Top 10 HIV Medical Stories of 2006
Table of Contents
In the world of HIV research, 2006 was the year of the head-to-head HIV clinical trial. Well-powered, randomized, controlled, comparative studies tend to be large, expensive and lengthy -- and, as a result, they have a hard time keeping pace with the rapidly evolving field of HIV clinical management. It takes years to go from the initial conceptualization of a study to the release of final results; thus, what at first made for an intriguing research question, frequently ends up becoming mundane -- or worse, irrelevant. Like the light from a distant star, the results of these period-piece trials often come to us long after the drugs under study have seen their market share go supernova.
However, last year, several well-designed, comparative studies were either presented or published that answered questions relevant to today's clinical practice. In several cases, the studies were conducted quickly and were prescient in their design (i.e., ACTG 5142). In more than one case, the relevance is due to the fact that the studies' findings were not at all what the investigators expected (i.e., the SMART study, ACTG 5142 and A5005s).
To a clinical trialist such as myself, this glut of topical research is just way too cool. I found 2006 to be a refreshing and reaffirming 12 months. Refreshing in that we had some excellent comparative studies that broke a stalemate among researchers and clinicians regarding an important clinical issue; reaffirming because some of these studies, with their well-considered design and impressive execution, reminded me of the reason I went into this field.
We design clinical studies to fill gaps in our knowledge, so we can learn how to better care for our patients. Ideally, the motivation for the study is a disagreement among researchers and clinicians as to the answer to a specific clinical question; some hold firmly that the result will lie in one direction, while their learned colleagues shout back that, on the contrary, it will point the other way. To find out the answer, we experiment. This year, our experiments yielded many important answers.
The trials of 2006 that I felt had the greatest immediate or potential impact on how we approach treating people affected by the HIV epidemic are included in the major HIV medical stories of 2006 listed below. Most are clinical treatment studies; others are not, but all, I think, still challenge us to reconsider how we practice HIV care and how we can prevent further expansion of the HIV epidemic. All are worth reading and appreciating.
A review of:
The study they designed was straightforward. HIV-infected individuals with a CD4+ cell count greater than 350 cells/µL were randomized to continuous antiretroviral therapy based on U.S. Department of Health and Human Services (DHHS) treatment guidelines2 (the viral suppression arm) versus episodic antiretroviral therapy based on CD4+ cell count (the drug conservation arm), in which therapy was discontinued after the CD4+ cell count rose to 350 cells/µL or greater and then reinitiated when the CD4+ cell count fell below 250 cells/µL.
Participants were eligible for the trial regardless of antiretroviral treatment experience; however, less than 5% of the cohort was treatment naive. As described in an editorial in the New England Journal of Medicine by Judith Currier and Lindsey Baden3 that accompanied the paper's publication, at the time the study was designed, there was considerable concern in some quarters of the scientific community regarding the wisdom of interrupting effective HIV therapy. However, there was also a counterbalancing intense interest among clinicians, as well as patients, in strategies to reduce antiretroviral-related adverse effects and treatment fatigue.
The study, which was initiated by the U.S. National Institutes of Health (NIH)-funded Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA), opened in early 2002 in 33 countries and eventually enrolled 5,472 participants. Over 80% of the participants lived in North America or Europe; the median CD4+ cell count was 597 cells/µL and 72% had a plasma HIV RNA level less than 400 copies/mL. The trial had been expected to run until at least an average of six years of participant follow-up had been achieved. However, the trial was abruptly halted by the study's Data and Safety Monitoring Board (DSMB) with a mean on-study follow-up time of only 16 months.
What happened has become the most recent entry into the "Annals of Completely Unanticipated Study Results." Rather than protect patients, the withdrawal of HIV therapy increased the risk of death from any cause, including that of opportunistic infections. In addition, the rates of cardiovascular, hepatic and renal disease -- which are often considered antiretroviral-related -- were greater in the treatment interruption group (see Table 1).
Click to enlarge
Adapted from Wafaa El-Sadr, M.D., M.P.H., et al. N Engl J Med. 2006;355(22):2283-2296.
During study follow-up, the CD4+ cell count was an average of 206 cells/µL lower in the drug conservation arm than in the viral suppression arm. Not unexpectedly, the proportion of participants with virologic suppression plummeted with treatment cessation in the drug conservation arm, falling from 72% to 6% during the first two months of study participation.
Adjustment for the latest on-study CD4+ cell count and viral load reduced the hazard ratio for mortality classified as not related to AIDS-associated opportunistic conditions in the drug conservation group compared to the viral suppression group, implying that immunosuppression and uncontrolled viremia provide at least a partial explanation for the differences between the study groups.
According to data presented at the XVI International AIDS Conference (AIDS 2006) in Toronto, Canada, this past summer by the SMART investigators, the superiority of the viral suppression strategy was observed in all subgroups of participants that were analyzed.4,5 While the greatest risk for opportunistic disease or death was, as expected, observed among patients who had CD4+ cell counts of less than 350 cells/µL, continuous HIV therapy was found to be more beneficial than treatment interruption even among those participants with CD4+ cell counts that remained above 350 cells/µL and who, therefore, would be considered at less risk for adverse outcomes.
These analyses also demonstrate that, in fact, patients who had higher CD4+ cell counts and suppressed viral loads experienced particularly worse outcomes during their treatment interruptions. Thus it turned out that patients who were doing well on their antiretroviral therapy had more to lose from stopping therapy than those who were receiving relatively less benefit from their HIV treatment due to non-adherence (i.e., self-directed treatment interruptions) or other factors.
The Bottom Line
This CD4+ cell count-directed-intermittent HIV treatment interruption was found to be hazardous across the board. Patients who may seem most attractive for such a break -- such as those who have higher CD4+ cell counts and lower viral loads -- were most likely to experience an adverse event compared to patients who remained on therapy.
The findings of this trial are significant on several levels. First, the inferiority of the drug conservation strategy has chilled any notion of structured treatment interruptions. In their editorial, Currier and Baden suggest that long-term treatment interruption may be safe only for people who did not need HIV therapy to begin with (e.g., patients who have extremely high CD4+ cell counts when initiating therapy) and that there remain few circumstances in which a treatment interruption may continue to have a role. This would include interrupting treatment following therapy initiated during acute HIV infection and in women with high CD4+ cell counts who received HIV therapy to prevent HIV transmission during pregnancy, as well as brief interruptions (i.e., a few weeks) due to serious side effects in those with chronic HIV infection.
Some people may argue about the appropriateness of the CD4+ cell count "floor" that the investigators used to trigger resumption of antiretroviral therapy. The investigators felt a CD4+ cell count threshold for reinitiating HIV therapy of 250 cells/µL was safe given historic data demonstrating low rates of disease progression among HIV-infected patients with counts above 200 cells/µL.
Those who disagreed countered that a threshold of 250 cells/µL was cutting things too close. They would have preferred that therapy restart at a higher count. It is likely that any future studies of treatment interruption -- were they to be conducted -- would not tolerate the degree of immunosuppression experienced by participants in the drug conservation arm of this study.
Without a doubt, the superiority of continuous HIV therapy brings us back to the altar of treatment-worship. Concerns over the perceived complications of HIV therapies are giving way to a newfound appreciation for the toxicity of unchecked HIV viremia and progressive CD4+ cell count depletion.
The finding that sustained treatment holidays are largely a bad idea is absolutely clear from the findings of the SMART study. Clinicians must now share these results with their patients, who also need to understand the potential perils of their dabbling in therapy interruptions. Given the popular prejudices regarding HIV therapy, more than a few patients will be surprised to find that survival and cardiovascular outcomes were better in the continuously treated participants of this large study.
The intriguing finding of excess cardiovascular, renal and hepatic disease among patients assigned to the treatment interruption arm adds to an emerging body of evidence suggesting a role for both immunosuppression and HIV viremia in the etiology of some of these conditions. Higher rates of renal, hepatic and neurological adverse effects were also reported in untreated HIV-infected patients compared to those starting therapy in the HIV Outpatients Study (HOPS).6 It is possible that viremia, and its attendant immune activation, may provoke inflammatory reactions that can lead to end-organ damage. The suggestion that HIV therapy may actually be protective against cardiovascular disease will require careful examination within this study, as well as other studies.
Lastly, the trial was a coup for the scientific method and a testament to the value of studies on research questions that many people think they know the answer to in the absence of sufficient data. The investigators' hypothesis (that had been shared by many smart people) was proved to be false. The surprising results have led to a dramatic shift in our thinking about antiretroviral treatment and the complications of HIV and its therapies.
A review of:
Often the ACTG, as a federally funded research network, is uniquely able to conduct comparative trials that the pharmaceutical companies have no interest in performing. Such was the case for ACTG 5142,7 which pitted the two longest-standing, guideline- recommended antiretroviral treatment regimen anchors, efavirenz and lopinavir/ritonavir, against one another.
Naturally, there were the expected ACTG idiosyncrasies. Although independent from the pharmaceutical industry, the ACTG relies on drug companies to provide study medications free of charge. In this case, not everyone from big pharma was eager to play. As a result, for a large proportion of the study participants, not all the HIV medications were supplied at no cost -- leading to an over reliance on extended release stavudine (d4T, Zerit), which was provided gratis to the investigators.
In addition, there are the usual ACTG statistical quirks, such as undetectable being defined as less than 200 copies/mL, composite endpoints and non-traditional P-values. But these probably just indicate that ACTG statisticians are simply smarter than the rest of us.
The design of this trial may be well known to most readers since it snagged a lot of attention this past summer at AIDS 2006, where it was first presented. Efavirenz and lopinavir/ritonavir were compared to each other in combination with a pair of nucleoside/nucleotide reverse transcriptase inhibitors (NRTIs). The NRTIs were chosen by the clinicians and patients, and 753 HIV treatment-naive participants were enrolled.
Importantly, during the initial phase of the study, only one of the NRTI options was provided at no cost to the participant, and this was the extended-release stavudine. Later, tenofovir (TDF, Viread) was also made available by the study at no cost to patients. This led to overuse of stavudine, which is now considered too toxic for use when alternative NRTIs are readily available. A third, "nuke-sparing" study arm consisted of only efavirenz + lopinavir/ritonavir without accompanying NRTIs.
At present, everything we know about the findings of this study was presented at AIDS 2006 during a 15-minute late breaker session by Sharon Riddler of the University of Pittsburgh. Essentially, the study demonstrated that at week 96 a significantly greater proportion of participants who had been assigned to receive efavirenz + two NRTIs reached the virologic endpoints of a plasma HIV RNA level of less than 200 copies/mL or less than 50 copies/mL, compared to participants who had been randomized to receive lopinavir/ritonavir + two NRTIs (Table 2). The nuke-sparing regimen produced results that were, generally, in between those of the other two study arms.
Click to enlarge
Adapted from Sharon A. Riddler, et al.
Although limited data were presented on toxicity, those that were shown were tantalizing. In the two NRTI-containing arms, about one third of participants experienced a moderate or severe laboratory abnormality, and approximately 20% had a new high-grade sign or symptom. But, there were no major differences in the rates of these events between these two arms. The nuke-sparing arm did produce a high rate (45%) of moderate or severe laboratory abnormalities. Off-the-chart hypertriglyceridemia was rare in both NRTIs arms (less than 6%), but was high (14%) in the nuke-sparing arm -- perhaps accounting for much of the excess laboratory adverse events seen in this arm. Data on lipid-lowering drug use was not presented.
Additional lipid and body shape data were reserved for presentation at the 14th Conference on Retroviruses and Opportunistic Infections (CROI 2007) in Los Angeles, Calif.
Click to enlarge
Adapted from Sharon A. Riddler, et al.
Particularly noteworthy was that the time to first treatment-limiting toxicity was observed to be no different among the three study arms at 96 weeks. More than 80% of the participants remained free of such an adverse event.
However, when examining the composite study endpoint of "regimen completion" -- which was defined as virologic failure (lack of viral suppression or rebound) or toxicity requiring discontinuation of any component of the regimen -- at 96 weeks, 56% of the patients who were assigned to lopinavir/ritonavir + two NRTIs, compared to 40% of those randomized to efavirenz + two NRTIs, reached regimen completion (P = .02). The proportion of participants meeting the study endpoint of regimen completion was similar in the nuke-sparing arm and the efavirenz + two NRTIs arm.
Therefore, 50% to 60% of the participants assigned to these state-of-the-art therapies experienced either a virologic failure or an adverse event requiring a change in one or more of their antiretrovirals (note: a quarter of the participants were taking extended-release stavudine).
Granted, the protocol-specified definition of virologic failure was more stringent than what is typically utilized in clinical practice. To what extent patients with low-level viremia were considered treatment failures was not included in Riddler's brief presentation. Likewise, still unreported is whether treatment modification was related to problems with the NRTIs (i.e., stavudine) rather than the protease inhibitor (PI) or non-nucleoside reverse transcriptase inhibitor (NNRTI) under study. The specific reasons for drug discontinuations also remain unreported.
Less surprising were the viral resistance data. No resistance to lopinavir/ritonavir was observed in participants who were assigned to this boosted PI plus two NRTIs. In contrast, resistance to efavirenz was detected in almost half of the patients who failed efavirenz + two NRTIs (however, resistance testing results were not available for everyone). NRTI resistance was observed in more patients with virologic failure while on efavirenz + two NRTIs than in patients in the other two study arms.
The Bottom Line
There was no clear "winner" in this trial, yet it was extremely successful in achieving its goal -- which was to provide clinicians and patients with a side-by-side comparison of two of the most potent and popular antiretrovirals.
Different people will take home differing messages regarding the results according to what is most important to them. For some people, the virologic data will reinforce confidence in efavirenz-based regimens. For others, the resistance and immunologic results will bolster faith in lopinavir/ritonavir and, perhaps by extension, in other boosted PIs.
However, the significance of the results of this trial extends beyond the study aims to compare the efficacy and safety of these regimens. In my view, the beauty of this investigation's conclusions is found in how completely unexpected the results turned out to be. In many of today's HIV clinical treatment trials, the outcome is all but assumed: Powerful drug X is pitted against a known weaker drug Y and, voila!, X wins. Who among you would have predicted that efavirenz would lead to significantly greater rates of viral suppression than lopinavir/ritonavir? And who would have guessed that the rates of high-grade hypertriglyceridemia would be similar between these two drugs? Planned analyses of adverse events and metabolic complications may yield even further surprises. More data from this trial was presented at CROI 2007, and surely a manuscript cannot be far behind.
While no other single trial in 2006 was, in my view, as monumental as either the SMART study or ACTG 5142, there were a few other noteworthy HIV treatment trials, which pitted popular treatment regimens against one another to reveal their relative strengths and weaknesses, that collectively merit inclusion in this list.
A review of:
In the industry-sponsored KLEAN trial, the two boosted PIs were compared over 48 weeks in 878 treatment-naive patients at 131 sites across the United States, Canada and Europe. Fosamprenavir was dosed as 700 mg along with 100 mg of ritonavir, both twice a day. Lopinavir/ritonavir in the original soft-gel formulation was also administered twice daily. All study therapy was open-label.
Here again, the results defied the oddsmakers. Across all parameters by which these two regimens were compared -- virologic, immunologic, metabolic -- there were no significant differences between the study arms. There was not even a trend. Both boosted PIs were able to lead to profound reductions in viral load and impressive increases in CD4+ cell count. In the intent-to-treat analysis, in which all those exposed to drug were considered, the rates of viral suppression below 50 copies/mL at 48 weeks were 89% and 88% in the fosamprenavir + ritonavir group and the lopinavir/ritonavir group, respectively (P = NS).
There were no differences in virologic responses by treatment arm when the analysis was stratified by baseline viral load above and below 100,000 copies/mL. The CD4+ cell count increased by a median of 176 cells/µL in the fosamprenavir + ritonavir group (interquartile range [IQR] 106-281) compared to 191 cells/µL in the lopinavir/ritonavir group (IQR 124-287) (P = NS).
Both regimens were also well-tolerated. Only 12% of the patients who were assigned fosamprenavir + ritonavir and 10% of those randomized to lopinavir/ritonavir prematurely discontinued study medication due to adverse events (some of these events were due to abacavir hypersensitivity reactions). The rate of diarrhea, a typical adverse effect with both of these PIs, was similar between the two arms (11% for lopinavir/ritonavir and 13% for fosamprenavir + ritonavir).
Of interest were the lipid data from this trial. There were modest increases in low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol and fasting triglyceride levels in both arms but, again, no differences between the two arms with regard to these parameters (Table 4). This is notable as there has been a perception that lopinavir/ritonavir is particularly deleterious when it comes to on-treatment lipid changes and conflicting data regarding the relative effects of these PIs on lipid parameters.
Click to enlarge
Adapted from Joseph Eron Jr., M.D., et al. Lancet. 2006;368(9534):476-482.
A review of:
Further, in a comparison with lopinavir/ritonavir in treatment-experienced patients, patients who switched from a failing combination antiretroviral therapy to ritonavir-boosted atazanavir saw a mean reduction in triglyceride, LDL cholesterol and total cholesterol levels.10 In contrast, 48 weeks after the switch, study participants who were assigned to lopinavir/ritonavir experienced an increase in triglycerides and total cholesterol and no change in LDL cholesterol. Although these results demonstrated less of an impact by ritonavir-boosted atazanavir on most lipid parameters in comparison with lopinavir/ritonavir in the treatment-experienced, little information on the effects of boosted atazanavir itself on lipids was provided. The absence of these data made it difficult to appreciate the precise degree to which ritonavir offsets atazanavir's beneficial aspects vis-à-vis lipids, which is exactly what the study known as BMS-089 set out to find out.
BMS-089 was a 96-week, open-label trial of 300 mg of once daily atazanavir boosted with 100 mg of once daily ritonavir versus 400 mg of unboosted atazanavir, both administered with extended release stavudine and lamivudine (3TC, Epivir), in 199 treatment-naive patients. The investigators aimed primarily to examine virologic efficacy, but also were able to detail the lipid effects of atazanavir and ritonavir co-administration.
At 48 weeks, the proportion of participants with a viral load less than 50 copies/mL was not significantly different statistically: 75% in the boosted arm versus 70% in the unboosted arm (95% CI: -7.0, 17.0). However, in the unboosted arm, there were 10 virologic failures compared to only three in the ritonavir-boosted atazanavir arm. Although there were few resistance mutations detected during the study, more people (n = 7) in the unboosted arm were observed to have drug resistance, mostly at the 184 codon. Only one patient in the boosted arm had a resistance mutation (M184V).
Interestingly, there were more adverse effects (mostly related to hyperbilirubinemia/jaundice) in the ritonavir-containing arm compared to the unboosted arm (8% versus less than 1%). Almost a quarter of the participants who were taking ritonavir-boosted atazanavir experienced jaundice compared to 7% in the unboosted-atazanavir group.
In terms of lipids, a ritonavir-effect was evident. Changes in total cholesterol and triglyceride levels were significantly higher in the ritonavir-boosted atazanavir arm (Figure 1).
Click to enlarge
Adapted from Niel Malan, et al.
Virologically, although boosted and unboosted atazanavir both led to similar proportions of patients achieving viral suppression, the study was small. The excess virologic failures and detectable resistance in the unboosted arm indicate a relative lack of potency with regard to this (and probably every) PI when it is not boosted with ritonavir. However, the added virologic potency achieved with ritonavir boosting does come with a price -- beyond the monetary type -- and that is the added risk of significant hyperbilirubinemia and jaundice. As was the case for fosamprenavir, ritonavir-boosted atazanavir was recently added to the U.S. DHHS list of recommended initial therapies.2
Tenofovir + Emtricitabine + Efavirenz Versus Zidovudine/Lamivudine + Efavirenz -- Gilead 934
A review of:
The NRTI components of this combination drug have been most rigorously studied in Gilead 934, a head-to-head study of treatment-naive patients randomized to receive open-label tenofovir + emtricitabine (FTC, Emtriva) or fixed-dose zidovudine/lamivudine (AZT/3TC, Combivir) along with their efavirenz. The 48-week results of this study were first published in the New England Journal of Medicine11 and the 96-week data followed in JAIDS.12
After 96 weeks, the proportion of participants with a viral load less than 400 copies/mL (the primary study endpoint) was found to be significantly greater in the tenofovir + emtricitabine arm. However, the proportions of participants within the study arms with viral suppression below 50 copies/mL were not significantly different. These viral endpoints were analyzed using the time-to-loss of viral response (TLOVR), an approach that is preferred by the U.S. Food and Drug Administration (FDA), and that includes treatment discontinuation for virologic reasons as well as for reasons related to toxicity. Therefore, excessive toxicity in one arm could tip the analysis in favor of the other. This was clearly the case at 48 weeks, when higher rates of treatment-limiting adverse effects (largely anemia) in the zidovudine/lamivudine arm led to a significantly greater proportion of participants in the tenofovir + emtricitabine arm achieving the primary endpoint of a viral load below 400 copies/mL. At week 96, this was still the case; however, at this time point there were also significant differences between the arms with regard to the rates of virologic rebound (5% versus less than 1%), which contributed to the superiority of tenofovir + emtricitabine.
Overall, treatment-limiting adverse events were not common, but were more likely with zidovudine/lamivudine (11%) than with tenofovir + emtricitabine (5%) (P = .008). There were no clinical renal events leading to treatment discontinuation in either arm. The median glomerular filtration rate (GFR), estimated by the Cockcroft-Gault equation, declined in both groups over 96 weeks, with no significant difference between the arms seen. However, the Modification of Diet in Renal Disease (MDRD) equation, considered a more sensitive estimate of renal function, did find a greater decline in GFR in the tenofovir + emtricitabine arm. As in other studies of tenofovir, the clinical significance of small, but statistically significant, changes in GFR estimates is unclear.
DEXA (dual energy X-ray absorptiometry) scans were performed at weeks 48 and 96 (but were not performed at baseline!) to assess regional fat. Interestingly, a significant decline (-700 grams) in limb fat was observed in the zidovudine/lamivudine-assigned participants, while patients who were randomized to tenofovir + emtricitabine experienced a median gain (300 grams) in total limb fat. The observation of a drop in limb fat with zidovudine/lamivudine + efavirenz stands in stark contrast to results from ACTG study A5005s,13 a rigorous metabolic substudy of ACTG 384, which found no change in peripheral fat among patients receiving this combination (this study will be examined further later in this paper). In A5005s, DEXA scanning was performed at baseline and at six-month intervals over 144 weeks.
The Bottom Line
These three compare-and-contrast antiretroviral studies provide useful data that clinicians can mull over before reaching for the prescription pad. The unexpected parity of lopinavir and fosamprenavir in the KLEAN study and the performance of atazanavir in BMS-089 suggest that the differences between ritonavir-boosted PIs may not be as vast as we may have once thought.
The 934 study makes it loud and clear that tenofovir + emtricitabine works (duh!) and that it can more than hold its own against zidovudine/lamivudine. The discordance of the DEXA results between the 934 and ACTG studies is interesting and it will take other studies, including A5142, to better address the affect of zidovudine/lamivudine + efavirenz on body fat.
In terms of further qualification of atazanavir, in the future, we can look forward to seeing the results of an ongoing direct comparison of lopinavir/ritonavir and ritonavir-boosted atazanavir in treatment-naive patients.
A review of:
Circumcision and HIV transmission. TC Quinn. Current Opinion in Infectious Diseases. February 2007;20(1):33-38.
Epidemiologic studies, mostly conducted in sub-Saharan Africa, have overwhelmingly observed an association between male circumcision and protection from infection with several sexually transmitted infections, including HIV.14,15 Even after adjustment for confounding variables, cross-study analyses suggest that circumcised men are approximately 70% less likely to acquire HIV than uncircumcised men.
Given the consistency of the results of these studies, all that was left to clinch this association was a randomized clinical trial. For those of us conducting drug treatment trials, studies that randomize participants to a surgical procedure can seem rather extreme. However, there is a long history of such studies, including recent trials, some even blinded with the use of sham procedures.
Three major randomized controlled studies of male circumcision to prevent HIV infection of men have been initiated in South Africa, Kenya and Uganda. The first of these was published at the end of 200516 and the other two were ended in 2006.17 Each of these three trials was halted when interim analyses demonstrated that medically performed circumcision protected men against HIV infection.
At that point, there were 20 infections in the intervention group compared to 49 in the control group. This corresponded to a rate ratio of 0.4 (95% CI 0.24, 0.68%; P < .001) -- that is, in this study of South African men, there was a 60% reduced risk of HIV infection with circumcision seen compared to the control group. When these data were analyzed to account for men who were assigned to circumcision and opted out of the procedure and men in the control arm who went and were circumcised on their own during the study, the protective effect was even stronger, at 76%.
The studies conducted in Kenya and Uganda had a similar design to that employed in the South African trial. They too were halted early as a result of the early benefits seen in the circumcision arms. The Kenyan study was conducted in an urban setting and enrolled 2,784 men while the Ugandan trial, based in a more rural region, enrolled 4,996 men.
The Bottom Line
Even a cynic can appreciate how impressive these results are. Given the difficulty found in motivating behavioral change, there is room for alternative methods to assist in preventing the transmission of HIV and other sexually transmitted diseases that can facilitate HIV acquisition. For sure, there are limitations to the application of circumcision as a global HIV preventative method -- as one international HIV researcher said to me, "We are not going to circumcise ourselves out of the HIV epidemic." This procedure clearly has some finite associated morbidity. And, outside of research settings, the risk of infection and other complications may be higher.
Additionally, the protection afforded by circumcision is hardly absolute; some circumcised men were still infected. While rates of new HIV infection among men undergoing the procedure were lower than those of the control groups, the follow-up time was short due to the halting of these trials. Longer-term study will be needed to understand the durability of the protection circumcision provides.
Further, given the limitations of the protection of circumcision, over time one could imagine a false sense of security among circumcised men that could actually lead to risky behavior and a paradoxical increased exposure to infection.
Additionally, there are few data to indicate whether circumcision will be of much effect in protecting men who have sex with men (MSM). In a study of MSM in the United States, acquisition of HIV was twice as likely in uncircumcised men compared with men who were circumcised.18
That said, in a world where policymakers tell us that the choices for preventing HIV can be summarized by the first three letters of the alphabet, even the limited protection of circumcision is more than welcome.
A review of:
Since the beginning of the HIV epidemic, there have been obvious advances in HIV care. This began with the expansion of the antiretroviral armamentarium from one agent to more than 24 agents, spanning at least four drug classes today. In addition, sophisticated tests for the detection and quantification of HIV viremia, as well as drug resistance (and soon co-receptor tropism), have become part of the standard of HIV clinical care in the United States and elsewhere. At the same time, among HIV-infected patients, hospitalization rates have declined, as have serious incapacitating illnesses.19 This has led to prolonged disease-free survival.
To quantify the benefits and costs of HIV medical care in the United States today, Schackman et al created a hypothetical cohort of patients presenting for HIV care. The investigators modeled the characteristics of the patients, their utilization of health services (medication, outpatient and inpatient services and laboratories) and their disease progression. They used data from a number of sources, including the HIV Research Network (HIVRN), a consortium of U.S. HIV clinics, the HIV Cost and Services Utilization Study (HCSUS), the Multicenter AIDS Cohort Study (MACS), U.S. DHHS treatment guidelines, and previously devised models.
Under this model, patients entered care with a mean CD4+ cell count of 310 cells/µL at 39 years of age. The distribution of HIV RNA levels of new patients was based on data from a Boston HIV clinic. Therapy initiation followed the U.S. DHHS guidelines and treatment consisted of a sequence of regimens starting with tenofovir plus one other NRTI and efavirenz and ending with an enfuvirtide (T-20, Fuzeon)-containing salvage regimen. Medication toxicity and virologic failure rates were based on clinical trial data. In all, the investigators aimed to create a realistic model of the U.S. HIV-infected patient population and were inventive in their inclusion of details that could influence care and its costs.
Click to enlarge
Adapted from Bruce R. Schackman, Ph.D., et al. Med Care. 2006;44(11):990-997.
Analyses were conducted in which the timing of HAART initiation and its effectiveness were looked at. One of the most interesting findings of one of these analyses was that when HAART was initiated when the CD4+ cell count dropped below 200 cells/µL, both cost and survival were reduced, the latter by 1.7 years.
On the other end of the spectrum, if HAART was started at the time of infection, survival jumped to 32.1 years, with a total lifetime cost of only $303,100. The low cost in this case was due to anticipated continued discounting of HIV medications at a rate of 3%.
The Bottom Line
Modeling studies such as this one do what cannot be done with actual human beings. But a model is only as good as the data it is based on and here the investigators -- a dream team of experts in this field -- worked painstakingly to account for everything that may influence cost and care. There are inevitably holes in the virtual reality they created and there are assumptions that may not prove to be true (such as the projected cost of HIV medications). Not included were other costs -- such as those for hepatitis C therapy; mental health and substance abuse treatment; as well as costs for the management of diabetes and cardiovascular disease.
Despite these limitations, these analyses were extremely instructive in that they provided an estimate -- based on all the considerations one can throw into the mix -- of the magnitude of the benefit of HIV therapy and care and, at least, a ballpark figure on the costs. The prognosis aspect of the findings is of great importance to our patients who understandably, want to know what they can expect. This study gives the clinician a reasonable basis from which to answer that toughest of questions.
A review of:
Viral load testing was ushered into clinical practice as a result of data from MACS, an ongoing prospective U.S. study of the natural and treated history of MSM. This cohort demonstrated that higher levels of plasma HIV RNA meant increases in the rate of progression to AIDS or death. Since the publication of these results in 1996,20 they have left an indelible impression on HIV care.
Extrapolating from these results, many clinicians refer to the actuarial-like tables derived from MACS and other cohorts to help patients appreciate the likelihood of disease progression given their CD4+ cell count and viral load. The importance of the pre-treatment plasma HIV RNA level is also reflected in its inclusion in the U.S. DHHS treatment guidelines2 as a parameter to consider when determining the appropriate time to initiate a patient on HIV therapy.
However, there are several observations that have called into question the preeminence of the pre-treatment viral load as a predictor of the rate of CD4+ cell count decline for an individual patient. Foremost, many clinicians have cared for patients who have low viral loads and unexpectedly experience rapid CD4+ cell declines. In addition, there is emerging evidence21 for the role of cell phenotype, co-receptor tropism, immune activation as well as individual HIV-specific immune responses in HIV disease progression.
A study by Benigno Rodríguez et al22 set out to examine the ability of pre-treatment HIV viral load to accurately predict the rate of an individual patient's CD4+ cell count decline. They analyzed data from 1,289 HIV-infected patients culled from several sources, including the Centers for AIDS Research (CFAR) Network of Integrated Clinical Systems sites (seven major clinical centers in the United States), the San Francisco Men's Health Study of MSM in that city and the Research in Access to Care for the Homeless (REACH) cohort (a cohort of homeless and marginally housed HIV-infected individuals in the San Francisco area).
All patients included in the analysis were HIV-therapy naive, and it was largely determined that they were not in the throes of acute HIV infection. The investigators were able to collect a contemporaneous viral load and CD4+ cell count and a repeat CD4+ cell count at least six months later, during which time the patients were still not receiving HIV treatment. These data were used to create a model in which a slope of the CD4+ cell count change per individual -- calculated using data from the cohort -- and viral load were included to determine how well the viral load predicted a patient's yearly CD4+ cell count decline.
As has been seen in other cohorts, when patients were grouped into broad categories based on viral load levels, the higher the presenting HIV RNA level strata, the greater the patients' overall subsequent CD4+ cell count decline. Specifically, the median annual-model derived CD4+ cell count decline for those with a viral load of less than 500 copies/mL was 20 cells/µL, for 501 to 2,000 copies/mL it was 39 cells/µL, for 2,001 to 10,000 copies/mL it was 48 cells/µL, for 10,001 to 40,000 copies/mL it was 56 cells/µL and for greater than 40,000 copies/mL it was 78 cells/µL. Therefore, a group of patients with similar viral load levels tends to lose CD4+ cells at a faster rate than a group of patients with a lower level of HIV viremia.
However, the pre-treatment viral load was found not to be an accurate predictor of the speed at which an individual's CD4+ cell count changes. The variability in CD4+ cell count decline between patients was poorly explained by baseline viral load, which indicated that there are other factors with greater influence on the decay of the pool of these cells.
In an attempt to validate these findings, the analysis was repeated -- this time using data from MACS. When this was done, the results were found to be remarkably similar.
To understand how viral load levels can be associated with changes in CD4+ cell count when categorized into broad groups, but not be predictive of the decline of these cells at an individual level, it is helpful to look at Figure 3, in which the distribution of CD4+ cell count changes within each quartile of viral load measured is displayed. There is clearly a wide range of CD4+ cell count changes within each viral load category. In addition, there is marked overlap in the changes in CD4+ cell count across the viral categories. Therefore, while the average CD4+ cell count drop may be steeper in the higher viral load quartile, there are a significant proportion of individuals with similar CD4+ cell count changes in each of these viral load strata.
Click to enlarge
Adapted from Benigno Rodríguez, M.D., M.Sc., et al. JAMA. 2006;296(12):1498-1506.
The Bottom Line
This is an extremely interesting analysis that dramatically challenges the existing belief system regarding the predictive value of quantitative HIV viremia. There are important limitations to Rodríguez et al's study, including the fact that patients who initiated HIV therapy within six months of their initial viral load were excluded from the dataset. In addition, the investigators relied on relatively few viral load determinations to predict CD4+ cell count decline over a fairly short period of time.
However, the rigor with which the investigators attempted to confirm their controversial findings -- reconstructing the analysis using data from the MACS cohort, the very population that was first used by John Mellors and colleagues20 to describe the associations between viral load and HIV disease progression -- makes it difficult to not be impressed by these results and their implications, despite the shortcomings.
Many clinicians have focused on the CD4+ cell count over the viral load when considering treatment initiation. Others, however, take a more aggressive approach and recommend antiretroviral therapy to patients who have high levels of HIV viremia even at relatively high CD4+ cell counts. This study suggests that given the wide variability in CD4+ cell count decline, HIV treatment of such patients cannot be justified as a means to prevent CD4+ cell count depletion.
The study also makes the point that there are other factors besides viremia itself that seem to drive disease progression. In an accompanying editorial,21 Keith Henry, Pablo Tebas and Clifford Lane list several potential influences on the fate of CD4+ cells in HIV-infected individuals and call for the development of therapies that go beyond the targeting of the virus and that enhance immune responses to HIV.
It can be expected that the issue of the predictive value of the viral load will continue to be studied and debated until a consensus is achieved through the replication of study results. Meanwhile, the prudent clinician should perhaps be a bit less dogmatic about what a patient's pre-therapy viral load foretells.
A review of:
The best data we have available regarding body shape during HIV infection come from two studies: the Fat Redistribution and Metabolic Change in HIV Infection (FRAM) study23 and the ACTG study A5005s.24 FRAM was a cross-sectional study that clearly demonstrated that fat wasting was the dominant body shape change in the HIV-infected men and women studied. A5005s is a longitudinal substudy of a cohort of patients who are participating in ACTG 384, a trial of six different antiretroviral regimens in treatment-naive participants (Figure 4). The ability of A5005s to rigorously evaluate changes in body shape over time makes it the best source of data regarding these changes currently available.
Click to enlarge
Adapted from Michael P. Dubé, M.D., et al. AIDS. 2005;19(16):1807-1818.
Click to enlarge
Adapted from Michael P. Dubé, M.D., et al.
During 2006, A5005s produced additional information of major importance for anyone seeking clarity regarding fat and HIV. In an article published in the Journal of Acquired Immune Deficiency Syndromes,25 the A5005s study team found that, during the first 64 weeks of antiretroviral therapy, changes in limb and trunk fat were directionally concordant, with increases in both limb and trunk fat observed in 36% of participants and decreases in both limb and trunk fat detected in 32% of participants. One third of patients had discordant results, and only 26% of the cohort had a decrease in limb fat accompanied by a gain in trunk fat -- the purported sine qua non of the lipodystrophy syndrome.
On top of these data, the investigators presented the 144-week DEXA results at AIDS 2006.26 Importantly, their poster examined the effects of assigning NRTI + NNRTI and NRTI + PI regimens in an "as-treated" analysis, wherein patients who switched assignments were censored from inclusion in subsequent analyses. Previously, the authors had presented data using an intent-to-treat approach, which was less than ideal given the rate of regimen modifications that took place during the study, especially in the stavudine + didanosine (ddI, Videx) arm, which was fairly toxic.
The meat of their presentation is found in the PI versus NNRTI comparison. Participants who were receiving nelfinavir had greater fat loss than those who were receiving efavirenz. Interestingly, there was a 7.9% annual decline in limb fat in the patients who were taking the PI + zidovudine (AZT, Retrovir) + lamivudine, compared to a 2.7% per year gain in the NNRTI group (P = .03). Therefore, patients taking zidovudine + lamivudine + efavirenz did not experience limb fat loss, while those taking these NRTIs with a PI did.
The Bottom Line
These data demonstrate a few things:
Regardless, these data should be better appreciated. The finding that trunk fat increases no matter what HIV therapy one starts, provided it works well, is not that much of a surprise as every antiretroviral study that has bothered to look has found likewise (repeat after me, "It is not only PIs that increase belly fat").
The lack of an effect of zidovudine + lamivudine + efavirenz on limb fat is touchier. The industry-sponsored 384 study, discussed above, found that this combination did reduce limb adipose volume. However, as mentioned previously, baseline DEXA scans were not done in that study, so the follow-up time is less and cannot be compared to pre-therapy scans. Again, A5142 provides the best opportunity to examine this combination's effects on regional fat.
Patients are making decisions about therapy based on their perceptions of which therapies are considered most likely to cause metabolic disturbances. At a minimum, they should be advised that fat wasting has become rarer with the U.S. demise of stavudine, and it seems to be even less likely with stavudine-free, NNRTI-based regimens.
Patients should also be told that bigger bellies occur following treatment with both NNRTI-containing and PI-containing regimens. Now that we finally have some clarity on this issue, it is time to plow ahead and discover precisely why some people accumulate visceral fat and others do not and how this complication can be prevented and treated.
A review of:
Despite the prevalence of this deadly viral infection, screening for HIV has historically targeted people considered to be at "high-risk." Sizing up an individual patient's risk, however, has been a tricky affair for clinicians -- who have been notoriously wanting when it comes to taking sexual histories. It is complicated by the movement of the epidemic to new populations, such as women, who may have relatively few obvious risk factors for infection other than occasionally having sex with a man.29
Clinicians had been previously advised by the CDC to offer HIV testing to only those patients who they considered at risk or more broadly if the prevalence of HIV in the area was 1% or higher.30 The drawback with this suggestion is that most primary care physicians have no ready access to local HIV prevalence data.
As HIV incidence has remained essentially stable over the past years, but has been on the rise in some groups, it has become evident that current screening practices are not working. Following a series of consultative meetings over the last few years, the CDC overhauled its HIV testing recommendations.31 The major revisions include:
For pregnant women, the CDC recommends:
Clearly, attracting the most attention are the recommendations that HIV testing should be offered widely and written informed consent need not be obtained prior to testing. As the epidemic has shifted into a diversity of populations, a significant proportion of HIV-infected persons do not necessarily practice behaviors that could be considered high-risk. Recent studies have described relatively few risk behaviors among some HIV-infected African Americans in the southern United States, especially those who are female.29 Therefore, broad testing will expand testing beyond those who consider themselves, or are deemed by their clinician, to be "at-risk."
The dropping of both the written consent and the extensive counseling prior to HIV testing reflects the experience of clinicians who feel that these procedures are an obstacle to testing.31 Importantly, it must be clear that it is not consent that is eliminated. According to the recommendations, patients still must be informed that they are being tested for HIV and provided an opportunity to opt-out of that testing. But, this information now can be provided verbally.
An additional major departure from prior versions of the CDC recommendations is the suggestion that the results of HIV testing be reported in the same manner as other diagnostic tests. Most clinicians, however, prefer to break the news of HIV infection in person in order to provide emotional support for the patient.
The Bottom Line
Expanding and simplifying HIV testing will surely lead to an increase in testing and is intended to reduce the number of persons unaware of their serostatus. The ultimate goal is to identify people with HIV infection so that appropriate counseling and treatment can be applied to reduce their infectiousness. These bold changes have, not unexpectedly, sparked some controversy.32 Although the potential benefits of expanded and routine HIV testing both to the individual (in terms of detection of HIV earlier in the disease course) and to public health (reduced HIV transmission) have been predicted by published models,33,34 there are concerns that these benefits come at the expense of an individual's autonomy.
Critics point out that HIV continues to be a stigmatizing infection and that the weakening of the consent process heightens the risk of patients being tested without being fully aware of the potential consequences and, in the worst-case, without their consent. There is an additional concern that the increased detection of HIV infection will tax the already overburdened U.S. health care system (including the AIDS Drug Assistance Program [ADAP], which already has capped enrollments and waiting lists in some states35) and lead to increased demand on the limited resources available to care for those currently already known to be infected.
My own view is that we can no longer ignore undetected HIV infection. "Routinizing" HIV testing, I think, is a step toward reducing the stigma associated both with HIV testing as well as being HIV-infected. Certainly, at least some of the people who remain undiagnosed are unintentionally infecting others.
The unwavering incidence of HIV infection in the United States justifies a revamped effort to identify those with chronic and acute HIV and testing is the lynchpin to making headway in reducing HIV transmission. While any resultant surge in HIV diagnoses may increase the short run utilization of care, if these efforts are successful, we will see a reduction in the numbers of persons living with the virus requiring care due to a decline in new HIV cases.
Unfortunately, at present, the CDC recommendations have had limited impact for most clinicians and patients. There are statutory or other regulatory impediments to their implementation, particularly the opt-out screening provision, and these will need to be dealt with before the fruits and/or fouls of the revised recommendations can be seen. But local or regional efforts are being made in emergency rooms and clinics across the country.
A review of:
The U.S. DHHS' revised Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents2 are almost always a shoe-in for inclusion on this top 10 list. But this year they take their place on the list because of how boldly they have embraced current HIV care in the United States.
As discussed above, for several years the guidelines have provided a list of preferred first-line HIV regimens that the clinical data best support. This list featured efavirenz and lopinavir/ritonavir as first-line therapy options -- along with certain combinations of NRTIs (zidovudine + lamivudine for either, tenofovir + lamivudine or emtricitabine with efavirenz). And so it remained until this latest update, which sparkles as if it has gone through a makeover on "Queer Eye for the Straight Guy."
Free from the shackles of evidence-based medical literalism, the guidelines cut some slack and recognize that some therapies, even if not yet rigorously tested in randomized, controlled studies against competitors, can produce similar results to the stalwarts of the preferred list. Added to the preferred list was fosamprenavir boosted with ritonavir (700 mg/100 mg twice daily) and atazanavir boosted with ritonavir (300 mg/100 mg). The "column B" NRTI choices have been simplified, appropriately, to either tenofovir/emtricitabine (TDF/FTC, Truvada) or zidovudine/lamivudine.
In addition to revamping the top regimen list, given the high rates of transmitted drug resistance being detected in cohort studies in the United States and Europe,36-38 the guidelines have also strengthened their recommendation that genotypic resistance testing be obtained in patients with chronic HIV infection. There were also cautions included regarding the use of lopinavir/ritonavir as monotherapy -- an experimental approach that has received much attention this past year39-41 -- as well as data regarding the use and adverse effects of the new PIs darunavir (TMC114, Prezista) and tipranavir (TPV, Aptivus).
The Bottom Line
Love them or hate them, they are our guidelines. Now there is more to love. The guidelines, always an excellent source of information on the principles of HIV therapy, have evolved to better reflect how HIV is, and should be, practiced.
The inclusion of fosamprenavir in "column A" of the preferred list is completely justified by the results from the KLEAN study (see above). The addition of ritonavir-boosted atazanavir was a bit more of a leap for the guideline committee. There has not yet been a large head-to-head trial of this boosted PI versus any of the other "column A" agents, although, as mentioned above, the maker of atazanavir is conducting just such a comparative study with lopinavir/ritonavir. Nevertheless, it is clear from BMS-089 (see above) that boosted atazanavir is at least as good as unboosted atazanavir, which was found to be similar to efavirenz in viral efficacy in one large study.42
In short, the U.S. DHHS guidelines are a must-read for anyone caring for HIV-infected patients (at least pages 5 through 28). The tables and figures alone are fantastic. The shot of reality that has been injected into this important document will help the many clinicians who rely on them to guide their care.
Click to enlarge
Adapted from "Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents," U.S. Dept of Health and Human Services. 2006.
It seems as though there can never be enough HIV medications. For one thing, the prevalence of HIV is increasing -- and there is nothing like expanding markets to motivate product development. In addition, none of the current batches of antiretrovirals have been able to provide long-term viral suppression for HIV-infected patients who live chaotic lives punctuated by periods of suboptimal adherence. As a result, regimens fail patients and patients fail regimens and the need for additional drugs that are more durable and can be active against resistant virus grows.
During the past year, the only new medication approved for the treatment of HIV infection was darunavir, a potent PI that is currently being used in patients with prior antiretroviral experience. In 2007, it is likely that three additional agents will be approved:
Each of these agents is currently available in the United States through expanded access programs and offers patients with multi-drug resistant virus another chance at controlling their HIV.
A review of:
Potent efficacy of MK-0518, a novel HIV-1 integrase inhibitor, in patients with triple-class resistant virus: 24-week data. B Grinsztejn, B-Y Nguyen, C Katlama, J Gatell, A Lazzarin, D Vittecoq, C Gonzalez, J Chen, C Harvey, R Isaacs, and Protocol 005 Team. In: Program and abstracts of the 46th Interscience Conference on Antimicrobial Agents and Chemotherapy; September 27-30, 2006; San Francisco, Calif. Abstract H-1670b.
MK-0518 (known generically now as raltegravir [Isentress]) inhibits HIV integrase, an enzyme that is tasked with the incorporation of viral genetic material into human cellular DNA. There is a long history to the development of inhibitors of HIV integrase. Previous efforts have been plagued by a variety of problems, including toxicity. MK-0518 has distinguished itself both in monotherapy and salvage treatment trials by its potency and tolerability.43,44 In addition, available data45-47 suggest that this agent will have few, if any, interactions with other HIV medications and, as if that was not enough, MK-0518 does not require boosting with ritonavir.
Two major clinical trials of MK-0518 studying treatment-naive and treatment-experienced patients were presented in 2006 and have generated a great deal of excitement.
As expected, few patients (14%) in the placebo arm achieved a viral load below 50 copies/mL by week 24. In contrast, 65%, 57% and 67% of patients who were taking MK-0518 at doses of 200 mg, 400 mg and 600 mg, respectively, achieved a viral load below this benchmark. As has been seen in other studies of salvage therapies, enfuvirtide use was associated with greater virologic response rates compared to when it wasn't used. In this study, the co-administration of enfuvirtide bumped the proportions of patients with a viral load below 400 copies/mL by about 20% in the MK-0518 arms. Once again, MK-0518 was observed to be extremely well tolerated, particularly at the 400 mg dose, which is the dose selected for U.S. FDA submission.
MK-0518 is an exciting agent belonging to a new class of antiretrovirals. Its potency, tolerability and lack of drug-drug interactions make it an attractive option for treatment-experienced patients who are in need of a critical mass of salvage agents. Further studies, including those presented at CROI 2007, will help identify the role of this drug in the management of HIV-infected patients with less treatment experience.
A review of:
Impact of NNRTI and NRTI resistance on the response to the regimen of TMC125 plus two NRTIs in Study TMC125-C227. B Woodfall, J Vingerhoets, M Peeters, I Peeters, G De Smedt, GD Miralles. In: Program and abstracts of the 8th International Congress on Drug Therapy in HIV Infection; November 12-16, 2006; Glasgow, United Kingdom. Abstract PL5.6.
These are encouraging results, particularly given a previous study known as TMC125-C227 with results presented by Brian Woodfall at the 8th International Congress on Drug Therapy in HIV Infection in Glasgow, United Kingdom.50 TMC125-C227 compared TMC125 to a PI-based optimized background regimen. It demonstrated that patients who were randomized to the PI arm had a significantly greater likelihood of suppression of viral load than patients who were in the TMC125 arm. This result led to the early halting of TMC125-C227.
A review of:
In early studies, this CCR5 antagonist was found to be effective in reducing viral load when administered as 100 mg twice daily to patients with virus that was CCR5 tropic.51 Maraviroc has no activity against virus that binds to the CXCR4 co-receptor to gain entry into the T cell. A major concern regarding CCR5 antagonists is whether blocking of the CCR5 co-receptor will apply selective pressure that would lead to an outgrowth of CXCR4 tropic virus in patients who have mixed tropic HIV populations. Such a shift from CCR5 to CXCR4 tropic virus has been associated with the progression of HIV disease.52
At AIDS 2006, Mayer et al presented a study53 of 186 treatment-experienced patients harboring a mixed population of CCR5 and CXCR4 tropic viruses who were treated with maraviroc at one of two doses (150 mg daily or 150 mg twice daily) versus placebo along with an optimized background antiretroviral regimen. After 24 weeks, no difference between the arms was seen when it came to viral suppression (15.5% to 26.9% of patients achieved less than 50 copies/mL). However, there was a trend favoring the twice-a-day maraviroc arm.
Importantly, 12 of 52 patients in the twice-daily maraviroc arm and 12 of 57 patients in the once-a-day maraviroc arm had only CXCR4 virus detected at treatment failure compared to only two in the placebo arm. Fortunately, the CD4+ cell count gains were actually higher in the maraviroc arms than the placebo (60 to 62 cells/µL versus 36 cells/µL). This was reassuring since the concern was that CXCR4 virus could be more pathogenic, which would lead to a more rapid CD4+ cell count decline. Why this was not the case in this study still remains unclear. Additional data on maraviroc was presented at CROI 2007.
These results are interesting, because maraviroc was not found to be particularly outstanding in suppressing HIV viremia among patients with mixed tropic virus. In addition, more patients treated with maraviroc had CXCR4 tropic virus emerge at the time of failure. The silver lining was that the CD4+ cell count responses of those on the drug were actually better than patients who were receiving placebo. That said, many clinicians will be reluctant to use maraviroc in patients other than those with only CCR5 tropic virus detected on tropism assays.
That brings me to a new subject. With the arrival of drugs targeting the T-cell co-receptors, it is a foregone conclusion that patients and clinicians will start to care which way their virus swings: "Are you R5, R4 or bi?" To find out, they will order a tropism test. That sound you hear is the ringing of the cash registers at the laboratories slated to market tropism assays. Ca-ching.
The Bottom Line
New drugs are generally good, new classes even better. There should be no complaints about the lack of industry investment in HIV therapeutics in 2007. Remarkably, the pipeline continues to flow. The agents that will become available are needed by the veterans of our days of sequential monotherapy and victims of their own all too human lapses. Even patients with really bad-ass resistant virus will likely benefit from one and perhaps all of the new antiretrovirals of 2006 and 2007. What better note to end on?
To ignore the spread of HIV in prisons is growing more and more difficult. In the first few weeks of 2007, editorials in the New England Journal of Medicine54 and the Lancet-Infectious Diseases55 went on the record castigating the policies and practices of U.S. correctional settings that promulgate the spread of HIV.
Despite the fact that prisoners are the only group of individuals in the United States for which the provision of HIV care is mandated by law, the quality of that care varies across facilities and HIV prevention efforts in prisons and jails are often spartan.
The problem of HIV in correctional settings should be of interest to everyone in the HIV community. Even HIV clinicians not directly involved in correctional health care are affected, since most will see patients who have been or will be incarcerated. It is estimated that one in five HIV-infected persons pass through a prison or jail annually.56 The prevalence of HIV infection in correctional systems is several fold greater than in the general population.56
However, when most people think at all about corrections and the HIV epidemic, they conjure an image of prisons as incubators of HIV infection. Yet, the available data and clinical experience indicate that the majority of inmates with HIV come into prison with their infection.57 Unfortunately, there are few studies that have examined the spread of HIV and other sexually transmitted diseases within correctional facilities.
In April 2006, the CDC reported on an investigation that had been conducted within the Georgia Department of Corrections following the identification of 88 men who acquired HIV infection during their incarceration from 1992 to 2005.57 For almost half of these men, HIV was detected during a pilot program that ran from 2003 to 2005 during which voluntary HIV screening was offered annually to inmates. (It is unfortunate, if not tragic, that this pilot program has been subsequently halted.)
In the series of case control studies conducted by the CDC, it was found that the following factors were associated with HIV seroconversion during incarceration:
While the overwhelming majority of sex reported by the inmates studied was consensual, the association of HIV and low BMI may suggest the victimization of those less able to resist sexual assault. Sex with prison staff was reported by both cases and controls. Condoms (typically makeshift as they are considered contraband in almost all U.S. prisons) were rarely used.
The Bottom Line
These results are disturbing. Even though the CDC focused on one state during more than 13 years, we can extrapolate these numbers to the hundreds of correctional systems in the United States. These numbers suggest that the intramural spread of HIV within prisons is probably not as rare as once believed and is largely missed. That a brief pilot HIV testing program in one state prison system was able to quickly uncover HIV seroconversion during incarceration is telling, as was the response -- the testing program was discontinued.
Further, this study makes plain that we need to do more to protect men and women within prisons and jails, foremost by providing them with the same effective HIV and sexually transmitted disease prevention tool that is available to people who are not incarcerated: condoms. It is the experience of our colleagues in Canada and in Europe that condoms and other prevention interventions, including even needle exchange, can be used safely and effectively in correctional settings.
We need to get with the program. State and federal policy makers must take a leadership role in making U.S. correctional facilities models of HIV prevention interventions rather than dramatic examples of how to avoid dealing with HIV infection. The answer lies in ethical HIV testing procedures and sensible HIV prevention for all 2.3 million people in our country who are incarcerated. Maybe Georgia, where this study was conducted, should take the lead.
It was not that long ago that one could hear concerns that the provision of HIV therapy to resource-limited nations would be doomed to failure as people in these troubled countries would not take their medications, resulting in the development and spread of HIV drug resistance.
Of course, this turned out to be rubbish. Lots of people living with HIV in the United States, Europe and Australia were found to have their fair share of trouble adhering to their medications, and those in Asia and Africa were able to take their HIV drugs just fine, thank you, as long as it was affordable.
In an instructive analysis58 of the first 1,004 HIV-infected patients initiating triple-drug antiretroviral therapy in Port-au-Prince, Haiti, HIV treatment was found to produce results similar to those seen in developed nations. After one year of therapy, 87% of adult and adolescent patients remained alive and the median increase in CD4+ cell count was 163 cells/µL (IQR 77 to 251 cells/µL).
The majority of patients were treated with zidovudine + lamivudine + either efavirenz or nevirapine. Viral load data were available for only 100 patients who had been followed for at least a year. In 76% of these patients, viral loads were less than 400 copies/mL. The median weight gain at one year was 5.5 kg. Therapy was well-tolerated, with only 25% of patients requiring a change in their initial regimen. In over half of these patients, treatment was discontinued or modified for reasons not having to do with toxicity (such as disruption of supply, sexual activity by women of reproductive age, treatment failure and tuberculosis).
The Bottom Line
HIV therapy works. It works in resource rich countries such as the United States, Canada and countries in Europe, and, we are learning, it works everywhere else. It does have to be free or nearly so for the many people living in poverty -- whether in the Bronx or Burkina Faso -- so they can keep taking their therapies. This study demonstrates that the benefits of HIV treatment in resource-limited nations, despite poverty and conflict, were similar to those seen in resource-rich nations. These findings underline the absolute need for government and private funds to continue to invest in HIV therapy for the millions of persons living with HIV across the world.
Surgeons are often reluctant to operate on HIV-infected patients, citing concerns that such individuals are at increased risk for surgical complications. This is a continuing cause of frustration for many HIV clinicians who try in vain to explain to their surgical colleagues that patients with adequate CD4+ cell counts should have little or no greater risk of adverse consequences with surgery than HIV-uninfected patients.
Over the past several years, an increasing number of surgeons have recognized the benefits of HAART and have agreed to perform scheduled surgeries in patients with controlled HIV infection. However, the criteria for acceptance by surgeons varies. Some surgeons insist that they'll only conduct surgical procedures on HIV-infected patients whose CD4+ cell counts are above 300 cells/µL, while other surgeons only accept patients whose counts are greater than 500 cells/µL for the same procedures. Unfortunately, until now there have been few data to which either surgeons or HIV clinicians can point to in order to make their case.
Michael A. Horberg et al from Kaiser Permanente in northern California conducted a case-control retrospective study of 332 HIV-infected and HIV-uninfected pairs of patients59 -- all of whom underwent the same surgery. Pairs were also matched by sex, age at surgery, medical center where the procedure was performed and year of the operation. Outcomes included complications and mortality through 12 post-operative months of follow-up.
The most common surgery was hernia repair (44%) followed by appendectomy (17%), cholecystectomy (14%) and a variety of other procedures. The median CD4+ cell count of the HIV-infected patients was 379 cells/µL and over 60% of the patients had a viral load less than 500 copies/mL.
Surgery-related complications were found to be no more frequent among the HIV-infected patients than the controls (11.1% versus 10.2%). However, non-Pneumocystis pneumonia was observed in eight HIV-infected patients compared to one control patient (P = .04), and 10 HIV-infected patients died compared to only two HIV-uninfected patients (P = .04).
The mean time from surgery to death for the HIV-infected patients who died was 180 days (range: 6 to 318 days). Of the 10 deaths among the HIV-infected patients, none appeared to be a consequence of the surgery, and only two patients died within 30 days of their operation (one as a result of multiple organ failure secondary to lymphoma and the other sudden death was due to arrhythmia).
A CD4+ cell count less than 50 cells/µL was associated with a greater risk of post-operative complications compared to a count above this level. There was no evidence of an increased complication rate among those with a CD4+ cell count of 50 to 199 cells/µL compared to those with higher counts. Similarly, a viral load above 30,000 copies/mL was also predictive of complication postoperatively.
The Bottom Line
This study suggests that the majority of HIV-infected patients can undergo surgical procedures with little increased risk of surgical complications. Patients with the most profound immunosuppression, as evidenced by a CD4+ cell count below 50 cells/µL, however, do face a heightened risk of problems after a surgical procedure. In short, except in the case of such advanced patients, HIV status should not be a criterion used to determine whether a patient should have a surgical procedure. Keeping a copy of this study handy to share with a surgical consultant may be prudent.
Copyright © 2007 Body Health Resources Corporation. All rights reserved.
This article was provided by TheBodyPRO. It is a part of the publication HIV JournalView.