Geometric diffusions as a tool for harmonic analysis and structure definition of data: multiscale methods.

In the companion article, a framework for structural multiscale geometric organization of subsets of R(n) and of graphs was introduced. Here, diffusion semigroups are used to generate multiscale analyses in order to organize and represent complex structures. We emphasize the multiscale nature of these problems and build scaling functions of Markov matrices (describing local transitions) that lead to macroscopic descriptions at different scales. The process of iterating or diffusing the Markov matrix is seen as a generalization of some aspects of the Newtonian paradigm, in which local infinitesimal transitions of a system lead to global macroscopic descriptions by integration. This article deals with the construction of fast-order N algorithms for data representation and for homogenization of heterogeneous structures.

Geometric diffusions as a tool for harmonic analysis and structure definition of data: diffusion maps.

We provide a framework for structural multiscale geometric organization of graphs and subsets of R(n). We use diffusion semigroups to generate multiscale geometries in order to organize and represent complex structures. We show that appropriately selected eigenfunctions or scaling functions of Markov matrices, which describe local transitions, lead to macroscopic descriptions at different scales. The process of iterating or diffusing the Markov matrix is seen as a generalization of some aspects of the Newtonian paradigm, in which local infinitesimal transitions of a system lead to global macroscopic descriptions by integration. We provide a unified view of ideas from data analysis, machine learning, and numerical analysis.

Hypersensitivity reactions during therapy with the nucleoside reverse transcriptase inhibitor abacavir.

BACKGROUND: Hypersensitivity reactions consist of a variable group of clinical findings and have been described for a wide variety of chemical compounds. OBJECTIVE: This review characterizes the clinical profile of hypersensitivity to the nucleoside reverse transcriptase inhibitor abacavir sulfate. METHODS: We performed a retrospective medical review of pooled adverse events data from approximately 200,000 patients who received abacavir in clinical trials, through expanded-access programs, or by prescription from 1996 through 2000. Screened cases of hypersensitivity were classified as either definitive or probable. Definitive cases were identified when initial symptoms resolved on interruption of abacavir therapy and returned on reintroduction of abacavir therapy. RESULTS: A total of 1803 cases were identified, 1302 in the 30,595 patients participating in clinical trials or the expanded-access program and 501 in patients from the post-marketing experience. On review, 176 (9.8%) of these cases were considered definitive and the remainder probable. Based on the 1302 cases identified in clinical trials or the expanded-access program, the calculated incidence of hypersensitivity was 4.3%. Symptoms reported in > or = 20% of cases of this multiorgan reaction included fever, rash, malaise/fatigue, and gastrointestinal symptoms such as nausea, vomiting, and diarrhea, among others. Respiratory symptoms occurred in 30% of cases and included dyspnea (12%), cough (10%), and pharyngitis (6%). In 90% of cases, hypersensitivity reactions occurred within the first 6 weeks after initiation of abacavir (median time, 11 days); after an initial reaction, rechallenge with abacavir resulted in the reappearance of symptoms within hours of reexposure. Hypotension was present in 25% of these rechallenge reactions. Among patients who received abacavir in clinical trials, the mortality rate was 0.03% (3 per 10,000 patients). CONCLUSIONS: Hypersensitivity to abacavir is an idiosyncratic reaction and a distinct clinical syndrome characterized predominantly by systemic involvement. It can be expected to appear as a treatment-limiting event in approximately 5% of patients. The appearance of clinical symptoms consistent with this syndrome mandates immediate discontinuation of abacavir. Hypersensitivity to abacavir is an absolute contraindication to subsequent treatment with any formulation that includes this agent.

Use of drug effect interaction modeling with Monte Carlo simulation to examine the impact of dosing interval on the projected antiviral activity of the combination of abacavir and amprenavir.

The delineation of optimal regimens for combinations of agents is a difficult problem, in part because, to address it, one needs to (i) have effect relationships between the pathogen in question and the drugs in the combination, (ii) have knowledge of how the drugs interact (synergy, antagonism, and additivity), and (iii) address the issue of true between-patient variability in pharmacokinetics for the drugs in the population. We have developed an approach which employs a fully parametric assessment of drug interaction using the equation of W. R. Greco, G. Bravo, and J. C. Parsons (Pharmacol. Rev. 47:331-385, 1995) to generate an estimate of effects for the two drugs and have linked this approach to a population simulator, using Monte Carlo methods, which produce concentration-time profiles for the drugs in combination. This software automatically integrates the effect over a steady-state dosing interval and produces an estimate of the mean effect over a steady-state interval for each simulated subject. In this way, doses and schedules can be easily evaluated. This software allows for a rational choice of dose and schedule for evaluation in clinical trials. We evaluated different schedules of administration for the combination of the nucleoside analogue abacavir plus the human immunodeficiency virus type 1 protease inhibitor amprenavir. Amprenavir was simulated as either 800 mg every 8 h (q8h) or 1,200 mg q12h, each along with 300 mg q12h of abacavir. Both regimens produced excellent effects over the simulated population of 500 subjects, with average percentages of maximal effect (as determined from the in vitro assays) of 90.9%+/- 11.4% and 80.9%+/-18.6%, respectively. This difference is statistically significant (P<<0.001). In addition, 68.8 and 46.0% of the population had an average percentage of maximal effect which was greater than or equal to 90% for the two regimens. We can conclude that the combination of abacavir plus amprenavir is a potent combination when it is given on either schedule. However, the more fractionated schedule for the protease inhibitor produced significantly better effects in combination. Clinicians need to explicitly balance the improvement in antiviral effect seen with the more fractionated regimen against the loss of compliance attendant to the use of such a regimen. This approach may be helpful in the preclinical evaluation of multidrug anti-infective regimens.

Combination therapy with amprenavir, abacavir, and efavirenz in human immunodeficiency virus (HIV)-infected patients failing a protease-inhibitor regimen: pharmacokinetic drug interactions and antiviral activity.

Patients with plasma viral RNA >50,000 copies/mL, despite a protease-inhibitor regimen, received abacavir, amprenavir, and efavirenz to assess efavirenz-amprenavir drug interactions and to evaluate safety and antiviral response. Patients first received amprenavir with abacavir and other nucleoside analogs. Amprenavir levels were measured before and after adding efavirenz. Patients then received a second protease inhibitor. There was evidence of genotypic and phenotypic resistance at study entry. No patient had study drugs discontinued because of toxicity. Efavirenz decreased the steady-state area under the curve, maximum plasma concentration, and minimum plasma concentration of amprenavir by 24%, 33%, and 43%, respectively. Three of 10 patients had >1.5 log10 viral response to abacavir and amprenavir. All 8 patients who added efavirenz had >0.5 log10 decline in viral load, and this response lasted >24 weeks for 3 of the patients. A combination regimen that included abacavir, amprenavir, and efavirenz was well tolerated and had sustained activity in some patients. Concomitant efavirenz therapy decreases amprenavir concentrations.

Atovaquone suspension in HIV-infected volunteers: pharmacokinetics, pharmacodynamics, and TMP-SMX interaction study.

STUDY OBJECTIVE: To evaluate the pharmacokinetics and safety of atovaquone suspension in volunteers infected with the human immunodeficiency virus ((HIV). DESIGN: Open-label, nonrandomized study. SETTING: Two clinical research centers. PATIENTS: Twenty-two HIV-infected volunteers with a median CD4 cell count of 37 cells/mm3. INTERVENTIONS: Patients received atovaquone suspension fasting or fed for 2-week periods with crossover at dosages of 500 mg/day, and randomization to fasting or fed at dosages of 750 and 1000 mg/day. A subset of patients also received 750 mg twice/day with food, and a subset of those who received 1000 mg/day fasting also received 1000 mg with food. During a long-term dosing phase, a subset of subjects were evaluated for an interaction between atovaquone and trimethoprim-sulfamethoxazole (TMP-SMX). MEASUREMENTS AND MAIN RESULTS: Average steady-state atovaquone concentrations at 500 mg were 6.7 +/- 3.2 microg/ml fasted and 11.3 +/- 5.0 microg/ml with food; at 750 mg, 9.9 +/- 7.1 microg/ml fasted and 12.5 +/- 5.9 microg/ml with food; at 1000 mg, 9.7 +/- 4.3 microg/ml fasted and 13.6 +/- 5.0 microg/ml with food; and at 1500 mg, 21.1 +/- 5.0 microg/ml with food. Thus, plasma concentrations were not proportional to dose. Concomitant food ingestion resulted in a 1.3- to 1.7-fold increase in values. Average steady-state concentrations were less than 10 microg/ml in 21% and more than 15 microg/ml in 36% of patients at 1000 mg/day with food; at 750 mg twice/day, all five patients had levels above 15 microg/ml. Atovaquone suspension was well tolerated; diarrhea, nausea, fatigue, and rash were the most common adverse events. Concomitant administration of TMP-SMX did not change atovaquone concentrations and resulted in small decreases in concentrations of TMP (16%) and SMX (10%). CONCLUSION: Plasma concentrations are significantly higher when atovaquone suspension is administered with food compared with fasting. Total doses of 1500 mg/day are likely to achieve concentrations effective for prophylaxis of Pneumocystis carinii pneumonia.

Safety and pharmacokinetics of abacavir (1592U89) following oral administration of escalating single doses in human immunodeficiency virus type 1-infected adults.

Abacavir (1592U89) is a nucleoside analog reverse transcriptase inhibitor that has been demonstrated to have selective activity against human immunodeficiency virus (HIV) in vitro and favorable safety profiles in mice and monkeys. A phase I study was conducted to evaluate the safety and pharmacokinetics of abacavir following oral administration of single escalating doses (100, 300, 600, 900, and 1,200 mg) to HIV-infected adults. In this double-blind, placebo-controlled study, subjects with baseline CD4+ cell counts ranging from < 50 to 713 cells per mm3 (median, 315 cells per mm3) were randomly assigned to receive abacavir (n = 12) or placebo (n = 6). The bioavailability of the caplet formulation relative to that of the oral solution was also assessed with the 300-mg dose. Abacavir was well tolerated by all subjects; mild to moderate asthenia, abdominal pain, headache, diarrhea, and dyspepsia were the most frequently reported adverse events, and these were not dose related. No significant clinical or laboratory abnormalities were observed throughout the study. All doses resulted in mean abacavir concentrations in plasma that exceeded the mean 50% inhibitory concentration (IC50) for clinical HIV isolates in vitro (0.07 microgram/ml) for almost 3 h. Abacavir was rapidly absorbed following oral administration, with the time to the peak concentration in plasma occurring at 1.0 to 1.7 h postdosing. Mean maximum concentrations in plasma (Cmax) and the area under the plasma concentration-time curve from time zero to infinity (AUC0-infinity) increased slightly more than proportionally from 100 to 600 mg (from 0.6 to 4.7 micrograms/ml for Cmax; from 1.0 to 15.7 micrograms.h/ml for AUC0-infinity) but increased proportionally from 600 to 1,200 mg (from 4.7 to 9.6 micrograms/ml for Cmax; from 15.7 to 32.8 micrograms.h/ml for AUC0-infinity. The elimination of abacavir from plasma was rapid, with an apparent elimination half-life of 0.9 to 1.7 h. Abacavir was well absorbed, with a relative bioavailability of the caplet formulation of 96% versus that of an oral solution (drug substance in water). In conclusion, this study showed that abacavir is safe and is well tolerated by HIV-infected subjects and demonstrated predictable pharmacokinetic characteristics when it was administered as single oral doses ranging from 100 to 1,200 mg.

Antiretroviral effect and safety of abacavir alone and in combination with zidovudine in HIV-infected adults. Abacavir Phase 2 Clinical Team.

OBJECTIVES: To evaluate, over 12 weeks, the antiretroviral activity and safety of abacavir, used alone and in combination with zidovudine (ZDV), as treatment for HIV-1-infected subjects who had limited or no antiretroviral treatment. DESIGN: Seventy-nine HIV-1-infected subjects, with CD4 cell counts 200-500 x 10(6)/l and <12 weeks of previous treatment with ZDV were enrolled in a multicenter study. Subjects were randomly assigned to one of four cohorts receiving abacavir monotherapy for the first 4 weeks (200, 400, or 600 mg every 8 h daily, or 300 mg every 12 h daily) and, thereafter, combination therapy of abacavir with 600 mg ZDV or ZDV placebo, administered in a double-blind manner for an additional 8 weeks. METHODS: Antiretroviral activity was assessed by measuring changes in plasma HIV-1 RNA levels and CD4+ cell counts. Safety was assessed by monitoring clinical adverse events and laboratory abnormalities during the 12-week period and for 4 weeks post-treatment. RESULTS: Treatment with abacavir, alone or in combination with ZDV, produced marked decreases in plasma HIV-1 RNA loads and increases in CD4+ cell counts in all groups. At week 4, median plasma HIV-1 RNA loads decreased by 1.11-1.77 log10 copies/ml and median CD4+ cell counts increased by 63-111 x 10(6)/l in all groups. At week 12, median HIV-1 RNA loads decreased by 1.02-2.24 log10 copies/ml (abacavir monotherapy) and by 1.81-2.01 log10 copies/ml (abacavir-ZDV); median CD4+ cell counts increased by 79-195 x 10(6)/l (abacavir monotherapy) and by 93-142 x 10(6)/l (abacavir-ZDV). At week 12, the percentage of subjects who had plasma HIV-1 RNA levels below 400 and 40 copies/ml were 28 and 11%, respectively (abacavir monotherapy) and 69 and 22%, respectively (abacavir-ZDV). Eight subjects (10%) discontinued the study prematurely because of adverse events; nausea (n = 4) and hypersensitivity (n = 3) were the most common reasons for withdrawal. There were no deaths among the study subjects. CONCLUSIONS: In HIV-infected subjects who have received little or no prior antiretroviral therapy, treatment with abacavir alone or in combination with ZDV is well tolerated and resulted in sustained improvements in key immunologic and virologic efficacy parameters through 12 weeks.

Effects of aerosolized synthetic surfactant, atovaquone, and the combination of these on murine Pneumocystis carinii pneumonia.

An immunosuppressed rat model was used to determine the pharmacokinetics of aerosolized atovaquone (administered with and without a synthetic surfactant) and to evaluate the efficacy of inhaled atovaquone in the prevention and treatment of Pneumocystis carinii pneumonia (PCP). After a single dose by aerosol, mean peak concentrations of atovaquone averaged 52 microg/mL in plasma and 31 microg/g in lungs of rats infected with P. carinii. When atovaquone was combined with surfactant, mean peak concentrations of 94 microg/mL in plasma and 51 microg/g in lung were achieved. Aerosolized synthetic surfactant alone significantly increased survival of rats with PCP and, when combined with atovaquone, increased plasma and lung concentrations of the drug and eradication of the organism.

A simple Trypanosoma cruzi enzyme-linked immunoassay for control of human infection in nonendemic areas.

An enzyme linked immunosorbent assay (ELISA) was developed for detecting IgM and IgG antibodies against Trypanosoma cruzi in blood bank donors from endemic or nonendemic areas. A crude extract of trypomastigotes from cultures was used as antigen. A total of 494 serum samples from patients with acute, congenital, or chronic form of Chagas' disease, and from healthy French individuals were studied. The sensitivity of the ELISA was determined with 89 serum samples from chagasic patients and was evaluated to 98.8%. The specificity was determined with 405 serum samples from French blood transfusion centers donors and evaluated to 98.3%. Two hundred and eighty-five serum samples from blood donors from Argentina and Brazil were also tested. Furthermore, in order to assess the absence of cross-reactivity with other protozoan infections, we studied 86 serum samples including (i) 32 individuals with cutaneous leishmaniasis living in a T. cruzi endemic region of Bolivia, and (ii) 54 patients from nonendemic area for Chagas' disease, 19 of them with kala-azar and 35 others with malaria.

Leishmaniasis in Texas: epidemiology and clinical aspects of human cases.

Twenty-seven autochthonous cases of cutaneous leishmaniasis in Texas were identified by contact with dermatologists and State Health Department officials, and by a review of medical records, pathology reports, and previously published case reports. Fifteen cases were previously unreported. Although the date of onset of the first recognized case was 1903, in 20 of the cases the date of onset of the lesion(s) was in 1980-1989. Twelve cases were female; 15 were male. Age at diagnosis ranged from two to 86 (median 37) years. The disease was identified significantly more frequently in younger males and older females. The distribution of cases closely followed the distribution of Neotoma micropus, a rodent host for Leishmania mexicana. The most common risk factor appeared to be residence or activity in close proximity to woodrat habitat. Two cases lived in central Texas; the remainder had a residence in, or history of travel to, southern Texas. A majority of cases were first recognized during the cooler months of the year. Most lesions began as papules or nodules that subsequently ulcerated. In 20 cases, a single lesion was present. Five cases had resolution of their lesions without receiving specific anti-leishmanial therapy; lesions of 17 resolved after treatment with a variety of therapies. One life-long case of disseminated disease failed to respond to treatment, and four cases were lost to follow-up. A Leishmania-specific lymphocyte proliferation assay gave a positive response for four of five cases tested. Screening of 13 family members found no evidence of subclinical infection. These 27 cases, and two recently recognized cases reported in a note added in proof, indicate that cutaneous leishmaniasis may be more common in Texas than previously thought.

Prevalence of anti-R-13 antibodies in human Trypanosoma cruzi infection.

Infection with Trypanosoma cruzi develops in three phases: acute, indeterminate or asymptomatic, and chronic phase (with cardiac or digestive manifestations). Moreover, transmission may occur from infected mothers to newborn, the so-called congenital form. In the present study, humoral responses against T. cruzi total extract and against the 13 amino acid peptide named R-13 derived from the parasite ribosomal P protein, previously described as a possible marker of chronic Chagas heart disease, were determined in chagasic patients and in blood bank donors from endemic areas. While in sera from acute phase, only IgM anti-T.cruzi response was observed, both IgM and IgG anti-T. cruzi antibodies were detected in sera from congenitally infected newborns. The percentage of positive response in sera from blood bank donors was relatively high in endemic regions. Antibodies against the R-13 peptide were present in a large proportion of cardiac chagasic patients but were totally lacking in patients with digestive form of Chagas' disease. Furthermore, anti-R-13 positive responses were detected in congenitally infected newborns.

Clinical experience with atovaquone on a treatment investigational new drug protocol for Pneumocystis carinii pneumonia.

The clinical experience of human immunodeficiency virus (HIV) + patients treated with oral atovaquone for acute Pneumocytstis carinii pneumonia (PCP) under a Treatment Investigational New Drug (IND) protocol (mild or moderate PCP) and an Open-Label Study protocol (severe PCP) was evaluated. A total of 940 patients intolerant of or unresponsive to trimethoprimsulfamethoxazole were enrolled from private practices, clinics, and institutional HIV treatment centers in the United States. Demographics data and the history and severity of PCP were collected at enrollment. The number of therapy days, adverse experiences, clinical response to therapy, and mortality were collected at day 21. Reporting of serious, unexpected adverse experiences attributable to therapy was required. Of the 760 (96%) patients with mild to moderate disease for whom follow-up observation was complete, 591 (78%) responded clinically to treatment, 177 patients (23%) discontinued treatment prematurely, and 50 patients (7%) died. Of the 140 patients (95%) with severe PCP with follow-up data, 79 (56%) responded to treatment, 45 (32%) discontinued treatment early, and 53 patients (38%) died. Adverse events that resulted in temporary or permanent discontinuation of therapy included diarrhea, vomiting, elevated liver enzyme levels, nausea, and fever. No serious unexpected adverse events attributable to the drug were reported. The treatment IND mechanism enabled a large number of patients with acute PCP to be treated with this experimental therapy while the drug was under regulatory view.

Adverse events associated with trimethoprim-sulfamethoxazole and atovaquone during the treatment of AIDS-related Pneumocystis carinii pneumonia.

Atovaquone was compared to trimethoprim-sulfamethoxazole (TMP-SMZ) for the relationship of time receiving therapy, plasma drug concentrations, and incidence of adverse reactions in patients with AIDS-associated Pneumocystis carinii pneumonia. Treatment-limiting adverse events occurred in 9% of atovaquone-treated patients and 24% of TMP-SMZ-treated patients. Adverse events usually did not occur before day 7 for either treatment. Only the incidence of rash increased with increasing plasma concentrations of atovaquone. The incidence of anemia, neutropenia, and azotemia increased with increasing trimethoprim plasma concentration, while other adverse events (gastrointestinal disorders, rash, fever, and liver function abnormalities) were independent of plasma drug concentration.

The chronic presence of the parasite, and anti-P autoimmunity in Chagas disease: the Trypanosoma cruzi ribosomal P proteins, and their recognition by the host immune system.

Molecular expression cloning techniques revealed that patients with the severest clinical form of Chagas disease, chronic Chagas heart disease, presented a strong humoral response against the cloned C-terminal portion of a Trypanosoma cruzi ribosomal P protein. Parasite P antigens identification led to characterize the ribosomal P protein system in T. cruzi. Their exposed location on the ribosome, and the "amplification" of their parasite specific, serine free C-terminal domain, generate a strong parasite specific anti-P response, that in certain cases may induce anti-P autoimmunity.

The chronic presence of the parasite, and anti-P autoimmunity in Chagas disease: the Trypanosoma cruzi ribosomal P proteins, and their recognition by the host immune system

Molecular expression cloning techniques revealed that patients with the severest clinical form of Chagas disease, chronic Chagas heart disease, presented a strong humoral response against the cloned C-terminal portion of a Trypanosoma cruzi ribosomal P protein. Parasite P antigens identification led to characterize the ribosomal P protein system in T. cruzi. Their exposed location on the ribosome, and the ®amplification® of their parasite specific, serine free C-terminal domain, generate a strong parasite specific anti-P response, that in certain cases may induce anti-P autoimmunity

The 70-kDa heat-shock protein is a major antigenic determinant in human Trypanosoma cruzi/Leishmania braziliensis braziliensis mixed infection.

Five sera from Bolivian individuals chronically infected by Trypanosoma cruzi, and suffering an active Leishmania braziliensis braziliensis metastatic mucocutaneous lesion were characterized. They reacted with the T. cruzi recombinant antigens that are currently used as Chagas diagnostic reagents, and with several L. b. braziliensis proteins as assessed by Western blot. These sera showed an intense reaction with a T. cruzi and an L. b. braziliensis polypeptide of about 70 kDa. Expression cloning techniques demonstrated that the target of this immunologic reaction was a cross-reactive antigen, the 70-kDa heat-shock protein (HSP 70). High levels of anti-HSP 70 reactivity and positive reactions with all or some of the T. cruzi recombinant antigens JL7, JL8, and JL5, defined a serologic pattern that was characteristic of the T. cruzi/L. b. braziliensis mixed infection.

Recombinant Trypanosoma cruzi antigens and Chagas' disease diagnosis: analysis of a workshop.

A workshop organized by the Ibero-American Project of Biotechnology evaluated the diagnostic potential of several cloned Trypanosoma cruzi recombinant antigens for Chagas' disease serodiagnosis. A set of recombinants, Antigen 2, Antigen 13, SAPA, H49, A13, JL5, JL7, JL8, JL9, and RA1 provided by three different South American laboratories were probed with a panel of 236 South American serum samples. Antigens JL7, H49, Antigen 2, and A13 scored as the best diagnostic recombinant reagents. The results suggested that the main advantage of using cloned peptides for chronic Chagas' disease diagnosis resided in their highly specific immunoreactive properties.

A preliminary evaluation of 566C80 for the treatment of Pneumocystis pneumonia in patients with the acquired immunodeficiency syndrome.

BACKGROUND: The drug 566C80 is an investigational hydroxynaphthoquinone that is active against Pneumocystis carinii in vitro and in animal models. Initial studies in humans indicate that 566C80 is safe and has adequate bioavailability after oral administration. METHODS: We conducted an open-label trial of 566C80 in 34 adults with the acquired immunodeficiency syndrome (AIDS) and untreated pneumocystis pneumonia. All the patients had a partial pressure of arterial oxygen of at least 60 mm Hg while breathing room air. They were enrolled sequentially in three cohorts taking 566C80 at different dosages, all administered orally: 750 mg three times daily for 5 days, then twice daily for 16 days; 750 mg three times daily for 21 days; and 750 mg four times daily for 21 days. RESULTS: All 34 patients survived, and 27 (79 percent) were successfully treated with 566C80 alone. The mean partial pressure of oxygen in 33 patients was 78 mm Hg at entry and 93 mm Hg after the course of 566C80 (P less than 0.001). In five patients (15 percent) the drug was discontinued because of lack of response. In four patients (12 percent), the drug was discontinued because of toxicity (fever and rash in two patients each). In two of these, treatment was considered to have succeeded because 566C80 was not discontinued because of toxicity until after day 14. Five of the successfully treated patients had rashes that resolved despite continued therapy. In nine patients, serum alanine aminotransferase levels rose above 100 U per liter. During the first three months after the completion of therapy, pneumocystis pneumonia recurred in 4 of the 27 successfully treated patients, and another 3 patients had recurrences between month 3 and month 6 of follow-up. The mean (+/- SEM) steady-state plasma levels of 566C80 were similar in the three cohorts: 16.3 +/- 2.10, 20.4 +/- 2.48, and 18.9 +/- 3.08 micrograms per milliliter in the patients taking the drug twice daily, three times daily, and four times daily, respectively. CONCLUSIONS: From these preliminary data, the investigational compound 566C80 appears to be a safe, effective, and well-tolerated therapy for P. carinii pneumonia of mild-to-moderate severity in patients with AIDS.