Atorvastatin as a Potential Antimalarial Drug: In Vitro Synergy in Combinational Therapy with Dihydroartemisinin

Atorvastatin (AVA) is a synthetic inhibitor of 3-hydroxy-3-methylglutaryl-coenzyme A. AVA has lipid-lowering ability, a long plasma half-life, and an excellent safety record. In addition to the previously mentioned positive aspects of AVA, this drug has also shown in vitro antimalarial activity (10). AVA showed no in vitro cross-resistance with the commonly used antimalarials. In addition, the 50% inhibitory concentrations (IC₅₀s) of AVA were found to be unrelated to mutations occurring in transport protein genes involved in quinoline antimalarial drug resistance, including the pfcrt, pfmdr1, pfmrp, and pfnhe-1 genes (8). Nevertheless, AVA treatment alone, at a dose of 20 mg/kg of body weight, failed to prevent death due to cerebral malaria and was not able to affect parasitemia in infected mice (1). One possibility is that AVA could act as an adjuvant therapy. Wong and Davis recently showed that AVA did not potentiate the in vitro activity of dihydroartemisinin (DHA) in the 3D7 strain of P. falciparum (13) and also posited that AVA had no clinical relevance because AVA IC₅₀s were well above the plasma concentrations achievable in humans taking high-dose AVA (0.1 to 0.5 μM) (2).

In the present study, we tested the effects of AVA at six concentrations (0.1, 0.5, 1, 2, 4, and 8 μM). We used 13 well-established strains of P. falciparum for the analysis (Table 1) (6). The methodology of the in vitro potentiation test was previously described (5). The combination of AVA and DHA had synergistic effects (Fig. 1). The differences in DHA IC₅₀ between AVA concentrations groups have first been tested using analysis of variance for repeated measures to take into account the fact that observations made within each strain were not independent. Using the most conservative correction for interdependence between observations (i.e., Box's conservative epsilon), the differences in DHA IC₅₀ were highly significant (P < 0.001) in the range of AVA plasma concentrations achievable in patients taking 80 mg of AVA daily (Table 1). Using a random-effect linear regression approach, the regression coefficients for the log-transformed DHA IC₅₀ indicated that the mean fold change in the DHA IC₅₀ when adding AVA at concentrations of 0.1, 0.5, and 1.0 μM (0.90, 0.82, and 0.72, respectively) were also highly significant (P < 0.001). The DHA IC₅₀ was reduced by 10% (range, 0 to 24%; 95% confidence interval [CI], 5 to 15%), 18% (range, 8 to 38%; 95% CI, 14 to 22%), and 28% (range, 15 to 43%; 95% CI, 24 to 32%) in the presence of AVA at concentrations of 0.1, 0.5, and 1.0 μM, respectively. These reductions were all significant (P < 0.001). Another finding was that the synergy of the effects of DHA was AVA dose dependent. The reductions in DHA IC₅₀ were also significant (P < 0.001) between 0.1 and 0.5 μM AVA and between 0.5 and 1.0 μM AVA.

TABLE 1.

In vitro susceptibilities of 13 P. falciparum strains to AVA alone, DHA alone, and combinations of DHA and different concentrations of AVA

Strain or parameter	IC50
	AVA alone (μM)	DHA alone (nM)	DHA + 0.1 μM AVA (nM)	DHA + 0.5 μM AVA (nM)	DHA + 1.0 μM AVA (nM)
3D7	9.5	2.50	1.97	1.55	1.42
PA	8.1	1.30	1.25	1.10	1.04
FCR3	8.9	1.62	1.23	1.13	0.97
FCM29	6.7	1.15	1.05	0.89	0.74
W2	5.5	2.83	2.75	2.52	2.38
IMT K2	9.4	1.05	0.95	0.88	0.78
IMT K14	7.5	2.07	1.65	1.58	1.41
IMT L1	8.2	0.92	0.92	0.85	0.78
IMT Vol	8.4	1.10	1.07	0.90	0.74
IMT 31	6.1	1.70	1.46	1.36	1.21
IMT 9881	7.4	3.45	3.31	3.21	3.07
IMT 10500	6.5	3.33	3.26	2.98	2.53
IMT 16332	6.0	0.94	0.82	0.85	0.65
Geometric mean IC50 (95% CI)	7.45 (6.74-8.22)	1.65 (1.24-2.21)	1.49 (1.11-1.99)	1.35 (1.01-1.81)	1.19 (0.87-1.63)
Avg % IC50 diminution (95%CI)			10 (5-15)	18 (14-22)	28 (24-32)
P value (vs DHA alone)			<0.001	<0.001	<0.001

FIG. 1.

In vitro synergy of AVA with DHA against 13 strains of P. falciparum.

In contrast to previously published studies on 3D7 (13), we showed that AVA had a synergistic effect on DHA activity. These controversial findings could be explained by differences in the isotopic method used; Wrong and Davis used a modified microdilution isotopic method with 50% plasma (versus 10% in commonly used in vitro test). In addition, their IC₅₀ of DHA alone was 12.5 nM, while the commonly accepted IC₅₀ on 3D7 was <4 nM. We also illustrated that this synergy occurred at AVA plasma concentrations achievable in patients taking 80 mg of AVA daily (0.1 to 0.5 μM) (2). Furthermore, the doses of AVA administered to humans could be increased to 120 mg daily with only limited additional side effects (9). A dose of 120 mg of AVA increased the maximal plasma concentration (C_max) after oral administration by 4- to 10-fold. Cerebral malaria has pathophysiological features in common with sepsis, especially with regard to the pathology of the endothelium (4), and critically ill patients with sepsis had a significantly higher C_max than healthy volunteers (110.5 versus 5.9 ng/ml) (7).

Artemisinin-based combination therapy is beginning to fail in southern Cambodia and Thailand, particularly the mefloquine-artesunate drug combination (3, 11, 12). AVA could improve the activity of artemisinin derivatives.

All of these observations support calls for both an in vivo evaluation with a pharmacokinetic component and clinical trials of AVA as an antimalarial therapy.