Many adverse reactions have been discovered by spontaneous reporting, and this has led to important public health issues and drugs being removed from the market, in some cases issues are not so clear-cut as they may seem: in this respect the case of nimesulide hepatotoxicity is quite interesting.
The European authorities asked for a study of drug-related acute liver failure leading to registration for transplantation (ALFT), the Study of Acute Liver Transplantation (SALT). The advantage of ALFT is that compared to less severe cases of hepatotoxicity, all cases are in specific centres and on lists, so it is possible to identify every single case in a country, and be sure not to miss any. This can be compared to the sales and usage statistics of drugs in the same country, in what is called a case-population design.
SALT was thus a multinational case-population study in France, Greece, Ireland, Italy, Netherlands, Portugal, and UK between 2005-2007 in adults. Patients with acute liver failure were identified through national/local liver transplant (LT) registries. Possible clinical causes (ICC) for the liver failure, and exposure to NSAID or paracetamol or other drugs was determined within 30 days prior to index date (ID), the date of the first clinical or laboratory symptoms of liver disease. Overall and per-drug sales for NSAIDs and for paracetamol were obtained in each country, so that per country rates of NSAID-exposed ALFT were estimated, expressed in cases per million treatment-years (MTY).1,2
Of the 62 LT centres in these seven countries, 57 were eligible, and 52 contributed data. Among the 9479 patients identified from LT lists for the 3-year period, 600 (6%) were ALF: 219 (36% of all-cause ALF) had an ICC, 62 (11%) were not drug-exposed and had no ICC, and 301 (52%) were drug-exposed without ICC. Among the latter, 40 were exposed to NSAIDs. Event rates per MTY were 1.59 (95% confidence interval (CI), 1.14-2.17) for all NSAIDs pooled, 2.28 (1.21-3.90) for ibuprofen, 1.88 (0.81-3.70) for nimesulide, 1.55 (0.57-3.38) for diclofenac. Event rate for non-overdose paracetamol was 3.31 (2.63-4.11), and 7.84 (6.77-9.04) when intentional or non-intentional overdoses were included. Event rates for all NSAID pooled were almost four (95%CI 1.2-12.5) times higher in Ireland than all countries pooled.3
The risk of ALFT was very rare and no differences were found among the most used NSAIDs such as nimesulide, diclofenac or ibuprofen. So why was there such a difference between spontaneous reports and epidemiological studies? There might be per-country differences in patient sensitivity related to genetic predispositions or environmental factors (such as beer?), or there might be more human factors, such as notoriety bias where media communication on a problem increases its reporting, or in countries with few specialists the intimate conviction by one that a given drug is dangerous, thereby increasing targeted reporting of cases exposed to the culprit drug rather than to other drugs.
In pharmacovigilance as in other matters, healthy scepticism needs to take over: without neglecting the precautionary principle, the possibility of biases must be considered, and properly designed studies need to take over.
In the case of nimesulide, the chances are it isn't worse than the others. But nothing in medicine is certain, and certainly the topic must be kept under surveillance. In particular, the finding of a two to three-fold higher event rate with normal dose paracetamol also certainly needs scrutiny.4
The publications from the SALT study are referenced 3-5
- Eur J Clin Pharmacol 2013;69:605-16. CDI #nnn#
- Pharmacoepidemiol Drug Saf 2012;21:851-6. CDI #nnn#
- Drug Saf 2013;36:135-44. CDI #fnn#
- Brit Med J 2013;346:f1519. CDI #nnn#
- Pharmacoepidemiol Drug Saf 2013;22:160-7. CDI #nnn#
NIMESULIDE AND HEPATOXICITY
Nimesulide has been evaluated several times at European level because of its hepatoxicity.
1998-1999 First alerts for hepatoxicity from Israel, where the drug had been suspended and then reintroduced on the market in a second time. At the same time some cases in Portugal and some others in one centre in Belgium were reported.
2002 Report of cases concentration (66) in Finland, which decides nimesulide suspension and requires a safety profile evaluation to the CHMP of EMA. In the same year Spain withdraws nimesulide.
2002-2003 The CHMP concludes that the benefit/risk profile of nimesulide is still positive, provided that usage recommendations are modified.
2004 The CHMP conclusion is sustained by European Commission that in April 2004 publishes its legally-binding decision. The product information is modified adding counter-indication for patients with hepatic insufficiency and warnings for risks of hepatitis, acute hepatitis (including fatal cases), jaundice and cholestasis. The daily dose is now limited to a maximum of 200 mg/day.
2007 The Irish Agency of Medicines decides to suspend the nimesulide commercialization after 6 cases of severe hepatic insufficiency that required transplant (reported by the National Liver Transplant Unit of the St Vincent Hospital) and an overall revision of all case-reports collected by the Irish Agency.
The CHMP and the European Commission conclude that the nimesulide benefits exceed the risks but indications must be further restricted: the drug must be used only as second-line treatment and a new procedure must be opened to assess all potential risks, not the hepatic ones only. During the evaluation, the CHMP requires that a study is conducted on European Transplant Centres. Meantime, Italy decides to modify the prescription modality from repeatable to not repeatable, meaning it can be used only once.
2012 The Committee concludes that the nimesulide benefit/risks profile is not favourable anymore to chronic use for “symptomatic pain treatment of osteoarthritis” and therefore its use must be exclusively limited to acute pain treatment and primary dysmenorrhea always as a second-line drug, for the minimum possible time.