Mechanisms of action

COX is present in most body tissues and can become upregulated with a variety of stimuli. Two forms have been recognized traditionally, COX-1 and COX-2, and both are expressed in the spinal cord [1–5]. Although both enzymes have similar functions, their patterns of expression are very different. COX-1 is expressed in many cell types and functions as a “housekeeping” enzyme with important roles in vascular hemostasis and gastroprotection. In contrast, COX-2 expression is primarily induced by factors such as endotoxins and cytokines, is expressed at sites of inflammation and produces prostaglandins that mediate inflammatory and pain responses. NSAIDs inhibit the biosynthesis of prostanoids though the inhibition of the COX. They can target both COX-1 and COX-2 or be COX-2 selective. COX-2 selective NSAIDs have been developed to reduce the adverse effects, especially when administered over a long period of time. Experimentally, prostanoid concentrations in gastric and duodenal tissues are not significantly affected by COX-2 selective drugs [6]. However, there are no studies demonstrating that they also reduce the renal or hepatic adverse effects [2].

Experimental animal and clinical human studies

In contrast to a number of studies evaluating steroids in patients with neurologic deficits due to spinal cord injury, studies of NSAIDs in patients with intervertebral disc herniation focus primarily on inhibition of the hyperalgesia caused by the sensitization of the peripheral afferent nerve and not on the spinal cord injury itself [7]. The role of the different epidural inflammatory components is still a subject of study both in animal models and humans. Studies of human patients reveal a predominant prevalence of macrophages and a minor role for mast cells in both acute and chronic disc herniation [8]. Furthermore acute lateral disc herniations are linked with the presence of T and B lymphocytes. Based on these observations, the type of inflammation might be dependent on the kind of herniation. One study evaluating the inflammatory properties of contained versus noncontained lumbar disc herniation in human patients demonstrated a higher concentration of leukotriene B4 and thromboxane B2 in cases of noncontained herniated disc material [9]. Similar observations have been made in animal studies. In a porcine model, disc injury without disc prolapse resulted in a modest inflammatory response dominated by T lymphocytes, followed by a macrophage response that peaks 1 month after disc injury [10]. In rats, COX-2 contributes to an increased excitability of the spinal cord with peripheral inflammation [5].

In dogs, the cytologic and histopathologic appearance of extruded degenerate disc material is variable and can include some degree of inflammation and presence of dysplastic spindyloid cells [11]. Approximately 82% of the dogs with cervical or thoracolumbar disc extrusion have an inflammatory reaction in the epidural space at the site of extrusion [12]. This ranges from acute invasion of neutrophils to formation of chronic granulation tissue. The mononuclear inflammatory infiltrates consisted mostly of monocytes and macrophages and only few T and B cells. The degree of inflammation correlates with the degree of disc extrusion and associated epidural hemorrhage and the extent of calcification of the extruded disc material. Also, there is an inverse correlation between intensity of the epidural inflammation and outcome [12].

In light of these observations, experimental studies evaluating the effect of NSAIDs on nerve root dysfunction in model animals have been performed. In pigs, a nucleus pulposus-induced nerve root injury was created. Nerve root dysfunction was reduced with the administration of diclofenac but not ketoprofen. The authors suggested that the reason for this disparity in the response might be due to the different selectivity for the two COX types [13]. Similarly in dogs, incision of the nucleus pulposus decreased blood flow to the nerve root and dorsal ganglion and reduced nerve function, an effect that was blocked by simultaneous administration of indomethacin [14]. In another study, local application of nucleus pulposus to the nerve roots and/or mechanical compression resulted in functional and histological changes in nerve roots and the dorsal root ganglia. The combination of mechanical compression and chemical irritation from the nucleus pulposus caused more nerve root injury than each factor alone [15].

Canine clinical studies

There are only a few clinical studies of NSAIDs as part of a conservative treatment for IVDD in dogs. In a retrospective study of conservative treatment of dogs with presumptive thoracolumbar IVDH, dogs receiving NSAIDs were more likely to have higher quality of life than those who had not received these medications [16]. In another retrospective study, dogs with back pain and mild neurological deficits due to presumed thoracolumbar intervertebral disc herniation were managed medically with steroids or NSAIDs. Patients treated with NSAIDs had a lower rate or recurrence (33%) compared to those treated with steroids (66%) [17]. In a retrospective study of dogs with presumed cervical disc herniation treated conservatively, 49 per cent of patients recovered and administration of NSAIDs was significantly associated with a successful outcome [18].

Recommendations

NSAIDs are valuable as analgesics, either alone or as part of a multimodal analgesic plan. A multimodal analgesic approach is preferred for an adequate management of neuropathic pain [19]. To date, published studies of NSAID administration in dogs with disc disease are retrospective in nature, and there are no randomized controlled clinical trials evaluating efficacy and safety. However, based on clinical experience in human and veterinary patients as well as laboratory animal studies, NSAIDs can be considered as one component of conservative therapy in appropriate patients. The authors would recommend their use as part of a conservative treatment only in cases with acute signs of neck or back pain, with no to mild neurological deficits. These patients should be closely monitored for any worsening of the neurological status, and specific instructions about physical restraint should be given to the owner.

Commonly used NSAIDs in dogs include aspirin, carprofen, etodolac, meloxicam, ketoprofen, deracoxib, firocoxib, meclofenamic acid, tepoxalin, and tolfenamic acid. Selectivity of COX-2 versus COX-1 is often expressed as the COX-1/COX-2 inhibitory ratio. Because different studies use different techniques, there is considerable variability with respect to the COX-1/COX-2 inhibitory ratio for the various products. And in veterinary studies, there is no convincing evidence that drugs with higher COX-1/COX-2 ratios produce fewer gastrointestinal or renal adverse effects than drugs with low ratios [20]. NSAID administration should be only considered in normotensive, well-hydrated patients with a normal hepatic, renal, gastrointestinal, and hemostatic function. In addition, NSAIDs should not be combined with corticosteroids and two or more NSAIDs should not be administered concurrently.