Evaluation of anti-inflammatory drugs by orthopaedic surgery in dogs

Abstract

Both traditional and modern surgery have emperically given importance to various anti-phlogistic measures, following the rationale that the inflammatory process often overshoots its objective as a defense and repair process and becomes excessive and detrimental. The anti-inflammatory drugs used in human surgery have generally been selected according to their performance in patients with rheumatoid arthritis, and essentially the same drugs have been adopted for use in veterinary surgery. Rheumatoid inflammation, however, differs markedly from an acute post-traumatic inflammatory reaction. There is a lack of reliable models for clinical assessment of anti-inflammatory effects. Recent research in human oral surgery with a rather unique model that allows well controlled studies on how antiinflammatory drugs may modulate a post-operative course, has for several drugs challenged the common view regarding their efficiency and suitability in controlling post-traumatic sequelae. It has remained an open apply body, question whether the findings in oral surgery also to surgery and traumata of other parts of the e.g. the extremities. Appropriate and well - xxvii - controlled clinical models for such studies are lacking in human medicine. The aim of the present work was to establish a properly controlled model for investigations on how steroidal and non-steroidal anti-inflammatory drugs (NSAID) might modulate the signs of the inflammatory reaction and the healing process following orthopaedic surgery. The experiments were designed on a placebocontrolled crossover basis, with two "identical" surgical interventions being performed on the forelimbs of each dog with an interval of 28 days, to enable a paired comparison of the post-operative courses. In a standardized way, under general anaesthesia, the 3rd metacarpus was transected with an oscillating saw. The fracture was then stabilized with a mini dynamic compression plate before the wound was sutured. A special device was designed to allow measurements of post-operative swelling, while pain and limb function were assessed by the use of visual analogue scales. Abnormalities in the wound healing were recorded as well as clinical signs that could be indicative of adverse drug effects. Radiographs taken 2, 4, 6 and 8 weeks after the two operations were interpreted and compared for bone union, callus formation, signs of infection and foreign body acceptance. The dogs were - xxviii - euthanized 8 weeks after the 2nd operation and the two 3rd metacarpi of the forelimbs excised. They were later cut in a cryo-microtome and the stained sections assessed for bone healing. The three anti-inflammatory drugs selected for the investigations were a glucocorticoid, betamethasone; and two NSAID, phenylbutazone and indomethacin. Glucocorticoids are recognized as the most powerful anti-inflammatory drugs, but their place in surgery is disputed. Phenylbutazone was selected since it is probably still the most widely used NSAID in veterinary practice, while a main reason for including indomethacin was that it has been reported to delay or inhibit fracture healing. In the first trial, a single pre-operative injection of 3mg betamethasone was tested against placebo in each of 8 dogs. The drug proved to significantly reduce the post-operative swelling. On the 3rd day the reduction was 43~. Less pain and limping were assessed after the glucocorticoid was injected, but the differences did not reach a level of significance. No adverse effects of the glucocorticoid on wound or fracture healing were detected. This trial included measurements of the endogenous cortisol levels. A marked decline in the serum cortisol levels followed the glucocorticoid injection. The levels remained low for about 3 days and - xxix - then returned to normal. It was concluded that the results of the study support the view that short-term glucocorticoid administration can efficiently curb an excessive post-traumatic inflammatory reaction and is essentially safe. In the next trial 8 dogs were given 300 mg phenylbutazone by oral administration twice daily for 8 days starting on the day before surgery. Phenylbutazone did not reduce the swelling significantly as compared to placebo, although the drug gave a significant pain relief. The clinical observations indicated somewhat better wound healing after the operation when placebo was given, and that also applied to the fracture healing as evaluated by radiographs and bone sections. In another group of 8 dogs, 25 mg indomethacin was to be administered orally twice daily for 8 days starting on the day before surgery. This medication had to be discontinued after 2 1/2 days when they had received a total dose of 125 mg indomethacin, since signs of toxicity became evident, e.g. vomiting, bloody stool and lethargy. One indomethacin-treated dog died on the 5th post-operative day. Swelling measurements showed no consistent difference, but the pain assessments were significantly lower after the operation when indomethacin was administered. No noticeable differences were observed - xxx - in wound healing, but the radiological evaluation revealed tendencies in disfavour of indomethacin. A trial was then undertaken in another group of 8 dogs with a lower dosage of indomethacin. They each received 5 mg indomethacin twice daily for 8 days starting on the day before surgery. Even at this dosage one of the dogs developed bloody stool on the 5th post-operative day. With indomethacin there was a tendency towards less swelling, and the reduction became significant after one week. The pain assessments showed no consistent difference and there appeared to be no difference in wound and fracture healing. The difficulties encountered in selecting an appropriate dosage of indomethacin provide a striking example of how differences in pharmacokinetics may explain differences in drug response both within as well as between species. It was difficult to obtain consistent and reliable assessments of pain and limping even if the dogs served as their own controls. These results should therefore be cautiously interpreted. The present studies provide evidence that the drug effects on post-operative swelling observed in oral surgery, also apply to acute traumatic swellings in other parts of the body, since the recordings with limb volumetry showed a remarkably good correlation with - xxxi - corresponding results obtained in oral surgery. This conclusion was recently also reached by another researcher in tests on paracetamol using this model with limb surgery. In addition to significantly reducing the swelling, paracetamol also proved to efficiently reduce pain without any signs of adverse effects. Indomethacin does not appear to be recommendable in dogs, while phenylbutazone presents a relatively wide safety margin. The anti-phlogistic potential was, however, not impressive for any of the two NSAID. A short-term glucocorticoid administration or paracetamol seem to be better choices for curbing the sequelae of an acute post-traumatic inflammatory reaction.