Audio With Slides Audio Only (69 min., 27.5MB)

Table of Contents

• Introduction
• Questions & Answers
• Post-Test

Introduction

Moderator: Greetings and welcome to the continuing education webcast titled New HIV Guidelines. This program is jointly sponsored by the Annenberg Center for Health Sciences and Clarus Health. This program is supported by an independent educational grant from Bristol-Myers Squibb. In order to receive continuing education credit for this activity, participants must complete the program evaluation and post-test. I would now like to turn the conference over to your speaker, Dr. John Bartlett. Thank you, you may now begin.

Dr. Bartlett: Welcome everyone. I'm going to talk about the DHHS federal guidelines for HIV care that were released as a new edition on Dec. 1. These are available on the Web at AIDSinfo, the source of all the federal guidelines. I'm going to try to highlight those things that have changed since the last version, which was probably six months earlier. We'll probably have another version not too deep into 2008.

Let me give credit to the people who did the work. These are the people that are on the DHHS panel. It's composed of about 25 people that do HIV care, and that's the majority of the people that are on the list. There are five community representatives and there are seven from various federal agencies, like the FDA, the CDC, the NIH, HRSA (Health Resources and Services Administration) and so forth. All of these people have a three-year term. All these people are on a conference call once a month. Various parts of the guidelines are distributed to somebody that agrees to chair a subcommittee that will update a section and then they do it, and then the entire group reviews it on a conference call and a vote. Eventually it makes its way into the final document, or the revised document, as it were. So what I'm going to talk about are the things that have happened in the last six months that are now official. What you see on the slide are the people that actually did the great majority of the work.

So we'll start off with a question. I started my remarks by saying what you're going to see were the changes. The main changes were [regarding] the issue of when to start therapy, and then we dealt with the new drugs (maraviroc and raltegravir) and put them into all the tables and what not. We talked about the tropism assay for maraviroc use or CCR5 inhibitors. We also dealt with the HLA-B*5701 genetic test for susceptibility to abacavir. We had a lot of discussion about when to change therapy and how to change therapy and use of resistance testing. There is a separate committee that deals with pregnancy and a separate committee that deals with antiretroviral agents for children. I'm not going to dwell on those, but I do have one slide on the issue of pregnancy.

Let's get started. This is a polling question. It's about a 40-year-old man who has recently-detected HIV infection and a CD4 count of 400/mm³. He wants to start therapy. And the question is which of the following would be the clearest rationale for him to start therapy?

The options given here are:

1. Viral load >100,000/mm³

2. HIV-associated nephropathy

3. His wife is seronegative, so he is [in] a "discordant couple."

4. A CD4 done 3 months ago was 480. The CD4 negative slope is pretty fast, and it turns out to be about 320/year.

Moderator: Dr. Bartlett, the results are in:

• About 30% said option 1 (viral load >100,000).

• 41% said option 2 (HIV-associated nephropathy).

• 5% said option 3 (his wife is seronegative).

• About 23% said option 4 (CD4 count done 3 months ago was 480).

Dr. Bartlett: We really split the group didn't we?

Moderator: Yes.

Dr. Bartlett: Let me go over them. I'll tell you what the right answer is according to the guidelines. Listen, believe me, this is a judgment call and there is not a right answer, but the one that's on the guidelines is HIV-associated nephropathy. Everybody with HIV-associated nephropathy ought to be treated regardless of the CD4 count.

Let's look at the others. Viral load of 100,000. We sort of abandoned this as an indicator for early therapy on the basis of the work done from Case Western and a number of other places that have shown that the viral load really has very little correlation with the negative CD4 slope and, therefore, thought that the CD4 count was the main driver of the decision and not the viral load. So that has kind of gone down a lot. It appears that the most important factor in the CD4 slope is really immune activation and right now, the theory I think is that much of that comes from leakage across the gut wall and people are starting to talk about antisepsis agents. That has nothing to do with actually the guidelines per se, except that the part on the viral load probably does. I think it's important to point out that that's not any longer thought to be a major driver, which is a quantum shift from thoughts of several years ago. HIV-associated nephropathy is a clear reason to treat.

Wife is seronegative, so that makes him [part of] a discordant couple. I think everybody that's hearing this would probably agree that having a seropositive partner is probably one of the highest risks that you [could] probably reduce substantially by HAART (highly active antiretroviral therapy). There are three studies -- I don't know if Tom Quinn talked about them when he gave his talk. But there is one published in JAIDS, and two others. They all show that the probability of transmission is reduced by 80% to 98%. Not surprising since the main driver of risk is the viral load.

So my own view is that number 3 is also right. Number 2 is right. Number 1 is maybe very, very slightly right. And then number 4 is the CD4 slope. The reason I didn't like that is because there is so much noise in the CD4 counts. I gave you two values of 480 and 400, and the variation around either one of them is about 30%. So I'm not sure we have any idea whether this person is going down, or how fast this person is going down. Two points on a curve that is so filled with noise probably wouldn't be legitimate.

At any rate, let's go on to one more [question]. [For] this, I will apologize in advance. This is not a bad question, but it is not very well written by me. So I'll take the full responsibility. A patient has a CD4 count of 400/mm³ (just like before), has hepatitis B that is HBsAg positive, HBeAg negative, and [has a] hepatitis B viral load that is 10,000 IU/mL, which is high. What is your best option?

1. ABC/3TC/LPV/r

2. TDF/FTC/EFV

3. 3TC and entecavir

4. Entecavir

5. Peg INF

So those are the options, and the issue here really is this: What do you do with the person that has hepatitis B and HIV and either requires treatment for HIV or [requires treatment for] hepatitis B, or [requires treatment for] both. That's the issue that is being raised here and of course, it has gotten very complicated with the drugs we currently have for both diseases. So let's now vote.

Patrick, are you going to tell me how these come in?

Moderator: It looks like a majority is saying choice 2. We have over 50% that have chosen that with about a quarter choosing choice 1 and quarter choosing choice 3.

Dr. Bartlett: So I think number 2 is the best answer. Number 2 gets us above the threshold to start therapy for HIV infection, but quite frankly, I think people on the committee really had the feeling that we should be probably trying to make it as simple as possible and treat this person for HIV and hepatitis B concurrently. So number 1 is wrong. And it's wrong because 3TC monotherapy is good therapy for HIV and it's bad therapy for hepatitis B coinfected patients because of the high rate of resistance to 3TC by hepatitis B -- not HIV -- but hepatitis B. So 3TC should not be used alone in someone that has chronic hepatitis B. I think that's the point to make regarding this answer being wrong.

The second one provides two agents active against hepatitis B and three agents active against HIV. They're both well recognized good treatments for each disease and quite frankly, this would be the tactic I would take.

I think an obvious question would be what would you do if the CD4 count was 800, not 400? In other words, not even in the neighborhood where people would recommend therapy. And then I think what you would do is you would say we need a treatment that will be directed against hepatitis B without any activity against HIV. Now how can we get there? We can't get it with 1 obviously. We can't get it with 2. We can't get it with 3. Both drugs are active, but it's an inadequate regimen for HIV because it doesn't have enough clout. Entecavir is the new kid on the block for having activity against HIV. This was the study that actually came out of Hopkins showing that entecavir has activity against hepatitis B, as well as HIV. Actually, it was one of those really neat observations where actually a PA who was doing hepatitis B and HIV care, went to the clinic and had a patient that was getting entecavir for hepatitis B, was not treated for HIV and noted that the HIV viral load went down. Of course, entecavir was not supposed to be active against HIV. But it was a significant drop, and then went ahead and subsequently we had Bob Silicano and others measure it in vitro. It had activity. And then we found a patient and subsequently a second patient who had an M184V that emerged during entecavir therapy. So it not only had activity against HIV, but it also brought out an HIV resistance mutation. So entecavir for 3 and entecavir for 4 are wrong. Peg interferon would be the right answer if you wanted to treat for hepatitis B. Telbivudine would be another possible answer. I'm not sure that we would want to use that drug that way. Nevertheless that would be another correct answer; it would treat hepatitis B without touching HIV. So that's how I would read that one.

Here are the recommendations of the DHHS committee. What they have said is you treat for AIDS, you treat for CD4 count that's <350. Now remember what the old one was -- the old recommendations where the data were robust to show that treatment should be given when patients have a CD4 count <200. Then we took the 200 to 350, and we said that's a gray zone, and it should be "considered." A lot of people sort of took that to mean that you could treat it, but you didn't have to treat it, and a lot of people didn't get treated. And I think that's probably, out of all of the changes, that's probably the major one. But I think it reflects the spirit of what's going on in the field and everybody agrees that we ought to be treating more patients earlier in the course, and I guess that's the bottom line of this slide, and it has got perhaps the most important point raised in the revised guidelines and it is reflected in the new European guidelines and virtually all of the guidelines that have been published in 2007.

There are three situations where you would treat HIV with a CD4 count of >350, and that's pregnancy, HIV-associated nephropathy, and patients who require treatment for hepatitis B -- the case I just gave. So I gave you two out of the three that were the exceptions and both were patients that had CD4 counts of 400.

Then there is this sort of waffling statement of you don't necessarily have to wait until it crosses the 350 threshold. You're AIDS providers and therefore, you know the obvious thing here. Nobody is saying if it's 380, don't treat and if it's 340, do treat. You have got to use some judgment here. And there is some noise in this test, as well as in the CD4 count that is just as great. It's a half log in viral load and it's 30% for the CD4 count. So there is a lot of noise in the test we rely on. But anyway, for this one, those are the statements that are made.

It does say that the viral load may influence -- and it says some experts in the field. So there are still people that talk about the viral load as driving this decision. Discordant couples were also mentioned. I think we're waiting for more of the studies. There are a bunch of those discordant couple studies that are being done in Africa, but I don't think anybody thinks that you're not going to prevent HIV transmission with treatment of people regardless of the CD4 count.

Acute HIV infection is another situation where there is no consensus agreement. Nobody knows whether we should be treating or not treating. So those are the guidelines that were just published in terms of when to start.

Now I thought I would say a few things about that decision. Why is there so much weight now given to an earlier start? And I would have to say that these telephone conference calls [with panel members] have 35 people on the line. I would have to say that when people started talking about starting earlier, that the starts are too late, there was a rapid consensus. I don't think there was one person that actually disagreed that we're starting people too late. There is also the thought that there may be a limited CD4 recovery. That's not necessarily etched in stone. There are a few papers on it. There is another paper from Europe that shows that the CD4 count maybe -- when it starts low -- never stops going up, but the incremental gains after four or five years are pretty slow.

The adherence demands are something that we thought were overpowering to a lot of patients that didn't necessarily have to start early. Of course, we've now said that adherence demands are not the 95% rule that we used to quote all the time. We're a long way from that. It's not something that we necessarily want to wear on our sleeve, but something on which I think there is consensus.

The regimens are much easier, they're strong, and they're less toxic. Resistance was always a fear. It continues to be a fear. I put stabilized, maybe. Actually in Europe, it's going down. In the United States, it's going up. But I think people would say that the rate of increase is not as great. Also we're not running out of drugs. We've got a rich pipeline. And the PI (protease inhibitor) data would suggest that the great majority of people who fail PI-based HAART with boosted PIs don't have resistance. So that may be a very important exception to something that we thought was etched in stone five years ago.

Early treatment is supported by large cohort studies. It possibly will prevent transmission, and I've already talked about that. One of the big surprises in the last few years is that the non-AIDS-related complications -- heart disease, renal disease, liver disease -- seem to be progressive and more common in people that aren't treated. The non-OI (opportunistic infection) complications, including malignancy.

I'll start to expand on a couple of these. This is the Egger presentation from CROI (Conference on Retroviruses and Opportunistic Infections), and it shows the median CD4 count in patients that were part of cohort studies. It was 30,000 patients that were in these various cohorts. If you say that you've got to treat before it gets to 200, you can see nobody is really doing that very well, except for maybe Australia. The United States were down. In Africa, they're way down but of course, they're coming up now. But I think the point of the slide is that we're starting and starting too late. We're discovering people too late.

These are data from our own clinic, the Moore Clinic, and it shows the CD4 bounce after initiating therapy out to actually six years. What it shows is that the patients who are seen very late in the course may never get that high CD4 count. Now you know that this is highly individualized -- you'll be able to quote a patient who had a CD4 count of 50 and now has a CD4 count of 800. We all have those. But these are the average, and you can see that they have reached a plateau. I think there are two caveats here. One is that it is inconsistent, that is other studies have not shown such a flat curve after four or five years, but when they don't show the flat curve, they show very small incremental gains. The second caveat is that I'm not sure we know what CD4 count you need in order to be immunologically good or competent. Even the patients that have high CD4 counts may have holes and we all know about these reports of CMV (cytomegalovirus) retinitis with a CD4 count of 500 after somebody has had a good immune reconstitution and that's because of a hole in the stuff that came back. So this was another argument in favor of an earlier start.

Then the big cohort from Europe, the UK Collaborative HIV Cohort with an analysis of 30,000 patient-years, has shown that the risk of going on to AIDS or death is a continuum. And the earlier you start, the better you are. So in other words, you can justify starting at a CD4 count of 650 or higher, something that most people on the call probably would not endorse right now. And to be fair, what I would say is that these differences as you go down these strata that are very high, higher than we're recommending, are statistically significant but they're statistically significant only because the denominator is so huge, the number of observations are so huge, and nobody has shown that this is necessarily cost-effective. So if you start everybody when they have a CD4 count of 650, you can reduce the rate of AIDS and death; however, the number of events is relatively small. All that is to say that these big cohorts can achieve very powerful statistical differences but there is a bit of a problem in the way we interpret those simply because the numbers are so big.

This is the SMART (Strategies for Management of Antiretroviral Therapy) study, and it shows the thing that I think surprised everybody more than any other part of it. We weren't really surprised that progression and death were lower in those that had continuous therapy as compared to those that had discontinued therapy or structured treatment interruption. What was a big surprise was the difference in complications that are often related to the medicines, that is liver disease, renal disease, and cardiac disease, including cardiac disease expressed in the form of a stroke, myocardial infarction, or the need for a bypass surgery. And that was statistically greater in those that had discontinuation of treatment, rather than continuous treatment. And I think that has now all of a sudden attracted a great deal of attention in the field. And that is cardiovascular disease, liver disease and renal disease that are associated with HIV infection per se and untreated HIV infection appear to be a particular risk, or a significant risk.

I picked this slide out. It's not necessarily in the guidelines, but I thought it was one of the more interesting studies that was presented at CROI. It was from the D:A:D. study, which is that huge database that is actually an attempt to look at cardiovascular disease. But here they looked at cancer. And on the left-hand side of the slide, what you see is AIDS-related cancer -- that's Kaposi's sarcoma or non-Hodgkin's lymphoma. That's not going to surprise anybody. What it shows is that the risk of non-Hodgkin's lymphoma or Kaposi's sarcoma is 175 times more common with a CD4 count of less than 50, than it is with a CD4 count of more than 500. Nobody is going to be surprised. The surprise is on the right hand side of the slide, and that is non-AIDS cancers -- that's cancers like lung cancer, breast cancer, stomach cancer, colon cancer -- those kinds of cancers. In there, the risk ratio is somewhere between 15 and 20 with a low CD4 count. So I think the interest here is not only in the importance of treating AIDS to prevent that kind of malignancy. Again the number of events is probably relatively small, but also one of the things that's really interesting about this is it gets the oncologists interested in what we're doing. So new surveillance is now very interesting to cardiologists, to oncologists and so forth.

We also talked about some of the concerns about adherence, and we had the 95% rule, that's the Paterson paper out of Pittsburgh, which said that a patient had to take 95% of their pills or they were going to have virologic failure. And then we had the concern about resistance, the long-term toxicities were unknown, and the benefit of early therapy was not very well established.

The next slide kind of addresses those issues [slide 13]. The Paterson study was done with nelfinavir-based HAART. We don't advocate nelfinavir-based HAART anymore. We all have boosted PIs or an NRTI-based HAART. There the rules for adherence are much, much different. Bangsberg, who has been a good student of this, says that for non-nucleoside reverse transcriptase inhibitors (NNRTIs), that curve starts to go down in terms of continued success or loss of viral potency and resistance when adherence is reduced to 70%, and that accounts for the 70% that's on the slide. I don't want to tell a patient -- I would never tell a patient listen, you take 7 out of 10 of these pills, and you're going to be okay. I would never do that, but I think it does change our mindset in terms of the demands of the regimens. Resistance right now in the United States, the nearest we can tell, is 6% to 16% for recent seroconversion, and it's usually around 15% or 16%. So it's going north. We don't like it, but we would say that it's not going as fast as we thought it would and there are a lot of options left. The long-term toxicity, of course, now we got surprised with lipodystrophy. We sort of knew about the mitochondrial toxicity back in the early 1990s actually with AZT and then d4T. I think what I'm trying to say here is that it's not that we know these drugs are safe, it's the fact that we kind of know what to expect after at least 10 years of using them. And then finally, the benefit of early therapy was not established -- that's still true.

Why does nobody do a study to find out when we would start? The study that was done of when to start therapy was ACTG 019, a famous old study started in 1987. The issue has never been addressed since.

There are a couple of problems and reasons that people don't do this -- i.e., randomize somebody to start when the CD4 count is 500 versus 350 versus 200 or something like that. The reason is that first of all, it would take a huge database. It would be a very expensive study. The second is you would probably have an answer after starting it for 3 to 5 years, however, there is a great deal of doubt that it would necessarily be applicable at the time that the study stopped. In other words, you would have an answer, but you wouldn't be able to apply it by the time you finished the study because we'd have new drugs and new patterns of care and so forth. So we sort of so far counted on the cohort studies to give us that answer. So it's impure science, but in a way, it's the best show in town.

Here is another question. A patient has newly detected HIV, has a CD4 count of 450 and has cryptosporidiosis. When should he start HAART? I'm not sure I like this question, but anyway, let's ask it.

There are only two answers -- so we're making the assumption that you're going to start HAART, and that's why I don't like it. I'm not sure it's clear that this man is not like 95% of the people that have cryptosporidiosis. They don't have AIDS and they get over it on their own. So quite frankly, I'm not sure that this patient should even be treated. So let's get by and say we've decided to treat and the question [becomes: Do we treat] now or two weeks from now?

Patrick, tell me how we're doing.

Moderator: We've got about three-quarters of the audience that are saying two weeks from now and about a quarter who are saying to treat now.

Dr. Bartlett: Now we're going to do the same kind of question. A patient has newly detected HIV -- and this is kind of ridiculous; I apologize -- with a CD4 count of 450 and is incompatible with HIV-associated cryptococcal meningitis. So let's get rid of the CD4 count. Let's not have a CD4 count. We have a patient who has AIDS, has cryptococcal meningitis, has not been treated and the question is: When should he start HAART; now or two weeks from now?

This is a fine question without the CD4 count and the previous one was a good question without the CD4 count.

Patrick, how are we doing?

Moderator: In this case we have about two-thirds saying to treat now and a third saying to wait two weeks.

Dr. Bartlett: Let me tell you what the panel said. We don't really have good data on this. A number of studies are examining the issue, most are about tuberculosis. The whole issue is immune reconstitution syndrome. I'll tell you what the panel said. I don't think anybody has the "right answer." What they said is that if you have an HIV-related complication, and you can't treat it, then you have to [initiate] HAART now, and the examples of that were cryptosporidiosis, microsporidiosis, PML (Progressive multifocal leukoencephalopathy), HIV-associated dementia -- a number of complications for which HAART is probably the major therapy. So in other words, this patient that had cryptosporidiosis. We don't have any way to treat cryptosporidiosis, so let's start HAART.

The second situation is where we have both good therapy for HIV, the cause of the immune deficit that led to the disease, and we have antimicrobial treatment that can treat the complication. And the examples here were MAC (Mycobacterium avium complex), PCP (Pneumocystis carinii pneumonia), and cryptococcal meningitis. And there, the recommendation was for a "short delay," and the reason for the "short delay" is the hope that the anti-infective agent, the anti-pathogen therapy, would get rid of the microbial load enough so that there would not be immune reconstitution. This is in the guidelines.

And I'll tell you what else, the guidelines for this part have to be in sync with the guidelines for the opportunistic infection committee. Now the opportunistic infection committee is run by King Holmes, Henry Masur, Connie Benson, there is somebody else -- but anyway, they're rewriting their guidelines now. One of the things the federal guidelines do is that everything that is in one guideline that is relevant to another guideline is shared with the two committees to make sure that they're in sync. So when I talk to you about hepatitis B and the decisions there, then we talk to the hepatitis B committee on the OI guidelines to make sure that we said something that made sense to them and would support what is in their document.

So this one is commensurate with what we think will be in the OI guidelines, which are probably going to be available early in 2008. It's a massive document and timely because it has been needed to be updated for so long. At any rate, they said what is shown on the slide and that is if you can't treat it, except by getting back the CD4 count, then start treatment as quickly as possible. If there is a major risk of IRIS (immune reconstitution inflammatory syndrome), then it might be good to have a short delay.

That short delay is usually not in numbers, but is usually conceived of to be about two weeks. In other words, this patient with cryptococcal meningitis would have a delay of initiating HAART by two weeks. The patient that had cryptosporidiosis would start HAART now. For tuberculosis, there is a bunch of studies, and they will have an answer to this. One of the best ones being done in the Far East, there is about five of them being done in Africa, and that is nobody wants to start all seven drugs at once. The question is, do you delay for two weeks, do you delay until the first eight weeks of therapy is done -- in other words, you have the induction phase over -- or can you wait longer? A lot of it depends on the CD4 count. So if [the CD4 cell count is] <200, then we go two weeks; if it's more than that, well actually if it is <50, we go two weeks, and if it's 50 to 200, we may wait eight weeks, and if it's >200, then we may wait until they are finished altogether. At any rates, that's kind of a crude summary of where those overlapping guidelines stand.

Now the next question is: A patient has a negative test for HLA-B*5701. What is the safety of giving abacavir in terms of the hypersensitivity reaction? These are judgment calls. Is it completely safe, or is it relatively safe? This is not a silly question. It may sound like a silly question, but the real question is you and I know that when you have a patient that's going to get abacavir, we spend a substantial amount of the office time telling this person about abacavir hypersensitivity reaction and the fact that it can be serious, in fact life-threatening. In fact, there have been 19 deaths as a result of the abacavir hypersensitivity reaction.

So the question is, does the new test make it completely safe or relatively safe?

How are we doing, Patrick?

Moderator: We have 86% that have said it's relatively safe and 13% that say it's completely safe.

Dr. Bartlett: I would go with the majority. I think it's relatively safe, and I'll tell you why.

Let me show the next slide. This is the study that I wanted to mention. The study that was presented at the International AIDS Society meeting in Sydney, Australia, was by Mallal from Sydney, Australia, and he's the one that has sort of discovered this association and gets credit for it and has a strong belief in it. The study was the PREDICT study, and I think most of the people on the line know about it, and what they did was 1,660 patients who were going to get abacavir were tested or not tested, randomized to get tested or not tested. And what they showed is that if you got tested and you were HLA-B*5701-positive, then your risk of getting a hypersensitivity reaction was about 50%, and that could be confirmed with a skin test that's not commercially available. If you had a negative test, the negative predictive value was an amazing 100%. And that sounded like it was essentially a perfect test going in the direction that you would want it to go. In other words, if you had a negative test, you could use abacavir with complete comfort, and all of these were verified or not verified on the basis of a skin test, and that was the definitive test to say whether the genetic test for 5701 was valid or not valid.

So when you went away from PREDICT, it sounded like this was a slam dunk. This was 100%. This is one of the more recent papers, it was just published in AIDS, and here's what it showed. This is their experience with a positive result. They actually gave abacavir to two patients that had a positive 5701, and both of them had hypersensitivity reaction. The more interesting group is the group that was negative. [Some were] negative and got abacavir. They had 151. They had a suspected hypersensitivity reaction in 4. Now you and I know that calling something a hypersensitivity reaction becomes fairly difficult in some cases. However, they did have 1 or 2 that had a positive skin test, and the positive skin test is supposed to mean that the genetic test didn't call it. Now Mallal says he has never heard of that, and he has offered a case of Australian wine to anybody that can find a positive test or a hypersensitivity reaction that is confirmed by a skin test and is negative by the HLA typing. So this would be one where he would have to pay off.

I think the point I would emphasize is this, it's a damn good test, and I think the standard of care means that anybody that gets abacavir in 2007 or 2008 ought to have this test before they get abacavir. And if they have a negative test, that it's almost certain that they will not get a hypersensitivity reaction. However, we can't say never. There are a couple of exceptions of things that impressed scholars in the hypersensitivity reaction that they were real, and this happens to be one of those papers to support that contention. So what the guidelines say is this, they say it's a good test, standard of care, do it. If you don't have access to it, it's all right to give abacavir. However, it becomes the standard of care. Even if you give it in the presence of a negative test, you should warn the patient about hypersensitivity reaction and all that we used to say.

This, by the way, is the PREDICT study, and it shows the confirmed hypersensitivity reaction, those that got the test -- zero. So this was what led them to conclude that it has 100% negative predictive value. For positive predictive value, it was about 50%. In other words, about 50% had that allele but don't get a hypersensitivity reaction even though they have it. I'm not sure we understand that, but quite frankly, the way this test is going is the way you would want it to go. That is, it tells you when you have to be nervous and you may under treat with abacavir in a very small number.

Now the other test that is reviewed by the guidelines is the tropism assay for CCR5, used by maraviroc now as the only CCR5 inhibitor that's marketed. And what they essentially said is this is a test you've got to get if you're going to use a CCR5 inhibitor. It's far more common to have CCR5 exclusively in patients that are treatment naive. But, of course, the paradox is that that's not really a group that are candidates for therapy, at least not now. None of the new treatments were thought to be adequate for initial therapy, but that is not necessarily a done deal at the present time. But for patients who are more in need of salvage and could use maraviroc, the probability that they will have exclusively CCR5 is about 50%.

What is shown down here are kind of the variations in the test, and what it shows is that there is a bit of noise. This is an imperfect test. So maybe 5% of patients will be read as CCR5 only but have X4 either as mixed or dual isolates. What will happen with those patients if they're treated with maraviroc? Of course, we used to worry that that would bring out X4 virus, and that's true, it does. And we thought that bringing out X4 virus would make the disease more progressive, and that turns out to not be right. The more rapid progression seems to not be caused by X4, but is a product of the X4 virus. So we're not doing harm. We're not losing anything as a result of erroneous use of maraviroc, except time.

The resistance testing. What the guidelines have said now is that the patient should get a resistance test as soon as they enter care. So you may see a patient that has got acute HIV infection. Everybody would agree they ought to have a resistance test whether or not they're getting HAART therapy. And you'll see somebody that's early in seroconversion or maybe a year into the disease and has CD4 counts still of 800, you get a resistance test. And the reason is because the further out from the time of transmission and the more we're going to have resistance mutations in the quasispecies, or the minority pool, we would not be able to tell it, and we would like to have that information -- as much information as we can when we go into therapy. So you can repeat it. That's optional. But to get it when it first comes out.

Baseline resistance -- something I talked about; 6% to 16%. Of course, the two that we see the most are M184V and K103. The K103N is the one that we worry about the most because it knocks out the class. In the Rockefeller study in New York City, it showed that in recent seroconversions, the rate of the 103N is now up to about 7% to 10% in newly seroconverted patients.

One of the things that is mentioned in the guidelines -- it's not considered a standard, but it is considered sage advice -- and that is when you start someone on therapy, to have them come back for a viral load test at one to four weeks. It ought to be down by one log. If it isn't, then that's a good time to get resistance testing because of possibly inadequate virologic response in bringing out a resistant strain.

A couple of other guidelines that I thought were interesting that have addressed the issue of what nucleoside pairs to go with. This is the IAS (International AIDS Society) guidelines. They are still 2006. The DHHS guideline still 2006. The European guidelines -- they're new -- that's 2007. Swiss guidelines are 2007. WHO (World Health Organization) is 2006.

Here is what I put on the slide here, and that is what nucleoside pairs are people using [slide 22]. Of course, all of these three are coformulated. They're the lion share of prescribing at the present time. Those are the three that are shown. Naturally, all of the guidelines focus all of their attention, at least in the preferred category, on these three pairs. So I asked USA and DHHS had AZT/3TC and tenofovir/FTC and then IAS USA had abacavir/3TC and the DHHS guidelines didn't because of the hypersensitivity reaction, simple as that. Europe and Swiss added abacavir/3TC -- actually, Swiss have only that as their only option -- and I think that reflects the widespread use of the 5701 allele test that's now very commonly used in most of Europe and Australia. I think the United States is kind of slow in this. WHO is still on AZT and tenofovir as the major choices. A lot of that is driven by price.

In terms of the third drug [slide 23] -- so it's a nucleoside pair and then we've got DHHS guidelines and the IAS USA guidelines. For the NNRTI, we had efavirenz on the basis of better toxicity profile -- maybe a little edge in potency, but pretty close. IAS USA had both in NNRTI. And then the PI pairings, all of them so far have had boosted PIs, only boosted PIs. And the three that are on the DHHS panel are here -- the Gemini study was not available and that, of course, is the flagship for invirase or saquinavir study. IAS did have that one -- that went on -- otherwise they're pretty darn similar.

And then for pregnancy, I mentioned that there is a separate panel that does this. For pregnancy, the recommendation now is AZT/3TC as the nucleoside pair, and you know why -- because it goes back to the 076 study as the drug that has been used the most in pregnancy and has a good track record and huge numbers. And then lopinavir/ritonavir (Kaletra). So that is the choice for pregnancy. And then there are the alternatives -- indinavir is still on the list, which was a little bit of a surprise. Then there is a category called limited data where they just don't have good pharmacology and the big problem is the pharmacokinetics in the third trimester and that's holding back some of these drugs. And then two drugs that are contraindicated; efavirenz, you know about that for teratogenicity and then nelfinavir because of the carcinogen found recently in the capsule which has set it back in a way that we all know. So this is the scorecard right now for pregnancy. That's a separate document. It's a very scholarly, very long document that sort of adds to the DHHS guidelines in terms of having another dimension that covers that.

This is kind of the summary slide [slide 25]. It's the DHHS guidelines. When to start earlier, everybody wants to start earlier. Every guideline in the world that I know of is saying start earlier, WHO included. The new tests are the tropism assay for CCR5 inhibitors and the HLA-B*5701. The two new drugs I didn't talk about; they're spread throughout the guidelines in the updated tables. I always thought the tables were one of the best parts of the DHHS guidelines because they are so useful at the bedside. Resistance test -- get it as soon as possible in untreated patients. And for treated patients, get it while they're on treatment or within four weeks. And we say four weeks because we start to lose mutations at the end of four weeks. And what to start is not in the current document; it's going to be out in 2008.

These are the new guidelines from the European AIDS Society [slide 26]. I thought I would mention these. Notice that the NRTIs are the two that I mentioned. They use AZT/3TC and ddI/3TC as alternatives. The third drug is shown here. I'm not sure why they have atazanavir as an alternative drug but nevertheless, the rest of it sort of resonates with what we've been talking about so far.

That's the last slide, so I'll thank everybody for listening and be glad to answer questions if we have some. I think we're close to the one hour time limit. I'll be glad to answer questions if there are some.

Questions & Answers

Moderator: Thank you. We will now be conducting a question and answer session.

Dr. Bartlett, while we're waiting for questions to come over the phone, we have a couple that came in over the Web. The first is: Do you think the results of the SMART study have effectively killed interest in treatment interruption?

Dr. Bartlett: I think they have. Whether it is a fair killing or not, I don't know. There is still the Staccato study being done in Thailand. The one thing that worried me about the SMART study is, I mean the good news was that it came out with such a clean answer so that their statistics were very powerful in saying you know, it wasn't a close call. You didn't need a statistician to tell you. But the two things that bothered me about it were the fact that they waited until the CD4 count was 250 before they restarted. And most of us said we would restart at 350. So the way most of us did treatment interruption was different than the way they did it in the study. The second thing that bothered me was that they had huge numbers (5,400 people) that participated in that. And of course, you don't have to have a high frequency of an event to sort of drive the statistics so that it can make a very strong statistical statement on the basis of a relatively low frequency event. And what I don't know, and I can't get an answer to is, is continuing therapy in everybody cost-effective? They wound up with patients in the treatment interruption arm that were off of therapy half of the time. In other words, they spent half of their time off of treatment. And I think people are going to start to ask more and more questions and were given therapy that cost $15,000/year and now we're talking about continuing it with uninterrupted for a lifetime. That lifetime duration is starting to look longer and longer and longer. So I think eventually somebody is going to say well you can do that, but it's not going to be cost-effective. But I think for most purposes right now, most of us in the clinic are going to say no more treatment interruption unless they are in some sort of a study.

Moderator: We have a related question regarding the trend to start earlier and whether you think it will continue even possibly to CD4 counts of 400 or above.

Dr. Bartlett: That's a great question. It's strictly conjecture on my part, I expect it probably will. I think we're going to see more and more enthusiasm. You know, when you think about it, what other disease that's serious and potentially a lethal disease do we have in the field of infectious disease where we have good therapy -- really good therapy, and we say we're not going to treat right now, we're going to wait until they have more disease before we start therapy. I mean the precedent just is not there. So I think the answer to the question is we'll probably be treating earlier and earlier. The big question that a lot of people are going to ask, and I think it's a fair question, is: What is the cost effectiveness of treating somebody with a high CD4 count? We can get the numbers. The North American cohort that is under Richard Moore now has 75,000 patients in it. It's huge! And the ability to answer a question in that is just awesome! However, I guess the second part of it is what I raised before, and that is, is it cost-effective to do that?

What else do we have here? We've got some good questions.

Is there any particular reason we've been so slow to adopt HLA testing in the United States?

I don't know why. I thought we should have started doing this two years ago or three years ago. So I'm surprised that the United States didn't start earlier, like the Europeans have. Usually we start things much before Europe does in HIV care. This is one where I think we were slow. However, having said that, I think now when I go around to clinics, I find almost everybody is starting to do this and do this quite frequently. It's commercially available, costs about $100, takes about 10 days or two weeks to come back. It comes back as a yes/no answer. It's a very good test for what it is intended to do.

For those who are diagnosed with a questionable history of abacavir hypersensitivity, does emergence of a new genetic test allow us to revisit?

Yeah, I think it does. One of the interesting things was that in the ACTG trial that had abacavir/3TC/AZT as one regimen and AZT/3TC as another, it turned out that the frequency of abacavir hypersensitivity was the same in the group that didn't get abacavir as it was in the group that did get abacavir. And that's not going to surprise people that do this work. And I think what it shows is that yeah, we see a lot of bad things, and we're very sensitized to abacavir hypersensitivity reaction, and I think this question is a really good one. If you have a patient that had hypersensitivity call and that doesn't necessarily feel real good, and you feel like you really need abacavir, I think I would get the test and then give that under observation. That's what we always did before when we had these confusing patients. One thing you can say about abacavir hypersensitivity reaction every time you gave the drug, they got a worse reaction and it was progressive, so that if they didn't have a reaction, you would have to say you don't have it.

How is maraviroc going to be used when most clinics can't afford the tropism test?

Very good question. The sticker price on the test is $1,960. It's tough for a lot of places. There are some ADAPs (AIDS Drug Assistance Programs) that have agreed to buy it. The one in my state (Maryland) has not agreed to buy it. They don't want to offer that test. They say that ADAPs are for drugs not tests. So some places will have an ADAP that will do it, some third parties will do it, some won't. I think if you can't get the test, you can't use the drug. I think it's as simple as that. And it is quite checkered as to who will be willing to pay for the test. I expect that six months from now it will all be worked out. But right now we're in the transition zone. Only two labs in the world can do it. The one that we usually use is the test out of California. The cost is about $1,960 as I mentioned. It does have a bit of noise in it that I mentioned, but it's not hard to interpret the results. I mean it comes back R5 or X4 R5 and it's not hard to read their answer. What is a bit of a problem is a little bit of false-positives for R5 only, not the other way.

Are there going to be dramatic changes in the guidelines regarding what to start with?

I don't know what the panel will do. I can't answer that. I personally wanted to show the European guidelines because I think that's sort of the reflection of the field, [but] that doesn't necessarily mean the DHHS guidelines are going to be the same. But I think people are excited about some of the new drugs. I think the new NNRTIs that are probably going to get marketed next year -- raltegravir and maraviroc are generating a fair amount of interest. Darunavir did awfully well in the ARTEMIS study against lopinavir. There may be some modest changes. But when you come to new drugs, a lot of people have said it's got to be a big victory in order to get me, say Dr. Joe, to want to use them in first-line therapy because what we've got now is working awfully well and we've had these drugs for 5 to 10 years, and we know all of their pocks. If you ask me right now, I would guess that there is not going to be a lot of big changes, but there will probably be changes.

I noticed in reading the new guidelines there is nothing regarding treating lipoatrophy.

I think we've said all we can say. That part of the document, by the way, is going to be done over again. So that part has not been revised, and that is the part about drug toxicity. But if you think about what's going on in the field, I mean, except for avoiding the drugs that cause lipoatrophy, which is something that we can talk about, for people that already have it, I think we know that stopping d4T or ddI or AZT, if they are responsible, is slow and painful -- painful only because it is so slow. And I expect that that's going to the major change. And then the other is going to be plastic surgery. I'm just reflecting our own practice in the Moore Clinic, not anything on behalf of the DHHS panel. And that's pricey. Our insurance companies will not pay for an iatrogenic complication. So plastic surgery is a bill. I don't know what it's costing in the rest of the world, but for us, it's about $2,000 an injection and most of the patients require two to five injections. It's very pricey.

I noticed in the guidelines there is no upper limit regarding what CD4 count to start out with discordant couples. Is this to get their viral load undetectable to prevent transmission?

Yeah, I think that's exactly it. I think the goal with the discordant couple is to protect the uninfected partner. And I think someone on the line could say listen, you don't have to expose somebody to all that high-powered therapy with the risks and so forth if somebody would just use condoms. And I think that's a very valid criticism. But nobody has had very good luck in getting uniform compliance with condoms. You find individual exceptions, but you don't find population-based use of condoms on a regular basis. So for some people, it will be important (that is early therapy), for other people, it will not. I would also emphasize the fact that the guidelines did not come down strongly on that point, they mention it, as one of the issues to think about in the person that has a CD4 count above the threshold where we start because of the threshold.

I think we've answered all these questions; are there any on the phone?

Moderator: No, doctor, there are currently no questions on the audio portion.

Dr. Bartlett: Okay. Thanks.

Moderator: In order to receive continuing education credit for this activity, participants must complete the program evaluation and post-test. Do you have any closing comments, doctor?

Dr. Bartlett: No, but thanks everybody for listening. The questions were great. I hope the presentation was clear.

Moderator: Thank you everyone for your participation.

This presentation was also webcast on Dec. 13, 2007 and Dec. 17, 2007.