Karen Starko writes: Several basic questions related to Reye’s syndrome (RS) have come to me from readers of Mark’s book, Inside the Outbreaks. These show the importance of continued education on health issues. (For example, some physicians thought that fever was essential to getting RS). Again, thanks to Mark Pendergrast for a wonderful addition to our public health knowledge.
Is an influenza or chickenpox infection necessary to acquire RS?
The answer is no. RS generally has two phases: the antecedent illness and usually, within a few days of this, the syndrome of vomiting and encephalopathy. While viruses are the most common antecedent event, other illnesses may precede RS such as teething and connective tissue diseases. Some antecedent viruses are influenza A and B, adenovirus, Coxsackie virus, Epstein-Barr virus, ECHO virus, polio, parinfluenza, reovirus, rubella virus, measles, herpes simplex. (Corey, Am J Med, 1975)
Is fever in the antecedent illness necessary to acquire RS?
No. In the U.S. Public Health Service Study (Hurwitz, NEJM, 1985) fever was reported less often by RS cases (54%) than by any of the four control groups (59% to 89%).
Was RS truly more common after influenza B than other viruses?
Currently available data are not sufficient to answer this question. (Any data on this would be welcome). Outbreaks of RS associated with influenza B infection were reported to CDC starting in 1967 when 6 cases were reported from Tampa, Florida. Other subsequent outbreaks included: Oklahoma 1969 (6 cases), Atlanta 1969 (6 cases), Massachusetts and Vermont 1970 (24 cases), and Guam 1971(6 cases). (Hochberg, JAMA, 1975)
The largest study (Corey, Am J Med, 1975) addressing the association was performed between December 15, 1973 and June 30, 1974. In anticipation of substantial influenza B activity after identification of a new variant of influenza B, both CDC through state health departments and the American Academy of Pediatrics urged physicians to report cases. 379 cases fitting the CDC case definition were reported. While it was hoped that laboratory evidence of the type of infection would be obtained from many, attempts at viral isolation or properly spaced serology were conducted in only 21%. However, many cases did cluster temporally and geographically with influenza B. 340 cases reported an upper respiratory infection. 58 (17%) had evidence of influenza B infection. Serology or viral isolation identified influenza A/Port Chalmers (2 cases), adenovirus (1 case), parainfluenza virus (1 case), double virus infections (3 cases), and no evidence for a virus (5 cases). 28 patients had an antecedent varicella infection and 15 patients had a gastrointestinal infection.
Also, more RS cases were reported to CDC during influenza B years as follows: (Hurwitz, Pediatrics, 1982)
Year Major influenza activity Number of RS Cases
1973-74 B 379
1976-77 B 454
1977-78 A 237
1978-79 A 389
1979-80 B 548
Thus, the association with influenza B is a descriptive one. It may be real (influenza B is an independent risk factor), a confounder (influenza B is a reason to give aspirin), or an artifact of surveillance. Reporting RS to CDC was for the most part voluntary. Once the early small outbreaks associated with influenza B were reported and CDC and the AAP called for RS cases during an influenza B year (and found them!) the impression that the illness was associated with influenza B was furthered. Interestingly, after 1990, although the number of RS cases had fallen, no association was noted with influenza B (Belay, NEJM, 1999). Thus the question of whether or not influenza B predisposed to RS is still up in the air. To my knowledge, there are no data addressing the question of whether children with influenza B are at greater risk of RS than children with influenza A or other viral illnesses.
Are influenza and chicken pox independent risk factors for RS or are they simply a trigger for the administration of aspirin?
Unfortunately, this question has not been addressed directly. While temporal and geographic associations with viruses have been identified, these are only descriptive. It remains unclear if these viruses possess some characteristic that promotes the development of RS or whether they are simply the trigger for the administration of aspirin.
On the other hand, there is pharmacology evidence that aspirin alone can accumulate to toxic levels at the doses recommended at the time in children. In 1906 Langmead (Lancet, 1906) reported a child hospitalized with rheumatic fever who received 325mg of a salicylate every 6 hours and on day 4 developed vomiting, fever, dyspnea, cyanosis, and coma and died. He suggested the salicylate might have accumulated. Yet, it wasn’t until 1947 and the development of a test to measure salicylate in the blood that Erganian (J Pediatrics, 1947) showed that indeed salicylate levels progressively increased over several days in some children given commonly recommended doses. In 1965 Levy showed that when the body level of salicylate reaches 360 mg, metabolism changes from first order (when the body can metabolize whatever amount of drug it is given) to zero order (when only a certain amount can be metabolized at a time). In other words, metabolism becomes rate limited and the drug can easily accumulate. Any one of a number of factors will predispose to salicylate accumulation: too much, doses too close together, dehydration (as with fever), reduced renal excretion, low urine pH, and individual variation in speed of metabolism. Other factors such as acidosis and hypoproteinemia increase brain uptake. In addition, individual differences in level at which toxicity occurs have been noted. (Reviewed in Starko, CID, 2009)
Pinsky and colleagues (JAMA, 1988) at the CDC provided evidence that suggests that accumulation happens in RS. They showed that RS cases generally took more salicylate than did controls (although for the most part within recommended amounts) and more importantly that RS cases continued to take salicylate for more days than controls. On day 3 for example, 18 (78%) of 23 RS cases were still taking aspirin whereas only 12 (9%) of 134 controls were doing so. This pattern is consistent with accumulation similar to that reported by Langmead in 1906 and Erganian in 1947.