November 3, 2008
Listen to Audio (58 min.)
Please note: These files can be quite large. Allow some time for them to download.
Table of Contents
Note: This CME/CE activity expired on Nov. 3, 2009. For a list of currently available activities, click here.
|Benjamin Young, M.D., Ph.D.|
Bonnie Goldman: And this is Bonnie Goldman, Editorial Director of TheBodyPRO.com.
Ben Young: We're here at the 2008 ICAAC/IDSA meeting in Washington, D.C. I'd like to spend the next few minutes going over some of the key concepts that emerged from important presentations at this meeting. There were many themes discussed at ICAAC/IDSA 2008. I'll focus on some key areas.
This was a very big meeting, with many attendees. I found several areas of focus that brought together consensus, as well as areas of controversy. Let me give you some of those themes first, and then I'll talk about individual studies.
The first theme centered on a series of presentations that looked at the effectiveness of highly active antiretroviral therapy [HAART] in the parts of the world where people do have access to lifesaving therapies.1-3
There was a second group of studies presented that helped us define when to use therapies in specific clinical situations, namely, in patients who have active opportunistic infections [OIs].4,5 And there was a very fascinating presentation that readdressed the question about when specifically people with higher CD4+ cell counts should start therapy.6
A subject area that seems to be emerging, and which has been previously underappreciated at these meetings, looked at special populations of patients (e.g., women of color in the United States).7-9 I'll talk about a number of studies that addressed the disparity in access to care, as well as some of the important and, frankly, devastating complications of poor access to care.
Lastly, I'd like to discuss a theme that's very important to me in the clinical management of patients in my practice, and that is: How does one take care of older patients who are diagnosed with HIV, or HIV-infected patients who have survived into their fifth and sixth decades, with often many years of HIV therapy? This gets to the point on the emerging issues of primary care in HIV medicine.10-13
Bonnie Goldman: It sounds like you're going to cover a nice assortment of materials. Let's go back now and discuss the fact that HAART works. What were the studies that popped out at you in this category?
Ben Young: There were three major studies that to me highlight this point, and really provide an exclamation point to the fact that highly active therapies continue to work.
Ben Young: It's now been 11 years since the first published studies showed that we could suppress HIV. It was a study by Frank Palella and our colleagues at the CDC [U.S. Centers for Disease Control and Prevention] that showed, in 1998, that patients who received antiretroviral therapy actually had decreased rates of death and decreased rates of complications.14
At ICAAC/IDSA 2008, the CDC and John Brooks presented an update of this same cohort -- 10 years after the original publication appeared in The New England Journal of Medicine.1 In this study and presentation, Dr. Brooks showed us an analysis of the 7,800 patients who continue onward in the HOPS cohort. What he showed very clearly was that the death rate remains low in this cohort of patients.
Dr. Brooks also showed us that the incidence of high frequency infectious opportunistic infections in the HOPS cohort decreased in 1995 and 1996, and remained suppressed through the 2006 analysis point. So the rates of cytomegalovirus infections, Pneumocystis pneumonia, MAC [Mycobacterium avium complex] infections, etc., have remained low.
This was an important observation, because there was concern that highly effective therapies would be effective for a while, but then they would stop being effective. So, even in this population of patients -- with its issues of drug resistance and non-adherence -- therapies continue to work into the second decade of highly active regimens.
Bonnie Goldman: This reflects the reality that most clinicians are seeing in their practice.
Ben Young: Yes, and in fact, it's very much the case in my practice.
In the same presentation, Dr. Brooks showed us data about the frequency of malignant complications of HIV -- namely, Kaposi's sarcoma, non-Hodgkin's lymphoma, cervical cancer and CNS [central nervous system] lymphoma. The data showed persistent decreases in the frequency of these events; again, suggestive of the long-term benefit to patients who receive therapy.
An interesting analysis of this study looked at the CD4+ cell count at the time of the presentation of these opportunistic complications. What was a very interesting point here was that there seems to be a change in the CD4+ cell count at time of presentation of certain complications. So while the median CD4+ cell count at the time of diagnosis for cytomegalovirus and Pneumocystis remains the same through this chronological period, the CD4+ cell counts that we're seeing for non-Hodgkin's lymphoma, for Kaposi's sarcoma and for HIV encephalopathy actually increased over time. This suggests that there could be a changing natural history of these diseases in the U.S. population.
Bonnie Goldman: Can you give some examples?
Ben Young: For example, in the early HAART period, the median CD4+ cell count at the time of diagnosis of a Kaposi's sarcoma case patient was 38 cells/mm3. In 2001, it was 107 cells/mm3. In the 2002 to 2006 timeframe, it was 143 cells/mm3. So we have seen about a 100-cell increase in the median CD4+ cell count at the time of Kaposi's presentation. This compares to similar changes we've seen in CD4+ cell count at presentation with non-Hodgkin's lymphoma -- initially starting with a median CD4+ cell count of 73 cells/mm3, and most recently, with a median CD4+ cell count of 243 cells/mm3.
Bonnie Goldman: Do we understand the mechanisms at work here?
Ben Young: We don't. I think, again, it speaks to the unmasking of new complications, an evolving natural history of HIV disease and an evolving natural history of the management of HIV-infected patients. I'll actually talk more about this later in our discussion.
Ben Young: Yes, absolutely. Both of these studies are important. To put this into context, in the early days of highly active therapies, it was easy to show survival benefit because there were so many patients dying of AIDS complications -- anything that improved that death rate could be measured in prospective clinical studies. As we saw the frequency of death from AIDS complications decrease, it became increasingly difficult to demonstrate a survival benefit in prospective clinical studies. For the last 10 years, we really have not been able to demonstrate that critically important point in clinical studies. Here, at this meeting, we found two studies that actually demonstrated this, or showed us some signal of this happening.
Ben Young: The first of these studies is an analysis of the DUET study, presented by Richard Haubrich from the University of California in San Diego.2 In DUET15,16 patients received either etravirine [TMC125, Intelence] or placebo, and all patients received ritonavir [RTV, Norvir]-boosted darunavir [TMC114, Prezista] as part of their optimized background therapy.
To be eligible for DUET, all the patients had to be very highly drug resistant, with very limited treatment options. This represents the patient population that might be at very high risk of having an opportunistic infection.
In brief -- and I'll go through this in a little greater detail in a moment -- the study shows that there was a statistically significant survival benefit and a statistically significant benefit in preventing AIDS complications with therapy compared to placebo.
There were approximately 600 patients in each arm of the DUET study. In the overall patient population, almost 10% of placebo-receiving patients developed an AIDS-defining event or death in the 96 weeks of the study.
In contrast, only 5.8% of patients who were receiving etravirine had an AIDS-defining event or death. Again, it's a statistically significant improvement in the risk of having something very bad happen to patients. This is consistent with the idea that if you give people effective therapies, you can actually reduce their risk of serious complications and reduce their risk of death.
Bonnie Goldman: And that's because their CD4+ cell counts go up?
Ben Young: Presumably. This analysis doesn't control for CD4+ cell count; it doesn't even control for virologic response. It simply says that you can prevent complications.
There were several really important, and I think corroborative, analyses in Haubrich's presentation. The first showed that there was a strongly statistically significant difference in the time to having an adverse event or death. There was a decreased proportion of patients who were hospitalized in the study after the 48 weeks. In the placebo arm of the study, 23% of patients were hospitalized, versus 17% in the treatment arm.
Lastly, there was an analysis that looked at the cumulative number of hospital days over the 48 weeks of the study. It showed a strong, significant difference in the number of hospitalized days. Remember there were 600 patients in each arm; 2,700 hospital days were experienced by those in the placebo arm of the study. By contrast, 1,700 hospital days were spent by the etravirine-receiving patients.
To recap, the study showed a significant reduction in AIDS-related complications or death in the patients who received etravirine plus an optimized background regimen, when compared to patients who received placebo plus an optimized background regimen. There were fewer cumulative hospital days, and the time to having a complication was significantly prolonged if you received active treatment. These together suggest to me that if you give people active therapy, you will keep them out of the hospital, you will keep them healthier longer and you will prevent them from getting sick and dying.
Bonnie Goldman: We knew that before, based on experience in the clinic, but this actually demonstrates it for anyone who doesn't get to see what the reality is for most people in clinics.
Ben Young: I think that you're correct in that this is an observation that has been frequently made in parts of the world that have access to medicines, where doctors and the health care systems understand the value of antiretroviral therapies.
Bonnie Goldman: Right. I guess what I was trying to get at is, DUET was the registrational study for etravirine, and the drug was approved because patients' viral loads were suppressed;15,16 so this study seems like an afterthought, demonstrating an obvious point.
Ben Young: This was a secondary analysis in the study,2 for sure, because the study was designed with the intent of demonstrating that etravirine actually was able to lower viral loads and raise CD4+ cell counts. It's very different to say in a prospective clinical study that you can actually prevent AIDS, or you can prevent death. That's a much more powerful statement, a statement that has not been made in many recent clinical studies. So I think that this is a very important study. Haubrich didn't get enough credit for a presentation that I think makes an important point.
Bonnie Goldman: Was the raltegravir [MK-0518, Isentress] study similar?3
Ben Young: Yes. Joe Eron presented a follow-up analysis of the BENCHMRK 1 and 2 clinical studies. To remind you: BENCHMRK 1 and 2 were studies that gave highly treatment-experienced patients either raltegravir or placebo, plus an optimized background regimen.17
Dr. Eron showed in this analysis that there was a difference in the time it took for a patient to experience his or her first AIDS-defining complication, as well as a difference in the time to death, and a difference in the time to the first AIDS-defining complication or death, as an aggregate endpoint.
Overall, in this study, there were 37 patients in both arms who had a clinical endpoint of either an AIDS-defining complication or death; 660 of the trial participants did not. So not many patients had a clinical endpoint. In the end analysis, the differences between the two treatment arms did not meet the full criteria for statistical significance.
From a management perspective, it's reassuring that when we see new drugs and new drug classes come forward, not only do they work in reducing the surrogate markers that we always follow in the clinic, but they also actually do work in the larger sense of our main objective, which is improving life and preventing death.
Bonnie Goldman: It should be underlined that both DUET15,16 and BENCHMRK17 were in highly treatment-experienced patients. Looking at the BENCHMRK data for the people who did get sick, their average CD4+ cell count was 50 cells/mm3.
Ben Young: In fact, in the absence of these kinds of very aggressive diagnostic testing and access to therapies, many of these patients would have otherwise gone on to die.
I actually have very few patients -- I think there are only three patients in my practice now -- who have detectable virus while on HIV treatment.
Bonnie Goldman: So, a triumph of HIV medicine.
Ben Young: Absolutely. It is a triumph for HIV medicine and drug discovery. It is the biggest single improvement in the prognosis of a population with a life-threatening illness that we've seen in modern medical history.
Bonnie Goldman: Let's switch to the question of the year: When is the optimal time to initiate therapy?
Ben Young: Now that we've established and reestablished that HIV therapies work, the question is: Who gets those therapies, and when do you use them? I should certainly point out that this discussion is really about those patients who have access to medications. It would be, frankly, irresponsible to ignore the fact that the majority of HIV-infected people in the world who desperately need HIV medicines still, today, in 2008, don't have access to them, and as a result of that, millions of them are suffering and dying.18
Ben Young: With that gloomy backdrop, though, let's talk about what's going on here in the United States and Western Europe. I'd like to start the discussion with a presentation that was a review.4 Andrew Zolopa, from Stanford University, recapitulated the results of the AIDS Clinical Trials Group [ACTG] A5164 study.19
For many, many years, it's been assumed that patients who had active opportunistic infections were too sick to receive antiretroviral therapies. In fact, it was codified in older U.S. treatment guidelines20 and older international guidelines21 that you would initiate ART [antiretroviral therapy] only after patients had recovered from, or at least completed therapy for, their opportunistic infection. The question about starting antiretroviral therapy at this point, though, was until recently neither fully understood nor fully evaluated.
In ACTG A5164, patients were randomized to either receive early or immediate antiretroviral therapy during the time when they were receiving acute treatment for the opportunistic infection.19 Alternatively, antiretroviral therapy was deferred until after the completion of treatment for that acute opportunistic infection.
Overall, just under 300 patients were enrolled in the study. Again, they were randomized to receive either immediate or deferred treatment. Most of the patients in the study had Pneumocystis pneumonia. There were a few patients who had bacterial infections and other major AIDS-related complications, such as cryptococcal meningitis, histoplasmosis and so on.
It's important to note that these patients weren't just a little sick; they were really sick. Within 30 days of entry into the study, about a third of the patients had a second or multiple opportunistic infections. The baseline CD4+ cell count of these patients was just around 30 cells/mm3 -- so, again, a very sick patient population -- the very population of patients, actually, for whom this question of when to start antiretroviral treatment is a really urgent one.
Overall, the study team worked really well. Patients in the immediate group received antiretroviral therapy within 12 days of enrollment in the study. Patients who deferred treatment started therapy about a month later, at day 45. About 85% of the patients received a boosted protease inhibitor during this phase and 15% received a non-nucleoside-based therapy.
The primary analysis of the patients looked for the time to AIDS progression or death, and showed very strongly a statistically significant difference between these events, where patients who had immediate initiation of antiretroviral therapy experienced a much better prognosis, with decreased progression to AIDS, to a new AIDS event or to death.
Bonnie Goldman: A powerful message.
Ben Young: Absolutely. It flies in the face of the idea that these patients were so sick that they would have more complications from HIV medications, that they would have immune reconstitution reactions, and so on. In fact, the likelihood of having an IRIS [immune reconstitution inflammatory syndrome] event was the same in the immediate antiretroviral treatment group versus the deferred group.
In this country, where patients have this kind of distribution of complications -- mostly Pneumocystis, and not a lot of tuberculosis [TB] -- this strategy seems to have become very well-established.
Bonnie Goldman: You should mention that TB is a different story.
Ben Young: TB may be a different story. We actually don't know this. We know that TB immune reconstitution reactions are quite common and can certainly be very devastating,22 as I have seen in my travels to Eastern Europe. This is the number one most common complication23 and is a frequent cause of mortality among patients receiving therapy in Eastern Europe24 and Africa.25
There have been a series of smaller studies that have tried to address this question in populations of patients who have more tuberculosis. The data from those studies, while smaller in sample size and power, are consistent with this observation that the fear of IRIS and the fear of the other management complications that come along in these patients should not be a reason to withhold antiretroviral therapies.26,27
What this means for those patients who are in the hospital with severe Pneumocystis pneumonia, the patients that we used to wait to start on therapy until they finished their 21st day of treatment for Pneumocystis, these are now the patients that we initiate therapy in as soon as it's reasonable to do so. A5164 is one of the studies this year that has fundamentally changed the way I take care of patients in my practice.19 There aren't too many studies that actually do that, so this was one. It was good to see this.
The U.S. treatment guidelines,28 the British treatment guidelines29 and the International AIDS Society-USA treatment guidelines30 all now recommend the initiation of therapy for patients -- with the exception of patients who have tuberculosis, given the absence of data -- as soon as possible.
Ben Young: There was an instructive study that comes from Stephen Berry from Johns Hopkins University.5 This study looked at the risk of hospitalization as a function of time after starting antiretroviral therapy. This was a cohort of patients from the Johns Hopkins HIV group. Overall, there were about 1,200 patients in this analysis. So this was a large study from a very credible group.
The study makes two simple points. First, if patients did not have a response to highly active HIV therapy, then they continued to be at risk of hospitalization, presumably because of complications. That risk of complications was the same over the first year after the initiation of therapy.
By contrast, in the majority of patients who responded to therapy, the risk of having a hospitalization event continued for the first 45 days (the first month and a half of treatment), but thereafter, it dropped dramatically and was about half the risk in the following period of time.
Clinically, what that means to me is that we should keep a close eye on patients for the first month or two after they start therapy. Thereafter, the risk of hospitalization goes down. It reiterates the idea that patients remain at risk, not just for the first day after they start therapy, but for some period of time.
Bonnie Goldman: Could you describe some of the hospitalizations or complications seen during the first 45 days?
Ben Young: The group actually didn't elaborate in any detail regarding the kind of hospitalizations that occurred. It would be fair to assume that, in a population of patients whose average CD4+ cell count at the time of treatment initiation was 150 cells/mm3, a significant chunk of them were probably AIDS-related, or at least HIV-associated symptomatic, complications. But what the researchers did elaborate on was also an important point. This is a preamble to our later discussions on access to care. Certain patient groups were at significantly greater risk of being hospitalized, irrespective of time. Women in this study were about 40% more likely to be hospitalized during their first year of treatment. African Americans were 86% more likely to be hospitalized, more so even than injection drug users, who were about 56% more likely among responders to be hospitalized.
This means that, as doctors and as health care systems, we need to much more adequately address and figure out systems that get women into better care, African Americans into better care, injection drug users into better care and so on. It's a matter of trying to identify the patients who are at greater risk of having complications.
Bonnie Goldman: And then there was the most talked about study at the conference.
Ben Young: Right. This is the study from the North American AIDS Cohort Collaboration on Research and Design group, otherwise known as North American ACCORD [NA-ACCORD], which was presented by Dr. Mari Kitahata from the University of Washington.6
This study was a group of cohort studies from North America, and represented an extremely large population of patients in observational databases. The study, with its large sample size, asked: What is the optimal time to initiate antiretroviral therapy?
Over the years, we watched the recommended thresholds for starting therapy go from 500 cells/mm3 down to 350 cells/mm3,31 down to 200 cells/mm3,32 and then back up to over 350 cells/mm3 in the most recent version of the U.S. guidelines.28
The real question here is: Are we sure that 350 cells/mm3 is the optimal time to start therapy, given today's therapies and today's patients? Or is there a redefinition of that?
The group was interested in testing the hypothesis that starting treatment earlier might have a survival benefit or disease morbidity benefit. So North American ACCORD looked at all HIV-infected individuals in their cohorts with both a CD4+ cell count between 350 cells/mm3 and 500 cells/mm3 and an active follow-up for a 10-year period of time between 1996 and 2000. This is the kind of study that previously was very difficult to do, because you couldn't find enough of these patients in one large study.
The groups were compared then based on whether they started therapy immediately -- defined as starting therapy within a year and a half of the first CD4+ cell count in that window -- or deferred therapy, meaning they didn't begin treatment until their CD4+ cell counts fell below 350 cells/mm3.
As I alluded to, this was an immense study. There were overall almost 8,400 study patients that were followed for almost 25,000 patient-years of follow-up -- the largest cohort study to be conducted in North America.
The patient demographics are probably not surprising for a North American population: about 80% male, a young to middle-aged population, with a median age of around 40, and only 40% of the population was Caucasian.
The median CD4+ cell count of the patients when HAART was initiated follows the study definition. Patients who started therapy immediately had a median CD4+ cell count of 420 cells/mm3. Those who deferred therapy had a lower count, of course, and that count on median was 275 cells/mm3 -- a count, by the way, that is higher than what we'd see in many patients who present for therapy for the first time,33 but nevertheless a large enough group to help us answer this question.
Reflecting the time capture in the study, from 1996 to 2006, many of the patients started on therapy: Approximately 40% started on non-boosted protease inhibitors, as was done at the time; only around 10% initiated therapy with a boosted protease inhibitor; and about 35% started on a non-nucleoside-based therapy.
Again, this study looks at this question of whether to start or defer therapy in patients with CD4+ cell counts between 350 cells/mm3 and 500 cells/mm3. The endpoint of the study wasn't a soft endpoint; it was an absolute endpoint of mortality. So this is really looking at survival.
The results were, I think, very compelling. If patients deferred therapy -- as defined in this group -- the relative hazard of having a mortality outcome was increased by 70% -- strongly statistically significant in a cohort of nearly 8,400 patients.
An increased risk of death was also associated with being older, perhaps not surprising. Reassuringly, being female and having a lower baseline CD4+ cell count was not associated with an increased risk of death. The major point here was that waiting to start therapy, when someone's counts are in a range above where we currently recommend therapy, was associated with an increased risk of death.
Bonnie Goldman: Were you surprised by that number?
Ben Young: Personally, no. I think that the reason that we tend to wait to start antiretroviral therapy is based mostly on the older generation of HIV drugs that had a lot of toxicity, and a lot of risk for patient non-adherence.
We often talk about the idea that HIV is one of the few diseases that we wait for the disease to advance before we begin therapy. It would be like waiting for the tumor to metastasize before we start chemotherapy for cancer. That doesn't make any sense, and it doesn't make any sense for HIV. The only difference is the need to balance the risks of therapy with the risk of the disease.
Now, with newer therapies and a larger patient cohort -- where we have the opportunity to ask this question with some precision -- we can demonstrate that when CD4+ cell counts are below normal you can bring them back, presumably, to a more normal range by treating the HIV and thereby decrease the risk of even subtle HIV complications and, in this case, the hard endpoint of death.
Bonnie Goldman: So it's redefining what a healthy immune system looks like, and 350 cells/mm3 is not it.
Ben Young: Well, 350 cells/mm3 is the lower limit of normal, which means that only 5% of people have normal CD4+ cell counts below that. But that doesn't mean that's the average CD4+ cell count; that means it's the lower limit of normal. The difference is that in previous studies we could only show benefit of initiating therapy when the CD4+ cell counts were below 350 cells/mm3.34-36 Those studies were largely compromised in their strength because of their size. This is a study that, because of its sample size, and because of its analytical methodology, I think, really deserves a lot of attention. I think this was potentially one of the most important studies of the meeting.
Bonnie Goldman: Do you think they are going to do a randomized trial anytime soon? I know there is the START trial, where they want to look at when to start in a different kind of way. Or is NA-ACCORD the study that will convince people that patients should be on therapy earlier?
Ben Young: I think you raise some really important questions. The START study, hopefully, will be a well-powered study that will answer this question in a prospective fashion. It's really a question of what the omission of treatment might lead to. Delaying therapy increases the risks associated with HIV. Starting therapy too early adds a risk of toxicity from therapies. It's a matter of balancing those risks, not just in a single individual but in large populations.
It will take us a long time to get this question answered in a rigorous scientific fashion. Personally, what this means for me, if I were a patient or if I was counseling a friend or a family member, is that there is clear guidance that when CD4+ cell counts are below 350 cells/mm3 in asymptomatic patients, antiretroviral therapy should be started, and that there is now compelling, though perhaps not confirmed, data that there may be a benefit to starting sooner.
It gets back to some of the older discussions around the question: Shouldn't you just start therapy in patients who are willing to be adherent, who understand the risks and benefits of therapy, and who are properly monitored by medical professionals who have been well-trained? I think that the risks in those patients of starting treatment earlier today, in 2008, are much less than they were even five years ago, given the kinds of treatment options that we have and the better understanding about management. If it were me, short answer: I'd probably start sooner, and probably start somewhere around 500 cells/mm3.
Bonnie Goldman: Until the next study that shows that you have to start at 800 cells/mm3, or some other impossibly high number, considering that there are no patients who walk into a clinic with a very, very high T cell count -- I mean, it's unusual.
Ben Young: It's unusual, but certainly not unheard of. We focus a lot of attention on the minority of patients who present with CD4+ cell counts of 50 cells/mm3. I would point out that there are patients who present to my practice with CD4+ cell counts of 700 cells/mm3, and this study provides some guidance for how to counsel those patients.
It says that it may not be so bad to start sooner rather than later, for patients who are really concerned about that. As we start to talk about the depletion of gastrointestinal immunity37 and long-term complications related to HIV viremia (whether that's cardiovascular disease, neurocognitive disease, etc.), I'm not willing to say that earlier treatment is such a bad thing.
Don't forget that malignancies occur because of sustained immunodeficiency. So anything that decreases the total time that a patient has decreased immunodeficiency makes sense to me, at least at first principle. It's just a matter of balancing that theoretical benefit with the theoretical risk of toxicity and complications. Until we have defining clinical studies, we have to use clinical judgment, best intellect and discussion that's individualized to the needs of one patient at a time.
Bonnie Goldman: In fact, the guidelines don't recommend against starting treatment at any T-cell count.28
Ben Young: Absolutely.
Bonnie Goldman: They say it depends on the discussion between a patient and his or her doctor.
Ben Young: Exactly. Often, most of us who are busy clinicians will just look for the summary tables, but the summary tables only give you the bare minimum. It's in the details of the text of the treatment guidelines where you really find out about management at the individual level. And it is an individualized decision. There are plenty of circumstances where the guidelines fully endorse the initiation of therapy in patients with very high CD4+ cell counts -- namely, pregnancy and after high-risk occupational exposures -- and so, in a sense, it doesn't say that you shouldn't treat people; it's just a matter of balancing those risks and allowing patients to make those decisions in an educated fashion.
We have seen in these three studies further refinement in the discussion around when is the best time to start, what kinds of conditions might influence those starting factors, and so on. I think these represent significant improvements and advances in the way we think about this very important question.
Ben Young: Let's focus attention now on some of the emerging, or underappreciated, areas of HIV clinical management. Among these many areas, the ones that I found most interesting at ICAAC/IDSA 2008 were discussions on:
Bonnie Goldman: Let's look first at the issue of women and HIV. I'm gratified to see a growing number of studies presented about women. Five years ago there was hardly anything, and now there are more and more interesting studies.
Ben Young: Yes, the increased number of studies about women and HIV are reflective of the changing demographics of HIV here in the United States, as well as in Western Europe. Of course, it's well-recognized that the HIV epidemic in many parts of the world affects women to a large extent -- in fact, most populations of HIV-infected patients in the world have a roughly 50-50 gender distribution.38 It's really the North American39 and Western European epidemics40 that have this gender inequity.
Let's talk about some of the presentations here that focused on issues related to health care discrepancies and gender inequalities, as they relate to women in the United States.
The first poster presentation I'd like to review is one from Sally Hodder at the University of Medicine and Dentistry in New Jersey, in collaboration with investigators from across the United States.7 This study was a survey of about 700 HIV-infected women who were recruited through a national network of AIDS services organizations.
What the study showed was that it was not uncommon for HIV-infected women in the United States to feel that there were significant cultural, ethnic or language barriers between them and their health care providers. This difference was even further heightened among women of color.
African-American women or Latinas were more likely than Caucasian women to feel that culture, ethnicity or language impacted their health care. This problem was the least common in the West. It was more common among Latinas and African-American females who lived in the Northeast, South and Midwest. Nevertheless, overall, about 40% of women in the study said that these cultural, ethnic and language issues impacted their health care.
Bonnie Goldman: I thought it interesting that this is all in the HIV world. I think that if we looked at disparities in the same population in the non-HIV world, we'd probably find the same thing.
Ben Young: I think you raise a really excellent point, Bonnie; this study only looked at HIV-infected women. If you asked this question, my assumption would be that, yes, you're right that similar issues and problems arise, irrespective of HIV status. In fact, I think the interesting study would be to compare HIV-infected women to HIV-uninfected women with regard to their perceptions of health care access.
Bonnie Goldman: The question that then comes to mind is: Is it a function of the stigma of HIV that practitioners don't have an understanding of HIV-infected women? Or is there still a great divide between health care providers and their patients of color, who might also be poor?
Ben Young: I don't think the study answers that question, but I think it highlights that there probably is a problem. It's just a question of how much it influences any particular circumstance. Perhaps related to that question, though, is a second analysis that looked at the likelihood of a woman reporting that there were these problems of culture, language and such, but stratified now by the type of doctor that the patient called her care provider. Namely, was the care provider an infectious disease specialist or a general practitioner?
What the study showed was that there was a statistically significant improvement in the likelihood of these so-called bad feelings if the provider was an infectious disease specialist -- namely, an HIV specialist. That suggests that HIV specialists are a little better at addressing these concerns than the general practitioner who may see fewer patients with HIV, or fewer patients with the same demographical makeup of women of color living with HIV.
Perhaps an extension of Sally Hodder's presentation that said that there were some challenges to the way women access care, and how doctors talk and interact with women, was a disturbing study that came from a collaboration between Marsha Lillie-Blanton and Valerie Stone at George Washington University and the Massachusetts General Hospital, with other collaborators.8 What this study did was look at the influence of race, drug use and health insurance on the receipt of highly active therapy in HIV-infected women. Again, the effect of these demographical factors on whether the doctor actually prescribes treatment.
In short, what they showed was that, yes, it did matter. Women of color were less likely to get therapy than those who were Caucasian. Women who were using drugs were much less likely to receive therapy than those who were not.
The analysis looked at 445 patients in 2002 and 390 women in 2005. To me a very depressing statistic was the proportion of women, who were otherwise eligible to receive treatment, who were not receiving treatment. Thirty percent of women who should have been on therapy in 2002 were not receiving therapy, and essentially the same number in 2005. So, nearly a third of women who might otherwise be receiving therapy, for whatever reasons -- doctors' perceptions, patients' unwillingness to take medicines, access care issues and things like that -- still don't receive therapy.
These are the very patients that we just talked about earlier that at the least should be on therapy.6 They would benefit from therapy from not just a quality-of-life standpoint, but actually from a mortality standpoint. We need to do a better job here.
Dr. Stone concluded that there were persistent and substantial disparities in the use and access of antiretroviral therapy among racial and ethnic minority women, compared to white women, and this to me is a real area where we must do better. We must understand the basis of it. That's a systematic challenge, but as individual doctors in the clinic, it really says that we should look carefully at ourselves when we are offering therapy and counseling people of color and counseling women about the need to be on therapy.
Ben Young: Right, and I'm sad to say that I'm not surprised. My colleagues and I work very hard to try to provide the most egalitarian and the least gender-stigmatizing or discriminatory environment in our clinics. I don't think that the clinics that were involved in this study feel any differently. But even in those kinds of university environments, where we have women investigators trying to find the answers to the HIV care issues for women, we still find these discrepancies.
Bonnie Goldman: It's sort of ironic, because in national health statistics, women access care a lot more than men.43
Ben Young: Yes. More frequent visits per patient, per year, and despite that, we see these kinds of problems. But I think the problem lies in the perceptions, both from the patient perspective and from the provider perspective.
The last study that I'll talk about that focused on women -- again, there were many studies around women's access to care and demographics -- was I think a very sobering study conducted in Mississippi that was presented by Dr. Rana from Brown University.9
What this study did was look through a retrospective chart review of all of the women who were enrolled in the perinatal service at the University of Mississippi Medical Center in Jackson.
The study identified 275 women who, from 1999 to 2006, delivered 297 babies. The demographics of the women in this study again reflect the underappreciated demographics of the HIV epidemic in the south of the United States -- namely that the median age of these women was 25; 99% of the patients reported heterosexual transmission as the route of acquisition of HIV; 89% of these women were black and non-Hispanic; 40% of the women had attended grade school and high school, but had not graduated from high school; and only 4% of these women were college graduates.
This snapshot is one of a very different demographical makeup than that of the populations of patients with HIV in California, New York and Florida, the traditional epicenters of HIV in the United States.
Bonnie Goldman: Dr. Young, you forgot to mention, I thought, a most important point, and that was that a high percentage of these women were unemployed.
Ben Young: Absolutely. In fact, in a country where most people receive their third-party insurance through their employers, the very fact that most of these patients were either unemployed or underemployed, really speaks to this. In their analysis, 31% of the women were unemployed and for 53% there is no data. I presume that probably means they are unemployed, as well, but we don't know that for sure. Only 5% of the women stated that they were employed full time; 8% said that they were employed part-time.
We could look at this in a number of ways, but really, we're talking about a poorly educated population of women who are also poorly employed. One can only imagine what that implies for their overall health risk, never mind the specifics of having HIV or being pregnant with HIV, and all of the additional medical management concerns that this brings up with the prevention of mother-to-child transmission.
Bonnie Goldman: I believe all the women were Medicaid patients, so they had no real source of income. They were having a child without any money, which is just a striking picture in itself.
Ben Young: Absolutely.
Bonnie Goldman: Add HIV to that mix, plus the stigma of that in the South, and it's a heady mixture.
Ben Young: Absolutely. More to this point: 37% of the women in the study were diagnosed with HIV during their pregnancy. In a sense, their HIV-defining event was their pregnancy. Imagine the social and medical turmoil that this has to create in a newly diagnosed individual.
Aside from the demographical characteristics of this population of patients, which I think is very sobering, a more sobering analysis looked at how they did, what was their access to antiretroviral therapy and how successful was that therapy. I think, in this country, we all appreciate that access to HIV therapies during pregnancy effectively prevents transmission to child. So the goal of medical systems and the goal of health care providers should be to get these women under care and initiate combination therapies as soon as it's feasible to try to drive viral loads to undetectable levels at the time of delivery.
That was only achieved in 67% of these women. There were 14 transmissions among the nearly 300 deliveries, an overall transmission rate that's just under 5%. The overall transmission rate in the United States is far less than 1%. Those differences to me represent a significant area of unmet need and, frankly, individual catastrophe that we need to try to understand and do a lot more for.
The systems failed those 14 children, at the very least. But I think it really speaks to a very troubling patient population that is underappreciated in the United States by the medical community. It's a significant area for social support and governmental recognition. Systems really do need to work on this. It's perhaps a sentinel, sociological symptom of the simple fact that an awful lot of African-American women here in Washington, D.C., have HIV, with a prevalence rate that actually matches that of Port-Au-Prince, Haiti -- a frequency that's higher than that of any other region in the United States and higher than that of many of the capitals of developing world countries, where we accept that HIV is epidemic and pandemic.44,45
I'm very concerned that we are missing the ball here. We have this very shiny silver train that's highly active antiretroviral therapy. Those patients aboard the train will in all likelihood live longer and well. But the train is leaving the platform and there's a significant population of people in the developing world and also here in the United States who don't have tickets.
There are actually some other important points to this study. The first is that 70% of the pregnancies concluded with a tubal ligation, a rate that's far higher than that of the general population.46 So the issues and notions around family planning and such clearly are different than that of the general population. I have my concerns about how those discussions were proffered to women, potentially many of whom were probably diagnosed with HIV very recently.
Moreover, many of the children were lost to follow-up or had suboptimal postpartum care. Of the 195 women who were known to have HIV at the time of their pregnancy diagnosis, 27% of their children had fewer than three postnatal visits in the year following care. These are children born to HIV-infected women, children that should be theoretically receiving postnatal antiretroviral therapy. You don't get antiretroviral therapy with only three or fewer visits. So the children are getting suboptimal care.
Lastly, a significant number of these women are just lost to follow-up. So the systems don't address the social needs, and clearly don't, therefore, address the unmet medical needs for this patient population. The systems, terribly to me, seem to be reiterating a lot of the challenges that we have and continue to face in Southern Africa, Central Europe, Eastern Europe and Central Asia -- parts of the world where a lot more work has to be done. The very fact that this is happening here in the United States to me is a very, very sad point.
Bonnie Goldman: It's a lesson to all the people who think that we're very successful here in America and we now only have to help people in resource-poor countries. I think that this is just one of many, many studies that take a picture of a part of America that is not being looked at.
Ben Young: A part of America that many of us don't want to look at. It's perhaps not surprising, or maybe it is a commentary, that most of the research being done in these patient populations is being done by people from the North, and not by investigators and doctors within the region that's most affected. I'd like to see the day when we encourage local providers to recognize problems and become part of both the investigative solution and the on-the-ground treatment solution. I hope that day comes soon.
Ben Young: We have now talked about a very difficult set of circumstances and areas of unmet need. Let me try to somehow elegantly transition to another area of emerging concern, an area that I think is going to receive a lot of attention, and that is namely this interface between HIV medicine and an aging population of surviving patients.
Certain kinds of non-AIDS malignancies -- such as lung cancer, colorectal cancer, adenocarcinoma of the colon, non-Hodgkin's lymphoma and squamous-cell carcinoma of the anal canal -- seem to occur in greater frequency among the HIV infected.47 In that context, it's probably notable that there have been a number of studies in the last couple of months that have addressed the question about how good are we, as medical providers, at doing routine screening for these very common cancers that we actually know how to screen for.
My colleagues and I presented a paper looking at colorectal cancer screening utilization.48 Only two-thirds of patients in our center where colorectal cancer screening was recommended had actually received it; so a third of patients were being under-screened for colon cancer.
Similarly, a study presented here by a group from the Miriam Hospital at Brown University looked at breast cancer screening in the HIV clinic. There were 223 women in this clinic who would have been eligible for a screening mammogram.10 Only 23% of them had that done. I think that my colleagues and I provide pretty good care in our center in Denver, and I know that the group in Providence is an exemplary HIV clinic. Nevertheless, in these two centers where we try to provide high quality care, a significant number of patients aren't getting the recommended background cancer screenings for people without HIV, let alone those who have HIV and have a higher risk for cancer.
Bonnie Goldman: Do you think this is a function of no insurance for it? Or no time to do it? Or the patient doesn't want yet another procedure?
Ben Young: There are clearly multiple reasons why people fail to get cancer screenings. The first is that the doctors don't talk about it and don't offer it; that turned out to be probably the biggest reason why patients in our clinic weren't getting their cancer screenings.
In patients who are very sick and have AIDS complications, neither the patient nor the doctor want to talk about, or are concerned as much about, cancer screening as they are about taking care of the Pneumocystis, for example. It turned out that this accounted for a minority of the cases in these studies. Patients are reluctant to talk about cancer because they are worried about other things. They're too busy. Paradoxically, it's the patients who survive for decades with HIV, who often become less engaged in preventive health measures as they become busy with things that are not related to their health, such as jobs and family.
Again, it says that individual patients probably have a different constellation of reasons, but that doesn't obviate the need to make sure that people get just the routine stuff. I actually think that we will be seeing an increased, or a heightened sense of, need for screening for colon cancers and other types of cancers, not less. So, if anything, we need to be doing a more stringent job in screening patients with HIV for these kinds of malignancies. As we move away from being AIDS doctors and now that antiretroviral therapy has become much better tolerated, we need to devote more attention to the implications of long-term survival with HIV, and the complications that that brings.
Finally, I would like to conclude with a discussion about bone disease. Bone disease did not capture the attention of medical researchers in HIV until very recently, but it is a problem that very clearly has been around for some time. Some of the first investigations for this were done by Pablo Tebas almost 10 years ago, when he noted problems with bone mineral density in patients living with HIV.49
Over the last two or three years, there have been an increasing number of clinical observational studies that have examined the frequency of abnormalities in bone mineral density -- namely, osteopenia and osteoporosis.
One of the more compelling stories came from Turner Overton via the CDC SUN study.50 This study was presented last year. It looked at over 500 patients living with HIV, and found that the overall prevalence today of abnormal bone density was over 60%.
So a very large population of patients today have problems with bone mineral density. Previously, it was questioned whether this was of any clinical significance. I have now seen unexplained vertebral fractures in patients in my practice. When these patients' bone density was examined using DEXA [dual energy X-ray absorptiometry] scans, previously undiagnosed osteoporosis was found.
Consistent with our clinic's observation, was a publication from Steve Grinspoon's group in Boston that showed that, in a population of about 8,500 HIV-infected individuals, the overall frequency and risk of having a vertebral, wrist or hip fracture was about 60% higher than that of the general population.51 While we didn't see these kinds of fractures five or 10 years ago, we're now starting to see them, and the clinical cohorts are now beginning to report these things. So this is a very real disease. I'm very concerned that we need to improve our ability to diagnose and treat this disorder.
Bonnie Goldman: Does anyone yet understand the mechanisms?
Ben Young: We know that abnormalities of bone mineral density come from many factors, much like cardiovascular disease.52 There are non-modifiable risk factors, such as one's gender and ethnicity. There are modifiable risk factors; things like smoking and tobacco use are associated with decreased bone mineral density.
The question, in a sense, is: Why do HIV-infected patients seem to have a much greater risk of these events than HIV-uninfected patients? Is it HIV? Is it HIV therapy? Is it some mixture of all of those things? At this conference, there were three presentations that I'd like to review that addressed this question.
The first was a study from the Edward Hines Jr. VA [Veteran Affairs] Medical Center in Illinois, a study that, in a sense, gave us an expected outcome.11 In brief, what that study showed was that HIV-infected patients who have low testosterone levels have a greater likelihood of having abnormally low bone mineral density. It just reaffirms what we have known from the so-called healthy population who doesn't have HIV, that low testosterone levels, hypogonadism, is one of the risk factors associated with low bone mineral density. What that means in my practice, is that what I have been doing for many years is probably a good idea -- namely, every one of my patients gets testosterone levels checked at baseline and, in symptomatic patients with fatigue or issues with hyposexuality, I check their testosterone levels annually.
Now we have another reason to check testosterone levels, namely, a risk of abnormal bone density.
There was a small study from the Mt. Sinai Medical Center in New York, authored by Kathryn Childs et al, that looked at vitamin D levels and bone mineral density.12
This study reiterates the idea that multiple factors are associated with abnormal bone mineral density. It also suggested that an evaluation of vitamin D levels, and perhaps parathyroid hormone levels, would be useful in the early assessment and diagnosis of abnormal bone health. The investigators further postulated that there may be some relationship between these abnormal conditions and tenofovir [TDF, Viread] -- a drug that has been associated in studies with the accelerated loss of bone mineral density, but which has not been associated with fractures to date.
Overall, these two studies look at circulating markers that we can use in the chemistry lab to identify patients that might be at enhanced risk, or greater risk, of having bone mineral density issues and/or fractures.
That takes us to a study that generated a lot of discussion at this conference; it is a subgroup analysis that was presented by Birgit Grund and collaborators from the University of Minnesota for the SMART study group.13
Most of you will recall that the SMART study was a study of continuous viral suppression versus drug conservation or intermittent therapy.53 The strength of the SMART study was its very large sample size of patients. Patients were followed in order to ascertain the effect of treatment versus no treatment on various markers. In the case of this subgroup analysis, the investigators were interested in looking at the effect of intermittent therapy on bone mineral density.
One might postulate, as is often popular, that since antiretrovirals are known to have toxicities, patients off therapy should have better bone health. In fact, that was what was seen in the study. The patients in the subgroup analysis were those patients who received DEXA scans through the four years of the SMART study.
What the authors showed by looking at changes in the bone mineral density in the spine, either by DEXA scan or by CT [computed tomography] scan, or changes in bone mineral density in the hip, was that patients who received continuous therapy had greater degrees of bone mineral loss than those who had intermittent CD4+ cell count guided treatment.
The basis for this is not known. It certainly suggests that antiretroviral drugs may play a role in the loss of bone mineral density. It's important to point out that patients in the study received a variety of treatments and the investigators were unable to find a single causative drug among all of the various treatment regimens.
I think this study raises some very challenging questions with regard to bone mineral density. While it may be better to be off therapy, in the SMART study, being off therapy was also very strongly associated with increased risk of death, cardiovascular disease events and so on. So it's a matter of balancing a lot of complex issues if this conclusion is correct.
Bonnie Goldman: So, a lot of very interesting studies.
Ben Young: Yes. It's reassuring to me that, in the second decade of highly active therapies, we continue to try to refine both the short-term implications of treatment and the long-term potential complications. We've seen very good studies show that we can improve the mortality and morbidity of patients, even those who are at the greatest risk of death -- namely, those patients with multidrug resistance. That to me is very important because, again, those patients represent some of the most challenging patients that I take care of in my practice.
The data from the ACTG19 and from the North American ACCORD groups,6 I think, provide the impetus to ask more detailed questions about when we start therapy. Some of those answers today, I think, are much more compelling than they were only a few months ago.
The story about HIV-infected women, particularly women of color, is very disturbing and needs much more attention, both from a research standpoint and from a clinical, day-to-day perspective.
Lastly, my patients are clearly living longer, and they're living in a time when all of us are becoming more concerned about the long-term implications of preventive medicine on the fifth, sixth and seventh decades of life for our patients. There's an emerging awareness of those new needs and challenges, as they affect patients with HIV. As care providers, those new needs and challenges inform how we take care of those people and how we try to provide optimal management and preventive medicine for patients as they live even longer.
Bonnie Goldman: Great review! Thank you very much.
Ben Young: Thanks, Bonnie, for the opportunity. I'd like to especially thank the many investigators who shared with us their poster presentations and their slide presentations. Without those contributions, we couldn't have this discussion.
This transcript has been lightly edited for clarity.
|Please note: Knowledge about HIV changes rapidly. Note the date of this summary's publication, and before treating patients or employing any therapies described in these materials, verify all information independently. If you are a patient, please consult a doctor or other medical professional before acting on any of the information presented in this summary. For a complete listing of our most recent conference coverage, click here.|