http://www.medscape.com/viewarticle/450729

Opportunistic Infections: They Still Exist, Even in North America and Western Europe

Henry Masur, MD   

At the 10th Conference on Retroviruses and Opportunistic Infections (CROI), there was broad and outspoken recognition that opportunistic infections (OIs) continue to cause morbidity and mortality in patients who are unaware of their HIV infection, individuals who have inadequate access to healthcare, and patients whose CD4+ T lymphocytes remain low because they can no longer benefit from available antiretroviral agents. However, the pathogens that received the most attention were not Pneumocystis carinii, Cryptococcus, and the well known OIs that characterized the first 2 decades of AIDS. Most attention was focused on tuberculosis because of its worldwide importance, and on hepatitis B and C, human herpesvirus 8 (HHV-8), and hepatitis G virus (GBV-C).

In a plenary talk, Sebastian Lucas^[1] from St. Thomas' Hospital, London, United Kingdom, gave an overview on global differences in the clinical pathology of AIDS patients. He emphasized that some of the apparent differences in the incidence of specific OIs are due to reporting bias, ie, diagnostic facilities are not comprehensive in many areas, and thus certain diseases are rarely diagnosed even though they are occurring. In addition, certain diseases may not be occurring because patients do not live long enough to develop advanced immune disease. He also emphasized that some pathogens are geographically restricted, such as measles pneumonia, malaria, leishmaniasis, and trypanosomiasis. Some of these regionally restricted pathogens cause clinical manifestations in HIV-infected patients which are no different from what is seen in non-HIV-infected patients, while some cause disease which is more severe (eg, malaria), and some cause unique manifestations in patients with HIV infection, such as trypanosomiasis (cerebral chagomas) and penicilliosis (disseminated disease).

Tuberculosis

Tuberculosis received considerable attention from investigators in both the developed and developing world. It is obviously a huge problem in many parts of Asia, Africa, and Latin America. In the United States, coinfection with tuberculosis and HIV continues to be a problem, especially among immigrants and residents of group facilities such as prisons and shelters.

Early in the era of highly active antiretroviral therapy (HAART), it was recognized that rifampin could be difficult to use in HIV-tuberculosis-coinfected patients because of its potent induction of the cytochrome p450 system, which would substantially lower drug exposure to most protease inhibitors and nonnucleoside reverse transcriptase inhibitors. The Centers for Disease Control and Prevention (CDC) and others emphasized the use of rifabutin (United States Public Health Service-Infectious Disease Society of America Guidelines for Antiretroviral Therapy: www.hivatis.org), with the expectation that drug interactions would be less substantial than with rifampin. Healthcare providers have been seeking more clinical data to confirm that the recommended regimens were successful for treating both HIV infection and tuberculosis.

A group in India led by Patel^[2] reviewed their records retrospectively to compare the CD4+ T-cell response to efavirenz and 2 nucleoside analogues in a group of HIV-infected patients with tuberculosis (n = 99) who were started on antituberculous therapy, and a group of patients who did not have tuberculosis (n = 98) and were treated for HIV infection alone. Patients with tuberculosis entered into therapy with a lower CD4+ T-lymphocyte count (104 cells/microliter [mcL]) compared with patients without tuberculosis (130 cells/mcL, P < .0001), but at 9 months patients with tuberculosis had a higher CD4+ T-lymphocyte count (306 cells/mcL vs 270 cells/mcL, P < .0001). There were no major differences in drug-related toxicity. Of note, 8 patients (8%) in the tuberculosis group had a paradoxical worsening of their disease, possibly due to the well-recognized phenomenon of immune reconstitution. This study did not address viral load response, and did not address whether patients with tuberculosis would have done even better with another antituberculous regime, eg, one containing rifabutin.

Bill Burman and colleagues^[3] from the CDC provided another piece of evidence on the efficacy of antiretroviral therapy for patients being treated for tuberculosis. His group asked the question: Does HAART improve the prognosis of patients receiving rifabutin-containing antituberculosis regimens compared with patients in an earlier era, prior to the development of HAART? He compared outcomes of 169 patients enrolled in the TB Trials Consortium (TBTC) Study 23, who received any antiretroviral drug except delavirdine, with those of patients in a pre-HAART cohort of CPCRA 019/ACTG 222. The TBTC Study cohort had a lower death rate at 12 months (5%) than the CPCRA/ACTG cohort (15%). Thus, it would seem that rifabutin-based antituberculous regimens can also be used safely and effectively with HAART. Disappointingly, this study did not compare CD4+ T-lymphocyte counts or viral loads between the 2 groups.

Thus, when rifampin-containing regimens and rifabutin-containing regimens are used to treat tuberculosis in HIV-infected patients, antiretroviral regimens can be quite effective, although more data are needed to determine which regimens, and which antituberculosis regimes, provide the best efficacy and safety.

Nettles and colleagues^[4] compared rifampin- and rifabutin-based therapy regimens for all patients (HIV-infected and noninfected) at the Baltimore City (Maryland) Tuberculosis Clinic from 1993-2001. This retrospective cohort study included 431 patients: 25% were HIV-infected, 42% were HIV-negative, and a surprising 33% had unknown HIV infection status. Patients received directly observed therapy consisting of 15 daily tablets/capsules, followed by a standard twice-weekly regimen, even though the CDC does not recommend the use of intermittent rifamycin regimens due to resistance concerns. There was no standard protocol regarding which HAART regimen to receive, when to start, or which CDC-approved rifamycin to use.

Tuberculosis recurrences (ie, treatment failures) were noted in 9/109 HIV infected patients (8.3%) vs 7/322 others (2.2%; P = .007). Relapse was associated with a lower CD4 cell count for HIV-infected patients (51 vs 137 cells/mcL; P = .02). Among HIV-infected patients, recurrences occurred in 3/81 patients who received rifampin vs 0/27 who received rifabutin (P = .57). Thus, even though this experience was retrospective and uncontrolled, and even though it used intermittent rifamycin regimens that are no longer recommended (there are more relapses with intermittent rifabutin compared with daily rifabutin), this report provides some useful data supporting the concept that rifabutin is no less effective than rifampin.

Cytomegalovirus

There were several studies of the immune response to cytomegalovirus (CMV) infection or disease, showing correlates of CD4+ and CD8+ T-cell responses and HIV infection.^[5,6] Preliminary work suggested that it might be feasible to measure immune correlates that would predict reactivation of disease. These studies are interesting but do not yet provide assays that are feasible in most laboratories, or are sensitive and specific enough for predicting outcome to warrant clinical use.

Measurement of CMV viral load is becoming more available at commercial laboratories, and can be useful in a variety of settings for predicting clinical events and for monitoring the efficacy of therapy or chemoprophylaxis. Charles van der Horst and colleagues^[7] from the University of North Carolina expanded on work previously published by others. His group reported on a cohort of 200 patients with HIV infection and CD4+ T-lymphocyte counts < 200 cells/mcL. They performed 4 different CMV assays on serial blood samples (pp65 antigenemia, NucliSens pp67 antigenemia, Roche CMV DNA PCR, and whole-blood CMV DNA hybrid capture [Digene]). They observed no mortality due to CMV disease, but interestingly they found that HIV viral load and CMV DNA PCR viral load were both independent predictors of death (CD4+ T-cell counts were not predictors in this group probably since all had low CD4+ T-cell counts). Thus, CMV viremia was an independent predictor of death: whether preemptive therapy for CMV would improve outcome, or whether this viremia is merely a surrogate marker for poor immune function remains to be determined.

In another abstract related to CMV, Douglas Jabs and colleagues^[8] from Johns Hopkins reported on 176 prospectively studied patients with CMV retinitis on behalf of the CMV Retinitis and Viral Resistance Study. They found, not surprisingly, that when viremia occurred, the isolates were often foscarnet-resistant (at 3,6, and 9 months, resistance occurred in 13%, 24%, and 37% of isolates, respectively). This should remind clinicians that long-term foscarnet therapy is not uniformly and indefinitely effective: resistance develops at rates and times comparable to those seen for ganciclovir.

Enhancing Lean Body Mass for Patients With OIs

When patients develop OIs, they often experience distressing and debilitating weight loss that includes substantial loss of muscle mass. Morris Schambelan and colleagues^[9] recruited 74 patients to determine whether a 4-week course of growth hormone or a tumor necrosis factor inhibitor (thalidomide) would enhance lean body mass or functional performance. Disappointingly, this study showed no sustained benefit from either intervention. Growth hormone did produce better early accumulation of lean body mass, but by 8 weeks there was little difference compared with placebo.

Sexually Transmitted Diseases

New manifestations deserve attention from clinicians. An interesting report commented on the occurrence of ocular syphilis. Gail Phadungchai and colleagues^[10] at Georgetown University Hospital Infectious Diseases Clinic noted that over the past year, 40/350 (11.4%) of patients were diagnosed with syphilis: 6 had neurosyphilis and 3 had ocular syphilis. Patients with ocular syphilis presented with blurred vision and vitreal inflammation. All had CSF pleocytosis, but only 2/3 had a positive CSF VDRL, reminding clinicians that this latter test may not always be very sensitive. These patients had high CD4+ T-cell counts (388-594 cells/mcL) and none had recently initiated antiretroviral therapy. Whether ocular syphilis represents a well-known (but relatively infrequent) manifestation of an increasingly common STD, or whether this is a new complication such as a type of late immune reconstitution syndrome, remains to be seen.

Prevention of OIs: Chemoprophylaxis and Immunization

At the 9th CROI in 2002, there was considerable discussion of the indications to start and to stop primary and secondary prophylaxis for OIs. At this meeting, those issues seemed to have been resolved with the publication of large studies as well as the publication of the revised United States Public Health Service-Infectious Disease Society of American Guidelines on Prevention of Opportunistic Infections in Persons with HIV Infection (www.hivatis.org).

An issue that has not been widely addressed is the effects of stopping trimethoprim-sulfamethoxazole prophylaxis on "secondary" targets, such as respiratory infections. Furrer and colleagues^[11] from the Swiss Cohort looked at the incidence of pneumonia in patients who discontinued trimethoprim-sulfamethoxazole prophylaxis following a CD4+ T-cell response to antiretroviral therapy. They found that not only was it safe to stop this prophylaxis in terms of a very low occurrence rate of Pneumocystis carinii pneumonia, but the incidence of bacterial pneumonia was also very low as long as the CD4+ T-lymphocyte count remained > 200 cells/mcL.

Prevention of respiratory diseases for patients receiving HAART, and for those without access to antiretroviral therapy, can be provided by immunization as well as chemoprophylaxis. The use of the 23-valent pneumococcal vaccine in all HIV-infected patients, with revaccination if the patient's CD4+ T-lymphocyte count rises to > 200 cells/mcL after the initial immunization, is considered a standard of care. However, more data would be useful regarding the likelihood and durability of antibody response as well as the clinical benefit in terms of reduced morbidity and mortality. A study from Uganda by French and colleagues^[12] in 2000 caused considerable consternation by showing that, for reasons that are difficult to explain, HIV-infected patients who were immunized with pneumococcal vaccine had a higher, rather than lower, mortality.

In terms of measuring antibody responses to this vaccine, a group from Barcelona re-examined the serologic response to the 23-valent pneumococcal vaccine in HIV-infected patients who had been treated with HAART.^[13] They studied 3 groups of patients: HIV-uninfected individuals, HIV-infected individuals who had never had a CD4+ T-lymphocyte nadir < 200 cells/mcL, and HIV-infected patients who had a low nadir but who had responded to HAART and currently had a count > 200 cells/mcL. They found no substantial difference in serologic response among these groups.

In an oral presentation, Hung and colleagues^[14] from Taiwan assessed the incidence of community-acquired pneumonia and AIDS-defining OIs in 305 patients who received vaccine and 203 patients who had not. This trial was observational and not randomized, and unlike the African study, almost all patients were receiving HAART. The adjusted odds ratio for developing pulmonary disease was 0.081 (95% confidence interval [CI] 0.009-0.724, P = .02) for vaccinees compared with nonvaccinees. The adjusted odds ratio for death favored vaccinees (0.144, 95% CI 0.041-0.404, P = .002). There could be other explanations for why patients in an observational trial who received vaccine did better (eg, they were receiving better care). While this study was too small to produce statistical significance for several comparisons, and while it differed from the study in Gambia in that patients were receiving HAART, it is comforting to see data supporting the safety and efficacy of immunization, an inexpensive and highly available preventative measure.

Herpes and Hepatitis Viruses

Kaposi's sarcoma has been decreasing in incidence in North America and Western Europe over the past 5 years, but there is still considerable interest in the biology and epidemiology of human herpesvirus-8 (HHV-8), its causative agent.

Robert Biggar^[15] presented a poster about transmission of HHV-8 in rural Tanzania. He demonstrated that most children acquire this infection during the first few years of life in Tanzania, with some additional transmission among individuals who are sexually active. Transmission appeared to occur from mother to children, probably through saliva, and from husbands to wives, probably through sexual contact.

GB virus type C (GBV-C, formerly called hepatitis G, but renamed because the virus does not cause hepatitis) received considerable attention. A study published last year in the New England Journal of Medicine suggested that patients with GBV-C are relatively protected from HIV disease progression.^[16] Ways in which this virus affects certain cell receptors provide a plausible biological basis for such a clinical effect.

The finding of the New England Journal of Medicine study was contradicted by a study from Sweden of 230 HIV-infected patients who were followed for a mean of 4.33 years, which found that GBV-C status(viremia or antibody status) at diagnosis did not predict the natural history of HIV infection.^[17] However, in a larger study from the ACTG, Williams and colleagues^[18] reported that GBV-C viremia, when assessed 5-6 years after HIV seroconversion, was a strong predictor of survival, and that loss of viremia predicted death. The significance of GBV-C is thus still unresolved: Whether viremia is truly protective (and whether a therapeutic immunization would be beneficial), or whether the viremia is a marker for some other beneficial immunologic process, remains to be determined.

Hepatitis B and C were the subjects of numerous presentations. Tenofovir therapy for hepatitis B showed promising results in lamivudine-experienced and in treatment-naive patients.^[19] Tenofovir appears to be well tolerated and effective, although tenofovir-related toxicities in adefovir-experienced patients were the subject of some attention. For hepatitis C, more trial results were presented. One study indicated that ribavirin administered 3 times per week is not as effective as daily alpha interferon although this result is less relevant now that pegylated alpha interferon is being widely used.^[20] The long-term benefit of these regimens and their cost:benefit ratio^[21] require further study.

Thus, at the 10th CROI, it was clear that considerably more work is needed to develop new approaches to characterizing and managing infections caused by the plethora of well-known and emerging pathogens that continue to cause morbidity and mortality in patients with HIV in North America, Western Europe, and the developing world.