NPC Archive Item: Less anaemia with celecoxib vs diclofenac plus PPI, but no difference in GI complications (and don’t forget CV risks)

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

6 July 2010a

The CONDOR study found a lower risk of anaemia, mainly of presumed occult GI origin, with celecoxib compared to diclofenac plus PPI. However, rates of GI haemorrhage, obstruction or perforation were identical. Taking into account the increased CV risks of many NSAIDs, including celecoxib and diclofenac, using lower dose ibuprofen or naproxen plus a PPI is likely to present the lowest overall risk of serious adverse events for most patients.

Level of evidence:
Level 3 (other evidence) according to the SORT criteria.

Action
Health professionals should continue to follow existing NICE and other guidance and use a proton pump inhibitor (PPI) for gastroprotection in obligate NSAID users, especially those at increased risk of adverse gastrointestinal (GI) events. The MHRA has warned about an increased risk of cardiovascular (CV) events with coxibs (including celecoxib) and many traditional NSAIDs (including diclofenac). Considered together, if an NSAID is essential, the lowest GI and CV risks are associated with ibuprofen 1200mg/day or less, especially if a PPI is co-prescribed. The next lowest CV and GI risks overall are probably associated with naproxen plus a PPI.

What is the background to this?
NSAIDs are well known as causes of GI adverse effects including dyspepsia and, more seriously, erosions including ulcers and haemorrhage. Some NSAIDs are more toxic to the gut than others: ibuprofen at 1200mg/day or less is associated with the lowest risk of all traditional NSAIDs. Cox-2 selective NSAIDs (‘coxibs’) were developed because, on pharmacological principles, selectivity for the Cox-2 enzyme would be expected to confer greater GI safety. As a class, coxibs have been found to have a lower risk of upper GI complications than traditional NSAIDs as a class. However, coxibs (including celecoxib) and many traditional NSAIDs (including diclofenac) are associated with an increased risk of adverse CV events, particularly thrombotic events (e.g. myocardial infarction and stroke). Ibuprofen 1200mg/day or less and naproxen 1000mg/day appear to be associated with a lower CV risk.

PPIs have been shown to reduce the risk of NSAID-related gastric and duodenal ulcers, and dyspepsia. They are widely recommended to reduce the risk of GI adverse effects in obligate NSAID users, especially those at increased risk (see NICE guidance on osteoarthritis and rheumatoid arthritis).

This study set out to compare celecoxib 200mg twice daily with diclofenac modified release 75mg twice daily plus omeprazole 20mg once daily, over six months. The intention was to look at the effects of these treatments on the small bowel and colon as well as the stomach and duodenum. It recruited 4,484 patients with osteoarthritis or rheumatoid arthritis, aged 60 years or older or 18 years or older with previous gastroduodenal ulceration or GI haemorrhage.

What does this study claim?
The primary outcome was a composite of gastroduodenal, small bowel or large bowel haemorrhage or perforation; gastric outlet obstruction; clinically significant anaemia of defined GI or presumed occult GI origin; and acute GI haemorrhage of unknown origin.

This primary endpoint occurred in 3.8% of patients receiving diclofenac plus omeprazole and 0.9% of patients receiving celecoxib (hazard ratio [HR] 4.3, 95% confidence interval [CI] 2.6 to 7.0; P<0·0001, number needed to harm [NNH] 34). However, this result was driven entirely by differences in rates of anaemia, mainly of presumed occult GI origin rather than documented lower GI bleeds (see study details). Rates of gastroduodenal and large bowel haemorrhage were identical in the two study groups (0.13% and 0.04% respectively) and there were no episodes of gastroduodenal, small-bowel or large-bowel perforation; small-bowel haemorrhage; or gastric-outlet obstruction.

So what?
This is not a study to change routine practice from existing NICE and other guidance. The usual approach to reduce the risk of adverse GI effects in obligate NSAID users, especially those at increased risk, should be to consider co-prescribing a PPI. The study has a number of limitations. In particular, the main driver of the composite outcome, anaemia, is not a true patient-oriented outcome: it does not provide direct evidence as to whether or not patients are likely to live longer or live better with one treatment rather than the other. Moreover, ‘anaemia of presumed occult GI origin’ was a diagnosis of exclusion, i.e. the source of blood loss and the reason for the anaemia was not confirmed and assumptions were made that it was of GI origin. In addition, the study excluded patients taking aspirin: the lower GI risks of celecoxib are diminished when it is co-prescribed with aspirin. In addition, the study lasted only six months which makes drawing conclusions about long term safety more difficult.

Health professionals and patients should weigh carefully the lower risk of anaemia (largely of presumed occult GI origin) seen in this study against the increased cardiovascular risks associated with coxibs, taking into account the person’s baseline risk of GI or CV adverse events. The MHRA has issued warnings about the CV safety of NSAIDs, including diclofenac and celecoxib. A recent observational study has suggested that diclofenac is associated with a similar excess CV risk to rofecoxib (withdrawn from the market worldwide due to an increased risk of serious thrombotic events). In that study, celecoxib was not associated with a statistically significantly increased risk of CV death. However, as discussed the the NPCi blog on this study, this was based on relatively few numbers of events, and the study neither rules out nor confirms that celecoxib has a lower CV risk than either diclofenac or rofecoxib.

As we discuss on the musculoskeletal section of NPC, many people with pain can be managed without an NSAID. However, if an NSAID is necessary, GI and CV risks can be minimised by using ibuprofen 1200mg/day or less, especially if a PPI is added. If the person does not respond adequately to ibuprofen at this dose, naproxen would seem to be preferable to other NSAIDs, including celecoxib, insofar as minimising the CV risk. Naproxen carries a greater GI risk than lower dose ibuprofen or celecoxib, but this can be mitigated substantially by co-prescribing a PPI.

Direct comparative studies of the GI risks of traditional NSAIDs plus PPIs versus coxibs are scarce. An RCT of 224 patients with a history of NSAID-related complicated peptic ulcers found no significant difference between groups in the proportion of patients who developed recurrent ulcer complications on celecoxib or on naproxen plus PPI (lansoprazole) over a median of 24 weeks (4% vs. 6%, respectively, P=0.37). However, significantly more patients receiving celecoxib developed dyspepsia (15% vs. 6%, P=0.02). There is no robust evidence that prescribing a coxib plus a PPI offers any significant advantage over prescribing a traditional NSAID plus a PPI in preventing GI complications.

Finally, according to the June Drug Tariff, naproxen tablets 500mg twice a day plus omeprazole 20mg capsules daily costs the NHS around £6.24 per 28 days, whereas celecoxib 200mg twice a day costs £40.23 per 28 days.

Study details
Chan FKL et al. Celecoxib versus omeprazole and diclofenac in patients with osteoarthritis and rheumatoid arthritis (CONDOR): a randomised trial. Lancet early online publication, 17 June 2010 doi:10.1016/S0140-6736(10)60673-3

Design Double-blind, triple-dummy, parallel group randomised controlled trial

Patients 4,484 patients with osteoarthritis or rheumatoid arthritis, aged 60 years or older or 18 years or older with previous gastroduodenal ulceration or GI haemorrhage.

Intervention and comparison Celecoxib 200mg twice daily, compared with diclofenac modified release 75mg twice daily plus omeprazole 20mg once daily, over six months.

Outcomes and results The primary outcome was a composite of gastroduodenal, small bowel or large bowel haemorrhage or perforation; gastric outlet obstruction; clinically significant anaemia of defined GI or presumed occult GI origin; and acute GI haemorrhage of unknown origin. Clinically significant anaemia was defined as a decrease in haemoglobin of 2g/dL or more, or a decrease in haematocrit of at least 10 percentage points.

This primary endpoint occurred in 3.8% of patients receiving diclofenac plus omeprazole and 0.9% of patients receiving celecoxib (HR 4.3, 95%CI 2.6 to 7.0; P<0·0001, NNH approximately 34). This result was driven entirely by differences in rates of anaemia:  0.22% vs 1.07% for anaemia of defined GI origin (RR 4.9, NNH 118), and 0.45% vs 2.36% for anaemia of presumed occult GI origin (RR 5.24, NNH 52), Rates of gastroduodenal and large bowel haemorrhage were identical in the two study groups (0.13% and 0.04%, respectively) and there were no episodes of gastroduodenal, small-bowel or large-bowel perforation; small-bowel haemorrhage; or gastric-outlet obstruction.

Sponsorship Pfizer Inc. Two of the six authors are Pfizer employees and the others are listed as consultants to Pfizer.

More information on NSAIDs and gastroprotection can be found on the musculoskeletal pain section of NPC.

Please comment on this blog in the NPC discussion rooms, or using our feedback form.

Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you