AIDS Treatment Update, Issue 56/57, August/September 1997
Interviews by Edward King

Concorde found that there was no advantage to starting early treatment with AZT monotherapy, as opposed to waiting until symptoms or other signs of advancing immune deterioration developed. Some clinicians argued that Concorde proved that treatment needed to be more intensive than AZT monotherapy; others reasoned that the study demonstrated that the available antiviral drugs were of extremely limited value and should thus be reserved for use in symptomatic disease, when their benefits were most obvious and most needed. Individual doctors and even whole clinics acquired reputations as being more or less `aggressive' in their approach to antiviral treatment.

How have views changed?

There have been many developments in AIDS research since then - notably, the availability of more potent drugs, the effectiveness of combination therapy, and the use of viral load tests. The publication of various sets of treatment guidelines over the last year has added to the impression that there is a new consensus among AIDS researchers and doctors on the `correct' way to treat HIV and AIDS. Even in this issue's front-page article in which UK clinicians disagree with many of the details in the recent American treatment guidelines, it is striking how our British respondents broadly agreed with each other.

But in reality, there remains much valid debate among specialists over practical treatment questions. To illustrate the range of expert opinion, we asked the same questions in separate interviews with two clinicians who by reputation occupy opposite ends of the interventionist/conservative spectrum.

Dr Mike Youle is HIV Clinical Research Director at the Royal Free Hospital, and until recently coordinated research at the Kobler Centre. He still treats patients at both centres. Professor Ian Weller is Professor of Genito-Urinary Medicine at University College London Medical School, and sees patients at the Mortimer Market Centre.

Starting Treatment

--------------------------------------------------------- The new US guidelines recommend that anyone with a CD4 count below 500 and a viral load greater than 20,000 should be offered treatment. The International AIDS Society-USA suggest treatment for everyone with a viral load greater than 5,000 to 10,000. Do you agree? ---------------------------------------------------------

Mike Youle: Unfortunately I worry about our ability to assess true viral load, especially since in the UK people are infected with such a wide range of HIV sub-types (clades). At the moment I think it is only really possible to say that an individual's viral load is high, medium and low, and I am sure that there are many as yet undetermined factors which may make our assessment of what a viral load result shows less valid in a year's time. In addition the viral load only gives one side of the equation - the immune system may be doing an adequate job of sustaining T-cell levels even in the face of viral activity.

Finally, I worry that people may run out of therapies if they start too quickly and fail to suppress viral activity to very low levels. There is a difference between people who have never taken anti-HIV therapies but have a certain viral load, and those who are drug-experienced with the same viral load.

Ian Weller: I'm very uncomfortable with the emphasis being placed on viral load in the recommendations. My worry is this - viral load is just into our hands; we're just starting to get used to it and its variability. A fair proportion of patients with CD4 counts above 500, whatever their viral load, are going to remain well for a fairly long period of time. Therefore the question is this: do you start treatment then, or do you leave it later? I really don't know the answer.

I'm also uncomfortable about the emphasis on early intervention when we really don't know that that's the right thing to do. Biologically it makes sense, but do we have the right drugs? I`m concerned about exposing patients to our current drugs for a long period of time, beyond the time about which we know about safety. We're already starting to see new problems emerge such as diabetes (see AIDS Treatment Update issue 55).

--------------------------------------------------------- Do you take the view that it is `too late' to delay treatment until symptoms develop, or the CD4 count is low enough to put the individual at significant risk of symptoms? ---------------------------------------------------------

MY: Yes, although I think it's important to distinguish between those symptoms which may be a side-effect of `normal' immune responses to HIV, such as sweats, rashes and swollen lymph glands, versus those which reveal true immune deterioration such as candida and oral hairy leukoplakia.

I recommend treatment to everyone whose T-cell count is below 350, or whose T-cell count is between 350 and 500 plus viral load greater than the 50,000 to 100,000 range, and to anyone whose T-cell count is falling by more than 100 each year.

IW: I've certainly moved to starting therapy before people get sick, but I'm still fairly conservative.

At our clinic we currently start talking to patients about antiviral therapy when their CD4 count drops below 350. Personally, though, I would advise patients to wait until their count is just above 200, providing they're perfectly well and their CD4 count is not falling too fast.

I give them co-trimoxazole (Septrin) as PCP prophylaxis, so although it's fairly late they're not really at risk of developing symptoms.

--------------------------------------------------------- How do you advise patients to use viral load results in deciding whether and when to start treatment? ---------------------------------------------------------

MY: Get several tests done at a time when you haven't had any recent illness or heavy nights out, and follow the trend over time. Try to get a test that can detect the less common sub-types (non clade B) such as the Chiron or the improved Roche viral load test. Then use the guidelines I said above.

For people who are already on therapy, I'd consider a switch in therapy if viral load is increasing even if it has only risen to relatively low levels of viral load, especially for someone taking a protease inhibitor when any viral activity is undesirable.

IW: For patients with CD4 counts greater than 500 and high viral load, I would just see how things go for six months to a year. If that viral load really means something, their CD4 will start to drop towards 350 quickly, in which case I might well consider treatment.

I wouldn't recommend people with high CD4 counts to start treatment based on their viral load alone, because I want to maximise their years of life without drugs. I would say, `Look, I don't know how many years you'll end up taking these drugs if you start now. I don't know whether if you start treatment now you may compromise yourself later on.'

But for someone with a low CD4 count, I would start treatment, and if their viral load is very high I might recommend hitting it hard with triple therapy, as opposed to double.

--------------------------------------------------------- Do you think that someone who begins treatment when their CD4 count is 500 faces a significant risk of eventually running out of antiviral options? ---------------------------------------------------------

MY: This depends on so many different factors: how well currently available therapies work over the longer term; what new therapies we will have, including new types of drugs with different sites of action; what drugs someone starts with, since cross-resistance may limit their future options; how well they respond to and how well they tolerate the initial treatments, which will determine how rapidly they work through the available options.

I feel that most people will get good results if they are starting treatment today, but it's those that fail who may have limited future options.

IW: I think that's a real worry if you start too early, just like quite a lot of people who started early AZT monotherapy have lost their chance of benefiting from that drug.

We see cross-resistance with the PIs, cross-resistance between some of the nucleoside analogues and cross-resistance with the NNRTIs. And I can't see a new class of drugs arriving for two or three years at least.

--------------------------------------------------------- On the basis of recent advances, do you think that people with HIV can realistically expect to live a normal length of life? ---------------------------------------------------------

MY: Not at the moment, although I do think we have a valuable breathing space for the vast majority of patients. However, we only have two to three years' data on protease inhibitors and already we are seeing people in whom they are failing to keep viral activity suppressed.

In addition I think the impact of therapy failure on future options becomes increasingly problematic if we don't get new drugs with a greater range of target sites. Multi-drug resistant HIV is already with us, and we don't know how fast people with multi-drug resistance will deteriorate. We're also already seeing people becoming infected with drug-resistant strains from the outset, which will make it much harder to predict how well someone starting treatment will respond.

IW: There's no doubt that we have made some tremendous advances in the last few years, especially the impact of double and triple therapy when you use it. But we don't know how long the effect will last.

I don't for a minute believe that we're totally suppressing viral replication with the treatments we've got, even when we've pushed viral load below the limit of detection. Therefore resistance will emerge; it may take longer but it's going to emerge.

So I think the blunt answer to the question is `No'; we don't have the evidence.

Assessing response to treatment

--------------------------------------------------------- Do you agree that the ideal response to a new treatment regimen is for viral load to fall below the limit of detection? ---------------------------------------------------------

MY: Yes. All the data we have suggest that this is best. However, we do not have information on the relative benefit of dropping to very low - but still detectable - levels. Early data suggest that a rebound in viral load occurs earlier among people whose viral load drops least on therapy.

It's important to remember that the limit of detection varies depending on what test you use and what sub-type of virus you have; someone who has undetectable viral load with one test may have measurable virus using a more sensitive test.

IW: Biologically that makes an awful lot of sense - that the more you can inhibit the virus the better the patient will be. But the drugs we've got may cause longer term toxicities; there's the potential for resistance even if you've gone below the limit of detection; they're difficult to take and they impair quality of life.

When you take those practical issues into account, I don't know whether it's best to start treatment lightly in someone who is free of disease and then hit hard later, or to hit hard from the start.

--------------------------------------------------------- Would you consider modifying a new regimen after a few weeks or months solely because viral load remained detectable? ---------------------------------------------------------

MY: This would very much depend on what the patient wanted, what stage of disease they were at, and what antiviral options they had left. I would be more likely to recommend a switch for people at a relatively early stage of disease, but I would usually want to wait until they had been on drug for at least six months, since on average it takes at least four months for viral load to reach its lowest level. The only time I think a very early switch would be logical is for people who are very obviously failing early, with a very marked rebound in viral load.

IW: No. I think the point to think about changing therapy is if your viral load returns to baseline and at the same time you're not feeling well and your CD4 count's gone down. I wouldn't change therapy just because the viral load pops up a bit. I'd live with a bit of virus for a while and save those big heavy weapons for later.

--------------------------------------------------------- What would you regard as a satisfactory response to treatment - and how would you aim to achieve it - in (a) a treatment-naive person with a viral load of 40,000 and a CD4 count of 350? ---------------------------------------------------------

MY: I'd want to get their viral load below the level of detection, and see a T-cell count rise of at least 50. I think a good initial combination would be d4T/ddI/nevirapine. We know that d4T/ddI is a good combination of nucleosides and carries little risk of cross-resistance to other nucleosides should you fail. Adding nevirapine significantly improved responses in several studies, and it means that you can save all the protease inhibitors for second-line regimens.

IW: First up, I'd try to talk them out of starting treatment. But if they did, I would like to see at least a one log reduction in viral load and since their CD4 count is already 350, any further rise would keep me happy. If there wasn't a viral load response, you start wondering whether the patient was truly treatment-naive or might be infected with a resistant virus, or whether problems with absorption were preventing the drugs from reaching adequate levels.

At this early stage I would start with AZT/ddI. I want to save the protease inhibitors in reserve until later.

--------------------------------------------------------- (b) a person who had failed on AZT/ddI with a viral load of 100,000 and a CD4 count of 200? ---------------------------------------------------------

MY: Again, I'd be aiming to reduce viral load below the level of detection and boost the T-cells by 50 or more. You need to change the nucleoside analogues - 3TC/d4T would be a good option - and I'd also want to add a protease inhibitor, either nelfinavir or indinavir depending on the patient's wishes and the latest data on nelfinavir resistance.

IW: In that person, particularly if they have had an illness, I'd change both the nucleosides and add a protease inhibitor.

I'd want the CD4 count to rise above 200, and at least a one log reduction in viral load; since I'm using a triple combination I would actually hope to get a much better viral load response than just one log.

--------------------------------------------------------- Is too much emphasis being placed on the virological (such as viral load), at the expense of the immunological (such as studies of immune function)? ---------------------------------------------------------

MY: Yes, but that is a result of the improvement in our understanding of the virus and the ease with which new antiviral agents have been developed. As yet there are few immune-modulating agents although things are improving.

Now that the prime function of suppressing virus has been partly achieved, the new challenge will clearly be evaluating ways to enhance immune function. At recent conferences AIDS researchers have been showing a lot of interest in what can be learned about this from other fields of medicine, such as transplant medicine.

IW: Absolutely. If we're comparing treatments that make the virus undetectable, there's not much virological left to measure unless you wait until the viral load comes up again. But there could be huge differences between different combinations in terms of their effect on immune function. In trials we need to start measuring things like the rise in different types of CD4 cells and fall in activated CD8 cells, the ability of CD4 cells to respond to antigens and other more sophisticated immunological markers.

My concern is that with some of these combinations, as you blast the virus could you possibly also be blasting the immune system and impairing it with toxicities. It's quite possible that some drugs have toxicities on the immune system that we're not currently measuring.

Protease inhibitors

--------------------------------------------------------- The US guidelines suggest that a combination of two nucleoside analogues plus a protease inhibitor is the optimal regimen for anyone starting treatment. Do you agree? ---------------------------------------------------------

MY: No, since we have no head-to-head comparisons of people who start treatment with other combinations, such as a triple combination using an NNRTI. Is the degree of viral suppression the most important factor affecting the outcome of treatment, both in the short term and over the longer term? If so, a potent combination of drugs that all target reverse transcriptase should be just as good as a potent combination of reverse transcriptase inhibitors and protease inhibitors. Or is the important thing to attack the virus at more than one site at once? This is one of the main questions we have to answer.

What is clearer is that with NNRTIs you probably only get one chance because of cross-resistance, although there is increasing evidence that this is also true of protease inhibitors. We need this information, and fast. The problem is how to identify which approach is going to be more effective as part of a long term treatment strategy in which people will need effective second-line and third-line regimens too, rather than just to look at the success of the initial combination.

IW: I think that triple therapy is the optimal regimen for people who are more advanced, but I'm not confident that I want to start triple therapy for people earlier on. With the drugs that we've got there's a lot of uncertainty and I don't want to take that gamble. The problem of cross-resistance between the protease inhibitors is looking murkier every day, and it could be that you only have one bite of the cherry with the PIs. The question is, when do you want to take that bite?

Having said that, if a patient wants to start early of course I'm not going to withhold therapy.

Some people are saying that even when you do decide to use triple therapy you might use an NNRTI rather than a protease inhibitor. That's a rationale that's worth exploring, but we need trials to sort it out. In the meantime, people feel that we need to make recommendations, but in my view they're just pulling them out of the air.

--------------------------------------------------------- Under what circumstances do you currently recommend protease inhibitors to your patients? ---------------------------------------------------------

MY: When they have failed their first line therapy, especially if this contained an NNRTI. For people who start treatment at a very late stage, once their T-cell count has fallen below 50, I would probably start with 2 nucleoside analogues and a protease inhibitor, rather than hope that d4T/ddI/nevirapine would be enough, but that's still my greatest uncertainty.

IW: Only for patients who have failed on initial dual therapy, or who are starting treatment very late.

--------------------------------------------------------- Will the protease inhibitor bubble burst? ---------------------------------------------------------

MY: That depends what you mean. If you're asking whether people will fail, the answer is `Yes'. If you're asking how soon they'll fail, that's more complex and probably depends on a range of factors such as their viral load at baseline, the extent of their viral load response, compliance and their remaining immune response.

However, I think that for most people the bubble will not burst for several years and that theoretically gives time for a further generation of drugs to be developed. What it argues is that drug development is still vital for when failure occurs and that the blind optimism of the initial reports was dangerous.

IW: Yes. Hopes are always too high. The bubble is already leaking in three respects. First, there are the toxicities, and I'm concerned that as more time passes, more toxicities will emerge.

Then there's the difficulty in taking these drugs, and finally the worsening problems with cross-resistance.

Resistance

-------------------------------------------------------- Do you think that the development of resistance is the most important reason why treatment fails? --------------------------------------------------------

MY: It would appear that this is the main reason why the current drugs fail. This may not be true of future agents.

There may be other reasons for drug failure which are related to problems getting adequate levels of drug in the body, especially treatment adherence, poor absorption or interaction with other drugs or substances such as blood plasma proteins. Overall though, the two major reasons for failure are that the drug isn't there, or that the drug isn't suppressing viral activity to a low enough level to prevent resistance mutations from emerging.

IW: Looking into the causes of treatment failure is the most important question to address now. I think it's too simplistic to say that it's only down to patients not taking the pills, or to the emergence of resistance. Another cause of failure could be a toxicity on the immune system from some of these drugs. Or there may be some drugs with which we'll get a good antiviral effect but a very different response in terms of immune function.

We've still got a lot to learn about resistance. We don't yet understand the significance of all the mutations that emerge during protease inhibitor therapy. I think that resistance as we currently measure it is not going to be the whole story.

-------------------------------------------------------- How useful do you think that tests to detect HIV mutations associated with viral resistance will be for making treatment choices? --------------------------------------------------------

MY: Not particularly. It's becoming clearer that mutations can affect sensitivity to different drugs in quite a varied fashion. It's going to become more and more complex to ascertain which mutations are really associated with decreased sensitivity to drugs.

Phenotypic assays, which measure the degree of actual resistance to drugs rather than just looking for gene mutations, may be more useful although there are questions about these too. Tests will also be useful only if they are easily available and results are produced in a timely manner.

IW: There may be occasions when it might be useful to do a quick screen for mutations associated with the common drugs, such as when you have a patient who doesn't respond when starting treatment and you're wondering whether he or she might have been infected with a resistant strain.

But at this stage those tests are still a research tool. The worst thing would be to flood the market with tests when we don't know what they mean. We're not ready to start using resistance assays routinely in clinical practice.

-------------------------------------------------------- Do you agree that the threat of resistance provides a reason to delay treatment for as long as possible? --------------------------------------------------------

MY: I think it is a logical reason to delay for a time, because if you start treatment and the viral load is not suppressed completely then resistance will develop.

However, you could argue that resistance may be more likely to develop among people with late stage disease, because their own immune responses against the virus are less, and their viral load is likely to be higher and harder to suppress.

In other words, one way to minimise the threat of resistance may be to start treatment relatively early.

IW: Yes, I really do. We don't know how many patients who start treatment now will be able to benefit for a reasonable period of time.

Biologically I agree that early treatment makes sense and may reduce the risk of developing resistance. It's obvious that suppressing the virus is a good thing, but you have to pause to think hard about the drugs we've got. I'm worried that we don't know enough about their longer term effects.

It's terribly important to separate those two issues. Without any drugs, 50% of people will still be perfectly well ten years after they become infected. We may not have the right drugs to commit people to many years of therapy.

Trials and opportunistic infections

--------------------------------------------------------- Do you think that clinical endpoint trials are now unnecessary and/or impossible? ---------------------------------------------------------

MY: So few people in trials reach clinical endpoints now that I think that it is not as easy to conduct such studies. Perhaps what we need now are more strategic studies, looking at questions such as should you start with a triple combination containing a protease inhibitor or an NNRTI? When newer agents with different sites of action come along, it may be logical to compare them with current regimens and see which offers patients the longest `time to failure', but once again I think the prevailing mood would be that they should be based on virologic criteria for failure, rather than waiting for people to reach clinical endpoints.

IW: Not at all. After Delta, people said that would be the last clinical endpoint study we could do, but the licensing authorities asked for clinical endpoint trials of the protease inhibitors. Consequently, in the last two years we've had a wonderful series of clinical endpoint trials that have demonstrated that triple therapy has a better impact than double therapy in later disease.

As a result of those trials I'm happy to give triple therapy to the populations that were studied in those trials. If we only had studies looking at treatment effects on viral load and CD4 counts I'd still be hovering about prescribing triple therapy.

I don't for a minute underestimate the difficulty we now face in doing clinical endpoint trials, but it will be even harder if the licensing authorities completely remove the requirement for anything more than viral load and CD4 data. What if those markers are inadequate for some drugs? We won't know unless we have clinical endpoint data.

--------------------------------------------------------- What are the most problematic OIs that your patients are now facing? ---------------------------------------------------------

MY: One of the most difficult is tuberculosis at later stage disease, as drug interactions usually require modification of antiviral treatment. CMV is still problematic. People who've had CMV disease appear to remain at risk of further episodes until at least 6-12 months after their T-cell counts rise above 50.

Progressive multifocal leukoencephalopathy (PML) is still a wild card and really doesn't seem to be inhibited by HIV therapy.

However, suffice it to say that if anti-HIV therapy boosts your T-cell count above 100, opportunistic infections are very uncommon.

IW: We've obviously seen a real reduction in the number of people who are coming in with OIs, although it remains to be seen whether the effect will last.

The group that's most worrying right now is those who haven't been tested and show up at casualty with PCP. They're largely gay men who have decided not to be tested so haven't been getting prophylaxis or anti-virals. Those cases haven't gone away.

--------------------------------------------------------- For which patients, if any, do you recommend primary prophylaxis against CMV or MAI? ---------------------------------------------------------

MY: No routine prophylaxis for either. But for CMV I would use aggressive pre-emptive therapy for those who are CMV PCR positive (only available in limited centres). For MAI I'd rather screen people thoroughly and start treatment at the earliest signs of disease, especially as you can't really use protease inhibitors with anti-MAI agents except rifabutin.

IW: None. I am not convinced by the evidence that it's really worthwhile. There also problems with interactions between MAI drugs and triple therapy.

Conclusion

--------------------------------------------------------- What is the most important thing that AIDS researchers have learned during the last two years? ---------------------------------------------------------

MY: That suppression of HIV to very low levels can lead to reversal of even severe disease and results in marked improvement in health.

IW: That combination therapy is the way forward, although we've still got a lot more to learn about resistance.

--------------------------------------------------------- What is the most important thing that AIDS researchers have failed to learn? ---------------------------------------------------------

MY: That there are two sides to the equation and that more aggresive research into immunomodulatory agents is also required.

IW: We haven't learned the lesson of AZT monotherapy. It was shown to be effective in symptomatic disease and so was assumed to be effective in early infection, but then we showed that it wasn't the best thing to intervene early. I would predict that if we do the trials we'll show that exactly the same is true for triple therapy with the drugs we have.

The trouble is that people are so sold on the idea of early intervention with triple therapy that they won't start trials to test their beliefs. I have a real concern that we'll live to regret it.

970801
ATU05602

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