Session II - Spine

The Effect of Postoperative Nonsteroidal Anti-Inflammatory Drug Administration on Spinal Fusion

Steven D. Glassman, MD, S. Matthew Rose, BE, John R. Dimar, MD, Rolando M. Puno, MD, Mitchell J. Campbell, MD, John R. Johnson, MD

Louisville, Kentucky, USA

Purpose: Nonsteroidal anti-inflammatory drugs (NSAIDS) are widely used compounds which are known to inhibit osteogenic activity, but no prior studies have examined the influence of NSAIDs on spinal fusion in clinical practice. The purpose of this study was to examine the effect of postoperative ketorolac (Toradol®) administration on spinal fusion. We attempted to determine the extent to which inhibition of fusion, seen in animals using indomethacin, occurred with an NSAID in humans at clinically relevant doses.

Methods: Two-hundred eighty-eight patients with minimum 2 year follow-up who underwent an instrumented posterior spinal fusion from L4 to the sacrum between 1991 and 1993 were reviewed. Patients with a previous attempt at fusion or any supplementary procedure such as interbody fusion were excluded. Hospital records were examined for demographic data, and postsurgical medication regimen. Ketorolac was administered intramuscularly based on an assessment of pain control or the desire to limit narcotic analgesia. No other NSAID was utilized in the perioperative time frame. Office records and radiographs were used to evaluate the status of the fusion. Nonunion was defined strictly based upon surgical exploration, hardware failure, or tomograms. Differences in fusion rate were statistically analyzed based on ketorolac administration as well as potential confounding factors such as smoking history, gender, age and weight.

Results: One-hundred twenty-one patients received no NSAIDS and 167 patients received between 1 and 39 (mean 10) doses of ketorolac postoperatively. The two groups were found to be statistically equivalent for all demographic characteristics, except for a slightly greater body weight (182 vs. 173 lbs., p<.05) in the ketorolac treated group. Nonunion was identified in 5 of the 121 patients (4%) who received no NSAID. In contrast, nonunion was identified in 29 of 167 patients (17%) who received ketorolac postoperatively. This statistically significant difference in nonunion rate (p<.001) demonstrates an approximately 5 times greater likelihood of developing pseudarthrosis with ketorolac administration (odds ratio 4.9, 95% confidence limits 1.8 - 16.6). The influence of cigarette smoking on fusion was independently assessed. There were nine nonunions identified among the 129 nonsmokers (7%) and 22 nonunions in the group of 122 smokers (18%). The difference in nonunion rate was statistically significant (p<.01) as nonunion occurred almost 3 times more frequently in cigarette smokers (odds ratio 2.8, 95% confidence limits 1.2 - 6.6). The rate of nonunion was 2% among those patients who neither smoked nor received ketorolac; it was 7% for those who smoked but did not receive ketorolac; it was 10% for those who received ketorolac but did not smoke; and it was 25% for those who both received ketorolac and smoked cigarettes.

Discussion and Conclusion: Nonsteroidal anti-inflammatory drugs are commonly used medications with the potential to diminish osteogenesis. Recently, indomethacin was shown to decrease spinal fusion rate in an animal model. Despite the implication of these studies, the effect of NSAIDs on spinal fusion has not been assessed in any prior clinical study. The results of this study clearly demonstrate inhibition of spinal fusion following the postoperative administration of an NSAID, ketorolac. Statistical analysis revealed that patients treated with ketorolac were 5 times more likely than controls to develop nonunion. A statistically significant effect of ketorolac on fusion rate was also seen in subgroups of males, females, smokers, and nonsmokers. Smokers demonstrated a statistically significant increase in nonunion rate (p<.01) as has been shown in prior clinical studies. Smokers were 2.8 times more likely than nonsmokers to develop nonunion. This suggests that the influence of postoperative ketorolac administration on spinal fusion is at least equivalent, and possibly greater, than the known inhibitory effect of cigarette smoking. At present our recommendation, based on this data, is that nonsteroidal anti-inflammatory drugs be avoided, at least in the early postoperative period, after spinal fusion.