The world's scientific and social network for malaria professionals
9165 malaria professionals are enjoying the free benefits of MalariaWorld today

Chronic toxicity of artesunate and artemether

July 28, 2014 - 10:10 -- Pierre Lutgen

This is an addition to the document posted on in September 2013

A very recent toxicity study (Yin J-y, Wang H-m, Wang Q-j, Dong Y-s, Han G, et al. 2014. Subchronic Toxicological Study of two Artemisinin Derivatives in Dogs. PLoS ONE 9.4: e94034) should have been done 30 years ago ! Frightening ! Not necessarily for the majority of Africans who don’t have the money to buy several ACT treatments per year. But if an European working in Africa decides to take regular doses of ACT (Coartem or Coarsucam) to prevent or cure suspected malaria attacks he may end up with serious damage in his brain, in his liver, his blood and so on.. This paper uses relatively low doses of 6mg/kg/day for 3 months, inferior to the doses prescribed by WHO. The WHO/MAL/98.1086 document has never been cancelled « The recommended dose of ART is 20 mg/kg as loading dose and 10 mg/kg for the following days »

Of course we don’t have clinical data studying chronic toxicity for regular consumption of Artemisia annua as WHO strictly forbids clinical trials with the plant For Artemisia afra there is however the thesis of J T Mukinda of Western Cape University. Doses of 100 mg/kg of the aqueous extract (20 times more than the doses in the Chinese paper) administered for 92 days show no symptoms of toxicity, no effect vs. control on body mass, growth, food intake, RBC, hematocrit, platelet count, hemoglobin, WBC.


Haemolytic, hepatotoxic, cytotoxic, neurotoxic, cardiotoxic, genotoxic, ototoxic, embryotoxic, spleenotoxic, hemolytic, atherosclerosis, immunodepressive effects of chemical artemisinine derivatives (ACTs) at high doses.

Sept 2013


Some recent research, mostly in relation with the resistance to ACT pills and/or artesunate injections, has highlighted serious secundary health effects at the doses prescribed by WHO.

Fears of emerging artemisin resistance in western Cambodia have promoted a series of clinical trials investigating if resistance can be overcome by increasing doses of drug. After 3 to 5 doses neutrophil counts were reduced in all groups and several patients required artesunate to be discontinued because of neutropenia. This study demonstrates that the dosing limit may have been reached (D Bethell et al., Clin Infect Dis, Nov 2010). A high risk of neutropenia in HIV-infected children following treatment with artesunate was also noticed in Uganda (AF Gasasira et al., Clin Infect Dis 1. 2008, 992-993). US patent 5578637 states that artesunate stimulates the immune system at low doses and inhibits it at high doses.

Higher doses are also used in fixed formulations because in climatic zones III and IV artemisinin and derivatives are unstable and it is important that the amount of active drug present does not fall below the designated effective therapeutic dose before the expiry date (RK Haynes et al., ChemMedChem, 2007,2, 14448-1463).

One of the side effects of the higher doses is the dormancy effect induced in plasmodium. The parasite encapsulates itself against the aggressive peroxide artesunate and reawakens at the end of the treatment. It has recently been confirmed in vivo ( AN LaCrue et al., PLoS ONE 6(10) e26689). This may be one of the causes of resistance .This dormancy effect is similar, but not related to the sequestration and cytoadherence of falciparum parasitized erythrocytes. Sequestration of these erythrocytes in capillaries or placenta may protect the parasite from destruction (D Roberts et al. Nature 318, 1985, 64-66).

This massive administration of high doses of artemisinin derivatives is in contradiction with the dose of 34 mg for a human recommended for the new ozonide antimalarial OZ439 (JJ Moehrle et al., Br J Clin Pharmacol, 75:2, Jan 2012). The artemisinin dose recommended by WHO/MAL/98-1086 on the first days of treatment is 1 400 mg. But the doom of OZ439 is probably going to be the same as for all monotherapies: resistance. In Nigeria (O Omotuyi et al, Afr J Biochem Res, 2, 2008, 107-110) in a study on rats it was found that hepatoxicity and hemolysis were clearly associated with artesunate. The hepatoxicity effect was monitored in the rats as a function of aspartase transaminase (ASAT), alanine transaminase (ASAT) which both increased drastically. The hemolytic effect was monitored by the packed cell volume, bilirubin, hematocrit and serum albumin. Artemisinin significantly increases ALAT in rats (A Udobre et al. Asian J Biochem 4.2, 2009, 55-59). Acute hepototoxicty following administration of artesunate in guinea pigs was also noticed in other trials in Nigeria ( HU Nwanjo et al., The International J of Toxicol., 4, 2007, 1-5). In humans hepatotoxicity can be particularly severe if artesunate is used in combination with HIV antiretroviral drugs ( P.German et al., Clin Inf Diseases44, 2007, 889-891) The document WHOPAR 06/2011 MA058 recognizes that artesunate/amodiaquine 100/270 mg tablets may result in severe hepatotoxicity.

The histological effect of oral administration of artesunate on the liver in wistar rats confirmed a toxic effect on liver cells and mild inflammation of the portal tracts. The generation of ROS leads to the accumulation of lipid peroxides leading to change in permeability of the cells. The authors conclude that self-medication with artesunate should be discouraged. (AM Izunya et al., Res J Appl Sci eng and Technology, 2-4, 2010, 314-318). Another recent study from Nigeria (AO Abolaji et al., Hum Exp Toxicol. Nov 2012) shows that the co-administration of artemether and lumefantrine causes significant increase in glutathione peroxidase and malondialdehyde levels and may thus increase the risk of atherosclerosis. Artemether can aggravate anemia when administered to malaria patients ( Osonuga I et al., Asian Pac J Trop Biomed, 2012, 2-6, 493-5). The liver is considerably more affected than the kidney by administration of artemether or Coartem ( Adarmoye OA et al. Basic Clin Pharmacol Toxicol. 2008 102:4, 412-8). A Brazilian study showed that liver cells treated with artemisinin and artesunate showed a significant dose-dependent increase in the cells with DNA damage at all concentrations tested ( I Aquino et al., Genet Mol Res, 2013, 12-3, 2517-27).

Recently a strong spleenotoxic effect was discovered (OT Soniran et al., Malaria Research and Treatment, 2012). Proliferation of megakaryoblasts was induced by artesunate in the spleen of Albino mice, but not by chloroquine. The authors claim that self-medication with artesunate should be prohibited. Spleen enlargement due to artemether is known since decades ( PY LIN et al., Acta Pharmaceutica Sinica, 1985-03). High doses of artemether cause progressive degeneration of the renal tissue in rats (RO Akomolafe et al., Afric J Biotechnology, 10:20, 4226-33, 2011).

A study from Nigeria shows that artesunate dramatically increases ALAT and ASAT values, indicative of serious liver toxicity ( A Udobre et al., Asian Journal of Biochemistry, 4, 2009, 55-59). Another Nigerian study (JS Aprioku et al., J Appl Pharm Sc 2-10, 054-59, 2012) shows that artemether-lumefantrine increases urea by 53.2 % and may cause toxicity to renal and reproductive functions.

An electrocardiographic study performed on Ghanaian children in 2012, 5 years after the introduction of ASAQ, showed that artesunate-amodiaquine showed a high incidence of brachycardia( G Adjei a et al., Malaria Journal, 2012, 11:420.). Another study in Nigeria on rats ( CA Ouechere et al., Malaria Research and Treatment, 2012, ID 257986) showed that ASAQ afforded a 27.2% heart weight decrease when compared with control. The authors also state that caution is required and that chronic doses may predispose to renal oxidative stress. Arteether has never been introduced mainly because of cardiologic concerns (Andrea Bosman CDS/RBM, 18 March 2002)

Derivatives of artemisinin exhibit potent immunosuppressive activity (JX Wang et al., Br J Pharmacol150, 2007, 652-661). Artesunate concentrations between 0.1-1.5 microg/ml reduced lymphocyte production in a generally dose dependant manner ( P Veerusubramanian et al., Southeast Asian J Trop Med , 37, 2006, 838-847). The authors claim that further work is warranted to define the mechanisms involved and wether this affects malaria treatment. Another research work confirmed the genotoxic and cytotoxic effect of artesunate in cultured human lymphocytes ( TC Mota et al., Env Mol Mutagen 52, 2011, 590-594.) But the mechanism of this inhibitory effect on lymphoproliferation is unknown. TNF-α levels were also reduced in patients treated with artesunate (W Ittarat et al., Southeast Asian J Trop Med, 30, 1999, 7-10) which obviously has an effect on the inflammatory process. In normal mice artemisinin and artether exhibit a pronounced inhibition of the humoral reponse( AF Tawfik et al., Int J Immunopharmacol 12, 1990, 385-389). Artesunate suppressed phagocytosis of peritoneal macrophage in mice ( PY LIN et al, Acta Pharmacol Sin. 16, 1995, 441-444).

The majority of these studies on immunity also find a reduction of CD4+ and CD8+ T cells which could be detrimental in HIV infections. Several recent papers (JK Kirinyet et al., Intern.J. Adv Res. 2013, 1:3, 140-149, SN Shah et al., HIV infection and Antimalarial Treatment, 2006, 194, 1519-28) have shown that in patients with low CD4 count it was impossible to lower the high plasmodium load to zero, even with the strongest antimalarial drug known: dihydroartemisinin (YM Tatfeng et al., J Vect Borne Dis , 44. June 2007, 111-115). After a malaria attack the CD4 count remains low for several months in persons which had a low CD4 baseline at the moment of the infection.

The haemolytic effect is well described in the following paper ( A Corpolongo et al., Malaria Journal, 11-91, 2012) The patient described in their article presented with fever, headache after returning from Central African Republic. He had been treated with oral artemether and lumefantrine. The blood analysis revealed acute renal failure Hereditary or auto-immune disorders were excluded .There are a few other cases described in the same paper of haemolytic anaemia during or after treatment with i.v. artesunate alone or combined with mefloquine: the case of a Nigerian male with severe P. falciparum malaria initially treated with mefloquine. After one day of treatment, because of the worsening clinical condition of the patient and the increase of the parasitaemia, therapy with i.v. artesunate was initiated. Parasite clearance was obtained within 20 hours after the first administration of artesunate, but fever persisted for a further seven days and haemolytic anaemia was observed, requiring blood transfusion; the case of a young woman with P. falciparum malaria who was successfully treated with i.v. artesunate, but showed worsening anaemia after artesunate administration ; severe haemolytic anaemia and jaundice on day 11 after i.v. artesunate administration in a 68-year-old Japanese woman affected by severe malaria; a series of 25 travellers with severe malaria treated with i.v. artesunate. Among them, six patients developed haemolytic anaemia week after treatment possibily related to artesunate or had persistent signs of haemolytic activity until six weeks after the first dose of i.v. artesunate. In this case series, patients with post-treatment haemolysis had received a higher cumulative dose of i.v. artesunate and were treated for longer periods. This appears to be the case also for several patients treated at the University Medical Center at Hamburg. After having received 3-4 intravenous doses of 150 mg artesunate they suffered delayed haemolysis , on the average 14 days later. Further studies are necessary to assess the pathophysiological background of this complication (Malaria Journal 2012, 11:29). Most antimalarials have a hemolytic effect. The strongest are known for quinine, chloroquine and primaquine. At high doses Fansidar® and Coartem® cause significant hemolysis ( F C Anaba et al., Pak J Pharm Sci 25:4 2012, 851-55). A recent study in Nigeria ( PC Chikezie et al., Iranian J of Blood and Cancer, 2009, 4, 151-157) compares the effect of five antimalarial drugs on methemoglobin concentrations in erythrocytes. The five drugs showed a concentration dependent elevation of plasma methemoglobin. Coartem® showed the highest propensity. Another study by the same authors (Afr J Biochem Res. 4:3 57-64, 2010) ascertained the osmotic fragility of erythrocytes after administration of the 5 same antimalarial drugs (Fansidar-SP, Halofantrine, Quinine, Coartem). All caused a rapid accumulation of ROS overwhelming the antioxidant defense capacities of the erythrocytes and caused depletion of erythrocyte glutathione concentration. Again the most aggressive effect on the average was noticed for Coartem®. Dihydroartemisinin is known to cause embryonic erythrocyte depletion and a significant delay in erythroid differentiation ( S Finaurini et al., Toxicology 276, 2010, 128-134).

The derivatives of artemisinin are strong peroxides. They create radicals and reactive oxygen species which at high doses might damage cortical and brain stem neurons (G Schmuck et al., Antimicrob Agents and Chemother. 46, 2002, 821-827) and are cytotoxic to many other cells. The genotoxicity of artesunate was studied in Brazil on mice ( I Aquino et al., Food Chem Toxicol. 2011 Jun;49(6):1335-9.) The artesunate was administered by oral gavage at doses of 5, 50 and 100 mg/kg. Artesunate showed weak genotoxic effects at low doses and severe (clastogenic effects) at high doses. The authors claim that the data obtained suggest caution about either continuous or high-dose use of artesunate by humans. Similar results were obtained by another research team in Brazil. Their results showed that artesunate is a genotoxic and cytotoxic compound in cultured human lymphocytes. A significant increase in the frequency of apoptotic and necrotic cells was observed. (T Mota et al., Environ. Mol. Mutagen., 2011).

A study from Iran shows that pure artemisinin leads to cell injury in MDCK kidney cells in concentration dependent manner ( AA Shahbazfar et al., Interdisc Toxicol 2012, 5, 30-37).

There is also increasing evidence that the oxidative stress induced by artemisinin derivatives plays a major role in the pathogenesis of diabetes (C F M Lima, Tese de Doutoramento, Universidade do Minho, Nov 2006). This in turn has an important role in causing the secondary complications of diabetes, such as atherosclorosis, nephropathy, retinopathy and neuropathy. Artemisinin reduces intracellular ATP levels and the potential of the inner mitochondrial membrane below the cytotoxic concentration range, with these effects being most dominant in the brain stem cultures. Surprisingly, there were substantial effects on cortical neurons after 7 days and on astrocytes after 1 day. Toxic brainstem encephalopathy was noticed after artemisinin treatment for breast cancer. (Panossian LA,. Et al., Ann Neurol. 2005 Nov;58(5):812-3.) Artesunate and artemether rapidly degrade into dihydroartemisinin which is claimed by pharmaceutical companies to be 10 times more effective against malaria than artemisinin. Artemisinin itself does not metabolize into dihydroartemisinin and stays much longer in the plasma. This might be one of the reasons why Artemisia annua tea infusions, despite the fact that they contain much less active molecules than ACTs, might be as effective. A very recent paper from Nigeria confirms all the concerns listed before. (ET Olayinka et al., IOSR J of Pharm and Biol Sc Mar-April 2013) :” Our results suggest that the therapeutic doses of P-ALAXIN (DHA-piperaquine) induced marked renal and hepatic failure and decreased the levels of antioxidant defese systems, with the effect more pronounced in double therapeutic doses. As such caution should be taken in administering the drug…”

All these findings are especially worrisome for the use of higher doses of artesunate required for pregnant women because immunity is temporarely disrupted in pregnancy. Non-immune or immunodepressed patients require higher doses for an adequate therapeutic response. Artesunate also causes embryo death in animals by causing severe anemia with higher drug concentrations ( RL Clark et al., Birth Defect Res B Dev Reproduc Toxicol 92, 2011, 52-68). Artesunate increased embryolethality and the incidence of limb long bone malformations on the absence of overt maternal toxicity (AC Boareto et al., Reprod toxicol Oct 2012). Embryotoxicity is due to dihydroartemisinin, the main metabolite of artesunate and artemether, but not of artemisinin ( S D’Alessandro et al., Toxicology 241, 2007, 66-74). In cancer treatment doses of artemisinin 10-20 times higher than for malaria are required. It was found (S.D’Alessandro et al., Biochem Pharmacol. 82, 2011, 466) that these lead to inhibition and apoptosis of endothelial cells and lipid peroxidation. A recent study from Mali ( I Sagara et al., AM J Trop Med Hyg 87, 2012, 50-56) shows that all ACT treatments, but mainly Coartem lead to creatinemia (high creatinine). Inflammatory effects have been several times reported for ACTs. In Ethiopia ( A Assefa et al., Parasites & Vectors , 2010, 3:1) a study revealed the development of oral inflammation (oral ulcer) in 7.2% of the patients treated with Coartem. Clinicians in the area had reported the same side effect in the past. In Mozambique audiometric changes were noticed after the treatment with co-artemether. A significantly greater hearing loss than for controls was noticed on 150 subjects receiving artemether-lumefantrine treatment of uncomplicated falciparum malaria. (S Toovey et al., Trans R Soc Trop Med Hyg, 2004, 98:5, 261-7). On the basis of these otometric results the authors recommend that the neurotoxicity of artemisinins be more fully evaluated. Worrisome is also the finding of research in Nigeria which indicates that high doses of artesunate may cause inflammation of the testicles and aggravate male infertility. ( AM Izunya et al., Biology and Medicine, Vol 2(2), 49-56, 2010). A previous study in Nigeria had shown that artesunate decreases the testosterone level by 37%., even at subclinical and clinical doses, an alarming situation ( AW Obianime et al., Nigerian Journal of Physiological Sciences, 24-2, 2009, 101-106). This effect on reproductive parameters has also been described in guinea pigs (AW Obianime et al., Int J Pharmacology, 7, 2011, 84-95). The effect on sperm count and androgenic deficiency has also been described ( H Nwanjo et al., Internet J of Endocronoloy, 4:1, 2007). It was also found that long-term administration of artesunate could induce reversible infertility in rats which may act via distortion of blood-testis barrier formed by Sertoli cells ( SA Oumide et al., J Reproduc Infertil., 2011, 12-4, 249-60)

Caution is also required during co-administration of artemether/lumefantrine with lopinavir/rotinavir (anti HIV drugs) where severe interactions have been noticed ( P Byakika-Kibwika et al., J Antimicrob Chemother 67, 2012, 1217-1223). The clinical significance of this needs to be evaluated. Artemisinin induces CYP3A4 and this might influence the efficiency of other drugs like lumefantrine or other molecules which are metabolized by this cytochrome. Artemisin induces resistance against the anti-cancer drug doxorubicin ( C Riganti et al., Br J Pharmacol., 2009, 156:7, 1054-66). Coadministration of artemether and lumefantrine may increase the risks of atherosclerosis as well as liver and renal function impairments in the users. The drugs also promote oxidative stress in the erythrocytes A O Abolaji et al., Human & Exp Toxicol 2013, 32:2, 206-215).

A recent study from UK on the risk of developing eye disorders after the use of antimalarial drugs by travellers provides evidence that there was an increased risk compared to non-users. (C Schneider et al., Travel Medicine and Infectious disease, 2014, 12, 40-47)

But none of these hepatotoxic, haemolytic, cytotoxic, immunodepressive, cardiotoxic , atheriosclerosic effects has ever been observed by drinking Artemisia annua tea.

Several studies ( A Muzemil, PhD thesis, AddisAbeba 2009; JT Mukinda, Thesis, University of Western Cape, 2005; G Chuipet, Thesis, Université des Montagnes, 2012, University of Calabar, Phytotherapy Research, iSSN 1099-1573, Aug 2013)) have confirmed that Artemisia annua tea is completely innocuous up to 5000 mg/kg of dried plant extracts and although it administers doses of artemisinin 100 x lower than those of the ACT pills, it cures >95% of the malaria infections.

Many studies have shown that Artemisia annua stimulates the immune system, increasing for example the monocyte count. (PE Ogwang et al., Brit J Phar Res., 1, 2011, 124-132) This effect is probably due to other constituents than artemisinin: essential oils, flavonoids, coumarins, polysaccharides, saponins. The research work from the University des Montagnes in Cameroon even indicates that Artemisia annua tea lowers the alanine aminotransferase (ALAT) and could be hepatoprotective.


Submitted by Mary Vanderkooi M.D. (not verified) on

In August of '08 I was diagnosed with Stage 3C endometrial cancer. I had radiation and chemo, finishing in April of '09 with some improvement, then deterioration. By June '09 I was almost bedridden, getting weaker by the day. Mid-June I started taking Artemisia annua tea (leaves and all), 5 grams a day. In two weeks I was putting in 4 hour days. My sed rate dropped from 35 to 2. After 3 months I travelled alone to London, walked all over central London for a week and returned with a full luggage allowance. By Jan '10 I began to be ataxic. In April '10 I had a repeat CA-125 which was 14, so I stopped the Aa with immediate, marked improvement in the ataxia. My CA-125 has been normal since, and I'm in good health. (Post-surgical it was 475.)

Submitted by David Warhurst on

Way back, there was so much convincing evidence of fetotoxicity of artemisinins in animal studies (mainly with intramuscular injection)that the deployment of these agents in humans outside China was delayed.
There is now good evidence to support the use of artemisinins in the current dosing regimens (Efferth & Kaina 2010) but the crucial part of their summary is even handed with regard to toxicity. "The lesson learned from animal and human studies is that long term availability rather than short term peak concentrations of artemisins cause toxicity. Rapid elimination of artemisinins after oral intake represents a relatively safe route of administration compared to delayed release after intramuscular addition there are drug-related toxicities .....Although there is no need to increase dose of artemisinins for uncomplicated malaria, this has to be taken into account for cerebellar involvement in severe malaria.
Clark 2009 has a relatively optimistic review: "Single oral doses of artesunate, dihydroartemisinin, arteether and artemether administered to rats during
a sensitive period of organogenesis caused embryo deaths and malformations (malformed long bones and ventricular septal defects). Extended oral dosing (12 days or more) of monkeys once daily with 12 mg/kg-d artesunate also caused embryo deaths. The initial embryotoxic effect in both species was to kill primitive erythroblasts which are present in the embryo for a few days of gestation in rats and several weeks in
primates. The malformations that occurred in rats are attributed to a transient depletion of the primitive erythroblasts. In monkeys, when treatment at 12 mg/kg-d was shortened to 3 or 7 days, the embryos survived but likely suffered a transient loss of primitive erythroblasts. Limited clinical data including 123 first trimester pregnancies have not indicated any adverse effects on pregnancy. However, in rats and monkeys, the embryonic erythroblasts are much more sensitive to artemisinins than are erythroblasts in the adult bone marrow."
However these worries have not been supported in general use, though increases in exposure associated with more persistent new drugs or any drastic increase in dosing frequency or duration may have uncertain effects. Malaria risk of death to the mother is (probably sensibly) taken more seriously than risk of fetal loss.