Vesikari: I wanted to share some data on the viral aetiology of diarrhoea, which are relevant to the question of whether the rotavirus vaccine is needed. We have data from a cohort of 1200 children that we are following to the age of 2 years (Fig. 1 [Vesikari]). 764 episodes of gastroenteritis were detected during this follow-up and were searched for all gastroenteritis viruses by PCR. We scored the episodes using the traditional 20 point score (Pang et al 2000). Approximately 60% of the episodes were mild. In this group, we have a miscellaneous collection ofviruses but most of the episodes remain aetiologically unresolved. On the other hand, we may not really need to resolve them, because these are not significant illnesses; the children will not see a doctor. Also, because the threshold for detection of a case was low, we were probably also detecting other childhood diseases which are associated with either diarrhoea or some vomiting. The picture changes completely when we go to a score of 8 or higher, which is moderately severe to severe gastroenteritis. About 40% of the episodes fell into this category. This is the same score that was used in the US rotavirus vaccine studies as a definition of moderately severe disease. Here rotavirus is responsible for 42% of episodes, and a little bit more where it is associated with co-infection with other viruses. There are two points here. First, human caliciviruses are an important group of aetiological agents in the moderately severe category. These children may or may not see a doctor. Second, only 20% of the episodes in this category remain unresolved.
Finally, I would like to show the really severe episodes, which represent the top 10% of all gastroenteritis cases we saw. These are children who are likely to end up in hospital; if they don't, they should. In this category we can find a virus in virtually all of the cases. Rotavirus is responsible for 72%. We still have some caliciviruses and enteric adenoviruses.
Is there anything that we could do to improve the detection rate still beyond this into the unknown category? I don't think we have to presume totally unknown viruses. The causative agents are probably among those that we already looked for. I suspect that one of the culprits is enteric adenoviruses: we are not detecting them all and we have some evidence that by using a combination of primers we can detect more. There is a hope that we can narrow the aetiological void still further, and that there won't be much room left for novel gastroenteritis viruses as aetiological agents in human gastroenteritis.
Greenberg: Is the attributable risk from the managed care facilities of about 1 in 12 000 excess cases per year after being vaccinated?
Kapikian: Yes, this is the most recent figure that the CDC has calculated from an expanded cohort study of participants enrolled in 10 managed care organizations.
Offit: As a member of the ACIP, I would love to be able to go back to them and convince them to use this vaccine. Here is what I need from this group to help me convince them. Let us say that we use your lowest attributable risk of 1 in 12 274, with a birth cohort of roughly 4 million this would mean potentially at least 320 excess cases of intussusception per year with a universal recommendation. In the USA, roughly 55 000 children are hospitalized with severe gastroenteritis. So we are comparing 55 000 versus 320 hospitalizations, 20—40 deaths versus say 4 deaths. There are some assumptions here that are probably false: we assume that all deaths would be preventable by vaccination, and ignore the fact that now we are aware of intussusception, deaths are less likely to occur from this. The best way to compel people to still give this vaccine — and I think that it is going to have to be given in the USA if it is going to be given in developing countries — is to convince people that it is safe, with safety being defined as the benefits clearly outweighing the risks. The most compelling way to do this is not going to be with these data. Any upper middle class parent is not going to want their child to take this vaccine because of intussusception. I don't think the members of the ACIP are going to give a permissive recommendation if they are not going to give it to their own children. Rotashield was withdrawn from the market in July of 1999. We have a year's worth of information. If you can show that there is a decrease in the rate of intussusception because the child was vaccinated, which I think is a decent formal way of proving that natural infection does cause intussusception, and therefore vaccination prevents it, this would be the most compelling set of data to convince people that the benefit outweighs the risk. I would argue very strongly before June 20th that we should do whatever we can do to make that information available. The way it looks now, this is not going to happen.
Kapikian: The increased risk during the first week after the first dose appears to be significant and deserves further evaluation. Is it possible that the vaccine had a triggering effect in an infant who would have developed intussusception later if they had not been vaccinated? It should also be noted that the package insert notes intussusception as a possible adverse reaction. Is it possible that some of those early cases resulted from recall or reporting bias? For example, a mother calls the paediatrician and says that her child is having abdominal complaints, the paediatrician says that since the child had rotavirus vaccine three days ago, she should bring the child in for further evaluation. An X-ray suggests to the paediatrician that it is intussusception. Another mother calls her paediatrician with the same complaint. Because the paediatrician didn't vaccinate this child, he suggests watchful waiting to see if the condition will improve. These are scenarios we can only speculate about at this time. The only way to answer this issue would be in a double-blind placebo-controlled study, which may no longer be possible.
Offit: If we can get those data before June that is helpful, or if anyone can provide me with some other piece of information that convinces this group that makes decisions about vaccines in our country that the benefits of Rotashield clearly outweigh its risks, I would love to hear it.
Matson: I think that the June meeting is too soon. That date is simply asking people to crush data that need to be cleaned up. The October meeting might be the better time. With regard to the issue about the upper- and middle-class mother who says that her child doesn't die from intussusception, that child also won't die from diarrhoea. The risk factors for mortality occur in a different population than the risk factors for intussusception in the USA.
Glass: One thing that has come from all of these studies, whether they are biased or not, is that there is an excess risk in the week after immunization with the first dose of Rotashield. Regardless of what the attributable risk is, the relationship is present.
Greenberg: There is a spectrum of'informedness' about this issue. Our colleagues from Europe are probably as not yet fully on top of the issue. I simply want to say that as you try to think this through, the USA, like any country, is complicated. There is the FDA that has the legal power to licence and unlicence; there is the ACIP which is an advisory committee that consists of a group of experts in the area of vaccines; and there is the CDC that is doing the epidemiological analysis. Paul asked a question for the ACIP group of experts, what should we do? Everyone listens to what they say. I am an FDA advisor, but I am not on the ACIP committee. My feeling is that you can only deal with the data provided, but the ACIP did not, in its initial discussion, present to the public a process ofhow to weigh the problem in terms of risk and benefit. They simply said that there was a risk and that the risk seemed unacceptable. They didn't discuss in detail the potential benefit also. If we can't figure out the benefit, we can't make an informed decision. The US public may decide that if there is a risk, they do not care about the benefit. This is a public decision. But they have to be presented with the data about the potential benefits as well as the risks, and how one goes about weighing these two factors.
Offit: Your point is excellent, in that we lost an opportunity in the ACIP in discussing issues of risk—benefit. For this or any vaccine, that is the point: do the benefits clearly outweigh any potential risk? In defence of the ACIP, two things happened. One is that the vaccine was withdrawn by the company two days before we met, so to some extent it was a fait accompli. Secondly, the antivaccine groups sit in that meeting and are very vociferous. People are scared to have the discussion that they should have. I agree with you, we should have this vaccine if it is what is best for children in our country.
Desselberger: We saw that various calculations lead to different results in assessing the risks. But you pointed out clearly that the clustering of cases after the first dose of the vaccine is significant and has to be looked at in respect of all the other factors. What has been done further with these 56 intussusception cases in terms of trying to analyse what is going on? There is an epidemiological suspicion of association with the vaccine. About 20% of the cases have been operated on. What do the surgical resection specimens look like? Have they been investigated by ELISA, RT-PCR or immunofluorescence?
Kapikian: This is in progress, and can be answered better by Roger Glass, since the CDC is coordinating this activity.
Glass: The epidemiological investigation is being handled by the National Immunization Program at CDC. This has been one of the largest adverse event investigations they have ever conducted, covering 19 states with lots of analyses. We are in the process of trying to get the pathology tissue from some of those patients who had surgery. This has been very slow, because we have had to go back and get informed consent.
Desselberger: Paul Offit made some interesting comments at the WHO meeting in February 2000 that there might be some link between this event early after the first infection and viral replication. It is certainly worth looking at this.
Offit: I'd like to add to the comments that I made at the WHO meeting. Three to seven days is certainly when one sees replication, but it is not clear that natural infection causes intussusception — it may, but if it does, it doesn't do it at a high and obvious rate. Yet in natural infection of children with human rotavirus, the virus replicates very well. When you look at the simian viruses — the rhesus x human reassortants — these viruses replicate much less well, and clearly cause intussusception at a low but reproducible rate. There was a case of intussusception in the WC3 x human reassortant in Finland. This was in a 7 month old child who got a vaccine. Stools at days 3 and 5 and biopsy material at the time of surgery didn't show any evidence for virus. Initially this was somewhat surprising, but there is a very compelling animal model of transient intussusception in mice that are parenterally inoculated with lipopolysaccharide. What is compelling about this is that one can see this transient intussusception occurring in the absence in inflammation and a clear lead point. It is also an intussusception that can be modified by giving agents that decrease the effects of cytokines. One could argue, then, that there is something about immunization— whether it is with polio virus, simian x human rotavirus reassortants or with bovine x human rotavirus reassortants — where one gives a lot of virus at one time and it is taken up in a site or processed by an antigen-presenting cell (APC) type that is different from that which occurs after natural infection. It is associated with a profile of cytokines that causes this increase in motility and the subsequent intussusception event. One could even argue that there would be a preventive strategy where one could give an anticytokine-specific agent and prevent this. If it is true that any oral bolus of a lot of virus causes intussusception and is prohibited for use as a vaccine, this doesn't bode well for rotavirus vaccines. It is hard to prime the intestine by a non-mucosal route. I hope we can get around this.
Vesikari: I want to comment on the cases of intussusception that we have seen in Finland. The first one was associated with the Rotashield vaccine, and the question in retrospect is could the subsequent events in the USA have been predicted from our experience? I always have an uneasy feeling when I look at the table compiled by Rennels and others, because it puts together cases from very different studies (Rennels et al 1998). In our series we had one case in 1191 children in the vaccine group, and one case in 1207 children in the placebo group. Thus the two populations were very comparable. The one case in the vaccine group had intussusception six days after the third dose, given at the age of 5 months, and the placebo case was 44 days after the second dose. We reported that this was possibly associated, but because there was the other case in the placebo group, the external safety monitor group said it was all right.
Intussusception can happen with other vaccines. We have had one case of intussusception in a child who received a WC3-based reassortant vaccine. It was a 7 month old child, 8 days after the first dose, out of about 1600 who received the vaccine in Finland. We had another case in a UK-based bovine reassortant vaccine recipient, which was a five month old child six days after administration. In these more recent studies, we are giving the vaccine to children at this age, because the rhesus-based rotavirus vaccine cannot be given to such children due to a high febrile reaction rate. We thought that with the use of the bovine vaccine strains we could extend rotavirus vaccination to older children. However, it turns out that this approach may also run into problems, because 6—9 months is the peak age of naturally occurring intussusception. The children are susceptible to developing intussusception, whether this is due to another virus or the rotavirus vaccine — the two can even coincide, and it is difficult to rule out one or the other. The response to this situation has been that anyone who is contemplating a new candidate vaccine has moved down with the age of vaccination. In the future vaccine studies, the first dose is about to be given at the age of six weeks. One reason might be that this age fits with the EPI schedule, but the real reason is to avoid intussusception. Probably the safest way would be to give rotavirus vaccine to neonates. The reason why such studies are not being conducted is because neonatal immunization is not part of the US immunization schedule.
Arias: Are there any data about the rate of intussusception in developing countries?
Kapikian: The incidence of intussusception in developing countries was discussed at the WHO meeting on rotavirus vaccines in Geneva in February 2000. Although the rates seemed to be characteristically lower in developing countries than in developed countries, it appeared that the issue needed to be studied further to exclude the possibility of incomplete reporting of cases.
Glass: We have tried to look at the intussusception rates in developing countries, and we don't have population-based data. We have two experiences which are quite contrasting. One is that in Dehli, where Dr M. K. Bhan at PIIMs looked at the largest hospitals in the city. In the last five years, they have had very few cases of intussusception, so it is rare. In Ho Chi Min City in Vietnam, Dr Ngoun takes care of two cases per day, which is several hundred a year from a huge area. The rates seem to differ greatly. This could all be irrelevant: one of the great discoveries that Al Kapikian's vaccine has made is that the approach of using a live oral vaccine really works well and can protect children from subsequent severe disease. Even with a low rate of intussusception, it could be a life-saver in these other countries.
Kapikian: Unfortunately, a recommendation for evaluation of the rotavirus vaccine in developing countries may turn out to be merely a Pyrrhic victory because the vaccine will not be used in developing countries as long as it is not recommended in the USA. It is a tragedy that every day about 2000 infants and young children die from rotavirus diarrhoea in the developing countries. It will be another 4-7 years before another rotavirus vaccine becomes available for licensure, and there is no guarantee that this vaccine will not have a similar problem. In the meantime, about 4 million children will have died. Wyeth-Lederle, the producers of the rotavirus vaccine have indicated that they would consider staying in the business of making the vaccine if three conditions were met: (1) WHO recommends use of the vaccine; (2) the attributable risk becomes rarefied; and (3) the ACIP would at a minimum grant a permissive licence for its use. If a permissive licence is not granted in the USA, I do not believe that developing countries will use the vaccine.
Pollok: Is there pathological evidence that rotavirus infection is associated with Peyer's patch hypertrophy? This is critical: is it an association or is there a causal link?
Greenberg: There are not enough data. It is anecdotal.
Koopmans: Laying aside the issue of the risk of intussusception, in the USA there may be a consensus that it is useful to vaccinate every child for rotavirus, but this discussion is still being held in Europe. In The Netherlands the issue is not so much one of informing the public: rather, the paediatricians are not all convinced that we need to vaccinate against rotavirus.
Estes: I have one other point I would like to raise about intussusception. One of the things that is most interesting to me about this story is the fact that a virus that is clearly associated with intussusception is respiratory adenovirus, and not the enteric adenovirus. Why does the enteric virus not seem to be associated? Barbara Coulson's papers describe integrins as the rotavirus receptor, and the receptor that is clearly identified for the respiratory adenovirus is an integrin (Coulson et al 1997, Hewish et al 2000). There has been some discussion that perhaps these integrins are found in the intestine in the region of the ileal—caecal junction. I don't know whether there is any connection here, but it is interesting to think about this.
Greenberg: I think the receptor for the respiratory adenovirus is also the receptor for some of the enteroviruses. This receptor is more broadly used in the enteric environment than just for those adenoviruses.
Prasad: It is also the receptor for foot-and-mouth disease virus. Glass: One of the other differences with the adenoviruses is that pathologists tell us that the adenovirus inclusions in the mesenteric lymph nodes are longer and more well defined. This provides a nice leading edge from which intussusception can occur. This is the only one that is well defined besides polyps.
Estes: Intussusception is a very important clinical problem for children. Paediatric surgeons say that this is the most important problem that they deal with. How people resolve it is very different in different countries and even in different cities in the USA. In some cities it is resolved by the radiologists; in other cities, particularly in today's litigious environment, the children always go to surgery. This is one area where we as virologists might consider interacting with the physicians in our institutions to see whether we can help resolve this difficulty in children, regardless of its association or not with the rotavirus vaccine.
Was this article helpful?