Regulatory Forum Opinion Piece*

Abstract

Necropsies conducted to support medical device development may be conducted in facilities that do not have the infrastructure in place to support Good Laboratory Practice for Nonclinical Laboratory Studies (GLP) studies and for a variety of reasons they may be conducted without a pathologist present at necropsy. However, when a novel medical device or one that is expected to significantly alter tissues is deployed, or when the surgical model confounds interpretation of device effects, it is the opinion of the authors that an experienced pathologist should be present at necropsy.

Nonclinical safety studies supporting medical device development differ somewhat from those supporting small- or large-molecule pharmaceutical compounds. Necropsies conducted to support development of small and large molecules are typically conducted according to the study protocol and standard operating procedures (SOPs) by technical staff trained in documentation of macroscopic observations and accustomed to routinely differentiating normal from abnormal organ changes, which may lessen the need for a pathologist’s attendance at necropsies. Medical device studies are generally smaller studies with fewer animals on a staggered schedule due to limitations in the number of devices that can be implanted each day. These studies may be conducted at sites that do not have pathologists experienced in safety assessment on staff. For logistical reasons and to decrease costs, necropsies are often conducted without a pathologist’s presence. However, for the following reasons, an experienced pathologist should be present at necropsy for novel medical devices: (1) by their nature, medical device studies are often a combination of safety and efficacy studies and gross necropsy observations may be as important as microscopic findings in assessing safety and efficacy, (2) device studies may be conducted at a non-GLP testing facility such as a university or research hospital that does not have staff trained in documentation of macroscopic observations in safety studies, (3) placement or effects of the device may alter normal tissue borders or cause local effects such that sampling must be modified from standard procedures, and (4) effects are often local and may be confounded by tissue response to invasive techniques used for implantation, so understanding the orientation and extent of findings may be critical to safety evaluation.

Device Study Conduct

In the United States, Food and Drug Administration (FDA)’s Center for Devices and Radiological Health approves medical devices. Premarket approval (PMA) is FDA’s process to evaluate the safety and efficacy of Class III medical devices, which are those that support or sustain human life, are of substantial importance in preventing impairment of human health, or present a potential, unreasonable risk of illness or injury. Examples of Class III devices include stimulators implanted in the brain, endosseous implants, and replacement heart valves. PMA applications contain the results of nonclinical toxicology studies and must include a statement that each study was conducted in compliance with 21CFR58 or, if not compliant, a statement of the reason for noncompliance. Medical device studies may be conducted at a non-GLP testing facility due to the need for specialized imaging technology, surgical equipment and skills, or other technologies that are not available at Contract Research Organizations (CROs) conducting more routine, regulated safety studies. In these situations, attendees at necropsy may consist of a medical doctor and/or an engineer who are keenly interested in the device’s efficacy and technical staff who are not trained by a pathologist in necropsy techniques and documentation. In addition, because these testing facilities are non-GLP, SOPs for necropsy may not exist and anticipated procedures are listed in the protocol. These anticipated procedures cannot predict all possible complications and often do not give adequate guidance as to what should be sampled or how (i.e., what the pathologist will need to examine) when complications occur. Furthermore, there are several CROs that conduct medical device studies but do not have a pathologist on staff. At these regulated facilities, SOPs are in place, but the necropsy is typically overseen by a clinical veterinarian or study director, often in the presence of, and with input from, the investigator/engineer, and by staff not trained by a pathologist.

Appropriate and Adequate Lesion Sampling

If a pathologist is not present at necropsy, the (histo)pathologist may subsequently be presented with poorly described observations or with inadequate sampling such that the pathologist may not be sure of the full extent of the findings, thus precluding accurate safety evaluation. For example, the pathologist may be presented with a small tissue sample exhibiting locally extensive ulceration but without adequate description to know the true severity or extent of the finding in the organ. If the gross necropsy description of an organ such as the stomach, intestine, or bladder is “thickened,” and the pathologist receives a 2 cm × 1 cm sample with mucosal ulceration and an edematous wall, how can he/she determine if the ulcer was confined to the sample submitted, which presumably was selected because it appeared to be the most affected area, or if the majority of the mucosa was ulcerated? Gross descriptions even by well-trained, experienced staff may be unreliable in these situations, and the appropriate action at necropsy is adequate sampling, which may include fixation of the entire organ. However, a technician holding a protocol that states “two samples of the organ, one proximal and one distal, will be fixed in formalin and submitted for histopathology” is often reluctant to modify the approach when left to his or her own judgment to conduct the necropsy.

Well-trained staffs with adequate SOPs are often able to conduct necropsies and adequately record findings when effects are systemic and organs are in their normal site and maintain their usual, discrete borders, and orientation to other tissues. By their nature, devices and their use may alter tissues. Examples include placement of a device to create a urinary bladder stoma in the abdominal wall, use of a device or thermal injury to ablate an entire gland or a portion of an organ, or placement of a mesh or other substrate to promote wound closure or tissue repair. In these situations, the normal tissue borders and the borders of the device effects may be unclear. Without a pathologist present, the amount and quality of sampling that will allow the study pathologist to subsequently make a safety assessment based on microscopic findings and the technician’s description will be up to someone who has not been trained in histopathology or safety assessment of devices.

Determination of Device versus Implantation Reaction and Importance of Macroscopic Data for Preclinical Development of Medical Devices

Orientation and distance from the device placement site or any surgical manipulation can be critical to evaluating the safety and efficacy of a device. Surgical scars and responses to surgical procedures must be differentiated from tissue reaction due to the device. For example, is a granulomatous response due to an implanted, absorbable matrix or is it the consequence of an osteotomy to place the matrix? Of course, control samples are very helpful, but their usefulness can be limited if sampling or the quality of sample labeling varies across animals.

In many situations, the highest quality data are obtained when the surgeon or person who manipulated the device is present at necropsy along with the pathologist who will subsequently evaluate the tissues microscopically. Communication of the pathway of device placement and surgical procedures such as use of electrocautery, absorbable sutures, or osteotomies to place the device and appreciation of other complicating factors such as the amount of tissue handling and exposure to thermal energy or physical or chemical agents is facilitated when both are present. The pathologist can sample the tissues with a clear understanding of orientation and distance relative to device placement. Adequate samples and descriptions are obtained to clearly define margins or extent of a finding and thus place the local severity into context.

Additionally, many devices intended for human use require nonrodent laboratory animal species approximating human size for testing, such as small ruminants or pigs. The time required for device placement or intraoperative or postoperative data collection in these animals may require staggered surgeries over days, consequently requiring necropsies over several days. The surgeons learn as they adapt their surgical procedures to a nonhuman species and as they become accustomed to using the device, so they typically become more proficient at device placement over time. Subtle differences in surgical technique may become apparent to an experienced technician but more probably to the experienced pathologist trained to look for, and sort through, subtle macroscopic changes and their effects on microscopic findings. Feedback to the sponsor from the pathologist regarding observations on surgical technique, even if given informally during conversation rather than formally in a written report, may prove helpful in planning future studies and may help the sponsor understand a seemingly greater efficacy over time.

A pathologist also has an understanding of background disease processes that may be seen in these larger, “nonlaboratory” species, particularly grazing ruminants and farm pigs. The pathologist is able to appropriately describe observations related to these diseases such that they can be understood as background findings and not ascribed to the device or surgical placement, as well as understand the impact, if any, subclinical disease may have on the healing (recovery) from surgical device placement.

Finally, in nonclinical device studies, the study pathologist is frequently called upon to make a statement regarding efficacy as well as an assessment of safety. In some cases, efficacy and the extent or lack of complications are better appreciated on a macroscopic than microscopic basis. For both macroscopic and microscopic results, the pathologist is in an untenable position if he or she cannot rely upon the gross observations and the adequacy of sampling. For this reason, an experienced pathologist should be present at gross necropsy for novel device studies.

When an Experienced Pathologist Should Be Present at Necropsy

There are situations where well-trained staff can conduct the necropsy for a medical device study and the presence of a pathologist would not add value. These situations often involve repetitive studies of similar devices where the extent of findings and complications are predictable. For these studies, protocols or SOPs that provide appropriate guidance for sampling and recording of observations are readily written. Also, for minimally invasive procedures where tissue changes can be predicted to be highly localized and not alter the organ on a macroscopic basis, a trained technician armed with a protocol that directs appropriate sampling can conduct the necropsy without the pathologist present. On the other hand, when a novel device or one that is expected to significantly alter tissues is deployed or when the surgical model complicates interpretation of device effects, it is the opinion of the authors that an experienced toxicological pathologist, preferably the pathologist who will evaluate the histopathology, should be present at necropsy.