Dynamics of G6PD activity in patients receiving weekly primaquine for therapy of Plasmodium vivax malaria

Kinetics of glucose-6-phosphate dehydrogenase (G6PD) activity during Plasmodium vivax infection: implications for early radical malaria treatment

BACKGROUND

Plasmodium vivax relapses due to dormant liver hypnozoites can be prevented with primaquine. However, the dose must be adjusted in individuals with glucose-6-phosphate-dehydrogenase (G6PD) deficiency. In French Guiana, assessment of G6PD activity is typically delayed until day (D)14 to avoid the risk if misclassification. This study assessed the kinetics of G6PD activity throughout P. vivax infection to inform the timing of treatment.

METHODS

For this retrospective monocentric study, data on G6PD activity between D1 and D28 after treatment initiation with chloroquine or artemisinin-based combination therapy were collected for patients followed at Cayenne Hospital, French Guiana, between January 2018 and December 2020. Patients were divided into three groups based on the number of available G6PD activity assessments: (i) at least two measurements during the P. vivax malaria infection; (ii) two measurements: one during the current infection and one previously; (iii) only one measurement during the malaria infection.

RESULTS

In total, 210 patients were included (80, 20 and 110 in groups 1, 2 and 3, respectively). Data from group 1 showed that G6PD activity remained stable in each patient over time (D1, D3, D7, D14, D21, D28). None of the patients with normal G6PD activity during the initial phase (D1-D3) of the malaria episode (n = 44) was categorized as G6PD-deficient at D14. Patients with G6PD activity < 80% at D1 or D3 showed normal activity at D14. Sex and reticulocyte count were statistically associated with G6PD activity variation. In the whole sample (n = 210), no patient had severe G6PD deficiency (< 10%) and only three between 10 and 30%, giving a G6PD deficiency prevalence of 1.4%. Among the 100 patients from group 1 and 2, 30 patients (26.5%) were lost to follow-up before primaquine initiation.

CONCLUSIONS

In patients treated for P. vivax infection, G6PD activity did not vary over time. Therefore, G6PD activity on D1 instead of D14 could be used for primaquine dose-adjustment. This could allow earlier radical treatment with primaquine, that could have a public health impact by decreasing early recurrences and patients lost to follow-up before primaquine initiation. This hypothesis needs to be confirmed in larger prospective studies.

Case Series of Primaquine-Induced Haemolytic Events in Controlled Trials with G6PD Screening

Primaquine for radical cure of Plasmodium vivax malaria poses a potentially life-threatening risk of haemolysis in G6PD-deficient patients. Herein, we review five events of acute haemolytic anaemia following the administration of primaquine in four malaria trials from Indonesia, the Solomon Islands, and Vietnam. Five males aged 9 to 48 years were improperly classified as G6PD-normal by various screening procedures and included as subjects in trials of anti-relapse therapy with daily primaquine. Routine safety monitoring by physical examination, urine inspection, and blood haemoglobin (Hb) assessment were performed in all those trials. Early signs of acute haemolysis, i.e., dark urine and haemoglobin drop >20%, occurred only after day 3 and as late as day 8 of primaquine dosing. All patients were hospitalized and fully recovered, all but one following blood transfusion rescue. Hb nadir was 4.7 to 7.9 g/dL. Hospitalization was for 1 to 7 days. Hb levels returned to baseline values 3 to 10 days after transfusion. Failed G6PD screening procedures in these trials led G6PD-deficient patients to suffer harmful exposures to primaquine. The safe application of primaquine anti-relapse therapy requires G6PD screening and anticipation of its failure with a means of prompt detection and rescue from the typically abrupt haemolytic crisis.

A Review of the Current Status of G6PD Deficiency Testing to Guide Radical Cure Treatment for Vivax Malaria

Plasmodium vivax malaria continues to cause a significant burden of disease in the Asia-Pacific, the Horn of Africa, and the Americas. In addition to schizontocidal treatment, the 8-aminoquinoline drugs are crucial for the complete removal of the parasite from the human host (radical cure). While well tolerated in most recipients, 8-aminoquinolines can cause severe haemolysis in glucose-6-phosphate dehydrogenase (G6PD) deficient patients. G6PD deficiency is one of the most common enzymopathies worldwide; therefore, the WHO recommends routine testing to guide 8-aminoquinoline based treatment for vivax malaria whenever possible. In practice, this is not yet implemented in most malaria endemic countries. This review provides an update of the characteristics of the most used G6PD diagnostics. We describe the current state of policy and implementation of routine point-of-care G6PD testing in malaria endemic countries and highlight key knowledge gaps that hinder broader implementation. Identified challenges include optimal training of health facility staff on point-of-care diagnostics, quality control of novel G6PD diagnostics, and culturally appropriate information and communication with affected communities around G6PD deficiency and implications for treatment.

Variation in Glucose-6-Phosphate Dehydrogenase activity following acute malaria

Primaquine and tafenoquine are the only licensed drugs with activity against Plasmodium vivax hypnozoites but cause haemolysis in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency. Malaria also causes haemolysis, leading to the replacement of older erythrocytes with low G6PD activity by reticulocytes and young erythrocytes with higher activity. Aim of this study was to assess the impact of acute malaria on G6PD activity. Selected patients with uncomplicated malaria were recruited in Bangladesh (n = 87), Indonesia (n = 75), and Ethiopia (n = 173); G6PD activity was measured at the initial presentation with malaria and a median of 176 days later (range 140 to 998) in the absence of malaria. Among selected participants (deficient participants preferentially enrolled in Bangladesh but not at other sites) G6PD activity fell between malaria and follow up by 79.1% (95%CI: 40.4 to 117.8) in 6 participants classified as deficient (<30% activity), 43.7% (95%CI: 34.2 to 53.1) in 39 individuals with intermediate activity (30% to <70%), and by 4.5% (95%CI: 1.4 to 7.6) in 290 G6PD normal (≥70%) participants. In Bangladesh and Indonesia G6PD activity was significantly higher during acute malaria than when the same individuals were retested during follow up (40.9% (95%CI: 33.4-48.1) and 7.4% (95%CI: 0.2 to 14.6) respectively), whereas in Ethiopia G6PD activity was 3.6% (95%CI: -1.0 to -6.1) lower during acute malaria. The change in G6PD activity was apparent in patients presenting with either P. vivax or P. falciparum infection. Overall, 66.7% (4/6) severely deficient participants and 87.2% (34/39) with intermediate deficiency had normal activities when presenting with malaria. These findings suggest that G6PD activity rises significantly and at clinically relevant levels during acute malaria. Prospective case-control studies are warranted to confirm the degree to which the predicted population attributable risks of drug induced haemolysis is lower than would be predicted from cross sectional surveys.

Comparative pharmacokinetics and tissue distribution of primaquine enantiomers in mice

Background

Primaquine (PQ) has been used for the radical cure of relapsing Plasmodium vivax malaria for more than 60 years. PQ is also recommended for prophylaxis and prevention of transmission of Plasmodium falciparum. However, clinical utility of PQ has been limited due to toxicity in individuals with genetic deficiencies in glucose 6-phosphate dehydrogenase (G6PD). PQ is currently approved for clinical use as a racemic mixture. Recent studies in animals as well as humans have established differential pharmacological and toxicological properties of the two enantiomers of PQ. This has been attributed to differential metabolism and pharmacokinetics of individual PQ enantiomers. The aim of the current study is to evaluate the comparative pharmacokinetics (PK), tissue distribution and metabolic profiles of the individual enantiomers in mice.

Methods

Two groups of 21 male Albino ND4 Swiss mice were dosed orally with 45 mg/kg of S-(+)-PQ and R-(−)PQ respectively. Each of the enantiomers was comprised of a 50:50 mixture of 12C- and 13C- stable isotope labelled species (at 6 carbons on the benzene ring of the quinoline core). Three mice were euthanized from each group at different time points (at 0, 0.5, 1, 2, 4, 8, 24 h) and blood was collected by terminal cardiac bleed. Liver, spleen, lungs, kidneys and brain were removed, extracted and analysed using UPLC/MS. The metabolites were profiled by tandem mass (MS/MS) fragmentation profile and fragments with 12C–13C twin peaks. Non-compartmental analysis was performed using the Phoenix WinNonLin PK software module.

Results

The plasma AUC0-last (µg h/mL) (1.6 vs. 0.6), T1/2 (h) (1.9 vs. 0.45), and Tmax (h) (1 vs. 0.5) were greater for SPQ as compared to RPQ. Generally, the concentration of SPQ was higher in all tissues. At Tmax, (0.5–1 h in all tissues), the level of SPQ was 3 times that of RPQ in the liver. Measured Cmax of SPQ and RPQ in the liver were about 100 and 40 times the Cmax values in plasma, respectively. Similar observations were recorded in other tissues where the concentration of SPQ was higher compared to RPQ (2× in the spleen, 6× in the kidneys, and 49× in the lungs) than in the plasma. CPQ, the major metabolite, was preferentially generated from RPQ, with higher levels in all tissues (> 10× in the liver, and 3.5× in the plasma) than from SPQ. The PQ-o-quinone was preferentially formed from the SPQ (> 4× compared to RPQ), with higher concentrations in the liver.

Conclusion

These studies show that in mice, PQ enantiomers are differentially biodistributed and metabolized, which may contribute to differential pharmacologic and toxicity profiles of PQ enantiomers. The findings on higher levels of PQ-o-quinone in liver and RBCs compared to plasma and preferential generation of this metabolite from SPQ are consistent with the higher anti-malarial efficacy of SPQ observed in the mouse causal prophylaxis test, and higher haemolytic toxicity in the humanized mouse model of G6PD deficiency. Potential relevance of these findings to clinical use of racemic PQ and other 8-aminoquinolines vis-à-vis need for further clinical evaluation of individual enantiomers are discussed.