Re: Re: Neonatal HIV 30 September 2004
Previous Rapid Response Next Rapid Response Top
Peter J Flegg,
Blackpool, UK FY3 8NR

Send response to journal:
Re: Re: Re: Neonatal HIV

It is refreshing to see Mr Tyler rejoin this debate. Unlike the Perth Group, he raises clear and specific concerns and does not seem to make up his own rules as he goes along. He quite appropriately questions the use and utility of drugs in pregnancy such as zidovudine and nevirapine. Perhaps I can answer some of his queries. Firstly, I must say I agree with many of his points- there is considerable concern about the toxicity of these drugs in pregnancy and infants. Their use is a trade off between their risks and benefits, and if there were safer alternatives that were as effective we would not hesitate to use them. The “current standard of care”, as he puts it, is not ideal, but is still very much better than doing nothing.

Tyler raises the issue of rapid disease progression in infants who had also been exposed to zidovudine in utero. This undeniably can occur, as his citations indicate. However, he fails to explain why there is this correlation, and as he will be the first to accept, correlation does not necessarily imply causation. One of the main explanations given by the study authors is that those infants who progressed rapidly may have been infected with maternal HIV early in utero rather than perinatally. The natural history of HIV in infants infected in this manner differs from that in infants infected perinatally (the period during which zidovudine prophylaxis acts). In other words, this group of infants may have been “self-selected” as rapid progressors as compared to infants infected perinatally, and the use of zidovudine is irrelevant.

In the De Souza study (1) there were two comparative groups, mothers who received zidovudine, and mothers who did not. The percentage of infants who became HIV-infected and then rapidly progressed was similar in both arms – if this was going to happen anyway because these infants were infected early on in-utero, then zidovudine (nor any other drug for that matter) would not have altered anything. The apparent link with zidovudine and rapid progression is only evident because far fewer HIV transmissions occurred in the zidovudine group overall. Infants were far less likely to be infected if mothers took zidovudine, but equally as likely to become infected with a rapidly progressive HIV whether their mothers took zidovudine or not. So it does not induce rapid progression, it just does not prevent it.

Another explanation for more rapid progression in those on zidovudine is given by the authors of the Italian study – this is that transmission in the face of zidovudine is linked with high viral load, and by implication, with zidovudine resistance. Infants with high viral loads do worse, and if an infant has drug-resistant HIV, then zidovudine alone may not prevent infection, and the infant will respond less well to subsequent HIV treatment. The study authors are quite clear in their final conclusion:“Our findings should not be misinterpreted as a reason not to use ZDV prophylaxis, which is effective in preventing perinatal HIV-1 infection.” (2). Tyler is guilty of just this misinterpretation.

Another point that escapes Tyler is that zidovudine obviously does not “increase the likelihood of morbidity and mortality” of infants as he claims. It actually lowers overall morbidity and mortality, because it helps prevent HIV infection in the majority of infants who would have become infected but for the use of the drug.

Tyler’s point about utility of nevirapine in “field conditions” is an appropriate one. Studies have shown both zidovudine and nevirapine containing regimens can significantly reduce perinatal HIV transmission. However, a significant minority of these infants will go on to acquire HIV through breast feeding. Total abstinence from breast feeding is ideal, but in the tropics this can expose infants to other infection risks. Mixed feeding (part breast, part bottle) has been shown to be worse than exclusive breast feeding alone, probably because gut infections in bottle fed infants also predispose to HIV infection via the breast. So much of the benefits induced by preventing perinatal transmission with drugs can be subsequently lost, as the study by Quaghebeur indicated (3). At 6 weeks post partum, HIV infection was at half the rate of historical controls, but by 14-16 weeks had risen to nearly the same prevalence. 56% of the mothers were reported to be breastfeeding, and it is not stated how many of these were exclusively so. Instead of serving as a reason to question the prophylactic benefits of perinatal HIV drugs, this study should prompt workers in the field to find more acceptable ways of limiting the necessity of Mothers in Africa to breast feed by providing safe alternatives.

(1) de Souza R; Gómez-Marín O, Scott G, Guasti S; O'Sullivan M; Oliveira R; Mitchell C.† Effect of Prenatal Zidovudine on Disease Progression in Perinatally HIV-1–Infected Infants. JAIDS 2000; 24(2);154-161.

(2) The Italian Register for HIV Infection in Children. Rapid disease progression in HIV-1 perinatally infected children born to mothers receiving zidovudine monotherapy during pregnancy. AIDS 1999; 13(8); 927-933.

(3) Quaghebeur A; Mutunga L, Mwanyumba F, Mandaliya K, Verhofstede C; Temmerman M. Low efficacy of nevirapine (HIVNET012) in preventing perinatal HIV-1 transmission in a real-life situation. AIDS 2004; 18(13); 1854-1856.

Competing interests: None declared