Dr. Adhikari referred to the report by Wairagkar and colleagues on the isolation of measles virus from children with acute encephalopathy(2). He has rightly pointed out the need for epidemiological studies and for follow-up of surviving children to document the development of IgG antibodies against measles virus, so that additional evidence could be gathered(1). Although Wairagkar, et al. had stated that "geographic clustering was not observed"(2) in reality there was clustering in time and space, as may be seen in the earlier and more complete report on the ‘epidemic’ itself, by Ghosh and colleagues(5). There were 129 cases, occurring in October and November of 1997, in the rural or suburban areas of 10 contiguous districts in Haryana, Uttar Pradesh, Punjab and Chandigar(5). It was a short and sharp epidemic with 92 cases in October and 37 in November with none before or after. The attack rate was 54 cases per 1000 children below 15 years(5). In at least 7 families more than one child was affected(2,5). Of the 129 cases, only the 51 admitted in the Postgraduate Institute of Medical Education and Research, Chandi-garh, constituted the subjects of the Wairagkar study(2). Therefore, the lack of geographic clustering was an artifact of hospital-based selection of cases during an outbreak.

Of the 51 children, the liver tissues of 19 had shown microvesicular steatosis and brain tissue in 18 of them showed edema without any encephalitis or meningitis, both of which are classical features of Reye’s syndrome(5). ln summary, there was an epidemic of Reye’s syndrome. Of the same 51 hospitalised cases, 21 were subjected to virological studies. Measles virus was isolated from the cerebrospinal fluid (CSF) of 4 children and the sera of two children(2). Thus, the conclusion of the Wairagkar study was that measles virus caused the outbreak of Reye’s syndrome. However, none of the children had contact with clinical measles and there was no measles in the community during October and November(2,5). The conclusion in the report by Ghosh et al was that the finding of measles and varicella-zoster virus infections in some cases "may indicate subclinical infection as the prodromal illness for development of Reye’s syndrome"(5). This was one inter-pretation of the unexpected laboratory finding of measles and varicella-zoster infections. However, subclinical measles and chickenpox are uncommon in the age group under study; there was no source of infection. Therefore it could have been a laboratory artifact in finding IgM antibody by enzyme-linked immunoassay, which is more common than outbreaks of subclinical measles and chickenpox.

Among the 21 children who were investigated for measles virus infection, 6 yielded virus and IgM antibody was found in the serum or CSF in 8 more children, thus adding up to 14 children with evidence of measles without rash(2). The first report mentioned that 12 of 33 sera tested were positive for varicella-zoster antibody (presumably IgM), for which reason this agent was also implicated in subclinical ‘prodromal illness’ triggering Reye’s syndrome(2,5). Moreover, although detailed clinical, laboratory and histopathology data were available on 51 hospitalised children with Reye’s syndrome(5), no correlation was presented between measles virus isolation and IgM antibody to varicella-zoster virus, and also to the clinical or laboratory features of the children(2,5). In other words, it was not clarified if any of the 6 children with virologically diagnosed measles virus infection or the additional 8 with measles IgM antibodies overlapped with the 12 children with varicella-zoster antibodies or if any of them were among the children who had laboratory confirmation of Reye’s syndrome.

In support of their finding of measles virus in the CSF of children with Reye’s syndrome, Wairagkar et al. cited their earlier report of the isolation of measles virus from the CSF of patients without features of encephalitis(3). The report was on the detection of measles virus in the CSF of 3 adults with acute renal failure and neurological deficits including cranial nerve palsies and peripheral nerve axonal pathology(3). These patients had no rash and no contact history with measles cases and they fell ill at a time when there was no measles in the community(3). Thus, the second clinical syndrome attributed to measles virus infection was very different from the first.

The third clinical syndrome was the Siliguri outbreak, which has not been described in detail. It involved adults, including doctors and nurses, and more than 20 persons died. Once again, there was no contact with measles and there was no measles in the community at the time(1,4).

The detection of measles virus from body fluids is almost incontrovertible evidence of causative association, no doubt. However, we know much about Reye’s syndrome and measles virus and such knowledge must also find a place in our deliberations in seeking an etiological role for the virus. Measles virus is known to cause central nervous system (CNS) disease; it causes three specific CNS syndromes, such as acute measles encephalitis (AME), subacute sclerosing panencephalitis (SSPE), and subacute measles encephalitis (SME).

The AME follows clinical measles close on its heels, and is often a demyelinating ‘post-infectious’ complication. The SSPE is a late sequel to measles in early childhood. The SME occurs in children or adults with deficient cell mediated immunity (CMI) and the disease is severe, progressive and frequently fatal, and without rash and the typical prodrome of measles. The rash in measles requires normal immune response. In SME the measles virus behaves like an ‘opportunistic’ invader of the brain(6). In a review of 33 reported cases of SME, 10 had no history of earlier clinical measles(7). When there was a history of earlier clinical measles, the interval from it to SME was one to seven months, showing that in some cases the virus had persisted after clinical measles. The predisposing diseases causing immune deficiency were acute lymphocytic leukemia (23 cases), neuroblastoma(3), lympho-sarcoma(2), rhabdomyosarcoma, ganglio-neuroma, Whipple’s disease, renal trans-plantation and human immunodeficiency virus infection (one case each)(7). The age range of subjects were 2 to 21 years. Although histology, electron microscopy or polymerase chain reaction (PCR) of the brain tissue did confirm measles virus etiology the virus was not easily isolated from the CSF or brain tissue(7). When detected the virus had defective mutational changes(7), rather reminiscent of the persisting virus in SSPE. Thus the differences between SME and the reported cases of acute renal disease and CNS deficits(3) include an older age spectrum in the latter (37-43 years), the absence of predisposing disease with defective CMI, the ease of isolating measles virus from the CSF and the occurrence of acute renal failure. Some of these differences apply also to the Siliguri outbreak disease.

Reye’s syndrome without any brain or meningeal inflammation does not resemble AME, SSPE or SME, all of which have distinct histopathology. Did the four children with measles virus in their CSF have brain tissue studies and did they show a lack of inflammation? If they did not have inflammation in the CNS, but had only brain edema, the detection of virus in CSF must be viewed with some respectful skepticism. That the authors were silent on such an obvious and crucial piece of information was indeed very unfortunate. While measles virus infection without inflammation but with only edema is not impossible to happen, it is very unlikely to occur according to our present state of knowledge. The onus of proof of etiological association rests on those who put forward such a proposition. Such credible evidence is lacking in the report.

The incubation period of measles is about two weeks, for which reason measles antibody, even IgG, is usually present even on the day the rash appears. The rash is a manifestation of the dissemination of virus in the body and immune-mediated reaction in the skin. In other words, it is the result of, and nearly concomitant with, the early rise in immune response following secondary viraemia. If the encephalopathy occurred even before the development of brain inflammation, that is, before virus invasion in the tissue, then it would have been prior to the secondary viraemia. This surmise is supported by the absence of IgG antibody in a few children with virus in the serum or CSF. In that case the presence of virus coincided more with primary viraemia than with secondary viraemia. Usually, the primary viraemia consists of very small amounts of virus and is not usually detectable.

Acute disease due to measles virus infection occurs only in immunologically naive children; therefore it would have been crucial to document the past history of whether or not these children had measles. If we assume that the infection was the first episode, where did the virus come from? There was no measles virus circulation at that point in time and place. This would suggest that the disease was due to reactivation of latent virus infection, rather like in SSPE, but then how do we reconcile an epidemic of reactivation? SSPE is always sporadic. Were all these children having persistent measles virus infection in their brain or other tissues, and were going to develop SSPE if they had not developed Reye’s syndrome? Do the surviving children continue to have latent infection? These questions have to be addressed if we accept the subclinical measles etiology of an epidemic of Reye’s syndrome. If the disease were indeed due to reactivation, there would have been significant pathology as well as IgG antibody. Both these elements were missing. Thus, more evidence and more information are essential to interpret the finding of measles virus in unexpected diseases, without any epidemiological links of transmission.

All of the above discussions point to a direction opposite to the proposed etiological association of a subclinical measles epidemic without evidence of exogenous source of infection, causing an outbreak of Reye’s syndrome. Yet, the finding of measles virus in CSF or serum is clear evidence in support of measles etiology. More evidence for causation and more information on the reported diseases and outbreaks are obviously urgently needed in order to accept the ‘discovery’ of new forms of measles disease or to reject the proposed aetiology by measles virus.

It would appear that the adult renal disease with CNS involvement (claimed to be caused by measles virus) and the Siliguri disease had some common features(3,4). Both were in adults, who were presumably immune to measles virus, based on the known epidemiology of measles in our country. We know from many studies that the vast majority, if not all, adults have had measles as children and are measles antibody positive. Moreover, the median age of measles was between 2 and 3 years in the pre- vaccine period. Once again the question where the virus came from is relevant in both situations. If it was first infection, which was most unlikely in adults, there was no known source. If it was endogenous latent infection being reactivated, how did cases occur in clusters?

Very reluctantly and with trepidation as well as due respect to my virological colleagues who have reported the new forms of measles, I must raise the possibility of some inadvertent problem somewhere, in the virus isolation results, essentially for the sake of ‘science’. The purpose here is only to raise this question, but let me clarify that I have no evidence, except the arguments presented above, to refute the finding of measles virus causing these outbreaks.

First, all virus isolation reports have come from a single laboratory. In addition to the data discussed above, measles virus was isolated from the sera of children in two other outbreaks of ‘encephalitis without rash’, one from southern and the other from western India(2). Thus, this one laboratory has been successful in isolating measles virus from five outbreaks, whereas no other laboratory in the world has reported such a finding in outbreaks of Reye’s syndrome or in adult CNS disease with or without renal disease. Obviously, independent confirmation from another center is necessary before we can accept the evidence as proven.

There is no need to question the identity of measles virus, which has been properly clinched, at least in the report of the Reye’s syndrome study(2). However, the isolation itself needs to be further scrutinized. The authors stated, "cytopathic effect (CPE) was observed on day 3 post infection (meaning post inoculation)" in cultures with four CSF and 2 serum samples of 6 different individuals(2). This is extraordinary. Usually in the first isolation of measles virus CPE develops slowly, over 7 or more days. Only in later passages with heavy virus inoculum, especially with cell culture adaptation of the isolated virus, we see uniformly rapid development of CPE, as reported in the study(2). Although the report clearly stated that all usual precautions had been taken to avoid laboratory contamination, the above finding does suggest the possibility of a laboratory-adapted virus in large inoculum somehow getting into the laboratory system.

Since the epidemiology and pathology of the diseases cannot be reconciled with measles virus invasion of tissues, hence the proposed etiology, and since more than one clinical syndrome is claimed to be caused by one virus in one laboratory, the possibility of some laboratory problem must be kept in mind before accepting the claim of the investigators. This is by no means a slur on the investigators, but to point out the inescapable onus on their part to give convincing evidence by way of checking for antibody response of surviving children to the isolated virus; genotyping the isolates to see if they resemble vaccine measles viruses which are readily available everywhere; checking the access of the laboratory personnel to any measles virus; virus titration of original specimens of CSF and sera; sharing specimens and cell cultures with other laboratories and re-inoculation of the same specimens by independent personnel.

Until another laboratory confirms that measles virus causes outbreaks Reye’s syndrome and adult neurological diseases without rash, the proposed etiological association must be viewed with scientific interest, but also with caution. It is too premature to accept this association to have been convincingly established.

T. Jacob John,
439, Civil Supplies Godown Lane,
Kamalakshipuram,
Vellore, TN 632 002, India.
E-mail: [email protected]