A randomized, controlled trial of vitamin A in children with severe measles

BACKGROUND. Measles kills about 2 million children annually, and there is no specific therapy for the disease. It has been suggested that vitamin A may be of benefit in the treatment of measles. METHODS. We conducted a randomized, double-blind trial involving 189 children who were hospitalized at a regional center in South Africa because of measles complicated by pneumonia, diarrhea, or croup. The children (median age, 10 months) were assigned to receive either vitamin A (total dose, 400,000 IU of retinyl palmitate, given orally; n = 92) or placebo (n = 97), beginning within five days of the onset of the rash. At base line, the characteristics of the two groups were similar. RESULTS. Although clinically apparent vitamin A deficiency is rare in this population, the children's serum retinol levels were markedly depressed (mean [+/- SEM], 0.405 +/- 0.021 mumols per liter [11.6 +/- 0.6 micrograms per deciliter]), and 92 percent of them had hyporetinemia (serum retinol level less than 0.7 mumols per liter [20 micrograms per deciliter]). Serum concentrations of retinol-binding protein (mean, 30.1 +/- 2.0 mg per liter) and albumin (mean, 33.4 +/- 0.5 g per liter) were also low. As compared with the placebo group, the children who received vitamin A recovered more rapidly from pneumonia (mean, 6.3 vs. 12.4 days, respectively; P less than 0.001) and diarrhea (mean, 5.6 vs. 8.5 days; P less than 0.001), had less croup (13 vs. 27 cases; P = 0.03), and spent fewer days in the hospital (mean, 10.6 vs. 14.8 days; P = 0.01). Of the 12 children who died, 10 were among those given placebo (P = 0.05). For the group treated with vitamin A, the risk of death or a major complication during the hospital stay was half that of the control group (relative risk, 0.51; 95 percent confidence interval, 0.35 to 0.74). CONCLUSIONS. Treatment with vitamin A reduces morbidity and mortality in measles, and all children with severe measles should be given vitamin A supplements, whether or not they are thought to have a nutritional deficiency.     

A controlled trial of early adjunctive treatment with corticosteroids for Pneumocystis carinii pneumonia in the acquired immunodeficiency syndrome. California Collaborative Treatment Group

BACKGROUND. Pneumocystis carinii pneumonia remains a common cause of serious morbidity and mortality in patients with the acquired immunodeficiency syndrome (AIDS). The extensive lung injury that accompanies pneumocystis-associated respiratory failure and the reports of clinical benefit from the use of adjunctive corticosteroids provided the rationale for this prospective multicenter trial. METHODS. A total of 333 patients with AIDS and pneumocystis pneumonia received standard treatment and were randomly assigned to receive either corticosteroids (beginning with the equivalent of 40 mg of prednisone twice daily) or no additional therapy. The primary end points in this unblinded trial were the occurrence of respiratory failure (hypoxemia ratio [partial pressure of arterial oxygen divided by fraction of inspired oxygen] less than 75, intubation, or death), death, and dose-limiting toxicity of the initial standard therapy. RESULTS. Of the patients with confirmed or presumed pneumocystis pneumonia (n = 225 and n = 26, respectively), those assigned to treatment with corticosteroids had a lower cumulative risk at 31 days of respiratory failure (0.14 vs. 0.30, P = 0.004) and of death (0.11 vs. 0.23, P = 0.009), as well as a lower risk of death within 84 days (0.16 vs. 0.26, P = 0.026). The frequency of dose-limiting toxicity of the standard therapy was similar in the two treatment groups. Intention-to-treat analyses of the entire cohort confirmed these findings. Clinical benefit could not be demonstrated, however, for patients with mild disease (hypoxemia ratio, greater than 350), equivalent to a partial pressure of oxygen greater than 75 torr on room air. The patients assigned to corticosteroid treatment had an excess of localized herpetic lesions (26 percent vs. 15 percent, P = 0.04) but not of other infections or of neoplasms. CONCLUSIONS. Early adjunctive treatment with corticosteroids reduces the risks of respiratory failure and death in patients with AIDS and moderate-to-severe pneumocystis pneumonia. Because the adverse effects are few, corticosteroids should be included as part of the initial treatment for persons with AIDS who have moderate-to-severe pneumocystis pneumonia.     

Circulating thymic hormone levels in severe combined immunodeficiency.

Twenty-three patients with severe combined immunodeficiency disease were studied for circulating thymic hormone levels (facteur thymique serique, FTS), 21 prior to treatment by transplantation of bone marrow, thymus or fetal liver. Thirteen showed undetectable FTS activity. Only two had normal levels of this hormone. In serial determinations of FTS activity prior to and after transplantation, patients given bone marrow transplants developed sustained increments of serum FTS activity early in the course of their immunological reconstitution. However, patients given transplants of fetal liver alone or fetal liver plus thymus from fetuses of less than 12 weeks gestation generally did not show an increment of FTS activity during the period of observation. Transplantation of irradiated thymus derived from fetuses of more than 14 weeks gestation produced sustained increases of thymic hormone activity. These observations suggest that a cell of haematopoietic origin provides a stimulus necessary for differentiation or maturation of thymic secretory activity and that this cell(s) is present in post-natal marrow, but is either undeveloped or immature in the early fetal liver or fails to migrate to the thymus of an allogeneic host.     

Pulmonary sarcoidosis associated with acquired humoral and cellular immunodeficiency

A previously healthy 27-year-old man with class II pulmonary sarcoidosis developed severe humoral immunodeficiency within the course of the disease with an IgG of less than 250 mg/ml and undetectable levels of IgA and IgM. Repeated skin tests were negative for seven common recall antigens. Cellular blood test demonstrated normal numbers of B cells and slight T-cell lymphopenia with a normal T-helper/suppressor subset distribution (ratio 1.6). In contrast, parallel examination of the bronchial alveolar lavage fluid (BAL) demonstrated highly elevated numbers of T cells with a subset ratio of 3.1 and significant numbers of activated T cells as revealed by the expression of Ia and Tac antigens. Functional in vitro assays showed a greatly decreased mitogenic response of blood T cells and diminished production of immunoglobulins. These data indicate that, despite a severely depressed systemic humoral and cellular immune system, T-cell activation can take place at the inflammatory site, potentially causing the lesions characteristic of sarcoidosis.     

Evaluation of functional thymic hormones in Arabian horses with severe combined immunodeficiency.

Arabian horses with severe combined immunodeficiency disease (SCID) were evaluated for thymic hormone activities using thymic extracts and sera. Extracts prepared from thymus of SCID horses were able to increase the number of spleen cells responding to sheep red blood cells in irradiated, bone marrow-reconstituted mice. In addition, ultrafiltrates prepared from sera of these immunodeficient horses, which contained material with molecular weight of less than 50,000 Daltons could (a) induce a population of human bone marrow precursor cells to differentiate into cells bearing SRBC receptors and form spontaneous E-rosettes, a characteristic of T lymphocytes, and (b) stimulate cyclic adenosine monophosphate (AMP) synthesis in mouse thymocytes. Based on in vivo and in vitro effects, it was concluded that the defect of these Arabian horses with severe combined immunodeficiency disease did not involve a complete thymic hormone inadequacy.     

Evidence for a humoral thymic factor in rabbits

Rabbits thymectomized between 12 and 36 hours after birth subsequently show reduced levels of circulating lymphocytes together with a lowered immunological response to human γ-globulin. These effects of thymectomy can, to a large extent, be prevented by means of intraperitoneal auto-transplants of thymus in a Millipore diffusion chamber. It appears possible that the epithelial-reticular cells of the transplant persist in the diffusion chamber and elaborate a humoral factor, or factors, which take part in the maturation of antibody-producing cells.     

Oral therapy for Pneumocystis carinii pneumonia in the acquired immunodeficiency syndrome. A controlled trial of trimethoprim-sulfamethoxazole versus trimethoprim-dapsone

BACKGROUND. Antimicrobial drugs that can be taken orally are needed for the treatment of Pneumocystis carinii pneumonia in patients with the acquired immunodeficiency syndrome (AIDS). Preliminary data indicate that dapsone with trimethoprim may be an effective alternative to trimethoprim-sulfamethoxazole, which is frequently toxic. METHODS. In a double-blind trial, 60 patients with AIDS and mild-to-moderately-severe first episodes of P. carinii pneumonia (partial pressure of oxygen in arterial blood, greater than 60 mm Hg while breathing room air) were randomly assigned to 21 days of treatment with either trimethoprim-sulfamethoxazole (20 and 100 mg per kilogram of body weight per day, respectively) or trimethoprim-dapsone (20 mg per kilogram per day and 100 mg per day). RESULTS. The orally administered treatment failed because of progressive pneumonitis in 3 of the 30 patients assigned to trimethoprim-sulfamethoxazole and in 2 of the 30 assigned to trimethoprim-dapsone (P greater than 0.3). Major toxic effects required a switch to intravenous pentamidine for 17 patients (57 percent) in the trimethoprim-sulfamethoxazole group, as compared with 9 (30 percent) in the trimethoprim-dapsone group (P less than 0.025). With trimethoprim-sulfamethoxazole, there were more instances of severe chemical hepatitis (six, as compared with one in the trimethoprim-dapsone group) and marked neutropenia (five vs. one). Intolerable rash (three in each treatment group) and severe nausea and vomiting (two in each group) occurred with equal frequency with both drug combinations. Methemoglobinemia occurred in most of the patients treated with trimethoprim-dapsone, but it was asymptomatic and the level exceeded 20 percent in only one patient. Mild hyperkalemia (serum potassium level, 5.1 to 6.1 mmol per liter) also occurred in 53 percent of the patients treated with trimethoprim-dapsone. CONCLUSIONS. In patients with AIDS, oral therapy with trimethoprim-sulfamethoxazole and with trimethoprim-dapsone are equally effective for mild-to-moderate first episodes of P. carinii pneumonia, but with trimethoprim-dapsone there are fewer serious adverse reactions than with trimethoprim-sulfamethoxazole.     

A controlled trial comparing foscarnet with vidarabine for acyclovir-resistant mucocutaneous herpes simplex in the acquired immunodeficiency syndrome. The AIDS Clinical Trials Group.

BACKGROUND AND METHODS. Most strains of herpes simplex virus that are resistant to acyclovir are susceptible in vitro to both foscarnet and vidarabine. We conducted a randomized trial to compare foscarnet with vidarabine in 14 patients with the acquired immunodeficiency syndrome (AIDS) and mucocutaneous herpetic lesions that had been unresponsive to intravenous therapy with acyclovir for a minimum of 10 days. The patients were randomly assigned to receive either foscarnet (40 mg per kilogram of body weight intravenously every 8 hours) or vidarabine (15 mg per kilogram per day intravenously) for 10 to 42 days. In the isolates of herpes simplex virus we documented in vitro resistance to acyclovir and susceptibility to foscarnet and vidarabine. RESULTS. The lesions in all eight patients assigned to foscarnet healed completely after 10 to 24 days of therapy. In contrast, vidarabine was discontinued because of failure in all six patients assigned to receive it. The time to complete healing (P = 0.01), time to 50 percent reductions in the size of the lesions (P = 0.01) and the pain score (P = 0.004), and time to the end of viral shedding (P = 0.006) were all significantly shorter in the patients assigned to foscarnet. Three patients had new neurologic abnormalities while receiving vidarabine. No patient discontinued foscarnet because of toxicity. Although initial recurrences of herpes simplex infection after the index lesion had healed tended to be susceptible to acyclovir, acyclovir-resistant infection eventually recurred in every healed patient, a median of 42.5 days (range, 14 to 191) after foscarnet was discontinued. CONCLUSIONS. For the treatment of acyclovir-resistant herpes simplex infection in patients with AIDS, foscarnet has superior efficacy and less frequent serious toxicity than vidarabine. Once the treatment is stopped, however; there is a high frequency of relapse.     

Corticosteroids as adjunctive therapy for severe Pneumocystis carinii pneumonia in the acquired immunodeficiency syndrome. A double-blind, placebo-controlled trial

BACKGROUND. Preliminary reports suggest that patients with the acquired immunodeficiency syndrome (AIDS) and Pneumocystis carinii pneumonia may benefit from the addition of corticosteroid treatment to antibiotic therapy. METHODS. We conducted a double-blind, placebo-controlled trial to assess the efficacy of adjunctive corticosteroids in patients with AIDS and severe P. carinii pneumonia. Patients with marked abnormalities in gas exchange who had been treated with antibiotics for less than 72 hours were randomly assigned to receive either methylprednisolone (40 mg) or placebo every 6 hours for 7 days, in addition to treatment for 21 days with trimethoprim-sulfamethoxazole. The primary outcome measures were survival until hospital discharge and the development of respiratory failure. RESULTS. Twenty-three patients were enrolled in the study; there were no significant differences in base-line clinical or laboratory measures between the two treatment groups. Of 12 patients treated with corticosteroids, 9 (75 percent) survived until hospital discharge, as compared with only 2 of 11 placebo recipients (18 percent) (P less than 0.008). Respiratory failure developed in nine placebo recipients, as compared with only three patients treated with corticosteroids (P less than 0.008). No patient required the interruption or discontinuation of corticosteroid or antibiotic treatment because of toxicity or a complicating event. Because of the marked difference in survival, it was deemed unethical to continue the trial, and the study was terminated. CONCLUSIONS. Early adjunctive corticosteroid therapy can improve survival and decrease the occurrence of respiratory failure in patients with AIDS and severe P. carinii pneumonia.     

A randomized controlled trial of a reduced daily dose of zidovudine in patients with the acquired immunodeficiency syndrome. The AIDS Clinical Trials Group

BACKGROUND. The initially tested dose of zidovudine for the treatment of patients with advanced disease caused by the human immunodeficiency virus type 1 (HIV) was 1500 mg. Although this dose is effective, it is associated with substantial toxicity. METHODS. To evaluate the efficacy and safety of a reduced dose, we conducted a randomized controlled trial in 524 subjects who had had a first episode of Pneumocystis carinii pneumonia. The subjects were assigned to receive zidovudine in either a dose of 250 mg taken orally every four hours (the standard-treatment group, n = 262) or a dose of 200 mg taken orally every four hours for four weeks and thereafter 100 mg taken every four hours (the low-dose group, n = 262). RESULTS. The median length of follow-up was 25.6 months. At 18 months the estimated survival rates were 52 percent for the standard-treatment group and 63 percent for the low-dose group (P = 0.012 by the log-rank test). At 24 months the estimated survival rates were 27 percent for the standard-treatment group and 34 percent for the low-dose group (P = 0.033). In both groups, 82 percent of the subjects had another opportunistic infection, and the length of time to that infection was similar in the two groups (P = 0.56 by the log-rank test). CD4 T-lymphocyte counts improved transiently in both groups, and serum levels of HIV antigen decreased in the subjects with antigenemia. The hemoglobin level declined to less than 5 mmol per liter (80 g per liter) in 101 subjects in the standard-treatment group and in 77 in the low-dose group (39 vs. 29 percent, P = 0.0009 by the log-rank test). The neutrophil count declined to less than 0.750 x 10(9) per liter in 134 subjects in the standard-treatment group and in 96 in the low-dose group (51 vs. 37 percent, P = 0.0001). CONCLUSIONS. The reduced daily dose of zidovudine used in this study was at least as effective as the standard dose and was less toxic; however, with the use of a four-week induction period with a high dose followed by low-dose treatment, severe anemia and neutropenia were common complications of treatment with zidovudine.     

A placebo-controlled trial of maintenance therapy with fluconazole after treatment of cryptococcal meningitis in the acquired immunodeficiency syndrome. California Collaborative Treatment Group

BACKGROUND AND METHODS. In patients with the acquired immunodeficiency syndrome (AIDS), the rate of relapse after primary treatment for cryptococcal meningitis remains high. We conducted a controlled, double-blind trial to evaluate the efficacy of maintenance therapy with fluconazole. At entry into the study, all participants had sterile cultures of cerebrospinal fluid, blood, and urine after following a standardized course of therapy for culture-proved cryptococcal meningitis. The patients were randomly assigned to take either fluconazole or placebo as maintenance therapy. The dose of fluconazole was 100 mg daily in the first phase of study and 200 mg daily in the second phase. RESULTS. Of 84 patients initially enrolled, 16 (19 percent) were found to have silent, persistent infection on the basis of cultures that became positive after entry into the study; 7 other patients were lost to follow-up shortly after entry. Of the remaining 61 patients, 10 of 27 assigned to placebo (37 percent) and 1 of 34 assigned to fluconazole (3 percent) had a recurrence of cryptococcal infection at any site (difference in risk, 34 percent; 95 percent confidence interval, 15 to 53). Of the 11 recurrent infections, 7 were detected in urine obtained after prostatic massage. There were four recurrent meningeal infections in the patients taking placebo, but none in those taking fluconazole (mean duration of follow-up, 164 days) (P = 0.03). In multivariate analyses, the best predictors of recurrence-free survival were fluconazole treatment (P = 0.02; relative hazard, 13.2), a lower serum cryptococcal-antigen titer (P = 0.05; relative hazard, 1.2), and more prolonged primary therapy with flucytosine (P = 0.09; relative hazard, 1.1). Survival and toxicity were similar in the two maintenance-treatment groups. CONCLUSIONS. In patients with AIDS, silent persistent infection is common after clinically successful treatment for cryptococcal meningitis. Maintenance therapy with fluconazole is highly effective in preventing recurrent cryptococcal infection.     

Low levels of human immunodeficiency virus replication in the brain tissue of children with severe acquired immunodeficiency syndrome encephalopathy.

The authors examined the autopsy brain samples of nine children infected with human immunodeficiency virus (HIV) at birth by histology, immunologic staining, and in situ hybridization. Surprisingly, although seven of these children presented with typical AIDS encephalopathy, the authors could detect a multifocal HIV infection in the brains of only three of these patients. The authors could not detect any significant HIV replication in the brain of four other children despite severe neurologic disease. However, HIV DNA was detected by polymerase chain reaction (PCR) in the central nervous system (CNS) of all patients. In addition, the authors found associated lesions in the brains of three of these four patients. This study shows that severe AIDS encephalopathy exists in children and therefore might exist in adults with few signs or without any signs of HIV replication or inflammation in the CNS. Understanding the pathogenesis of this neurologic disease and the kinetics of HIV replication in brain tissue of children with AIDS encephalopathy is essential to determine the best therapeutic strategy.     

A placebo controlled clinical trial of transfer factor in lepromatous leprosy.

The effects of repeated injections of transfer factor over a period of 20 weeks were investigated in fourteen bacteriologically positive patients at the lepromatous side of the leprosy spectrum. All patients showed negative (0 mm induration) skin tests to M. leprae antigens (i.e. leprolin and lepromin). Of these patients, seven were treated with transfer factor with a total of 9 units (1 unit being equivalent to 5 x 10(8) lymphocytes) and seven with a placebo. Maintenance treatment with clofazimine was continued. Transfer factor was prepared from the lymphocytes of donors who showed positive skin tests to M. leprae antigens (i.e. leprolin greater than or equal to 12 mm induration, average 15.5 mm or lepromin greater than or equal to 8 mm induration, average 13.6 mm), as well as a positive lymphocyte transformation in vitro to M. leprae (the average transformation being higher than the average transformation of lymphocytes of tuberculoid leprosy patients). No differences were found between the two groups as regards the clinical course of the disease, the histopathological and bacteriological evaluation of skin biopsies, changes in skin test reactivity to various antigens (i.e. lepromin, leprolin, PPD, Mumps, C. albicans, Tr. rubrum and Varidase), as well as the lymphocyte transformation in vitro to various mitogens (i.e. PHA, PWM, Con A) and antigens (i.e. M. leprae, leprolin, PPD, BCG, Mumps, C. albicans, Trichophyton and Varidase). No evidence was found to suggest that transfer factor is a valuable adjuvant in the treatment of lepromatous leprosy patients or that it increases cell-mediated immune reactivity towards M. leprae.     

A randomized controlled trial of cognitive-behavioral treatment for posttraumatic stress disorder in severe mental illness

A cognitive-behavioral therapy (CBT) program for posttraumatic stress disorder (PTSD) was developed to address its high prevalence in persons with severe mental illness receiving treatment at community mental health centers. CBT was compared with treatment as usual (TAU) in a randomized controlled trial with 108 clients with PTSD and either major mood disorder (85%) or schizophrenia or schizoaffective disorder (15%), of whom 25% also had borderline personality disorder. Eighty-one percent of clients assigned to CBT participated in the program. Intent-to-treat analyses showed that CBT clients improved significantly more than did clients in TAU at blinded posttreatment and 3- and 6-month follow-up assessments in PTSD symptoms, other symptoms, perceived health, negative trauma-related beliefs, knowledge about PTSD, and case manager working alliance. The effects of CBT on PTSD were strongest in clients with severe PTSD. Homework completion in CBT predicted greater reductions in symptoms. Changes in trauma-related beliefs in CBT mediated improvements in PTSD. The findings suggest that clients with severe mental illness and PTSD can benefit from CBT, despite severe symptoms, suicidal thinking, psychosis, and vulnerability to hospitalizations.     

Increased reactivity of responding cells in the mixed lymphocyte reaction by a thymic humoral factor.

A thymic humoral factor (THF) induces competence in a certain sector of the lymphoid cell population. The target cell for THF activity was studied using the one-way mixed lymphocyte culture (MLC). Cells from various lymphoid tissues (responding cells) were preincubated in THF for 1 hour and then mixed with mitomycin-C-treated cells (stimulating cells). The results obtained indicate that THF increases MLC competence of lymphocytes from spleen and thymus origin of intact mice and that of spleen of thymus-deprived mice. In contrast, no enhancing effect in the MLC response by THF could be observed when bone marrow-derived spleen cells, cortisone-resistant thymocytes or lymph node cells were used as responding cells. These results confirm our hypothesis that THF leads to differentiation of young thymus-derived (T) cells, promoting them to maturation and thus to the acquisition of full immune competence.     

A controlled trial of early versus late treatment with zidovudine in symptomatic human immunodeficiency virus infection. Results of the Veterans Affairs Cooperative Study.

BACKGROUND. Zidovudine is recommended for asymptomatic and early symptomatic human immunodeficiency virus (HIV) infection. The best time to initiate zidovudine treatment remains uncertain, however, and whether early treatment improves survival has not been established. METHODS. We conducted a multicenter, randomized, double-blind trial that compared early zidovudine therapy (beginning at 1500 mg per day) with late therapy in HIV-infected patients who were symptomatic and had CD4+ counts between 0.2 x 10(9) and 0.5 x 10(9) cells per liter (200 to 500 per cubic millimeter) at entry. Those assigned to late therapy initially received placebo and began zidovudine when their CD4+ counts fell below 0.2 x 10(9) per liter (200 per cubic millimeter) or when the acquired immunodeficiency syndrome (AIDS) developed. RESULTS. During a mean follow-up period of more than two years, there were 23 deaths in the early-therapy group (n = 170) and 20 deaths in the late-therapy group (n = 168) (P = 0.48; relative risk [late vs. early], 0.81; 95 percent confidence interval, 0.44 to 1.59). In the early-therapy group, 28 patients progressed to AIDS, as compared with 48 in the late-therapy group (P = 0.02; relative risk, 1.76; 95 percent confidence interval, 1.1 to 2.8). Early therapy increased the time until CD4+ counts fell below 0.2 x 10(9) per liter (200 per cubic millimeter), and it produced more conversions from positive to negative for serum p24 antigen. Early therapy was associated with more anemia, leukopenia, nausea, vomiting, and diarrhea, whereas late therapy was associated with more skin rash. CONCLUSIONS. In symptomatic patients with HIV infection, early treatment with zidovudine delays progression to AIDS, but in this controlled study it did not improve survival, and it was associated with more side effects.     

A controlled trial of trimethoprim-sulfamethoxazole or aerosolized pentamidine for secondary prophylaxis of Pneumocystis carinii pneumonia in patients with the acquired immunodeficiency syndrome. AIDS Clinical Trials Group Protocol 021.

BACKGROUND. Pneumocystis carinii pneumonia (PCP) continues to be the most common index diagnosis in the acquired immunodeficiency syndrome (AIDS), but it is not clear which of several available agents is the most effective in preventing a recurrence of PCP. METHODS. We conducted a comparative, open-label trial in 310 adults with AIDS who had recently recovered from an initial episode of PCP and had no treatment-limiting toxic effects of trimethoprim-sulfamethoxazole or pentamidine. All the patients were treated with zidovudine and were randomly assigned to receive either 800 mg of sulfamethoxazole and 160 mg of trimethoprim once daily or 300 mg of aerosolized pentamidine administered every four weeks by jet nebulizer. The participants were followed for a median of 17.4 months. RESULTS. In the trimethoprim-sulfamethoxazole group (n = 154) there were 14 recurrences of PCP, as compared with 36 recurrences (including 1 extrapulmonary recurrence) in the aerosolized-pentamidine group (n = 156). The estimated recurrence rates at 18 months were 11.4 percent with trimethoprim-sulfamethoxazole and 27.6 percent with pentamidine (P < 0.001). The risk of a recurrence (adjusted for initial CD4 cell count) was 3.25 times higher in the pentamidine group (P < 0.001, 95 percent confidence interval, 1.72 to 6.16). There were no significant differences between the groups in survival or in hematologic or hepatic toxicity. Crossovers from trimethoprim-sulfamethoxazole to aerosolized pentamidine were more common than the reverse (27 vs. 4 percent), partly because of the study protocols for the management of leukopenia. There were 19 serious bacterial infections in the trimethoprim-sulfamethoxazole group and 38 in the pentamidine group. The time to a first bacterial infection was significantly greater for those assigned to trimethoprim-sulfamethoxazole (P = 0.017). CONCLUSIONS. In patients with AIDS who are receiving zidovudine, trimethoprim-sulfamethoxazole is more effective than aerosolized pentamidine in conventional doses for the prevention of recurrent pneumocystis infection.