Chapter 49: Chemotherapy of Malaria

Mechanisms of erythrocyte invasion include initial binding by merozoites to specific red blood cell surface ligands. P. falciparum has a family of binding proteins that can recognize a variety of host cell molecules. P. falciparum invades all stages of erythrocytes and therefore can achieve high parasitemias. P. vivax selectively binds to the Duffy chemokine receptor protein as well as reticulocyte-specific proteins. Thus, P. vivax does not establish infection in Duffy-negative individuals and only invades reticulocytes. (However, P. vivax has reportedly mutated in Madagascar to enable infection of Duffy-negative individuals [Ménard et al., 2010].) Because of this restricted subpopulation of suitable erythrocytes, P. vivax rarely exceeds 1% parasitemia in the bloodstream. P. ovale is similar to P. vivax in its predilection for young red blood cells, but the mechanism of its erythrocyte recognition is unknown. P. malariae parasitizes senescent red blood cells, maintains a very low parasitemia, and typically causes an indolent infection.

P. falciparum assembles cytoadherence proteins (PfEMP1s, encoded by a highly variable family of var genes) into structures called knobs that are presented on the erythrocyte surface. Knobs allow the P. falciparum-parasitized erythrocyte to bind to postcapillary vascular endothelium, so as to avoid spleen-mediated clearance and allow the parasite to grow in a low oxygen, high carbon dioxide microenvironment. For the patient, the consequences are microvascular blockage in the brain and organ beds, and local release of cytokines and direct vascular intermediates such as nitric oxide. These lead to severe complications such as cerebral malaria, pulmonary edema, acute renal failure, and placental malaria. These can result in low birthweight and translocation of bacteria from the gastrointestinal (GI) tract to the blood (septic complications, so-called algid malaria).

Clinical Manifestations of Malaria.

The cardinal signs and symptoms of malaria are high, spiking fevers (with or without periodicity), chills, headaches, myalgias, malaise, and GI symptoms. Severe headache, a characteristic early symptom in malaria caused by all Plasmodium spp., often heralds the onset of infection, before fever and chills. P. falciparum causes the most severe disease and may lead to organ failure and death. Placental malaria, of particular danger for primigravidae, is due to P. falciparum adherence to chondroitin sulfate A (CSA) in the placenta. This often leads to severe complications, including miscarriage. When treated early, symptoms of malarial infection usually improve within 24-48 hours.

Acute illness due to P. vivax infection may appear severe due to high fever and prostration. Indeed, the pyrogenic threshold of this parasite (i.e., parasite burden associated with fever) is lower than that of P. falciparum. Nonetheless, P. vivax malaria generally has a low mortality rate. P. vivax malaria is characterized by relapses caused by the reactivation of latent tissue forms. Clinical manifestations of relapse are the same as those of primary infection. In recent years, severe P. vivax malaria from Oceania (Papua New Guinea, Indonesia) and India possess important similarities to severe malaria caused by P. falciparum. These include neurological symptoms (diminished consciousness, seizure) and pulmonary edema. Rare but life-threatening complications can occur, including splenic rupture, acute lung injury, and profound anemia.

P. ovale causes a clinical syndrome similar to that of P. vivax but may be milder with lower levels of parasitemia. It shares with P. vivax the ability to form the hypnozoite (dormant liver stage) that may relapse after months to 2 years later. P. ovale is more common in sub-Saharan Africa and some islands in Oceania.

P. malariae causes a generally indolent infection with very low levels of parasitemia and often does not produce clinical symptoms. This parasite can be found in all malaria-endemic areas but is most common in sub-Saharan Africa and the southwest Pacific. Interestingly, P. malariae prevalence increases during the dry season and can be found as a co-infection with P. falciparum. Although uncommon, a potentially fatal complication of P. malariae is a glomerulonephritis syndrome that does not respond to antimalarial treatment.

P. knowlesi infection is often misdiagnosed as P. malariae by light microscopy. This infection is distinguished by a shorter erythrocytic cycle (24 hours compared with 72 hours for P. malariae) and higher levels of parasitemia. Like P. malariae, P. knowlesi is generally sensitive to chloroquine, but patients presenting with advanced disease nonetheless may progress to death despite adequate drug dosing.

Asymptomatic P. falciparum and P. vivax infections are common in endemic regions, and they represent important potential reservoirs for malaria transmission. Although different studies are not entirely consistent in the definition of "asymptomatic," generally this state implies a lack of fever, headache, and other systemic complaints, within a defined time period prior to a positive test for malaria parasitemia. Migration of asymptomatic individuals, to areas where malaria is not present but vector mosquitoes are (i.e., anophelism without malaria), is an important mechanism for the introduction or reintroduction of malaria, in addition to facilitating the spread of drug-resistant isolates. Novel approaches to preventing transmission from asymptomatic reservoirs—whether through new drugs or vaccines—will be essential for future malaria control, elimination, and eradication strategies.

History. Artemisinin is a sesquiterpene lactone endoperoxide derived from qing hao (Artemisia annua), also called sweet wormwood or annual wormwood. The Chinese have ascribed medicinal value to this plant for >2000 years (White, 2008). As early as 340 A.D., Ge Hong prescribed tea made from qing hao as a remedy for fevers, and in 1596, Li Shizhen recommended it to relieve malarial symptoms. By 1972, Chinese scientists had identified the major antimalarial ingredient, qinghaosu, now known as artemisinin.

Chemistry. The structures of artemisinin and its three major semisynthetic derivatives in clinical use, dihydroartemisinin, artemether, and artesunate, are as follows:

The derivatives display improved potency and bioavailability and have largely replaced the use of artemisinin. Dihydroartemisinin is a reduced product, artesunate is the water-soluble hemisuccinate ester of dihydroartemisinin, and artemether is a lipophilic methyl ether. Extensive structure-activity studies have confirmed the requirement for an endoperoxide moiety for antimalarial activity.

The mechanisms by which ACTs exert their antimalarial activity remain contentious (Golenser et al., 2006). Nevertheless, most studies concur that the activity of artemisinin and its potent derivatives results from reductive scission of the peroxide bridge by reduced heme-iron, which is produced inside the highly acidic digestive vacuole (DV) of the parasite as it digests hemoglobin. In addition to the formation of potentially toxic heme-adducts, activated artemisinin (for which the site of action remains unclear) might in turn generate free radicals that alkylate and oxidize proteins and possibly lipids in parasitized erythrocytes (Eastman and Fidock, 2009).

Artemisinins do not display significant clinical cross-resistance with other drugs. Indeed, sensitivity to artemisinins may even be increased in at least some strains of chloroquine-resistant parasites. Recent evidence has nonetheless suggested the emergence of P. falciparum isolates with an increased tolerance to artemisinins, manifesting as longer parasite clearance times (Dondorp et al., 2009). This has triggered significant efforts to elucidate the mechanistic basis of resistance and implement means to limit its spread. The World Health Organization (WHO) and most authorities strenuously recommend using artemisinins only in combination therapy, both to increase treatment efficacy and prevent the emergence of drug resistance.

Absorption, Fate, and Excretion. The semisynthetic artemisinins have been formulated for oral (dihydroartemisinin, artesunate, and artemether), intramuscular (artesunate and artemether), intravenous (artesunate), and rectal (artesunate) routes. Bioavailability after oral dosing typically is ≤30%. Although artemisinins rapidly achieve peak serum levels; intramuscular administration of the lipid-soluble artemether peaks in 2-6 hours, due to a depot effect at the injection site. Both artesunate and artemether have modest levels of plasma protein binding, ranging from 43% to 82%. These derivatives are extensively metabolized and converted to dihydroartemisinin, which has a plasma t_1/2 of 1-2 hours (German and Aweeka, 2008). Rectal administration of artesunate has emerged as an important administration route, especially in tropical countries where it can be lifesaving. However, drug bioavailability via rectal administration is highly variable among individual patients (Medhi et al., 2009).

With repeated dosing, artemisinin and artesunate induce their own CYP-mediated metabolism, primarily via CYPs 2B6 and 3A4. This may enhance clearance by up to 5-fold. Recent studies have found no clinically significant pharmacokinetic or toxic interactions between artemisinins and its partner drugs.

Toxicity and Contraindications. The increasing usage of ACTs has focused attention on the safety profile of artemisinins, especially with regard to potential toxic effects in infants and during the first trimester of pregnancy. In pregnant rats and rabbits, artemisinins can cause increased embryo lethality or malformations early postconception. Preclinical toxicity studies have identified the brain (and brainstem), liver, bone marrow, and fetus as the principal target organs. In patients, the many neurological changes that accompany severe malaria confound the evaluation of drug neurotoxicity; however, no systematic neurological changes were attributable to treatment in patients >5 years of age. As in small animals, patients may develop dose-related and reversible decreases in reticulocyte and neutrophil counts and increases in transaminase levels in patients. About 1 in 3000 patients develops an allergic reaction. Studies of artemisinin treatment during the first trimester have found no evidence of adverse effects on fetal development (Eastman and Fidock, 2009). Nonetheless, out of general concern, it is recommended that ACTs not be used for the treatment of children ≤5 kg or during the first trimester of pregnancy.

This combination has proven to be highly effective for the treatment of uncomplicated malaria and is the most widely used first-line antimalarial across Africa. The pharmacokinetic properties of lumefantrine include a large apparent volume of distribution and a terminal elimination t_1/2 of 4-5 days. Human pharmacokinetic and pharmacodynamic studies correlate the risk of clinical failure (the likelihood to recrudesce) with plasma lumefantrine concentrations falling below 280 ng/mL. These findings, combined with a report of up to 15-fold variability in lumefantrine plasma concentrations in clinical trial volunteers, highlight the importance of appropriate dosing with lumefantrine-containing combinations (Checchi et al., 2006). Administration with a high-fat meal is recommended because it significantly increases absorption. Recently a sweetened dispersible formulation of artemether-lumefantrine (coartem Dispersible) has been approved for treatment of children. This formulation has comparable pharmacokinetics to crushed tablets and provides substantially improved ease of administration.

In vivo, amodiaquine is rapidly converted by hepatic CYPs into monodesethyl-amodiaquine. This metabolite, which retains substantial antimalarial activity, has a plasma t_1/2 of 9-18 days and reaches a peak concentration of ~500 nM 2 hours after oral administration. By contrast, amodiaquine has a t_1/2 of ~3 hours, attaining a peak concentration of ~25 nM within 30 minutes of oral administration (Eastman and Fidock, 2009). In vivo clearance rates of amodiaquine, however, display a variation between individuals that ranges from 78 to 943 mL/min/kg.

Piperaquine has a large volume of distribution and reduced rates of excretion after multiple doses. This lipophilic drug is rapidly absorbed, with a T_max (time to reach the highest concentration) of 2 hours after a single dose. In clinical trials, the combination of piperaquine and dihydroartemisinin produced cure rates and cleared fever and parasites in a time frame similar to artesunate-mefloquine (Wells et al., 2009). Piperaquine has the longest plasma t_1/2 (5 weeks) of all ACT partner drugs, suggesting that piperaquine-dihydroartemisinin might also be effective in reducing rates of reinfection following treatment.

Pyronaridine is well tolerated and highly potent against both P. falciparum and P. vivax, causing fever to subside in 1-2 days and parasite clearance in 2-3 days. Clinical data from trials of artesunate-pyronaridine should soon be available (Wells et al., 2009).

Resistance to atovaquone alone in P. falciparum develops easily in vitro and in vivo, and it is conferred by single non-synonymous nucleotide polymorphisms in the cytochrome b gene located in the mitochondrial genome (Kessl et al., 2007). In the Saccharomyces cerevisiae cytochrome bc₁ complex, atovaquone binding is inhibited by the introduction of resistance mutations found in P. falciparum. Similar mutations have been reported in resistant isolates of rodent malarial species, as well as in T. gondii and perhaps P. jiroveci. Addition of proguanil markedly reduces the frequency of appearance of atovaquone resistance, as based on in vivo treatment cure rate. However, once atovaquone resistance is present, the synergy of the partner drug proguanil diminishes. In the key mutation associated with clinical atovaquone-proguanil resistance, tyrosine is replaced by serine, cysteine, or asparagine at codon 268 (Y268S/C/N) of the cytochrome b gene. Reports of atovaquone-proguanil treatment failures are rare. However, when atovaquone-proguanil treatment fails because of parasite resistance, the recurrent malarial attack commonly occurs ~20-30 days after treatment initiation. Treatment with different drugs should then be started. Earlier treatment failures (during the first 2 weeks after treatment initiation) are more likely related to lack of compliance or inadequate drug levels (Musset et al., 2006; Rose et al., 2008).

Absorption, Fate, and Excretion. Atovaquone absorption after a single oral dose is slow, erratic, and variable due to its highly lipophilic nature and low aqueous solubility. Absorption improves when the drug is taken with a fatty meal (Boggild et al., 2007). More than 99% of the drug is bound to plasma protein, and cerebrospinal fluid levels are <1% of those in plasma. Profiles of drug concentration versus time often show a double peak, albeit with considerable variability. The first peak appears in 1-8 hours, whereas the second occurs 1-4 days after a single dose. This pattern suggests an enterohepatic circulation. In the absence of a CYP-inducing second medication, humans do not metabolize atovaquone significantly. The drug is excreted in bile, and >94% of the drug is recovered unchanged in feces; only traces appear in the urine. Atovaquone has a reported elimination t_1/2 from plasma of 2-3 days in adults and 1-2 days in children. Clearance is unaffected by dose or co-administration of proguanil (Boggild et al., 2007).

A tablet containing a fixed dose of 250 mg atovaquone and 100 mg proguanil hydrochloride, taken orally, is highly effective and safe in a 3-day regimen for treating mild-to-moderate attacks of chloroquine- or sulfadoxine-pyrimethamine-resistant P. falciparum malaria (Looareesuwan et al., 1999). Although the same regimen followed by a primaquine course is effective in treatment of P. vivax malaria (Lacy et al., 2002), evidence for atovaquone-proguanil efficacy as treatment of non–P. falciparum malaria is limited (Boggild et al., 2007). The CDC does not recommend the combination for non-P. falciparum treatment, except for malaria caused by chloroquine-resistant P. vivax. Atovaquone-proguanil is a standard agent for malaria chemoprophylaxis. It can be discontinued 1 week after leaving the endemic area because both components are active against liver stage parasites. Experience in prevention of non–P. falciparum malaria is also limited (Ling et al., 2002). P. vivax infection may occur after drug discontinuation, indicating imperfect activity against exo-erythrocytic stages of this parasite.

Toxicity. Atovaquone may cause side effects (abdominal pain, nausea, vomiting, diarrhea, headache, rash) that require cessation of therapy. Vomiting and diarrhea may decrease drug absorption, resulting in therapeutic failure. However, readministration of this drug within an hour of vomiting may still be effective in patients with P. falciparum malaria. Atovaquone occasionally causes transient elevations of serum transaminase or amylase.

Precautions and Contraindications. Although atovaquone is generally considered to be safe, it needs further evaluation in children <11 kg, pregnant women, and lactating mothers. Accordingly, although the drug is not formally recommended for these individuals, the risks and benefits of its use (both as chemoprophylaxis and treatment) should be considered carefully for individual patients, taking known toxicities into account. Preclinical evaluations for carcinogenicity, mutagenicity, and teratogenicity have been negative. Atovaquone may compete with certain drugs for binding to plasma proteins, and therapy with rifampin, a potent inducer of CYP-mediated drug metabolism, can reduce plasma levels of atovaquone substantially, whereas atovaquone may raise plasma levels of rifampin. Co-administration with tetracycline is associated with a 40% reduction in plasma concentration of atovaquone.

Dietary p-aminobenzoic acid or folate may affect the therapeutic response to antifolates. Both of these metabolites can be imported by the malarial parasite and can reduce drug efficacy substantially (Hyde, 2007). Resistance to pyrimethamine has developed in regions of prolonged or extensive drug use and can be attributed to mutations in dihydrofolate reductase that decrease the binding affinity of pyrimethamine (Gregson and Plowe, 2005). Interestingly, the pattern of amino acid substitutions is different in parasites resistant to cycloguanil, even though cross-resistance can occur between these structurally related inhibitors of plasmodial dihydrofolate reductase. The key mutation associated with pyrimethamine resistance is the substitution of asparagine for serine at codon 108 (S108N). The stepwise accumulation of additional single-amino-acid changes at codons 51, 59, and 164 is associated with progressively increasing resistance, as determined with P. falciparum isolates and with recombinant proteins. The triple mutant N51I/C59R/S108N (predominant in Africa) and the quadruple mutant N51I/C59R/S108N/I164L (predominant in Southeast Asia) exhibit high levels of pyrimethamine resistance and contribute to pyrimethamine-sulfadoxine therapeutic failure. Remarkably, studies exploring the mutational trajectories leading to the most resistant quadruple mutant N51I/C59R/S108N/I164L suggest a specific ordering of mutational events that balance the acquisition of resistance with the maintenance of pathogen fitness (Lozovsky et al., 2009). These studies also suggest that this quadruple mutant is substantially less fit in the absence of drugs when compared to its wild-type counterpart. Compensatory mechanisms might therefore be necessary to alleviate this fitness cost. An increase in the number of copy of the GTP cyclohydrolase I gene, which encodes the first enzyme in the Plasmodium folate biosynthesis pathway, is closely associated with the dihydrofolate reductase I164L mutation, suggesting a potential compensatory function (Nair et al., 2008).

Absorption, Fate, and Distribution. After oral administration, pyrimethamine is slowly but completely absorbed, reaching peak plasma levels in 2-6 hours. The compound is significantly distributed in the tissues and is ~90% bound to plasma proteins (German and Aweeka, 2008). Pyrimethamine is slowly eliminated from plasma with a t_1/2 of 85-100 hours. Concentrations that are suppressive for responsive Plasmodium strains remain in the blood for ~2 weeks, but levels are lower in patients with active infection. Several metabolites of pyrimethamine appear in the urine; however, their identities and antimalarial properties have not been fully characterized. Pyrimethamine also enters the milk of nursing mothers.

Sulfonamides or sulfones, rather than pyrimethamine, usually account for the toxicity associated with coadministration of these antifolate drugs. The combination of pyrimethamine and sulfadoxine is no longer recommended for malaria chemoprophylaxis because in ~1 in 5000 to 1 in 8000 individuals, this combination causes severe and even fatal cutaneous reactions, such as erythema multiforme, Stevens-Johnson syndrome, or toxic epidermal necrolysis. It has also been associated with serum sickness–type reactions, urticaria, exfoliative dermatitis, and hepatitis. Pyrimethamine-sulfadoxine is contraindicated for individuals with previous reactions to sulfonamides, for lactating mothers, and for infants <2 months of age. Administration of pyrimethamine with dapsone (maloprim), a drug combination unavailable in the U.S., has occasionally been associated with agranulocytosis.

History. Proguanil is the common name for chloroguanide, a biguanide derivative that emerged in 1945 as a product of British antimalarial research. The antimalarial activity of proguanil eventually was ascribed to cycloguanil, a cyclic triazine metabolite and selective inhibitor of the bifunctional plasmodial dihydrofolate reductase-thymidylate synthetase that is crucial for parasite de novo purine and pyrimidine synthesis. Indeed, investigation of compounds bearing a structural resemblance to cycloguanil resulted in the development of antimalarial dihydrofolate reductase inhibitors such as pyrimethamine. Accrued evidence also indicates that proguanil itself has intrinsic antimalarial activity independent of its metabolite's effect on parasite dihydrofolate reductase-thymidylate synthetase (Fidock and Wellems, 1997).

Chemistry. Proguanil and its triazine metabolite cycloguanil have the following chemical structures:

Proguanil has the widest margin of safety of a large series of antimalarial biguanide analogs examined. Dihalogen substitution in positions 3 and 4 of the benzene ring yields chlorproguanil, a more potent prodrug than proguanil. Cycloguanil, the metabolite, is structurally related to pyrimethamine.

Cycloguanil selectively inhibits the bifunctional dihydrofolate reductase–thymidylate synthetase of sensitive plasmodia, causing inhibition of DNA synthesis and depletion of folate cofactors (Kamchonwongpaisan et al., 2004). As referred to earlier, a series of amino acid changes near the dihydrofolate reductase–binding site have been identified that cause resistance to cycloguanil, pyrimethamine, or both (Gregson and Plowe, 2005). Specifically, resistance to cycloguanil (and chlorcycloguanil) can be linked to mutations leading to paired A16V/S108T substitutions in Plasmodium dihydrofolate reductase. In addition, cross-resistance between cycloguanil and pyrimethamine has been observed in vitro for both the triple (N51I/C59R/S108N) and quadruple (N51I/C59R/S108N/I164L) dihydrofolate reductase–thymidylate synthetase mutants, prevalent in Africa and South America, respectively (Peterson et al., 1990). Malarial isolates demonstrate varying degrees of cross-resistance to cycloguanil and pyrimethamine, indicating that the evolution of mutation patterns in individual organisms during treatment may be quite complex.

The presence of Plasmodium dihydrofolate reductase is not required for the intrinsic antimalarial activity of proguanil or chlorproguanil (Fidock and Wellems, 1997); however, the molecular basis for this alternative activity remains enigmatic. Proguanil as the biguanide accentuates the mitochondrial membrane-potential-collapsing action of atovaquone against P. falciparum but displays no such activity by itself (Vaidya and Mather, 2009) (see section on atovaquone). In contrast to cycloguanil, resistance to the parent drug, proguanil, either alone or in combination with atovaquone, is not well documented.

The metabolism of proguanil in mammals cosegregates with mephenytoin oxidation polymorphisms, controlled by isoforms in the CYP2C subfamily. Only ~3% of whites are deficient in this oxidation phenotype, contrasted with ~20% of Asians and Kenyans. Proguanil is oxidized to two major metabolites, cycloguanil and an inactive 4-chlorophenylbiguanide. On a 200 mg daily dosage regimen, plasma levels of cycloguanil in extensive metabolizers exceed the therapeutic range, whereas cycloguanil levels in poor metabolizers fail to reach a therapeutic level. Proguanil itself does not accumulate appreciably in tissues during long-term administration, except in red blood cells where its concentration is about three times that in plasma. The inactive metabolite 4-chlorophenyl-biguanide is not readily detected in plasma but appears in increased quantities in the urine of poor proguanil metabolizers. In humans, 40-60% of the absorbed proguanil is excreted in urine, either as the parent drug or as the active metabolite.

Toxicity and Side Effects. In chemoprophylactic doses of 200-300 mg daily, proguanil causes relatively few adverse effects, except occasional nausea and diarrhea. Large doses (≥1 g daily) may cause vomiting, abdominal pain, diarrhea, hematuria, and the transient appearance of epithelial cells and casts in the urine. Gross accidental or deliberate overdose (as much as 15 g) has failed to produce lasting sequelae. Doses as high as 700 mg twice daily have been taken for >2 weeks without serious toxicity. Proguanil is considered safe for use during pregnancy. It is remarkably safe when used in conjunction with other antimalarial drugs such as chloroquine, atovaquone, tetracyclines, and other antifolates.

History. Chloroquine (aralen) is one of a large series of 4-aminoquinolines synthesized as part of the extensive cooperative program of antimalarial research in the U.S. during World War II. Chloroquine proved most promising and was released for field trial. When hostilities ceased, it was discovered that the compound had been synthesized and studied as early as 1934 under the name of Resochin at the Bayer laboratories in Germany but had been rejected because of toxicity in avian models.

Chemistry. The d-, l-, and dl- forms of chloroquine (see structure in Figure 49–2) have equal potency against P. lophurae malaria (a duck parasite), but the d-isomer is somewhat less toxic than the l-isomer in mammals. A chlorine atom attached to position 7 of the quinoline ring confers the most potent antimalarial activity in both avian and human malarias. Research on the structure-activity relationships of chloroquine and related alkaloid compounds continues in an effort to find new effective antimalarials with improved safety profiles that can be used successfully against chloroquine- and multidrug-resistant strains of P. falciparum. One example is the synthesis of short-chain derivatives that can demonstrate full efficacy against chloroquine-resistant strains of P. falciparum. Another important example is piperaquine, a bisquinoline that was extensively used in China to treat chloroquine-resistant malaria. It is highly efficacious in combination with dihydroartemisinin (Davis et al., 2005); see earlier section on artemisinin-based combination therapies.

Hydroxychloroquine, in which one of the N-ethyl substituents of chloroquine is β-hydroxylated, is essentially equivalent to chloroquine against P. falciparum malaria. This analog is preferred over chloroquine for treatment of mild rheumatoid arthritis and lupus erythematosus because, in the high doses required, it may cause less ocular toxicity. Care should be taken when these compounds are administered to patients with known G6PD deficiency.

Resistance of erythrocytic asexual forms of P. falciparum to antimalarial quinolines, especially chloroquine, now is common in most parts of the world (Figure 49–3). Reports of chloroquine failures in the treatment of P. vivax are fairly common in Indonesia, East Timor, and Papua New Guinea, in which case alternative treatments (mefloquine, or quinine plus doxycycline or tetracycline, or atovaquone-proguanil) should be considered (Price et al., 2009). Chloroquine-resistant P. vivax has also been recently detected in Ethiopia, where P. vivax is an important cause of morbidity (Yeshiwondim et al., 2010). Chloroquine-resistant P. vivax isolates do not appear to have significant alterations in their pfcrt ortholog and may have a different resistance mechanism.

By studying the progeny of a genetic cross between a chloroquine-sensitive clone and a chloroquine-resistant clone, Wellems, Fidock, and colleagues identified a polymorphic gene (pfcrt, for P. falciparum chloroquine resistance transporter) that segregates with chloroquine resistance. Geographically distinct pfcrt alleles, possessing 4-8 point mutations compared to the invariant chloroquine-sensitive allele, were found to be highly associated with chloroquine resistance in clinical field isolates from across the malaria-endemic areas of the globe (Fidock et al., 2000). Multiple mutations are needed to confer resistance, including the K76T mutation that is ubiquitous to chloroquine-resistant strains. The pfcrt gene encodes a putative transporter that resides in the membrane of the acidic digestive vacuole, the site of hemoglobin degradation and chloroquine action. Its physiological function, however, remains unknown. Chloroquine binding and accumulation studies, with genetically modified isogenic lines differing in their pfcrt allele, suggest that mutant pfcrt confers chloroquine resistance by actively effluxing chloroquine away from its heme target in the digestive vacuole, seemingly via an energy-dependent process (Sanchez et al., 2007). Heterologous expression of codon-adjusted pfcrt alleles in Xenopus laevis oocytes has recently provided compelling evidence that mutant PfCRT can efflux chloroquine, and has found that this property was not solely dependent on the K76T mutation (Martin et al., 2009). Genomic studies suggest that pfcrt has been under intense selection pressure in recent decades, consistent with its role as the primary resistance determinant and the prevalence of chloroquine use in populations exposed to parasites (Mu et al, 2010). Studies from Malawi have found that mutant pfcrt alleles essentially disappear upon prolonged cessation of chloroquine use. This implies that mutant pfcrt alleles can impart a significant fitness cost to the organism. This cost of fitness results in attrition of mutant forms in areas where polyclonal infections are common and subjects are sufficiently immune such that a substantial proportion of infections are not subject to the pressure of drug selection.

In addition to chloroquine, variant pfcrt alleles may impact parasite susceptibility to other antimalarials. As an example, the 7G8 allele, representative of pfcrt sequences from South America and the Pacific region, imparts low-level cross resistance to monodesethyl-amodiaquine, the active metabolite of amodiaquine (Sidhu et al., 2002). Cross-resistance to quinine has also been observed with some pfcrt alleles in certain genetic backgrounds. In contrast, mutant pfcrt increases parasite susceptibility to lumefantrine and artemisinin derivatives, which bodes well for the use of artemether-lumefantrine in areas where the prevalence of mutant pfcrt is high (Sisowath et al., 2009). In addition to PfCRT, the P-glycoprotein transporter encoded by pfmdr1, and other transporters including PfMRP, may play a modulatory role in chloroquine resistance (Duraisingh and Cowman, 2005, Raj et al., 2009), and the glutathione system also could contribute (Ginsburg and Golenser, 2003).

Absorption, Fate, and Excretion. Chloroquine is well absorbed from the GI tract and rapidly from intramuscular and subcutaneous sites. This drug extensively sequesters in tissues, particularly liver, spleen, kidney, lung, melanin-containing tissues, and, to a lesser extent, brain and spinal cord. Chloroquine binds moderately (60%) to plasma proteins and undergoes appreciable biotransformation via hepatic CYPs to two active metabolites, desethylchloroquine and bisdesethylchloroquine (Ducharme and Farinotti, 1996). These metabolites may reach plasma concentrations of 40% and 10% of that of chloroquine, respectively. The renal clearance of chloroquine is about half of its total systemic clearance. Unchanged chloroquine and desethylchloroquine account for >50% and 25% of the urinary drug products, respectively, and the renal excretion of both compounds is increased by acidification of the urine.

Both in adults and in children, chloroquine exhibits complex pharmacokinetics such that plasma levels of the drug shortly after dosing are determined primarily by the rate of distribution rather than the rate of elimination (Krishna and White, 1996). Because of extensive tissue binding, a loading dose is required to achieve effective concentrations in plasma. After parenteral administration, rapid entry into the bloodstream together with slow exit from this compartment can result in transiently high and potentially lethal concentrations of the drug in plasma. Hence, parenteral chloroquine is given either slowly by constant intravenous infusion or in small divided doses by the subcutaneous or intramuscular route. Chloroquine is safer when given orally because the rates of absorption and distribution are more closely matched. Peak plasma levels are achieved in ~3-5 hours after dosing by this route. The t_1/2 of chloroquine increases from a few days to weeks as plasma levels decline, reflecting release of drug from extensive tissue stores. The terminal t_1/2 ranges from 30 to 60 days, and traces of the drug can be found in the urine for years after a therapeutic regimen.

Chloroquine is inexpensive and safe, but its usefulness has declined across most malaria-endemic regions of the world because of the spread of chloroquine-resistant P. falciparum. Except in areas where resistant strains of P. vivax are reported, chloroquine is very effective in chemoprophylaxis or treatment of acute attacks of malaria caused by P. vivax, P. ovale, and P. malariae (Table 49–3). Chloroquine has no activity against primary or latent liver stages of the parasite. To prevent relapses in P. vivax and P. ovale infections, primaquine can be given either with chloroquine or used after a patient leaves an endemic area. Chloroquine rapidly controls the clinical symptoms and parasitemia of acute malarial attacks. Most patients become completely afebrile within 24-48 hours after receiving therapeutic doses, and thick smears of peripheral blood generally are negative by 48-72 hours. If patients fail to respond during the second day of chloroquine therapy, resistant strains should be suspected and therapy instituted with quinine plus tetracycline or doxycycline, or atovaquone-proguanil, or artemether-lumefantrine, or mefloquine if the others are not available. Although chloroquine can be given safely by parenteral routes to comatose or vomiting patients, quinidine gluconate usually is given in the U.S. In comatose children, chloroquine is well absorbed and effective when given through a nasogastric tube. Tables 49–2 and 49–3 provide information about recommended chemoprophylactic and therapeutic dosage regimens involving the use of chloroquine. These regimens are subject to modification according to clinical judgment, geographic patterns of chloroquine resistance, and regional usage.

Chloroquine and its analogs are also used to treat certain nonmalarial conditions, including hepatic amebiasis. Chloroquine and hydroxychloroquine have also been used as secondary drugs to treat a variety of chronic diseases because both alkaloids concentrate in lysosomes and have anti-inflammatory properties. Thus, these compounds, often together with other agents, have clinical efficacy in rheumatoid arthritis, systemic lupus erythematosus, discoid lupus, sarcoidosis, and photosensitivity diseases such as porphyria cutanea tarda and severe polymorphous light eruption.

Acute chloroquine toxicity is encountered most frequently when therapeutic or high doses are administered too rapidly by parenteral routes. Toxic manifestations relate primarily to the cardiovascular system and the CNS. Cardiovascular effects include hypotension, vasodilation, suppressed myocardial function, cardiac arrhythmias, and eventual cardiac arrest. Confusion, convulsions, and coma may also result from overdose. Chloroquine doses of >5 g given parenterally usually are fatal. Prompt treatment with mechanical ventilation, epinephrine, and diazepam may be lifesaving.

Doses of chloroquine used for oral therapy of the acute malarial attack may cause GI upset, headache, visual disturbances, and urticaria. Pruritus also occurs, most commonly among dark-skinned persons. Prolonged treatment with suppressive doses occasionally causes side effects such as headache, blurring of vision, diplopia, confusion, convulsions, lichenoid skin eruptions, bleaching of hair, widening of the QRS interval, and T-wave abnormalities. These complications usually disappear soon after the drug is withheld. Rare instances of hemolysis and blood dyscrasias have been reported. Chloroquine may cause discoloration of nail beds and mucous membranes. This drug has also been reported to interfere with the immunogenicity of certain vaccines (Pappaioanou et al, 1986).

High daily doses of chloroquine or hydroxychloroquine (>250 mg) leading to cumulative total doses of >1 g/kg, such as those used for treatment of diseases other than malaria, can result in irreversible retinopathy and ototoxicity. Retinopathy presumably is related to drug accumulation in melanin containing tissues and can be avoided if the daily dose is ≤250 mg (Rennie, 1993). Prolonged therapy with high doses of chloroquine or hydroxychloroquine also can cause toxic myopathy, cardiopathy, and peripheral neuropathy. These reactions improve if the drug is withdrawn promptly (Estes et al., 1987). Rarely, neuropsychiatric disturbances, including suicide, may be related to overdose.

Precautions and Contraindications. Chloroquine is not recommended for treating individuals who have epilepsy or myasthenia gravis. This drug should be used cautiously if at all in the presence of advanced liver disease or severe GI, neurological, or blood disorders. In individuals with decreased renal function, dosage should be adjusted to avoid elevated plasma concentrations. In rare cases, chloroquine can cause hemolysis in patients with G6PD deficiency (see later section on primaquine). Concomitant use of gold or phenylbutazone (no longer available in the U.S.) with chloroquine, in the treatment of rheumatoid arthritis, should be avoided because of the tendency of all three agents to produce dermatitis. Chloroquine should not be prescribed for patients with psoriasis or other exfoliative skin conditions because it can cause severe reactions. Because of the danger of cutaneous reactions, it should also not be used to treat malaria in patients with porphyria cutanea tarda; however it can be used in lower doses for treatment of manifestations of this form of porphyria. Chloroquine inhibits CYP2D6 and interacts with a variety of different drugs. It attenuates the efficacy of the yellow fever vaccine when administered at the same time. It should not be given with mefloquine because of increased risk of seizures and lack of added benefit. Most important, chloroquine opposes the action of anticonvulsants and increases the risk of ventricular arrhythmias when co-administered with amiodarone or halofantrine. By increasing plasma levels of digoxin and cyclosporine, chloroquine also can increase the risk of toxicity from these agents. Patients receiving long-term, high-dose therapy should undergo ophthalmological and neurological evaluations every 3-6 months.

History. The medicinal use of quinine dates back >350 years (Rocco, 2003). Quinine is the chief alkaloid of cinchona, the powdered bark of the South American cinchona tree, otherwise known as Peruvian, Jesuit's, or Cardinal's bark. It had been used by indigenous Peruvians to treat shivering. In 1633, an Augustinian monk named Calancha, of Lima, Peru, first wrote that a powder of cinchona "given as a beverage, cures the fevers and tertians." By 1640, cinchona was used to treat fevers in Europe. The Jesuit fathers were the main importers and distributors of cinchona in Europe.

For almost two centuries the bark was employed for medicine as a powder, extract, or infusion. In 1820, Pelletier and Caventou isolated quinine from cinchona. Quinine still is a mainstay for treating attacks of chloroquine- and multidrug-resistant P. falciparum malaria (Table 49–3). However, combination therapy (now standard artemisinin-derivative combinations) with other antimalarials is supplanting quinine regimens because of increasing resistance of P. falciparum to quinine in Southeast Asia and parts of the Amazon basin. Quinine toxicity, usually seen in overdose, can include pulmonary edema, immune thrombocytopenic purpura, irreversible deafness, or arrhythmias.

Oral quinine is FDA approved for the treatment of uncomplicated P. falciparum malaria and is currently available from one manufacturer in the U.S. Intravenous quinine is not available in the U.S. The more potent enantiomer, quinidine, is the preferred intravenous form.

Chemistry. Cinchona contains a mixture of >20 structurally related alkaloids, the most important of which are quinine and quinidine. Both compounds contain a quinoline group attached through a secondary alcohol linkage to a quinuclidine ring (Figure 49–2). A methoxy side chain is attached to the quinoline ring and a vinyl to the quinuclidine. They differ only in the steric configuration at two of the three asymmetrical centers: the carbon bearing the secondary alcohol group and at the quinuclidine junction. Although quinine and quinidine have been synthesized, the procedures are complex; hence they still are obtained from natural sources. Quinidine is both somewhat more potent as an antimalarial and more toxic than quinine (Griffith et al., 2007). Structure-activity analysis of the cinchona alkaloids provided the basis for the discovery of more recent antimalarials such as mefloquine.

The antimalarial mechanism of quinine is thought to share similarities to chloroquine in being able to bind heme and prevent its detoxification. The basis of P. falciparum resistance to quinine, nonetheless, is complex. Patterns of P. falciparum resistance to quinine correlate in some strains with resistance to chloroquine yet in others correlate more closely with resistance to mefloquine and halofantrine. Gene amplification of pfmdr1 in P. falciparum, implicated in resistance to mefloquine and halofantrine, can contribute to reduced quinine susceptibility in vitro. Similarly, pfmdr1 point mutations can also contribute to quinine resistance, in particular the N1042D mutation (Duraisingh and Cowman, 2005; Sidhu et al., 2005). Despite their close chemical similarity, quinine and quinidine sensitivity can also diverge in some strains harboring novel PfCRT haplotypes (Cooper et al., 2002). Recent evidence suggests that other transporter genes participate in conferring resistance to quinine, potentially including the sodium-hydrogen exchanger PfNHE (Hayton and Su, 2008; Nkrumah et al., 2009).

Action on Skeletal Muscle. Quinine and related cinchona alkaloids exert effects on skeletal muscle that can have clinical implications. Quinine increases the tension response to a single maximal stimulus delivered to muscle directly or through nerves, but it also increases the refractory period of muscle so that the response to tetanic stimulation is diminished. The excitability of the motor end-plate region decreases so that responses to repetitive nerve stimulation and to acetylcholine are reduced. Thus, quinine can antagonize the actions of physostigmine on skeletal muscle as effectively as curare. Quinine may also produce alarming respiratory distress and dysphagia in patients with myasthenia gravis. The effect on skeletal muscle can be augmented by concurrent administration of gentamicin. Quinine may also cause symptomatic relief of myotonia congenita. This disease is the pharmacological antithesis of myasthenia gravis, such that drugs effective in one syndrome aggravate the other.

Absorption, Fate, and Excretion. Quinine is readily absorbed when given orally or intramuscularly. In the former case, absorption occurs mainly from the upper small intestine and is >80% complete, even in patients with marked diarrhea. After an oral dose, plasma levels reach a maximum in 3-8 hours and, after distributing into an apparent volume of ~1.5 L/kg in healthy individuals, decline with a t_1/2 of ~11 hours. The pharmacokinetics of quinine may change according to the severity of malarial infection (Krishna and White, 1996). Values for both the apparent volume of distribution and the systemic clearance of quinine decrease, the latter more than the former, such that the average elimination t_1/2 increases to 18 hours. In patients with severe infection, standard therapeutic doses may produce peak plasma levels of quinine as high as 15-20 mg/L without causing major toxicity. In contrast, levels >10 mg/L can produce severe drug reactions. The high levels of plasma α₁-acid glycoprotein produced in severe malaria may prevent toxicity by binding quinine and thereby reducing the free fraction of drug. Concentrations of quinine are lower in erythrocytes (33-40%) and CSF (2-5%) than in plasma, and the drug readily reaches fetal tissues.

The cinchona alkaloids are metabolized extensively, especially by hepatic CYP3A4; thus only ~20% of an administered dose is excreted in an unaltered form in the urine. These drugs do not accumulate in the body upon continued administration. However, the major metabolite of quinine, 3-hydroxyquinine, retains some antimalarial activity and can accumulate and possibly cause toxicity in patients with renal failure. Renal excretion of quinine itself is more rapid when the urine is acidic.

In a series of studies over the past two decades, White and associates designed rational regimens, including the institution of loading doses, for the use of quinine and quinidine in the treatment of P. falciparum malaria in Southeast Asia (Krishna and White, 1996). Between 0.2 and 2.0 mg/L has been estimated as the therapeutic range for "free" quinine. Regimens needed to achieve this target may vary based on patient age, severity of illness, and the responsiveness of P. falciparum to the drug. For example, lower doses are more effective in treating children in Africa than adults in Southeast Asia because the pharmacokinetics of quinine differ in the two populations, as does the susceptibility of P. falciparum to the drug (Krishna and White, 1996). Dosage regimens for quinidine are similar to those for quinine, although quinidine binds less to plasma proteins and has a larger apparent volume of distribution, greater systemic clearance, and shorter terminal elimination t_1/2 than quinine (Griffith et al., 2007; Miller et al., 1989). Thompson et al. (2003) suggest that the dose of quinidine currently recommended by the CDC (10 mg salt/kg initially, followed by 0.02 mg salt/kg per minute) may be too low and should be 10 mg salt/kg and 0.02 mg/kg of base/minute (60% of the salt is base). Clinical data on which to base a firm recommendation are lacking.

Treatment of Nocturnal Leg Cramps. It is commonly believed that night cramps might be relieved by quinine taken at bedtime at a dose of 200-300 mg (available until 1995 in products that did not require a prescription). Some consider this condition severe. The degree of symptomatic relief appears to vary substantially between individuals. In 1995, the FDA issued a ruling that required drug manufacturers to stop marketing over-the-counter quinine products for nocturnal leg cramps, stating that data supporting safety and efficacy of quinine for this indication were inadequate and that risks outweighed the potential benefits.

Hearing and vision are particularly affected. Functional impairment of the eighth nerve results in tinnitus, decreased auditory acuity, and vertigo. Visual signs consist of blurred vision, disturbed color perception, photophobia, diplopia, night blindness, constricted visual fields, scotomata, mydriasis, and even blindness (Bateman and Dyson, 1986). These visual and auditory effects probably result from direct neurotoxicity, although secondary vascular changes may have a role. Marked spastic constriction of the retinal vessels leads to retinal ischemia, pale optic discs, and retinal edema, with the potential for severe optic atrophy.

GI symptoms are also prominent in cinchonism. Nausea, vomiting, abdominal pain, and diarrhea result from the local irritant action of quinine, but the nausea and emesis also have a central basis. Cutaneous manifestations may include flushing, sweating, rash, and angioedema, especially of the face. Quinine and quinidine, even at therapeutic doses, may cause hyperinsulinemia and severe hypoglycemia through their powerful stimulatory effect on pancreatic beta cells. Despite treatment with glucose infusions, this complication can be serious and possibly life threatening, especially in pregnancy and prolonged severe infection. Hypoglycemia is seen occasionally in uninfected patients who take quinine.

Quinine rarely causes cardiac complications unless therapeutic plasma concentrations are exceeded (Krishna and White, 1996). QTc prolongation is mild and does not appear to be affected by concurrent mefloquine treatment. Acute overdosage also may cause serious and even fatal cardiac dysrhythmias such as sinus arrest, junctional rhythms, AV block, and ventricular tachycardia and fibrillation (Bateman and Dyson, 1986). Quinidine is even more cardiotoxic than quinine. Cardiac monitoring of patients on intravenous quinidine is advisable where possible.

Severe hemolysis can result from hypersensitivity to these cinchona alkaloids. Hemoglobinuria and asthma from quinine may occur more rarely. "Blackwater fever"—the triad of massive hemolysis, hemoglobinemia, and hemoglobinuria leading to anuria, renal failure, and in some instances death—is a rare type of hypersensitivity reaction to quinine therapy that can occur during treatment of malaria. Quinine may cause milder hemolysis upon occasion, especially in people with G6PD deficiency. Thrombotic thrombocytopenic purpura is a rare but clinically significant adverse effect. This reaction can occur even in response to ingestion of tonic water, which has ~4% the therapeutic oral dose per 12 oz ("cocktail purpura"). Other rare adverse effects include hypoprothrombinemia, leukopenia, and agranulocytosis.

This drug should be avoided in patients with tinnitus or optic neuritis. In patients with cardiac dysrhythmias, the administration of quinine requires the same precautions as for quinidine. Quinine appears to be safe in pregnancy and is used commonly for the treatment of pregnancy-associated malaria. However, glucose levels must be monitored because of the increased risk of hypoglycemia.

Because parenteral solutions of quinine and quinidine are highly irritating, the drug should not be given subcutaneously. Concentrated solutions may cause abscesses when injected intramuscularly, or thrombophlebitis when infused intravenously. Antacids that contain aluminum can delay absorption of quinine from the GI tract. Quinine and quinidine can delay the absorption and elevate plasma levels of digoxin and related cardiac glycosides. Likewise, these alkaloids may raise plasma levels of warfarin and related anticoagulants. The action of quinine at neuromuscular junctions enhances the effect of neuromuscular blocking agents and opposes the action of acetylcholinesterase inhibitors. Prochlorperazine can amplify quinine's cardiotoxicity, as can halofantrine. The renal clearance of quinine can be decreased by cimetidine and increased by urine acidification and by rifampin.

History. Mefloquine (lariam) is a product of the Malaria Research Program established by the Walter Reed Institute for Medical Research in 1963 to develop promising new compounds to address the alarming growth of drug-resistant malaria. Of the many 4-quinoline methanols tested based on their structural similarity to quinine, mefloquine displayed high antimalarial activity in animal models, and emerged from clinical trials as safe and effective against drug-resistant strains of P. falciparum. Mefloquine was first used to treat chloroquine-resistant P. falciparum malaria in Thailand. However, the slow elimination of mefloquine fostered the emergence of drug-resistant parasites.

Chemistry. The structure of the mefloquine raceme is illustrated in Figure 49–2. Recent work has demonstrated that the (–)-enantiomer is associated with adverse CNS effects, whereas the (+)-enantiomer retains antimalarial activity with fewer side effects. Mefloquine can be paired with artesunate, thereby reducing the selection pressure for resistance. This combination has proved efficacious for the treatment of P. falciparum malaria, even in regions with high prevalence of mefloquine-resistant parasites.

Earlier work showed that mefloquine associates with intra-erythrocytic hemozoin, suggesting similarities to the mode of action of chloroquine (Sullivan et al., 1998). However, evidence that this association might be secondary to a primarily cytosolic mode of action comes from studies with transgenic P. falciparum lines expressing different pfmdr1 copy numbers. Increased pfmdr1 copy numbers are associated with both reduced parasite susceptibility to mefloquine (and artemisinin) and increased PfMDR1-mediated solute import into the digestive vacuole of intraerythrocytic parasites (Rohrbach et al., 2006). Therefore, if the drug's target resides outside of this vacuolar compartment, increased import of mefloquine into the digestive vacuole, potentially driven by PfMDR1 activity, would be beneficial for the parasite and reduce its susceptibility to the drug.

A comprehensive clinical and molecular epidemiological study conducted with parasites from Thailand identified pfmdr1 gene amplification as a major determinant of mefloquine treatment failure and in vitro mefloquine resistance (Price et al., 2004). Some individuals in that study nonetheless failed mefloquine treatment despite not having parasites with multiple pfmdr1 copies, implying secondary mechanisms of resistance. Mefloquine susceptibility in vitro can also be affected by the presence of point mutations in pfmdr1 or pfcrt (Valderramos and Fidock, 2006).

Absorption, Fate, and Excretion. Mefloquine is taken orally because parenteral preparations cause severe local reactions. The drug is rapidly absorbed, with marked variability between individuals. Probably owing to extensive enterogastric and enterohepatic circulation, plasma levels of mefloquine rise in a biphasic manner to their peak in ~17 hours. Mefloquine has a variable and long t_1/2, 13-24 days, reflecting its high lipophilicity, extensive tissue distribution, and extensive binding (~98%) to plasma proteins. Mefloquine is extensively metabolized in the liver. CYP3A4 has been implicated in the metabolism of mefloquine; this CYP can be inhibited by ketoconazole and induced by rifampicin (German and Aweeka, 2008). Excretion of mefloquine is mainly by the fecal route; only ~10% of mefloquine appears unchanged in the urine. The stereoisomers of mefloquine exhibit quite different pharmacokinetic characteristics that relate to their biodisposition (Hellgren et al., 1997).

Estimates for frequency of severe CNS toxicity after mefloquine treatment can be as high as 0.5%. Adverse reactions include seizures, confusion or decreased sensorium, acute psychosis, and disabling vertigo. Such symptoms generally are reversible upon drug discontinuation. At chemoprophylactic dosages, the risk of serious neuropsychiatric effects is estimated to be ~0.01% (about the same as for chloroquine). Mild-to-moderate toxicities (e.g., disturbed sleep, dysphoria, headache, GI disturbances, and dizziness) occur even at chemoprophylactic dosages. Whether these symptoms are more common than with other antimalarial regimens is debated. Adverse effects usually manifest after the first to third doses and often abate even with continued treatment. Reports of cardiac abnormalities, hemolysis, and agranulocytosis are rare.

Contraindications and Interactions. At very high doses, mefloquine is teratogenic in rodents. Studies have suggested an increased rate of stillbirths with mefloquine use, especially during the first trimester (Taylor and White, 2004). The significance of these data has been debated, but it is fair to say that the evidence for mefloquine's safety in pregnancy is not convincing, and mefloquine may be used if it is the only available treatment option. Pregnancy should be avoided for 3 months after mefloquine use because of the prolonged t_1/2 of this agent.

This drug is contraindicated for patients with a history of seizures, depression, bipolar disorder and other severe neuropsychiatric conditions, or adverse reactions to quinoline antimalarials such as quinine, quinidine, halofantrine, mefloquine, and chloroquine. Mefloquine may counteract seizures in epileptic patients. Although this drug can be taken safely 12 hours after a last dose of quinine, taking quinine shortly after mefloquine can be very hazardous because the latter is eliminated so slowly. Treatment with or after halofantrine or within 2 months of prior mefloquine administration is contraindicated.

Recent studies do not indicate that mefloquine compromises the performance of tasks that require good motor coordination, e.g., driving or operating machinery. Although some advise against the use of mefloquine for patients in occupations that require focused concentration, dexterity and cognitive function in safety-sensitive settings, such as pilots, controlled studies suggest that mefloquine does not impair performance in persons who tolerate the drug.

History. The weak anti-Plasmodium activity of methylene blue, first discovered by Ehrlich in 1891, led to the development of the 8-aminoquinoline antimalarials, of which pamaquine was the first introduced into medicine. During World War II the search for more potent and less toxic 8-aminoquinoline antimalarials resulted in the discovery of primaquine (Figure 49–2).

Primaquine, in contrast to other antimalarials, acts on exoerythrocytic tissue stages of plasmodia in the liver to prevent and cure relapsing malaria. The striking hemolysis that may follow primaquine therapy led directly to the landmark discovery of G6PD deficiency, the first genetic disorder associated with an enzyme. Hemolysis remains notoriously identified with primaquine. Patients should be screened for G6PD deficiency prior to therapy with this drug. The pressing need for alternatives to this important drug has resulted in the evaluation of multiple 8-aminoquinoline analogs (Vale et al., 2009). These include tafenoquine, a compound under clinical evaluation that is active against relapsing liver stage forms of P. vivax, and multidrug resistant asexual blood stage forms of P. falciparum (Wells et al., 2009).

The mechanism of action of the 8-aminoquinolines has not been elucidated. Primaquine may be converted to electrophilic intermediates that act as oxidation-reduction mediators. Such activity could contribute to antimalarial effects by generating reactive oxygen species or by interfering with mitochondrial electron transport in the parasite (Vale et al., 2009). Some strains of P. vivax can exhibit partial resistance to primaquine (Baird, 2009).

The apparent volume of distribution is several times that of total-body water, due to extensive tissue distribution. Primaquine binds preferentially to the acute-phase reactant protein α1-glycoprotein, which may alter the distribution of free drug depending on its concentration in the blood. Primaquine is metabolized rapidly, and only a small fraction of an administered dose is excreted as the parent drug. Importantly, primaquine induces CYP1A2. Thus, caution should be taken in administering primaquine with drugs metabolized by CYP1A2 (including warfarin) (Hill et al., 2006). The major metabolite, carboxyprimaquine, is inactive.

For terminal chemoprophylaxis, primaquine regimens should be initiated shortly before or immediately after a subject leaves an endemic area (Table 49–2). Radical cure of P. vivax or P. ovale malaria can be achieved if the drug is given either during an asymptomatic latent period of presumed infection (based on travel to or residence within an endemic region) or during an acute attack. Simultaneous administration of a schizonticidal drug plus primaquine is more effective in radical cure than sequential treatment. Limited studies have shown efficacy in prevention of P. falciparum and P. vivax malaria when primaquine is taken as chemoprophylaxis (Taylor and White, 2004). Gametocytocidal activity of primaquine should also reduce the transmission potential during drug treatment of P. falciparum. However, this approach to prophylaxis is not routine clinical practice. Primaquine is generally well tolerated when taken for up to 1 year.

Toxicity and Side Effects. Primaquine has few side effects when given to most whites in the usual therapeutic doses. Primaquine can cause mild to moderate abdominal distress in some individuals. Taking the drug at mealtime often alleviates these symptoms. Mild anemia, cyanosis (methemoglobinemia), and leukocytosis are less common. High doses (60-240 mg daily) worsen the abdominal symptoms and cause some degree of methemoglobinemia in most subjects. Methemoglobinemia can occur even with usual doses of primaquine and can be severe in individuals with congenital deficiency of NADH methemoglobin reductase (Coleman and Coleman, 1996). Chloroquine and dapsone may synergize with primaquine to produce methemoglobinemia in these patients. Granulocytopenia and agranulocytosis are rare complications of therapy and usually are associated with overdosage. Other rare adverse reactions are hypertension, arrhythmias, and symptoms referable to the CNS.

Therapeutic or higher doses of primaquine may cause acute hemolysis and hemolytic anemia in humans with G6PD deficiency (Vale et al., 2009). This X-linked condition, primarily owing to amino acid substitutions in the G6PD enzyme, affects >200 million people worldwide. More than 400 genetic variants of G6PD produce variable responses to oxidative stress. About 11% of African Americans have the A variant of G6PD, and are therefore vulnerable to hemolysis caused by pro-oxidant drugs such as primaquine. G6PD deficiency is uncommon in Latin America but may be present in those residents of African descent. Primaquine-induced hemolysis can be even more severe in white ethnic groups, including Sardinians, Sephardic Jews, Greeks, and Iranians; these populations have a G6PD variant in which two amino acid substitutions impair both enzyme stability and activity. Because primaquine sensitivity is inherited through an X-linked gene, hemolysis is often of intermediate severity in heterozygous females who have two populations of red blood cells, one normal and the other deficient in G6PD. Owing to "variable penetrance," these females may be affected less frequently than predicted. Primaquine is the prototype of >50 drugs, including antimalarial sulfonamides, that cause hemolysis in G6PD-deficient individuals.

History. Shortly after their introduction into clinical practice, the sulfonamides were found to have antimalarial activity, a property investigated extensively during World War II. The efficacy of sulfones was demonstrated in the first trial of dapsone against P. falciparum in 1943. The sulfonamides combined with pyrimethamine have been used to treat chloroquine-resistant P. falciparum malaria, particularly in parts of Africa. The sulfonamides and sulfones are slow-acting blood schizonticides and more active against P. falciparum than P. vivax.

Mechanism of Action. Sulfonamides are p-aminobenzoic acid analogs that competitively inhibit Plasmodium dihydropteroate synthase. These agents are combined with an inhibitor of parasite dihydrofolate reductase to enhance their antimalarial action. The synergistic "antifolate" combination of sulfadoxine, a long-acting sulfonamide, with pyrimethamine has been extensively used to treat malaria. As observed for pyrimethamine, sulfadoxine resistance appeared quickly in vivo after the introduction of this treatment as monotherapy.

Drug Resistance. Sulfadoxine resistance is conferred by several point mutations in the dihydropteroate synthase gene; the more widespread is the substitution Ala437Gly. These sulfadoxine-resistance mutations, when combined with mutations of dihydrofolate reductase and conferring pyrimethamine resistance, greatly increase the likelihood of sulfadoxine-pyrimethamine treatment failure. Sulfadoxine-pyrimethamine, given intermittently during the second and third trimesters of pregnancy, is a routine component of antenatal care throughout Africa. Intermittent preventive treatment strategies might also benefit infants (Aponte et al., 2009). However, there are substantial concerns about these strategies in the face of the expanding profile of resistance to pyrimethamine-sulfadoxine. A recent report found that intermittent preventive treatment in pregnancy exacerbated placental inflammation and caused increased levels of parasitemia and increased selective pressure for drug-resistant infections (Harrington et al., 2009). Thus, there is a clear need to identify alternative intermittent preventive treatment regimens. Generally, one can anticipate that, in the absence of novel antifolates effective against existing drug-resistant strains, the use of these antimalarials for either prevention or treatment will continue to decline.

Tetracyclines are a group of antibiotic agents initially derived from Streptomyces. The pharmacological properties of tetracyclines are presented in Chapter 55. Two members of this group, tetracycline and doxycycline, are useful in malaria treatment. In addition, clindamycin, a lincosamide antibiotic (Chapter 55), is also recommended. These agents are slow-acting blood schizonticides that can be used alone for short-term chemoprophylaxis in areas with chloroquine- and mefloquine-resistant malaria (only doxycycline is recommended for malaria chemoprophylaxis). All of these antibiotics act via a delayed death mechanism resulting from their inhibition of protein translation in the parasite plastid. This effect on malarial parasites manifests as death of the progeny of drug-treated parasites, resulting in slow onset of antimalarial activity (Dahl and Rosenthal, 2008). Their relatively slow mode of action makes these drugs ineffective as single agents for malaria treatment. However, their efficacy as treatment is increased when they are used as an adjunct to quinine, quinidine, or artesunate, and in this context, they can be used to treat uncomplicated or severe P. falciparum malaria. These antibiotics are not clinically used to eliminate liver stage infection.

Dosage regimens for tetracyclines and clindamycin are listed in Tables 49–2 and 49–3. Because of their adverse effects on bones and teeth, tetracyclines should not be given to pregnant women or to children <8 years of age. In these subjects, clindamycin is a reasonable alternative. Photosensitivity reactions or drug-induced superinfections may mandate discontinuation of therapy or chemoprophylaxis with tetracyclines. As an alternative antibiotic, the macrolide azithromycin also displays antimalarial activity through a similar delayed-death mechanism and is undergoing further evaluation. As with clindamycin, azithromycin is safe for use in young children and pregnant women.

Genetic studies indicate that isolates of P. falciparum from patients in highly endemic areas contain many parasite clones with different drug resistance phenotypes (Druilhe et al., 1998). A patient with severe malaria may have up to 10¹² parasites. Because single point mutations occur with an estimated frequency of 10^–9, based on in vitro studies (Cooper et al., 2002), single point mutations that confer resistance could potentially arise in virtually every patient and double point mutations would arise occasionally. Most of these new mutant strains will not survive because of immunity, decreased fitness, and additional factors; nonetheless, successful drugs or drug combinations must ideally not be susceptible to resistance resulting from single point mutations. Mutations by gene amplification are expected to be even more frequent than single point mutations (Nair et al., 2008). Looking back in time, one notes that decades of intense chloroquine use preceded development of resistance to this agent. This delay was likely due to the requirement for at least four point mutations in the pfcrt gene to confer resistance (Valderramos and Fidock, 2006). Pharmacokinetics can also be a determinant in the generation of resistance. Drugs with a long t_1/2 are thought to generally be more likely to select for resistance (White, 2004).

On a population level, the fitness of resistant parasites is another important parameter: If mutation causes a reduced ability to survive or grow, the parasites are less likely to spread. However, in the case of antifolates, treatment actually induces gametocytogenesis, which promotes transmission; this effect, together with other forces (intense selection pressure from preventive drug use; parasite migrations driven by humans), may explain the spread of particular resistance alleles across Southeast Asia and Africa. These considerations strongly argue for regimens that combine two or more antimalarial agents with different mechanisms of action (and therefore, hopefully, with different mechanisms of resistance) to treat drug-resistant P. falciparum malaria (White, 2004). ACTs are the most important combination treatment currently available.

In those few areas where chloroquine-sensitive strains of P. falciparum are found, chloroquine is still suitable for chemoprophylaxis. It also remains the drug of choice for chemoprophylaxis and control of infections due to P. vivax, P. ovale, P. malariae, and P. knowlesi. In areas where chloroquine-resistant malaria is endemic, mefloquine and atovaquone-proguanil are the regimens of choice for chemoprophylaxis. There is more experience with mefloquine and more evidence for its efficacy against P. vivax, an important consideration in areas where this species co-exists with P. falciparum. However, there are more contraindications to the use of mefloquine and perhaps greater toxicity than for alternative regimens. Doxycycline is an alternative chemoprophylactic agent. In cases where mefloquine, atovaquone-proguanil, and doxycycline all are contraindicated, primaquine is a possibility for chemoprophylaxis. Primaquine, like atovaquone-proguanil, is active against liver stages and can be discontinued shortly after leaving the endemic area. Attempts at radical cure of P. vivax malaria with primaquine should be delayed until the patient leaves an endemic area.

Children and pregnant women are the most susceptible to severe malaria. With appropriate dosage adjustments and safety precautions, the treatment of children generally is the same as for adults (pediatric dose should never exceed adult dose). However, tetracyclines should not be given to children <8 years of age except in an emergency, and atovaquone-proguanil as treatment has been approved only for children weighing >5 kg (for current information, see the CDC website and Boggild et al., 2007).

Among the antimalarial treatments available in the U.S., many should not be used during pregnancy because of lack of sufficient formal safety data (antifolates, atovaquone-proguanil, and artemether-lumefantrine) or known potential risks for the fetus (tetracycline, doxycycline, primaquine, and mefloquine). The recommended treatment of uncomplicated malaria during pregnancy relies on chloroquine in areas with chloroquine-sensitive parasites, and quinine, alone (in P. vivax infections) or in combination with clindamycin (in P. falciparum infections), in areas with chloroquine-resistant parasites. Appropriate caution should be exercised when administering quinine, given the frequency of hypoglycemia as an adverse reaction in pregnant women. Pregnant patients with P. vivax or P. ovale infections should be kept under chloroquine chemoprophylaxis until delivery. After delivery, and after verifying that the patient is not G6PD deficient, treatment with primaquine should be initiated to eradicate hypnozoites. Atovaquone-proguanil and artemether-lumefantrine are not indicated because of lack of safety data. Mefloquine is also not indicated because of a possible association with increased risk of stillbirth. However, if recommended treatments are not available, these alternative options should be considered after balancing the potential benefits and risks for the fetus.

A recent study demonstrated better outcomes when pregnant women were treated with artesunate-atovaquone-proguanil compared with quinine; however, adequate safety data for the developing fetus are lacking (for review, see Dellicour et al., 2007). Another clinical trial recently showed that a standard six-dose oral artemether-lumefantrine regimen was well tolerated and safe in pregnant women with uncomplicated falciparum malaria, but efficacy was inferior to 7-day artesunate monotherapy and was unsatisfactory for general deployment along the Thai Cambodian border, probably resulting from low drug concentrations in later pregnancy (McGready et al., 2008). The use of antimalarial drugs in pregnancy has been well reviewed (Ward et al., 2007).

In lactating mothers, treatment with most compounds is acceptable, although chloroquine and hydroxychloroquine are the preferred agents. The use of atovaquone-proguanil is not recommended unless breast-feeding infants weigh >5 kg. Also, the breast-feeding infant should be shown to have a normal G6PD level before receiving primaquine (Schlagenhauf and Petersen, 2008).

New Targets, New Drugs. In the face of evolving drug resistance and the necessity to increase the useful life span of antimalarial drugs through combination therapies, novel and potent antimalarial drugs are urgently needed. The malaria research pipeline includes either compounds derived from known antimalarials through modification of their chemical structure (likely to retain activity against the original targets), drugs previously developed for other infectious organisms or diseases (and likely act on the same target in Plasmodium), or inhibitors of novel targets (exploiting differences in the biology of host and parasite). This latter strategy of drug discovery benefits from the genome sequencing projects of the different Plasmodium species and the human host. Perspectives in malaria chemotherapy (including new drug targets and antimalarial drugs in discovery, preclinical and clinical phases) have recently been extensively discussed (Biot and Chibale, 2006; de Beer et al., 2009; Gardiner et al., 2009; Olliaro and Wells, 2009; Wells et al., 2009).

Among the novel antimalarial drugs undergoing clinical development, two notable examples act on new metabolic pathways:

1. TE3, a prodrug of a bis-ammonium compound that acts on phospholipid metabolism through the inhibition of de novo phosphatidylcholine synthesis, combined with a putative activity on heme detoxification. This new class of compounds has shown potent in vivo antimalarial activity in the primate model Aotus and is not cross resistance in vitro with known antimalarials.

2. Fosmidomycin, which inhibits the 1-desoxy-D-xylulose-5-phosphate reductoisomerase in the mevalonate-independent pathway of isoprenoid synthesis in the apicoplast (Wiesner and Jomaa, 2007). The apicoplast, a specialized parasite organelle of algal origin, appears to be important for lipid and heme biosynthesis, and is a major focus of current drug development efforts (Wiesner et al., 2008). Fosmidomycin was initially developed as an antibacterial drug, but has shown efficacy in combination with clindamycin or azithromycin against P. falciparum malaria.

A nonexhaustive list of the molecular targets under current experimental or clinical investigation includes two clinically validated antimalarial drug targets (dihydrofolate reductase and cytochrome bc1 complex), for which inhibitors are being developed to overcome existing resistance (this strategy is also being explored to derive new 4-aminoquinolines and amino-alcohols) and a series of novel, experimental antimalarial drug targets: dihydro-orotate dehydrogenase (involved in pyrimidine synthesis), purine nucleoside phosphorylase (nucleotide synthesis), adenosine deaminase (regulation of intracellular nucleoside pools), subtilisin-like proteases (egress from erythrocytes), falcipains (hemoglobin proteolysis), histone deacetylase (DNA replication), protein kinases (signal transduction), and fatty acid biosynthesis (apicoplast lipid synthesis). The rationale focusing on these targets for antimalarial drug development can be found in the reviews cited.