IntroductionIntramuscular benzathine penicillin G (BPG) injections are a cornerstone of secondary prophylaxis to prevent acute rheumatic fever (ARF) and rheumatic heart disease (RHD). Uncertainties regarding inter-ethnic and preparation variability, and target exposure profiles of BPG injection are key knowledge gaps for RHD control.MethodsTo evaluate BPG pharmacokinetics (PK) in patients receiving 4-weekly doses in Ethiopia, we conducted a prospective cohort study of ARF/RHD patients attending cardiology outpatient clinics. Serum samples were collected weekly for one month after injection and assayed with a liquid chromatography-mass spectroscopy assay. Concentration-time datasets for BPG were analyzed by nonlinear mixed effects modelling using NONMEM.ResultsA total of 190 penicillin concentration samples from 74 patients were included in the final PK model. The median age, weight, BMI was 21 years, 47 kg and 18 kg/m², respectively. When compared with estimates derived from Indigenous Australian patients, the estimate for median (95% confidence interval) volume of distribution (V/F) was lower (54.8 [43.9–66.3] l.70kg^-1) whilst the absorption half-life (t_1/2-abs2) was longer (12.0 [8.75–17.7] days). The median (IQR) percentage of time where the concentrations remained above 20 ng/mL and 10 ng/mL within the 28-day treatment cycle was 42.5% (27.5–60) and 73% (58.5–99), respectively.ConclusionsThe majority of Ethiopian patients receiving BPG as secondary prophylaxis to prevent RHD do not attain target concentrations for more than two weeks during each 4-weekly injection cycle, highlighting the limitations of current BPG strategies. Between-population variation, together with PK differences between different preparations may be important considerations for ARF/RHD control programs.