Abstract
Dear Dr. Foster,
We would like to bring to your attention our concerns about an article entitled “Spontaneous Findings in the Heart of Mauritian-Origin Cynomolgus Macaques (Macaca fascicularis)” by J. D. Vidal et al., (Toxicol Pathol
The article by Vidal et al. is a retrospective review of heart lesions noted in Mauritian-origin cynomolgus monkeys used at the authors' facility in “recent years” (exact dates not specified). The abstract states that “[p]revious work from the authors' facility has described spontaneous cardiac findings in predominantly Indonesian-source animals; however, the authors have recently observed a novel spectrum of cardiac findings in Mauritian-source animals.” These findings included “a higher incidence of myocardial degeneration” as well as “macroscopic and microscopic subendocardial hemorrhage with hemosiderin, myocardial fibrosis, and arterial medial degeneration/hemorrhage.” Vidal et al. further state, “[i]t is unclear why these animals display a propensity to develop these lesions, but Mauritian-origin cynomolgus monkeys are a genetically distinct population…” and “. . . the relatively high incidences” of these findings “… could potentially complicate the use of this genetic source of cynomologus macaque in nonclinical safety testing… ”
We do not disagree with the descriptions and reported incidences of the various cardiovascular findings reported in this paper. We also do not dispute that some genetic variation exists among different geographic populations of Macaca fascicularis (Stevison and Kohn 2008). Instead, our concerns about the Vidal et al. paper include the lack of consistency with other data about cardiac pathology in macaques as well as specific methodological deficiencies of the study. We are also concerned that the data presented in the paper are too incomplete to support Vidal et al.’s interpretation of the study results, specifically, their implication that the reportedly “novel spectrum” of cardiac findings are attributable to genetic predisposition in Mauritian-origin cynomolgus monkeys.
Bioculture (Mauritius) Ltd., a leading supplier of Mauritian-origin cynomolgus monkeys, undertook an internal study in late 2009 to investigate the issue of cardiac pathology in Mauritian-origin cynomlogus monkeys and establish historical control data. As the initial phase of an ongoing investigation, five male and five female untreated control cynomolgus monkeys from the Bioculture (Mauritius) Ltd. facility in Mauritius were necropsied, and light microscopic examination was performed on a full complement of tissues with special attention to the cardiovascular system. None of the “spectrum of unique findings” in the heart described by Vidal et al. was noted in this initial study either by the study pathologist (M. Gruebbel) or the peer-review pathologist (A. Nyska) (unpublished observations).
In addition, recently published surveys of background pathology in control cynomolgus monkeys, including those of Mauritian origin, have not described heart lesions of the types described by Vidal et al., with the exception of very low incidences of myocardial fibrosis (geographic origin of affected animals not specified) (Drevon-Gaillot et al. 2006; Chamanza et al. 2006). Vidal et al. do not address why their results are inconsistent with the results of these other published studies.
The Vidal et al. study lacks appropriate concurrent controls, which in this case would be similar numbers of cynomolgus monkeys of different geographic origin but otherwise identical signalment, subjected to exactly the same experimental and environmental conditions over exactly the same time period as the Mauritian-origin monkeys. Vidal et al. did not include such a contemporaneous control group in their study, but they instead compared their results solely to those from a previous review of cardiac pathology in primarily Indonesian-origin cynomologus monkeys at the authors' facility (Keenan and Vidal 2006).
Such historical control comparisons might be useful in supporting Vidal et al.’s implied thesis of genetic predisposition if there was clear evidence that the “previous” primarily Indonesian-origin cohort had been treated exactly the same as the Mauritian-origin cohort reported in Vidal et al. This evidence would ensure that the only variable in the experiment was the geographic origin of the two cohorts.
Such clear evidence is not provided by Vidal et al. The Vidal et al. paper’s Materials and Methods section does include documentation that the general husbandry, euthanasia, necropsy, histology, and light microscopic methods applied to the Mauritian-origin cohort were similar to those used for the “previous” primarily Indonesian-origin cohort (Keenan and Vidal 2006).
However, numerous other variables can influence study results and comparisons between experimental groups. Indeed, Vidal has stated, “[f]actors that can affect type and incidence of spontaneous changes include origin of the animal, environmental conditions, provisions for routine health care, type of vehicle used, and the route of administration” (Keenan and Vidal 2006).
Aside from the Materials and Methods items already mentioned, the Vidal et al. paper does not provide data documenting that these other factors (besides geographic origin) were considered, evaluated, and conclusively ruled out as possible causative or contributing factors in the development of the described heart changes.
For example, the monkeys in the Vidal et al. study were noted to have “received vehicle only by either the oral, intravenous, or subcutaneous routes of administration.” But Vidal et al. do not identify the vehicle(s). Vidal et al. do not correlate vehicle type and route of administration to lesion occurrence on a per animal basis. Therefore, the reader cannot ascertain whether or not a particular vehicle material and/or route of administration might have been associated with the occurrence of some or all of the reported heart lesions.
Perhaps even more importantly, Vidal et al. do not provide evidence that these factors (vehicles and route of administration) were strictly controlled for when the subject Mauritian-origin cohort was compared to the “previous” reference cohort of primarily Indonesian-origin monkeys (which also “received vehicle only by either the oral, intravenous, or subcutaneous routes of administration” [Keenan and Vidal 2006]). Vidal et al. do not document whether the percentages of Mauritian-origin monkeys and Indonesian-origin monkeys receiving particular vehicles by particular routes were similar. In other words, there is no evidence that these potentially confounding variables were adequately controlled for in the evaluation of the study results. Thus, it is impossible for the reader to evaluate whether or not such possible extraneous disparities may have influenced the differences in cardiac lesion incidences in the two cohorts.
The same can be said of many other “environmental conditions” which might have changed in between the periods that the two temporally separated monkey cohorts were present in the authors' facility, including different quarantine facility conditions prior to arrival at the authors' facility; different holding periods in the quarantine and authors' facilities prior to necropsy; different levels of intercurrent disease; different technical personnel handling the animals, administering dosages, and performing euthanasia; and so on.
Thus, although Vidal et al. categorically state about the cardiac findings that “[a]n increased proportion of macaques being of Mauritian origin correlates with the increased incidence of these findings observed at our facility”, they do not address or conclusively rule out whether these increased incidences may also have had possible concurrent correlations with a number of other potentially causative or contributing factors.
In summary, we think that the above points constitute relevant concerns about the validity of the conclusions of Vidal et al., in particular the implied thesis that the observed cardiac lesions might be related to a genetic predisposition in Mauritian-origin cynomolgus monkeys.
Thank you for the opportunity to bring these concerns to the attention of the readership of Toxicologic Pathology and the general scientific community.
Footnotes
Both co-authors wrote the “Letter to the Editor” in their capacity as consultants to Bioculture (Mauritius) Ltd.