5. Short Comments on the Use of the Dog in Regulatory Toxicology

抄録

　The first attempts to bring some order to the toxicological testing of new drugs came in the 1940s. There was then a lull, before the publication of detailed regulatory guidelines to define test methodology began in the 1960s. The dog, almost exclusively for the last several decades, the beagle, became the principal non-rodent species used in toxicology, this being more formalised in writing (but not as official guidelines) in Europe in 1965 and in the USA in 1977 (see Zbinden [16]). It is not within the scope of this short paper to give a full description of the advantages and the disadvantages of the dog and there any many such papers already available (4, 9, 12) as well as papers reviewing animal/human concordance in toxicity prediction (see 8 and references therein). In this paper, I shall attempt to summarise and discuss some of the data available at Huntingdon Life Sciences and other topics of debate and recent advances surrounding the use of the dog. To provide a source of reference, I shall also make short comments on other reviews to which the researcher may wish to refer.

　Greaves et al (5) provide a recent summary of papers discussing the correlation between observed animal and human toxicity for major drug classes and organ systems. They conclude that the dog is better predictor of human adverse effects than rodents or monkeys based on the particular papers they have reviewed, being particularly useful in predicting gastroin­testinal toxicity but over-predicting hepatic and renal toxicity. It is interesting to note however, that a number of these key papers were published in the 1960s and there seems to be a relative lack of more recent critical reviews of this subject. One of the newer papers cited is by Olson et al (8) describing the outcome of a pharmaceutical industry initiative on this subject and an International Life Sciences Institute work­shop in 1999 at which working parties were ask to identify modified testing strategies which may improve non-clinical screening of drugs in development. They concluded that the dog is often selected with insufficient consideration of whether it is pharmacodynamically, physiologically, biochemi­cally or metabolically the best choice, but they could recommend no alternative species widely applicable in general toxi­cology to replace the dog.

　There is a notable change of emphasis between papers written prior to the mid-1990s and those written more recently concerning the dog. When discussing the disadvantages of the dog, older papers do so from the position that the dog was the only widely used non-rodent species (primates such as rhesus, cynomolgus and marmosets being used in regulatory toxi­cology but expensive and not easy to obtain) and so the disadvantages had to be allowed for and coped with. Newer papers make much greater reference to the need for the use of pharmacodynamically rele­vant species and to the newer available models, particularly the minipig.