Author + information
- Robert S. Schwartz, MD, FACC⁎ (, )
- Nicolas A. Chronos, MBBS and
- Renu Virmani, MD
- ↵⁎Minneapolis Heart Institute, Minnesota Cardiovascular Research Institute, 920 East 28th Street, Suite 300, Minneapolis, MN 55407
We read with interest the comments of Dr. Langeveld and colleagues concerning our recent review of preclinical restenosis models (1). These investigators write that stenting the rat carotid or abdominal artery provides a “simple, inexpensive, rapid, and accurate preclinical model for in-stent restenosis.” We have several comments in response regarding the utility of the rat model.
A useful in-stent restenosis animal model should accurately predict: 1) safety, 2) efficacy, and 3) pathophysiologic mechanisms. These are addressed as follows.
Safety. The major safety issues for stents are thrombosis (acute or subacute) and neointimal thickening causing luminal stenosis. Although the rat model sometimes induces stent thrombosis, it does so to a lesser extent than the porcine and rabbit models. Total occlusion and severe stent stenosis do not generally occur in the rat model.
Efficacy. Rat carotid restenosis models were abandoned years ago because virtually all therapies that were tested and effective in rats later proved ineffective in patients. Such studies included angiotensin-converting enzyme inhibition, heparin, and other anticoagulants, antiplatelet therapies, and corticosteroids. Much expense, time, and energy-conducting clinical trials were expended based on inaccurate results from rat models. This may occur as stent efficacy is determined by neointimal inhibition and prevention of negative remodeling. Rat vessels may develop enough neointima to cause arterial narrowing following stenting, but they do not appear to exhibit remodeling. Rat models of diabetes and hypertension exist that could theoretically test different aspects of human disease. Drug-eluting stents have been deployed in rat carotid arteries, and we are awaiting the outcome of these studies (R. Virmani, unpublished data, August 2005). Much current effort is directed toward establishing efficacy parameters in the pig and rabbit model, with promising early results by investigators such as Touchard et al. (personal communication, November 2005) or Finn et al. (personal communication, September 2005).
Simplicity, expense, rapidity. Stenting rat carotid arteries is as simple as stenting porcine or rabbit vessels. Whether the rat model is truly cheaper than the rabbit or porcine model is unclear. The purchase costs and per diem housing are higher with pigs than with rabbits or rats. This is partially overcome though by the ability to place three or sometimes four stents in a single pig, and two in the rabbit carotid or three to four stents in the rabbit aorta.
The porcine model offers the advantage of being an excellent coronary artery model for testing stent, guidewire, and catheter performance in arterial beds with size, shape, curvature, and tortuosity similar to humans. Flow dynamics in a small, noncoronary vessel such as in the rat differ dramatically from patients, with different shear and boundary layer flow. These may be important in the era of drug-eluting stents where elution is dependent on flow parameters. Elution studies have been performed in the rabbit iliac arteries and are similar to those in the pig (A. Finn, personal communication, September 2005).
Importantly, the major costs of preclinical restenosis studies relate to surgical time, histopathologic stent processing (requiring plastic embedding), and careful histomorphometric measurement. Using a rat model for this step is thus no simpler, and it saves neither time nor expense compared to rabbits or pigs.
Finally, it is certainly true that the available antibodies and knowledge of proteins and genes are greater for rats and mice compared to rabbits and pigs. However, given the desire by the interventionalist and the Food and Drug Administration to achieve parallel clinical trial results, elucidation of vascular response mechanisms in rodents may have limited translational value, although of scientific merit and interest.
Until stented rat models are validated against clinical trials for providing accurate results, we are hesitant to recommend this strategy, and we must continue to question scientifically the wisdom of their use in translational vascular research. Hence, large animal models continue to be the standard. In the meantime, vigorous scientific debate and discussion about animal models in vascular biology is both welcomed and exciting. There is no doubt that for safety studies we are limited to the pig coronary artery model, which also assesses the distal myocardial bed for ischemia and emboli.
Thus, we look forward to future scientific and validation studies of the stented rat model by Dr. Langeveld and others, and we thank them for their comments.
- American College of Cardiology Foundation