Di Gibb
The US guidelines for infected infants recommend initiating therapy for all babies diagnosed under twelve months old. The rationale is similar to treating primary infection in adults - but treating babies this young is complicated and not always very successful.
I thought I would just start by showing the estimated drop in the vertical transmission rate that we have seen in the UK shown in Figure 1, which was published in the BMJ at the end of 1999. We have tended to have more elected caesareans here compared to the US and perhaps less triple therapy. However certainly through 1999/2000 many more women were going onto triple therapy at least for the duration of pregnancy.
I am not going to say a lot about follow up of babies born to positive mothers who are not infected, or how you manage indeterminant babies. There are of course issues that have been raised, especially in France about mitochondrial toxicity, which has raised the need for longer-term follow-up, especially as more women are taking more drugs.
Using PCR tests we are now able to make an early diagnosis of whether a child is infected with HIV. In the first day or two of life only about 40% of babies will be PCR positive on these techniques, but as the sensitivity rises rapidly, they can virtually all be diagnosed by three months. We haven't completely sorted out how the use of potent antiretroviral therapy that might happen affects the accuracy of diagnosis.
There have been a lot of discussions recently about when to start treatment, and in adults there is a move to saying we should be starting later, maybe not until CD4 count approaches 200 cells/mm3. In children the issues are more difficult because we don't have such good surrogate markers in kids. Should we start in primary infection soon after birth? If all mothers knew about their HIV status in pregnancy, then we would have fewer babies who were HIV-infected, and we would certainly know about most of them. Should we base the decision on HIV viral load or CD4 count or perhaps their rate of change with age? Using the PENTA, the PACTG, and data from a number of cohorts we are trying to look at how quickly CD4 and viral load these have changed in untreated children with increasing age.
The US paediatric guidelines for starting antiretrovirals have suggested that therapy should be initiated in all infected infants if they are diagnosed less than twelve months of age. To some extent that has been followed in Europe, although with quite a lot of caveats. I think it is definitely not evidence-based but there is an enormous feeling that we have to do something when a baby is found to be HIV-positive, especially if the mother was taking ART. It is always a pretty devastating situation.
Rationale for early treatment is equivalent to primary infection in adults, as we know within a fairly tight window when babies were infected. Also, fro historical data we have evidence that about 20% of children have rapid disease progression and risk progression to AIDS or death in their first year of life. Control of HIV viral load, allowing normal development of the immune system in infants may also be very important because they acquire this virus when they are immunologically immature.
Children have different immune systems from adults; they have a very active functioning thymus, which may be good. They have higher levels and more variation of CD4 cells, which only really decline to adult levels by the age of five or six years. This means that we have more difficulty in predicting what is going to happen to those kids according to their CD4 count alone.
For example, children can have PCP with CD4 counts of 1400-1500 cells/mm3 (which is unheard of in adults). When you put children on HAART, you mostly see an increase in CD4 naive cells, and this again is a different pattern from the response in adults, it shows a more active thymus. They also have a different pattern of RNA decline, by over a log over the time from birth (a rise after birth and then a slow fall to 5 years). Certainly there is a relationship between viral load in babies under one year and risk of progression but the positive predictive value is not good. You can say that a child with a low viral load will do okay but a child with a high viral load may not or may do well. This is not very helpful and is an area in which we need more information.
Figure 2 lists antiretrovirals available at the end of 2000 and highlights the inadequacy of PK data in infants. We have rather a dearth of data for under two year olds for any of the PIs. Ritonavir has been most used in infants, but one of the problems with PK studies is that we have only small numbers eg 10, 12, 20 children across all the age ranges. This means only a few results are relevant for children under two. Getting PK in children under two is very difficult, but is also very important for the PIs and probably the NNRTIs as well.
In Katherine Luzuriaga's PACTG356 study, 15 out of 24 babies under three months of age who started on treatment saw their viral load fall to <50 copies and remain there to 2 years. She then studied these fifteen babies with two babies from a previous study who also had undetectable viral loads. Regimens used included: five children on triple with nevirapine, five on four drugs with nevirapine and abacavir and seven children on four drugs with nevirapine and nelfinavir. Sixteen out of seventeen became seronegative by sixteen months and HIV-1 specific lymphoproliferative responses were not detected, although proliferative responses against other antigens were. This may be worrying, because one would hope to preserve those responses against HIV after primary infection, and this is a different situation from that seen in adults where responses are preserved. This has resulted in lots of discussions about the possibility of giving vaccines to try to induce more HIV-1 responses after viral load has been suppressed to below 50 copies/mL in babies.
We have also done a similar study in Europe - the PENTA 7 study, where we looked at the toxicity, tolerability and activity of early therapy with d4T, ddI and nelfinavir. We also looked at the pharmacokinetics of nelfinavir in very young babies. We started with quite a high dose of nelfinavir - 120 mg/kg/day - which we increased to 150 mg/kg/day when the results from the first four babies showed low trough levels on 120 mg/kg/day (Note that this is five times the equivalent dose that you would give to an adult).
The data are fairly preliminary but the first thing we found was that nelfinavir powder was poorly tolerated and crushing up the tablets was much better. However even increasing the dose to an average of 150 mg/kg/day, we still observed inadequate trough levels. Also we have had quite disappointing results of activity. Of the twelve children out to week 24, only 50% have seen their viral load reduce to below detection to less than 400 copies/mL and only a quarter are below 50 copies/mL.
It is worrying that we were not more successful, especially given the data presented earlier (see Figure 3, and pages 12-13). Giving four drugs to five babies who were all quite sick with much lower CD4 count seems to produce a much better response, with four out of five of them achieving viral load levels below 50 copies/mL. Of course you don't want to wait until a baby gets an opportunistic infection with PCP or CMV before you start giving treatment, as it can happen very quickly with some babies. However it may be that adherence is much better in babies that have had symptoms.
What about clearing infection? That is one reason why people started treating primary infection both in adults and in children. The theory is that if you stop treatment after starting early, you could reset the viral set point? These are the reasons for thinking about starting, but what about the disadvantages? Some children may not need treatment for many years, and there are concerns about toxicity, poor tolerability, adherence and whether you are getting the right dose. It is easier to be surer that you are going to get the right dose when the child is a bit older, particularly for PIs and possibility NNRTIs.
Regimens need to be chosen based on tolerability, PK and support for adherence - and this is vital. Management for toxicity means planning for what happens if side effects become apparent and what do you do if the management of side effects is unsuccessful? You need to plan ahead. If you succeed virologically, how long are you going to keep children on the same treatment? Forever? If you decide to defer treatment and monitor, is your decision to start based on a rate of change in CD4 a percentage as well as absolute values?
These questions need to be addressed but we only have results from smaller studies and no randomised studies. With such small numbers, we all need to work together to pool our data.
If a child is just not responding and you decide to switch to another regime or even to stop - then having a blood sample before stopping for HIV resistance testing at the time could be important. Resistance in children is a critical issue that we need to become acutely aware of especially if you start with all 3 classes, which I'm not in favour of. The data from the Palella study in adults attending US clinics showed the average time on HAART regimens to be only twelve months for the first, eight months for the second, seven months for the third treatment regimen. By the time you got to your third regimen you are usually on four or five drugs after being on three or four for your first.
Mike Sharland reported similar sorts of data for children in Durban. Hopefully some of our children are going to live for twenty or thirty years and they could run out of all classes of drugs in two years. We know from these data that you are half as likely to respond to your second regimen as your first regimen.
Figure 1. Vertical transmission rates
in UK |
Fig 2. ARV therapies available in
Oct 2000 |
Figure 3. Studies with early treatment |
 |
 |
 |
Click on slide for a larger view |
Click on slide for a larger view |
Click on slide for a larger view |
Dr Diana Gibb currently works on paediatric trials at the newly formed MRC Clinical Trials Unit (CTU) in London, and also does clinical work at Great Ormond Street Children's Hospital, where she set up the paediatric HIV service. She was instrumental in setting up the PENTA network. She established the first family HIV clinic in the UK at GOS and a similar clinic at Newham General Hospital in East London. She is author to many publications on HIV-infection in children and has lectured widely nationally and internationally on paediatric HIV infection.
 |
 |
 |
 |