Introduction: Primaquine is the only widely available drug that targets Plasmodium vivax malaria parasites in the liver. However, primaquine is underused because of concerns that this drug can cause dose-dependent haemolysis in glucose-6-phosphate dehydrogenase (G6PD) deficient individuals. We developed and calibrated a within-host Bayesian pharmacodynamic model of red blood cell (RBC) production and turnover to explore the safety of primaquine regimens for individuals with G6PD deficiency.
Methods: The within-host RBC model captured deviations from the RBC normal process by assuming a dose-dependent effect of primaquine on the lifespan of circulating erythrocytes and the compensatory response of precursor cells and reticulocytes. The compartmental mechanistic RBC model was fitted to data from a regimen-adaptive trial of ascending primaquine doses in 23 G6PD hemizygote deficient volunteers using a Bayesian hierarchical framework. Detailed serial haemoglobin and reticulocyte count data were available, 1523 individual measurements over 656 unique time points.
Results: Our model estimated that primaquine doses of ~0.75 mg base/kg reduce the circulating lifespan of deficient erythrocytes by ~30 days in individuals with common Southeast Asian G6PD variants. We predict that a total dose of 5 mg/kg of primaquine can be administered safely to G6PD deficient individuals over 14 days with expected drops in haemoglobin of 2.7 to 6.5 g/dL from a baseline of 15 g/dL.
Conclusion: Our within-host RBC model captured the effect of primaquine on G6PD deficient volunteers’ haemoglobin and reticulocyte profiles. This mechanistic model suggests safe and effective primaquine dosing schemes can be administered within two weeks to G6PD deficient individuals.
Funding: Bill & Melinda Gates Foundation, Wellcome Trust, Australian National Health & Medical Research Council.