A STEP Conference Review
10th Conference on Retroviruses and Opportunistic Infections
Suniti Solomon, M.D. provided an eye-opening talk about HIV in women, particularly in the cultural setting of Madras, India, where she is from. Over 40 percent of HIV infections now occur among women, and a very high proportion of them are contracted through heterosexual sex. In India, women are highly vulnerable to HIV and other sexually transmitted diseases (STDs) due to economic, social, cultural, and legal discrimination as well as basic biological and anatomical differences. Women's economic dependence on men often results in poorer healthcare, at times delaying the treatment of genital tract infections that increase HIV susceptibility. The lack of equal opportunity for women to receive education regarding sex and HIV sometimes results in the reliance on cultural myths for guidance. Dr. Suniti used the example of the belief that "sex with a virgin cures a male of all STDs." Women with poor social and job skills offer sexual services in return for the social and economic support of men. Social norms encourage sexually "innocent" women to willingly participate in sex "for" their husband even if he is HIV-infected. Women do not have the power to enforce safe sexual practices. Plus, fertility pressures force women to abandon caution when having sex with a known HIV-infected partner, since barrenness is a cause for divorce in India. These women are more likely to stay within a marriage no matter how vulnerable they are to infection, since an unmarried female is considered worthless. According to Dr. Solomon, "marriages are saved at the cost of HIV." The impact of HIV on women in many countries around the world is much greater than that on men due to their lower status. HIV prevention programs need to focus on ways to empower women, change gender norms and recognize the barriers to safe sex options for women. Dr. Suniti concluded her talk by saying, "It is not flattering that it takes a ruthless epidemic to awaken the world to the needs and condition of women."
Focusing on women with HIV in the U.S., Ruth M. Greenblatt, M.D., discussed the findings from the Women's Interagency HIV Study (WIHS). AIDS cases among women in the U.S. continue to increase at a rate that exceeds that of other demographic groups. WIHS is the largest cohort study of women living with HIV in the U.S. and has, over the past 10 years, helped identify the significant features of HIV infection in women. The more than 2,500 WIHS participants are a racially diverse group of generally poor women with high rates of co-morbid conditions such as hepatitis C, herpes infection, depression, and domestic violence. The key findings in this female cohort were as follows: 1) the viral load was lower than among the male participants of the MACS study when stratified by CD4 cell count. 2) The viral load varied with the ovulatory cycle. 3) Loss of subcutaneous fat in the extremities is common among HIV-infected women, whether antiretroviral treated or untreated, but not among uninfected women. 4) Low-grade cervical intraepithelial lesions are as common as anal lesions. 5) Switching and discontinuing HAART are both common. 6) Adverse events and virologic rebound during HAART is common. Studying the female HIV-infected population will continue to broaden our knowledge of HIV and determine the unique differences of the infection in men and women.
Historically, women have been excluded from many large randomized controlled trials to avoid dealing with a variety of presumed difficulties related to being female. Kathleen E Squires, M.D., reported that the Women First cohort verified that women are willing to participate in antiretroviral therapy (ART) studies and can take ART just as easily as men. It also confirmed that men and women have a similar virologic and immunologic response to ART. Several clinical trials have shown no significant difference between men and women in the percentage achieving an undetectable viral load, the amount of increase in the CD4 cell count, and the time to virologic response or failure.
However, women may differ from men in respect to drug toxicities and HIV complications due to lower body weight, altered hepatic metabolism, higher body fat content, hormonal differences, pregnancy effects, and drug-to-drug interactions. For example, in a therapeutic drug-monitoring study involving indinavir, men and women had similar blood concentrations of drug, yet women had more adverse effects resulting in more dose reductions. ACTG 359 noted that Fortovase clearance was 50 percent less in women, which was directly related to weight. ACTG 175 revealed increased rates of dose modification for Videx, also due to weight differences between genders. Certain antiretrovirals such as Viramune, Viracept, and Norvir are not recommended in combination with oral contraceptives due to decreased contraceptive blood levels. It appears that women have increased rates of pancreatitis, decreased risk of triglyceride elevations, increased hepatic steatosis and lactic acidosis, increased risk of fat accumulation, possible decrease in bone mineral density and increased risk of rash (e.g., with NVP). While it is important to keep these differences unique to women in mind when selecting an ART regimen, these differences should not preclude women from participating in trials.
Jared M. Baeten, Ph.D., from the University of Washington, presented data collected from a cohort of female sex workers in Mombassa, Kenya. Since 1993, multiple research trials have been conducted by the University of Washington group to determine modifiable risk factors for HIV-1 acquisition, infectiousness and disease progression. At the 10th CROI, Dr. Baeten discussed the influence of vitamin A and hormonal contraception on HIV transmission and disease progression in women.
Vitamin A deficiency is a common nutritional deficiency in HIV infected individuals in Africa. Observational studies have associated vitamin A deficiency with decreased CD4 cell counts, faster disease progression, increased mortality, increased genital shedding of HIV, and increased risk of HIV transmission. In an attempt to validate this association, 400 HIV-infected women were randomized in a 6-week double blind, placebo-controlled trial of vitamin A supplementation. Unfortunately, vitamin A supplements had no effect on clinical outcomes, CD4 cell counts, or viral load in plasma or vaginal secretions, even among the 60 percent of persons who had documented vitamin A deficiency at baseline. There did not seem to be any impact of supplementation on infectiousness or disease progression. It was concluded that HIV disease is likely a cause of vitamin A deficiency (vitamin A levels may reflect more active HIV infection and be an acute phase reactant), but not visa versa. Dr. Baeten's research group also conducted a small trial using supplementation with a multivitamin, and found a surprising two-fold increase in vaginal and cervical shedding of HIV in the multivitamin group when compared to controls. This indicates that something as innocuous as a multivitamin may increase HIV transmission rates, and studies need to be done to assure the safety of all supplements. At this time, nutritional interventions are of uncertain benefit, despite compelling observational evidence suggesting potential.
Studies were also conducted to determine the effects of hormonal contraception on HIV. Since more than 100 million women use oral contraception worldwide, it is important to determine whether any systemic or genital tract changes secondary to hormonal contraception influence HIV disease. Previous studies reported that women using hormonal contraception had a higher risk of HIV acquisition. Dr. Baeten's research group collected data about contraceptive use in a prospective cohort study of more than 1,500 HIV seronegative female sex workers and explored whether oral contraceptives at the time of acquisition influenced subsequent HIV disease. Despite receiving counseling, the women showed high HIV seroincidence (8.5 per 100 persons), which led during the study period to a total of 250 seroconversions; of these, 161 women were eligible for the study. It was found that the use of hormonal contraception, both oral and injectable, is associated with an increased risk of HIV-1 acquisition. Hormonal contraceptive use was associated with early HIV genetic diversity, leading to a persistently higher viral set point, in turn resulting in a faster decline in the CD4 count. It was hypothesized that hormonal contraceptives may thin the genital epithelium, thereby increasing the risk of virus establishment, and may increase the number or permissiveness of susceptible cells. However, given the benefits of using oral or injectable contraception to avoid unwanted pregnancies, further studies to need to be done to validate the data and better determine the risk versus benefit ratio.
When and What Antiretroviral Therapy to Start
Martin S. Hirsch, M.D., from Massachusetts General Hospital in Boston, discussed what is currently known about starting antiretroviral therapy (ART). Everyone agrees that ART is indicated if a person has symptomatic HIV, a CD4 count less than 200 cells/ml, or a high HIV viral load (i.e., greater than 100,000 copies/mL). There is some support for the treatment of acute HIV infection, but the hope of maintaining HIV-specific responses has not been upheld in randomized controlled trials. The treatment of HIV during chronic asymptomatic infection, with a CD4 count greater than 200-cells/ml, however, still remains controversial.
The benefits of treating persons with a higher CD4 count include controlling viral replication, decreasing pathogenic mutants, potentially decreasing the risk of lipoatrophy due to HIV, and decreasing rates of immune reconstitution syndrome, but this is at the expense of decreasing quality of life, increasing drug resistance, creating more toxicities and increasing costs. Studies have clearly shown that waiting until the CD4 count is below 200 cells/mm3 results in worse clinical outcomes, but we do not know what CD4 count is optimal to start treatment. Some studies suggest a clinical outcome difference if a CD4 count of 350 cells/mm3 is used as the threshold for starting treatment, but other studies do not. The Swiss Cohort Study, a nested case control study of 283 asymptomatic ART-naive HIV-positive persons, found that using a CD4 count of 350 cells/ml as the threshold for initiating treatment delayed death, decreased the development of AIDS, and symptomatic clinical disease. The HOPS Cohort suggested a benefit of treatment when CD4 counts were between 200 and 350 cells/mm3. The Hopkins cohort study also showed a possible advantage of treating patients with CD4 counts below 350 cell/mm3.
The current International AIDS Society-USA guidelines state that in persons with asymptomatic HIV, if the CD4 cell count is less than 200 cells/mm3 then treat, if the CD4 cell count is between 200 and 350 cells/ml consider treatment, and if the CD4 cell count is above 350 cells/ml then delay treatment. Given the uncertain benefit, the decision of whether to start treatment when the CD4 count is between 200 and 350 cells/ml needs to be individualized. Factors to consider include the rate of CD4 decline (if it is greater than 100 cells per year, start treatment), a person's interest in starting ART, the HIV viral load (if it is greater than 100,000 copies/µL, start treatment), and the person's risk of toxicity based on other co-morbidities. The DHHS guidelines add that women who are pregnant should be treated regardless of their CD4 count to decrease vertical transmission rates. There clearly is a continuum of risk related to progressive viral replication and immune dysfunction. Dr. Hirsch concluded that "as therapies improve the pendulum will likely swing to earlier treatment."
What to start as the initial ART regimen in a treatment-naive HIV-positive person is a little more complicated. However, there is an increasing wealth of data comparing the current possible options available. ACTG 384 looked at 980 HIV-positive persons stratified by country and viral load and found that for the primary endpoint (time to salvage therapy) AZT + Epivir (Combivir) was better than ddI + d4T when used with Sustiva, but not with Viracept. Sustiva was better than Viracept only when used with AZT + Epivir. Overall the best combination found in this study was AZT + Epivir + Sustiva and Dr. Hirsch felt that this regimen should be the standard to which all other treatment regimens are compared. Toxicity data for the combination of AZT + Epivir was better than ddI + d4T and Sustiva was better than Viracept. Dr. Hirsch, therefore, stated that ddI + d4T should probably not be a first-line choice in the initial ART regimen. The other interesting finding in this study was that the time to virologic failure for the four-drug regimens (Sustiva + Viracept with AZT + Epivir or with ddI + d4T) was not significantly different than two consecutive three-drug regimens. From this data, Dr. Hirsch concluded that adding additional drugs to the traditional three drugs used for the initial antiretroviral regimen increases toxicity but not potency and therefore should be avoided.
The eagerly awaited 2NN study (see related article in this issue), presented by the International Antiviral Therapy Evaluation Center (IATEC), compared the efficacy and safety of Sustiva and Viramune in a large-scale, randomized head-to-head trial. In this 48-week analysis, 1,216 people were randomized to either Viramune 400 mg once daily, Viramune 200 mg twice daily, Sustiva 600 mg once daily, or Sustiva 800 mg + Viramune 400 mg once daily. All persons received a HAART backbone of Epivir + Zerit (d4T). There were no significant differences in treatment success (viral suppression or change in the CD4 count) among the single NNRTI arms. The two Viramune arms had more liver toxicity, but a better lipid profile than Sustiva. Surprisingly, Viramune and Sustiva alone not only had less toxicity, but also were clinically better than the combination. In conclusion, two NNRTIs are not better than one. Also, there was not a significant advantage in viral suppression with Sustiva compared to Viramune, but there was a trend towards better suppression. Additionally, it appears that once a day Viramune is as effective as twice a day, but may have more side effects.
The Phase III trial (M98-863) comparing Kaletra to Viracept compared these two antiretrovirals in combination with Zerit and Epivir. The virologic results favored Kaletra/r over Viracept at 48 weeks. The Viracept failures were due to resistance mutations. Kaletra though did have more elevations in triglyceride levels. The Gilead 903 study compared Viread with Zerit when combined with Epivir + Sustiva. At 96 weeks there was comparable antiviral activity and increases in CD4 counts between the two groups. However, the lipid profile favored Viread over Zerit. Zerit also had more toxicity, including peripheral neuropathy, Lipodystrophy, and lactic acidosis. The information we have learned to date from the existing head-to-head comparison trials of ART in treatment-naive persons is summarized as follows:
Other options for first-line therapy in ART-naive persons include a triple-NRTI regimen using Trizivir (AZT + Epivir + Ziagen), which has a low pill burden and few interactions, but concerns about the potency of this regimen still exist (see report in this issue on ACTG study 5095) and the risk of abacavir hypersensitivity. Atazanavir, which is likely to be approved this year, may also be an option for initial therapy (see "Ask Dr. Jeff" in this issue). BMS 034 found that Atazanavir has efficacy similar to that of Sustiva, but a better lipid profile. However, a prior study showed Atazanavir to have efficacy similar to that Viracept, and as noted above, ACTG 384 found Sustiva to be superior to Viracept. Therefore, further studies are needed, especially since the Sustiva group in the BMS study had poorer outcomes than those usually seen, possibly making Atazanavir look better than it actually is. It is important to keep in mind that the benefit of any single drug depends in part on how it is combined with other antiretrovirals in a HAART regimen.
When to Switch Antiretroviral Therapy
Stephen Deeks, M.D., from the University of San Francisco, presented his expert opinion about switching ART in persons experiencing virologic failure, because no clinical studies have specifically addressed this topic. We know that drug resistance emerges in treated persons with ongoing detectable viral replication (i.e., a viral load greater than 50 cells/mL). Therefore, if a patient has a detectable viral load and options exist that are likely to suppress the virus to undetectable levels, then the current regimen should be modified as soon as possible while trying to preserve future treatment options.
However, not everyone who is failing their current HAART regimen has enough remaining active drugs available (due to resistance mutations) to completely suppress the viral load, resulting in very difficult treatment decisions. Such persons have one of three broadly defined options: switch to an aggressive regimen with a goal of maximal viral suppression, interrupt all therapy, or continue a partially effective regimen despite the presence of drug-resistant HIV. There are many reasons why one would not want to switch a failing HAART regimen even if an optimal regimen does exist. Salvage regimens are often complex, with significant toxicities that can outweigh the benefits. Economic costs are high for these complex therapies. If a person has resistance to the three existing drug classes, it might be worth saving new drugs like T-20 until additional therapeutic classes are available to combine with. As seen in the T-20 trials, only 9 percent of people who added T-20 as the only active drug to their failing regimen attained an undetectable viral load. Plus, many persons do well for years despite incomplete viral suppression and often continue to have a good CD4 benefit.
Drug-resistant virus has been shown to have a decreased replicative capacity and depletes CD4 cells slower than wild-type (WT) virus. The assumption is that archived WT virus emerges once ART is stopped because it is more "fit" than the virus with multiple resistance mutations. People who remain on a failing regimen, often due to a lack of good options, have a sustained lower viral load set point, compared to pre-treatment. It is inferred from this persistent lower viral load that the existing resistance to the current HAART regimen is only partial. After HAART is discontinued, the viral load continues to be partially suppressed until the virus reverts back to WT, at which point there is a dramatic decrease in the CD4 count and an increase in the viral load. Continued treatment during incomplete viral suppression is of benefit by preventing WT virus from rebounding, thus maintaining less fit, less virulent virus. CD4 cell counts and HIV-specific T-cell activity are often preserved, perhaps due to the altered virulence of the drug-resistant variant. However, partial viral suppression is at the risk of continued viral evolution and increasing the number of resistance mutations.
Dr. Deeks proposed a very interesting possibility for patients who are currently on a failing regimen. A failing regimen usually consists of antiretrovirals from two or more drug classes. By simplifying the failing regimen to only one class of antiretrovirals, the prevention of WT rebound would still occur (maintaining decreased "fitness") while decreasing toxicity and cost. In addition, limiting the regimen to one drug class would remove the selective pressures of the other drug classes, minimizing the evolution of additional resistance mutations. Dr. Deeks found through a clinical trial that it is the NRTI class that accounts for the partial activity of a failing regimen and not the protease inhibitors. Persons on a failing regimen of Zerit + Ziagen + Sustiva + Viracept stopped the Viracept and for at least 24 weeks the viral load did not increase and there was no shift in the resistance panel (the PI mutations remained despite stopping Viracept). In conclusion, in persons on a failing regimen awaiting new treatment options, it may be safe to stop all antiretrovirals in the current regimen except the NRTIs. Further studies need to be conducted to verify the effectiveness and safety of this strategy.
When to Interrupt Antiretroviral Therapy
Huldrych Günthard, M.D., reviewed the current rationale for interrupting HAART therapy; this rationale includes strengthening immune responses to HIV, decreasing toxicity and cost, and possibly increasing salvage options. However, treatment interruption is also associated with risks that include increasing drug resistance, decreasing CD4 cell counts, increasing HIV transmission rates, reseeding viral reservoirs and developing acute antiretroviral syndrome. Treatment interruptions have been studied in persons who initiated treatment during both acute and chronic HIV infection.
Some researchers believed that if the immune system is repeatedly exposed to HIV, it would respond by creating a boosted immune reaction to HIV, keeping the virus better controlled without drug therapy. This hypothesis of using "autovaccination" to decrease the viral set point by increasing protective immunity unfortunately did not pan out in clinical studies. HIV-specific cellular immune responses are induced, but only to pre-treatment levels and they do not correlate with control of viremia. Therefore, this is no longer a good rationale for treatment interruption.
Given the current antiretroviral options and the increasing HIV epidemic in developing countries, the idea of interrupting HAART to reduce toxicity and cost is very enticing. J. Ananworach in the HIV-NAT 0014 study compared continuous therapy (Arm 1) with two types of structured treatment interruptions (STIs). The first type of STI is CD4-guided (stop if the CD4 count is greater than 500 cells/ml or rises 30 percent; restart if the CD4 count is less than 350 cells/ml or falls 30 percent) treatment (Arm 2). The second type is short-cycle (week on-week off) intermittent therapy (Arm 3). The 74 Thai participants had all previously received 1 year of dual NRTI therapy followed by 3 years of HAART, with currently well-controlled HIV disease. During the 48-week study period, the CD4-guided and week on-week off strategies resulted in clinical adverse events, and quality of life outcomes that are comparable to continuous antiretroviral therapy. The proportion of persons with a CD4 cell count greater than 350 cells/mm3 was similar in all arms, although the CD4-guided treatment had the largest CD4 count decrease by design (Arm 1: +5, Arm 2: -178, Arm 3: -6). There were high rates of virologic failure in the week on-week off arm (Arm 1: 4 percent, Arm 2: 17 percent, Arm 3: 65 percent); however, all had a viral load less than 50 cells/ml later after resuming the same regimen. The previous sub-optimal antiretroviral therapy prior to HAART likely contributed to the high viral failure rates of the STI strategies, especially in the week on-week off arm. The biggest difference between the participants was that the people in the continuous therapy group spent 100 percent of time on medication, whereas the short-cycle intermittent persons spent 59 percent, and the CD4-guided group only 33 percent. CD4-guided treatment was the best cost saving strategy and had similar virologic outcomes to continuous therapy, but at the expense of lower CD4 cell counts.
Lidia Ruiz compared continuous HAART with STI guided by CD4 cell counts and viral load in a multi-center, randomized controlled clinical trial. The 122 persons had well-controlled chronic HIV disease. The criteria for restarting antiretrovirals in persons who interrupted therapy were a viral load increase of 100,000 copies/ml a drop in the CD4 cell count below 350-cells/ml, or the appearance of any opportunistic infection. Treatment was interrupted again if the CD4 count rose to 500-cells/ml and the viral load dropped below 80 copies/mL. At 48 weeks, all but two individuals on continuous therapy maintained an undetectable viral load, and the median CD4 cell counts remained stable. In the STI group, 43 percent remained off therapy at 48 weeks, with a median viral load of about 10,000 copies/ml. Of the 57 percent of persons who had to reinitiate HAART (96 percent of these were due an increase in viral load, 4 percent were due to a rise in CD4 cell count, and 23 percent were due to both reasons), all achieved re-suppression and 66 percent were later able to re-interrupt therapy. A median CD4 decrease of 96-cells/ml was observed in continuous therapy group compared to 335-cells/ml in the STI group. Acute retroviral syndrome was observed in 6 (10 percent) of the persons discontinuing therapy. When these results are looked at in combination with the HIV-NAT study, we can conclude that CD4-guided intermittent therapy is an attractive approach, but the optimal criteria for stopping and starting therapy still need to be determined.
Mark Dybul, M.D., from the National Institute of Allergy and Infectious Diseases (NIAID), previously evaluated short-cycle intermittent therapy, and although this strategy was found to reduce the costs and side effects of HAART while maintaining virologic control, it was difficult to adhere to the approach in the real world. At the 10th CROI, Dr. Dybul presented data for a similar, but possibly more reasonable strategy of long-cycle STI. Participants were randomized to continuous therapy or seven cycles of 4 weeks off followed by 8 weeks on. The controlled trial evaluated drug toxicity, virologic and immunologic parameters in 90 persons with a suppressed viral load and a CD4 cell count greater than 300-cells/ml. Enrollment was terminated early due to the emergence of resistance in 3 of the 8 participants in the STI arm receiving a Sustiva-based regimen and one person receiving a PI-based regimen. By the end of the 8-week on-therapy period, none of these participants was able to completely re-suppress viral replication that had rebounded during the off-therapy period. The periodic reductions in certain lipid levels during the off-therapy period may have clinical benefit; however, there was no immunologic or virologic benefit of long-cycle STI.
Two studies regarding the response to salvage therapy after an STI seemed to have conflicting results. First, Jody Lawrence, M.D., reported the data from the multi-center CPCRA 064 study, the first prospective randomized clinical endpoint study for patients with multi-drug resistant (MDR) HIV to test whether a 4-month STI prior to changing therapy reduces progression of disease compared with immediately changing therapy. The 270 study participants were HIV-infected with fairly advanced disease, on stable HAART regimens with virologic failure (viral load greater than 5,000 copies/mL) and MDR virus documented by genotypic testing (48 percent of participants had resistance to all three classes of antiretrovirals). The mean number of drugs in the new regimen was 3.6 (2.7 active drugs based on resistance testing) in the STI arm and 3.8 (2.8 active) in the no-STI arm. There were 22 persons with progression of disease in the STI group and 12 in the no-STI group. Interestingly, the majority of clinical events did not occur during the 4 months off therapy, but later, after it was re-initiated. There were eight deaths in each arm. The mean difference in CD4 cell counts favored the no-STI group. The STI group showed no virologic advantage compared to the control group despite reversion to WT virus in 64 percent of persons. Quality of life measures were similar for each arm. The authors concluded that in treatment-experienced persons failing therapy with MDR virus, STIs as used in this study do not appear to confer clinical, immunologic, virologic, or quality of life benefits. While STIs have been postulated as a means of enhancing virologic response to subsequent treatment, this did not occur. Plus, there was a persistently poorer CD4 response and higher number of clinical events in the STI group, resulting in early closure of the study.
The GIGHAART ANRS 097 French study presented by Christine Katlama also evaluated the benefit of STI in 63 highly drug-experienced persons with very advanced HIV disease (average HIV viral load of 200,000 copies/mL and CD4 cell count of 27 cells/ml) and MDR virus, but came to a very different conclusion. Persons were randomized to either immediate mega-HAART therapy or mega-HAART deferred for 2 months. Mega-HAART was a very aggressive regimen consisting of 7-9 drugs: 3-4 NRTIs, plus one NNRTI with or without hydroxyurea, plus two ritonavir-boosted PIs. Even though there were few reversions to WT virus during the short 2-month treatment interruption period, after starting the new regimen, the STI group had a greater median drop in viral load at 12 and 24 weeks compared to the immediate mega-HAART group. Greater CD4 cell count increases also occurred in the deferred mega-HAART group (51cells/ml at week 24 and 69 cells /ml at week 48) versus immediate mega-HAART (7 cells/ml at week 24 and 48). Tolerance was acceptable in both groups (22 percent in the immediate and 47 percent in the deferred mega-HAART group were still on a regimen of more than 6 drugs at week 48). The three major factors associated with virologic success were treatment interruption with reversion of resistance, adequate drug concentration, and the use of lopinavir. The authors concluded that a 2-month treatment interruption followed by aggressive multi-drug salvage therapy induces a significant virologic and immunological benefit lasting up to 48 weeks and that complete reversion to WT is unnecessary as long as the virus gains back some sensitivity during the STI period.
The apparently contradictory results of these studies may be explained by the following key differences between the studies. The participants in the GIGHAART study usually had no active drugs available and were subsequently placed on eight- to nine-drug regimens, while those in the CPCRA study did not have as much MDR-HIV, so they often still had two or three active drugs used as part of only a three- or four-drug regimen. The availability of more than two active drugs may be a reason to start immediate salvage therapy, because waiting for reversion to WT or less resistant virus may be unnecessary and not worth the risk of a treatment interruption. How aggressive, and therefore how complex and difficult, the optimal salvage regimen needs to be is also brought into question. The other difference between these two studies is that CPCRA interrupted therapy for 4 months before starting the salvage regimen, while the GIGHAART study waited only 2 months. It is possible that a 4-month interruption in therapy is too long in persons with such advanced HIV disease. Further studies are clearly necessary to completely reconcile these opposing results.
GB virus type C (GBV-C), previously known as hepatitis G, was renamed because it does not appear to cause any human disease -- including hepatitis -- despite its similarity to the hepatitis C virus. Once infected, most immunocompetent people clear the infection, but 20 to 50 percent of people have persistent viremia. Since HIV and GBV-C share the same routes of transmission (blood, sex, and mother-to-child), 20 to 40 percent of HIV-infected persons are co-infected with both viruses. However, co-infection with GBV-C is thought to lead to increased survival because GBV-C inhibits HIV replication. Jack T. Stapleton, M.D. found that the greater the amount of GBV-C virus present, the more it inhibited the HIV virus. This inhibitory effect is thought to be related to changes in cytokines and chemokines (the body's chemical balance) that cause a down regulation in the expression of CCR5 co-receptors, which are vital for HIV cell entry. This has led researchers to consider actively infecting all HIV-positive individuals with GBV-C to slow HIV disease progression.
P. Bjorkman looked at whether GBV-C viremia at the time of HIV diagnosis is prognostic in HIV-infected persons off HAART. Of the 230 HIV-positive participants, 27 percent had active viremia (GBV-C RNA was present), 30 percent had cleared infection (GBV antibodies were present), and 43 percent had never been infected (no GBV-C markers were present). The results did not reveal an association between GBV-C viremia and HIV clinical outcome. Interestingly, the loss of GBV-C viremia without development of GBV antibodies (seroconversion) was found to be associated with HIV progression and worse outcomes.
Prior studies have not controlled for the amount of time an individual was HIV or GBV-C infected. Carolyn Masters Williams, Ph.D., studied a population of 271 HIV-positive persons within 1 year of seroconversion and then re-examined the 137 surviving participants 5 to 6 years later. There was a better prognosis with persistent GBV-C infection; but again, clearance of the GBV-C virus was associated with a poorer prognosis. They did not, however, adjust the data for those persons who cleared their GBV-C infection prior to the start of the study. Due to the results of this study, Dr. Williams rightly has reservations about conducting a trial that would actively infect a person with GBV-C to control HIV disease. Although persistent virus may be beneficial, if the virus is cleared, it seems to be detrimental. It is not known why some people clear GBV infection, nor is it predictable which people will clear GBV infection.
Dr. Stapleton, on the other hand, has fewer reservations and foresees a trial of actively infecting persons with advanced disease who have few treatment options remaining. Yet, his data showed that the survival advantage of GBV-C co-infection mostly occurs when a person is infected with GBV-C before they contract HIV. Therefore, it is unclear how effective this intervention would be, unless a high-risk population was purposely infected with GBV-C before they even acquired HIV. All agree that it would be important to evaluate the role of GBV-C infection in long-term non-progressors. However, a better understanding of the mechanisms of GBV-C infection is necessary before the GBV-C can truly be considered a "safe" virus.
Two basic approaches have been used to address concerns about the relationship between heart disease and the elevated cholesterol, triglyceride, and body fat changes associated with HIV treatment. There are significant concerns that people on treatment for HIV will have accelerated hardening of the arteries, or atherosclerosis, which could result in an increased number of strokes and/or heart attacks. One approach has been to look for blood vessel changes prior to any clinical manifestations, so called sub-clinical changes. The second approach has been to look at large groups of people treated for HIV to determine if there are more strokes and heart attacks than expected.
One simple way to determine sub-clinical changes in the blood vessels is to look for thickening of the carotid artery, the large artery in the neck, with a very sensitive ultrasound instrument. The AIDS Clinical Trials Group (ACTG) reported on the initial findings of such a study, A5078. This study had a unique trial design. Groups of three people (triads) were recruited and matched for the following characteristics: age, sex, race/ethnicity, smoking status, blood pressure, and menopausal status. Group I consisted of HIV-infected people with continuous use of PI therapy for more than 2 years, Group II consisted of HIV-infected people without prior PI use, and Group III was HIV-uninfected people. People were excluded if they had known heart disease, diabetes, a family history of heart disease, uncontrolled hypertension (elevated blood pressure), or a body mass index above 30.
For this study, trained ultrasonographers at six sites sent standardized images to a central reading site for measurement. Carotid images were compared within the HIV-positive groups (I and II) and between the HIV-positive and HIV-negative groups (I/II and III). In a matched analysis that controlled for known risk factors for heart disease, there were no clinically relevant differences in baseline arterial thickening between people with elevated cholesterol or triglycerides, people with over 2 years of PI exposure, people who are PI-naive, and people who are HIV-uninfected. Follow-up is ongoing to determine whether rates of progression of carotid thickness are influenced by PI exposure and HIV infection over time.
However, a second study presented at the 10th Conference of Retroviruses and Opportunistic Infections (CROI) found conflicting results using a similar technique. Carotid artery thickness as well as inflammatory markers, lipid and lipoprotein levels were measured and followed over 1 year in people on therapy for HIV infection. Also assessed were cardiac risk factors, HIV disease characteristics, fat distribution, and body fat measurements. Progression of carotid thickening in the subset with 1-year follow-up was accelerated by ten-fold compared to HIV-uninfected populations. Among HIV-infected people, carotid artery thickness was independently associated with the classic coronary risk factors (age, cholesterol, and high blood pressure); low CD4 counts (at or below 200), and PI use. These data suggest that both immunodeficiency and traditional risk factors in addition to PI use potentially contribute to atherosclerosis in HIV-infected people.
How is it that two studies using a similar technique could have such different findings? The first study measured the thickness of the back wall of the carotid artery on the right and left side of the neck. The second study averaged the measurement of the front and back walls of six different sections of the carotid artery of the right and left side of the neck. So the techniques were not identical. Another difference is in how the studies attempted to control for underlying risks factors such as family history, smoking, cholesterol level, and activity level. Unlike the second study, the first study excluded the highest risk people -- those with a family history of heart disease -- because of the difficulty in matching subjects for this factor in addition to all the other risk factors they did match for at baseline. The rapid thickening of the carotid artery observed in the second study over the course of just 1 year is very concerning. However, these results are counterbalanced by the negative results found in the first study, but these are only baseline results. Therefore the final results of this first study, A5078, revealing the progression of arterial disease are anxiously awaited see if the negative results hold up over a 2 year period of time.
Two additional studies presented at the 10th CROI looked at large groups of people with HIV to determine if there are higher numbers of heart attacks or strokes than would be expected. The challenge in these types of studies is to know what the expected rate of these events is in the group of people you are observing, which is influenced by all the factors mentioned above in the carotid thickness studies. The first study, the "D:A:D" study, is a prospective observational study of 23,490 people from 11 cohorts in 3 continents. Enrollment was from July 1999 to April 2001, and follow-up was completed in February 2002. HAART use was associated with a 27 percent relative increase in the rate of heart attacks per year of exposure over the first 7 years of use. Further follow-up is important to monitor trends in heart attack rates beyond 7 years of HAART use. Factors that influence this association are under investigation. The study concluded: "Although of concern, the absolute risk of heart attack remains low and should be balanced with the known beneficial effects of HAART in the prevention of HIV-related complications."
A second study used a different design to look for clinical events including stroke and heart attack. This nested case control study assessed factors associated with stroke and heart attack. The control group consisted of people randomly selected from the same cohort who did not have stoke or heart disease, and were matched for cohort enrollment date and duration of follow-up. Of 2,671 people followed for 7,330 person-years (PYs) after January 1, 1996, there were 43 heart attacks and 37 stroke events for an incidence rate of 5.9 events per 1,000 PYs and 5.0 events per 1,000 PYs, respectively. Based on the National Health and Nutrition Examination Survey Epidemiologic Follow-up Study, the age-sex-race population rates of heart attacks and strokes would be expected to be 2 per 1,000 PYs and 3 per 1,000 PYs, respectively. Compared to national rates, the incidence rates of heart attacks and strokes in this cohort were approximately 2 to 3 times higher than expected. This increase in event rates was associated not only with expected cardiovascular risk factors, but also with antiretroviral drug use.
Dr. Lara Strick is a member of STEP's Board of Directors, and is completing her Fellowship in Infectious Diseases at the University of Washington.
Dr. Jeffrey T. Schouten is a former general surgeon who has been living with HIV for over 16 years. He is chair of STEP's Publications Advisory Committee and a primary care provider at Harborview's HIV Clinic in Seattle, Washington.
This article was provided by Seattle Treatment Education Project. It is a part of the publication STEP Perspective.