Best Evidence Analyses and Commentary
Adina Solis, PharmD Candidate

Prostaglandins play a role in both bone resorption and bone formation. Prostaglandins are potent stimulators of cytokines and growth factors which mediate bone resorption. The inflammatory response to trauma leads to resorption of necrotic tissue and the production of bone forming cells, eventually leading to new bone. Without this primary inflammatory response, bone repair may not completely occur.

Nonsteroidal anti-inflammatory drugs (NSAIDs) decrease inflammation by inhibiting the synthesis of prostaglandins. They are proven to decrease pain and average length of hospital stay, allowing earlier mobilization, decreasing opioid consumption, and increasing patient satisfaction. Because prostaglandins play a role in bone repair, researchers have attempted to prove whether NSAIDs cause delay in fracture healing or an increase in nonunion.

In numerous animal studies, the use of NSAIDs has been shown to impede the healing of fractures. In 1976, Ro et al. compared the effects of indomethacin 2mg/kg/day given to rabbits with femur fractures. The researchers made a closed, standardized fracture of 129 rabbit femurs and randomly assigned the rabbits to either receive indomethacin via stomach tube or placebo. The rabbits were sacrificed at 6, 9, 12, 18, and 24 days after the fracture. The bones of the two groups were compared using mechanical strength testing, radiological examination, and histological examination. Researchers discovered that the treatment group had significantly less mineralization, cartilage, and bone bridging than the placebo group. This effect has been repeated in numerous studies with rats and rabbits, using indomethacin, ibuprofen, and ketorolac.

The question is whether the inhibitory action upon bone healing will be seen in humans. Unfortunately, there are very few quality studies, and of those there are conflicting data. In a randomized, double-blind study by Adolphson et al., there was no effect of piroxicam on bone mineral density or fracture healing. Researchers in this study randomly assigned 42 healthy postmenopausal women with displaced Colles’ fractures to either start taking 20 mg oral piroxicam daily within 48 hours of injury or to the placebo group. The fractures were analyzed radiographically initially, after reduction, 11 days, 4 weeks, and 8 weeks post-trauma. Although not significant, patients assigned to the treatment group experienced less bone mineral decrease when compared to placebo. Although this randomized controlled trial showed that NSAIDs may not have contributed to nonunions in humans and may have bone-sparing effects, several retrospective studies have suggested otherwise.


In a retrospective study by Burd et al., patients given indomethacin after the surgical treatment of acetabular fractures developed significantly more nonunions than those who did not receive indomethacin.  Researchers involved in this study evaluated 112 patients who had open reduction and internal fixation of an acetabular fracture and who were at risk for heterotopic ossification.  Of these patients, 36 patients needed no prophylaxis, 38 received focal radiation, and 38 received indomethacin. A total of fifteen patients developed non-unions, but the rate of non-union in patients receiving indomethacin was significantly higher versus those the who did not (26% v 7%; p = 0.004). In another retrospective study investigating the effects of posterior spinal fusion, patients given ketorolac 60mg IM loading dose followed by 30mg every 6 hours as needed, there was an odds ratio of 4.9 with a dose-dependent relationship between nonunion rates.  Researchers looked at the records of 288 patients who received spinal fusion. Nonunion was identified in five of the 121 patients (4%) who received no NSAID. In contrast, nonunion was identified in 29 of 167 patients (17%) who received ketorolac after surgery, demonstrating an approximately five times greater likelihood of developing nonunion after ketorolac consumption. Unfortunately, because ketorolac was given to patients for pain control, those who requested ketorolac might have been selected preferentially for further surgical exploration, possibly introducing confounding. Finally, in a 2010 meta-analysis of cohort and case-control studies looking at the effect of NSAIDs on bone union, a pooling of all studies showed an increased risk of nonunion with NSAID exposure; however, when only the high quality studies were used, there was no increased risk. Quality of the studies was assessed using the Newcastle-Ottawa Scale. After screening 158 studies for inclusion, researchers used a total of 11 studies for the initial pooling of results which showed odds ratio for nonunion of 3.0 (95% CI 1.6-5.6). However, metaregression showed a significant association between lower study quality and increased reported risk of nonunion (p = 0.0009). When only higher quality spine studies were considered, there was no increased risk of nonunion demonstrated (OR = 2.2, 95% CI 0.8-6.3).

Certainly, a need exists for more prospective trials investigating the effects of NSAIDs on bone healing. Although several retrospective cohort and case-control studies have found an association between NSAID use and effects on bone healing, no causal relationship has been established. NSAIDs have proven benefits, yet due to the possible association of NSAIDs and nonunions, consideration must be given to patient selection for NSAIDs. 

[1] Raisz LG, Martin TJ (1984). Prostaglandins in bone and mineral metabolism. In: Peck WA (ed) Bone and mineral research. Elsevier, Amsterdam, pp 286-310

2 Simmons DJ: Fracture healing perspectives. Clin Orthop 1985; 200:100-113

3 Garaj NM (2003) Cyclooxygenase-2 inhibitors. Anesth Analg 6:1720-1738

4 Ro J, Sudmann E, Marton PF: Effect of indomethacin on fracture healing in rats. Acta Orthop Scand 1976:47:588-599

5P Adolphson et al. No effects of piroxicam on osteopenia and recovery after Colles’ fracture. Arch Orthop Trauma Surg 1993 112:127-130.

6 TA Burd et al. Indomethacin compared with localized irradiation for the prevention of heterotopic ossification following surgical treatment of acetabular fractures. J Bone Joint Surg Am. 2001 83-A:1783-1788.

7 SD Glassman et al. The effect of postoperative nonsteroidal anti-inflammatory drug administration on spinal fusion. Spine 1998 23: 834-838.

8 Dodwell ER et al., NSAID Exposure and Risk of Nonunion: A Meta-Analysis of Case-Control and Cohort Studies            

Website Terms of Use Disclaimer | E-mail Terms of Use | Contact LECOM | RSS | Copyright © 2009 LECOM - All rights reserved