BHIVA clinicians roundtable feedback from the Eighth Annual Retroviruses Conference, Chicago - Wednesday
Dr Martin Fisher, Consultant in HIV Medicine, Royal Sussex County Hospital, Brighton
Dr Anton Pozniak, Consultant, Chelsea and Westminster Hospital, London
Caroline Sabin, Statistician, Royal Free and University College Hospital Medical School, London
Dr Ian Williams, Consultant, Mortimer Market Centre, University College Hospital, London
Interviewer: Keith Alcorn
- Viral blips and when to switch
- Viral reservoirs
- Metabolic complications and lipodystrophy
Individual abstract numbers link to the abstracts online at http://www.retroconference.org.
Where session titles are underlined, this link takes you to the list of all abstracts presented in that session.
Where presenter names are underlined, this presentation can be viewed online, and the link will take you to the list of sessions which can be viewed online. size=2>
Viral blips and when to switch
KA: What do you think was the significance of the studies on viral blips? (Session 60: Antiretroviral Chemotherapy, abstracts 519-528)[These studies looked at the consequences of occasional increases in viral load above 50 copies in people with long-term viral suppression on HAART].
AP: There’s a big question about whether the assays they were using to measure viral load were actually missing the real data. In other words, when it comes out as 200, it might be less than 50 copies – that is, the test is not done very well in the lab. There’s no data on matching adherence to blipping – that is, you’ve missed one or two doses and had a blip. Tony Fauci’s data suggest that that might not be true, because in the week on week off treatment interruption study, they didn’t have anyone blip during the seven days off treatment.
IW: There was only one study with data on whether blipping is affected by adherence, which looked at DOT versus self-administered therapy. There was an equal number of blips in those patients who were given therapy through DOTS and self administered therapy (abstract 528), despite the fact that the DOT group had higher baseline viral load, 100% adherence and better virological outcomes out to 80 weeks – all of this suggesting that adherence may not be the factor driving blips. The other thing, which came out of Diane Havlir’s study (abstract 521) - and it’s related to the issue around therapy interruption - is whether the cycles of blips drive new immune responses. The main finding in her study was that in fact there was no effect on stimulation indices [which show the levels of HIV-specific immunity]. There was no difference in those who had blips and those who didn’t have blips in their immune response to different HIV proteins.
MF: Which implies that short low level viraemia may not have an immune benefit.
IW: So maybe Fauci’s week on week off may not be good enough to drive an immunological response.
MF: The other thing people are concerned about with blips is whether you get resistance mutations. And they‘ve done very little work on that, but what work has been done suggests there are aren’t any, but then a man stood up and said that from his data you get resistance mutations to everything. And Anthony Fauci has one patient who’s developed mutations, but that was after two months on, one month off.
IW: I think it’s reasonable to suggest that you don’t get mutations with blips.... In the Havlir study they didn’t get any 3TC mutations at all in their blippers.
AP: The problem with that is that you may get them at a very low level in tiny minority species which are stored up, but the other drugs are enough to keep them suppressed.
MF: The second study certainly seemed to suggest that the higher your blip and the more sustained it was, the more likely your therapy was subsequently to fail. But again, most people with low level viraemia went back down below the limit of detection and I wondered to what extent that should change national guidelines which currently suggest if you have two consecutive viral load measurement over 50 copies then you should change drugs.
IW: While Diane Havlir had said there was no association with subsequent failure in those with blips and those who didn’t, in a much larger study involving many more patients – the Swiss Cohort and Frankfurt Cohort (abstract 522) – there was trend towards a 1.5-fold higher risk of subsequent failure with a blip. So that adds further data on a much larger population.
I think the frequency of monitoring is clearly important. The other point I think comes out from Alessandro Cozzi-Lepri’s study (abstract 522) is that you don’t really know how long it was from the first blip above 50 copies to the more sustained blip of between 300-500 – constant low level viremia. What time period was that? When do you repeat the measurement? How long do you wait? Once you get to 300-500 copies, there’s a much higher risk of it going on to greater than 500 copies and failure. So that’s when therapy modification had a definite impact on getting it down to less than 50.
CS: I think the other thing which I noticed from this study is that they didn’t give you the CD4 count at the same time. It would be interesting to see if the blip coincided with a sudden drop in the CD4 count and whether that was maintained.
IW : Coming back to the issue of resistance, about whether patients who blip are more likely to develop resistance, the thing which I would ask Alessandro is: those patients who blipped who had sustained blips, two or three times, was there a difference between those on NNRTIs versus PIs who went on to fail? I think that that’s an issue that affects clinical management of blips.
MF: So you may wish to switch different types of drugs earlier?
IW: It may well be that if you blip frequently with NNRTIs, there’s more likelihood of failure because the genetic barrier is lower, than if you blip with PIs where it may not matter so much.
AP: Our data, which is still unpublished, suggest that is not the case. We just looked at non-nukes, and in people that blip and don’t blip the rates of failure are exactly the same but what we are waiting for now is the resistance data during the blip.
IW: And people on PIs?
AP: We haven’t done that - we’ve just done people on NNRTIs and the reason for that was that it was a resistance driven study. During the blip do you get any resistance mutations?
MF: The other issue is that we are dividing people into blippers and non-blippers, and it may be that if we measured viral load frequently enough everybody is a blipper.
CS: That was one of the questions that was raised. I think that in the Frankfurt data where they measure viral load every month there were more blippers, compared to the Havlir study where they measured every three months or so. This is obviously a key area to research because if people switch drugs with a single blip, they run out of drugs very very quickly. So in reality we are going to have to deal with this.
IW: Coming back to the earlier question ‘should BHIVA guidelines change?’ –then I think they probably should. Changing therapy based on two viral loads above 50 copies/ml is probably too harsh.
AP: Well it depends how high it goes and what the trend is. But that needs a lot of discussion.
KA: Do you think we are quite there yet – is there enough data to say that the guidelines should change?
AP: I think it needs further discussion. The question about when to change drugs is crucial. As Ian and Caroline said, you just start running out of drugs.
KA: And the drug pipeline is probably not as full as we’d like to think.
IW: We need to look at the evolutionary resistance in different combinations of drug, in PIs and NNRTIs, and whether your subsequent response to therapy is going to be any different whether you change at 50, 500, or 5000.
MF: Also, if you change at such a low level, certainly with current technology you will never know what you were resistant to at the time that you changed.size=2>
Viral reservoirs
KA: There have been several good presentations on viral reservoirs at this conference from Robert Siliciano and a symposium with Tony Fauci and David Ho.
AP: Reservoirs of HIV are laid down in primary infection, Fauci went over that again today in the symposium on viral reservoirs, and they can’t eradicate them. He actually said that today. So he is really talking about long term control. He is trying to cut down on the amount of antiretroviral exposure you get during your lifetime.They’ve got 10 patients to whom they’ve given 22 cycles of 7 days on, 7 days off and they haven’t looked at all the toxicity problems. He didn’t answer the questions – if you have 7 days off, and you start again, do you get nausea, vomiting and diarrhoea again?
IW: They don’t get a viral load increase during that time off.
AP: During that 7 days [off treatment], but they did have two patients who had viral rebound, one who waited to 10 days before re-starting, so maybe it takes a week before [the viral load] comes up. And the other guy went on a cruise. Let me just say that the median CD4 was 940,and they are incredibly well looked after people, because there are very few in the study, and the motivation to take their drugs must be great.
MF: So they come in twice a week for viral loads –
[general laughter]
MF: - so the applicability of that is very dubious!
CS: I don’t know if you saw Andrew Phillips’ presentation at BHIVA on the idea of treatment interruption? His conclusion was that if you do something for seven days on, seven days off, there is a high risk for getting resistance. And maybe they haven’t seen it yet, but it is a real worry because it could generate resistance very very rapidly.size=2>
Metabolic complications and lipodystrophy
KA: Let’s talk about metabolic complications. There were quite a lot of posters on osteopenia, apparently conflicting in their conclusions (abstracts 626-638). What impression did you come away with?
IW: I think there were 3 or 4 posters which looked at background levels of osteopenia in patients who are naive to therapy or had very short exposure to therapy (less than four months) compared to those who have been treated, and all these were compared to HIV-negative controls, and the rate of osteopenia was higher in all HIV-infected individuals. He was looking at an expected background, for an age-matched population, of around 15-16% percent and found osteopenia in 25-26% of the treatment-naive patient population. It appears that maybe HIV disease has an effect on osteopenia which was then further increased in people on treatment.
AP: And one group had a mechanism for it, with ritonavir inhibiting osteoclasts and indinavir inhibiting osteoblasts (abstract 541).
IW: The other thing was Andrew Carr’s poster(abstract 631). He showed an association between the incidence of lactataemia and osteopenia in his cohort but I think that was an association rather than a causal factor.
MF: I thought that with Andrew Carr’s presentation, one of the most significant issues was around routine practice in terms of lactate measurement. He said there was no justification for routine lactate measurements. In the absence of symptoms an elevated lactate may be over-responded to.
AP: He gave a nice algorithm of what to do. He basically said, if lactate is greater than 10, stop treatment. If you are 5-10 with symptoms [Fatigue, tiredness, abdominal bloating or discomfort, abnormal LFTs] – stop the nucleosides. If you are asymptomatic 5-10, it’s probably an artefact of the test and repeat the measurement. If it is anywhere from 2-5 consider stopping if you’ve got symptoms, but really try to exclude other problems that might cause the raised lactate.
MF: He said that the high lactate took a long time to resolve when he took patients off the therapy – is this a production mechanism or is it a clearance mechanism? Nobody in the conference has addressed that as far as I can see. Are we talking about lactate not being cleared by the liver (and that’s the reason for the abnormal liver function test) or is it a problem with lactate production in abnormal mitochondria?
AP: I can’t believe lactate clearance takes weeks. If you go for a run, get lots of lactate, even if you have liver disease, you’re going to clear it fast. I think it must be a production problem.
IW: And I think he ended on a very cautious note - although the lactate may return to normal, albeit slowly, we don’t know what the long term consequences as far as liver damage might be.
MF: I think it’s a very important question as to whether slightly raised lactate might have a knock on effect as far as liver disease is concerned.
CS: Changing the subject, did anyone see the posters on cardiovascular disease? (abstracts 655-658). Because there was a very interesting contradiction between two posters virtually next to each other(abstracts 657 and 655) . There was one from the French database on myocardial infarction rates which showed that over time there was an increase in MI rates and that this was related to duration of time on PIs.
AP: Up to 18 months there was no difference with the expected age-matched incidence in the general population.
CS: But two posters along from that, one from the Kaiser Permanente data set in the States showed no difference. There was a small difference in the incidence rates in the PI patients versus non PI patients, but it was not significant. What was interesting there was that the rate generally in the HIV positives was significantly higher than in the HIV negatives. And methodologically what’s interesting about that, is their data, and there are a lot of problems with using data from the healthcare systems in the States, but [in this case] it is all objectively measured because it is based on hospitalisations so there is no reporting bias. So if somebody is hospitalised with an MI, it is in that dataset. Whereas in the French study there is always the possibility of reporting bias and it may just be an artefact of people looking for MI whereas before it wasn’t getting considered.
MF: Was that the study [abstract 655] where the HIV negatives had about 3 per 10,000 and on PIs it was double that?
CS: Yes, it was 5.5 per 10000 and that was significant. So their point was that maybe it is a HIV effect that is confounded by the use of PIs, because by the time you have started to see it in the HIV patients it was about the same time PIs came into routine use.
IW: The other interesting thing was the poster on indinavir which showed that there is an association between indinavir and hypertension. I’d hadn’t seen that before and I just wondered about the aetiology, what the mechanism behind that was.
MF: Steven Grinspoon did say [in his presentation] that if you decrease your waist measurement and decrease your insulin resistance and decrease your glucose, then your blood pressure goes down.
IW: So it’s more related to your insulin resistance and your triglycerides.
MF: - and your weight increase -
IW: - all driving your blood pressure up.
AP: The other question on this is: do they exclude renal disease with the indinavir in driving your blood pressure up?
IW: The last point I would make is the issue around the increasing incidence of avascular necrosis seen in the Baltimore cohort (abstract 638).
AP: 4.4% by MRI category, two actually had disease. If you look at them, they are on testosterone, or steroids, or lipid lowering agents or they are body builders. But the Baltimore group had an increased incidence from 1995 - when they had no cases - to the year 2000 when it was 4 per 1000 patient years. And the worry is there was a 15-fold increased incidence in children.
IW: In the San Francisco survey (abstract 628), treatment-naïve patients were screened by MRI , and this study suggested actual underlying disease incidence and prevalence in the patient population was higher than the clinical disease presentation.
MF: And they thought it was associated with the duration of HIV disease and low CD4 count.