RESEARCH - Largest NSAID database study ever casts doubt on entire coxib controversy

[Guest guest]

Largest NSAID database study ever casts doubt on entire coxib controversy

Rheumawire

Jun 10, 2005

Gandey

Vienna, Austria - The largest NSAID study to dateone that is reportedly

bigger than all other studies combinedhas put coxibs back in the limelight,

this time suggesting a class effect of all nonsteroidal anti-inflammatory

drugs (NSAIDs) [1]. Presenting here at the European League Against

Rheumatism (EULAR) 2005 meeting, the researchers have mapped out for

clinicians the odds ratios for myocardial infarction (MI) of various

products, and some of what they found is quite surprising. Lead author Dr

Gurkirpal Singh (Stanford University, CA) sat down with rheumawire to

discuss his group's findings.

" What we wanted to do is put this issue into perspective and look at all of

the drugs, " Singh said. " And what we found is that drug selectivity doesn't

really mattersome were selective and some were not, but many of the drugs

increased the risk of MI. " Among the top MI offenders, Singh and his team

identified indomethacin with a 71% increased risk, sulindac 41%, and

meloxicam 37%. Among the COX-2 inhibitors, rofecoxib showed the highest risk

at 32%, followed by celecoxib at 9%. The researchers point out that

high-dose valdecoxib could not be studied since most use in the current

analysis was at the dosing level of 20 mg or less.

Commenting about the study, session cochair and incoming EULAR president Dr

Tore Kvien (Diakkonhjemmet Hospital, Oslo, Norway) said, " This is an

important high-quality database study. It provides information on all NSAIDs

associated with an increased risk of MI and also provides us with

information about the magnitude of this risk. "

During an interview Singh said, " The hysteria in the media about COX-2

inhibitors has been a bit of an overreach. All of the news about how these

drugs cause heart attacks and kill has been excessive, because we simply did

not have data for the other drugs. " He added, " Absence of evidence is not

evidence of absence. "

Using data from the California-based Medicaid program, his group conducted a

nested case control study of all patients over 18 with physician-diagnosed

arthritis treated with an NSAID between January 1999 and June 2004. Cases of

acute MI were risk-set matched with four controls on age, gender, and date

of acute MI. The researchers compared current exposure to COX-2 selective

and nonselective NSAIDs with remote exposure to any drug. The group reports

that all analyses were adjusted for 38 confounding risk factors as well as

concomitant aspirin treatment.

" This does appear to be a class effect of the NSAIDs with some variation, "

Singh said during the presentation of his findings. Summarizing the results,

he told the audience, " MI risk appears to be common in a large number of

NSAIDs, and decisions regarding their use need to be made in consultation

with a physician who knows the patient well and can take into account their

individual risk factors. "

Included in his presentation was a disclaimer emphasizing that the work

conducted by Singh and colleagues in collaboration with Dr Graham,

dubbed by the media as a US Food and Drug Administration whistleblower, does

not represent the views of the FDA. When asked by rheumawire about the

disclaimer, something that had also been used during Graham's presentations

at recent FDA hearings, Singh responded, " We put in the disclaimer because

this represents the opinion of the scientists and not that of the FDA. The

agency has a different mechanism of reaching a formal decision, and we

wanted to make that distinction clear. "

What will this mean for the FDA, EMEA, Health Canada, and others?

Singh says that compared with other regulators, the FDA has done a fairly

good job of handling the NSAIDs. " On the whole, the FDA's response has been

balanced, and it has treated all NSAIDs equally. While I think the blanket

black-box warning is a bit of an overreach, on the whole, I think they have

done well. " But Singh gives Canadian regulators a different review. " They

have taken action only on the COX-2 inhibitors, and that just doesn't make

any sense. Just because the other NSAIDs haven't been looked at doesn't mean

the others are not a potential risk. "

Speaking about the European approach, Kvien said the European Agency for the

Evaluation of Medicinal Products (EMEA) has had a heavy-handed approach that

will likely need to be reversed somewhat. " We Europeans for the most part

see the FDA statement as more consistent with the data than the current EMEA

statement. " Kvien says that as more data become available, physicians and

regulators will need to be open to reconsidering their positions.

One of the theories discussed during the question period following the Singh

presentation was the COX-1/COX-2 imbalance effect. Does adding aspirin

negate or reinforce side effects? Singh noted that his group is currently

studying this question: " At this point, we really don't know. "

" This is an issue that is no longer clear, and it is one of the things that

really confuse physicians, " Kvien said. " If you believe in the imbalance

theory, then adding aspirin should eliminate risk, but randomized trials and

database studies are showing this is not the case, so this is something that

will need to be added to the research agenda. What are the mechanisms behind

this effect? "

Who's to blame?

During the session, a participant asked, " Is it our fault? Why didn't

physicians, and rheumatologists in particular, associate patient medications

with heart attacks? " Responding to the question, Dr Hawkey (University

Hospital, Nottingham, UK) said that associations that seem obvious are often

overlooked, and he cited Crohn's disease as an example where clinicians and

researchers failed to notice that outcomes were associated with high rates

of smoking in these patients. " Sometimes we are afflicted by tunnel vision. "

During a press conference following the session, Singh told reporters that

naproxen, ibuprofen, and diclofenac would be good choices for comparators in

future clinical trials. When asked whether he thought rofecoxib should be

returned to the market, Singh answered, " It is my personal opinion that it

is unlikely that it will come back. "

Putting his work into perspective for the press, Singh explained that while

daily smoking at least doubles the risk of MI, the risk from NSAIDs is only

about 12%, " which is not really very high, " he said.

" It's important to keep this in perspective, " Kvien told rheumawire.

" Smoking increases the risk of MI by about 200% or 300%, and people are

still allowed to buy cigarettes, but some patients are not allowed to use

these drugs, which can relieve pain. "

Source

1. Singh G, Mithal A, Triadafilopoulos G. Both selective

COX-2 inhibitors andnon-selective NSAIDs increase the risk of acute

myocardial infarction inpatients with arthritis: Selectivity is with the

patient, not the drugclass. EULAR 2005; June 8-11, 2005; Vienna, Austria.

OP0091.

Not an MD

I'll tell you where to go!

Mayo Clinic in Rochester

http://www.mayoclinic.org/rochester

s Hopkins Medicine

http://www.hopkinsmedicine.org