BHIVA clinicians feedback from Buenos Aires

The First International AIDS Society Conference on HIV Pathogenesis and Treatment took place in Buenos Aires between July 8 and 11 2001. We invited a group of UK clinicians and experts to give feedback on the key presentations from the conference.

The expert panel was:

Dr Margaret Johnson, Director of HIV Medicine, Royal Free Hospital, London

Dr Ian Williams, Mortimer Market Centre, University College Hospital, London

Dr Caroline Sabin, Reader in Medical Statistics and Epidemiology, Department of Primary Care and Population Sciences, Royal Free and University College Medical School

The interviewer was Keith Alcorn, editor, aidsmap.com

Abacavir and indinavir compared

KA: Forty eight week data were presented on the open label comparative study of abacavir and indinavir. Did you find those data convincing in terms of prescribing abacavir as first line therapy, given that this was a comparison with three times daily indinavir?

MJ: I think this study tried to address what impact the study design of CNAB 3005 had on the outcome, given that that was a double blind study with many pills per day [Vibhagool]. This study tried to re-look at the question [of equivalence] in a real life situation, and the results were pretty similar to 3005. The problem is that they were comparing abacavir with three times daily indinavir, which is not a regimen any of us would be prescribing as first line therapy today. What it showed yet again was much better adherence to the simple regimen, and that in patients with low viral loads [below 100,000 copies], the outcome appeared to be better [in the abacavir group].

KA: What about patients with high viral load?

MJ: I think you clearly need to be thinking about more potent therapy in that group. That might be four drugs, or that might be another triple regimen. If you looked at the as treated analysis indinavir appeared to be performing better than abacavir, but not in the intent to treat analysis, and they had good adherence data which clearly showed that patients adhered better with a simpler regimen.

CS: What I would have liked to have seen was an analysis which didn't treat switches as failures.

CNA 3014: Combivir plus abacavir 300mg bid or indinavir 800mg tid

As treated analyses: do they tell us anything useful nowadays?

KA: You raised the issue of the difference between the intent to treat analysis and the as treated analysis. In what circumstances should we pay attention to the as treated analysis?

CS: The as treated analysis is answering the question about what happens if you can keep the patient on the same drug - how good is it going to be? But in practice when you are making a decision about what combination to start a patient with, you can't tell how likely it is that that patient will stay on that drug, so it's very difficult [from that analysis] to tell the patient how likely it is that that combination is going to be effective.

MJ: And I don't know that that should be our strategy any more. We know that these regimens are all very complicated to take, and different patients have different problems. For me, it is a very important strategy to change early if it's not a combination that's going to suit that patient.

IW: I entirely agree. On treatment analysis gives you an idea of the antiviral efficacy of the drug in patients able to tolerate that drug. If you look at the BEST study (a comparison of tid indinavir with bid indinavir with ritonavir liquid), if you look at the intent to treat analysis there was no difference, but if you look at the as treated analysis, 41% of the RIT/IND arm came off, and the indinavir tid arm was more effective, probably because they were more adherent.

BEST: continued indinavir (800mg tid) versus switching to indinavir and ritonavir (800/100mg) (Cahn)

KA: They were using liquid ritonavir in that study, but there have been a number of other studies [at this conference] which have looked at ritonavir capsules with indinavir. What sort of performance have they shown?

IW: I think you're right that the BEST study was markedly disabled by the fact that they had ritonavir liquid, but I think it's also difficult to interpret that switch in people who are already effectively controlled - and therefore very adherent to a tds [indinavir] regimen. I think it is a very false clinical situation. But there was an interesting small study of 21 people presented by Christine Katlama, which looked at a 400/100 mg dose of indinavir and ritonavir, [in] patients who were stable on tid therapy and then switched to 400/100mg, and she showed that the [indinavir] Cmin increased two and half fold, and the C max came down two and half fold, so you are flattening the pharmacokinetic curve but increasing the Cmin. She showed that it was a very well tolerated combination, two pills twice a day, and there was no increase in triglyceride levels from adding in the ritonavir, so it might be a useful combination in patients who are treatment naive, although you could argue that in patients in salvage therapy, they may need the higher dose of indinavir [to overcome drug resistance].

MJ: What's very important in that study is to look how cheap it is! It may be an important strategy for countries where cost is an important consideration.

CS: - like the UK!

MJ: I was also very surprised by the high level of gastrointestinal toxicity reported in the BEST study, because in clinical practice that's something I don't see with ritonavir/indinavir - it tends to be very well tolerated from a gastrointestinal point of view and I wonder whether that was almost entirely driven by the nausea caused by ritonavir liquid.

KA: That level [of gastrointestinal side effects] was not seen in the Katlama study, was it?

MJ: No, although that was a smaller study. But it's not something you see in clinical practice.

IW: I wouldn't want to over-interpret the Katlama study. It's 21 patients, already well controlled on therapy, and it was mainly a pharmacokinetic study, so I don't think we can say that we should be using that as a starting dose.

CS: I think what we've been saying emphasises the fact that we do need to look at the incidence of adverse effects and you do need to have an idea of what is going to happen to the whole population of patients, not just the patients who can tolerate the drug. I think it also goes back to something that Brian Gazzard was saying in his talk yesterday - that we have also got to remember that while you may have a first treatment regime which performs particularly well in one group, it's actually what happens to the second treatment regime in those patients - maybe they are going to do worse. The strategy is what we need to be looking at, rather than just focusing on the first treatment regime.

Therapeutic drug monitoring

IW: It may well be in naive patients you don't need to drive your plasma levels too high because you have wild type virus, but with salvage patients you may need to drive plasma levels much higher to overcome resistance -

KA - which has implications for the use of therapeutic drug monitoring in salvage therapy to fine tune plasma levels and monitor the risk of side effects. Was there any further data at this meeting on therapeutic drug monitoring?

IW: The Athena study is an important study. In indinavir recipients they were able to adjust the dose of indinavir mainly for reasons of toxicity, rather than for treatment failure, which was mainly the reason for dose adjustment in the nelfinavir group.

Athena: Randomised study of effects of therapeutic drug monitoring on virological outcomes after 48 weeks therapy with indinavir or nelfinavir

CS: How many years have we been using TDM? Two years, three years? We still don't know whether it's beneficial or not -

MJ: - and where -

CS: - and whether it's cost-effective.

IW: Clearly there is a relationship between pharmacodynamics and effectiveness and toxicity for the protease inhibitors, so that's where I would expect the information gained from TDM to be useful.

KA: The thing that I found most striking about the Athena study, although that data wasn't presented yesterday, was the proportion of people who gained virological control in the nelfinavir arm simply after having a discussion with their doctor about dosing of nelfinavir with food, prior to any nelfinavir dose adjustment. That raises serious concern for me about the amount of time that doctors have with patients to discuss the requirements of therapy at the very beginning, and I was surprised that at centres in the Netherlands which are apparently well resourced, that discussion doesn't appear to have been taking place.

CS: Maybe it has, but at the end of the day you don't necessarily believe everything you are told until you see that the drug levels are not high enough and you think "maybe I was being a little too flexible with meal arrangements and maybe I should tighten up what I'm doing". Demonstrating that by looking at the drug levels is probably a good way of proving to patients that you really do need to follow these guidelines quite strictly.

MJ: I think you are giving a patient an enormous amount of information the day that he or she starts therapy - you are telling them about three drugs, you are telling them about how you are going to monitor them thereafter, you are talking to them about how to take the drugs - perhaps we are giving our patients too much information all at once, and we are going to have work much harder at getting starting therapy right.

IW: There is also marked inter-patient variability in protease inhibitor levels, and therefore taking it with food, if it has the effect of increasing plasma levels by 30% may be very important for some patients, but not so important for other patients.

Adherence

MJ: There was a variety of studies that focussed on adherence, but the Friedland study was particularly useful. It was a large, 1,000 person study of people which quite convincingly showed to me that a simple self-administered questionnaire correlated very well with virological outcome, and sometimes I think we might be making adherence monitoring very complicated with MEMScaps and bleepers, and something simple, in this study, seemed to give very good information. The data from that study clearly showed that adherence decays over time, and the difference between 80% and 99% and 100% adherence results in a big difference in virological outcome, that patients will be more adherent to non-nucleoside based regimens than PI-based regimens, and that naive patients appear to be more adherent to therapy than treatment-experienced patients.

Duration of antiretroviral adherence predicts biologic outcome in clinical trials: data from CPCRA 057 and 058 participants gathered from a seven day recall self-complete questionnaire

IW: the other interesting thing on adherence was Julio Montaner's data presented at the opening session which suggested that one of the reasons why patients with low CD4 counts tended not to respond to therapy as well as patients with CD4 cell counts above 50 was due to poor adherence, and that it is the poor adherence that is driving the poorer virological outcome rather than the fact that particular drugs are not potent enough in that patient population.

[In the

MJ: There was also a presentation from the same cohort showing that women appear to do less well after starting therapy [O`Connell], and were less likely to have virologic success after one year on therapy, and it was related to age and more injecting drug use in that group, but they may be busy with other things which may mean that their own medical problems are put down the list, such as looking after children.

CS: In terms of the actual effect of gender, it was clear that it was to do with adherence and lifestyle factors, and it wasn't to do with a biological difference between men and women and their response to therapy per se. Where men and women were equally adherent to therapy, their response was similar.

Switch studies

KA: What do you think were the key messages from the studies which looked at the effects of treatment switches on lipid levels and body composition?

IW: There was a study from Andrew Carr and David Cooper, following on from their PI-sparing study [presented at the Retroviruses conference in San Francisco, 2000], in which they withdrew the thymidine analogue, AZT or d4T, in a small number of patients. Nine patients continued and nine patients stopped. They showed that there was an increase in limb fat gain [after 24 weeks] in those who stopped therapy which suggested that there was an association with thymidine analogues.

[Five of nine patients who ceased d4T or AZT experienced viral load rebound within 24 weeks]

KA: What about the findings from the Atlantic study, where they looked at body composition changes after two years on treatment, and also the ritonavir/saquinavir study where they looked at the effect of adding nucleoside analogues on body composition over four years of follow-up? [abstract 488]

IW: The whole field of lipodystrophy is becoming more and more confusing! The associated factors become more and more difficult to tease out, but there do seem to be consistent data showing that the nucleoside analogues are associated with lipoatrophy, or peripheral fat wasting. In the Atlantic study, where there were three arms - a protease containing regimen, an NNRTI containing regimen and a NRTI-containing regimen, they carried out a cross-sectional survey after two years in a proportion of the patients, and also carried out DEXA scans and CAT scans in some, and found no difference in the rate of lipoatrophy or central fat accumulation between the three arms. How do you interpret that? You might have expected more fat accumulation in the indinavir arm than the other two arms. It may have been something about the nucleoside backbone, d4T/ddI, which was common to all arms.

The other study is a long cohort study following people receiving open label ritonavir/saquinavir, and they compared people who intensified their therapy with nucleoside analogues, and they clearly showed that there was a higher proportion of lipodystrophy and lipoatrophy in people who intensified their therapy - evaluated by various different measures.

KA: What was particularly interesting in that study was that the prevalence of fat accumulation was greater in those who added nucleoside analogues to ritonavir/saquinavir, suggesting that there may be an interaction between the two classes of drugs, not just in terms of lipoatrophy but also in terms of fat accumulation - although they didn't report on the effect of the duration of NRTI therapy, either within in this study or prior to joining it [half the participants were NRTI-experienced].

Effect of nucleoside intensification on prevalence of body fat changes after four years of ritonavir/saquinavir therapy (NRTIs added if virologic rebound or at clinician discretion after week 48)

Nevirapine toxicities

KA: There were several studies on nevirapine toxicity at this meeting - what were the key messages? Did they say anything new?

MJ: I don't think they said anything particularly new, but the messages bear repeating.

IW: There was a randomised study looking at prednisolone which once again showed that it doesn't seem to protect you against nevirapine rash - that study was in abacavir and nevirapine [abstract 92].

MJ: That study did have another important conclusion, which was: don't start those two drugs together, because it is so difficult to tease out which is causing rash, and you then lose your ability to use abacavir

KA: And they did show that using the two drugs together did not have a synergistic effect on the incidence of hypersensitivity reaction, as some had thought.

IW: The other thing that came out of that study was the finding that you are more likely to get a nevirapine hypersensitivity reaction at a higher CD4 count [the risk increased 1.27-fold per 100 cells), and that was also shown in the phase IV post-marketing study by Bennett et al, which showed that most of the reports had occurred in people with high CD4 counts, and that it is more common in women.

[Boehringer's own analysis of four of their pivotal studies found a one year hepatotoxicity rate of 2.9% in people with baseline CD4 counts below 350 and 8.1% in people with CD4 cell counts above 350 (abstract 44)]

[A meta analysis of 21 randomised ACTG studies looked at incidence and risk factors for hepatoxicity. No difference was found between NNRTI-containing regimens and PI-containing regimens, nor between nevirapine and efavirenz-containing regimens (although the confidence intervals in the case of efavirenz were much wider than those observed for nevirapine)] [abstract 43]

Structured treatment interruptions

KA: There's been less attention focused on structured treatment interruptions at this meeting. Did you learn anything new on the subject here?

CS: I agree that there's been little new - just more people showing the same pattern. I still don't think we're any further forward in showing whether they are safe or not.

IW: I think the Fauci study was interesting. It was designed to show whether or not there was any improvement in virological or immunological outcome with a seven days on, seven days off regimen, and there was no difference in immunologic markers or virological control, but what they did show was that all the lipids got better!

MJ: But don't you think these are still very small groups of patients? I still have a real worry that where for example you have two months on, one month off, the risk of resistance has already been proven, and that from a clinical point of view we shouldn't be going anywhere near it at the moment.

IW: I entirely agree it's a research strategy and you need highly motivated patients to take one week on, one week off.

KA: The main problem with the rationale which Diane Havlir presented for doing STIs in chronic HIV infection - that it reduces drug exposure and therefore reduces toxicity - is that in the group of patients which that applies to, people have extensive treatment experience and low CD4 counts, so we know what's going to happen - they are going to have rapid CD4 declines, and even though their lipids may improve, they are going to be at risk of clinical illness within six months.

MJ: There were also several posters showing that in patients with extensive treatment experience, although they revert to wild type [virus] during a treatment interruption, it makes no difference when they start treatment again. This is the group of patients with the greatest danger of rapid CD4 decline, a group of patients clearly at risk of opportunistic infections, and I think the data is beginning to not support that strategy.

IW: These were non-randomised studies, and I'd hate to pre-judge the outcome of the OPTIMA study, but you might get only a transient benefit [from a treatment interruption].

References

Bennett C et al. Hepatic and cutaneous toxicity attributed to nevirapine (NVP). First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 30, 2001.

Amiel C et al. Treatment interruptions in heavily antiretroviral pretreated patients: clinical and genotypic evolution. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 445, 2001.

Burger D et al. Therapeutic drug monitoring (TDM) of nelfinavir or indinavir in treatment naive patients improves outcomes after 1 year: results from the ATHENA study. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 30, 2001.

Cahn P et al. Continued indinavir (800mg tid) versus switching to indinavir and ritonavir (800/100mg bid) in HIV patients having achieved viral load suppression. A randomised study: the BID Efficacy and safety Trial. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 60, 2001.

Carr A et al. A randomised study of thymidine analogue withdrawal in lipoatrophic HIV patients, virologically controlled on protease sparing therapy - the PIILR extension study. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 96, 2001.

Cohen C et al. Effect of nucleoside intensification on prevalence of morphologic abnormalities (MOAS) at year 4 of ritonavir plus saquinavir therapy in an HIV-infected cohort. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 93, 2001.

Dieterich D et al. Analyses of four key clinical trials to assess the risk of hepatotoxicity with nevirapine: correlation with CD4+ levels, hepatitis B & C seropositivity and baseline liver function tests. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 44, 2001.

Fauci A. Host factors in the pathogenesis of HIV disease: implications for therapeutic strategies. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract PL4, 2001.

Friedland GH et al. Duration of antiretroviral adherence predicts biologic outcome in clinical trials. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 33, 2001.

Katlama C et al. Ritonavir (RTV)/indinavir (IDV) (100/400mg bid): a simple effective and well tolerated PI-containing regimen. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 214, 2001.

Montaner J et al. Further characterizing determinants of disease progression among HIV-1 infected patients initiating triple drug therapy. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract LB-10, 2001.

O'Connell J et al. Differences in virologic suppression among men and women enrolled in a population-based antiretroviral drug treatment program. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 62, 2001.

Reisler R et al. Incidence of hepatotoxicity and mortality in 21 adult antiretroviral treatment trials. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 43, 2001.

Slom T et al. Body composition changes in HIV-infected patients treated with NRTI, non-NRTI or PI-based therapy: preliminary results of the fat redistribution and metabolic sub-study (FRAMS) of the Atlantic study. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 488, 2001.

Vibhagool A et al. Abacavir/Combivir is comparable to indinavir/Combivir in HIV-1 infected antiretroviral naive adults: results of a 48 week open label study (CNA 3014). First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 63, 2001.

Wallace M et al. Structured treatment interruption: effect on triglyceride levels. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 448, 2001.

Wood R et al. Short term immunosuppressive therapy with prednisolone in HIV-1 infected patients receiving abacavir with or without nevirapine. First International AIDS Society Conference on HIV Pathogenesis and Treatment, Buenos Aires, abstract 63, 2001.