Aspirin and Reye syndrome

Introduction
The use of aspirin in children has been restricted by the medicines regulatory authorities in Europe. In the UK, for example, it may not be given to children and adolescents under 16 years old except for a specific medical indication such as Kawasaki's disease, preventing clotting after cardiac surgery or reducing the risk of stroke. Some European countries also make an exception for the treatment of mild to moderate pain in children provided they do not have fever or a viral illness.

The reason for limiting the use of aspirin in this way is a longstanding concern about a possible association between aspirin and Reye syndrome, a rare but potentially fatal metabolic disorder. However, the evidence behind this association is far from conclusive and the restrictions on using aspirin have long been controversial. The most recent comprehensive appraisal of the evidence, carried out by Professor Karsten Schrör of the Institut fur Pharmakologie und Klinische Pharmakologie, Heinrich-Heine-Universität in Dusseldorf,1 is summarised here.

What is Reye syndrome?
Reye syndrome is named after an Australian pathologist who, in 1963, described a previously unrecognised condition in children. It was preceded by a period of illness, usually associated with an upper airway infection. The signs and symptoms included rapid breathing and low blood sugar. followed by severe vomiting, stupor, coma and sometimes convulsions. Of the 21 children Reye described, 17 died; post mortems revealed fatty degeneration of the liver and oedema of the brain.

Possible causes
Reye did not know what caused the syndrome and he suspected it was not a single disorder. Today, Reye syndrome is considered a diagnosis of exclusion (i.e. all other possible causes have been excluded) covering a group of disorders possibly caused by infection (influenza or chicken pox are most frequently implicated), metabolic disorders, toxins and chemicals.

What was unknown when many cases were reported, but has since been increasingly recognised, is that a group of inherited disorders of metabolism can cause a Reye-like syndrome. Examples include cellular enzyme deficiencies resulting in faulty beta-oxidation of fatty acids and errors in the urea cycle: it is believed such disorders may account for 10 - 20 percent of cases previously attributed to Reye syndrome. Professor Schrör states that such genetic causes should be considered in all patients in whom the conditions is suspected.

Another possibility is an acquired (rather than inherited) mitochondrial metabolic failure (mitochondria are a cell's 'power plants'). This can be due to the effects of a virus (at least 19 viruses have been implicated), an environmental chemical (such as a pesticide or fungal toxin) or a medicine. It is also possible that some individuals metabolise some medicines abnormally, resulting in the accumulation of metabolites that are toxic to the liver.

It is still uncertain how frequently these many possible causes actually contribute to the development of Reye syndrome. What is clear, however, is that a wide range of disorders can result in metabolic changes, and in liver damage in particular, that are characteristic of Reye's.

Laboratory evidence implicating aspirin
Aspirin was first suspected as a possible cause of Reye syndrome because severe aspirin overdose is associated with fatty degeneration of the liver. However, the similarity is only superficial. Microscopy has shown that the structure and abnormalities of liver cells after severe aspirin overdose are different from those in patients with Reye's. Furthermore, the effects of aspirin overdose are transient and occur only when blood levels of salicylate, its metabolite, are very high; when salicylate levels have been measured in children with Reye's, they have been low. There is also no convincing evidence that a blood-borne toxin accounts for liver failure in patients with Reye syndrome.

One US study, involving 218 children diagnosed with Reye's between 1963 and 1980,2 reported that average salicylate levels were higher in 27 children who died or developed serious neurological complications (mean 150 mcg/mL, range 0 - 460 mcg/mL) than in 103 who recovered fully (mean 100 mcg/mL, range 0 - 480 mcg/mL), and much higher than in 27 other children who did not have Reye syndrome (<20 mcg/mL). The inference that aspirin has a causal role is weakened by a lack of information about the doses of aspirin taken, and not knowing the interval between the dose and taking the blood sample to measure salicylate. The assay used to measure salicylate was not specific to that compound and may have over-estimated its concentration as much as 100-fold. Further more, most other studies don’t suggest a direct relationship between salicylate levels and the severity of Reye syndrome.

Epidemiological evidence implicating aspirin
An established technique for investigating rare conditions such as Reye syndrome is to carry out a case-control study. This identifies a group of people who have the target condition, matches them to a second group (controls) who do not have the condition but are otherwise as similar as possible, and using statistical analysis to compare how often possibly contributory factors occur in each group. Such studies can only demonstrate an association between contributory factors and the target condition, not that the factor caused the condition (given the rarity of Reye's, a causal link may never be confirmed). The reliability of these studies is crucially dependent on adequately defining and diagnosing the target condition and accurately ascertaining exposure to the suspected causes.

Much of the evidence purporting to link aspirin with Reye syndrome comes from case-control studies. The available studies have serious flaws arising from failure to ensure proper diagnosis of patients (they used different diagnostic criteria) and difficulties in accurately measuring the duration of aspirin use and dosage. Given that Reye syndrome is now considered a diagnosis of exclusion, it's difficult to know whether some of the cases identified were actually due to inherited disorders or other conditions. Expert panels have estimated that as many as one-third of cases in the US and three-quarters of those in Canada 'definitely or probably' did not have Reye syndrome; in the UK, it has been suggested that 10 percent of cases were actually due to inborn errors of metabolism.

In the 1980s, the US Public Health Service planned a large study that would attempt to overcome these problems but even that did not avoid the possibility of bias in the results. The data were re-analysed and, comparing only 24 cases of Reye syndrome with 48 matched controls, found that aspirin use was more frequent among cases (88 vs. 17 percent).3 Of the 24 cases, the diagnosis was confirmed by liver biopsy in only 8 and by urine sample in 12 (though it's unclear how many patients gave both samples).

Reye syndrome has also been studied in the UK. Again, there is uncertainty about how closely the diagnosis in the UK corresponds with that in the US: in Britain, there was no peak coinciding with winter 'flu outbreaks and cases tended to be much younger (median age 14 months vs. 9 years). In one 1980s UK study involving 106 cases, 59 percent of patients diagnosed with Reye syndrome had taken aspirin compared with 26 percent of controls.

Estimates of the prevalence of Reye syndrome in continental Europe (sometimes based on very few cases found) range from 0.04 - 0.09/100,000 children, which is similar to current estimates for the US (<0.3 - 1/100,000). By contrast, the available data from Europe and other countries worldwide suggest that approximately 30 percent of cases are associated with aspirin use - much lower than the US data suggest.

It might be assumed that there is a link between the decrease in aspirin consumption following restrictions on its use and a decline in the occurrence of Reye syndrome. In fact, the condition was rarely reported prior to the 1950s when aspirin was probably widely used. The incidence of Reye's has declined at a similar rate in the US, Belgium and France, and Australia despite substantial differences in the use of aspirin in these countries.

Is there a downside to restricting aspirin use?
For children, the alternatives to aspirin are paracetamol (acetaminophen) and ibuprofen. Unlike aspirin, paracetamol causes dose-dependent liver damage and the difference between the therapeutic and toxic doses is relatively small; paracetamol overdose by children is a serious concern.5 Paracetamol has no anti-inflammatory activity and may be less effective in conditions associated with inflammation, such as laryngitis/pharyngitis and otitis/sinusitis. Ibuprofen is a well tolerated alternative to aspirin but is more expensive.

Conclusions
There is a lack of convincing evidence that aspirin causes Reye syndrome: it may be one of many possible factors but many cases currently reported are probably due to inborn errors of metabolism. It is unclear whether restricting aspirin use by children has a favourable risk/benefit ratio.

References
1. Schrör K. Aspirin and Reye syndrome. A review of the evidence. Pediatr Drugs 2007;9:191-200

2. Partin JS, Partin JC, Schubert WK et al. Serum salicylate concentrations in Reye’s disease. Lancet 1982;i:191-4

3. Forsyth BW, Horwitz RI, Acampora D et al. New epidemiologic evidence confirming that bias does not explain the aspirin/Reye’s syndrome association. J Am Med Assoc 1989; 261: 2517-24

4. Hall SM, Plaster PA, Glasgow JF et al. Preadmission antipyretics in Reye’s syndrome. Arch Dis Child 1988;63:857-66

5. Mahadevan SB, McKiernan PJ, Davies P, Kelly DA. Paracetamol induced hepatotoxicity. Arch Dis Child 2006;91:598-603