To use or not to use: the dilemma of NSAIDs and craniotomy

Postcraniotomy pain receives inadequate treatment, though studies have shown that 76% of patients report moderate to severe pain during the first 48 h after surgery [1,2]. Visual analog pain scores of at least 4, persist in 48% of patients even on the second postoperative day [2]. At present, there is no consensus on the management of pain following a craniotomy. Titrated doses of opioids are frequently administered to other surgical patients. However, the fear of complications such as sedation, hypoventilation, meiosis and seizures tends to limit their usage in adequate doses in neurosurgical patients. Scalp block at the end of surgery has been used to decrease the requirement for analgesic medications [3], but the duration of its action is limited to only 4–6 h. NSAIDs by themselves, even in large doses, are insufficient for postoperative pain relief after major surgery, but they are believed to offer a significant opioid-sparing effect. Addition of Vioxx, a cyclooxygenase-2 (COX-2) inhibitor NSAID, to opioids has been shown in a small study [4] to decrease the side effects associated with opioid pain medications, encourage earlier walking and reduce total hospitalization costs in neurosurgical patients. In this issue of European Journal of Anaesthesiology, Dolmetova et al. [5] compare the efficacy of ‘scheduled’ dosing of lornoxicam, a nonspecific NSAID, with that of an ‘on request’ regimen of the same drug following craniotomy. As in the earlier study [6] in nonneurosurgical patients, a scheduled analgesic regimen was more effective than ‘on demand’ analgesic administration. The authors report in this small experience no postoperative haematomas, renal failure or peptic ulcers in their study population. It is reassuring that lornoxicam, administered as per the schedule in this study, did not seem to have major side effects. However, the larger issue of benefits versus risks of short-term NSAID usage for postcraniotomy pain remains unclear. The vast majority of recent evidence on the risks associated with NSAIDs pertains to their long-term usage, and very few studies have addressed the issue of their short-term usage when started postoperatively. Three major effects of NSAIDs confront physicians in using these agents for postoperative pain relief, particularly after a craniotomy: effects on myocardial ischaemia; effects on platelet function and consequent increase in the incidence of postoperative haematoma; and effects on renal function. These concerns have their basis in studies conducted in patients receiving NSAIDs for a variable period of time, from a few days to several weeks or months preoperatively. COX enzyme is the major target of action of NSAIDs. Nonspecific NSAIDs block both COX-1 and COX-2 receptors. Bleeding and renal adverse effects are mediated through COX-1. Gastrointestinal bleeding is reduced with selective COX-2 inhibitors [7], but other concerns are far from negligible, at least with chronic use. Inhibition of prostacyclin without simultaneous inhibition of COX-1, which promotes platelet aggregation, has been proposed to increase the risk of myocardial ischaemia when COX-2 inhibitors are used on a long-term basis [8–10]. Two randomized controlled studies explored the cardiovascular safety of short-term perioperative usage of COX-2 inhibitors in patients undergoing cardiac surgery. Both the studies [11,12] reported higher rates of postoperative cardiovascular thromboembolic events with the use of COX-2 inhibitors. In contrast, in a study of noncardiac surgical patients, a similar incidence of cardiovascular thromboembolic events was reported in the placebo and the study groups. The study group in this investigation received parenteral parecoxib for 3 days followed by oral valdecoxib for another week. The patients underwent major orthopaedic, abdominal, gynaecological or noncardiac thoracic surgery. Neurosurgical procedures were particularly excluded from the study [13]. The differential cardiovascular effects of COX-2 inhibitors, between perioperative cardiac and noncardiac surgical patients, may be explained on the basis of the significant background cardiac risk in patients undergoing cardiac procedures. Until these mechanisms are better understood, anaesthesiologists may harbour apprehensions about using these drugs even for short-term postoperative pain relief in patients with or at high risk of cardiac disease. The effect of perioperative drugs on blood coagulation is a major concern after a craniotomy. The incidence of postoperative haematoma was 1.1% in a retrospective series of 6668 neurosurgical operations. NSAID use in the 2 weeks preceding surgery was reported as a possible cause of postoperative haematoma; all but one patient with postoperative wound haematoma received either aspirin or NSAIDs within 2 weeks before surgery [14]. The study, however, has many limitations. The percentage of patients who received NSAIDs is not clear, and the statistical evaluation of the observation is not reported. In their conclusions, the authors themselves state that ‘this study does not prove that antiplatelet agents are responsible for the occurrence of postoperative hematomas.’ There is a significant difference between nonselective NSAIDs and COX-2 inhibitors with regards to their potential to cause haemostatic complications. Continuing nonselective NSAIDs in the perioperative period has been documented to increase the risk of bleeding in orthopaedic surgery. The likelihood of developing various complications was higher by 5.8 times for patients using NSAIDs 24 h before surgery [15]. In contrast to nonselective NSAIDs, celecoxib in doses of 1200 mg per day administered for 10 consecutive days in healthy adults demonstrated no effect on platelet aggregation or bleeding time [16]. In a small series of patients undergoing total knee replacement, celecoxib did not increase perioperative blood loss [17]. In another series of 50 patients undergoing gynaecological surgery, a lesser degree of disturbance of platelet aggregation and less blood loss have been reported with one dose of rofecoxib compared with diclofenac [18]. All the above evidence suggests that COX-2 inhibitors are safer than nonselective NSAIDs with regards to their effects on coagulation. There are other potential concerns with the perioperative use of NSAIDS. Prostaglandins are essential to maintain renal blood flow and glomerular filtration rate. Therefore, COX inhibitors could cause renal dysfunction [19], especially in patients who are dehydrated from prolonged diuretic therapy or repeated vomiting due to raised intracranial pressure. However, there is no strong evidence for adverse renal effects when used for acute postoperative pain relief, especially in patients with normal preoperative renal function [20]. Other concerns still undergoing experimental investigation are the roles of COX-1 and COX-2 in neurodegeneration and in cerebral ischaemia [21,22]. More clarity is required with regards to the safety and efficacy of NSAIDs for postcraniotomy pain relief. What seems clear is that preoperative NSAIDs should be stopped prior to intracranial surgery, and that postoperative COX-2 inhibitors should not be used in patients with significant cardiac disease. These concerns contrast with the routine use of these agents in many centres, especially in Europe and Asia, for postoperative pain relief without any perceived or reported increase in complications. This experience suggests that a large randomized safety and efficacy trial of postoperative administration of NSAIDs in patients not taking them preoperatively would be entirely ethical. Such a study should be undertaken initially in healthy young patients and, if found safe and efficacious, the same may be extended to older patients.