Background: Primaquine has been the only widely available hypnozoitocidal anti-malarial drug for half a century. Despite this its clinical efficacy is poorly characterized resulting in a lack of consensus over the optimal regimen for the radical cure of Plasmodium vivax.
Methods: Published studies since 1950 of the use of primaquine regimens for preventing P. vivax relapse were reviewed. Data were extracted systematically from available papers. Primaquine regimens were categorized according to the total dose administered: very low (≤2.5 mg/kg), low (>2.5 mg/kg- < 5.0 mg/kg) and high (≥ 5.0 mg/kg). The risk of recurrent infection were summarized across geographical regions and the odds ratios between treatment regimens calculated after stratifying by total treatment dose and duration of study follow up.
Results: Data could be retrieved from 87 clinical trials presenting data in 59,735 patients enrolled into 156 treatment arms, conducted in 20 countries. There was marked heterogeneity in study design, particularly primaquine dosing and duration of follow up. The median rate of recurrence following very low dose of primaquine (n = 44) was 25% (range 0-90%) at 4-6 months, compared to 6.7 % (range 0-59%) following low dose primaquine (n = 82). High dose primaquine regimens were assessed in 28 treatment arms, and were associated with a median recurrence rate of 0% (Range: 0-15%) at one month. In 18 studies with control arms, the effectiveness of a very low dose primaquine regimen was no different from patients who did not receive primaquine (OR = 0.60, 95%CI 0.33-1.09, p = 0.09), whereas for the low dose regimens a significant difference was reported in 50% (6/12) of studies (overall OR = 0.14, 95%CI: 0.06-0.35, p < 0.001). Two studies enrolling 171 patients demonstrated high effectiveness of high dose primaquine compared to a control arm (OR = 0.03 (95%CI: 0.01-0.13); p < 0.0001).
Conclusions: Low dose regimens retain adequate efficacy in some areas, but this is not uniform. The efficacy and safety of pragmatic high dose primaquine regimens needs to be assessed in a range of endemic and geographical locations. Such studies will require a prolonged period of follow up and comparison with control arms to account for confounding factors.