Editorials
Neurology Pharmacology
Volume 195 - Issue 9

NSAIDs and stroke risk

Author:  David J Blacker

Med J Aust 2011; 195 (9): 488. || doi: 10.5694/mja11.11185
Published online: 7 November 2011
Download PDF (104 KB)
Share: Share on Blue Sky Share on Facebook Share on LinkedIn Share by email

Recent studies build a strong case to suggest that there is a clear risk

Clearly, stroke prevention is preferable to the currently available treatments, particularly for haemorrhagic stroke. The burden of stroke is substantial, so all strategies to reduce risk must be considered. The traditional risk factors, especially hypertension, are well recognised, but there is also increasing interest in identifying and modulating novel risks1 and precipitant causes. In this issue of the Journal, an important article by Caughey and colleagues2 adds to the growing literature concerning the risk of stroke related to the use of non-steroidal anti-inflammatory drugs (NSAIDs), particularly those with selective cyclooxygenase (COX)-2 inhibition. If the use of these agents is a clear and substantial risk, then avoiding such medications, particularly in high-risk patients, may be an important preventive strategy. But is the risk clear and substantial, or are other factors involved? Is there a potential for abandoning useful medications and also creating undue anxiety for patients currently using them to treat painful chronic conditions?

The lessons regarding the cardiovascular risk of rofecoxib must be heeded,3 and the pharmacological basis for increased risk of thrombosis and elevation of blood pressure (and thus haemorrhagic stroke risk) is well founded. Surprisingly, recent guidelines4 pertaining to patients with extracranial large arterial stenosis, who are at high risk of stroke, made no specific comment for or against the use of NSAIDs because of a lack of evidence. This was based on some earlier studies5 that did not show increased stroke risk. Two more recent studies6,7 from different populations demonstrated an increased stroke risk, especially for haemorrhagic stroke. Combined with the results of Caughey et al,2 these studies build a strong case to suggest that there is a clear risk. There are consistencies across the studies, including the observation that adverse outcomes seem to vary with different NSAID classes.

The population studied by Caughey et al2 was elderly, with comorbidities and, frequently, a combination of arthritis and vascular disease. This is precisely the group of patients where the dilemma commonly arises, making the study clinically valuable. The data appear robust, and feature a sensitivity analysis that strengthens the initial findings. It should be remembered, however, that the conclusions do not apply to younger and healthier populations. Additionally, the absolute stroke risk is small, and may be exceedingly small, particularly if NSAID exposure is brief.

What is not clear from these studies is the role of confounding variables. The use of a prescription medication database and hospitalisation codes might suggest an association, but will not provide all the answers to a complex clinical scenario. Although the crude sequence ratio is robust to confounders that are stable within individuals over time, cardiovascular and stroke risk are unlikely to be stable over time and probably fluctuate. Intercurrent infection, inflammation, immune response and blood pressure variability are dynamic factors that may precipitate vascular events, particularly in predisposed individuals. Blood pressure variability8 is under increasing scrutiny as a provoking factor for stroke events. Given the important impact of COX-2 inhibition on increased blood pressure,9 this may be the true link in the relationship.

The reasons for NSAID prescription would be of great interest. Suppose, for example, that an NSAID is prescribed for analgesia in an older patient with vascular disease who has a painful arthritic condition. The NSAID may well raise blood pressure, and fluctuations in pain may result in blood pressure fluctuations; thus providing two risks for stroke.

Several NSAIDs were shown by Caughey and colleagues2 to carry stroke risk similar to the cardiovascular risk of the now-withdrawn rofecoxib. Low-dose preparations of several NSAIDs are available in Australia without prescription. The findings of Caughey and colleagues further emphasise the need for great care in the use of these agents in patients with hypertension and other stroke risks.

Author

  • David J Blacker MB BS, FRACP1,2
  • 1. Sir Charles Gairdner Hospital, Perth, WA.
  • 2. University of Western Australia, Perth, WA.

Correspondence: david.blacker@health.wa.gov.au

Competing interests

I have received support for travel and accommodation at meetings from Boehringer Ingelheim and Sanofi–Synthelabo. I have also received advisory board fees from Novo Nordisk and per-patient payments to my institution from Servier for a clinical trial.

References

View this article on Wiley Online Library