Experiment assessment of mass effects in the rat: implications for small animal PET imaging

Abstract

In vivo imaging using positron emission tomography (PET) is important in the development of new radiopharmaceuticals in rodent animal models for use as biochemical probes, diagnostic agents, or in drug development. We have shown mathematically that, if small animal imaging studies in rodents are to have the same “quality” as human PET studies, the same number of coincidence events must be detected from a typical rodent imaging “voxel” as from the human imaging voxel. To achieve this using the same specific activity preparation, we show that roughly the same total amount of radiopharmaceutical must be given to a rodent as to a human subject. At high specific activities, the mass associated with human doses, when administered to a rodent, may not decrease the uptake of radioactivity at non saturable sites or sites where an enzyme has a high capacity for a substrate. However, in the case of binding sites of low density such as receptors, the increased mass injected could saturate the receptor and lead to physiologic effects and non-linear kinetics. Because of the importance of the mass injected for small animal PET imaging, we experimentally compared high and low mass preparations using ex vivo biodistribution and phosphorimaging of three compounds: 2-fluoro-2-deoxyglucose (FDG), 6-fluoro-L-metatyrosine (FMT) and one receptor-directed compound, the serotonin 5HT_1A receptor ligand, trans-4-fluoro-N-{2-[4-(2-methoxylphenyl) piperazino]ethyl}-N-(2-pyridyl) cyclohexane- carboxamide (FCWAY). Changes in the mass injected per rat did not affect the distribution of FDG, FMT, and FCWAY in the range of 0.6–1.9 nmol per rat. Changes in the target to nontarget ratio were observed for injected masses of FCWAY in the range of ∼5–50 nmol per rat. If the specific activity of such compounds and/or the sensitivity of small animal scanners are not increased relative to human studies, small animal PET imaging will not correctly portray the “true” tracer distribution. These difficulties will only be exacerbated in animals smaller than the rat, e.g., mice.

Introduction

The driving force for the development of small animal PET scanners is threefold: the availability of various disease models presently available in small animals, the advantage of paired statistics when using serial imaging compared to serial autopsy, and the need to replace, reduce, and refine animal use [1]. Most of the animal models of neurological disease [2], cardiovascular disease [3], and cancer [4] are in rodents. The choice of the animal species is determined by the general knowledge of the neuroanatomy and neurochemistry, the ability to investigate behavioral responses and the cost to purchase and house the species. Rats and mice meet many of these conditions.

PET radiopharmaceuticals fall into three categories: (1) those that trace nonsaturable systems (e.g. [¹⁸F] fluoride in bone), (2) intermediate saturable systems (e.g., 2-[¹⁸F] fluoro-2-deoxyglucose (FDG) to measure changes in glucose transporter (GLUT) and hexokinase activity, and 6-[¹⁸F]fluoro-metatyrosine (FMT) to measure dopamine metabolism), and (3) easily saturable, low-density systems (e.g., [¹⁸F] labeled FCWAY, a 5-HT_1A antagonist to measure changes in 5-HT_1A receptor density). Another example of the latter case is [¹⁸F] labeled dVIP, a radioligand to measure VIP receptor density in tumors [5].

These three classes of radiopharmaceuticals present increasing imaging challenges as the density of the target site decreases. If we assume that studies are performed in humans with minimal saturation of the target sites at the attainable specific activity, we can mathematically deduce the conditions necessary to obtain comparable images in rodents and then experimentally test the effect of these conclusions in rats using the three classes of radiopharmaceuticals. We show with a simple mathematical argument that comparable images are obtained only when the amount of radioactivity administered to the rat is roughly the same as to the human. Thus, we compared biodistribution data in rats obtained with milliCurie, or human, level doses of injected tracer (by autoradiography) to microCurie amounts of injected tracer (by dissection and well counting) to identify potential violations of the tracer principal within these classes. Studies using PiPET were also performed to show that comparable images can be obtained in the rat brain [6].