NGR-TNF, a novel vascular-targeting agent, does not induce cytokine recruitment of proangiogenic bone marrow-derived cells.

BACKGROUND: Administration of certain chemotherapy drugs at the maximum tolerated dose, vascular-disrupting agents (VDAs) and irradiation can induce mobilisation and tumour homing of proangiogenic bone marrow-derived cells (BMDCs). Increases in cytokines and chemokines contribute to such mobilisation that eventually promotes tumour (re)growth. NGR-TNF is a vascular-targeting agent in advanced clinical development, coupling the CNGRCG angiogenic vessel-homing peptide with tumour necrosis factor-alpha (TNF). We investigated whether NGR-TNF mobilises host BMDCs and growth factors. METHODS: Blood was obtained from Lewis lung carcinoma and 4T1 tumour-bearing mice at different time points following NGR-TNF, VDA or anti-VEGFR2/flk-1 antibody treatment. Levels of circulating growth factors were assessed by ELISAs. BMDCs were characterised by FACS. Circulating endothelial progenitor cells were defined as CD45(-)/CD13(+)/flk-1(+)/CD117(+)/7AAD(-), Tie2-expressing monocytes as CD45(+)/CD11b(+)/Tie2(+) and myeloid-derived suppressor cells as CD45(+)/CD11b(+)/Gr1(+) cells. RESULTS: NGR-TNF decreases tumour blood vessel density-inducing apoptosis of tumour and tumour endothelial cells. Unlike VDAs, or high-dose NGR-TNF, lower doses of NGR-TNF, comparable to those used in clinical trials, neither mobilise nor recruit to the tumour site proangiogenic BMDCs or induce host growth factors. CONCLUSION: Low-dose NGR-TNF exerts antitumour activity without inducing proangiogenic host responses, conceivably important for preventing/overcoming resistance and designing combination therapeutic strategies.

Evolutionary pattern of human immunodeficiency virus (HIV) replication and distribution in lymph nodes following primary infection: implications for antiviral therapy.

Evolutionary patterns of virus replication and distribution in lymphoid tissue during the early phases of HIV infection have not been delineated. Lymph node (LN) biopsies were excised from patients at different times after the estimated time of primary infection. Within 3 months of the acute viral syndrome, HIV was mostly present in individual virus-expressing cells in LNs; trapping of virions in the follicular dendritic cell (FDC) network was minimal or absent, but was the predominant form of HIV detected in LNs of subjects with chronic infection, either recent (4-20 months after primary infection) or long-term (>2-3 years after primary infection). Plasma viremia was significantly higher in patients during the first 3 months than in those recently infected; however, there were no significant differences in the number of virus-expressing cells per square millimeter of LN tissue in these two groups. Numbers of virus-expressing cells in lymphoid tissue were significantly lower in the subjects with long-term infection than in the other two groups. Therefore, during the transition from primary to chronic HIV infection, the level of HIV replication in lymphoid tissue remains elevated despite the fact that viremia is significantly downregulated. These findings have implications for therapeutic strategies in primary HIV infection and in recent seroconvertors.

Limited CD4+ T-cell renewal in early HIV-1 infection: effect of highly active antiretroviral therapy.

We show that the fraction of proliferating CD4+ lymphocytes is similar in HIV-infected subjects in the early stage of disease and in HIV-negative subjects, whereas the fraction of proliferating CD8+ lymphocytes is increased 6.8-fold in HIV-infected subjects. After initiation of antiviral therapy, there is a late increase in proliferating CD4+ T cells associated with the restoration of CD4+ T-cell counts. These results provide strong support for the idea of limited CD4+ T-cell renewal in the early stage of HIV infection and indicate that after effective suppression of virus replication, the mechanisms of CD4+ T-cell production are still functional in early HIV infection.

Predicting the duration of antiviral treatment needed to suppress plasma HIV-1 RNA.

Effective therapeutic interventions and clinical care of adults infected with HIV-1 require an understanding of factors that influence time of response to antiretroviral therapy. We have studied a cohort of 118 HIV-1-infected subjects naive to antiretroviral therapy and have correlated the time of response to treatment with a series of virological and immunological measures, including levels of viral load in blood and lymph node, percent of CD4 T cells in lymph nodes, and CD4 T-cell count in blood at study entry. Suppression of viremia below the limit of detection, 50 HIV-1 RNA copies/mL of plasma, served as a benchmark for a successful virological response. We employed these correlations to predict the length of treatment required to attain a virological response in each patient. Baseline plasma viremia emerged as the factor most tightly correlated with the duration of treatment required, allowing us to estimate the required time as a function of this one measure.

TH1 and TH2 cytokine production by peripheral blood mononuclear cells from HIV-infected patients.

OBJECTIVE: To study the TH1-->TH2 cytokine switch, thought to occur during the progression of HIV infection. DESIGN: We investigated interleukin (IL)-2, interferon (IFN)-gamma, IL-4, IL-6 and IL-10 production by phytohaemagglutinin (PHA)-stimulated and unstimulated peripheral blood mononuclear cell (PBMC) cultures from HIV-negative controls and HIV-positive subjects, stratified according to the Centers for Disease Control and Prevention (CDC) criteria. We correlated the above parameters with markers of disease progression. METHODS: Cytokine production was measured in supernatants using enzyme-linked immunosorbent assay (ELISA) in 40 patients and 17 controls. To evaluate disease progression, we also determined CD4+ cell counts, PHA-induced proliferative response, p24 release and spontaneous immunoglobulin (Ig) G and IgM production. RESULTS: In agreement with the TH1-->TH2 switch hypothesis, we found that in the course of HIV disease mitogen-stimulated IL-2 production decreased, spontaneous and stimulated IL-6 production and spontaneous IL-10 secretion increased; IL-4 showed an increasing trend, although it was reduced in HIV-positive subjects. Finally, immunoglobulin production increased over time. In contrast, mitogen-stimulated IFN-gamma and IL-10 production did not change among the CDC categories, although the former decreased and the latter increased in comparison with HIV-negative controls. CONCLUSIONS: Our data partially agree with the TH1-->TH2 switch hypothesis. Since IL-6 and IL-10 are produced by different cell types, whose proportions and functional features vary in the course of the disease, further experiments with purified lymphocyte subsets and monocytes are required. Nevertheless, as already suggested, we believe that a switch from a type 1 to a type 2 response occurs in HIV infection.

Plasma levels of soluble CD30, tumour necrosis factor (TNF)-alpha and TNF receptors during primary HIV-1 infection: correlation with HIV-1 RNA and the clinical outcome.

OBJECTIVES: The immunological and virological events associated with primary HIV-1 infection have a major impact on the course of HIV-1 disease, and the identification of early predictors during primary HIV infection is critical for the therapeutic strategy. DESIGN AND METHODS: Eighteen consecutive patients with primary HIV infection were followed for a median of 398 days. Clinical status, CD4+ T-cell counts, and plasma samples were obtained weekly from enrollment until week 6, then at weeks 12, 24 and 52, and every 6 months thereafter. Seroconversion was assessed by anti-HIV-1/2 antibodies and Western blot analysis. HIV-1 RNA in plasma was quantified by Amplicor HIV Monitor test. Samples were assayed for immune complex-dissociated p24 antigen, tumour necrosis factor (TNF)-alpha, soluble TNF receptor (sTNFR)-1, sTNFR-II, sCD30 and sCD8 by enzyme immunoassays. Outcome was defined as entering clinical category B or C according to the Centers for Disease Control and Prevention criteria. As a control group, we included 23 HIV-1-negative healthy blood donors. RESULTS: Plasma levels of sCD30, TNF-alpha and sTNFR were significantly higher in HIV-1-infected patients than in controls, and were positively correlated with each other and with values of HIV-1 RNA. Patients who developed an outcome (n = 4) had significantly higher levels of sCD30, TNF-alpha and sTNFR compared with those who did not. Multivariate logistic regression analysis showed that sCD30 and TNF-alpha were the best predictors of outcome independently of CD4+ T-cell counts. CONCLUSIONS: During primary HIV infection, a persistent immune activation may be associated with a poor clinical outcome. The identification of sCD30 and TNF-alpha levels in plasma as early predictors of outcome in primary HIV infection, may direct the implementation of early therapeutic strategies in patients with elevated risk of disease progression.

Virological and immunological responses to HAART in asymptomatic therapy-naive HIV-1-infected subjects according to CD4 cell count.

OBJECTIVE: When to start highly active antiretroviral therapy (HAART) in asymptomatic chronically HIV-1-infected subjects with CD4 cell counts of 300 x 10(6)-500 x 10(6)/l is debated extensively. Retrospective analyses of virological and immunological responses following HAART have been evaluated in both blood and lymph nodes according to pre-treatment levels of CD4 cells either above or below 500 x 10(6)/l. DESIGN: Open-label, observational, non-randomized, prospective study. SETTING: Outpatients attending the Centre of Clinical Investigation in Infectious Diseases, Centre Hospitalier Universitaire Vaudois, University of Lausanne, Switzerland. PARTICIPANTS: Fifty-four HIV-1-infected antiretroviral-naive subjects with CD4 cell count > or = 250 x 10(6)/l and plasma viraemia > or = 5000 copies/ml who had been treated with HAART for at least 48 weeks. Controls were 49 HIV-negative subjects. INTERVENTIONS: All patients received abacavir, nelfinavir, saquinavir soft gel capsules, and amprenavir in varying combinations for 72 weeks. MAIN OUTCOME MEASURES: The extent of immune reconstitution following HAART in 43 and 11 subjects with either more or fewer than 500 x 10(6) CD4 cells/l at baseline was evaluated in blood and lymph node, and compared with immunological measures observed in 49 HIV-negative controls. RESULTS: After 48 weeks of therapy, plasma viraemia was suppressed effectively in both groups of patients. Normalization of both CD4 cell count in blood, divided equally between memory and naive cells, and percentage of CD4 cells in lymph nodes occurred in the two groups. Consistently, the net increase over baseline in CD4 cell count and in memory and naive CD4 subsets was greater in patients with fewer than 500 x 10(6) CD4 cells/l at baseline. Recovery of HIV-specific responses was similar in the two groups. CONCLUSIONS: This study suggests that virological and immunological responses are comparable in asymptomatic therapy-naive HIV-1-infected subjects with CD4 cell counts above or below 500 x 10(6)/l.

In vitro production of type 1 and type 2 cytokines by peripheral blood mononuclear cells from high-risk HIV-negative intravenous drug users.

OBJECTIVE: To study type 1 and type 2 cytokine patterns in HIV-negative high-risk intravenous drug users (IVDU). DESIGN: We investigated interleukin (IL)-2, interferon (IFN)-gamma, IL-4, IL-6 and IL-10 production by phytohaemagglutinin (PHA)-stimulated and unstimulated peripheral blood mononuclear cell (PBMC) cultures from HIV-negative high-risk IVDU, HIV-negative controls and HIV-positive subjects. METHODS: Cytokine production was measured in supernatants using enzyme-linked immunosorbent assay (ELISA) in 10 HIV-negative high-risk IVDU, 25 HIV-negative controls, and 12 HIV-positive IVDU. We also determined spontaneous in vitro immunoglobulin (Ig) G and IgM production. RESULTS: HIV-negative high-risk IVDU showed increased IFN-gamma and decreased IL-4, IL-10 and IL-2, although the latter was not significant compared with HIV-negative controls. Further, HIV-negative high-risk IVDU had reduced IgG production and impaired IgM-IgG switch. CONCLUSIONS: The reduced IL-2 and IL-4 production suggest an impaired CD4+ T-cell function in HIV-negative high-risk IVDU. The increased IFN-gamma production along with the decreased type 2 cytokine profile is consistent with the hypothesis that protective immunity against HIV may reside in type 1 responses and cell-mediated immunity.

Interleukin-10-induced HIV-1 expression is mediated by induction of both membrane-bound tumour necrosis factor (TNF)-alpha and TNF receptor type 1 in a promonocytic cell line.

OBJECTIVE: To investigate whether the upregulatory effect of interleukin (IL)-10 on HIV expression in a model of latent HIV infection is mediated by induction of endogenous tumour necrosis factor (TNF)-alpha and TNF receptors (TNFR). DESIGN: The latently HIV-infected promonocytic cell line U1 was examined, because in this in vitro model IL-10 has been shown to synergize with multiple cytokines, including TNF-alpha, in enhancing HIV production. METHODS: Membrane-bound TNF-alpha, TNFR-1 and TNFR-2 surface expression were determined by flow cytometry. TNF-alpha mRNA was estimated by competitive polymerase chain reaction (PCR), and TNF-alpha, soluble TNFR-1 and soluble TNFR-2 supernatant content by enzyme-linked immunosorbent assay. HIV-1 expression was quantitated by reverse transcriptase assay and p24 antigen release. RESULTS: We demonstrated that IL-10 induces a time and cell-concentration dependent upregulation of HIV expression in U1 cells. This effect is mediated through the endogenous production of TNF-alpha as demonstrated by blocking experiments with anti-TNF-alpha antibodies and by detection of IL-10-induced increase of TNF-alpha mRNA by competitive PCR. More importantly, IL-10 is able to upregulate membrane-bound TNF-alpha and TNFR-1, along with a consistent increase in the shedding of soluble TNFR-1 without inducing detectable TNF-alpha secretion. CONCLUSIONS: IL-10 activates HIV expression through the membrane-bound TNF-alpha/TNFR-1 pathway, suggesting an amplification mechanism of HIV expression that might occur during cell-to-cell interaction. This positive regulatory effect of IL-10 in an in vitro model of chronic HIV infection is consistent with the inexorable progression of disease seen in advanced patients when both IL-10 and TNF-alpha are elevated.

Immunological and virological responses in HIV-1-infected adults at early stage of established infection treated with highly active antiretroviral therapy.

OBJECTIVE: To evaluate the immunological and virological responses to highly active antiretroviral therapy (HAART) in blood and lymphoid compartments of HIV-1-infected patients at an early stage of infection. DESIGN: An open-label, observational, non-randomized, prospective trial of outpatients attending the Centre of Clinical Investigation in Infectious Diseases, Centre Hospitalier Universitaire Vaudois, University of Lausanne, Switzerland. SUBJECTS: Forty-one antiretroviral-naive HIV-1-infected adults with 400 CD4 T cells/microl or greater and 5000 plasma HIV-1-RNA copies/ml or greater were enrolled, and 32 finished the study. Forty-nine HIV-negative individuals were included as controls. All subjects gave written informed consent. INTERVENTIONS: All patients received abacavir 300 mg by mouth every 12 h and amprenavir 1200 mg by mouth every 12 h for 72 weeks. MAIN OUTCOME MEASURES: The extent of immune reconstitution in blood and lymph nodes after 72 weeks of HAART was evaluated, and compared with immunological measures of 49 HIV-negative subjects. RESULTS: Virus replication was effectively suppressed (-3.5 log10 at week 72). Substantial increments of CD4 T cell count in blood and percentage in lymph nodes were observed over time, and these measures were comparable to HIV-negative subjects by week 24 in blood and by week 48 in lymph nodes. The increase was equally distributed between naive and memory CD4 T cells. Recovery of HIV-specific CD4 responses occurred in 40% of patients. CONCLUSION: The initiation of HAART at an early stage of established HIV infection induces systemic quantitative normalization of CD4 T cells, a partial recovery of HIV-specific CD4 cell responses, and effective and durable suppression of virus replication.

Cytokines and soluble receptor changes in the transition from primary to early chronic HIV type 1 infection.

We studied determinants of chronic inflammation and/or immune activation in plasma from patients in the transition from primary to early chronic HIV-1 infection. The following parameters were estimated in seven patients over time: plasma concentrations of soluble CD8 (sCD8), tumor necrosis factor alpha (TNF-alpha), soluble TNF receptor type II (sTNFRII), interleukin 6 (IL-6), soluble IL-6 receptor (sIL6R), IL-10, transforming growth factor beta1 (TGF-beta1), along with CD4- and CD8-positive T cell counts, p24 antigenemia, and clinical evaluation. Results showed that concentrations of sCD8, TNF-alpha, and sTNFRII, and peripheral CD8-positive lymphocyte counts, were significantly increased in patients, compared to HIV-negative controls, and showed a trend toward normal values over time. Levels of IL6, sIL6R, IL-10, and TGF-beta1 did not differ from those of controls and did not change over time. Heterogeneity was observed among the patients in terms of CD4-positive T cell depletion, levels of sCD8, concentrations of TNF-alpha/sTNFRII, and clinical outcome. These data indicate that in the transition phase from primary acute to chronic and asymptomatic infection the host immune activation in response to the virus is highly heterogeneous and that the sustained rise in TNF-alpha and its receptor may represent an important therapeutic target in early disease. The persistence of a state of chronic inflammation and/or immune activation could influence the progression of disease independently from CD4-positive T cell counts.

Beta-endorphin content in HIV-infected HuT78 cell line and in peripheral lymphocytes from HIV-positive subjects.

We investigated beta-endorphin (BE) content in an HIV-infected cell line and in peripheral blood mononuclear cells (PBM) from HIV-positive subjects. HIV infection increased BE content in HuT78 cell line compared to uninfected cells. Accordingly, BE content was greater in HIV-positive subjects than in healthy controls, both in fresh PBM and in mitogen-stimulated or unstimulated cultured cells. Further, in PHA-stimulated cultures, BE increase was correlated with disease progression. Opioids are known to decrease immune responsiveness in vivo, and it may be that the increased BE concentrations contribute to HIV-associated immune deficiency. In HIV-positive subjects, but not in healthy controls, intracellular BE concentration was positively correlated with PHA-induced PBM proliferation. The latter data suggest an alternative explanation: that the increased BE content represents a paradoxical response of the host in an attempt to balance virus-induced immunodepression. Thus, BE may be important in fine-tuning of the immune response with its up- and downregulation dependent upon differences in immune status.

Allogeneic murine melanoma cell vaccine: a model for the development of human allogeneic cancer vaccine.

In an attempt to induce an immune response against tumour antigens, several groups are transfecting cytokine and other genes into autologous tumour cells which are given to the patient as a vaccine. This process is labour-intensive, time-consuming and expensive. Allogeneic cells would offer a more convenient vehicle for the delivery of cytokines and other molecules. However, current dogma suggests that MHC-matched cells are a prerequisite for an effective immune response. Using murine melanoma models we compared allogeneic and autologous vaccination and showed that the survival of C56BL/6 mice (H-2b) was prolonged with some degree of protection achieved against an autologous B15-F10 (H-2b) cell challenge when the mice were vaccinated with allogeneic K1735-M2 (H-2k) cells but not when immunized with autologous B16-F10 cells. Both vaccination with live and irradiated allogeneic cells induced an anti-tumour effect using only one immunization and no boost or adjuvant. Protection was not observed after vaccination with another melanoma (S91; H-2d) or with a carcinoma (A9HT; H-2k). Allogeneic vaccination promoted a cytotoxic cellular response against both the allogeneic and the syngeneic melanomas. This allogeneic vaccination model will be useful for studying the underlying mechanisms of protection, in both pre- and post-challenge settings, as well as for developing whole cell vaccination systems using genetically modified allogeneic tumour cells.

Therapeutic perspectives in HIV-1 infection from recent advances in HIV-1 pathogenesis: it is time to move on.

The recent availability of highly active antiretroviral combination therapy (HAART) has significantly influenced the natural history of the infection, delaying the progression to overt AIDS and prolonging survival. At the same time, the increasing knowledge on the pathogenic mechanisms an the use of HAART have challenged the most widely accepted theories about HIV-1 disease, in particular about the feasibility to eradicate the virus. This review will outline what is and what is not achievable by HAART, and will discuss new concepts in the immunopathogenesis of HIV-1 infection that provide the rationale for the design of new therapeutic approaches in the management of HIV-1 disease and AIDS.

Potential role of immune modulation in the effective long-term control of HIV-1 infection.

Recent advances in HIV-1 pathogenesis, and in defining virological and immunological responses to highly active antiretroviral therapy (HAART), along with the identification of the numerous drawbacks of HAART, have clearly demonstrated that the eradication of the virus is not a feasible therapeutic goal, and that there is an urgent need to develop other approaches to fight HIV-1 infection. Novel therapeutic approaches of immune modulation have recently been evaluated in pilot clinical trials. First, treating primary HIV-1 infection with cyclosporin A (CsA) coupled with HAART to target massive immune activation extends the benefits achieved with HAART during primary HIV-1 infection and might contribute to the establishment of a more favourable immunological set-point affecting the ultimate pattern and rate of disease progression. Second, treating chronic HIV-1 infection in patients with long-term suppression of virus replication induced by HAART, with the addition of mycophenolate mofetil (MMF) reduces the pool of activated CD4+ T lymphocytes able to support productive HIV-1 infection, and might have an indirect impact on the pool of resting, latently infected CD4+ T cells, contributing to its depletion in vivo. The important question is clearly whether these results will have an impact on the clinical management of patients with HIV-1 infection, determining the precise therapeutic function of drugs like CsA and MMF, thus investigating the effects of these drugs on residual viral replication and the decay of the latent reservoir, on long-term immunological benefit, and, ultimately, on clinical benefit.

Cyclosporin A in combination with HAART in primary HIV-1 infection.

HAART is the cornerstone of HIV therapy, and has significantly reduced morbidity and mortality associated with HIV disease. The institution of HAART during the primary HIV-1 infection has a more profound influence on the ultimate pattern and rate of disease progression than therapy commenced later on. However, it also well demonstrated that HAART alone is not able to eradicate the virus, unless over a life-long period of time. There is therefore the need to develop alternative strategies aimed at modulating the immune responses in order to achieve the long-term control of HIV even once HAART is discontinued. Among immunomodulant agents, cyclosporin A in combination with HAART might play a role in the treatment of people with primary HIV-1 infection.

Soluble CD30, tumour necrosis factor (TNF)-alpha, and TNF receptors in primary HIV-1 infection: relationship with HIV-1, RNA, clinical outcome and early antiviral therapy.

The natural course of human immunodeficiency type 1 (HIV-1) infection varies considerably. The identification of laboratory disease markers has become critically important to patient management. This study, carried out on 37 patients with primary HIV-1 infection (PHI), shows that, along with plasma HIV-1 RNA and CD4+ T cell counts, evaluation of plasma levels of some immune activation markers (sCD30, TNF-alpha, and sTNFR-I) may help to identify patients at risk of a more rapid disease progression, suggesting that immune activation is among the factors who determine the rate of disease progression. Early combination antiviral therapy significantly decreased levels of virus load and of immune activation markers, suggesting that it may reduce the extent of immune activation through the suppression of HIV-1 replication. Among others, sCD30 could be a more sensitive marker of immune activation, and it might be also useful in the monitoring of the response to antiviral therapy.

Better efficacy of twice-monthly than monthly aerosolised pentamidine for secondary prophylaxis of Pneumocystis carinii pneumonia in patients with AIDS. An Italian multicentric randomised controlled trial. The Italian PCP Study Group.

The aim of this multicentric randomised controlled trial was to evaluate long-term efficacy and safety of once-monthly versus twice-monthly 300 mg aerosolized pentamidine (AP) as secondary prophylaxis of Pneumocystis carinii pneumonia (PCP). We randomised 205 patients with a previous confirmed episode of PCP (107 treated with 300 mg once-monthly AP, and 98 with 300 mg twice monthly AP); the median review period was 232 days. Kaplan-Meier method and Cox's hazard regression model were used for analysis. The main outcome assessments were PCP recurrence, survival and incidence of drug toxicity. The two groups were balanced for prognostic predictors. In the once-monthly AP group, 14 relapses of confirmed PCP were observed, while five occurred in the twice-monthly AP group; the crude relative risk (RR) was 2.69 (95% CI 1.002-7.236, P=0.0496) and the adjusted RR accounting for prognostic predictors was 2.62 (95% CI 0.92-7.5, P=0.071). Death occurred in 36 of 26 patients respectively (adjusted RR 1.32, 95% CI 0.8-2.18, P=0.28). Two patients interrupted the study because of intolerance to AP (one in each group), and severe coughing occurred in two patients (one in each group). At the end of the study, pulmonary function tests were not changed compared with baseline and were the same between the two groups. Our study suggests that 300 mg twice-monthly AP is more effective than 300 mg once-monthly AP as secondary prophylaxis of PCP.

Risks and benefits of aerosolized pentamidine and cotrimoxazole in primary prophylaxis of Pneumocystis carinii pneumonia in HIV-1-infected patients: a two-year Italian multicentric randomized controlled trial. The Italian PCP Study Group.

We randomized 220 HIV-1-infected subjects to receive aerosolized pentamidine (300 mg/4 weeks) or orally trimethoprim-sulfamethoxazole (320-1600 mg/day) for primary prophylaxis of Pneumocystis carinii pneumonia (PCP), and evaluated PCP and toxoplasmic encephalitis (TE) occurrence and survival. Patients developing toxicity switched to the other regimen. Analysis was on intention-to-treat. At 1 year of study, we observed in the pentamidine group a non-significant excess of PCP (4 vs. 1) and TE (7 vs. 3), and a significant increased death rate (15 vs. 2). After 2 years, no significant differences were observed: adjusted RR estimates for pentamidine vs. cotrimoxazole were 1.20 (95% CI, 0.33-4.37) for PCP (6 cases vs. 5), 1.23 (95% CI, 0.46-3.29) for TE (10 vs. 8) and 1.52 (95% CI, 0.83-2.79) for death (30 vs. 18). Crossovers were more frequent in the cotrimoxazole group (41 vs. 4, P < 0.001). Aerosolized pentamidine and cotrimoxazole were equally effective in preventing PCP, and no major differences were observed in TE occurrence and survival after 2 years follow-up.