Newsline

1. Simultaneous initiation of three-drug regimen more effective than adding drugs sequentially
2. Five-drug regimen produces sharp reductions in viral load
3. Genotype and disease progression
4. Reservoirs of HIV infection
5. Interleukin-2: The saga continues
6. Infectious HIV found in genital herpes lesions
7. Prevention of tuberculosis in patients with HIV
8. Panel Urges HIV Tests for All Pregnant Women

Simultaneous initiation of three-drug regimen more effective than adding drugs sequentially

More evidence favoring an aggressive approach to therapy

For more than a decade now, clinicians have been guiding their HIV-infected patients through uncharted waters as new antiretroviral agents have been approved by the F.D.A. As options to zidovudine became available, providers often opted to substitute these newer agents for ZDV or to combine recently approved drugs with older agents. By this process monotherapy was gradually extended to embrace treatment with two, then three, and, in some cases, four or more drugs (see the following news item for more information on this subject). In a study recently published in JAMA, Gulick et al. provide a useful comparison of this once-common practice with starting three drugs simultaneously.

The original study design randomized participants to one of three treatment arms: indinavir monotherapy; ZDV plus 3TC; or indinavir, ZDV, and 3TC. Investigators recruited 97 HIV-infected patients with CD4 counts between 50 and 400 cells/mm³ (median baseline count: 144 cells/mm³), viral loads greater than 20,000 copies/mL (median baseline viral load: 43,190 copies/mL), and at least six months of prior treatment with ZDV.

The protocol called for this study to continue for one year, but the trial was modified at 24 weeks when an interim analysis of the data showed that the triple-drug regimen was distinctly superior to the other two. Therefore, at week 24 all of the patients in the other treatment arms were offered the triple-combination regimen for the duration of the study, which lasted an additional 28 weeks. The trial continued on an open-label basis for a total follow-up period of 100 weeks. Of the original 97 patients, 70 were followed for the full 100 weeks.

After six months of treatment, patients taking the three-drug regimen experienced a median decrease in viral load of 2 logs. Patients in the indinavir monotherapy arm had a maximal median decrease in HIV RNA of 1.7 logs at week eight, after which this value started to increase again. At week 24 the median reduction in viral burden in this cohort was only 1 log lower than it had been at baseline. Corresponding values for patients randomized to ZDV plus 3TC were 1.4 log at week two and 0.6 log at week 24. Median increases in CD4 counts at week 24 were 127, 111, and 13 cells/mm³, respectively, for the triple-combination, indinavir monotherapy, and ZDV/3TC groups.

After the triple-combination regimen was offered to all patients, CD4 and viral-load parameters improved for those participants who had originally been assigned to the indinavir monotherapy and ZDV/3TC cohorts -- but these markers of disease progression were never as favorable for participants who added drugs as they were for those who began treatment with the three-drug regimen. After 100 weeks of treatment, fully 78% of the remaining participants who were originally randomized to three drugs had viral loads less than 500 copies/mL; the corresponding values were 45% and 30% for patients originally given indinavir alone or ZDV/3TC.

Although patients who did not receive triple-combination therapy at the outset of the study did experience rises in CD4 count and drops in viral burden after switching to the more potent regimen, improvements were of greater magnitude and durability in patients treated with three drugs from the beginning of the trial.

Gulick RM, Mellors JW, Havlir D, Eron JJ, Gonzalez C, McMahon D, et al. Simultaneous vs. sequential initiation of therapy with indinavir, zidovudine and lamivudine for HIV-1 infection. JAMA 1998; 280: 35-41.

Dr. Paul Volberding, the editor-in-chief of HIV Newsline, comments:

The 100-week data from this important study indicate that three-fourths of patients whose viral burden falls below 500 copies/mL after six months of treatment with the three-drug combination of indinavir, ZDV, and 3TC will remain below that level for at least two years. This is very good news indeed; in these patients, the insidious effects of HIV infection have been brought to a virtual halt.

What these data also tell us is that adding antiretroviral agents in a sequential manner yields results not unlike the ones we used to see when we switched patients from one single-agent regimen to another (or, later, from one two-nucleoside regimen to another): there is some improvement in virologic and immunologic parameters, but with each change of therapy the response is less pronounced and less durable. Although adding drugs in sequence made a certain amount of sense during the early days of antiretroviral therapy -- when both our options and our understanding of the pathophysiology of HIV infection were severely limited -- it is now an outmoded strategy. Indeed, Gulick and his co-workers suggest that this approach may actually encourage the eventual development of drug-resistant viral strains.

Five-drug regimen produces sharp reductions in viral load

Experimental combination may be useful in patients with high baseline HIV RNA levels

The potency of three-drug antiretroviral combinations is by now well established (see preceding news item). Some clinicians have taken it on faith that four or even five drugs might work even better, but few studies have examined such challenging regimens. A Dutch group decided to compare a standard three-drug regimen with an experimental five-drug combination in patients with high baseline viral loads, to explore the potential advantages of such an antiretroviral fusillade. Weverling et al. recruited 15 antiretroviral-naïve individuals to receive ZDV (300 mg twice daily), 3TC (150 mg twice daily), and ritonavir (600 mg twice daily). In an ongoing companion study, nine subjects were treated with the five-drug combination of ZDV and 3TC at the same doses, abacavir (300 mg twice daily), indinavir (1000 mg three times daily), and nevirapine (400 mg once a day). Only one of the patients assigned to the experimental protocol was treatment-experienced.

Blood was drawn for HIV RNA quantitation at weeks 1, 2, 4, 8, and 12. There was a statistically significant, but modest, difference in the time it took for 50% of patients to reach HIV RNA levels below 500 copies/mL: four weeks for those on the five-drug regimen and six weeks for those taking three drugs. However, a more impressive statistic emerged when the time to suppression of HIV RNA below 50 copies/mL was calculated. Half of those taking the five-drug regimen had viral loads in that range after four weeks of treatment, whereas 12 weeks elapsed before half of the subjects on the three-drug regimen achieved that goal.

Weverling GJ, Lange JMA, Jurriaans S, Prins JM, Lukashov VV, Notermans DW, et al. Alternative multidrug regimen provides improved suppression of HIV-1 replication over triple therapy. AIDS 1998; 12: F117-F122.

Dr. Harold Kessler, the author of "The Next Generation of Antiretroviral Agents -- An Update" (Vol. 3, No. 6) and a member of the editorial advisory board of HIV Newsline, comments:

Only 50% of patients with high baseline viral loads are able to maintain HIV RNA levels below 500 copies/mL after one year of treatment, so it is not unreasonable to consider, as these authors did, whether more aggressive therapy might result in a more rapid reduction in viral burden -- and whether that, in turn, might lead to more durable suppression of viral replication. Weverling et al. have certainly shown that the five-drug regimen they chose for this small pilot study does indeed produce a more rapid reduction in HIV RNA levels. More importantly, they have demonstrated that some of the subjects who are put on this regimen have their viral loads fall below the level of detection of the most sensitive of the currently available commercial assays. This is important, because we now have considerable evidence that patients whose HIV RNA levels drop below 500 copies/mL but remain above 50 copies do less well than those who achieve near-total suppression of viral replication. The unanswered question at this point is whether the rapid reductions in viral burden achieved with this five-drug regimen are of clinical consequence. We won't know that until Weverling and his colleagues can tell us if these reductions translate into more durable suppression of viral replication -- and, therefore, into improved survival.

What we do know at this point is that this experimental regimen is considerably more expensive than any of the three-drug therapies that are the current standard of care for people with HIV disease. More importantly, perhaps, we know that the complex dosing schedule imposed by this five-drug regimen will make adherence an issue -- and close compliance is already an issue in antiretroviral therapy, even with the least complex of daily dosing schedules (see "Strategies to Establish and Maintain Optimal Adherence," Vol. 3, No. 3, and "Compliance: How You Can Help," in the same issue).

Genotype and disease progression

A clearer picture of the role that CCR5 plays in HIV infection

Over the last few years, virologists have identified a number of proteins that facilitate HIV's entry into cells that are susceptible to infection by the virus. These proteins are classified as co-receptors, because HIV uses them in addition to CD4 to gain access to the intracellular compartment. One of these co-receptors, CCR5, has generated particularly avid interest because a genetic variant of this protein, referred to as CCR5-Delta32, has been found to be protective against HIV infection and to delay disease progression. (For more on the role of co-factors in HIV infection and disease progression, see "A New Approach to Therapy" in the February 1998 issue of HIV Newsline. For more specific information on CCR5 and CCR5-Delta32, see "Genetic mutation appears to confer immunity to HIV," Vol. 2, No. 5, and "More evidence that genes confer protection against AIDS progression," Vol. 4, No, 1.)

The keen interest in CCR5 led McDermott et al. to examine the promoter region of that protein, in an effort to determine if clinically significant genetic variants do exist. (The promoter region is a short span of DNA that upregulates transcription of a particular gene -- in this case, the gene that codes for the CCR5 protein.) Blood samples were obtained from 418 participants in the Multicenter AIDS Cohort Study, a four-city study that has recruited more than 5,000 men since 1984. All 418 of the individuals studied by McDermott and co-workers were seroconverters -- men who had entered the MACS study HIV-negative and later became infected.

Meticulous genetic evaluation of DNA from these blood samples -- and from samples taken from anonymous, HIV-negative blood donors -- did reveal one promoter region with a single nucleotide variation. At this site either adenine ("A") or guanine ("G") was present. Since everyone has two sets of genes, one from each parent, there are three possible genetic combinations at the site studied: AA, GG, or AG. These three genotypes were present in roughly the same frequency in all the samples tested, whether the blood came from HIV-positive patients or HIV-negative blood donors. This observation suggests that the genetic variation in question does not affect transmission of HIV.

However, significant differences were found in the rate of disease progression when the clinical courses of patients with the different alleles were compared: MACS patients with the GG genotype progressed to a diagnosis of AIDS more slowly than those with the AA genotype. This finding was true whether the 1987 or the 1993 C.D.C. definition of AIDS was used. There was a trend toward improved survival for patients with the GG genotype, but this result did not reach statistical significance. Progression and survival rates for patients with the AG genotype were intermediate between the GG and AA data.

The authors of the study took care to perform an analysis showing that their results involving the promoter region were independent of any advantage that might have been conferred by other genetic variations in the CCR5 protein, variations already linked to slower disease progression.

These findings have no direct bearing on the clinical management of patients with HIV disease, but this work does improve our understanding of the pathogenesis of HIV infection and offers new potential targets for pharmacological therapy. It is even possible that, at some future date, genetic analyses could inform decisions concerning when to start -- or modify -- antiretroviral therapy.

McDermott DH, Zimmerman PA, Guignard F, Kleeberger CA, Leitman SF, Murphy PM. CCR5 promoter polymorphism and HIV-1 disease progression. Lancet 1998; 352: 866-70.

Reservoirs of HIV infection

Replication-competent virus found in tonsillar tissue

No antiretroviral regimen yet devised completely eradicates HIV in so-called sanctuary sites, sequestered body compartments like lymphoid tissue and the central nervous system. Lymph nodes in particular are considered to be a significant reservoir for HIV; viral activity may persist at high rates in this tissue even when markers of HIV replication in the blood fail to detect viral particles (see "HIV found in lymph tissue of patients with 'undetectable' viral levels," Vol. 3, No. 6).

One possible explanation for this discordance is that HIV seems to behave differently in lymphoid tissue than it does in the bloodstream. A joint research effort by scientists at the Karolinska Institute in Stockholm and Rush Medical School in Chicago examined immunomodulatory changes in lymphoid tissue after the initiation of maximally suppressive antiretroviral therapy. Investigators were interested in assessing whether HIV replication in lymph nodes corrupts the immune system in such a way that viral replication can continue even after potent combination therapy has reduced HIV RNA in the blood to undetectable levels.

Andersson et al. obtained tonsillar tissue samples from three HIV-infected individuals before and after they began a regimen of ZDV, 3TC, and indinavir. These tissue sections were compared with samples taken from four uninfected volunteers. Investigators measured the following parameters of immune function: local cytokines, chemokine receptors, markers of cell activation, and viral load. At baseline, immune activation -- as reflected by levels of IL-2, IL-12, and interferon-gamma -- was 10% to 15% higher in the tonsillar tissue of the HIV-infected subjects than it was in the controls.

Combination antiretroviral therapy had a substantial down-regulatory effect on markers of immune function and viral activity. HIV RNA, initially above 3,000 copies/mL in all three HIV-infected patients, declined to less than 400 copies/mL after four weeks of treatment. Cellular expression of certain surface proteins that are known to facilitate the entry of HIV into lymphoid cells also declined following the initiation of combination therapy. In addition, certain markers of activation of macrophages and dendritic cells -- both of which are susceptible to HIV infection -- were reduced after four weeks of therapy.

That's the good news. The bad news is that cells containing proviral HIV DNA persisted at high levels in the tonsillar tissue of the HIV-infected subjects after a month of antiretroviral therapy. As others have found, such a pool of replication-competent cells poses a potential threat to patients who have no detectable HIV RNA in their blood, because these cells, when stimulated, will churn out new copies of HIV.

It has been suggested that high levels of immune activation may actually facilitate viral replication. (One such theory holds that immunomodulatory cytokines increase expression of CCR5, which in turn encourages cell-to-cell transmission of HIV.) This paper provides additional evidence of robust immune activity in HIV-infected lymphoid tissue. In addition, the finding that proviral DNA persists at high levels in latently infected lymphoid cells focuses attention on the challenge that researchers face in their efforts to completely eradicate HIV from the bodies of infected individuals.

Andersson J, Fehniger TE, Patterson BK, Pottage J, Agnoli M, Jones P, et al. Early reduction of immune activation in lymphoid tissue following highly active HIV therapy. AIDS 1998; 12: F123-F129.

Interleukin-2: The saga continues

Unfortunately, subcutaneous injections are not as effective as continuous infusion

Since the early days of the HIV epidemic, a naturally occurring cytokine, interleukin-2, has been investigated as a potential immunomodulatory agent in infected individuals -- and we have reported on the equivocal results of those clinical trials (see "Interleukin-2 boosts CD4 counts," Vol. 1, No. 2, pages 27-28, and "Update: Interleukin-2 to boost immune system," Vol. 2, No. 6).

IL-2 has the very desirable ability to stimulate production of CD4 cells and to enhance their functioning. For various reasons, it has not been shown to merit a clear role in treatment of immunodeficient patients to date. Part of the problem is that IL-2 must be administered every eight weeks, in cycles that last five days, by continuous IV infusion. Another problem is that IL-2 therapy produces fatigue and malaise in virtually all treated patients, and 33% of patients experience transient, asymptomatic elevations in serum bilirubin. Even so, IL-2 has not been abandoned by the HIV research community, and it may eventually find a niche in the treatment of AIDS patients.

A newly published Australian study compares two preparations of IL-2, one of them the standard formulation used for continuous IV infusion, the other a simplified preparation that can be injected subcutaneously. A control group included in the study was not treated with IL-2. The paper confirms this drug's ability to increase CD4 counts in treated patients, but the simplified preparation used in the study proved to be significantly less effective in this regard than continuously infused IL-2.

Carr et al. recruited 115 HIV-infected patients with CD4 counts between 200 and 500 cells/mm³ to participate in a three-armed, randomized study of the two different preparations of IL-2 versus no immunomodulatory treatment. The study was organized in eight-week cycles with a follow-up period lasting approximately one year. One group (n = 27) was treated with continuous IL-2 infusions for five days every eight weeks, with treatments administered on an outpatient basis. A second group (n = 53) received two subcutaneous injections of a preparation of IL-2 modified with a polyethylene glycol residue (IL-2-PEG), two days apart every eight weeks. The third group (n = 30) did not receive IL-2.

All patients received antiretroviral therapy. Protease inhibitors were not used in this study; patients took ZDV, ddI, or ddC in various combinations. During the fourth cycle, 3TC was started in all patients to see if further suppression of HIV might improve response to IL-2. Patients liberally changed their antiretroviral regimens throughout the study, but the investigators are persuaded that these modifications did not alter immunologic responses to IL-2.

Among IL-2 recipients, the most frequent adverse events were fever, fatigue, local erythema, gastrointestinal symptoms, and mood changes. One patient receiving IL-2 and four of the patients who were randomized to IL-2-PEG stopped treatment early because of adverse effects. Doses were escalated upward to the maximum dose that patients could tolerate -- which, in this study, was 11.5 million international units for subcutaneous injections of IL-2-PEG.

The median daily dose of IL-2 was 12 MU/day, but almost half of the patients in this arm required dose reductions to nine or even six MU/day. CD4 counts changed dramatically in patients who received IL-2: the median change was an increase of 359 cells/mm³ (Figure). Among recipients of IL-2-PEG, there was a median increase of 44 cells/mm³, and CD4 counts declined by a median value of 46 cells/mm³ in the control group. The potency of these cells was assessed by skin responses to recall antigens.

Delayed-type hypersensitivity scores improved substantially for all patients randomized to either form of IL-2 therapy, even as those scores declined significantly for all those who did not receive IL-2. However, most patients treated with IL-2 did not develop new responses to recall antigens; the improvements were in pre-existing responses. Like other investigators, Carr et al. observed a transient increase in HIV RNA levels following treatment with IL-2, but this therapy did not result in sustained increases in viral load.

Carr A, Emery S, Lloyd A, Hoy J, Garsia R, French M, Stewart G, et al. Outpatient continuous intravenous interleukin-2 or subcutaneous, polyethylene glycol-modified interleukin-2 in human immunodeficiency virus-infected patients: a randomized, controlled, multicenter study. J Infect Dis 1998; 178: 992-9.

Dr. William G. Powderly, the author of "Prophylaxis of OIs in the Age of Protease Inhibitors -- An Update," in this issue, and a member of the editorial advisory board of HIV Newsline, comments:

Although this study found that subcutaneously injected IL-2 is not nearly as effective as continuously infused IL-2 in boosting CD4 levels in treated patients, this does not mean that we should abandon hope for a convenient delivery system for IL-2 -- or that there is no role for this drug in the treatment of HIV disease. The Australian study was small and, more importantly, it was conducted prior to the introduction of potent combination antiretroviral therapy. A larger study, in patients receiving maximally suppressive therapy, might yield different results.

Improvements in immunologic function appeared to be modest in this study, although there was an impressive increase in absolute CD4 counts among patients treated with infused IL-2. The potential role for IL-2 in patients with HIV infection remains to be determined, but its potent immunomodulatory properties assure that research with this substance will continue.

Figure: Effects of Interleukin-2 Therapy   Mean changes from baseline in CD4 counts in patients receiving antiretroviral therapy plus either continuously infused IL-2 (solid black line), subcutaneously injected IL-2 (purple line), versus no IL-2 treatment (dotted black line). Carr et al. J Infec Dis 1998, with permission.

Infectious HIV found in genital herpes lesions

Reactivation of HSV leads to increased viral load on mucosal surfaces

Skin lesions resulting from sexually transmitted diseases are associated with more efficient transmission of HIV, and the prevailing wisdom has long held that these lesions increase transmission risk by presenting extra portals of entry for HIV. While that theory may still hold, at least in part, provocative new evidence suggests that there are additional factors. Schacker et al. studied the level of HIV present in herpes simplex virus-2 lesions in 12 HIV-infected gay men. They also compared the viral load found in samples taken from patients' HSV lesions with the subjects' plasma viral load.

The 12 participants had CD4 counts ranging from 18 to 541 cells/mm³, and the median baseline viral load was 35,000 copies/mL. Because this study took place in 1995, before triple-combination therapy was widely adopted and before experts began to urge that asymptomatic patients receive treatment, only three of the 12 were taking antiretroviral medications (ZDV plus saquinavir or ZDV plus ddI and d4T). Furthermore, none of the patients were taking suppressive therapy for HSV outbreaks, and none routinely took acyclovir or other anti-HSV drugs when lesions were present.

The study required patients to present to the clinic within 24 hours of the appearance of HSV lesions and to return every other day until the lesions healed. A total of 26 HSV episodes were reported by the 12 participants in the study. At each visit, lesions were swabbed for HSV culture as well as for HIV PCR assay. Most lesions were perirectal or along the inner gluteal fold; other sites included the penile shaft, outer buttock, sacral region, and face.

HIV RNA was detected in HSV lesions in 25 of the 26 documented episodes. In 75% of samples, the viral load exceeded 10,000 copies/mL. HIV RNA was much more likely to be detectable on days when HSV was also present on lesional swabs, but half of all samples that were negative for HSV nonetheless yielded HIV RNA.

No correlation was seen between plasma viral titers and the level of HIV RNA isolated from skin lesions. Reassuringly, investigators were unable to isolate HIV from healed, intact skin at the site of HSV lesions, and in only one of nine cases could HIV RNA be detected by PCR from those sites. Initiation of acyclovir treatment caused rapid declines in both HSV and HIV titers in mucosal lesions.

Further analysis of the HIV strains found in herpes lesions suggested that they were replication-competent. While this study was not designed to determine if HIV-infected herpetic lesions increase the risk that HIV will be transmitted during sexual contact, the authors postulate that these lesions are a contributing factor in transmission -- and that their role is underestimated. The investigators hold that reactivation of HSV, which leads to increased HIV replication on mucosal surfaces, makes men and women with genital herpes lesions more efficient transmitters of HIV.

Schacker T, Ryncarz AJ, Goddard J, Diem K, Shaughnessy M, Corey L. Frequent recovery of HIV-1 from genital herpes simplex virus lesions in HIV-1-infected men. JAMA 1998; 280: 61-66.

Dr. Judith A. Aberg, a member of the editorial advisory board of HIV Newsline, comments:

This paper presents a compelling argument for chronic suppressive treatment of HSV in patients with HIV infection. We do not know, at this juncture, whether maximally suppressive antiretroviral therapy, by reducing overall viral burden to very low levels, will also reduce the viral load in HSV lesions, but the lack of correlation between the viral levels in the plasma and lesions of these patients does not inspire confidence that anti-HIV therapy will make this mode of transmission less likely. Indeed, the high levels of replication-competent HIV found in HSV lesions suggest that modifications in current safer-sex guidelines may be warranted.

Prevention of tuberculosis in patients with HIV

Isoniazid therapy reduces risk of active infection by 70%

Seropositive patients who have been exposed to tuberculosis have a 30% or greater risk of developing active TB. It is routine in the U.S. to treat these patients with anti-TB drugs, usually isoniazid, for a prolonged period (see "The Challenge of Multidrug-Resistant Tuberculosis," Vol. 3, No. 5). Wilkinson et al. report on a meta-analysis of four trials conducted in different countries (Haiti, Kenya, Uganda, and the United States) which examined the effectiveness of standard TB therapies in HIV-positive patients.

All of these trials compared anti-TB medications to placebo in HIV-infected adults who had a positive purified protein derivative (PPD) skin test or who were deemed by investigators to be at high risk of infection. (The latter category included HIV-infected individuals with negative PPD results who live in communities where TB is endemic.) A PPD result was taken as positive if the induration was greater than 5 mm, which is the standard cutoff for HIV-infected patients in the U.S. Patients with active TB were excluded from all studies.

The following regimens were used:

• Pape et al. (Haiti): INH 300 mg/d vs. placebo for 12 months

• Hawken et al. (Kenya): INH 300 mg/d vs. placebo for 6 months

• Gordin et al. (U.S.A.): INH 300 mg/d vs. placebo for 6 months

• Whalen et al. (Uganda): INH 300 mg/d for 6 months, then INH plus rifampicin 600 mg/d for 3 months, then INH plus rifampicin plus pyrazinamide 2 g/d for 3 months vs. placebo

A total of 4,055 adults participated in these four trials. Mean follow-up time was 15 to 33 months. Adherence was by self-report, although some trials did include urine testing. Compliance appears to have been less than desirable. Hawken et al. report that 70% of their patients had positive urine results at least 50% of the time. In Gordin's study, only 63% of patients completed six months of treatment.

In spite of compliance problems, pooled results showed that these pharmacological interventions reduced the relative risk of active TB in treated patients who had a positive PPD at baseline. The reduction in active disease was approximately 70% in this population. The risk of death was not significantly different when treated patients were compared with controls, however. In patients with negative PPD results at baseline, treatment did not confer significant protection against developing an active case of TB (see "Isoniazid in HIV-positive patients at high risk for TB: Six months of prophylaxis produces no benefit," Vol. 3, No. 5).

Choice of drugs was not a parameter that these investigators set out to assess. They do note that isoniazid appeared to confer the greatest protective effect against TB in the one trial that employed combination therapy (Whalen et al.), but they also report that higher rates of adverse reactions were seen in that trial.

Wilkinson D, Squire SB, Garner P. Effect of preventive treatment for tuberculosis in adults infected with HIV: systematic review of randomised placebo controlled trials. BMJ 1998; 317: 625-29.

Dr. Judith A. Aberg, a member of the editorial advisory board of HIV Newsline, comments:

Although this analysis does not provide new information, it does underscore the importance of determining the PPD status of HIV-infected patients and treating those who have been exposed to M. tuberculosis. This intervention is especially important in communities where TB is prevalent or is on the rise (see "Predictors of TB infection: Geographical region and induration size of positive PPD test help identify those at risk," Vol. 3, No. 5).

Aggressive efforts to prevent the spread of active TB infection are an important aspect of the overall care of seropositive patients, because they are 100 times more likely to develop tuberculosis than uninfected individuals, chiefly as a result of reactivation of latent TB infection. And because TB is the one opportunistic infection in HIV-positive patients that is readily acquired by seronegative healthcare providers, family members, and others who come into contact with individuals coinfected with HIV and TB, prevention must also be a public health priority, to protect the community at large from this spreading scourge.

Providers should be aware that all of the currently approved protease inhibitors interact with rifamycin derivatives such as rifampin and rifabutin, drugs commonly used to treat tuberculosis or prevent recurrent infections. When protease inhibitors and rifamycin derivatives are taken concurrently, the latter drugs reduce serum concentrations of protease inhibitors to subtherapeutic levels. This insidious interaction creates a clinical dilemma for providers who treat patients infected with both HIV and TB. Before beginning TB treatment in such patients, clinicians should consult "Guidelines for Administration of Protease Inhibitors and Rifampin in HIV-Positive Patients with Tuberculosis," Vol. 3, No. 5.

Panel Urges HIV Tests for All Pregnant Women

Every pregnant woman in the country should be offered HIV testing, according to a panel of experts convened by the Institute of Sciences. This policy change would enable care providers to identify HIV-positive women during the early stages of pregnancy -- when the administration of antiretroviral drugs not only benefits expectant mothers but dramatically reduces the chances that the children these women are carrying will be born HIV-positive. Previously, the official policy was that any woman who was asked to agree to take an HIV test had to be offered counseling about the implications of a positive test result. This requirement imposed a burden on busy care providers, especially those in underfunded, understaffed inner-city clinics -- and a number of providers decided against offering HIV testing to pregnant patients simply because they didn't have the time to provide the mandated counseling to these women. The I.O.S. panel reaffirms the importance of counseling for all women who take an HIV test, but it recognizes the practical limitations of such a policy. In the end, the experts agreed, testing all pregnant women is more important than testing a lesser number and providing counseling to them.