Selective and non-selective cyclooxygenase inhibitors delay
stress fracture healing in the rat ulna

LJ, Cowling NR, Wu AC, Kelly WL, Forwood MR. J Orthop Res. 2012 Jul 30. doi:
10.1002/jor.22203. [Epub ahead of print]

anti-inflammatory drugs (NSAIDs) are a commonly used medication for pain
related to musculoskeletal injuries that can be obtained both over the counter
and with a prescription.  They act by
inhibiting the effects of cyclooxygenase (COX) enzymes, which play a role in
bone biology, specifically bone resorption by COX-2.  Stress fractures are common injuries that
affect many people of all ages and level of activity.  There are studies that show negative effects
of NSAIDs on complete fracture healing, but there is little data on the
relationship of NSAIDs and stress fractures. 
Kidd et al, set out to determine the effects of three different NSAIDs
on stress fracture healing and woven bone consolidation in an animal model
using the rat ulna.  Their hypothesis was
that a COX-2 inhibitor would have a greater effect on fracture line remodeling
and consolidation of periosteal woven bone than ibuprofen (non-selective COX
inhibitor) and PMX53 (a C5a receptor antagonist that targets the complement system; a key component of the immune system).  A total of 170 rats underwent repetitive cyclic
loading of their right ulna under anesthesia until a 10% increase in
displacement was reached (all treatments were approved by the authors’ animal
ethics committee).  Sixty rats were given
either a selective COX-2 inhibitor or a control daily over six weeks.  Eighty rats received ibuprofen or distilled
water over six weeks.  Thirty rats were
given PMX-53 and used the same control group as the rats given ibuprofen.  The rats were then euthanized at 2 (except
PMX-53 group), 4, or 6 weeks after the repetitive loading to examine the ulna using
histomorphometry (microscopic examination of the
cortical and woven bone area, as well as the area of bone resorption).  The authors found that ibuprofen
significantly reduced bone resorption and cortical formation 6 weeks after
loading and that the selective COX-2 inhibitor significantly reduced resorption
along the fracture line at 2 weeks only. 
There were no significant differences between the selective COX-2
inhibitor or PMX-53 treated groups and control groups at 6 weeks.  There was no difference in woven bone area
among any of the medications and control groups.

fractures are a painful injury and patients often look for some type of
medication for pain control, commonly an NSAID. 
The authors were surprised that ibuprofen caused a greater reduction in
bone healing than the COX-2 inhibitor because there have been many studies that
showed COX-2 plays the greatest role in fracture repair and resorption through
the production of a prostaglandin (PGE2). 
They conclude that maybe both COX-1 and COX-2 enzymes are needed for the
maximal amount of PGE2 production.   The authors noted that one possible limitation
that may have led to a lower than expected result with the COX-2 inhibitor was
the use of a different vehicle than the ibuprofen.  This may have created inconsistent drug levels
with peaks and troughs.  Although human
studies would be needed to make a definite recommendation against the daily use
of NSAIDs in stress fractures, we should be cautious when we prescribe a daily
dose of these pain medications to our patients. 
Ibuprofen and other non-selective NSAIDs are available over the counter
and thus are more commonly used than the COX-2 inhibitors, since they require a
prescription.  When an athlete or any
patient is looking for every possible way to return to play as soon as
possible, suggesting that they refrain from daily NSAIDs may be beneficial.
This study does not address the impact of intermittent use of NSAIDs.  Have you noticed any changes in healing rates
in your patients with stress fractures that use NSAIDs?  How do you counsel patients and athletes for
pain control with stress fractures?

by: Kris Fayock, MD
by: Jeffrey Driban


Kidd LJ, Cowling NR, Wu AC, Kelly WL, & Forwood MR (2012). Selective and non-selective cyclooxygenase inhibitors delay stress fracture healing in the rat ulna. Journal of Orthopaedic Research PMID: 22847634