https://orcid.org/0000-0003-4566-4030
Figures
Abstract
Introduction
The currently recommended 8-week daily benznidazole regimen for Chagas disease is poorly tolerated. While shorter benznidazole monotherapy and combination regimens have been explored, the pharmacokinetic/pharmacodynamic (PK/PD) relationship remains poorly understood.
Objectives
i) To describe the population pharmacokinetics of benznidazole and assess drug–drug interactions with fosravuconazole in patients with chronic Chagas disease, ii) to explore the relationship between benznidazole exposure and anti-trypanosomal treatment effects.
Methods
This was a secondary analysis based on data from the previously published BENDITA study (NCT03378661), a dose evaluation trial in adults with chronic indeterminate Chagas disease (n = 210). Patients were randomized to placebo, the standard benznidazole dose (300 mg/day for 8 weeks), or lower total dose regimens (300 mg/day for 4 or 2 weeks; 150 mg/day for 4 weeks alone or combined with fosravuconazole 300 mg/week; 300 mg/week for 8 weeks plus fosravuconazole 300 mg/week). Benznidazole pharmacokinetics were evaluated using nonlinear mixed-effects modeling. The relationship between individual benznidazole exposure and the pharmacodynamic (PD) endpoint was explored using beta binomial regression. The PD endpoint (qPCR positivity) was defined as the proportion of qPCR-positive blood samples collected post-treatment over 12 months of follow-up, capturing the frequency of detectable parasitemia per patient.
Results
Benznidazole pharmacokinetics were well described by a transit-absorption model with one-compartment disposition. Bioavailability was 13% lower in men than in women, and coadministration of fosravuconazole increased benznidazole clearance by 18% (both effects considered not clinically relevant). In the placebo arm, nearly all patients (97%) remained qPCR positive, with most showing qPCR positivity above 40%. Among patients receiving benznidazole, post-treatment qPCR positivity was substantially lower. In the 2-week arm, three patients had multiple positive qPCR samples (up to 43% PCR positivity). In contrast, individual qPCR positivity in the 4–8-week arms did not exceed 20% (i.e., one or no positive samples), with one non-adherent exception. The PK/PD analysis did not identify a significant pharmacokinetic driver of treatment response. While the study was not powered for between-arm comparisons, the findings suggest that lower total dose regimens (4 weeks daily or 8 weeks weekly) may provide efficacy comparable to the standard 8-week regimen.
Author summary
Chagas Disease, caused by the parasite Trypanosoma cruzi, remains a major public health challenge, particularly in Latin America. Benznidazole and nifurtimox have been the standard treatment for the past 50 years, but they have severe limitations, notably their lengthy 8-week regimen and poor tolerability. In this study, we analyzed data from the BENDITA trial, a randomised dose evaluation trial conducted in Bolivian in patients with chronic Chagas disease, which explored short-course benznidazole regimens and combination therapy with fosravuconazole. To gain a better understanding of individual treatment responses, we investigated the relationship between benznidazole plasma exposures and the risk of parasitological recrudescence, considering dose, adherence, and variability in pharmacokinetics. We successfully described benznidazole’s pharmacokinetic behavior and showed that fosravuconazole does not have a clinically relevant impact on benznidazole’s pharmacokinetics. While benznidazole was clearly efficacious, no clear association between drug exposure metrics and antiparasitic effects was observed among the benznidazole treatment arms. Daily dosing for 4 weeks or once weekly dosing for 8 weeks with fosravuconazole appeared as effective as the standard 8-week daily course. The results suggest that the standard regimen may be excessive. Shorter treatment courses or once weekly dosing could improve adherence, tolerability, and cost-effectiveness. Future well-designed studies, ideally using factorial randomisation, are needed to identify the optimal dose, duration, and frequency of benznidazole administration which improves tolerability without compromising efficacy.
Citation: Assmus F, Cruz C, Watson JA, White NJ, Adehin A, Hoglund RM, et al. (2025) Population pharmacokinetic-pharmacodynamic analysis of benznidazole monotherapy and combination therapy with fosravuconazole in chronic Chagas disease (BENDITA). PLoS Negl Trop Dis 19(9): e0013522. https://doi.org/10.1371/journal.pntd.0013522
Editor: Shaden Kamhawi, National Institutes of Health, UNITED STATES OF AMERICA
Received: April 15, 2025; Accepted: September 2, 2025; Published: September 22, 2025
Copyright: © 2025 Assmus et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Pseudonymised participant data used in this analysis are available for access through the Infectious Diseases Data Observatory (IDDO, www.iddo.org), which provides a secure platform for controlled access to clinical and pharmacokinetic datasets. IDDO is registered with the Registry of Research Data Repositories (re3data.org). Requests for access will be reviewed by IDDO Data Access Committee to ensure that use of data protects the interests of the participants and researchers according to the terms of ethics approval and principles of equitable data sharing. Guidance on how to request data is available at: https://www.iddo.org/chagas/data-reuse/accessing-data. Requests can be submitted by email to: dataaccess@iddo.org.
Funding: DNDi received funding from UK Aid, UK; and for its overall mission, from Médecins Sans Frontières International. This research is funded in part by the Wellcome Trust grant “Optimising pharmacometric assessment in phase 2 studies of Chagas disease” (222754/Z/21/Z). JAW is a Sir Henry Dale Fellow funded by the Wellcome Trust (223253/Z/21/Z). NJW is a Principal Research Fellow funded by the Wellcome Trust (093956/Z/10/C). This research was partly funded by the Wellcome Trust [220211] awarded to RMH and JT. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. For the purpose of Open Access, the author has applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Chagas disease (CD) is a neglected, chronic, potentially life-threatening disease caused by the protozoan parasite Trypanosoma cruzi [1]. Approximately 6–7 million people are infected worldwide, with an additional 75 million at risk [1]. While the largest burden of disease is found in Latin America, increasing population mobility has spread CD to non-endemic regions, turning it into a global public health concern [2,3]. CD progresses from an often asymptomatic acute phase to a chronic phase, marked by very low parasite densities in blood [4,5]. Most individuals remain asymptomatic, but after several decades about one-third develop severe cardiac or gastrointestinal complications, leading to substantial mortality and morbidity [1,4,6,7].
Current treatment options are limited to two drugs, nifurtimox and benznidazole, which were discovered over half a century ago and are far from ideal [8]. Benznidazole is commonly used, and its mechanism likely involves nitroreduction by parasite nitroreductases, generating reactive intermediates that disrupt vital biological functions [9]. Although both drugs are highly effective during the acute phase, their efficacy in the chronic phase – particularly in adults – and their ability to halt progression vary [6,8,10,11]. Treatment is lengthy (30–60 days [12]) and often associated with adverse drug reactions (ADRs), which are more frequent in adults than children [13]. Across studies, approximately half of adult patients experienced ADRs, mostly mild to moderate, with fewer than 1.5% classified as severe [8,13]. Treatment discontinuation was reported in 10–29% of patients, primarily due to skin reactions [8,13]. However, field programs by Médecins Sans Frontières have shown that with close treatment surveillance adherence improves and most patients complete therapy, highlighting the importance of early ADR management, patient counselling, and follow-up [14,15]. Despite this, improved regimens or new treatments with better tolerability remain urgently needed [16].
Combination therapy has been proposed to enhance efficacy and tolerability by combining drugs with different mechanisms of action [17,18]. Azole antifungals such as posaconazole and fosravuconazole (a ravuconazole prodrug) inhibit T. cruzi sterol 14α-demethylase (CYP51), essential for membrane integrity [19]. While these ergosterol inhibitors have shown trypanostatic effects in preclinical studies [20], they failed to achieve sustained parasite clearance in monotherapy trials [21–23]. However, combining azoles with benznidazole has demonstrated additive or synergistic effects in murine models, providing a rationale for clinical evaluation [17,24,25].
Recent initiatives have focused on re-evaluating benznidazole dosing regimens, as current protocols rely on limited data [26–28], and the optimal dose, duration, and frequency remain uncertain. Some studies suggest that the standard benznidazole dose of 5–8 mg/kg/day for 60 days may be excessive, so dose reduction could improve tolerability without compromising efficacy [29–31]. Alternative strategies, such as shorter treatment durations [32–37], intermittent dosing [38–41], combination therapy [24,32], and extended treatment durations [42], have been explored in preclinical studies and clinical trials.
The BENDITA trial, sponsored by the Drugs for Neglected Diseases initiative (DNDi), was a phase 2 clinical trial assessing various benznidazole dosing regimens in adult patients with chronic indeterminate CD. The regimens included shorter courses, reduced daily doses, either as monotherapy or in combination with fosravuconazole, and intermittent weekly combination therapy. Efficacy and safety data for these regimens have been reported previously [32]. Although the overall incidence of adverse events was not significantly different between treatment groups, the standard 8-week regimen (300 mg daily) had the highest rate of treatment discontinuations due to adverse events. Moreover, neutropenia, leukopenia, and headaches were reported slightly more frequently in this group. All benznidazole arms demonstrated higher antiparasitic responses compared to placebo. However, the relationship between benznidazole dose, plasma exposure, and variability in treatment response remains poorly understood.
Establishing pharmacokinetic/pharmacodynamic (PK/PD) relationships in CD research is necessary to guide dosing optimization, but efforts have been hampered by the lack of validated biomarkers for clinical efficacy and parasitological cure [43,44]. Since seroconversion typically takes years to achieve, most recent trials have relied on sustained PCR negativity (usually over one year) as a binary efficacy endpoint [44]. However, PCR is limited by the low parasite densities in blood which are very close to the limit of detection (circa 1 parasite/10mL). In previous work, we have shown the limitations of binary PCR endpoints and proposed a probabilistic interpretation to address the variability in PCR results at these low parasite densities [45].
In this study, we present a PK/PD analysis of benznidazole based on data from the BENDITA study to support the evaluation of new treatment regimens. Our specific aims are i) to describe the population pharmacokinetics of benznidazole and investigate potential interactions with fosravuconazole, and ii) to explore the relationship between benznidazole exposure and parasitological responses. We use the proportion of qPCR-positive follow-up visits as a non-binary pharmacodynamic (PD) endpoint, distinguishing between single, intermittent, and consistent PCR-positivity to provide a better description of individual treatment responses. By characterizing the PK/PD relationship - while considering dose, adherence, and individual variability - we aimed to enhance our understanding of individual treatment responses and optimal antiparasitic treatment.
Methods
Clinical data
(i) Ethics statement.
The data used for population PK and PK/PD modeling originated from the BENDITA study, which has been published previously [32]. The trial was approved by the following ethics committees: Comité de ética de la Fundación Colectivo de Estudios Aplicados y Desarrollo Social (CEADES) on 30-Nov-2015, Comite de bioetica de la Facultad de Medicina Universidad Mayor de San Simón on 07-Mar-2016, and Comité de ética e investigación SEDES Chuquisaca (001/2016) on 01-Apr-2016. Further approvals were obtained from the Comité Etico de Investigación Clínica del Hospital Clínic de Barcelona (HCB/2016/0678) on 21-Sep-2016, and AGEMED (former UNIMED) – Competent Authority Bolivia (MS/UNIMED/CFN/042/2016) on 14-Apr-2016.
The trial was conducted in accordance with local regulations, the Declaration of Helsinki and Good Clinical Practice guidelines. All participants provided written informed consent prior to enrollment. The study was registered at ClinicalTrials.gov (NCT03378661, https://clinicaltrials.gov/ct2/show/NCT03378661), and the protocol is available on the DNDi website: https://dndi.org/research-development/portfolio/new-benz-regimens/.
(ii) Study design and study subjects.
BENDITA was a double-blind, double dummy, placebo-controlled, randomized, multicenter, dose evaluation, proof-of-concept trial conducted between 2016–2018 in three specialized Chagas disease outpatient units in Bolivia (Cochabamba, Tarija, Sucre). A total of 210 adults (18–50 years, 50–80 kg body weight) with chronic indeterminate CD, confirmed by serological testing and positive qualitative PCR results, were enrolled in the study. Subjects with chronic health conditions and signs or symptoms of the chronic cardiac or digestive form of CD were excluded.
(iii) Treatment regimens.
Participants were randomly assigned to one of seven treatment groups, including six active treatment arms and one placebo arm, with 30 subjects per group. The active treatment arms (180 subjects) received benznidazole as follows: benznidazole 150 mg twice daily for 8 weeks (standard care), 4 weeks, or 2 weeks; benznidazole 150 mg once daily for 4 weeks, either as monotherapy or in combination with fosravuconazole; and benznidazole 300 mg (divided in two doses) once weekly for 8 weeks in combination with fosravuconazole. Benznidazole or benznidazole-matched placebo tablets (Abarax, Laboratorios ELEA, Buenos Aires, Argentina) were administered orally in two separate doses each day (morning and evening). Fosravuconazole (Eisai Co, Ltd., Tokyo, Japan) was administered orally at a loading dose of 300 mg once daily for 3 days, followed by 300 mg once per week for 8 weeks.
(iv) Follow-up visits and PCR.
Blood sampling follow-up visits took place daily on days 1–3, then at weeks 2, 3, 4, 6, 10 and 12, and subsequently at 4, 6, and 12 months. At each visit, three 5 mL venous blood samples were collected and assayed in triplicate for T. cruzi satellite DNA, using a previously validated quantitative real-time PCR (qPCR) method [46]. Parasitological response was determined qualitatively based on qPCR results, with the overall outcome at each visit classified as ‘PCR positive’ if at least one sample tested positive. Cycle threshold (Ct) values, a measure of parasite density, represent the mean of nine measurements from three blood samples assayed in triplicate at each study visit. Higher Ct values indicate lower parasite densities (maximum reported value: 40).
(v) Quantification of benznidazole.
Sparse whole blood samples for pharmacokinetic analyses were obtained pre-dose and at blood sampling follow-up visits up to week 10 (n = 9 protocol time points for each patient). Precise records of drug administration timing and blood draws were utilized for population pharmacokinetic modeling.
Concentrations of benznidazole (and ravuconazole) in dry blood spots (DBS) were quantified using a validated liquid chromatography tandem mass spectrometry method (HPLC-MS/MS), conducted in accordance with Good Laboratory Practice, as described previously [47]. Briefly, intra- and inter-day precision and bias were reported to be less than 15% (n = 9) and 10% (n = 27), respectively. The lower limits of quantification LLOQ) for benznidazole and ravuconazole were 50 ng/mL and 20 ng/mL, respectively.
Assessment of consistency in pharmacokinetic data
Initial exploration of individual benznidazole DBS concentration-time profiles revealed some inconsistencies. Some profiles had benznidazole levels which either exceeded or fell below plausible levels. For instance, benznidazole levels above the LLOQ were observed beyond 120 hours after dosing (‘TAD [time after dose] outlier’), which is not compatible with benznidazole’s reported half-life (~ 12–13 hours) [48]. As a result, a systematic and comprehensive outlier analysis was conducted (detailed in S1 Text). Additional PK outliers were identified through both the simeval and qa tools in PsN 5.2.6, supplemented by manual inspection of PK profiles. Other data sources (e.g., ADaM files, bioanalytical data) were reviewed to ensure data integrity and identify any further anomalies in PK data or uncertainties in drug exposure.
Subsequently, potential causes of these anomalies - including bioanalytical assay issues, non-adherence, physiological variability, and potential treatment misallocations - were considered to inform objective decisions regarding the inclusion or exclusion of subjects from subsequent analyses. Subjects were flagged for potential misallocation if their observed concentrations deviated markedly from the expected range for their assigned treatment, but aligned more closely with an alternative regimen. These discrepancies were identified using stochastic simulations in NONMEM, based on a preliminary population PK model. Benznidazole concentration-time profiles were simulated for 1,000 hypothetical adults per regimen, and 99% confidence intervals (CIs) were generated for each PK sampling time point. Subjects were flagged as potential treatment misallocations if they showed fewer discrepancies with an alternative treatment arm than with their assigned one. Documented non-adherence and treatment interruptions were considered. These participants, where treatment misallocation could not be ruled out, were excluded from all analyses.
Population pharmacokinetic analysis
(i) Model development.
All subjects allocated to active treatment arms were included in the final PK analysis dataset, except those flagged for potential treatment misallocation. Sparse DBS concentrations of benznidazole were pooled across all six active treatment arms. DBS concentrations were transformed into their natural logarithms, and concentration-time profiles modeled using nonlinear mixed-effects modeling in NONMEM, version 7.4.3 (Icon Development Solution, Ellicott City, MD, USA). The first-order conditional estimation method with interactions (FOCE-I) was employed. The model building process and model diagnostics were facilitated by the use of Pirana (v2.9.9), Pearl-speaks-NONMEM (PsN v5.2), R (v4.2.2) and TIBCO Spotfire (v 11.3.0) [49,50]. Discrimination between two competing, nested models was based on objective function values (OFV), with the difference in OFV being equivalent to a likelihood ratio test. A significant improvement of the structural model was indicated by a decrease of OFV by at least 3.84 (p < 0.05, one degree of freedom difference). Pharmacokinetic data were censored at 120 hours TAD (approximately 10 benznidazole half-lives), and concentrations below the LLOQ for the censored dataset were omitted.
One- and two compartment disposition models were evaluated, along with different absorption models (first-order absorption and transit compartment models). Relative bioavailability (F) was fixed to unity for the population, as all data were from oral administration and no intravenous data were available to estimate absolute F. Inter-individual variability (IIV) on F was incorporated into the base model, and covariate effects - such as sex - were modeled as relative differences in bioavailability. IIV was implemented using an exponential model, with IIV estimates below 10% fixed to zero in the final model. Residual unexplained variability was implemented as an additive error on the log-transformed observed concentrations (equivalent to an exponential residual error on an arithmetic scale).
The best performing structural model was carried forward to subsequent covariate model building. Body weight (standardized to a body weight of 65 kg) was implemented a priori as an allometric function on all clearance (exponent 0.75) and volume (exponent 1) parameters. Various additional demographic, pharmaceutical and parasitological factors were evaluated based on statistical significance and biological plausibility. Individual creatinine clearance was estimated from blood creatinine levels using the Cockcroft and Gault equation. Preselected covariates (age, sex, accumulated benznidazole dose and dose per day, duration of drug treatment, monotherapy/combination with fosravuconazole, markers of liver/kidney function, hematocrit, and serology) were further evaluated by a step-wise covariate modeling approach, using the scm functionality in PsN. For continuous covariates, linear, exponential, and power parameter-covariate relationships were explored. In the forward step, covariates were included at a statistical significance level of p = 0.05 (ΔOFV = −3.84), followed by a more stringent backward elimination step (p = 0.001, ΔOFV= −10.83).
Basic goodness-of-fit (GOF) diagnostics were used to identify potential model misspecifications and systematic errors. The predictive performance of the final model was assessed by prediction-corrected visual predictive checks (VPCs). Bootstrapping (n = 1,000 resampled bootstrap datasets) was performed to evaluate model robustness and obtain parameter precision estimates.
(ii) Definition of exposure metrics.
The following individual PK exposure metrics were extracted: peak concentrations (CMAX) and areas under the concentration-time curves to infinity (AUC∞) for benznidazole in blood. The time above the putative minimum target concentration (T > IC90) was estimated, with for benznidazole in blood set at 29.3 µM (7.61 mg/L). This target concentration is based on the in vitro
of benznidazole against the amastigote form of T. cruzi (Tulahuen strain) in 3T3 host cells, corrected for protein binding and scaled to human blood (DBS) samples. Further details are provided in S2 Text. A sensitivity analysis was also performed using previously suggested therapeutic ranges as thresholds (3–6 mg/L in plasma or 2.5-5 mg/L in DBS), instead of the IC90.
In addition to these exposure metrics, the duration of benznidazole treatment was described in several ways: total days of benznidazole treatment (actual number of days benznidazole was taken); weeks of benznidazole treatment (a week is counted if at least one benznidazole dose per week was taken); and total duration of benznidazole treatment, irrespective of intermittent drug interruptions. For example, if an individual randomised to 2 weeks benznidazole took active treatment only on days 1 and 12, this would be recorded as 2 total days of treatment, 2 weeks of treatment, and 12 days total duration of treatment.
Assessment of treatment failure
In this retrospective analysis, any positive T. cruzi qPCR result at the end of treatment (EOT) or during any follow-up post-EOT was considered indicative of treatment failure. EOT was defined according to the assigned treatment duration, allowing for a two-week grace period after the last active dose. Treatment failure was illustrated in both the modified intention-to-treat (mITT) and per-protocol (PP) populations. The mITT population included all participants with post-EOT follow-up data, excluding those with potential treatment misallocation. The PP population further excluded individuals with <80% benznidazole compliance. Compliance was assessed at the level of individual dosing occasions, accounting for partial dosing (e.g., one dose taken instead of two on a given day).
The proportion of patients with at least one qPCR-positive sample post-treatment was summarized per arm, with Clopper-Pearson 95% CIs. Fisher’s exact test (Holm-adjusted p-values) was used to compare active arms to placebo. As the study was not powered to differentiate between active treatment arms, no formal statistical comparisons were performed between them. Unlike the original BENDITA study, which standardized the number of follow-up visits across treatment arms by excluding specific visits (e.g., two in the 2-week arm), our analysis retained all available post-treatment qPCR data to support PK/PD modeling, specifically for the subsequent calculation of qPCR positivity (see below).
Exposure - parasitological response analysis
To assess the exposure–response relationship, a non-binary PD endpoint (qPCR positivity) was used, defined as the proportion of T. cruzi qPCR-positive blood samples post-EOT. qPCR positivity was calculated for each subject as (Equation 1):
The analysis was conducted in the modified intention-to-treat (mITT) population. Participants without post-EOT follow-up or with potential treatment misallocation were excluded but not considered treatment failures, as their post-treatment status could not be assessed. Excluding these individuals ensured the analysis focused on participants with reliable exposure and outcome data. Sensitivity analyses excluding additional PK outliers and individuals with uncertain exposure were conducted to assess the robustness of findings.
The relationship between benznidazole exposure in blood and qPCR positivity in the mITT population was explored using beta binomial regression in R (v4.2.2).
The model was specified as follows:
where p represents the probability that a qPCR sample is positive for the presence of parasites at a single follow-up visit, and logit (p) represents the log odds of qPCR positivity. In Equation 2, β0 is the intercept, while β1 and β2 represent the change in the log-odds ratio for qPCR positivity per unit increase in the predictor variable x and the Ct value at the screening visit (pre-dose), respectively. Various exposure variables (x) were investigated as predictors of the PCR positivity, including individual CMAX, AUC∞, T > IC90, and benznidazole treatment durations.
The probability of at least one follow-up sample being qPCR positive, denoted as p*, was estimated from the probability of qPCR positivity for a single follow-up visit (p), as follows:
where N is the number of follow-up visits. Predictions were made for a range of follow-up visits, and utilizing the median Ct value at baseline.
Model implementation was conducted using the ‘glmmTMB’ package [51], with the beta-binomial distribution parameterized according to Morris 1997 [52]. After excluding placebo data and one subject who only took 4 doses over 8 days, a subgroup analysis was performed using binomial regression modeling in the GLM family in R. This subgroup analysis did not support a beta binomial model and assumed a dispersion factor of 1. Model diagnostics, including tests for overdispersion and zero inflation, as well as simulated residuals, were performed using the ‘DHARMA’ package in R [53]. Model comparisons were based on the Akaike information criterion (AIC).
Results
Characterization of the study population
A total of 210 adults with chronic indeterminate CD were enrolled in the BENDITA study, with 30 patients randomly assigned to each treatment group (Fig 1). Prior to analysis, a systematic assessment of PK data consistency was conducted to evaluate its reliability. Details of inconsistencies by diagnostic criteria and likely causal factors are provided in S1 Text (S1.2 and S1.3). For most subjects, the specific reasons for these discrepancies remain unclear. However, five subjects exhibited significant discrepancies between observed and predicted benznidazole concentrations, warranting their exclusion from all analyses. Treatment allocation errors were one possible explanation, but other factors - such as incorrect sample labelling, non-adherence, or assay errors - cannot be ruled out.
Dark blue boxes indicate the primary analysis populations for the final PK and PK/PD analyses, following the exclusion of subjects with significant discrepancies between observed and predicted benznidazole concentrations, which may be explained by treatment misallocations, though the exact cause remains uncertain (n = 5). For the PK/PD sensitivity analysis, all subjects not meeting the PK/PD reliability criteria (PK outliers or uncertain exposure) were excluded from the mITT population, resulting in restricted population of 186 subjects. Abbreviations: EOT, end of treatment; BZN, benznidazole; E1224, fosravuconazole; wks, weeks; QD, once daily; BID, twice daily; mITT, modified intention-to-treat; PP, per-protocol.
The final PK analysis included 175 subjects allocated to active treatment. Four patients had no follow-up post-EOT and were excluded, leaving 201 patients in the mITT population (including placebo). The PP population comprised 183 patients after excluding those with <80% benznidazole compliance. For the PK/PD analysis, a sensitivity analysis was performed in a subset of 186 patients from the mITT population, excluding a total of 22 subjects with uncertain drug exposure who did not meet the PK/PD reliability criteria (e.g., PK outliers not attributable to an assay error). Further details on the exclusion criteria are provided in S1 Text (S1.4). A detailed flow chart of patient exclusions by treatment arm is provided in S1 Fig.
Baseline characteristics and adverse events for the entire BENDITA study population have been reported previously [32]. Here, demographic information for subjects included in the PK and PK/PD analyses is presented (Table 1). All patients were Bolivian and aged between 18 and 50 years, with a median body weight of 65 kg (range 50–80 kg). The majority of patients were female (70% female, 30% male, mITT). Demographic and clinical characteristics were similar across treatment arms.
The proportion of adherent subjects (those who took at least 80% of their allocated benznidazole dose) exceeded 90% in most study arms, except for the 8-week and 2-week treatment arms, where adherence was approximately 80%. As reported previously, treatment discontinuations due to adverse events were most common in the 8-week group (20%), followed by the combination therapy arms. Treatment interruptions due to adverse events were most frequent in the 2-week group (34%), but there were no treatment discontinuations in this group.
Pharmacokinetics
The 175 patients included in the pooled population PK analysis contributed 986 DBS benznidazole concentrations, of which 969 were above the LLOQ (1.7% BQL).
Benznidazole concentration-time profiles were best described by a transit-compartment absorption model, followed by a one-compartment disposition model, with linear elimination from the central compartment (Fig 2). A two-compartment disposition model did not significantly improve the model fit. A transit-compartment absorption model with one transit compartment was superior to all other absorption models.
Absorption from the gut compartment is described by a transit absorption model (with one transit compartment), followed by a one-compartment disposition model. F is the relative oral bioavailability, VC is the apparent volume of distribution of the central compartment, and CL is the apparent elimination clearance. kTR is the rate constant between absorption compartments and was calculated from the mean absorption transit time (MTT) as kTR = (1 + n)/MTT where n is the number of transit compartments.
Body weight was included as an allometric function on CL/F (exponent = 0.75) and Vc/F (exponent = 1.0), resulting in a substantial model improvement (∆OFV = -34.7). A potential time-dependency of CL/F was investigated using linear, exponential, and EMAX functions as described previously [55]. However, none of the models with time-varying clearance yielded a significant improvement in model fit. Co-administration of fosravuconazole resulted in an increase in benznidazole clearance by 18% (∆OFV = -15.7). Moreover, there was a trend towards lower relative bioavailability in men compared to women, with an estimated 13% lower bioavailability (F) in men (∆OFV = -15.5). None of the other covariates tested were retained after the backward elimination covariate step.
Parameter estimates from the final model, along with their relative standard errors and confidence intervals, are provided in Table 2. The variability in mean transit time, MTT, was high (62% coefficient of variation, CV). However, bootstrapping analysis indicated a robust PK model, demonstrating moderate to high precision in estimating PK parameters (RSE < 32% for all parameters). GOF plots and the prediction-corrected VPC are shown in Fig 3. The final model accurately described the observed concentration-time profiles, with no major model misspecifications. A sensitivity analysis, re-including the five subjects with severe PK outliers, showed minimal changes in PK parameter estimates, confirming the robustness of the PK model despite potential misallocations (S1 Text, S1.5).
a) Goodness-of-fit. Observations are represented by grey circles. Solid black lines represent the line of identity (upper panel) or zero line (lower panel), and the dashed lines represent the local polynomial regression fitting for all observations. (b) Prediction-corrected visual predictive check. Open circles represent observed plasma benznidazole concentrations. The solid red line represents the 50th percentile (median) and the blue dashed lines represent the 5th and 95th percentiles of the observed data. The shaded area represents the 95% CI around the simulated 5th, 50th, and 95th percentiles.
Benznidazole showed rapid absorption with a median time to maximum concentration (TMAX) ranging from 1.7 to 2.1 hours. The combination therapy arms demonstrated a slightly shorter elimination half-life (~14 hours) compared to the monotherapy arms (~17 hours), attributable to the effect of fosravuconazole co-administration on benznidazole clearance. AUC∞ and CMAX values were lower in men compared to women, reflecting a lower relative bioavailability, but this effect was minor (<20%). Secondary PK parameter estimates for benznidazole across the different active treatment arms, stratified by sex, are provided (S1 Table).
The distribution of selected exposure variables, not stratified by sex, is illustrated in Fig 4. Total exposure varied widely across treatment arms, with the standard treatment arm (300 mg benznidazole daily for 8 weeks) showing a 7.9-fold higher median AUC∞ compared to the weekly arm. AUC∞ exhibited approximate dose-linearity, with an approximate 2-fold increase when comparing 2 weeks to 4 weeks, and 4 weeks to 8 weeks of treatment in the monotherapy arms (300 mg daily). CMAX was about twofold higher in the twice-daily treatment arms (11.7 to 11.9 mg/L) compared to the once-daily or weekly treatment arms (6.2 to 6.9 mg/L).
Boxplots display the distribution of benznidazole exposures in terms of a) AUC∞, b) CMAX, and c) weeks of treatment (a week is counted if at least one dose of benznidazole was taken). Median (midline), interquartile range (IQR, box), and 1.5*IQR whiskers are shown. Abbreviations: BZN, benznidazole; QD, once daily; BID, twice daily; E1224, fosravuconazole.
Apart from AUC∞ and CMAX, the duration of treatment was assessed using different definitions (S2a Fig). Duration in weeks (counting any week with at least one dose) was highly correlated with total treatment duration (i.e., duration, irrespective of interruptions). However, in the weekly dosing arm, there was a clear difference between these parameters and the total number of days benznidazole was actually taken (see S2 Table for correlation matrix of exposure metrics). For consistency, treatment duration in terms of weeks is used throughout (Fig 4c).
Additionally, time above the target concentration was derived, showing high sensitivity to the choice of the threshold (S2b Fig). Using the lower bound of the previously suggested therapeutic range (3 mg/L in plasma, i.e., 2.5 mg/L in DBS), extended exposure above target concentrations was achieved in all treatment arms. In contrast, using the in vitro IC90,DBS (derived from activity against amastigotes), only twice-daily dosing regimens maintained levels above the target concentration. This is further illustrated in Fig 5, showing simulated median plasma concentration-time profiles for the different treatment arms. Simulated PK profiles, accounting for IIV and stratified by sex, are provided in S3 Fig.
Simulations were performed for 2000 patients (1000 women and 1000 men, body weight: 65 kg) using the final population PK model. The horizontal red line denotes the in vitro IC90, adjusted for protein binding and scaled to DBS concentrations. The grey shaded area marks the historically accepted therapeutic range: 3-6 mg/L in plasma, scaled to 2.5-5 mg/L in DBS samples.
Proportion of patients with treatment failure
Fig 6 presents the proportions of patients with at least one qPCR-positive result post-EOT (i.e., ‘treatment failure’) in both the mITT (n = 201) and PP (n = 183) populations. In both populations, treatment failure was significantly more frequent in the placebo arm (97%, 95% CI: 83–100, n = 29) compared to all active treatment arms (p < 0.0001). In the PP population, proportions were similar (15–19%) across most active treatment arms, except for the 2-week benznidazole regimen, where 30% (95% CI: 13–53, n = 20) had at least one positive PCR result post-treatment. Although the 2-week regimen showed a numerically higher failure rate, the 95% CIs across active treatment arms overlapped. No formal statistical comparisons were made between active regimens due to small sample sizes. Note that these proportions are based on all available follow-up samples, unlike the original BENDITA study, which excluded selected follow-ups to standardize visit numbers.
Color of barplots by benznidazole monotherapy (light blue), co-administration of E1224 (dark blue), or placebo (grey). Numbers above bars indicate patients with treatment failure/ total patients, along with the proportion of patients with treatment failure (%) and 95% Clopper Pearson CIs. No formal statistical comparisons were made between active treatment arms due to small sample sizes and differences in the number of follow-up visits.
Importantly, the binary endpoint ‘treatment failure’ is influenced by the number of follow-up visits - which varied by treatment arm (S4 Fig) – and was therefore used descriptively only, not for PK/PD modeling. Instead, it served to illustrate the proportion of patients with any detectable T. cruzi DNA post-treatment in a compliant population (PP) and in the PK/PD modeling population with available follow-up (mITT).
qPCR positivity: PD endpoint for PK/PD modeling
For PK/PD modeling, the proportion of qPCR-positive samples post-treatment (‘qPCR positivity’) was used as a non-binary PD endpoint in the mITT population (n = 201, Fig 7). This endpoint captured the frequency of detectable parasitemia per patient while avoiding bias due to differing numbers of follow-up visits across treatment arms. Unlike the PP analysis, participants with <80% benznidazole compliance were retained, as partially adherent individuals still provide valuable exposure–response information. Trends in qPCR positivity were consistent with the binary endpoint.
Each dot represents an individual’s proportion of post-treatment qPCR-positive blood samples (‘qPCR positivity’). Symbol color indicates whether patients met all PK/PD reliability criteria (black) or had problematic PK data or uncertain drug exposure (orange). Closed symbols indicate ≥80% benznidazole compliance; open symbols indicate <80% compliance.
In the mITT population, 29 of 30 patients in the placebo group had at least one positive qPCR result post-treatment. Among these, qPCR positivity was consistently high (≥ 40%) in all but one patient - i.e., at least 2 of 5 follow-up samples were positive. In contrast, patients in the 4- or 8-week treatment arms were either persistently negative or had only one single PCR-positive sample, corresponding to qPCR positivity ≤20%. An exception was observed in the 4-week once-daily treatment arm, where a patient discontinued treatment after 4 doses (over 8 days) and remained qPCR-positive in all follow-up visits.
In the 2-week arm, 10 of 29 patients had at least one positive sample post-treatment. Among these, three had multiple positive qPCRs, with qPCR positivity proportions of 29% and 43% (i.e., 2 or 3 positive qPCRs out of 7 follow-up visits). These values were higher than those observed in patients receiving longer benznidazole regimens. Two of these patients were likely non-adherent, or there may have been a labeling error or placebo mix-up, and they did not meet PK/PD reliability criteria. Caution is warranted in interpreting these results. However, notably, one fully adherent patient in the 2-week treatment arm showed a qPCR positivity of 29% (2/7 positive qPCRs).
Exposure - parasitological response model (PK/PD analysis)
Fig 8 presents observed qPCR positivity proportions, along with model-predicted probabilities of a positive qPCR result at a single follow-up visit for selected benznidazole exposure variables. Detailed diagnostics for this univariate (beta) binomial regression analysis are provided in Table 3.
Model fits are shown for the modified ITT population, including placebo (n = 201, grey lines and grey shaded area) and excluding placebo as well as one influential outlier (n = 170, red lines and red shaded area). The solid grey and red lines indicate the median, and the shaded areas represent the 95% CI around predicted probabilities. Points represent observed proportions of positive qPCR results after end of treatment (qPCR positivity) in the placebo arm (black) and active treatment arms. a) AUC∞, b) CMAX, c) weeks of treatment (a week is counted if at least one dose of benznidazole was taken). The insert in a) shows regression based on log-transformed AUC∞ (an increment of 1 was added to all AUC∞ values to allow for log-transformation of placebo data). The influential outlier, who took only four doses of 150 mg benznidazole over 8 days, was allocated to the 150 mg benznidazole once daily for 4 weeks treatment arm and is represented by the pink point with 100% qPCR positivity.
In the full mITT population (n = 201, including placebo), a significant relationship was observed between benznidazole exposure and the odds of qPCR positivity (p < 0.001) for all investigated PK/PD index parameters. The strength of this association was highly sensitive to the transformation of the exposure variable, as shown with the example of AUC∞. When using time above the target concentration as the predictor, parameter estimates and goodness-of-fit were also highly sensitive to the choice of the target concentration (S5 Fig), which remains unknown in vivo.
The exposure-response relationship was primarily driven by the placebo group (i.e., there was a clear difference in parasitological outcomes between no drug and all active regimens). After excluding placebo and the one patient who took only four doses of benznidazole, the exposure-response relationship was no longer significant (p > 0.05). While longer durations of treatment (whether expressed in weeks or total duration) suggested a marginal, non-significant trend towards lower probabilities of qPCR positivity, no clear PK/PD driver of qPCR positivity was identified in the remaining benznidazole treatment groups (n = 170 subjects). Goodness-of-fit diagnostics showed similar AIC values across different exposure parameters, with no improvement in model fit compared to a baseline model that included only Ct values. To aid interpretation of model estimates, probabilities of having at least one qPCR-positive follow-up for patients with multiple visits were calculated (S6 Fig). This transformation accounts for increased chances of detecting T. cruzi parasites at least once as the number of follow-up visits increases.
A trend towards a higher probability of post-treatment qPCR positivity was observed with lower Ct values (indicating higher baseline parasite loads). This trend remained significant or borderline significant even after excluding the placebo group and the influential subject, with an odds ratio consistently between 0.85 and 0.88 across models per unit increase in Ct (Table 3). Adjustment for combination therapy with fosravuconazole in this subgroup analysis did not show a significant effect on qPCR positivity.
Similar results were obtained in the refined mITT population (n = 186), excluding all subjects not meeting the PK/PD reliability criteria (S3 Table), though with reduced statistical power.
Discussion
Current efforts to improve the anti-parasitic treatment of chronic T. cruzi infection have focused on reducing benznidazole exposure to enhance tolerability while maintaining efficacy [29,30,32,35,37–39,57–60]. In our study, we characterized the population pharmacokinetic properties of benznidazole, and we explored the relationship between benznidazole exposure and parasitological recrudescence using data from BENDITA. To our knowledge, this represents the largest population PK and PK/PD analysis of benznidazole in CD patients to date.
Pharmacokinetic properties of benznidazole
Benznidazole concentration-time profiles were well described by a transit-compartment absorption model, followed by a one-compartment disposition model with linear elimination. The transit-compartment model provided a better fit and a more physiological representation of benznidazole’s absorption dynamics compared to the earlier first-order absorption models [26,27,29,57]. We observed substantial IIV in mean transit time (~62%), consistent with previous reports of high variability in absorption rates [48]. This variability is likely related to factors such as gastrointestinal pH, gastric emptying times, and food intake, which are potentially relevant because of benznidazole’s poor water solubility. The median TMAX in our study (approximately 2 hours) was slightly shorter than the 3–4 hours reported in earlier studies [27,29,48], possibly due to differences in tablet formulations (Abarax vs. Radanil) and food states.
The one-compartment disposition model found in this study aligns with previous population pharmacokinetic analyses in both adult [29] and pediatric [30,31] CD patients. Our parameter estimates—V/F of 31.6 L, CL/F of 1.3 L/h—are in good agreement with earlier studies [48,57] and suggest moderate distribution into body tissues. Wiens et al. conducted a meta-analysis of the single-dose pharmacokinetics of benznidazole in adults and reported similar values (V/F ~ 39 L, CL ~ 2 L/h). Our elimination half-life estimates (~ 17h for monotherapy and ~ 14h for combination therapy) are somewhat longer than the than the 13.3 hours reported by Wiens et al. [48] and the 12–15 hours observed by Raaflaub et al. [26]. Soy et al. reported an even longer half-life (36 hours), potentially explained by different characteristics in the study population and food effects.
Benznidazole is extensively metabolized, most likely in the liver and possibly involving cytochrome P450 enzymes, though the specific metabolic pathway remains largely unknown [43,61,62]. Coadministration of fosravuconazole significantly increased benznidazole clearance, resulting in decreased benznidazole exposure. However, the effect was small (< 20%) and likely not clinically relevant. This is consistent with previous findings from a drug-drug interaction study in healthy volunteers [61].
Men exhibited lower relative bioavailability of benznidazole compared to women. In contrast, Molina et al. reported a higher volume of distribution (V/F) in men compared to women [57], and. Soy et al. did not find significant covariate effects [29]. The sample sizes in these studies were much smaller. In our study, both covariate effects—fosravuconazole coadministration on elimination clearance and the difference in relative bioavailability between sexes - were likely not relevant clinically. Nonetheless, these effects were included in the final model to describe the pharmacokinetic properties of benznidazole fully.
PK/PD relationship for benznidazole
Evaluating antiparasitic drug efficacy in CD is remarkably challenging as there are no validated early biomarkers for clinical efficacy, and confirmation of parasitological cure is technically difficult as blood trypomastigote densities are often near or below the limit of detection [43]. In our study, we used the proportion of positive PCR results (qPCR positivity) as an endpoint for the PK/PD analysis. By modeling the outcome as a binomial random variable, we maximized information from multiple PCR samples and avoided bias against shorter treatment arms, which have more follow-up visits post-EOT and, therefore, higher chances of positive PCR results.
In the placebo arm, nearly all patients (29/30) did not clear their parasites, while the majority of patients in all benznidazole treatment arms had undetectable parasitaemia over one year follow-up. The clear difference in qPCR positivity between those receiving no treatment and those receiving any benznidazole highlights its potent anti-T. cruzi activity.
There was a non-significant trend suggesting higher post-treatment PCR positivity with the 2-week regimen compared to longer treatment courses, which has been examined in a recent probabilistic meta-analysis of these data [45]. However, the BENDITA trial was not powered to make definitive comparisons between active treatment regimens [32], and the 95% CIs for “treatment failure” proportions overlapped. Moreover, our analysis included all available post-treatment qPCR data, whereas the original BENDITA study excluded selected follow-ups to standardize visit numbers, limiting direct comparability. After excluding the placebo group and one patient who discontinued after only four doses of benznidazole (taken over 8 days), no clear pharmacokinetic predictor of treatment response was identified. It is an important finding that the once-weekly regimen, which provided the lowest total dose (AUC∞ 8-fold lower compared to the standard regimen), showed no apparent difference in efficacy compared to the 4-week or 8-week continuous treatment regimens. This suggests that time, in addition to benznidazole exposure, may be an important determinant of the therapeutic response. Together, these findings suggest that the standard 8-week regimen may be excessive.
Similar to the BENDITA trial, the MULTIBENZ trial explored shorter benznidazole treatment durations in the treatment of CD and found that a 15-day regimen of a higher dose (400mg instead of 300mg/day) was not inferior to the standard 60-day regimen, with fewer discontinuations resulting from adverse effects [35]. As in this study, the MULTIBENZ study did not observe a correlation between benznidazole concentrations and treatment outcomes. Together, these trials provide support for shorter or lower-dose regimens. However, the higher rate of qPCR positivity among patients with poor adherence emphasizes the importance of adhering to prescribed regimens, particularly for very short treatment durations.
In addition to exploring shorter treatment durations, lower doses have also been proposed. Soy et al. demonstrated through modeling and simulation that a reduced dose of 2.5 mg/kg/day could maintain plasma concentration within the historically accepted therapeutic range (3–6 mg/L) [29], supporting the idea that lower doses might be sufficient [30,63]. We confirmed that target concentrations (3–6 mg/L) could be achieved with lower doses, although results were sensitive to the threshold used. The historically accepted target range was extrapolated from tissue culture studies with limited methodological details [28], and while cited widely, the in vivo target concentration remains unknown. Furthermore, variability in drug susceptibility across different T. cruzi strains and regional differences in efficacy, as seen in MULTIBENZ, compromise generalisations over optimum treatment [64,65].
It has also been suggested that maintaining benznidazole levels above a target concentration is not essential, and that CMAX is the primary driver of efficacy [40,41,66]. The existence of spontaneous dormancy [67] or transient non-replicative states [68–70] remains a matter of debate and has been hypothesized to contribute to treatment failure due to their reduced susceptibility to drug-induced toxicity. Intermittent dosing of sufficiently high doses over an extended period could potentially address this challenge, while reducing total drug exposure and potentially lowering toxicity.
Given the complexity of CD, and the possibility of a transient non-replicative state, it is also plausible that more than one predictor of benznidazole efficacy - such as a combination of treatment duration and CMAX or AUC∞ - is needed to adequately capture the relationship between exposure and treatment efficacy. Several pre-clinical studies have emphasized the critical role of both benznidazole dose and treatment duration in achieving parasitological cure in T. cruzi infected mice [41,71–74]. However, high collinearity between PK/PD index parameters hampers the identification of the primary driver of efficacy or combination of drivers. Improved study designs, such as dose fractionation studies, have been suggested to help disentangle these factors and better understand the determinants of benznidazole efficacy [75].
The BENDITA trial assessed intermittent dosing of benznidazole with fosravuconazole co-administration, finding similar efficacy to the standard regimen. However, the design of the trial did not allow discrimination of the effects of fosravuconazole from those of intermittent treatment.
Strengths and limitations
A strength of our study was the application of nonlinear mixed-effects modeling to characterize the pharmacokinetic properties of benznidazole, which enabled a pooled analysis across treatment arms, effective handling of sparse PK data, and facilitated the identification of covariate effects. The modeling also enabled the identification of patients with outlying PK samples, allowing for the recognition of potential issues such as non-adherence or potential discrepancies in treatment allocation. Despite some inconsistencies in PK data, sensitivity analyses confirmed the robustness of our findings.
The study has several limitations. The BENDITA trial was not powered to differentiate between active treatment arms or to identify the primary driver of benznidazole efficacy, nor could it separate the effects of the azole drug from the different treatment regimens. PCR, while widely used in clinical trials, including BENDITA, has limitations in estimating absolute cure rates due to the low density of T. cruzi in blood near the detection limit and challenges in interpreting qPCR results from the few samples collected. New biomarkers of clinical efficacy and disease progression are urgently needed. Additionally, the study’s one-year follow-up may not be sufficient to assess long-term parasitological outcomes. A placebo arm was included to enable a robust comparison of the antiparasitic activity of benznidazole and new treatment regimens. The inclusion of a placebo arm may be debated in light of earlier observational studies suggesting long-term benefits of benznidazole in CD patients without cardiac involvement [10,11]. However, due to the chronic nature of CD and its slow progression, there is no evidence of significant short-term risk within the trial evaluation period. Moreover, all participants were offered rescue treatment after study completion. Since the study was conducted in Bolivia, the findings may not be generalizable to other regions with different T. cruzi strains [35]. Furthermore, the in vitro IC90 used as a reference in this study was measured against the amastigote form of T. cruzi (Tulahuen strain), whereas qPCR detects circulating trypomastigotes in blood. The PK/PD index parameter T > IC90 is highly sensitive to the choice of the target concentration, as demonstrated by comparisons with the often-cited range (3–6 mg/L). The in vivo target concentration, however, remains unknown. Lastly, exposure correlates of toxicity and adverse events have not yet been analyzed separately; this will be addressed in future work.
Conclusions
This study successfully described benznidazole’s population pharmacokinetics in adult patients with chronic CD, showing that fosravuconazole does not have a clinically relevant impact on benznidazole’s pharmacokinetics. Our analysis suggests that the standard 8-week dosing regimen may be excessive, as both weekly dosing for 8 weeks or daily dosing for 4 weeks appeared similarly effective. In contrast, there is an indication that the 2-week regimen may be less effective. No clear pharmacokinetic predictor of treatment failure was identified among those receiving benznidazole treatment. Future well-designed trials, ideally employing factorial randomization and complementary endpoints to qPCR, are needed to robustly evaluate the optimal balance of dose, duration, and frequency and to identify the most well-tolerated regimen that achieves maximal therapeutic efficacy.
Supporting information
S2 Text. Determination of in vitro antitrypanosomal activity.
https://doi.org/10.1371/journal.pntd.0013522.s002
(DOCX)
S2 Fig. Distribution of benznidazole exposure variables across treatment arms.
https://doi.org/10.1371/journal.pntd.0013522.s004
(DOCX)
S3 Fig. Simulated pharmacokinetic profiles, by sex.
https://doi.org/10.1371/journal.pntd.0013522.s005
(DOCX)
S5 Fig. Predicted probabilities of qPCR positivity for a single follow-up visit using different predictors.
https://doi.org/10.1371/journal.pntd.0013522.s007
(DOCX)
S6 Fig. Predicted probabilities of at least one follow-up visit being qPCR positive.
https://doi.org/10.1371/journal.pntd.0013522.s008
(DOCX)
S1 Table. Secondary pharmacokinetic parameter estimates.
https://doi.org/10.1371/journal.pntd.0013522.s009
(DOCX)
S2 Table. Correlation matrix of plasma variables for benznidazole DBS concentrations.
https://doi.org/10.1371/journal.pntd.0013522.s010
(DOCX)
S3 Table. Regression model diagnostics (sensitivity analyses).
https://doi.org/10.1371/journal.pntd.0013522.s011
(DOCX)
Acknowledgments
We would like to thank all trial participants and study site personnel. We also extend our gratitude to the BENDITA study group. The authors thank Danilo César Galindo Bedor for sharing the pharmacokinetic data and providing detailed information on the bioanalysis of benznidazole. I.S. wish to thank colleagues at DNDi for their valuable support and scientific input during this project.
References
- 1.
WHO World Health Organization. Chagas disease (American trypanosomiasis) [cited 15 August 2023]. Available from: https://www.who.int/health-topics/chagas-disease#tab=tab_1
- 2. Guhl F, Ramírez JD. Poverty, Migration, and Chagas Disease. Curr Trop Med Rep. 2021;8(1):52–8.
- 3. Gómez-Ochoa SA, Rojas LZ, Echeverría LE, Muka T, Franco OH. Global, Regional, and National Trends of Chagas Disease from 1990 to 2019: Comprehensive Analysis of the Global Burden of Disease Study. Glob Heart. 2022;17(1):59. pmid:36051318
- 4. Lidani KCF, Andrade FA, Bavia L, Damasceno FS, Beltrame MH, Messias-Reason IJ, et al. Chagas Disease: From Discovery to a Worldwide Health Problem. Front Public Health. 2019;7:166. pmid:31312626
- 5. World Health Organization = Organisation mondiale de la S. Chagas disease in Latin America: an epidemiological update based on 2010 estimates = Maladie de Chagas en Amérique latine: le point épidémiologique basé sur les estimations de 2010. Weekly Epidemiological Record = Relevé épidémiologique hebdomadaire. 2015;90(06):33–44.
- 6. Crespillo-Andújar C, Comeche B, Hamer DH, Arevalo-Rodriguez I, Alvarez-Díaz N, Zamora J, et al. Use of benznidazole to treat chronic Chagas disease: An updated systematic review with a meta-analysis. PLoS Negl Trop Dis. 2022;16(5):e0010386.
- 7. Prata A. Clinical and epidemiological aspects of Chagas disease. Lancet Infect Dis. 2001;1(2):92–100. pmid:11871482
- 8. Pérez-Molina JA, Molina I. Chagas disease. Lancet. 2018;391(10115):82–94. pmid:28673423
- 9. Hall BS, Wilkinson SR. Activation of benznidazole by trypanosomal type I nitroreductases results in glyoxal formation. Antimicrob Agents Chemother. 2012;56(1):115–23. pmid:22037852
- 10. Viotti R, Vigliano C, Lococo B, Bertocchi G, Petti M, Alvarez MG, et al. Long-term cardiac outcomes of treating chronic Chagas disease with benznidazole versus no treatment: a nonrandomized trial. Ann Intern Med. 2006;144(10):724–34. pmid:16702588
- 11. Hasslocher-Moreno AM, Saraiva RM, Sangenis LHC, Xavier SS, de Sousa AS, Costa AR, et al. Benznidazole decreases the risk of chronic Chagas disease progression and cardiovascular events: A long-term follow up study. EClinicalMedicine. 2020;31:100694. pmid:33554085
- 12.
Ministerio de Salud de la Nación BA. Guías para la atención al paciente infectado con Trypanosoma cruzi (enfermedad de Chagas). 2012.
- 13. Cruz CV, Rabinovich A, Moscatelli G, Moroni S, González N, Garcia-Bournissen F, et al. Adverse events associated with benznidazole treatment for Chagas disease in children and adults. Br J Clin Pharmacol. 2024;90(12):3334–47. pmid:39210623
- 14. Sperandio da Silva GM, Mediano MFF, Hasslocher-Moreno AM, de Holanda MT, Silvestre de Sousa A, Sangenis LHC, et al. Benznidazole treatment safety: the Médecins Sans Frontières experience in a large cohort of Bolivian patients with Chagas’ disease. J Antimicrob Chemother. 2017;72(9):2596–601. pmid:28645201
- 15. Yun O, Lima MA, Ellman T, Chambi W, Castillo S, Flevaud L, et al. Feasibility, drug safety, and effectiveness of etiological treatment programs for Chagas disease in Honduras, Guatemala, and Bolivia: 10-year experience of Médecins Sans Frontières. PLoS Negl Trop Dis. 2009;3(7):e488. pmid:19582142
- 16. Kratz JM. Drug discovery for chagas disease: A viewpoint. Acta Trop. 2019;198:105107. pmid:31351074
- 17. Echeverría LE, González CI, Hernandez JCM, Díaz ML, Eduardo Nieto J, López-Romero LA, et al. Efficacy of the Benznidazole+Posaconazole combination therapy in parasitemia reduction: An experimental murine model of acute Chagas. Rev Soc Bras Med Trop. 2020;53:e20190477. pmid:32049205
- 18. Bern C. A New Epoch in Antitrypanosomal Treatment for Chagas Disease. J Am Coll Cardiol. 2017;69(8):948–50. pmid:28231947
- 19. Lepesheva GI, Villalta F, Waterman MR. Targeting Trypanosoma cruzi sterol 14α-demethylase (CYP51). Adv Parasitol. 2011;75:65–87. pmid:21820552
- 20. Diniz L de F, Caldas IS, Guedes PM da M, Crepalde G, de Lana M, Carneiro CM, et al. Effects of ravuconazole treatment on parasite load and immune response in dogs experimentally infected with Trypanosoma cruzi. Antimicrob Agents Chemother. 2010;54(7):2979–86. pmid:20404124
- 21. Morillo CA, Waskin H, Sosa-Estani S, Del Carmen Bangher M, Cuneo C, Milesi R, et al. Benznidazole and Posaconazole in Eliminating Parasites in Asymptomatic T. Cruzi Carriers: The STOP-CHAGAS Trial. J Am College Cardiol. 2017;69(8):939–47.
- 22. Molina I, Gómez i Prat J, Salvador F, Treviño B, Sulleiro E, Serre N, et al. Randomized trial of posaconazole and benznidazole for chronic Chagas’ disease. N Engl J Med. 2014;370(20):1899–908. pmid:24827034
- 23. Torrico F, Gascon J, Ortiz L, Alonso-Vega C, Pinazo M-J, Schijman A, et al. Treatment of adult chronic indeterminate Chagas disease with benznidazole and three E1224 dosing regimens: a proof-of-concept, randomised, placebo-controlled trial. Lancet Infect Dis. 2018;18(4):419–30. pmid:29352704
- 24. Ribeiro I, Diniz F, Mazzeti AL, Mota LWR, Bahia MT. Efficacy of Oral Administration of E1224 in Combination with Benznidazole on Experimental Trypanosoma cruzi Infection. 9th European Congress on Tropical Medicine and International Health (ECTMIH), Basel. 2015.
- 25. Diniz L de F, Urbina JA, de Andrade IM, Mazzeti AL, Martins TAF, Caldas IS, et al. Benznidazole and posaconazole in experimental Chagas disease: positive interaction in concomitant and sequential treatments. PLoS Negl Trop Dis. 2013;7(8):e2367. pmid:23967360
- 26. Raaflaub J. Multiple-dose kinetics of the trypanosomicide benznidazole in man. Arzneimittel-Forschung. 1980;30(12):2192–4.
- 27. Raaflaub J, Ziegler WH. Single-dose pharmacokinetics of the trypanosomicide benznidazole in man. Arzneimittelforschung. 1979;29(10):1611–4. pmid:583230
- 28. Richle RW, Raaflaub J. Difference of effective antitrypanosomal dosages of benznidazole in mice and man. Chemotherapeutic and pharmacokinetic results. Acta Trop. 1980;37(3):257–61. pmid:6106364
- 29. Soy D, Aldasoro E, Guerrero L, Posada E, Serret N, Mejía T, et al. Population pharmacokinetics of benznidazole in adult patients with Chagas disease. Antimicrob Agent Chemo. 2015;59(6):3342–9.
- 30. Altcheh J, Moscatelli G, Mastrantonio G, Moroni S, Giglio N, Marson ME, et al. Population pharmacokinetic study of benznidazole in pediatric Chagas disease suggests efficacy despite lower plasma concentrations than in adults. PLoS Negl Trop Dis. 2014;8(5):e2907. pmid:24853169
- 31. Altcheh J, Moscatelli G, Caruso M, Moroni S, Bisio M, Miranda MR, et al. Population pharmacokinetics of benznidazole in neonates, infants and children using a new pediatric formulation. PLoS Negl Trop Dis. 2023;17(5):e0010850. pmid:37256863
- 32. Torrico F, Gascón J, Barreira F, Blum B, Almeida IC, Alonso-Vega C, et al. New regimens of benznidazole monotherapy and in combination with fosravuconazole for treatment of Chagas disease (BENDITA): a phase 2, double-blind, randomised trial. Lancet Infect Dis. 2021;21(8):1129–40. pmid:33836161
- 33. Chippaux J-P, Salas-Clavijo AN, Postigo JR, Schneider D, Santalla JA, Brutus L. Evaluation of compliance to congenital Chagas disease treatment: results of a randomised trial in Bolivia. Trans R Soc Trop Med Hyg. 2013;107(1):1–7. pmid:23296694
- 34. Streiger ML, del Barco ML, Fabbro DL, Arias ED, Amicone NA. Longitudinal study and specific chemotherapy in children with chronic Chagas’ disease, residing in a low endemicity area of Argentina. Rev Soc Bras Med Trop. 2004;37(5):365–75. pmid:15361952
- 35. Bosch-Nicolau P, Fernández ML, Sulleiro E, Villar JC, Perez-Molina JA, Correa-Oliveira R, et al. Efficacy of three benznidazole dosing strategies for adults living with chronic Chagas disease (MULTIBENZ): an international, randomised, double-blind, phase 2b trial. Lancet Infect Dis. 2024;24(4):386–94.
- 36. Alvarez MG, Vigliano C, Lococo B, Petti M, Bertocchi G, Viotti R. Seronegative conversion after incomplete benznidazole treatment in chronic Chagas disease. Trans R Soc Trop Med Hyg. 2012;106(10):636–8. pmid:22898619
- 37. Molina-Morant D, Fernández ML, Bosch-Nicolau P, Sulleiro E, Bangher M, Salvador F, et al. Efficacy and safety assessment of different dosage of benznidazol for the treatment of Chagas disease in chronic phase in adults (MULTIBENZ study): study protocol for a multicenter randomized Phase II non-inferiority clinical trial. Trials. 2020;21(1):328. pmid:32293523
- 38. Álvarez MG, Hernández Y, Bertocchi G, Fernández M, Lococo B, Ramírez JC, et al. New Scheme of Intermittent Benznidazole Administration in Patients Chronically Infected with Trypanosoma cruzi: a Pilot Short-Term Follow-Up Study with Adult Patients. Antimicrob Agents Chemother. 2015;60(2):833–7. pmid:26596935
- 39. Álvarez MG, Ramírez JC, Bertocchi G, Fernández M, Hernández Y, Lococo B, et al. New scheme of intermittent benznidazole administration in patients chronically infected with Trypanosoma cruzi: Clinical, parasitological, and serological assessment after three years of follow-up. Antimicrob Agent Chemo. 2020;64(9).
- 40. Bustamante JM, Craft JM, Crowe BD, Ketchie SA, Tarleton RL. New, combined, and reduced dosing treatment protocols cure Trypanosoma cruzi infection in mice. J Infect Dis. 2014;209(1):150–62. pmid:23945371
- 41. Bustamante JM, Sanchez-Valdez F, Padilla AM, White B, Wang W, Tarleton RL. A modified drug regimen clears active and dormant trypanosomes in mouse models of Chagas disease. Sci Trans Med. 2020;12(567).
- 42. Antunes AP, Ribeiro ALP, Sabino EC, Silveira MF, Oliveira CDL, Botelho AC de C. Benznidazole therapy for Chagas disease in asymptomatic Trypanosoma cruzi -seropositive former blood donors: evaluation of the efficacy of different treatment regimens. Rev Soc Bras Med Trop. 2016;49(6):713–20. pmid:28001218
- 43. Lascano F, García Bournissen F, Altcheh J. Review of pharmacological options for the treatment of Chagas disease. Br J Clin Pharmacol. 2022;88(2):383–402. pmid:33314266
- 44. Chatelain E, Konar N. Translational challenges of animal models in Chagas disease drug development: a review. Drug Des Devel Ther. 2015;9:4807–23. pmid:26316715
- 45. Watson JA, Cruz C, Barreira F, Forsyth C, Schijman A, Peploe R, et al. Quantifying anti-trypanosomal treatment effects in chronic indeterminate Chagas disease: an individual patient data meta-analysis of two proof of concept trials. medRxiv. 2024.
- 46. Duffy T, Cura CI, Ramirez JC, Abate T, Cayo NM, Parrado R, et al. Analytical performance of a multiplex Real-Time PCR assay using TaqMan probes for quantification of Trypanosoma cruzi satellite DNA in blood samples. PLoS Negl Trop Dis. 2013;7(1):e2000. pmid:23350002
- 47. Galindo Bedor DC, Tavares Cavalcanti Bedor NC, Viturino da Silva JW, Damasceno Sousa G, Pereira de Santana D, Garcia-Bournissen F, et al. Dried Blood Spot Technique-Based Liquid Chromatography-Tandem Mass Spectrometry Method as a Simple Alternative for Benznidazole Pharmacokinetic Assessment. Antimicrob Agents Chemother. 2018;62(12):e00845-18. pmid:30275095
- 48. Wiens MO, Kanters S, Mills E, Peregrina Lucano AA, Gold S, Ayers D, et al. Systematic review and meta-analysis of the pharmacokinetics of benznidazole in the treatment of Chagas disease. Antimicrob Agents Chemother. 2016;60(12):7035–42.
- 49. Lindbom L, Ribbing J, Jonsson EN. Perl-speaks-NONMEM (PsN)--a Perl module for NONMEM related programming. Comput Method Program Biomed. 2004;75(2):85–94.
- 50. Keizer RJ, van Benten M, Beijnen JH, Schellens JHM, Huitema ADR. Piraña and PCluster: a modeling environment and cluster infrastructure for NONMEM. Comput Methods Programs Biomed. 2011;101(1):72–9. pmid:20627442
- 51.
Brooks ME, Kristensen K, Benthem KJ, Magnusson A, Berg CW, Nielsen A, et al. glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. 2017;9:378.
- 52. Morris WF. Disentangling effects of induced plant defenses and food quantity on herbivores by fitting nonlinear models. Am Natural. 1997;150(3):299–327.
- 53.
Hertig F. DHARMa: Residual Diagnostics for Hierarchical (Multi-Level/ Mixed) Regression Models. R package version 0.4.6. 2022. Available from: https://CRAN.R-project.org/package=DHARMa
- 54. Wetzels JFM, Kiemeney LALM, Swinkels DW, Willems HL, den Heijer M. Age- and gender-specific reference values of estimated GFR in Caucasians: the Nijmegen Biomedical Study. Kidney Int. 2007;72(5):632–7. pmid:17568781
- 55. Würthwein G, Lanvers-Kaminsky C, Hempel G, Gastine S, Möricke A, Schrappe M, et al. Population Pharmacokinetics to Model the Time-Varying Clearance of the PEGylated Asparaginase Oncaspar® in Children with Acute Lymphoblastic Leukemia. Eur J Drug Metab Pharmacokinet. 2017;42(6):955–63. pmid:28349335
- 56. Savic RM, Jonker DM, Kerbusch T, Karlsson MO. Implementation of a transit compartment model for describing drug absorption in pharmacokinetic studies. J Pharmacokinet Pharmacodyn. 2007;34(5):711–26. pmid:17653836
- 57. Molina I, Salvador F, Sánchez-Montalvá A, Artaza MA, Moreno R, Perin L, et al. Pharmacokinetics of Benznidazole in Healthy Volunteers and Implications in Future Clinical Trials. Antimicrob Agents Chemother. 2017;61(4):e01912-16. pmid:28167552
- 58. Alonso-Vega C, Urbina JA, Sanz S, Pinazo MJ, Pinto JJ, Gonzalez VR, et al. New chemotherapy regimens and biomarkers for Chagas disease: the rationale and design of the TESEO study, an open-label, randomised, prospective, phase-2 clinical trial in the Plurinational State of Bolivia. BMJ Open. 2021;11(12):e052897.
- 59.
Drugs for Neglected Diseases. Study identifier: NCT04897516. Phase III Randomized,Multicenter Non-inferiority Study to Evaluate the Efficacy and Safety of Shorter Benznidazole Regimens Compared to the Standard Regimen to Treat Adult Patients With Chronic Chagas Disease [Internet]. 2021 [cited 5 August 2024]. Available from: https://clinicaltrials.gov/study/NCT04897516
- 60.
Evandro Chagas Institute of Clinical Research. Study identifier: NCT06339710. Double Blind, Phase III Randomized,Safety and Efficacy Non-inferiority Trial to Evaluate Two Short Benznidazole Regimens for the Treatment of Adults in the Chronic Phase of Chagas Disease in Its Indeterminate and Mild Cardiac Forms in Bolivia and Colombia [Internet]. 2024 [cited 5 August 2024]. Available from: https://clinicaltrials.gov/study/NCT06339710
- 61. Ribeiro I, Blum B, Fernandes J, Santina G, Asada M, Everson M, et al. Drug-Drug Interaction Study of Benznidazole and E1224 in Healthy Male Volunteers. Antimicrob Agents Chemother. 2021;65(4):e02150-19. pmid:33558286
- 62.
Schwartz DE, Hofheinz W. Metabolism of Nitroimidazoles. Nitroimidazoles Chemistry, Pharmacology,and Clinical Application. New York: Plenum Press; 1982.
- 63. Fernández ML, Marson ME, Ramirez JC, Mastrantonio G, Schijman AG, Altcheh J, et al. Pharmacokinetic and pharmacodynamic responses in adult patients with Chagas disease treated with a new formulation of benznidazole. Mem Inst Oswaldo Cruz. 2016;111(3):218–21. pmid:26982179
- 64. Zingales B, Miles MA, Moraes CB, Luquetti A, Guhl F, Schijman AG, et al. Drug discovery for Chagas disease should consider Trypanosoma cruzi strain diversity. Mem Inst Oswaldo Cruz. 2014;109(6):828–33. pmid:25317712
- 65. Zingales B, Miles MA, Campbell DA, Tibayrenc M, Macedo AM, Teixeira MMG, et al. The revised Trypanosoma cruzi subspecific nomenclature: rationale, epidemiological relevance and research applications. Infect Genet Evol. 2012;12(2):240–53. pmid:22226704
- 66. Moraes CB, Giardini MA, Kim H, Franco CH, Araujo-Junior AM, Schenkman S, et al. Nitroheterocyclic compounds are more efficacious than CYP51 inhibitors against Trypanosoma cruzi: implications for Chagas disease drug discovery and development. Sci Rep. 2014;4:4703.
- 67. Sánchez-Valdéz FJ, Padilla A, Wang W, Orr D, Tarleton RL. Spontaneous dormancy protects Trypanosoma cruzi during extended drug exposure. Elife. 2018;7:e34039. pmid:29578409
- 68. Ward AI, Lewis MD, Khan AA, McCann CJ, Francisco AF, Jayawardhana S, et al. In Vivo Analysis of Trypanosoma cruzi Persistence Foci at Single-Cell Resolution. mBio. 2020;11(4).
- 69. Ward AI, Olmo F, Atherton RL, Taylor MC, Kelly JM. Trypanosoma cruzi amastigotes that persist in the colon during chronic stage murine infections have a reduced replication rate. Open Biol. 2020;10(12):200261. pmid:33321060
- 70. Jayawardhana S, Ward AI, Francisco AF, Lewis MD, Taylor MC, Kelly JM, et al. Benznidazole treatment leads to DNA damage in Trypanosoma cruzi and the persistence of rare widely dispersed non-replicative amastigotes in mice. PLoS Pathog. 2023;19(11):e1011627. pmid:37956215
- 71. Khare S, Liu X, Stinson M, Rivera I, Groessl T, Tuntland T, et al. Antitrypanosomal Treatment with Benznidazole Is Superior to Posaconazole Regimens in Mouse Models of Chagas Disease. Antimicrob Agents Chemother. 2015;59(10):6385–94. pmid:26239982
- 72. Fonseca KDS, Perin L, de Paiva NCN, da Silva BC, Duarte THC, Marques FS, et al. Benznidazole treatment: time- and dose-dependence varies with the Trypanosoma cruzi strain. Pathogens. 2021;10(6).
- 73. Mazzeti AL, Diniz L de F, Gonçalves KR, Nascimento AFS, Spósito PAF, Mosqueira VCF, et al. Time and dose-dependence evaluation of nitroheterocyclic drugs for improving efficacy following Trypanosoma cruzi infection: A pre-clinical study. Biochem Pharmacol. 2018;148:213–21. pmid:29309767
- 74. Fesser AF, Braissant O, Olmo F, Kelly JM, Mäser P, Kaiser M. Non-invasive monitoring of drug action: A new live in vitro assay design for Chagas’ disease drug discovery. PLoS Negl Trop Dis. 2020;14(7):e0008487.
- 75. Assmus F, Adehin A, Hoglund RM, Fortes Francisco A, Lewis MD, Kelly JM, et al. Pharmacokinetic-pharmacodynamic modeling of benznidazole and its antitrypanosomal activity in a murine model of chronic Chagas disease. PLoS Negl Trop Dis. 2025;19(5):e0012968. pmid:40359193