Abstract
The Commentary by Schoeb et al. 8 in this issue of Veterinary Pathology raises specific questions concerning standards in the conduct of nonclinical carcinogenicity studies to support the safety assessment of commercial products or for products intended for commercialization. Three specific aspects of an adequate process include credentials for the pathologist evaluating the tissues, health standards for rodents at the testing laboratory, and peer review of the raw data.
We live in an era of globalism, with products flowing across international borders, necessitating concerns for safety standards. Furthermore, for industry and government, outsourcing or “offshoring” toxicology testing work today seems the norm. It is the responsibility of the regulatory agency and sponsoring company staff to ensure adequate quality of the toxicologic pathology product. This pathology product represents the assessment of responses of laboratory animals administered test articles of commercial interest to mimic and exaggerate the anticipated human or environmental exposure. The foundation for a high-quality toxicologic pathology product, of necessity, must be built on the primary evaluation by an appropriately credentialed veterinary pathologist (or other credentialed pathologist). In North America, the evaluations for the adverse effects of agents of commerce predominantly center on the evaluations by a board-certified veterinary pathologist (Diplomate of the American College of Veterinary Pathologists [ACVP]). The ACVP sets high standards for pathologist certification and has a strong interest in international standards, which is reflected in the decision to collaborate with the European and Japanese Colleges of Veterinary Pathologists in publication of Veterinary Pathology. However, in certain countries, significantly different standards exist.
Similarly, in North America, standards for the care and health status of laboratory animals in regulatory research studies are uniformly high, including exclusion of infectious disease from the research colony. Intercurrent disease generally has been recognized for several decades as an undesirable complicating factor in successful determination of the health hazard of commercial products in normal individuals.
In the United States, the regulatory response to a carcinogenicity study by the U.S. Food and Drug Administration (FDA) would depend on its own rigorous review of any relevant data. In this specific circumstance that Schoeb et al. 8 addresses, concerning an aspartame bioassay for carcinogenic hazard identification, the FDA would request all raw data as it did in the premarket phase for this agent, including the pathology slides. This information was requested by the FDA from the laboratory for the aspartame study in question, but the laboratory staff did not supply the detailed data or the pathology slides, precluding independent agency judgment, thus representing insufficient evidence for the study conclusions. 3
As professionals in 2009, it behooves us to participate in good science and to publish well-conceived, designed, and analyzed research studies. Most of our work is reviewed by someone in our organization, by journal reviewers, other professionals, or audience members for our presentations. We should be prepared to take criticism of any type if we strongly believe in our own work and to defend it by allowing others to review it in all aspects. In the specific circumstance described in the commentary by Schoeb et al., 8 in which a contract research organization rat aspartame carcinogenicity study was reported as showing that the food additive caused lymphomas after lifetime exposure to aspartame, questions arose.
The commentary by Schoeb et al. 8 in this issue of Veterinary Pathology raises the particular possibility that some or many of the reported lymphomas may not, in fact, be true lymphomas but rather severe inflammatory lesions caused by respiratory infections, possibly caused by Mycoplasma pulmonis. Their case is based on published reports and publically available data but not on review of the histology slides. A complete pathology peer review of appropriate cases would be in order, as well as polymerase chain reaction analysis of paraffin sections to support or refute the position of Schoeb et al. 8 Apparently no dose response, no dose-related mortality caused by lymphomas, and no relevant postulated mechanism of carcinogenesis by the chemical were reported. Evidently rats were allowed to die, producing perhaps severe autolytic changes that compromised diagnostic efforts with frequency inconsistent with good toxicologic pathology practice. 4
The possibility of M. pulmonis infection has not been addressed by the institute that performed the study, because the institute has not yet allowed others to review diagnostic records and verify the absence of M. pulmonis and other respiratory tract infections. Because aspartame is a popular food additive, it is important that the study be audited and the pathology be peer reviewed. Pathology peer review changed regulatory considerations for an important food additive over 25 years ago in the United States. 6 Veterinary pathology peer review procedures are now well defined and widely practiced by most laboratories involved in safety studies of test articles of commercial importance. 1,2,5,7,9,10
The article by Schoeb et al. 8 generated some controversy in the approval process. The editors have carefully considered this controversy but do not consider controversy alone as sufficient reason to refuse publication. We hope this commentary by Schoeb et al. 8 will foster public debate in the search for the truth in this bioassay. We believe that this commentary also illustrates the potential challenges in contract research organization selection and pathologist credentialing, and in the accurate determination of cancer hazard of articles in commerce.