Life without COX 2 inhibitorsBMJ 2006; 332 doi: https://doi.org/10.1136/bmj.332.7553.1287 (Published 01 June 2006) Cite this as: BMJ 2006;332:1287
All rapid responses
Shaugnessy and Gardener1 are to be commended for emphasising the non-
pharmacological treatment of arthritis – the daily mantra of
rheumatologists. Unfortunately, they misrepresent the data from the meta
–analysis of Kearney and colleagues2 that generated their leading article.
Their introductory statement that COX 2 inhibitors “have been associated
with an increased risk of myocardial infarction with prolonged use
compared with placebo or other NSAIDs” is both misleading and inaccurate,
as it is not supported by published data. The conclusions from Kearney’s
meta-analysis was actually that, with the exception of naproxen, there was
no difference between COX 2 inhibitors and NSAIDs in relation to vascular
events. The only surprise in this study was the apparent reduction in risk
from naproxen as this has not been observed in large population and case
controlled studies. The results from a retrospective observational study
in the Ontario population did not find any reduction in risk of a
myocardial infarction from naproxen compared with COX 2 inhibitors3. A
prospective large population based study conducted by the FDA in
California4 found a significant association between naproxen and
myocardial infarction, in addition to a number of other NSAIDs and
rofecoxib, but not with celecoxib, an observation that led to a black box
warning on all NSAIDs and COX 2 inhibitors in the USA. This finding was
similar to that from a nested case control study from a GP data-base in
the UK5. The important conclusion from all this data is that, with the
exception of high dose rofecoxib, there is no greater risk of myocardial
infarction from COX 2 inhibitors than non-selective NSAIDs. Whether this
observation is from the drugs or the arthritis however is not clear as
there is an established association between arthritis and cardio-vascular
Shaugnessy and Gardner also dismiss the argument for prescribing COX
2 inhibitors: “ The common assumption that COX 2 inhibitors are safer than
other NSAIDs has not been borne out.” This statement is referenced with a
critical review of the CLASS study yet they fail to reference a range of
other studies that demonstrate increased safety from COX 2 inhibitors. It
could be argued that the data from a case control study in a UK based GP
register6 supports their argument as the authors concluded that there was
no strong evidence for any enhanced benefit from COX 2 inhibitors.
However, although the study included symptomatic ulceration in addition to
serious complications, they observed that naproxen had the highest risk
and did not find an association between the use of celecoxib and GI
complications. The VIGOR7 and TARGET8 studies of rofecoxib and lumiracoxib
respectively both found an increased rate of GI complications from non-
selective NSAIDs. This potential benefit of COX 2 inhibitors was an
observation also supported by the Ontario population study9.
The major concern is the significant toxicity from non-selective
NSAIDs. They are arguably the most toxic drugs prescribed in primary care
and their toxicity is a major public health problem. Of course we need
more and better outcome data to determine which group of patients should
be considered for COX 2 inhibitors. Reducing the use of all these drugs is
an important message, but switching patients from a COX 2 inhibitor to a
non-selective NSAID confers no benefit to the patient and may increase the
risk of a serious and potentially life threatening GI complication.
1.Shaugnessy AF Gordon AE Life without COX 2 inhibitors BMJ
2006;332:1287-8 (3 June)
2. Kearney PM Baigent C Godwin J Halls H Emberson JR Patrono C. Do
selective cyclo-oxygenase inhibitors and traditional NSAIDs increase the
risk of atherothrombosis? Meta-analysis of randomised trials. BMJ
3. Mamdani M, Rochon P, Juurlink DN, Anderson GM, Kopp A, Naglie G, Austin
PC, Laupacis A.Effect of selective cyclooxygenase 2 inhibitors and
naproxen on short-term risk of acute myocardial infarction in the elderly.
Arch Intern Med. 2003;163:481-6
4. Graham DJ,Campen D, Hui R, Spence M, Cheetham C, Levy G, Shoor S, Ray
WA. Risk of acute myocardial infarction and sudden cardiac death in
patients treated with cyclo-oxygenase 2 selective and non-selective non-
steroidal anti-inflammatory drugs: nested case-control study. Lancet 2005;
5. Hippisley-Cox J, Coupland C Risk of myocardial infarction in patients
taking cyclo-oxygenase-2 inhibitors or conventional non-steroidal anti-
inflammatory drugs: population based nested case-control analysis. BMJ.
6. Hippisley-Cox J, Coupland C, Logan R Risk of adverse gastrointestinal
outcomes in patients taking cyclo-oxygenase-2 inhibitors or conventional
non-steroidal anti-inflammatory drugs: population based nested case-
control analysis BMJ 2005;331:1310-1316
7. Bombardier Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R,
Davis B, Day R, Ferraz MB, Hawkey CJ, Hochberg MC, Kvien TK, Schnitzer TJ;
VIGOR Study Group Comparison of upper gastrointestinal toxicity of
rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study
Group. N Engl J Med. 2000 Nov 23;343(21):1520-8
8. Hawkey Schnitzer TJ, Burmester GR, Mysler E, Hochberg MC, Doherty M,
Ehrsam E, Gitton X, Krammer G, Mellein B, Matchaba P, Gimona A, Hawkey CJ;
TARGET Study Group Comparison of lumiracoxib with naproxen and ibuprofen
in the Therapeutic Arthritis Research and Gastrointestinal Event Trial
(TARGET), reduction in ulcer complications: randomised controlled trial.
9. Muhammad Mamdani, Paula A Rochon, David N Juurlink, Alex Kopp, Geoffrey
M Anderson, Gary Naglie, Peter C Austin, and Andreas Laupacis
Observational study of upper gastrointestinal haemorrhage in elderly
patients given selective cyclo-oxygenase-2 inhibitors or conventional non-
steroidal anti-inflammatory drugs. BMJ 2002; 325:624-8
Clinical investigator COX 2 inhibitors and occasional paid lectures.
Competing interests: No competing interests
EDITOR—In their editorial “Life after coxibs” Shaughnessy and Gordon
deliver a narrative clinical review of alternatives to coxibs in arthritis
and chronic pain management (1). We agree with the authors that little is
lost if coxibs disappear. However, narrative reviews have the inherent
methodological weakness that selection of therapies and recommendations
may be circumstantial, and this editorial about alternatives to coxibs is
no exception to the rule. The editorial lacks substantial information
about priority in based on effect sizes, time-effect profiles and power of
evidence behind each of the 13 recommended therapies. And it does not seem
to make a difference which therapy is chosen. In our opinion, it actually
makes a difference which one of these therapies is chosen, and some of the
authors´ recommendations do not seem justified by evidence.
In the editorial on-drug options, such as exercise and weight loss are
rightfully claimed to be efficacious, but we think that they should be
mentioned as firstline alternatives in knee osteoarthritis as adviced by
Felson & Hunter in a recent clinical BMJ-review (2).
The authors recommend combining non-selective cox inhibitors with
misoprostol and other recommendations include topical NSAIDs,
acetaminophen (paracetamol), opioids and glucosamine sulphate (GS). The
above pharmacotherapies in addition to steroid injections and chondrotin
are the seven most common pharmacotherapies in Europe (3), and they have
recently been systematically reviewed and subjected to meta-analysis (4).
Baseline pain levels, effect sizes, time-effect profiles and power of
evidence have been thoroughly analysed, and some of these results
contradict Shaughnessy & Gordon´s recommendations.
Both GS and chondroitin were found in this meta-analysis to exhibit
rather small effect sizes of questionable clinical relevance. The Cochrane
review cited by the editorial did not find unequivocal positive evidence
from GS in general, as 8 out of 15 trials were negative. The positive
results came from 7 trials funded by a specific GS manufacturer. The large
independently-funded GAITT trial published in February 2006 (5) did not
find significant pain-relieving effects from neither GS nor chondrotin.
Paracetamol recommendations are common in guidelines in spite of a very
poor effect size (4).
The authors also claim that S-adenosylmethionine (SAMe) reduces pain
in osteoarthritis, but their own review reference and another meta-
analysis (6) actually found no pain-relieving effect from SAMe over
Some complementary therapies like therapeutic taping and braces are
also claimed to be effective, and acupuncture is suggested to give a small
pain-relieving effect, while the authors point out limited evidence in
favour of therapeutic touch (TT), and electrical stimulation.
There are several question-marks associated with the assessment of
non-drug therapies in the editorial. We have found 36 placebo-controlled
trials in knee osteoarthritis with eight physical therapies. So why are
the authors recommending non-drug interventions with only one or two
published trials? For instance, the evidence behind TT is a single Medline
-indexed randomized controlled trial with TT with 25 knee osteoarthritis
patients (7), and thus has not been subjected to assessment of
Therapeutic taping was also recommended in the editorial. But it has
only been tested in a single positive monotherapy trial (8), and the
effect was lost when combined with other interventions in a later trial by
the same group (9).
Electrical stimulation is also given a limited recommendation by the
authors while citing a Cochrane review with 3 trials (10). This cited
reference is in fact a review of electromagnetic fields therapy in
osteoarthritis, and not electrical stimulation.
The authors also fail to mention non-drug interventions like TENS and
low level laser therapy which have been systematically reviewed by
Cochrane (11) (12) and others (13) finding clinically relevant pain-
relieving effects from samples of 8 and 6 trials respectively.
It is becoming increasingly difficult for a general practitioner to
navigate towards “best practice” in the plethora of effectiveness claims.
The unfortunate truth is that several drug interventions are largely
ineffective. This is highlighted by findings where 1 in 4 osteoarthritis
patients uses self-medication, 1 in 2 osteoarthritis patients uses two
drugs or more for their osteoarthritis pain (OA Nation survey UK,
Arthritis Care 2003). In addition, more than half of chronic pain patients
(where arthritis is the largest diagnostic group) do not feel that their
pain is adequately controlled by drugs (14).
Patients expect GPs and other clinicians to give meaningful advice
about medical as well as physical and complementary treatment. A
comprehensive systematic review comparing physical and complementary
therapies in OAK management is clearly needed. We agree that arthritis
management needs to be individually tailored, and in some cases therapies
which seem ineffective on a group level, could be of value to the
individual patient. But it would have been more helpful if this editorial
had offered evidence-based guidance for osteoarthritis management based on
group level efficacy. As it stands the editorial adds to the chaos of life
after coxibs by their (almost) “anything goes” advice.
1. Shaughnessy AF, Gordon AE. Life without COX 2 inhibitors. Bmj
2. Hunter DJ, Felson DT. Osteoarthritis. Bmj 2006;332(7542):639-42.
3. Mazieres B, Schmidely N, Hauselmann HJ, Martin-Mola E, Serni U,
Verbruggen AA, et al. Level of acceptability of EULAR recommendations for
the management of knee osteoarthritis by practitioners in different
European countries. Ann Rheum Dis 2005:ard.2003.009431.
4. Bjordal JM, Klovning A, Ljunggren AE, Slordal L. Short-term
efficacy of pharmacotherapeutic interventions in osteoarthritic knee pain:
A meta-analysis of randomised placebo-controlled trials. Eur J Pain 2006.
5. Clegg DO, Reda DJ, Harris CL, Klein MA, O'Dell JR, Hooper MM, et
al. Glucosamine, chondroitin sulfate, and the two in combination for
painful knee osteoarthritis. N Engl J Med 2006;354(8):795-808.
6. Witte S, Lasek R, Victor N. [Meta-analysis of the efficacy of
adenosylmethionine and oxaceprol in the treatment of osteoarthritis].
7. Gordon A, Merenstein JH, D'Amico F, Hudgens D. The effects of
therapeutic touch on patients with osteoarthritis of the knee. J Fam Pract
8. Hinman RS, Crossley KM, McConnell J, Bennell KL. Efficacy of knee
tape in the management of osteoarthritis of the knee: blinded randomised
controlled trial. Bmj 2003;327(7407):135.
9. Bennell KL, Hinman RS, Metcalf BR, Buchbinder R, McConnell J,
McColl G, et al. Efficacy of physiotherapy management of knee joint
osteoarthritis: a randomised, double blind, placebo controlled trial. Ann
Rheum Dis 2005;64(6):906-12.
10. Hulme J, Robinson V, DeBie R, Wells G, Judd M, Tugwell P.
Electromagnetic fields for the treatment of osteoarthritis. Cochrane
Database Syst Rev 2002(1):CD003523.
11. Osiri M, Welch VV, Brosseau L, Shea B, McGowan J, Tugwell P, et
al. Transcutaneous electrical nerve stimulation for knee osteoarthritis
(Cochrane Review). Cochrane Database Syst Rev 2000;4.
12. Brosseau L, Robinson V, Wells G, Debie R, Gam A, Harman K, et al.
Low level laser therapy (Classes I, II and III) for treating rheumatoid
arthritis. Cochrane Database Syst Rev 2005(4):CD002049.
13. Bjordal JM, Lopes-Martins RAB, Klovning A. Is Quality Control of
Cochrane Reviews in Controversial Areas Sufficient? Journal of Alternative
and Complementary Medicine 2006;12(2):181-183.
14. Breivik H, Collett B, Ventafridda V, Cohen R, Gallacher D. Survey
of chronic pain in Europe: prevalence, impact on daily life, and
treatment. Eur J Pain 2006;10(4):287-333.
Competing interests: No competing interests
The authors do not mention those patients who suffer severe
anaphylactoid reactions to aspirin and conventional NSAIDs (non steroidal
anti inflammatory drugs) acting on cyclo-oxygenase-1 pathways. Such
patients with arthritis could possibly be disadvantaged by a total
withdrawl of these newer anti-inflammatories.
Competing interests: No competing interests
Shaughnessy and Gordon write in their editorial on the loss of COX 2
inhibitors that “there is good evidence that other pharmacological and non
-drug options may be reasonably effective, equally safe and less costly”¹.
They go on to give examples of both drug and non-drug measures which have
been shown to be effective in osteo-arthritis. However, their discussion
of opioids was not referenced. We find the omission of studies supporting
the use of opioids surprising, particularly when references for the non-
drug measures discussed were included, even when the effect sizes were
reported as small, or the data limited by small numbers.
Two systematic reviews of opioids in chronic non-cancer pain²,³
report several papers showing efficacy of opioids (morphine and oxycodone)
in osteoarthritis, demonstrating an average 30% reduction in pain
intensity, a level generally considered to be clinically meaningful4.
Whilst the authors of the Kalso systematic review note the worries of
addiction and drug diversion (presumably the reason they are referred to
as “a last pharmacological resort” by Shaughnessy and Gordon) and caution
that not all patients respond to opioids, in a more recent BMJ editorial
Professor Kalso observes that the British Pain Society has published
recommendations for the appropriate use of opioids in persistent non-
cancer pain5. The guidelines offer a framework for the safe prescribing of
opioids in conditions such as osteoarthritis6. A recent paper highlighted
that a quarter of GP’s sampled did not prescribe opioids for patients with
persistent chronic pain, and that prescription patterns were influenced by
the doctor’s beliefs about the appropriateness of opioids in chronic pain,
in spite of these guidelines7.
We are currently conducting a trial focusing on the patient’s
acceptability of opioids for osteoarthritis pain, which will add to this
debate and highlight the patient’s experience of opioid treatment. In
addition, one of us (CR) has recently completed a qualitative study
examining the views of patients with cancer pain when offered morphine.
Interestingly, the phrase most commonly used by these respondents was
“last resort” and this association meant that these patients delayed the
use of drugs like morphine for as long as possible, suffering from
uncontrolled pain as a consequence. Given the prejudice of this BMJ
editorial, perhaps we should not have been surprised that some of these
patients seemed to be reflecting the views of their physicians.
1. Shaughnessy, A.F. Gordon, A. E. Life without COX 2 inhibitors BMJ
2. Kalso, E. Edwards, J.E. Moore, R.A. and McQuay, H.J. Opioids in
chronic non-cancer pain: systematic review of efficacy and safety. Pain
2004; 112: 372-380
3. Bloodworth D. Issues in opioid management. American Journal of
Physical Medicine and Rehabilitation 2005;84:S42-S55.
4. Farrar, J.T. Portenoy, R.K. Berlin, J.A. Kinman, J.L. Strom,
B.L.Defining the clinically important difference in pain outcome measures.
Pain 2000;88: 287-94
5. Kalso, E. Opioids for persistant non-cancer pain BMJ 2005;330: 156
6. The Pain Society. Recommendations for the appropriate use of
opioids in persistent non-cancer pain. A consensus statement prepared on
behalf of the Pain Society, the Royal College of Anaethetists, the Royal
College of General Practitioners and the Royal College of Psychiatrists.
March 2004. www.britishpainsociety.org/pdf/opioids_doc_2004.pdf
7. Hutchinson, K. Moreland, A. Williams, A, Wienman, J. and Horne R.
(2006) Exploring beliefs and practice of opioid prescribing for persistent
non-cancer pain by general practitioners. European Journal of Pain. In
Press, Corrected Proof, Available online 17 February 2006
Competing interests: No competing interests
The doubts about the safety of NSAIDs are having a significant impact
on the ability of doctors to treat acute and chronic pain. It is a pity
therefore that Messrs Shaughnessy and Gordon did not mention that there is
a specialty of medicine dedicated specifically to managing pain.
A large proportion the work of my pain management colleagues and I is
spent working with older patients, helping them manage their problems
which are primarily degenerative musculo-skeletal. Our focus is partly on
the use of therapeutic agents (such as analgesics, acupuncture etc), but
more on the understanding of the pain problems, the reality of the
underlying situation (we are all getting old) and then addressing the
lifestyle changes that need to be made, dealing with the associated
depression and so on.
Perhaps one of the difficulties is the failure to see chronic pain as
a “chronic disease” like COPD, asthma, diabetes, stroke etc. Patients and
doctors often think that the pain can be cured with a well-placed
injection or a simple treatment. However, commonly the causes not only
don’t go away but also may progress. Therefore patients need help with
pain management rather than fruitless attempts to find a cure. Pain can
also be lethal disease. As examples, depression and despair can lead to
suicide and NSAIDs can lead to gastro-intestinal catastrophe.
Life without COX2 inhibitors will not change my personal practice.
Over the years I have spent much more of my time taking patients off these
drugs rather than initiating them. I now worry though that the reduced
usage of NSAIDs by GPs is going lead to an increased referral rate to our
Pain Clinics as they struggle to help their patients.
The authors are wrong to consider that opioids are "the last
pharmacological resort". There many other treatments that would qualify
for this title. Used appropriately, opioids are probably much safer than
NSAIDs even over the long term. However, doctors need to have the
appropriate skill and training in their use. This highlights the serious
deficiency in teaching about pain management in the curricula of medical
schools and postgraduate education programs (1). Personally I feel that no
doctor should qualify without being able to demonstrate the ability to use
the drug morphine for acute and chronic pain. Yet day after day I and my
colleagues see the most basic mistakes being made.
Thomas Sydenham, a 17th century physician wrote “Among the remedies
which it has pleased Almighty God to give to man to relieve his
sufferings, none is so universal and so efficacious as opium.” William
Serturner first purified morphine in May 1805. Both must be turning in
their graves knowing that even now in the 21st century we still can’t get
1. Notcutt WG. Purchasers should require providers to set standards for
BMJ, Feb 1997; 314: 442
Competing interests: No competing interests
The theme of this Editorial is that COX 2 inhibitors are not a
superior option for the management of musculoskeletal pain, and that
paracetamol can be equally effective and much safer. From personal
experience I suggest that for some of us, COX 2 inhibitors provide
significantly better pain relief than can be secured by conventional
NSAIDs or by paracetamol.
As a result of sports activity I suffer from bilateral shoulder pain
of variable severity. It was partially alleviated by ibuprofen in a dose
of 800 mg four times a day. I never suffered gastrointestinal side effects
from ibuprofen, but when it was demonstrated that the drug could nullify
the cardioprotective effects of low-dose aspirin, I switched to rofecoxib,
then thought to be a safe drug. I took two tiny tablets of 12.5 mg each,
once a day. This more convenient medication produced complete pain relief.
Furthermore, the onset of action was in approximately 30 minutes. Thus on
three counts - convenience, pain relief and immediacy of action -
rofecoxib was definitely superior to ibuprofen.
Then came alarming reports (1) that rofecoxib in particular, and
perhaps COX 2 inhibitors in general, were associated with an increased
risk of myocardial infarction, so I switched to paracetamol. I am now
constrained to take two 500 mg tablets four times a day and the control of
pain is incomplete. I conclude that rofecoxib is a highly effective drug.
Oh for a COX 2 inhibitor that does not promote myocardial infarction!
1 Kearney PM, Baigent C, Godwin J, Emberson JR, Patrono C. Do coxibs
and traditional non-steroidal anti-inflammatory drugs increase the risk of
atherothrombosis? Meta-analysis of randomised trials. BMJ 2006;332:1302-5.
Competing interests: No competing interests