Tafenoquine treatment of Plasmodium vivax malaria: suggestive evidence that CYP2D6 reduced metabolism is not associated with relapse in the Phase 2b DETECTIVE trial

Participants’ samples were from Part 1 of the seamless Ph2b/3 TAF112582 study, a multi-centred, double-blind, randomized, parallel-group, placebo-controlled study to evaluate the efficacy, safety and tolerability of TQ in subjects infected with P. vivax [2]. TAF112582 Part 1 consisted of six treatment arms: CQ (600 mg days 1 and 2, 300 mg day 3) plus TQ and PQ placebos; CQ (doses as above) combined with single dose TQ 50, 100, 300 or 600 mg plus PQ placebo; or CQ (doses as above) combined with PQ 15 mg daily for 14 days plus TQ placebo. Protocol approval was obtained from each site’s ethics committee or institutional review board and prospective written informed consent was obtained for all subjects involved in this PGx study, which was funded by GlaxoSmithKline (GSK) and the Medicines for Malaria Venture (MMV).

A population PK model was developed to characterize systemic TQ concentrations in TAF112582 Part 1 subjects treated with TQ. Model-predicted individual post hoc clearance estimates were utilized to generate the individual exposure (AUC) values for the analyses [14].

Venous blood was collected into an EDTA vacutainer for each of the subjects who consented to PGx research. Genomic DNA was extracted from peripheral blood using the Gentra Puregene kit on the Autopure LS (Qiagen, Valencia, CA, USA) by Quest Diagnostics (Valencia, CA, USA or Heston, UK). CYP2D6 genotyping was performed by BioProcessing Solutions (Piscataway, NJ, USA) using the Affymetrix^® DMET-Plus array (Santa Clara, CA, USA). All genotype calling and quality control was performed in accordance with the manufacturers’ protocols.

Genetically predicted CYP2D6 metabolizer status was determined from this array. The translation of genotype information into metabolizer phenotype is challenging given the range of activity possible for an allele [12] and many classification schemes have been proposed or used in practice [5, 7, 12]. Here, subjects were classified as (1) poor metabolizers (PM) if they carried two ‘null’ (no enzymatic function) alleles; (2) intermediate metabolizers (IM) if they carried one null and one functionally deficient allele, two deficient alleles, or one null allele and one normal (wildtype function) allele; and, (3) extensive metabolizers (EM) if they carried two normal alleles or one normal allele and one deficient allele. In addition, the CYP2D6 ‘activity score’ (AS) [12], which translates genotype into a qualitative measure of enzyme activity, was explored in this study. Briefly, for each study subject, AS was determined by summing the per-allele scores; a null allele having a score of 0, a deficient allele a score of 0.5 and a normal allele a score of 1. Thus, PMs could have an AS of 0, IMs an AS of 0.5 or 1 and EMs an AS of 1.5 or 2.0.

TAF112582 showed poor efficacy for TQ50 and TQ100 mg treatment arms, while higher dose treatment arms, TQ300 and TQ600 mg, showed good efficacy [2]. Therefore, to facilitate interpretation of results, the TQ low dose (50 and 100 mg) and high dose (300 and 600 mg) treatment arms were combined. An exact logistic regression model adjusting for country of origin was used to assess the effect of inferred CYP2D6 metabolizer phenotype or AS on clinical outcome within treatment arm. Because only 1 PM was identified in the clinical study, only subjects with IM or EM phenotypes were included in the analysis. Given published data indicating an effect of CYP2D6 metabolism on PQ efficacy [3‐7], a one-sided test was applied to test the one-sided alternative hypothesis that reduced metabolism increased the risk of relapse due to P. vivax infection 6 months post-dosing. For the PK assessment, a post hoc analysis using a non-parametric method to compare median TQ AUC between IM and EM subjects was conducted separately in the TQ low dose (50 or 100 mg) and TQ high dose (300 or 600 mg) groups. All analyses were conducted using SAS/STAT (SAS Systemv9.2, SAS Institute Inc, Cary, NC, USA).

Tafenoquine was supplied by Chemical Development, GlaxoSmithKline Research and Development Ltd. (Stevenage, UK). Glucose-6-phosphate (G6P), glucose-6-phosphate dehydrogenase (G6PD) and β-nicotinamide adenine dinucleotide phosphate (NADP) were obtained from Sigma Chemical Company (St Louis, USA). All other commercially obtained chemicals and solvents were of high-performance liquid chromatography (HPLC) or analytical grade.

Supersomes™, containing individually over-expressed human CYP enzymes, derived from baculovirus-infected insect cells, and control Supersomes™ (lacking any native human CYP activity) were obtained from BD Biosciences, Woburn, MA, USA. The enzymes used in this study were: CYP1A2, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP3A4 (Lot# 72600, 76739, 73030, 71916, 73755, and 70741, respectively). Supersomes™ expressing CYP2C8, CYP2C9, CYP2C19, and CYP3A4 co-expressed CYP reductase and cytochrome b₅, while Supersomes™ expressing CYP1A2 and CYP2D6 co-expressed CYP reductase only. TQ was incubated at 5 and 10 μM with Supersomes™ (BD Gentest, Woburn, MA, USA), overexpressing individual cytochrome P450 enzymes 1A2, 2C8, 2C9, 2C19, 2D6, or 3A4 at 300 pmol/ml P450, in phosphate buffer (50 mM, pH 7.4) at 37 °C in a shaking water bath. Following an equilibration period of 5 min at 37 °C, reactions were started by the addition of pre-warmed cofactor solution [a NADPH regenerating system containing 1.7 mg of NADP, 7.8 mg of G6P, and six units of G6PD per ml of 2 % (w/v) sodium hydrogen carbonate]. Incubations were conducted under conditions which limited the exposure of samples to light. All incubations were terminated after 120 min by addition of a volume of acetonitrile equal to the incubation volume. Control incubations were also performed in the absence of Supersomes™, in the absence of cofactor and in the absence of TQ. Samples were centrifuged (Eppendorf, Hauppauge, NY, USA; approximately 13,000g, 3 min, room temperature), and supernatants removed for analysis by ultra performance liquid chromatography-mass spectrometry (UPLC-MS).

The human hepatocyte relay method used was similar to that previously reported by Di et al. [15]. Pooled cryopreserved human hepatocytes from 26 donors were obtained from Life Technologies (MA, USA). Following thawing (as per supplier guidance), hepatocytes were resuspended in Williams’ E medium (Sigma) supplemented with cell maintenance pack (Life Technologies). The cells were counted using the trypan blue exclusion method and 12-well hepatocyte plates containing 1 million cells/ml were spiked with TQ at a final concentration of 5 and 10 μM in a final incubation volume of 0.50 ml. The plates were incubated at 37 °C with 95 % oxygen/5 % carbon dioxide, 75 % relative humidity for 24 h on an orbital shaker in an incubator. At 24 h the hepatocyte suspension in the incubation plate was centrifuged (13,000g, 3 min, room temperature); a total of 600 µl of the supernatant was transferred to a clean 12-well plate and stored at −80 °C until the next relay experiment. For the next relay experiment, the supernatant plates were pre-warmed to 37 °C for approximately 20 min, and further hepatocytes (prepared as described previously) were added to the samples to yield a final cell density of 1 million cells/ml. The plates were incubated at 37 °C for 24 h and processed as described earlier. Six relays were undertaken to give a total incubation time of 6 days and samples were protected from light wherever possible. A drug-only control plate was run under the same conditions to investigate any non-metabolic drug breakdown. At the end of the six relay method period the entire hepatocyte suspension was collected and centrifuged (Eppendorf, Hauppauge, NY, USA) at 13,000g for 3 min at room temperature, and transferred to vials prior to UPLC-MS analysis.

Analyses of reaction phenotyping samples for TQ-related components were carried out by UPLC-MS using a Waters Acquity UPLC connected to a Waters Synapt G2S mass spectrometer using an Acquity UPLC BEH C18 column (100 × 2.1 mm, 1.7 µm; Waters Corporation, Milford, MA, USA). The mobile phase consisted of 50 mM ammonium acetate (native pH; solvent A) and acetonitrile (solvent B) at a flow rate of 0.3 ml/min. A gradient was used, starting at 5 % B then linear phase to 95 % B by 12 min, held for 1 min before returning to the initial conditions by 13.5 min with these conditions being maintained for an additional 1.5 min before subsequent injections. TQ ([M + H]⁺) was detected at m/z 464 using positive ion electrospray ionization. Drug-related components were identified based on charged molecular ions, mass accuracy and their collision-induced dissociation fragmentation. Authentic standards, when available, were used to compare chromatographic retention times and fragmentation patterns. Supporting data from previous studies were also used in the assignment of structures (unpublished data on file within DMPK (Drug Metabolism and Pharmacokinetics) Department, GSK).