AIDS Treatment Update, Issue 56, August/September 1997
Edward King

Both sets of guidelines endorse an approach to anti-HIV therapy which is substantially more aggressive than that currently employed at most British clinics. To check whether British doctors share the views of their American colleagues, AIDS Treatment Update singled out three key elements of the new recommendations and sought responses from members of our medical advisory panel. The three topics were:

* should treatment be recommended for everyone with a viral load above the 5,000 to 10,000 range, regardless of their CD4 count?

* should everyone who starts treatment receive a triple combination that includes one of the three most highly active protease inhibitors - indinavir, ritonavir or nelfinavir?

* should anyone whose viral load remains above the limit of detection during anti-HIV therapy consider switching to a new or more aggressive regimen?

Starting treatment

The US guidelines make it clear that ongoing HIV replication is the driving force behind HIV disease progression. The recommendation to start treatment even at relatively low viral load levels is based on studies among groups of people with HIV who are not currently on treatment, which show that an individual's viral load strongly predicts the subsequent risk of disease progression (see table overleaf).

As Professor Tony Pinching of St Bartholomew's Hospital pointed out, this provides a tool for distinguishing between people whose risk of disease progression in the future may be quite different, even though their CD4 cell count at the moment may be identical. "Virus load testing is a method of identifying the patients who start off with a normal CD4 count but have a worse prognosis. The PACT Framework (described in AIDS Treatment Update issue 53) already covers this by suggesting that treatment should be considered for anyone with a virus load above 10,000 so we're not much at variance."

Professor Jonathan Weber of St Mary's Hospital agreed that "every single study has shown that viral load is an extraordinary sensitive prognostic indicator. I don't think you can ignore that."

But while British clinicians agree that baseline viral load is one of the most important tools in deciding whether to start treatment, they sounded a note of caution over the level of viral load that should be viewed as grounds for treatment. "Very few of my patients who aren't currently taking any treatments have a viral load less than 10,000 copies", said Dr Margaret Johnson of the Royal Free Hospital. "From the natural history of HIV infection we know that some of them will have a very good prognosis without treatment, and I don't think it would be the right advice to recommend them all to start therapy immediately. Even 20,000 is a low viral load, and I think you have to balance having therapy now against the potential advantage of waiting until there's more information about how to use the drugs most effectively. I'd use a viral load cut-off of 50,000 for firmly recommending therapy, rather than any lower."

"We only have definite evidence of short-term benefits of current antiretroviral therapies because no studies have last more than three years", added Dr Ian Williams of the Mortimer Market Centre. "It therefore seems reasonable for individuals to start therapy when their risk of clinical disease progression in the short term is at a level which is unacceptably high for them. For most people this would be between CD4 count of 200-350 and influenced within this range by their viral load. Individuals with very high viral load may wish to consider starting earlier, since their short-term risk of progression is higher."

Professor Weber noted that "I think we can debate where the cut-off should be. By and large even people with very low viral load will progress - they are all really slow progressors, rather than non-progressors. But our current practice is to treat people with viral load above 20,000, since at that level the great majority of people you treat are at significant risk of progression. Over time I wouldn't be surprised if everyone with detectable viral load ended up being treated."

First-line protease inhibitors

British doctors were much more doubtful about the US advice that the best option for people starting treatment was a triple combination including a protease inhibitor.

This does not mean that they still necessarily favoured treatment with only two drugs. "Starting with three drugs definitely seems better than two," said Professor Weber. Dr Johnson agreed that "Most of my patients are now starting on triple therapy. It's been suggested that people with very low viral load may only need two drugs to get a good viral load response, but I think those patients may be better advised simply to wait rather than to start with only two drugs."

Professor Pinching noted that "Patients who opt for treatment generally want to achieve as low a viral load as possible and minimise their chances of resistance, and choosing triple therapy increases the chance of them succeeding in that objective. That logic leaves very few people opting for dual therapy first up. However, for people who are already on dual regimens that are working, I certainly wouldn't recommend them to add something else just because that's what most people are currently starting with."

Still, Dr Williams pointed out that for some people there are very valid reasons for choosing to start with only two nucleoside analogues. "Some patients accept that starting with dual therapy may be sub-optimal in terms of its effects on viral load (although it will offer some clinical benefit in the short-term), but may have the advantages of being easier to take than triple therapy, and unlikely to strongly prejudice your response to other more potent regimens containing protease inhibitors or NNRTIs in the future."

Dr Martin Fisher in Brighton agreed that "I have a substantial number of patients who are keen to start with only two drugs because of tolerability, and a good proportion have achieved undetectable viral load on d4T/ddI."

Moving to triple therapy for everyone also has significant cost implications, noted Dr Brettle of City Hospital in Edinburgh. "Our treatment budget would need to more than double if we were to start treating everyone with viral load greater than 10,000 with triple therapy. Our own recent experience suggests that we would end up having to find the money from within the current AIDS budget and slash other services."

For those who do opt for triple therapy, the debate centres around the choice of third drug. Several clinicians made the case for choosing a non-nucleoside reverse transcriptase inhibitor (NNRTI) such as nevirapine, rather than a protease inhibitor (PI). "There are no data that suggest that starting treatment with three drugs that include a PI is better than three drugs that include an NNRTI," said Professor Weber.

Professor Pinching noted that "choosing an NNRTI rather than a PI might be more common if treatment decisions weren't so driven by fashion, because it lets you keep the PIs up your sleeve until we know better how to use them". Dr Brettle agreed, noting "I'm a pessimist and think resistance to the first regimen is likely, so it is important to keep a follow-up option". Dr Brian Gazzard of the Chelsea & Westminster Hospital pointed out that it is relatively easy for resistance to NNRTIs to emerge, making it "terribly important that such drugs are only used in completely suppressive regimens".

"I wouldn't be so emphatic as the Americans in suggesting a PI rather than an NNRTI when starting treatment," summarised Dr Johnson. "But for someone who already has advanced disease when starting therapy, I'd be more keen to suggest using a PI-based combination."

The saquinavir debate

The American guidelines advise using indinavir, ritonavir or nelfinavir - and specifically not saquinavir - as part of a triple combination when starting treatment, arguing that saquinavir is the least potent PI in terms of suppressing viral load, and is poorly absorbed in its current formulation. "At first look that stood out as a very strange recommendation", said Professor Pinching. "If you look at clinical endpoint studies, all the PIs in three-drug combinations show similar reductions in disease progression; they're within a whisker of each other. The exception really is nelfinavir, for which we don't yet have clinical endpoint trials and we don't have enough feel for the risk of cross-resistance.

The guidelines seem to me to be weighted too much towards the ability of different combinations to reduce viral load. Their analysis doesn't take account of the possible benefit from starting with saquinavir, in that it still looks less likely to produce cross-resistance to other PIs than any of the other drugs; so it may be the best choice to keep your options open. I'd be happier for all the PIs to be put in as equivalent, albeit for different reasons, rather than excluding the agent that's been around for longest and about which we know most."

Other clinicians shared the US concerns. "My view is that if you are going to use a PI as part of an initial regimen you probably want to use one that's as effective as possible at suppressing viral load", said Dr Johnson. "I don't use saquinavir first-line, although I have a lot of patients taking saquinavir with ritonavir as a back-up regimen."

"I was surprised to see first-line saquinavir excluded so bluntly", said Professor Weber, "but I must confess that I'm not using it first-line at the moment except in combination with ritonavir." He drew attention to research released at the International Resistance Workshop in Florida in late June, in which several different researchers presented results suggesting that many (but not all) people with prior saquinavir experience may not have good or lasting viral load responses when they subsequently switch to one of the other PIs or add ritonavir, even if they show no signs of saquinavir resistance.

Dr Fisher shared the concerns about the current formulation of saquinavir, but pointed out that the new `soft gel' formulation has shown promising early results in trials. However, it is not expected to be available in the UK for many months, and people who have already developed resistance to the current form of saquinavir will not benefit.

"The main message is to use the PIs appropriately, where there is a high likelihood of getting viral load to below detectable levels and sustaining a response", concluded Dr Williams. "The data emerging on cross-resistance suggest that if you fail on any of the currently licensed PIs, including saquinavir, you substantially reduce your response to the others."

Aiming for 'undetectable'

Whatever regimen one chooses, the US guidelines suggest that the primary aim of treatment should be to reduce viral load below the limit of detection. If viral load remains at measurable levels despite six months of therapy, or rises persistently during therapy, they recommend switching to a new or more aggressive regimen.

"There are valid reasons why some patients do not adopt `undetectable' viral load as a main treatment aim, such as the difficulties of compliance and side-effects with triple therapy", said Dr Williams. "If you start with two nucleosides you're accepting that you are unlikely to suppress your viral load below detection, but even partial suppression (particularly to below 10,000) will result in clinical benefit compared with no suppression."

"If you can do it, the theoretical argument for suppressing the virus below detection is clearly sound on the basis of the evidence we have to date", said Professor Pinching. "But what doesn't automatically follow is that undetectable virus load is the only valid objective, and that you should keep changing drugs if you fail to achieve it. That attitude is already causing considerable distress, to the extent of requiring serious psychological support, for patients who don't achieve this over-exalted goal. I see many patients whom we can't reduce to `undetectable' but whose response in terms of clinical health and CD4 count is clearly very good."

"Undetectable viral load simply isn't achievable for many patients", agreed Dr Johnson. "If I'm prescribing initial therapy with a PI I want to see a 1.5 to 2 log reduction, but I have a problem with switching people just because their virus remains detectable."

Professor Weber pointed out that "pushing viral replication below the threshold of detection is not the same as stopping replication altogether. Most patients with undetectable viral load probably still have a low level of ongoing replication; it's just too low for our current tests to be able measure it. So the question is whether dropping viral load to a low but detectable level such as 1200, rather than 200, actually makes any difference?".

"I agree that the lower the viral load is, the better," said Dr Fisher. "But undetectable viral load shouldn't be set up as the holy grail, and patients shouldn't be left feeling that they've failed if they haven't achieved it."

Conclusion

In summary, British clinicians are moving in a similar direction to their American colleagues on issues such as using viral load thresholds to guide decisions on starting treatment, a steady shift from dual to triple therapy regimens, and the theoretical desirability of undetectable viral load.

But their opinions differ on details such as precisely which viral threshold to use; which triple regimen to prefer (and the fact that dual therapy may remain a preferable option for some individuals); and the achievability of undetectable viral load.

Table 1. Percentage of people who develop AIDS within 3 years (assuming

no treatment)
Viral Load CD4<200 201-350 351-500 201-750 750
Below 3000 ** ** ** 3.7 0
3001-14,000 ** ** 2.0 2.0 2.0
14,001-41,000 ** 8.1 8.1 8.1 3.2
41,001-110,000 40.1 40.1 16.1 16.1 9.5
Above 110,000 85.5 64.4 42.9 32.6 32.6

"One very particular perspective"

Despite their disagreements with specific aspects of the guidelines, the clinicians we spoke to did not dispute that the US recommendations are internally consistent. They flow logically from a perspective that places the greatest emphasis on viral load as the engine that drives the disease process: even low levels of viral replication are deemed worthy of treatment; the best regimens are deemed to be those that have the greatest effect on viral load; and the aim of therapy is deemed to be the maximal and continual suppression of viral load.

However, this perspective was not necessarily shared wholeheartedly by our respondents. "It's foolish to base all our treatment decisions around viral load", said Dr Johnson. "I think you have to look at all the parameters - clinical state, CD4 count, viral load, the rate of change in all of those things. Every patient is different."

Professor Pinching noted that "the guidelines come from one very particular perspective on a limited set of data. They represent a certain spin on the data with certain prior assumptions, but not everyone will share those assumptions. Perspectives are continuing to evolve, and additional data will continue to come through that may change specific recommendations.

"People shouldn't feel that what comes out of North America is necessarily right. It's not very long ago that they told us all that anyone with a CD4 count below 500 should be treated with AZT monotherapy, and it turned out that they were wrong! The field is in a very interventionist mode at the moment, but the fact that these guidelines are more interventionist than others should not be assumed to mean that they are automatically correct. They may be, but history doesn't insist on it."

970801
ATU05601

Copyright © 1997 - AIDS Treatment Update. Permission granted for noncommercial reproduction, provided that our address and phone number are included if more than short quotations are used. Subscription lists are kept confidential. NAM Publications 16a Clapham Common Southside, London, England SW4 7AB; TEL: 01-71-627-3200 (from outside the UK: +44-171-627-3200); FAX: 01-71-627-3101 (from outside the UK: +44=171-627-3101)  info@nam.org.uk  http://www.nam.org.uk