Ex vivo activity of the ACT new components pyronaridine and piperaquine in comparison with conventional ACT drugs against isolates of Plasmodium falciparum

During the past 20 years, many strains of Plasmodium falciparum have become resistant to chloroquine and other anti-malarial drugs [1]. This has prompted a search for an effective alternative anti-malarial drug with minimal side effects. The emergence and spread of parasites resistant to anti-malarial drugs has caused an urgent need for novel compounds to be discovered and developed. One strategy for reducing the prevalence of malaria is the combinatorial use of drugs. The combination protects each drug from the development of resistance and reduces the overall transmission of malaria [2]. The official first-line anti-malarial policy is now artemisinin-based combination therapy (ACT) [3]. The artemisinin derivative causes rapid and effective reduction of the parasite biomass and gametocytes carriage, while the partner drug, which has a longer duration of action, achieves effective clinical and parasitological cure. Different formulations of ACT were evaluated: artesunate-sulphadoxine-pyrimethamine [4], artesunate-amodiaquine [5], artemether-lumefantrine [6], artesunate-mefloquine [7], artesunate-chlorproguanil-dapsone [8], dihydroartemisinin-piperaquine [9] and artesunate-pyronaridine [10].

However, suspected decreased susceptibility of ACT or, at least, longer parasite clearance times have been described in Cambodia [11‐14]. In addition, prior therapy by amodiaquine-containing ACT selected reduced response to monodesethylamodiaquine, suggested that amodiaquine-containing regimens may rapidly lose efficacy in Africa [15]. This emergence of parasite resistance to ACT indicates that novel compounds and combinations need to be discovered and developed.

The aim of the present work was to assess i) ex vivo activity of two recent ACT partner drugs, pyronaridine (PND) and piperaquine (PPQ) to define the susceptibility baseline, ii) comparison with standard components of ACT, such as monodesethylamodiaquine (the active metabolite of amodiaquine) (MDAQ), lumefantrine (LMF), mefloquine (MQ), artesunate (AS) or dihydroartemisinin (DHA), and iii) in vitro cross-resistance with other quinoline drugs, such as chloroquine (CQ) and quinine (QN).

One hundred and eighty one P. falciparum isolates were tested for their ex vivo susceptibility to PPQ, PND, CQ, QN, MQ, MDAQ, LMF, DHA and AS. The mean IC₅₀s for these nine anti-malarial drugs are presented in Table 1. The IC₅₀ values for PND ranged from 0.55 to 80.0 nM (geometric mean = 19.9 nM, 95% = CI 18.0-022.0) and from 11.8 to 217.3 nM for PPQ (geometric mean = 66.8 nM, 95% CI = 61.8-72.3). The distributions of the IC₅₀ for the nine anti-malarial drugs are presented in Figure 1.

In vitro cross-resistance was measured by pair-wise correlation of IC₅₀ values of 181 isolates (Table 2). A significant positive correlation was shown between responses to PPQ and PND responses (rho = 0.46; P < 0.00001) and between PPQ and MDAQ (rho = 0.30; P = 0.0001). No significant correlation was shown between PPQ IC₅₀ and responses to other anti-malarial drugs. A significant positive correlation was shown between responses to PND and MDAQ (rho = 0.37; P < 0.00001), LMF (rho = 0.28; P = 0.0002), QN (rho = 0.24; P = 0.0012), AS (rho = 0.19; P = 0.0108), DHA (rho = 0.18; P = 0.0173) and CQ (rho = 0.16; P = 0.0365). The highest coefficients of correlation were shown for correlation between responses to DHA and AS (rho = 0.84; P < 0.00001), LMF and MQ (rho = 0.58; P < 0.00001), CQ and MDAQ (rho = 0.51; P < 0.00001), LMF and DHA (rho = 0.46; P < 0.00001), LMF and AS (rho = 0.45; P < 0.00001) or CQ and QN (rho = 0.41; P < 0.00001).

Isolates with high IC₅₀ to at least one of the anti-malarial drug are listed in Table 3. One originated from Thailand with IC₅₀ = 71.5 nM for PND associated with IC50 = 91.1 nM for PPQ, IC50 = 1,131 nM for QN, IC50 = 166 nM for MQ, IC50 = 114 nM for LMF, IC50 = 21.2 nM for DHA and IC50 = 16.3 nM for AS.

Ex vivo analysis of the susceptibility of P. falciparum isolates to anti-malarial drugs has three important attributes. This approach allows firstly to assay the response of clinical isolates to individual drugs that are unmodified by important host factors that influence drug efficacy in vivo. Secondly, the progressive decline in drug susceptibility of isolates from the same site may identify incipient resistance in the parasite population. Finally, strains with reduced anti-malarial susceptibilities can then be established in continuous culture to provide the material needed to investigate novel molecular mechanisms of resistance as well as for tests of susceptibility to other anti-malarial agents.

The continued spread of P. falciparum drug resistance to monotherapies has forced a shift toward the use of ACT. Nevertheless, resistance to at least one component of many of the different formulations of ACT currently in clinical use has been documented, and it is feared that the widespread use of ACT will gradually reduce its clinical efficacy. Clinical failures or at least longer parasite clearance times have been described in Cambodia [12]. In addition, prior therapy by amodiaquine-containing ACT selected reduced response to monodesethylamodiaquine, suggested that amodiaquine-containing regimens may rapidly lose efficacy in Africa [15].

One hundred and eighty one P. falciparum strains were tested for their in vitro susceptibility to PPQ, PND, CQ, QN, MQ, MDAQ, LMF, DHA and AS. The IC₅₀ values for PND ranged from 0.55 to 80.0 nM (geometric mean = 19.9 nM, 95% CI = 18.0-22.0). These values are in accordance with previous studies on P. falciparum strains (1.9-47.8 nM and 15-49 nM, respectively) [19, 20] or in isolates from patients in Thailand cured with PND (geometric mean = 15.7 nM) or that recrudesced after PND treatment (geometric mean = 23.0 nM) [21] but higher than those found in isolates from Cameroon (geometric mean = 3.58 nM), Senegal (geometric mean = 3.8 nM and 4.52 nM), Gabon (geometric mean = 3.0 nM and 1.87 nM) [22‐26] and in Indonesia (geometric mean = 1.92 nM) [27]. In addition, PND is also effective in vitro against Plasmodium vivax isolates (geometric mean = 2.58 nM) [27].

Antagonistic in vitro drug interactions between PND and artemisinin derivatives have been described [22, 28, 29]. Previous studies have demonstrated in vitro cross-resistance between PND and DHA or CQ, with coefficients of determination (r²) of 0.84 and 0.19-0.46, respectively [19, 21, 23‐25]. A low r ² (0.20) was determined between PND and AS in P. falciparum strains [20]. In the present study, a positive correlation was shown between PND and DHA or AS responses with a coefficient of correlation rho = 0.18 and 0.19, respectively. To suggest common mechanisms of action or resistance for two compounds, that could induce cross-resistance, the coefficient of determination must be high, such as the one for AS and DHA (rho = 0.84 and r² = 0.68) corresponding to 68% of the variation in the response to DHA is explained by variation in the response to AS. A coefficient of determination of 0.027 or 0.038 means that only 2.7% and 3.8% of the variation in the response to PND is explained by variation in the response to DHA and AS. These data suggest that there is no cross-resistance between PND and artemisinin derivatives. In addition, the combination of PND and AS has undergone successful clinical evaluation in Africa [10, 30]. In the present study, a positive correlation was shown between PND and CQ responses with a coefficient of correlation rho = 0.16 and r ² = 0.017. This means that only 1.7% of the variation in the response to PND is explained by variation in the response to CQ. These data are consistent with the efficacy of the combination of PND and AS for areas in which parasites are resistant to CQ [10, 30]. There have been conflicting reports on the correlations between P. falciparum responses to PND and CQ. Previous studies showed weak (from 0.003 to 0.26) [20, 23‐25, 27] to modest (0.40 and 0.46) [19, 21] coefficients of determination for correlations between PND and CQ. PND appeared to be equally effective in vitro against 37 isolates from two areas of Thailand with different CQ resistance levels [31]. Similarly, Basco and Le Bras showed no correlation between resistance to PND and CQ for 31 isolates from Central and West Africa [32]. These results suggest that no cross-resistance exists between PND and CQ. In addition, an isolate collected in a patient who took part in trekking along the Mekong from the south of Laos to the north of Thailand showed high susceptibility to CQ and MDAQ and very low susceptibility to PND (71.5 nM), PPQ (91.1 nM), QN (1,131 nM), MQ (166 nM), LMF (114 nM) and to the artemisinin derivatives DHA (21.2 nM) and AS (16.3 nM) [33]. This multi-resistant isolate was suspected of being resistant to ACT. All of the other significant positive correlations between PND and QN, MDAQ or LMF are too low (rho < 0.37, r² < 0.10) to suggest cross-resistance between PND and quinoline drugs. This absence of cross-resistance may be explained by the absence of association between PND and genes involved in quinoline resistance, such as pfcrt, pfmdr1, pfmrp or pfnhe-1 [20].

The highest coefficient of correlation was observed between PND and PPQ (rho = 0.47 and r² = 0.23). This means that 23% of the variation in the response to PND is explained by variation in the response to PPQ. This result is also too low to suggest cross-resistance between the two drugs. This result was in accordance with previous data (r² = 0.20) [27].

In vitro drug interaction between PPQ and artemisinin derivatives was indifferent or antagonistic [34‐37]. However, PPQ has been combined with DHA and it has undergone successful clinical evaluation in Africa, Asia and South America [9, 38‐40]. In addition, PPQ-DHA is also effective to treat P. vivax malaria [41]. But unfortunately clinical failures to PPQ have been reported in areas of China where it has been deployed in monotherapy for P. falciparum [42].

The IC₅₀ values for PPQ ranged from 11.8 to 217.3 nM (geometric mean = 66.8 nM, CI95% 61.8-72.3). These data are in accordance with previous studies on P. falciparum strains [35, 43] or isolates from Cameroon (geometric mean = 39 nM) [44], Thai-Burmese border (49 nM) [45] and Kenya (50 nM) [46], but superior to geometric mean of isolates from Uganda (6.1 nM) [47], Indonesia (21.8 nM) [48] or Papua New Guinea [49]. The isolate with the highest IC₅₀ to PPQ (217.3 nM) was also resistant to CQ (1029 nM) and MDAQ (240 nM).

Encouragingly, PPQ has not shown any correlation with the other quinoline drugs, i.e. CQ, QN, MDAQ, LMF or MQ. These results suggest that no cross-resistance exists between PPQ and CQ and the other quinoline anti-malarial drugs. These data are in accordance with previous studies, which showed weak coefficients of determination included between 0.015 and 0.14 for correlation between PPQ and CQ [43, 45‐47]. No significant correlation was identified between PPQ and DHA or AS. These data are in accordance with previous results [47]. PPQ has shown no cross-resistance with any of the anti-malarial drug tested. The absence of cross-resistance may be explained by the absence of association between PPQ and genes involved in quinoline resistance, such as pfcrt, pfmdr1, pfmrp or pfnhe-1 [43]. A copy number variation event on chromosome 5 could be associated with PPQ resistance [50]. Transgene expression studies are underway with individual genes in this segment to evaluate their contribution to PPQ resistance.

DHA-PPQ is an inexpensive, safe and highly effective treatment for uncomplicated falciparum and vivax malaria [41, 51, 52]. DHA-PPQ seems to offer a better post-treatment prophylactic effect following therapy compared with artemether-LMF [6, 53, 54] or AS-amodiaquine [55]. The significant lower risk of recurrent parasitaemia after treatment with DHA-PPQ is likely to be explained by differences in pharmacokinetics of the non-artemisinin partner drugs. PPQ, a bisquinoline, is estimated to have elimination half-life of 17-33 days [38, 56, 57] while LMF and MDAQ have an estimated elimination half-life of two to six days and one-10 days [58], respectively.

The PND-AS combination (Pyramax^®) is one of the latest ACT combinations currently under development by the not-for-profit organisation Medicines for Malaria Venture (Geneva, Switzerland) and the pharmaceutical company Shin Poong Pharmaceuticals (Seoul, Republic of Korea) for the treatment of uncomplicated P. falciparum malaria and for the blood stages of P. vivax malaria. Pyramax^® has recently completed phase III trials in humans. A five-day regimen of PND alone (total dose = 1800 mg) produced a better cure rate than AS, artemether or MQ used alone in the same conditions in Thailand [21]. Efficacy of PND-AS was non-inferior to that of artemether-LMF for treatment of uncomplicated falciparum malaria in Africa and Southeast Asia [10].

In this study, the excellent in vitro anti-malarial activities of the ACT components PPQ and PND on P. falciparum were confirmed. The absence of cross-resistance with quinolines and artemisinin derivatives confirmed the relevant use of the combinations of PPQ and DHA or PND and AS in areas where parasites are resistant to conventional anti-malarial drugs.