Gastrointestinal (GI) safety has been the main selling point of some NSAID and COXIB preparations. Not surprisingly, it has been made to appear that this is the only thing we should be watching out whenever we prescribe an NSAID. But truth be told, safety concerns over NSAIDs reached an all time high when the cardiovascular (CV) safety of Rofecoxib (Vioxx, Merck) was questioned and this lead to its eventual withdrawal from the market. Some believe that whatever NSAIDs are still being marketed today had survived the scrutiny for CV risks and are considered “safe”.
However, I still believe that CV safety remains a relevant issue today as it was several years back. No completed randomized trial up to today has looked into the CV safety of NSAIDs as a primary endpoint (Although, there is one presently ongoing). What we instead have are attempts to quantify CV risks by performing meta-analyses on safety which is a common secondary end point. For the EBM user or the epidemiologist, we may argue that such trials are not empowered to detect significant differences in safety. So the results of such meta-analyses may not be considered “hard” conclusions. But I guess until such data becomes available, it’ll be worth our time to look into the results of such studies.
The study by McGettigan and Henry (PLOS Medicine, 2013) derived the relative risk (RR) for cardiovascular events of NSAIDS based on meta-analyses of controlled observational and randomized trials. Likewise, it compared the usage/ sales of these NSAIDs included in Essential Medicine Lists (EMLs) of 15 countries (the Philippines included).
Based on the derived RR, three NSAIDs had consistently higher CV risks – Rofecoxib, Etoricoxib (Arcoxia, MSD) and Diclofenac. Naproxen was judged to have the lowest risk – being considered risk-neutral in 5 of 6 metas. Indomethacin and Meloxicam had moderately increased RR that were significantly greater than Naproxen. While Celecoxib (Celebrex, Pfizer) and Ibuprofen were associated with elevated RR values when used in clinical trials in high doses but not in the lower doses typically used in the community. It would appear then that the dose of the NSAID appears to affect CV risks. But as pointed out by the same authors in a separate paper, for Diclofenac such “no-risk doses” were not found unlike those for Celecoxib and Ibuprofen. So, I guess, the old rule that we use the lowest effective dose of the NSAID for the shortest duration of time still holds.
Among the three high-risk NSAIDs, the RR for CV events of Diclofenac approximated those of Rofecoxib. While an updated 2011 meta-analyses found a doubling of CV risk when using Etoricoxib compared with non-use, a large head-to-head RCT found it to have similar CV risks to that of Diclofenac.
These results are troubling considering that almost 1/3 of total NSAID use across the 15 countries surveyed were for Diclofenac and Etoricoxib. In fact, Naproxen, Celecoxib and Ibuprofen accounted for only 9.4%, 7.2% and 11%, respectively of total NSAID use. Another surprising finding was culled by the authors from the WHO site on the EMLs of 100 countries. Diclofenac was listed in the EMLs of 74 countries while the safe Naproxen was only included in the EMLs of 27 countries.
As a side note, the European Medicine Agency (EMA) finalized a review in October 2012 that concluded that there was a small increase of cardiovascular side effects with Diclofenac compared to other NSAIDs.
Practical Note. How then should we proceed to use NSAIDs? I would like to quote a statement from the EMA which was included in the 2008 OARSI Guidelines on the Management of Knee and Hip Osteoarthritis.
The important note here is that, aside from knowing details about the of NSAIDs, we should likewise profile our individual patients and select the drug that least increases their risks for adverse events. I had previously posted on this blog about the choice of NSAIDs based on CV risk (see here). It would be really helpful if we use those CV risk calculators not only when we are evaluating patients for surgery but likewise when we are prescribing NSAIDs. If the 10 year risk for CV events is >10% (considered high risk), for now, we should choose Naproxen. GI risk is still a concern but risk mitigation by concomitantly using proton pump inhibitors with NSAIDs has been accepted. No such risk mitigation has been developed for the heart.