CardioExchange Archive

CardioExchange, an NEJM practice community for medical professionals dedicated to improving cardiac patient care, was active from 2009 to 2015. CardioExchange fostered discussion between clinicians from across the globe.

September 29th, 2011

Meta-Analysis Explores Real World CV Risk of NSAIDs

A new meta-analysis sheds additional light on the cardiovascular risk of NSAIDs as used in the real world, including low doses of the most popular drugs, over short periods, and in low-risk populations.

In a paper published online in PLoS Medicine, Patricia McGettigan and David Henry analyzed data from 30 case-control studies including 184,946 CV events and 21 cohort studies with outcomes in more than 2.7 million people.

Among the NSAIDs with extensive data, rofecoxib and diclofenac had the highest CV risk while ibuprofen and naproxen had the lowest risk:

  • rofecoxib, 1.45 (95% CI 1.33-1.59)
  • diclofenac, 1.40 (1.27-1.55
  • ibuprofen, 1.18 (1.11-1.25)
  • naproxen, 1.09 (1.02-1.16)
The authors point out that diclofenac was associated with elevated risk even at over-the-counter doses and that etoricoxib “is similar to drugs that have been withdrawn because of safety concerns.”  Indomethacin also showed high risk; the authors write that the drug “has a range of gastrointestinal and central nervous system effects that, combined with the evidence presented here on cardiovascular risk, should lead to questioning of its continued clinical use.”

Categories: General
Tags: , , , , , ,


You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.

4 Responses to “Meta-Analysis Explores Real World CV Risk of NSAIDs”

  1. Barry Smith, Ph.D. says:
    September 30, 2011 at 10:56 am

    Husten’s meta-analysis is generally consistent with the NSAID-CVD risk literature more generally. In their recent meta-analysis, Trelle et al. (2011) at the University of Bern employed data from 31 trials and reached largely the same conclusion: NSAIDS as a group consistently elevate CVD risk. They also found that refecoxib presents the greatest risk of MI (2.12), but ibuprofen yields the highest association with stroke risk (3.36), while etoricoxib presented the highest risk of cardiovascular death (4.07).

    Although further research is needed to sort out important details, this literature has reached a point at which there is substantial consistency and widespread agreement concerning the CVD toxicity of NSAIDs and related drugs. This situation raises the question of how pain is to be effectively treated without incurring substantial health risk. One obvious alternative is acetaminophen. However, it lacks the anti-inflammatory properties of NSAIDS, is the single greatest cause of liver failure and is, more generally, hepatotoxic. That leaves largely opioids as effective pain killers, but they are, of course, associated with addiction.

    Given the problems with these common pain killers, it strikes me that it may be time to devote considerable effort to developing more new analgesics that do not structurally resemble either NSAIDS or acetaminophen and constitute a whole new class (or classes) of safe, effective agents for pain.

    Competing interests pertaining specifically to this post, comment, or both:
    None

  2. David Powell , MD, FACC says:
    September 30, 2011 at 9:44 pm

    Another issue is the pharmacodynamic interaction with aspirin. IF NSAIDs are necessary, I advise naproxen. If the patient or other caregiver is committed to another agent and the patient has CAD, I will advise at least ttemporally separating intake of the NSAID and aspirin (and use a PPI….and I guess check bone density if long term). I believe the ranking of the NSAIDS correlates with cox 2 to cox 1 inhibition.

  3. Dan Hackam, MD PhD says:
    October 1, 2011 at 11:36 am

    Agreed. Ensure any ASA hits the system 2 hours before the first NSAID of the day is popped. This includes COX-2 inhibitors like celecoxib which share the same PD interaction with platelet COX as ASA (recent report in PNAS). Make sure a PPI is on board or at a minimum a BID H2RA.

  4. Preciosa Coloma, MD, MSc says:
    October 7, 2011 at 7:24 am

    Some issues come to mind:

    • The NSAIDs have been with us for a long time, certainly much earlier than 1985 (the starting year of publications covered in the meta-analysis) and mostly available over-the-counter. How many NSAID-related cardiovascular events have we been missing all this time?
    • The practice of medicine is always about balancing the benefits with the risks. I think it always comes down to: is the 8-55% overall increased risk of CV events – for NSAIDs that are still in the market – worth it for the subgroup of patients who might not have alternatives for pain relief? (not to mention other risks that come with NSAID therapy) This is a decision that has to be made on an individual basis.
    • Maybe it is high time indeed for developing new analgesics that are structurally different from the NSAIDs. Question is, will drug companies be willing to take the exceptionally challenging task of establishing that their new drug does not have the same adverse effect profile as the NSAIDs?

    Competing interests pertaining specifically to this post, comment, or both:
    None