There are many examples where a clever observation of an association between two diseases has led to an increased understanding of the pathological mechanism underlying both disorders. Just occasionally, the opposite happens. A chance finding, a rare complication in a clinical trial, leads to the scrapping of a promising line of advance. This happened when a few cases of intussusception occurred in infants who were offered vaccination against rotavirus infection, a common and serious cause of diarrhoea in infancy in many countries. Even though only one extra case of intussusception occurred for over 11 000 children vaccinated, this would represent a lot of children if rotavirus vaccination were to be provided to hundreds of millions of children worldwide.
This situation can cause great frustration. The clinical scientists, probably from academia, have looked forward to ‘making a difference’ and find their hopes dashed by a result that seems to be trivial at best and minute in effect, in comparison with the big picture of thousands of very sick children from rotavirus. The pharmaceutical firm (one of the big pharmas, in all probability, since vaccine development is expensive) sees its investment wiped out in a flash. Vaccines are not that profitable in any case (compared to yet another, more or less equivalent, statin, selective serotonin reuptake inhibitor, baldness treatment or viagra − we must try to see things from their point of view). A serious side effect, even in only a small number of individuals, will sink a new product in the highly risk-adverse climate of our decade. So the pharma drops the product.
The medical researcher (who reluctantly relinquished control of the project because they needed the experience of the pharma to take things to market) is now angry as the project is worse off than before. The company will release some information about the problems, but side effects never make good press, even during a stage 3 trial, and shareholders are not impressed by failures. Other pharmas are unlikely to ‘start again’ now they know there is a problem.
The researchers still care passionately about the children who are ill, perhaps dying, from rotavirus, and they hunt for a rational explanation so that vaccine development can proceed. What could explain the rare case of intussusception?
The study reported by Barnes et al.1 in this issue of the journal began with an interesting hypothesis. A proportion of infants with cystic fibrosis (CF) are born with meconium ileus, a ‘sticky intestinal mucus’, which might be thought to predispose to intussusception. Perhaps not only the rare infant with CF, but carriers of CF as well (roughly one in 25 of the Caucasian population in Australia) have a predisposition to intussusception when challenged with rotavirus vaccine. It would have been attractive were the result positive, since CF carriers are common in countries such as Australia, UK and the US (where rotavirus infection is common but rarely severe), but rare in South-East Asia and Africa, where the problem is worse and the need is greatest.
The research team obtained small discs cut from the blood spots (‘Guthrie cards’) which were collected at birth to test for phenylketonuria (PKU) of about 80 infants who later developed intussusception, and a similar spot from the adjacent card from an infant to act as a control. They found no difference in the frequency of CF carriers between the affected and the controls. Being a CF carrier does not result in an increased risk of intussusception, at least in this study of Australian infants. Like any negative study, this report rules out one possible confounder and not all causes. Also, the test infants were those who presented clinically to the gastrointestinal unit, and not those who developed intussusception after vaccination. Nonetheless, it seems unlikely that screening out infants who carry CF will reduce the risk, however small, of developing intussusception when given this vaccine.
In the course of the study, another issue arose, that of conflict between privacy in the use of samples and a need for accurate research using existing specimens. Every infant in Australia is tested for PKU, hypothyroidism and CF at birth, by measuring levels of blood analytes, such as the high levels of byproducts from phenylalanine metabolism that only occur in those with the disease. Those with PKU, in particular, benefit as this serious disease is prevented by strict removal of phenylalanine from the diet. This is an essential public health screening program.
The blood spots (‘Guthrie cards’) are retained for quality control; Canberra suggests they be kept for 30 years. This was uncontroversial until it became clear that these cards also could be a valuable resource for research, and also used for DNA fingerprinting by the police. The cards are kept under lock and key, and in Victoria there is a formal agreement between the police and the Genetic Health Service that they will only be given to a judge on the basis of a court order. The cards are sometimes requested by the Coroner to identify human remains where there is doubt, but rarely in the context of a criminal case.
In the study by Barnes et al.1, the samples were completely anonymized to ensure privacy, a strategy encouraged by the Human Research Ethics Committee. This now leaves the researchers in what I believe to be a strange (and unethical) position. Five of the 160 or so babies tested were shown by the DNA test to be CF carriers, but we don’t know who they are because all identifying data has been removed! Three infants came from the group of children who developed intussusception. Each of those children has parents who are at 16-fold increased risk of having a CF child next pregnancy (approximately 1/100 as against 1/1600), but we cannot warn them thanks to the decision of the Ethics Committee. What will happen if one of them has a CF child next time around? Will they be able to associate the members of the Ethics Committee with the lawsuit if they find out the facts, when they point out that an obsession with ‘privacy’ has taken precedence over a duty to warn of a real risk to health?
We must ensure that concerns over theoretical or philosophical risks to privacy do not take precedence over the very real problems posed by the risk of having a seriously ill child with rotavirus diarrhoea or CF. Let us also hope that Barnes and his colleagues find a way to implement their program of offering rotavirus vaccination to those who need it, in a safe and effective way.