- Caroline A Sabin, lecturer in epidemiology and medical statisticsa,
- Andrew N Phillips, reader in epidemiology and biostatisticsa,
- Christine A Lee, director haemophilia centreb
- a HIV Research Unit, Department of Public Health, Royal Free Hospital School of Medicine, London NW3 2PF
- b Haemophilia Centre and Haemostasis Unit, Department of Haematology, Royal Free Hospital and School of Medicine, London NW3 2PF
- Correspondence to: Dr Sabin.
In 1991 Duesberg challenged researchers to provide either data on “controlled epidemiologic studies comparing matched hemophiliacs, with and without HIV, or epidemiological evidence that the mortality of hemophiliacs is increased by HIV.”1 We and Darby et al have provided that evidence.2 3 Duesberg's commentary4 requires further comment.
HIV may be enough to cause AIDS
It is incorrect to conclude that HIV is not sufficient to cause AIDS simply because some of the infected patients in our study had not developed AIDS by 10 years after seroconversion. Only longer term follow up studies will finally establish whether all HIV positive patients would, given enough time, ultimately develop AIDS. In the United States around 3800 haemophilic patients have reportedly developed AIDS out of 9000 who have been infected5 (World Federation of Hemophilia, personal communication, 1995), a far higher proportion than that quoted by Duesberg in arguments against the HIV hypothesis.
Duesberg's foreign protein-zidovudine hypothesis predicts that haemophilic patients will not develop non-immunodeficiency diseases such as dementia. Given the low prevalence of some of the 26 different AIDS defining conditions it would not be expected that we should witness all conditions among our 17 patients. However, dementia is well documented in HIV positive haemophilic patients and occurs with a similar prevalence to that in other exposure categories (Xen Santas, Centers for Disease Control, personal communication, 1995). Among all 111 HIV positive haemophilic patients at this hospital, dementia occurred in six.
Duesberg points out that lifetime usage of concentrate may be expected to be different between a 60 year old and a 14 year old (our pair 3). Unfortunately, lifetime usage of concentrate was not available in these patients and therefore usage patterns over 10 years were used. However, it is important to remember that clotting factor concentrate was introduced in our centre in 1978 on average, so that age differences in the pairs may suggest larger differences in lifetime usage than actually existed. Even when the analysis was restricted to pairs in whom the HIV positive patient was younger than or the same age as the HIV negative patient (eight pairs) the results remained similar: four of eight HIV positive patients developed AIDS defining diseases compared with none of the eight HIV negative patients. Furthermore, since 1980 none of 400 HIV negative haemophilic patients registered at this hospital has developed AIDS despite having received clotting factor concentrates on average since 1978, and CD4 counts in these patients have been similar to those of HIV negative heterosexual subjects.6
Contrary to Duesberg's assertion, sexual transmission of AIDS has been observed at our centre. At the Royal Free Hospital sexual transmission of HIV to partners with no other risk factors for HIV has occurred in three cases. Of these infected partners, one developed AIDS and died (the haemophilic partner of this patient also died with AIDS), one was symptomatic with a CD4 count of zero but remained AIDS free, and the third remained asymptomatic but with a CD4 count of 0.2x109/l. No wives of any other haemophilic patients at our centre have developed AIDS.
Patients are given zidovudine because they are ill
It is not true that most British haemophilic patients infected with HIV have been given zidovudine since 1987. Initially patients were given zidovudine after the development of AIDS. Subsequently, since around 1989, patients have been given zidovudine once their CD4 count has fallen below 0.2x109/l or after the development of symptomatic disease. Similar recommendations are made for pentamidine or co-trimoxazole as prophylaxis against Pneumocystis carinii pneumonia. Consequently, by the time patients begin zidovudine and pentamidine they have low CD4 cell counts and are usually symptomatic.
Observational studies often show that patients given zidovudine have a worse prognosis than untreated patients.7 Patients receiving zidovudine are selectively treated because they are ill. The interpretation of findings from these studies should not therefore be that zidovudine increases the risk of AIDS. Of the nine patients developing AIDS in our study, seven received zidovudine only after an initial AIDS diagnosis when immunological deterioration had already occurred. There is no possibility, therefore, that either zidovudine or pentamidine had a causal role in the initial development of symptomatic disease in these patients.
Finally, though there may be some beneficial effect of high purity clotting factor concentrates on the immune systems of patients with haemophilia,8 there is little evidence that this has translated into clinical benefit for these patients.7 Conversely, a recent paper has suggested that increased usage of intermediate purity clotting factor concentrates may be beneficial for HIV positive haemophilic patients.9
Despite the provision of new data which support the HIV hypothesis for the development of AIDS, the arguments proposed by Duesberg in his commentary remain unchanged and contradict the “foreign proteinzidovudine” hypothesis. For the benefit of patients infected with HIV it must now be time to move on to enable researchers to devote time to the real issues at hand.