HIV in genital secretions
An important reservoir of transcription-competent virus
As suppression of viral burden to levels so low that HIV RNA cannot be detected by even the most sensitive commercial assays has become commonplace, researchers have begun to focus more attention on the dynamics of HIV replication in so-called sanctuary sites such as lymph tissue, seminal cells, and semen itself (see "HIV found in lymphoid tissue of patients with 'undetectable' viral levels," Vol. 3, No. 6, page 141).
A number of groups have compared viral-load levels and resistance patterns in isolates taken from both the semen and blood of study subjects. These reports suggest that antiretroviral therapy can significantly reduce the viral burden in the seminal cells of a large percentage of the men who respond well to combination therapy -- but the data also indicate that the response rate is never 100%. For example, Wong et al. found that HIV RNA was detectable in the genital secretions from 3 of 23 subjects treated with a combination of ZDV, 3TC, and indinavir for at least two years.1 All of these patients had serum RNA levels less than 500 copies/mL blood for at least 72 weeks.
Eron et al. evaluated 37 men (median baseline HIV RNA: 4.4 logs in blood and 3.45 logs in semen) randomized to receive amprenavir (APV) alone or APV with ZDV and 3TC.2 By eight weeks, 14 of 19 evaluable men on monotherapy had undetectable HIV RNA in semen, compared to 7 of 9 patients on triple-drug therapy. Four men had baseline viral load values in excess of 1,000,000 copies/mL semen, and they continued to have detectable viral levels after 16 to 24 weeks of treatment.
These authors found poor correlation between viral-load values in blood and semen -- as did Schacker et al., who studied 34 men with untreated HIV infection and found that viral levels in plasma did not correspond to levels found in genital secretions.3 Eron's group concludes that maximally suppressive antiretroviral therapy can indeed lead to dramatic reductions in the levels of HIV in genital secretions, but that a subset of patients are "hyper-excretors." Such patients present a particular therapeutic challenge -- and a particular risk to their sexual partners.
By contrast, two other studies did find a positive correlation between blood and semen viral-load levels. However, one of the two studies also found no association between semen viremia and disease stage, CD4 count, age, or race.4 The other study showed that viral loads in semen tend to be roughly one log lower than levels in blood.5 While this group showed that combination therapy suppresses viral load in semen, the researchers also observed that multidrug treatment does not decrease the rate at which HIV can be cultured from seminal cells: HIV was cultured at similar rates from the seminal cells of men with detectable and undetectable viral loads.
Yet another group of researchers has isolated viral strains from the genital tracts of men who have a predilection for macrophages.6 This is an ominous finding, as macrophage-tropic strains of HIV are implicated in sexual transmission of the virus.
Dr. William G. Powderly, a member of the editorial advisory board of HIV Newsline, comments:
Taken together, these small studies underscore the importance of safer-sex counseling for all sexually-active seropositive patients. These investigators have conclusively demonstrated that men with serum viral loads that are below the level of detection of any of the widely used commercial assays may nonetheless have plenty of transcription-competent virus in the genitourinary tract. There is growing evidence that this system represents a reservoir for HIV, one that antiretroviral agents penetrate with widely varying degrees of success. These patients may harbor macrophage-tropic strains of HIV in latently infected cells, strains that are capable of producing primary infection. Many seropositive individuals assume that drastic reductions in viral burden bring with them concomitant reductions in transmission risk, and practitioners must make a particular point of disabusing their patients of this dangerous notion.
The fact that we are witnessing a nationwide rise in the incidence of all sexually transmitted diseases is compelling evidence of increasingly laxity in observing safer sex guidelines. It is therefore imperative that care providers reinforce public health messages concerning HIV transmission, especially in the communities that continue to bear the brunt of the epidemic.
Taken together, these small studies underscore the importance of safer-sex counseling for all seropositive patients. Individuals with serum viral loads that are below the level of detection of any of the widely used commercial assays may nonetheless have plenty of transcription-competent virus in the genitourinary tract. There is growing evidence that this system represents a reservoir for HIV, one that antiretroviral agents penetrate with widely varying degrees of success. Public-health messages need to reflect this new and sobering information.
Ritonavir/saquinavir as an alternative for patients who fail nelfinavir
A second-line option in treatment-experienced individuals
Although three-drug antiretroviral regimens that contain a protease inhibitor typically produce prompt and profound suppression of viral replication, there is growing evidence that a significant proportion of these patients do not achieve durable viral suppression with these therapies. Prior treatment, advanced immunodeficiency, and poor adherence are the most common causes of virologic failure -- which, in some studies of heavily pretreated patients, has affected as many as half of those individuals (see "Protease Inhibitor Resistance and Salvage Therapy," Vol. 3, No. 6, pages 132-136).
The question of what successor therapy to choose for patients who break through on their initial protease-inhibitor-containing regimen has been especially problematic, because individuals who develop high-level resistance to one of the protease inhibitors generally exhibit broad resistance to this entire class of antiretroviral agents. Early evidence suggested that HIV strains which had become resistant to indinavir and ritonavir would be cross-resistant to nelfinavir, but nelfinavir-resistant strains might remain susceptible to the other protease inhibitors -- and this led many clinicians to assign nelfinavir as first-line therapy in protease-inhibitor-containing regimens.
The first pilot studies that investigated the benefits of switching nelfinavir-resistant patients to the two-protease combination of ritonavir and saquinavir produced equivocal results, but the data from several recent trials suggest that this combination may be an effective successor regimen in patients who fail nelfinavir therapy. When the two protease inhibitors are taken in combination, ritonavir acts as both a pharmacotherapeutic and a pharmacokinetic agent, boosting the bioavailability to chemical saquinavir as much as forty-fold by inhibiting the breakdown of that protease inhibitor (see "Enhancing saquinavir levels with ritonavir," Vol. 3, No. 3, pages 60-62).
Tebas et al. used ritonavir/saquinavir as a salvage regimen in patients who had rebound in HIV viral load while on nelfinavir.1 The median time of nelfinavir treatment prior to entry into the study was 55 weeks for the 26 patients enrolled. At baseline, median viral load was greater than 46,000 copies/mL. All of the study subjects were recruited from two clinical trials, one of which examined the combination of nelfinavir with d4T, the other, nelfinavir with ZDV and 3TC. Inclusion criteria stipulated two consecutive viral-load measurements above 5000 copies/mL while on nelfinavir therapy. In this open-label study, patients received the following dosages:
Two patients dropped out of the study due to GI intolerance and were not included in the analysis. A dramatic decrease in viral load was seen in all of the remaining participants. At the 24 week mark, 17 of 24 patients had HIV RNA measurements less than 500 copies/mL. This degree of viral suppression was maintained in 14 individuals over 48 weeks -- a response rate of 58% in patients who had failed nelfinavir therapy. Not surprisingly, a baseline viral load less than 30,000 copies/mL was associated with a greater likelihood of response to ritonavir/saquinavir. The mean CD4 cell increase was approximately 120 cells/mm3 at 48 weeks.
A corollary question is how ritonavir/saquinavir compares to indinavir as first-line therapy. Florence and coworkers addressed this question in a trial in which 168 patients who were randomized to one of the following regimens: ritonavir (400 mg b.i.d.), saquinavir (400 mg b.i.d.) and d4T (30-40 mg b.i.d.) or indinavir (800 mg t.i.d.) with d4T (30-40 mg b.i.d.) and 3TC (150 mg b.i.d.).1 All patients were protease inhibitor-naïve at baseline, and a few patients in both arms took slightly different combinations of nucleoside analogs. At 48 weeks, 32 patients were still being followed in the ritonavir/saquinavir arm, versus 36 in the indinavir arm.
The two regimens achieve similar degrees of viral suppression: in each cohort baseline viral loads decreased by a median value of roughly two logs. A slightly higher percentage of patients in the indinavir arm achieved undetectable viral loads (< 400 copies/mL), but quality of life scores were similar for both regimens. However, of the 30 patients who discontinued study medications, 22 were receiving ritonavir/saquinavir.
Dr. Harold A. Kessler, a member of the editorial advisory board of HIV Newsline, comments:
A serious limitation of ritonavir/saquinavir used in combination is that both of these protease inhibitors have poor CNS penetration. Indeed, one small study has demonstrated that the combination of ritonavir and saquinavir does not suppress viral load in CSF unless d4T is added. However, the evidence gathered by Florence et al. suggests that these two drugs, when taken in combination, achieve approximately the same reductions in viral load as are seen with indinavir. Furthermore, this two-protease combination may be a viable alternative for patients who break through on nelfinavir-containing regimens.
Gisolf E, Jurriaans S, Van Der Ende M, Portegies P, Hoetelmans R, Danner S. The Prometheus Study: double protease inhibitor (PI) treatment only is unable to suppress detectable viral load in the CSF. 5th Conference on Retroviruses and Opportunistic Infections, Chicago, IL, January 31-February 5, 1998. Abstract 403.
Secondary prophylaxis for CMV retinitis
When is it safe to discontinue maintenance therapy?
The apparent reconstitution of immune function that clinicians are seeing in patients who respond to maximally-suppressive antiretroviral therapy has raised the tantalizing possibility that it may be feasible to terminate at least some long-term (read "life-long") maintenance therapies for recalcitrant infections like CMV retinitis (see "Prophylaxis of OIs in the Age of Protease Inhibitors -- An Update," Vol. 4, No. 5, pages 89-92).
CMV disease is exceedingly rare in patients with CD4 counts above 150 cells/mm3. In theory, at least, it should be possible to halt secondary CMV prophylaxis in patients whose CD4 counts rebound to 150 cells or more on combination antiretroviral therapy. In clinical practice, however, practitioners have been extremely reluctant to take the untested step of stopping maintenance therapy in patients who have been through induction treatment for CMV retinitis. Not unreasonably, they have chosen to err on the side of caution until they had some data to guide their decision-making.
Those data are now at hand, as a result of two European studies that assessed the risk for reactivation of CMV disease in patients who had responded well to maximally suppressive antiretroviral therapy. Jouan et al. recruited 52 patients at 18 centers in France, Spain, Italy, and Belgium who had received induction treatment for CMV retinitis and were on long-term maintenance therapy.1 All participants were treated with multidrug antiretroviral regimens for a median time of 18 months. At baseline, all subjects had been free of clinically significant retinal lesions for a median of 17 months. The median CD4 count was 239 cells/mm3 in these individuals, and the median plasma HIV RNA level was 2.3 logs. None of the patients had CMV viremia.
Follow-up included an ophthalmologic exam, CD4 count, and CMV plasma viremia assay every two weeks. Wide-angle fundoscopic photographs, plasma RNA levels, and CMV PCR assays were done every three months, and retinal angiography was performed at 48 weeks. In 48 evaluable patients, there were two CMV events over 12 months of observation. These authors conclude that it is safe to discontinue maintenance therapy for CMV retinitis in selected patients, but they warn that it is important to follow patients very carefully and watch for events such as macular edema.
Another collaborative European effort, this one involving centers in Germany and Spain, reports similar findings.2 Tural et al. enrolled 16 patients with CMV retinitis. When stable CD4 counts above 150 cells/mm3 and viral load measurements below 200 copies/mL were achieved through combination antiretroviral therapy, patients were given the option of stopping their maintenance anti-CMV therapy. Seven participants, who had been treated for CMV for a median duration of 11 months, opted to stop therapy. All had negative plasma CMV PCR assays at the time they made this decision.
Patients underwent weekly ophthalmologic exams for the first month, then had monthly examination for the duration of the study (median follow up: 24 months.) In addition, CD4 counts, HIV RNA measurements, and CMV PCR studies were done every three months. Over a median follow up of two years, no new CMV events have occurred in these seven patients. All of them have maintained CD4 counts above 150 cells/mm3 and all have undetectable HIV RNA as well as negative plasma CMV PCR assays.
Dr. William G. Powderly, a member of the editorial advisory board of HIV Newsline and the author of "Prophylaxis of OIs in the Age of Protease Inhibitors -- An Update" (Vol. 4, No. 5, pages 89-92), comments:
These two studies reinforce a number of earlier reports, which also suggest that stopping CMV treatment in selected patients may be feasible and safe (1-5). Although these two newer studies are encouraging, both involved intense monitoring of study subjects -- which will be necessary until we have considerably more data on the advisability of stopping secondary prophylaxis in patients who have undergone induction therapy for CMV retinitis.
It will be interesting to see if it proves possible to recognize early surrogate markers of disease recrudescence in patients who elect to discontinue maintenance therapy. For now, however, there are no official recommendations concerning discontinuation of secondary prophylaxis for CMV retinitis.
It is important to bear in mind that CMV therapy is not harmless. Secondary prophylaxis for CMV retinitis puts ganciclovir-treated patients at risk for neutropenia, and individuals who receive intraocular ganciclovir implants can suffer retinal detachment, which necessitates repair of the retina and replacement of the implant. Nephrotoxicity is a worrisome side effect of long-term therapy with either cidofovir or foscarnet. I therefore recommend that clinicians consider halting maintenance therapy for CMV retinitis only in patients who have healed retinitis, whose CD4 counts are consistently above 100 cells/mm3, whose viral burden is well controlled with multidrug antiretroviral therapy, and who fully understand the risks involved.
Ideally, this withdrawal of maintenance can occur in the context of a clinical trial, where regular monitoring is built into the study protocol. In any case it is important that these patients have CD4 counts and viral-load measurements taken at regular intervals, and that they receive periodic ophthalmologic assessments. Secondary CMV prophylaxis should be resumed in any patient who experiences sustained immunologic deterioration (falling CD4 count, rising viral load, or both).
New treatment for chronic diarrhea
A proprietary botanical dietary supplement helps patients to normalize bowel function
Diarrhea is a common, noisome, irritating, activity-limiting problem in HIV-positive patients, especially those who are taking a protease inhibitor as part of their antiretroviral regimen. With the exception of indinavir, drugs of this class frequently induce gastrointestinal distress, and when this distress becomes chronic it creates the potential for acute dehydration and electrolyte loss. In its more moderate forms, diarrhea merely compromises quality of life, but protracted, severe diarrhea can threaten life itself.
Individuals with advanced HIV disease are often malnourished, in part because of reduced food intake but also because of alterations in intermediary metabolism as well as malabsorption secondary to chronic diarrhea (see "Management of Wasting Syndrome in Late-Stage HIV Infection," Vol. 1, No. 5, pages 91-97, and "Clinical Evaluation of Food Intake and Malabsorption," the pull out and save feature in that issue.)
Malabsorption is a result of specific injuries to the GI tract -- injuries stemming from infections or from mucosal damage -- and is not simply a function of immune deficiency itself. Consequently, diagnosis and treatment of the injury should lead to structural and functional improvement. However, many patients with advanced HIV infection have small intestine injury that cannot be assigned an etiology, even by electron microscopy.
To evaluate diarrheal illnesses in patients with HIV disease and wasting, the clinician must determine whether the small intestine or the colon is the site of injury. This determination is crucial, since diseases of the small intestine are often localized and lead to malabsorption, whereas colonic diseases are often disseminated and inflammatory in nature. As a general rule, irritations of the small bowel are distinguished by large, watery, irregular stools (Table).
Intestinal biopsy is sometimes the only way to determine the cause of a particular patient's diarrhea. This is an expensive and painful procedure -- which is why most clinicians elect to treat diarrhea symptomatically, resorting to biopsy only when the condition fails to respond to empirical measures. For practitioners and their patients, the therapeutic options to control chronic diarrhea effectively have been limited.
One classical approach has been to add bulk to the stool, either in the form of clay (Kaopectate) or fiber (Metamucil). The problem with these over-the-counter products is that they don't work very well for patients with chronic diarrhea -- and clay-based agents cannot be taken indefinitely anyway, because they eventually alter the patient's electrolyte balance.
The other option is to slow motility. Commonly used products like Imodium-AD, Lomotil, and paregoric can achieve this objective, at least in individuals with mild-to-moderate, episodic diarrhea. Prescription medications work better, with that old standby of turn-of-the-century pharmacies, tincture of opium, working best of all. However, many modern pharmacies do not stock tincture of opium, and even so this old remedy is not the optimal agent to treat chronic diarrhea -- because of its sedating side effects, its potential interaction with other agents, and the fact that patients eventually develop a certain degree of tolerance to it, which limits its efficacy.
A newly launched product, SB-Normal Stool Formula, offers practitioners and their patients another option. This patented compound, derived from the latex of the Croton lechleri, a plant that is widely distributed throughout the Amazon rain forest, reduces excess water flow in the bowel and thereby promotes the formation of normal stools. Although the precisely cellular mechanism of this compound remains to be fully elucidated, it has been found to decrease chlorine ion secretion in the gastrointestinal tract -- and elevated chlorine levels in the stool are associated with some types of diarrhea.
The active ingredient in SB-NSF was studied in a double-blind, randomized, placebo-controlled Phase II clinical trial designed to assess the safety and efficacy of this compound. In all, 51 patients with AIDS and chronic diarrhea were enrolled in the study. At the time they entered the study, approximately half of the subjects had diarrhea that was classified as moderate-to-severe (5 to 8 abnormal stools per day) or severe (9 or more abnormal stools per day); the others had moderate but chronic diarrhea (up to 5 abnormal stools per day).
Significantly, 48 of the 51 participants (94%) had diarrhea of unknown etiology, even after cultures, microscopic examination, and toxin assays had been conducted. Most of these individuals (86%) were taking at least one protease inhibitor as part of their antiretroviral regimen, and given the incidence of diarrhea associated with this class of agents, it is logical to assume that at least some of the chronic diarrhea seen in this cohort was a result of protease inhibitor therapy.
Participants were randomized to receive either orally administered SP-303, the active ingredient in SB-NSF (500 mg every six hours, a total of 16 doses over four days) or an identical placebo on the same dosing schedule. The patients who received SP-303 experienced a mean reduction from baseline in daily stool weight of 451 grams by the fourth day of treatment, as compared with a mean reduction of 151 grams in the placebo arm (p = 0.14) (Figure). Patients treated with SP-303 also experienced a statistically significant reduction in stool frequency, from a mean of 5.2 abnormal stools per day to 2.2, versus a reduction from 5.2 to 3.2 in the placebo arm (p = 0.30).
Stool chlorine concentration declined by a mean of 7.1 mEq/g after four days of treatment, as compared with an increase of 3.4 mEq/g in untreated subjects (p = 0.41). Study participants reported no adverse effects of therapy over the brief course of this Phase II clinical trial.
Holodniy M, Koch J, Mistal M, Maimares Schmidt J, O'Hanley P, Pennington J, Porter S. A double-blind, randomized, placebo-controlled Phase II study to assess the safety and efficacy of orally administered SP-303 for the symptomatic treatment of diarrhea in patients with AIDS. Am J Gastroent 1998; (in press).
Dr. David Hardy, a member of the editorial advisory board of HIV Newsline, comments:
As all clinicians who treat patients with advanced HIV disease are well aware, chronic diarrhea is ubiquitous in this patient population. In most cases there is no obvious pathogen. We used to suspect disseminated MAC or CMV infections when we saw patients with severe diarrhea -- but that was back when we saw a lot of disseminated MAC and CMV. With the advent of highly suppressive antiretroviral therapy, the incidence of these once-common OIs has dropped by as much as 70% (see "Where have all the OIs gone?" in Vol. 4, No. l, pages 10-11). We no longer see many patients with PCP, MAC, or CMV disease, but we still see lots of patients with chronic diarrhea of unknown or protease-inhibitor-related etiology.
My guess is that much of this diarrhea is a result of multidrug therapy. After all, diarrhea is now listed as one of the most common side effects of many antiretroviral agents, and it is not unreasonable to suppose that the adverse effects of these concomitantly administered drugs are synergistic.
Encouraging patients to remain adherent to their assigned antiretroviral therapy is a continuing challenge for all of us, and we know that compliance is compromised when side effects become intolerable (see "Compliance: How You Can Help," Vol. 3, No. 3, pages 67-72, and "Strategies to Establish and Maintain Optimal Adherence," the PULL OUT AND SAVE feature in that same issue). It is therefore incumbent upon all of us who prescribe these multidrug combinations to search for remedies that help patients tolerate therapy better by reducing the adverse events associated with the administration of these potent drugs, thereby improving both compliance and quality of life.
Shaman Pharmaceuticals -- which is marketing SB-NSF through its on-line division, ShamanBotanicals.com -- offers clinicians and their patients a new approach to the management of chronic diarrhea. Their new product, which is currently available only through their web site or by calling 1-800-987-9920, is a promising new addition to our anti-diarrheal armamentarium. Significantly, it appears to be a good candidate for chronic administration -- because unlike clay-based agents it does not alter electrolyte balance, and unlike opiate-based agents it does not have narcotic side effects and does not lead to drug-drug interactions. Finally, unlike prescription antidiarrheals, which can cost $10 or more a day, SB-NSF costs roughly $3.50 for a day's supply of pills.
More on post-exposure prophylaxis
At long last, guidelines for PEP following possible exposure through sex or shared needles
Most studies of post-exposure prophylaxis have focused on accidents in the workplace (see "Reducing the Risk of Occupational Exposure to HIV," Vol. 1, No. 4, pages 69-73, and "New data on workplace exposure to HIV: ZDV prophylaxis reduces risk of infection by 81%," Vol. 4, No. 3, page 42). More adventurous public-health advocates are finally expanding this dialogue to include the infection routes that account for the vast majority of exposures: sexual contact and the sharing of intravenous drug equipment.
Clinicians routinely encounter patients who may have been exposed to HIV by these routes, and there are practically no guidelines to help them assess the utility of antiretroviral agents in preventing HIV infection in these individuals or blunting its impact on the immune system. What little we do know strongly suggests that early, aggressive intervention can reduce the risk of transmission or alter the course of infection (see "Treating Primary Infection," Vol. 2, No. 6, pages 135-141, and "Highlights of the 5th Conference on Retroviruses and Opportunistic Infections: Immune reconstitution," Vol. 4, No. 1, pages 9-10).
A team from the Center for AIDS Prevention Studies at the University of California San Francisco tackled this all-important topic in a recent issue of JAMA. The group reviewed existing guidelines for occupational exposure to HIV and applied those recommendations to exposures in the community.
The C.D.C. currently recommends that post-exposure prophylaxis be given to persons who have sustained percutaneous occupational exposures involving large volumes of blood or other fluids with high concentrations of HIV because these accidents are associated with high probability of transmission. PEP should be offered to workers who experienced other percutaneous or mucous membrane exposures involving visible blood, tissue, or other fluids that could contain HIV. For all other accidents involving mucous membranes or cutaneous exposures, PEP is not recommended.
Lurie et al. constructed an algorithm to help care providers develop rational treatment plans for patients who have been newly exposed to HIV through sex or shared needles. According to the CAPS team, three factors should be taken into account when one is stratifying the risk of infection for an individual patient: frequency of exposure, the probability that the source is HIV-positive, and the probability of transmission if the source is indeed infected. The team makes a clear delineation between sporadic exposure and ongoing exposure.
For the latter patients -- defined as persons who have more than one exposure a month -- PEP would, for practical purposes, need to be continuous, since the recommended length of treatment with PEP drugs is 28 days. Therefore, the authors do not recommend PEP for patients in this category; they suggest that such patients are better served by referral to a state-of-the-art prevention program.
For patients who are exposed only sporadically to HIV -- which the authors define as less frequently than once a month -- PEP is held to be warranted when the risk of transmission is equal to or greater than 0.3%. Activities that meet this criterion and are considered to confer a high probability of transmission include unprotected anal intercourse or the sharing of syringes with someone known to have HIV infection, or someone of unknown HIV status who comes from a population that has a high prevalence of infection (> 30%).
Activities that are associated with an intermediate risk of infection include unprotected anal intercourse or the sharing of syringes with someone of unknown HIV status who comes from an intermediate-risk population (10-30% prevalence) as well as unprotected penile-vaginal sex with a partner known to be HIV-positive. All patients in this category should be informed of the availability of PEP, and the decision to treat or not treat should be made in collaboration with the patient.
According to Lurie et al., sporadic exposures associated with a low probability of infection (oral sex, penile-vaginal sex, unprotected anal intercourse or sharing of syringes with a partner of unknown HIV status who comes from a low-prevalence population) do not warrant PEP, although these patients should be informed that it is available.
Which drugs to use? The C.D.C. recommends ZDV plus 3TC for all exposures where PEP is warranted -- with the addition of a protease inhibitor when the risk of exposure is high.
Lurie P, Miller S, Hecht F, Chesney M, Lo B. Postexposure prophylaxis after nonoccupational HIV exposure. JAMA 1998; 280 (20): 1769-73.
Dr. James O. Kahn, a member of the editorial advisory board of HIV Newsline, comments:
Lurie et al. have ventured into waters that many researchers and clinicians would prefer to avoid. I for one am glad they have done so, because our obligation to our patients -- including those who are disinclined to follow safer-sex guidelines and/or participate in needle-exchange programs -- requires us to assess the risks and benefits of PEP in the community setting. These authors have performed a great service to the medical community by suggesting a rational approach for assessing the utility of PEP in a wide range of "real life" circumstances. I encourage other researchers and the C.D.C. to build on this pioneering work.
A key point made by the authors of this paper is that care providers must learn to factor their own moral standards out of the decision-making process where PEP is concerned. Lurie et al. point out that practitioners routinely treat overweight, out-of-shape patients with heart disease and lung cancer patients who continue to smoke -- all without allowing their feelings about the life-style choices made by these individuals to cloud their clinical judgment. These principles should not be altered when providers find themselves dealing with patients who may have been exposed to HIV through unsafe sex or unclean needles.
Finally, the choice of medications should be assessed. Although most of the studies of post-exposure prophylaxis conducted to date have focused on zidovudine monotherapy, other drugs and drug combinations may also be effective. For the present, the PEP project at San Francisco General Hospital routinely recommends dual-nucleoside therapy with either ZDV/3TC or ddI/d4T. A third drug is rarely used. NNRTIs, with their excellent tissue penetration and long half-life, may prove to have a role to play in PEP, but for now other antiretrovirals should be restricted to the research setting.
The most important part of any PEP regimen is not drug therapy but counseling. Care providers must resist the temptation to substitute pharmacotherapy for prevention education. In the long run, drug therapy alone is a poor strategy to prevent infection, and as the CAPS team points out, the only way that such therapy can reduce the risk of infection for individuals who are repeatedly exposed to HIV is if that therapy is given as multidrug prophylaxis -- hardly a workable proposition. To make PEP programs truly effective, drug therapy must be combined with HIV testing, education regarding the ways in which HIV is transmitted, and counseling to reduce the likelihood of subsequent exposures.
Fine-tuning the treatment of MAC
Clarithromycin better than azithromycin when either is combined with ethambutol
Since their arrival in the late 1980s, the macrolides clarithromycin and azithromycin have become cornerstones of the treatment of disseminated MAC infections. Both drugs are effective against MAC and both are well tolerated. But is one superior for treatment of established MAC? A small head-to-head study provides preliminary evidence that clarithromycin has the edge.
Several institutions, among them a number of Veterans Administration hospitals, participated in this study, which compared clarithromycin plus ethambutol to azithromycin plus ethambutol. Ward et al. recruited 59 patients with HIV who had documented MAC bacteremia. For various reasons, 22 of the original recruits were excluded, leaving a total of 37 evaluable patients. These subjects were randomly assigned to receive clarithromycin (500 mg b.i.d.) or azithromycin (600 mg daily) with either 800 or 1200 mg of ethambutol daily depending on body weight. All drugs were given on an open-label basis.
Clinical symptoms, Karnofsky scores, quantitative blood cultures, and certain laboratory parameters (hemoglobin, alkaline phosphatase levels, and LFTs) were followed regularly during the first 16 weeks of the study. At that point the trial was stopped, because the interim data showed a statistically significant difference between the groups in the percentage of patients with blood cultures negative for MAC. Of the patients randomized to the clarithromycin arm, 85.7% had negative blood cultures, whereas only 37.5% of those receiving azithromycin had cultures negative for MAC.
In addition, clarithromycin appeared to help clear the blood of MAC more quickly than azithromycin: the median time to clearance of MAC bacteremia among those receiving the former drug was 4.38 weeks, while the corresponding figure was more than 16 weeks for the patients randomized to azithromycin. Adverse events, which were mostly gastrointestinal in nature, were reported in roughly 29% of patients in each arm. More importantly, adherence was found to be about the same in both study arms. No significant differences in clinical symptoms or laboratory values were found between the two groups at 16 weeks.
Ward TT, Rimland D, Kauffman C, Huycke M, Evans TG, Heifets L. Randomized, open-label trial of azithromycin plus ethambutol vs. clarithromycin plus ethambutol as therapy for Mycobacterium avium complex bacteremia in patients with human immunodeficiency virus infection. Clin Infect Dis 1998; 27: 1278-85.
Dr. David Hardy, author of "Recent Advances in Prophylaxis for MAC" (Vol. 1, No. 3, pages 52-57) and a member of the editorial advisory board of HIV Newsline, comments:
This interesting study provides us with preliminary evidence that clarithromycin may be superior to azithromycin when these two drugs are combined with ethambutol to treat active MAC infections in HIV-positive patients. The rapidity with which clarithromycin cleared bacteremia in the majority of recipients is impressive. The results of this small study have yet to be corroborated by other researchers, however. Moreover, the sample size is small and 16 weeks is a fairly short follow-up time. It is possible that, over a longer treatment period, azithromycin could prove as efficacious as clarithromycin. It is also possible that higher doses of azithromycin might improve the drug's performance in treatment of MAC. Azithromycin has the advantage of a very long half-life, making once-daily dosing possible. In theory, this advantage should improve adherence, although in practice it did not enhance compliance in this study.
Itraconazole for Penicillium marneffei
Effective treatment for a rare -- and deadly -- fungal infection
The fungus Penicillium marneffei is the third most common opportunistic infection among HIV-infected individuals in Thailand. Elsewhere it remains very rare, with only 100-odd cases reported in the medical literature, most of them since the advent of the AIDS epidemic. American clinicians should keep this infection in mind, however, because cases have been reported in the United States, in individuals who have traveled to Southeast Asia. The symptoms of P. marneffei, which include fever, weight loss, and anemia, can easily mimic the clinical presentation of other infections seen in HIV-positive patients, and has been confused with tuberculosis, cryptococcosis, PCP, and histoplasmosis, as well as other entities. It produces skin lesions that resemble Molluscum contagiosum.
It has been reported that roughly 50% of patients successfully treated for P. marneffei suffer relapses within six months. Supparatpinyo et al. conducted a randomized, double-blind, placebo-controlled study of itraconazole to prevent such recurrences following initial treatment with amphotericin (0.6 mg/kg for two weeks, followed by itraconazole 200 mg twice weekly for 10 weeks).
In vitro studies have shown fluconazole to be less active against P. marneffei than other available antifungals. Close to 100% of patients who presented with P. marneffei from October 1993 through January 1996 were enrolled in this maintenance treatment study. Seventy-one patients were randomized to receive either itraconazole (n=36) or placebo (n=35). Compliance was in the mid-90% range for both arms. Infection was determined by blood or skin culture. None of these patients -- who had mean CD4 counts of 71.3 cells/mm3 in the treatment arm and 64.8 cells/mm3 in the control arm -- received antiretroviral therapy, which is not generally available in Thailand.
The results were striking (Figure). Twenty relapses occurred, all in the placebo arm. Positive cultures were obtained from these patients using blood, skin, lymph node, or sputum specimens. Median time to relapse was 24 weeks, consistent with other studies of the natural history of P. marneffei following acute treatment. Three of the patients who had relapses died while being re-treated for P. marneffei, which was considered to be the cause of death. The only adverse event that occurred with a statistically significant variation between arms was an elevation in aspartate aminotransferase -- which was, oddly enough, noted in more patients receiving placebo.
Supparatpinyo K, Perriens J, Nelson KE, Sirisanthana T. A controlled trial of itraconazole to prevent relapse of Penicillium marneffei infection in patients infected with the human immunodeficiency virus. New Engl J Med 1998; 339: 1739-43.
Sirisanthana T. A major opportunistic infection of HIV disease in Thailand: Penicillium marneffei. Harvard AIDS Institute Website.
Dr. William G. Powderly, a member of the editorial advisory board of HIV Newsline, comments:
Penicillium marneffei is an exceedingly rare clinical event in the U.S. However, outcome hinges on rapid diagnosis and treatment, and for that reason American clinicians should suspect this pathogen in patients with non-specific symptoms such as fever and weight loss who have recently traveled in Southeast Asia, particularly Thailand, or patients who have migrated from there. This well-designed and rigorously conducted study demonstrates the utility of secondary prophylaxis against P. marneffei infections with itraconazole. The treatment was effective and well tolerated in profoundly immunosuppressed patients not receiving antiretroviral therapy. Citing the expense of itraconazole, the authors plan a second study using ketoconazole for secondary prophylaxis.
Special Report: The Next Generation of Antiretroviral Agents
1. Amprenavir, the newest protease inhibitor
The F.D.A. has approved two new antiretroviral agents in the last two months, both made by Glaxo Wellcome: the protease inhibitor amprenavir (Agenerase®) and abacavir (Ziagen®), a potent addition to the oldest class of antiretrovirals, the nucleoside analogs (see the following section for more information on abacavir). A team of researchers in Europe administered amprenavir (1200 mg twice a day) with abacavir (300 mg twice a day) in an open-label fashion to 41 HIV-infected individuals with CD4 counts greater than 400 cells/mm3 and viral loads higher than 5000 copies/mL (1). A number of so-called long-term non-progressors were also evaluated by these investigators, strictly for comparison with the treated subjects, but did not receive the study drugs.
Treated patients experienced a rapid and dramatic decrease in viral load; 77% had undetectable viral levels (< 500 copies/mL) within two to four weeks. Of 11 patients who had completed 72 weeks of observation, 77% remained undetectable at a level below 50 copies/mL, and 54% of these patients had fewer than five copies of HIV RNA per mL. The CD4/CD8 ratio in samples taken from lymph node biopsies appeared to have improved significantly in treated patients. Among treated patients, two discontinued therapy due to rash and three others dropped out for reasons deemed unrelated to treatment: pancreatitis, Kaposi's sarcoma, and hepatitis B infection.
A group from the Aaron Diamond AIDS Research Center in Manhattan combined amprenavir, abacavir, and ZDV/3TC in newly infected and chronically infected patients (2). All drugs were given twice daily, and all subjects were protease inhibitor- and 3TC-naïve. By eight weeks, the mean decrease in viral load was 2.37 logs in the newly infected group and 2.04 in the chronically infected group. Of the patients remaining in the study at 52 weeks (9 and 10, respectively), all had HIV RNA levels below 500 copies/mL. Mean increases in CD4 counts were approximately 200 cells/mm3 in both groups. Comparable increases were seen in both total and naïve CD4 cells in both groups of patients after one year of treatment, suggesting that some degree of immune reconstitution had occurred in these individuals.
Gastrointestinal side effects are not uncommon in patients who receive amprenavir. A review of data on 606 patients treated with amprenavir for at least 12 weeks found that just over half of these patients experienced nausea (3). The next most frequent events were diarrhea (37%), rash (28%), oral paresthesias (25%), headache (24%), fatigue (23%), and vomiting (23%).
Long-term administration of multidrug regimens that include a protease inhibitor is thought to be associated with the appearance of unusual fatty deposits on the back, belly, and buttocks -- deposits informally known as "buffalo hump" or "protease paunch." It is therefore noteworthy that only four of the subjects treated with amprenavir in these small trials evidenced any lipodystrophy. At least one in vitro study has found that amprenavir and indinavir (as opposed to nelfinavir, saquinavir, and ritonavir) do not seem to cause the changes in fat cells that may result in lipodystrophy (4). Whether these findings will be borne out in vivo remains to be seen.
An ongoing pediatric trial of amprenavir in combination with two nucleoside analogs has shown sustained viral response at eight weeks in children 3 to 17 years old (5). Of 37 participants, 65% and 41% had viral loads less than 10,000 and 400 copies/mL, respectively. The median viral load decrease at eight weeks was 0.84 log. Unsurprisingly, the best responses were seen in protease-inhibitor-naïve children.
Dr. Harold A. Kessler, a member of the editorial advisory board of HIV Newsline and the author of "The Next Generation of Antiretroviral Agents -- An Update" (Vol. 3, No. 6, pages 143-151), comments:
The role of amprenavir in patients with genotypic resistance to the available protease inhibitors remains to be determined. One in vitro study found that 55% of viral isolates with "extensive" resistance to protease inhibitors (i.e. 5 or more mutations) retained susceptibility to amprenavir (1). A salvage study, conducted in patients with multiple genetic mutations conferring resistance to nucleoside analogs and protease inhibitors, found that these heavily pretreated patients responded to the combination of amprenavir, abacavir, and efavirenz (2). Over half of the patients with baseline sensitivity to at least two of the drugs had either a 1-log decrease in viral load or a viral load measurement less than 2.6 logs after 16 weeks of treatment. However, only 18% of those with baseline sensitivity to only one of the drugs had positive responses.
2. Abacavir, a potent nucleoside analog
A central tenet of the long-term management of HIV infection is that when patients break through on their assigned antiretroviral regimen, they should be switched to a successor regimen that includes at least two new anti-HIV agents. For clinicians and patients alike, this pharmacological ideal has posed a practical dilemma -- because most heavily pretreated individuals have already been exposed to all of the older nucleoside analogs. Although the protease inhibitors and the NNRTIs all suppress viral replication more effectively than the older nucleosides, this class of agents remains a mainstay of therapy, its unique mechanism of action an important component of any multifront assault on HIV.
For this reason Glaxo Wellcome's newest nucleoside analog, abacavir, is an especially welcome addition to the clinician's anti-HIV armamentarium. A water-soluble guanosine analog that is triphosphorylated in cells by a unique mechanism of action, abacavir has not only demonstrated strong in vitro synergy with other reverse-transcriptase inhibitors, it has proven to be the only drug in its class that can match ZDV's ability to penetrate the central nervous system.
In general, abacavir is well tolerated, with headache and GI disturbances being the most common side effects of therapy. However, this agent can induce a dramatic hypersensitivity reaction in patients, a reaction characterized by fever, malaise, nausea, vomiting, and an erythematous, generalized, and macular rash (see "The Next Generation of Antiretroviral Agents -- An Update," Vol. 3, No. 6, pages 143-151). This reaction, which typically occurs during the first or second week of therapy, is seen in less than 5% of patients who are treated with the drug, and symptoms usually resolve promptly once therapy is halted.
Because rechallenge results in an even more exaggerated reaction, patients who prove hypersensitive to abacavir should not be restarted on this drug. (In a handful of instances, rechallenge has proven fatal for hypersensitive patients.) Healthcare providers should be alert to the possibility that patients who are assigned an abacavir-containing regimen may prove hypersensitive to the drug -- and when patients are given their prescription for abacavir they should be instructed to report any and all adverse symptoms immediately.
Fischl et al. compared the two-nucleoside regimen of ZDV and 3TC with and without the addition of abacavir (ABC) (1). Their double-blind study enrolled 173 patients who were randomized to either the two-drug or the three-drug regimen. At 16 weeks, 75% of those receiving ABC had viral loads less than 400 copies/mL, versus 35% of those on ZDV and 3TC alone. At this point all of the participants had the option of adding open-label ABC or other available antiretrovirals. After 48 weeks of treatment 74% of the subjects who were originally assigned to ABC, ZDV, and 3TC continued to have HIV RNA levels below 400 copies/mL; only seven of these patients had added another antiretroviral. Some 70% of the patients who were originally assigned to the two-drug combination of ZDV and 3TC added another drug at 16 weeks (with 73% of them adding ABC); the addition of another agent or agents resulted in a 1-log drop in viral load among these patients.
Several other studies shed light on the benefits of adding ABC to existing antiretroviral regimens. An open-label French study examined the effects of adding ABC after 12 weeks of ZDV/3TC in 52 patients with median baseline HIV RNA of 2.88 logs and median CD4 count of 543 cells/mm3. At entry, 29% of these subjects had HIV RNA levels less than 20 copies/mL. After 16 weeks of treatment with the three-drug regimen, the percentage with less than 20 copies/mL had increased to 75% -- and it remained at 72% after 48 weeks. The median CD4 count had increased by 188 cells/mm3 at 48 weeks.
Most of the participants in this study experienced some side effects of ZDV-3TC-ABC therapy, chief among them nausea, vomiting, sleep disturbances, eczema, diarrhea, and malaise. Only one serious adverse event, a case of acute hepatitis, was believed to be due to the study medications.
Another European study randomized therapy-experienced patients to continue their current regimen ("stable background therapy") or to add ABC (300 mg twice daily) (2). After 16 weeks of treatment, 39% of those individuals randomized to ABC had significant decreases in viral load (< 400 copies), versus a mere 8% of patients who continued their baseline therapy alone. Only two cases of ABC-related hypersensitivity were recorded by the investigators, but in a study in which subjects combined ABC with one of several different protease inhibitors, 5% of patients experienced hypersensitivity reactions (3). More than half of patients in each arm experienced nausea or diarrhea, which were the most frequent adverse events.
Dr. Harold A. Kessler, a member of the editorial advisory board of HIV Newsline, comments:
It seems increasingly clear that patients either can, or cannot, tolerate abacavir -- and intolerance, if it does occur, manifests early, generally within a week of the initiation of therapy. (The median time to reaction has been 11 days.) However, there have been anecdotal reports of isolated cases of abacavir-induced hypersensitivity after as much as six weeks of treatment with this potent nucleoside analog, so healthcare providers -- and patients themselves -- must remain vigilant against the possibility that this hypersensitivity reaction will develop. Further, it cannot be stressed often enough that patients who do experience hypersensitivity reactions while taking ABC should not be rechallenged with the drug: the chance for an exaggerated, potentially life-threatening reaction is too great.
These important caveats should not discourage clinicians from assigning abacavir as part of a multidrug antiretroviral regimen. This newest nucleoside analog is at least three times as potent as the oldest nucleoside, ZDV. When ABC was given as monotherapy in early efficacy trials, it produced reductions in HIV RNA of as much as two logs, and it is highly effective when taken in combination with other nucleosides and other classes of anti-HIV agents.
3. Tipranavir, first of a new class of non-peptitic protease inhibitors
Selecting an initial antiretroviral regimen for treatment-naïve patients is becoming increasingly easy: the therapeutic options are numerous, and the clinical trial data on these regimens demonstrate that between 75% and 85% of treated patients experience reductions in circulating HIV RNA below 400 copies/mL after 16 to 24 weeks of treatment. There are differences in the incidence and severity of adverse effects of therapy, and pill counts suggest that adherence is better with certain regimens than with others, but it is a virtual certitude that patients will respond well to almost any initial therapy.
By contrast, constructing a second, third, or even fourth therapy, for patients who break through on their initial regimens, is quickly becoming the most challenging and frustrating clinical decision that HIV healthcare providers face today. There are precious few effective options, there are few data to support those limited choices, and there is, at best, only a 35% to 65% chance that such patients will respond to the chosen regimen (see "Protease Inhibitor Resistance and Salvage Therapy," Vol. 3, No. 6, pages 132-136, and "Ritonavir/saquinavir as an alternative for patients who fail nelfinavir: A second-line option in treatment-experienced individuals," on page TK of this newsline section).
Under these circumstances, new drugs or drug combinations that seem even marginally effective in patients with protease-inhibitor-resistant isolates are avidly welcomed by practitioners and patients alike. These second-, third-, and fourth-line regimens are usually regarded as salvage therapies, because until recently no one expected them to suppress viral replication as effectively, or as long, as any initial multidrug antiretroviral regimen.
In this increasingly common clinical context, the early data on tipranavir, the first of a new class of protease inhibitors, seem almost too good to be true. In findings presented this summer at the 3rd International Workshop on HIV Drug Resistance, Larder et al. showed that tipranavir, which is currently in Phase II clinical trials, appears to be active against the majority of HIV isolates that are highly resistant to all four of the currently available protease inhibitors: saquinavir (Fortovase®), ritonavir (Norvir®), indinavir (Crixivan®), and nelfinavir (Viracept®).
In this in vitro study 135 clinical isolates were taken from patients known to have developed resistance to at least one protease inhibitor. Phenotypic assays for the four protease inhibitors cited above were run on all of the viral isolates, and 107 of 135 proved to have greater than 10-fold resistance to at least three of the marketed protease inhibitors. Remarkably, 90% of these 107 highly-resistant isolates remained completely susceptible to tipranavir (< 4-fold increase over the IC50 of the specific protease inhibitor). Another 8% demonstrated intermediate resistance to Pharmacia & Upjohn's investigational agent, and only 2% showed a higher degree of resistance (> 10-fold increase).
In all of the Phase I clinical trials that evaluated tipranavir, this novel agent was generally well tolerated. The most frequently reported side effects were gastrointestinal in nature. Diarrhea, which headed the list of patients' complaints, usually resolved with standard over-the-counter medications.
Larder B, Bloor S, Hertogs K, Van den Eynde C, DeCian W, Yang Y, Freimuth W. Tipranavir is active against a large selection of highly protease inhibitor resistant HIV-1 clinical samples. Antiviral Therapy 1999; 4 (Suppl 1): 5. Abstract 5.
Dr. David Hardy, a member of the editorial advisory board of HIV Newsline, comments:
Discovering and developing antiretroviral drugs that possess in vitro activity against drug-resistant strains of HIV -- and that maintain that activity in clinical use -- is indisputably the most critical challenge facing the pharmaceutical industry in its campaign to find agents that will achieve enduring suppression of HIV replication.
Patients who develop resistance to their first protease inhibitor are in a potentially perilous position -- because resistance to one agent in the class tends to confer some degree of cross-resistance to other drugs in the same class. Even with nelfinavir, the agent that was initially reported to have the least cross-resistance to other protease inhibitors, we now see cross-resistance developing despite the absence of the D30N mutation that was once thought to be essential to nelfinavir resistance. To date, the only viable therapeutic option in patients who break through on a nelfinavir-containing regimen is a multidrug regimen that includes the combination of ritonavir and saquinavir -- and in the salvage study that has yielded the most promising results to date, only 58% of patients responded to this two-protease-inhibitor successor regimen (see "Ritonavir/saquinavir as an alternative for patients who fail nelfinavir").
At this juncture we do not know why tipranavir does not induce the degree of cross-resistance that is seen with the other protease inhibitors; perhaps its "flexible" non-peptitic molecular structure -- which purportedly allows it to avoid mutated binding sites and bind at non-mutated sites -- explains its activity against viral isolates that are highly resistant to one or more of the peptitic protease inhibitors. Whatever the explanation, if the remarkable in vitro data gathered by Larder et al. hold up in subsequent clinical trials in heavily pretreated patients with high-level resistance to most or all of the currently approved protease inhibitors, a genuine advance in antiretroviral therapy may be realized. Indeed, it is possible that providers will opt to use tipranavir as first-line therapy, knowing that its use will not compromise the addition or substitution of other protease inhibitors at a later date. At this point, however, it would be prudent to wait for clinical trial data confirming this assumption, so as not to prematurely and inadvertently introduce tipranavir-resistant HIV strains into the population of treated patients.