor and equipment intensive. The pathologist has to be involved throughout the whole study, from protocol design through trimming and imaging, in addition to histopathology, to ensure thorough documentation and adequate methodology.