Safety of primaquine given to people with G6PD deficiency: systematic review of prospective studies

Abstract

Background: Haemolysis risk with single dose or short course primaquine was evaluated in glucose-6-phosphate dehydrogenase (G6PD) deficient people.

Methods: Major electronic databases (to August 2016) were searched for single or short course 8-aminoquinolines (8-AQ) in (1) randomized comparisons against placebo in G6PD deficient people; and (2) observational comparisons in G6PD deficient compared to replete people. Two authors independently assessed eligibility, risk-of-bias, and extracted data.

Results: Five randomized controlled trials and four controlled observational cohorts were included. In G6PD deficient individuals, high-dose (0.75 mg/kg) PQ resulted in lower average haemoglobin levels at 7 days (mean difference [MD] -1.45 g/dl, 95% CI -2.17 to -0.74, 2 trials) and larger percentage fall from baseline to day 7 (MD -10.31%, 95% CI -17.69 to -2.92, 3 trials) compared to placebo. In G6PD deficient compared to replete people, average haemoglobin was lower at 7 days (MD -1.19 g/dl, 95% CI -1.94 to -0.44, 2 trials) and haemoglobin change from baseline to day 7 was greater (MD -9.10%, 95% CI -12.55 to -5.65, 5 trials). One small trial evaluated mid-range PQ dose (0.4-0.5 mg/kg) in G6PD deficient people, with no difference detected in average haemoglobin at day 7 compared to placebo. In one cohort comparing G6PD deficient and replete people there was a greater fall with G6PD deficiency (MD -4.99%, 95% CI -9.96 to -0.02). For low-dose PQ (0.1-0.25 mg/kg) in G6PD deficient people, haemoglobin change from baseline was similar to the placebo group (MD 1.72%, 95% CI -1.89 to 5.34, 2 trials). Comparing low dose PQ in G6PD deficient with replete people, the average haemoglobin was lower in the G6PD deficient group at 7 days (-0.57 g (95% CI -0.97 to -0.17, 1 trial)); although change from baseline was similar (MD -1.45%, 95% CI -5.69 to 2.78, 3 trials).

Conclusions: Falls in average haemoglobin are less marked with the 0.1 to 0.25 mg/kg PQ than with the 0.75 mg/kg dose, and severe haemolytic events are not common. However, data were limited and the evidence GRADE was low or very low certainty.

Keywords: Glucose-6-phosphate dehydrogenase (G6PD); Malaria; Primaquine.

Fig. 1

G6PD classification

Fig. 2

Study selection flow diagram

Fig. 3

Risk of bias of included studies

Fig. 4

High dose PQ (0.75 mg/kg) compared to placebo in G6PD deficient people

Fig. 5

High dose PQ (0.75 mg/kg) in G6PD deficient compared to G6PD replete people

Fig. 6

Mid-range dose (0.4–0.5 mg/kg) PQ compared to placebo in G6PD deficient people

Fig. 7

Mid-range dose (0.4–0.5 mg/kg) PQ in G6PD deficient compared to G6PD replete people

Fig. 8

Low dose (0.1–0.25 mg/kg) PQ compared to placebo in G6PD deficient people

Fig. 9

Low dose (0.1–0.25 mg/kg) PQ in G6PD deficient compared to G6PD replete people