Pharmacogenetic tools for malaria and TB in the Developing World

627ISSN 1741-054110.2217/17410541.5.6.627 © 2008 Future Medicine Ltd Personalized Medicine (2008) 5(6), 627–639

Review

Pharmacogenetic tools for malaria and TB in the Developing World

Malaria & TBMalaria and TB are two major global lethal diseases responsible for hundreds of millions of clinically reported infections per year, mainly in the Developing World and in a significant group of middle-income nations. The high incidence of these diseases is associated with a large num-ber of chemotherapeutic exposures, particularly under the current global strategy of combina-tion therapy formulations [1,2,101–103]. Many of the applied drugs have long been associated with non-negligible side effects, some with serious consequences [3–5]. In principle, pharmacoge-netics can be of considerable use in individual dose adjustments, avoiding the ‘one dose fits all’ chemotherapeutic approach, leading to decreased adverse events, while optimizing drug efficacy.

TB, a disease known for thousands of years, has witnessed a resurge in incidence in the last decades, to a large extent owing to the emergence of HIV/AIDS as a global pandemic. In parallel, a rise of multidrug resistant forms of the disease has been witnessed. Approximately 30% of the world population is infected with Mycobacterium tuberculosis, and 5–10% of those infected are predicted to develop symptomatic TB. This rise in incidence has been mainly concentrated in sub-Saharan Africa, also home of the most severe HIV/AIDS burden [6].

Malaria is globally the most important parasitic disease, present in all tropical regions. The malaria agents are members of the Plasmodium genus, with five different species being able to infect man: P. falciparum, P. vivax, P. malariae, P. ovale and

P. knowlesi. Of the five different species, P. falci-parum malaria represents the deadly form of the disease, focused in young children (aged <5 years), as well as the most capable of developing drug resistance [7]. With hundreds of millions of new clinically registered infections and a total of over 1 million deaths per year, this disease represents, along with TB, one of the three major infectious diseases in the Developing World [8].

Pharmacogenetic markersThe translation of pharmacogenetics towards the routine clinical world is still in its infancy, with its practical demands under discussion [9].

By principle, the potential utility of a pharmaco-genetic marker has to be well defined in terms of an expected significant therapeutic improve-ment. Assuming its potential usefulness, a robust phenotype/genotype correlation is demanded, in order for the genetic data to be a sensitive source of specific information. This implies that the specific gene product is the main/sole factor responsible for a pivotal step of metabo lism or transport affecting the pharmacological characteristics of the com-pound under inspection. It is also convenient that the gene shows the least possible environmental influence on its expression, such as for CYP2D6, and, unlike CYP3A4, which is highly inducible by drugs such as the anti-TB drug rifampicin. These characteristics decrease the influence of external factors in the desirable clear phenotype–genotype relation. Finally, low polymorphic complexity is advantageous, allowing the testing to be compre-hensive and practically/economically possible [10].

Some of the largest therapeutic drug exposures in the planet involve drugs employed against malaria and TB, two main global infectious diseases. Amodiaquine for malaria and isoniazid for TB are two pivotal drugs in the management of these diseases. Both drugs have been associated with severe adverse events. Amodiaquine and isoniazid are metabolized polymorphically by CYP2C8 and N-acetyltransferase 2, respectively. The polymorphic genes coding for these enzymes presently represent the best candidates for the application of personal pharmacogenetics for these diseases. We review the main reasons for this view, while asking the pivotal question of whether it is presently possible for pharmacogenetic-based personalized medicine to be applied in the malaria and TB settings of the Developing World.

Keywords: adverse events, CyP2C8, developing world, malaria, N‑acetyltransferase 2, NAT2, pharmacogenetics, TB, translational medicine

Pedro Eduardo Ferreira1,2, Isa Cavaco2,3 & José Pedro Gil1,2†

†Author for correspondence 1Malaria Research, Department of Medicine, Karolinska Institutet, Rätzius väg 10, plan 5, 171 77 Stockholm, SwedenTel.: +46 852 486 826Fax: +46 852 486 [email protected] of Biotechnology and Bioengineering, Centre of Molecular and Structural Biomedicine, University of Algarve, Portugal3Section of Pharmacogenetics, Department of Physiology and Pharmacology, Karolinska Institutet, Sweden

Personalized Medicine (2008) 5(6)628 future science group

Review Ferreira, Cavaco & Gil

In the context of TB and malaria, we focus on the polymorphisms that, in our view, presently offer the best possibilities of clinical application:

Nn -acetyltransferase 2 (NAT2), coding the cen-tral enzyme in the inactivation of the central anti-TB drug isoniazid (INH) (a part of the WHO directly observed treatment short course strategy [DOTS]);

Cytochrome P450 2C8 (n CYP2C8), essentially the factor responsible for the hepatic metabo-lism of the antimalarial amodiaquine (AQ), a central component of artemisinin deriva-tive-based combination therapy (ACT) in the African continent.

TB, isoniazid & N‑acetyltransferase 2 polymorphisms

TB, isoniazid & fast & n slow acetylatorsTB is presently being chemotherapeutically man-aged at the first-line stage through the combi-nation of several drugs, including streptomycin,

rifampicin, ethambutol and pyrazinamide, asso-ciated with the central mainstay, INH. These are present in alternative combination of 2–5 com-ponents, and spread in three different treatment categories (I–III). These six-to-eight-month treat-ments are denoted as ‘short course’, as opposed to the 12- to 18-month treatment periods associated with the regimens available before the introduc-tion of rifampicin and pyrizinamide [11]. Among these five anti-TB drugs, the metabolism of INH has been studied in particular detail.

When introduced more than half a century ago, INH showed an outstanding capacity for the treatment of terminal TB [12]. The drug has remained a pivotal drug in the management of TB until today, showing an outstandingly long useful lifespan. This longevity is associated with the remarkable capacity of INH for killing active infecting bacteria located in the walls of TB cavi-ties in the first 48 h of combined treatment [13], a clinically important feature for the manage-ment of the disease [14] and in the inhibition of drug resistance development [15]. This efficacy is

EliminationGlycine conjugates

INARFP

INH

NAT2

AC-INH

RFP

Amidase

HZ

NAT2

Amidase

NAT2

diAc-HZ

Ac-HZ

Elimination

Detoxificationpath

Detoxificationpath

CYP2E1

Spontaneous

Reactivemetabolites

GSTM1

Elimination

Detoxificationpath

Adverseevents

+

+

–

Figure 1. Brief representation of the isoniazid metabolic pathways. HZ and AC-HZ represent the gateways for the further production of reactive metabolites believed to be the direct responsible entities for the adverse effects of the INH, namely the hepatotoxicity events. NAT2 slow acetylators have increased INH plasma levels leading to a more likely production of HZ. The spontaneous conversion of HZ into reactive metabolites is the probable main basis for the observed INH toxicity. Another, probably minor but non-negligible route has been proposed involving the CYP2E1-catalyzed activation of Ac-HZ towards reactive metabolites. See Figure 4 for relevant chemical structures. Ac-HZ: Acetyl hydrazine; Ac-INH: Acetyl isoniazid; diAc-HZ: Diacetyl-hydrazine; GSTM1: Glutathione-S-transferase M1; HZ: Hydrazine; INA: Isonicotinic acid; INH: Isoniazid; NAT2: N-acetyltransferase 2; RFP: Rifampicin.

www.futuremedicine.com 629future science group

Pharmacogenetic tools for malaria & TB in the Developing World Review

reflected in its significant early bactericidal activ-ity (EBA: decline in the log

10 of the number of

sputum-sourced colony-forming units during the first 2 days of treatment), still the highest among the presently used anti-TB drugs [13]. In these piv-otal first hours of treatment, the EBA is positively correlated with the INH dosing until a plateau of approximately 300 mg/patient [16,17], presently the standard daily dose for adults. Accordingly, the EBA is also significantly correlated with the drug area under curve (AUC)

2–5h values [17].

Isoniazid is extensively metabolized through N-acetylation, undergoing mainly renal elimina-tion (Figure 1). The drug was early understood to be polymorphically metabolized, with a number of premolecular biology studies clearly distribut-ing patients in slow, intermediate and fast drug acetylator groups. This classical Mendelian dis-tribution pointed to an essentially monogenic factor [18].

Slow N-acetylators are associated with an approximately threefold decrease in the clear-ance of INH, an observation related to signifi-cant changes in pharmacokinetic parameters as compared with fast acetylators [19,20]. In reverse, fast acetylators are expected to experience a low-ered INH AUC

0–24, with likely implications in

the EBA of this drug [17,21,22].From its introduction, INH was associated

with serious dose-dependent side effects, mainly peripheral neuritis [23], affecting a fraction of the treated patients. During the 1970s, further clinical studies formally recognize hepatotoxicity

as another serious side effect of INH [18]. This currently represents the main toxicity concern in anti-TB therapy, occurring in 2–28% (a varia-tion mainly dependent on the criteria used for the definition of hepatotoxicity and ethnicity) of the patients under standard INH administration (5 mg/kg) [3,101]. The biochemical mechanisms of this toxicity are still under debate, although a major role of hydrazine, derived from INH metabolism, is becoming accepted. The acetyl-ation of INH towards acetyl-INH represents a detoxification process (Figure 1).

NAT2nThe molecular genetics basis of these differential phenotypes was clarified during the 1990s [24]: INH low acetylators are homozygous for specific SNPs in the gene coding for the NAT2, which is responsible for the N-acetylation of this drug (Figure 2). These SNPs are associated in vitro with expressed proteins, which are less effi-cient, or unstable enzymes leading to decreased N-acetylation capacity [25,26].

This in vitro data was confirmed in vivo soon after the discovery of NAT2. In an early study, Southern Japanese subjects provided the first correlation between NAT2 genotypes and INH acetylation capacity [27]. Several more recent studies have clearly established the association between the long known INH acetylation tri-modal polymorphism and the NAT2*4 (‘wild-type’) carrier status (Table 1) [20,28–30]. The NAT2 SNPs I114T (341 T>C), R197Q (590 G>A),

Table 1. NAT2 genotype/INH acetylation association: set of representative studies.

subjects relevant‡ parameter and genotype – phenotype association

Location INH dose#/sampling ref.

12 healthy subjects (8 males + 4 females)

AcINH/INH (% dose/% dose):NAT2*4/*4 = 7.39 ± 1.99§

NAT2*4/*X¶ = 4.26 ± 1.26NAT2*X/*X = 1.16 ± 0.21

Kobe, Central Japan

Single dose: 225–450 mgMultiple urine sampling 0–24 h after administration

[28]

46 TB patients(28 males + 18 females)

AcINH/INH (µM/µM):NAT2*4/*4 = 1.69 ± 0.66§

NAT2*4/*X¶ = 0.88 ± 0.40NAT2*X/*X = 0.67 ± 0.34

Nanjing, Central China

300 mg/day, 7 daysPlasma sample, 2 h after administration

[29]

18 healthy subjects(13 males + 5 females)

AUC/dose (h/l):NAT2*4/*4 = 0.031 (0.027–0.037)§

NAT2*4/*X¶ = 0.048‡‡

NAT2*X/*X = 0.091 (0.067–0.132)

Cologne, Germany

Single dose: 300 mgMultiple plasma sampling0–24 h after administration(pharmacokinetic profile)

[20]

102 TB in- and out-patients(28 males + 74 females)

AcINH/INH (µg·ml-1/µg·ml-1):NAT2*4/*4 = 9.181 ± 4.129NAT2*4/*X¶ = 3.465 ± 1.389§

NAT2*X/*X = 0.938‡‡

Sendai, Central-North Japan

300 mg/day, more than 2 weeksPlasma sample 3 h after last INH administration

[30]

‡According to the objectives of this review;§Difference statistically significant (p < 0.01); ¶*X = any allele other than *4; #Oral administration; ‡‡Two samples only.AcINH: Acetyl-isoniazid; AUC: Area under the curve; INH: Isoniazid; NAT2: N-acetyltransferase 2.

Personalized Medicine (2008) 5(6)630 future science group

Review Ferreira, Cavaco & Gil

G286E (857 G>A), and probably, R64Q (191 G>A), define the main alleles associated with decreased INH acetylation – NAT2*5, *6 and *7, respectively (Figure 3).

In accordance with its association with higher serum INH, individuals homozygous for NAT2 low acetylator alleles have been consistently observed to be significantly associated with increased risk of adverse events, particularly hepatotoxicity [30–35].

Although subsequent research has pointed to a probable more complex, multigene sce-nario (Figure 1) [36], SNPs in NAT2 remain the single most important genetic factor modulat-ing INH adverse reactions, its polymorphism presently being the only one predictive of such events [32].

In terms of clinical efficacy, recent work has pointed to the fact that the present ‘one dose fits all’ administration strategy for INH [101] might be insufficient for fast acetylators to fully ben-efit from this drug’s characteristic EBA, while placing the low acetylator subjects at risk of toxicity [20,37].

In conclusion, NAT2 genetics represents a likely important modulator of both the risk of development of adverse events and therapeutic efficacy, supporting the view of using pharmaco-genetic information for a certain degree of the individualization of INH dosing [20].

The molecular genetics of the INH metabo-lism gathers important basic hallmarks for the definition of a ‘clinically translatable’ pharmaco genetic tool:

Signif icant correlations between n NAT2 alleles and a number of associated pheno-types: altered exposure to INH, individual risk of adverse events, as well as of decreased therapeutic efficacy;

The fact that the NAT2 enzyme is the sole n

factor responsible for the acetylation of INH;

The gene has shown low induction capacity (in n

this aspect, like CYP2D6), decreasing the influ-ence of environmental factors in the desirable clear phenotype–genotype relation;

In other practical considerations, exception-n

ally for a human gene, NAT2 does not present introns, enhancing the possibility of analysis of several SNPs simultaneously [24];

The number of SNPs of interest (four) is rela-n

tively limited, with amplifications or other com-plex rearrangements of the gene not documented (unlike, for example, CYP2D6).

Malaria, amodiaquine & CYP2C8 polymorphisms

Malaria & amodiaquinenCurrently, the main malaria treatment in endemic areas is based on combination therapy (ACT). The strategy recommended by the WHO [103], stands in the use of short-acting/fast parasite reduction ratio artemsinin semi-synthetic deriva-tives (e.g., artesunate and artemether), together with a long half-life aminoquinoline anti malarial drug (e.g., amodiaquine, lumefantrine and mefloquine), in a typical 3-day period [38].

The 4-aminoquinoline AQ is one of the central long half-life partners for ACT, when combined with artesunate in a 3-day dose regimen. Recently, a fixed artesunate–amodi-aquine (ASAQ) formulation was introduced (Coarsucam®/Winthrop®, Sanof i-Aventis, Paris, France/Drugs for Neglected Diseases ini-tiative) [104], reinforcing the importance of this combination under public health system use in Africa, where it has been adopted by the national malaria control programs of a significant number of countries [105].

The prophylactic use of AQ was found two decades ago to be associated to rare (1:2000) but serious side effects, including severe neutro-penia [39,40] and hepatic toxicity [40,41]. Subsequent in vitro-based research pointed for AQ toxicity

Other minor metabolites

Urine?

CYP1A1CYP1B1

Extra hepatic?

AQ

CYP2C8(hepatic)

Desethylamodiaquine(DEAQ)

P450s

Myeloperoxidases

Adverse events

Reactivequinoneimine

Figure 2. summary of the metabolism of amodiaquine. Amodiaquine is rapidly transformed towards its main metabolite DEAQ [53]. In parallel with this main pathway, minor parallel pathways are present, including the formation of several minor probably nontoxic metabolites (BisDEAQ and 3-hydroxiDEAQ) [78] and – most importantly – a pathway towards short-lived, highly reactive quinoneimine derivatives [42,79]. Although DEAQ is also prone to be transformed in quinoneimine products [44], it seems to be significantly less prone then AQ. See Figure 4 for relevant chemical structures. AQ: Amodiaquine; DEAQ: Desethylamodiaquine.

www.futuremedicine.com 631future science group

Pharmacogenetic tools for malaria & TB in the Developing World Review

to be linked to the production of a highly reac-tive quinoneimine compound able to disrupt cell functions through irreversible protein bind-ing [42], further associated with the induction of an immune response [43].

CYP2C8 metabolizes AQ towards its active main metabolite desethylamodiaquine (Figure 2), a compound with a significantly decreased ability to generate the highly reactive quino-neimine (Figure 4) [44]. Subjects with less active CYP2C8 would be expected to be at higher risk of adverse events.

The aforementioned prophylaxis toxicity data have prompted the WHO to the controversial decision of removing AQ from its list of rec-ommended antimalarials for the management of uncomplicated malaria in the Developing World for a period in the late 1980s and early 1990s [45].

Although the present view is that the use of AQ in the context of ACT does not involve higher risks of toxicity than the mainstay use of chloroquine (25 mg/kg, 3-day treatment course) [46,47], recent data points to a more complex scenario: in a randomized efficacy trial comparing AQ with ASAQ performed among children in Kenya, Senegal and Gabon, 6% of the involved subjects were found to carry asymptomatic neutropenia at day 28 of the

follow-up [48]. In one case, asymptomatic hepati-tis was detected [49]. In addition, in a recent AQ versus ASAQ pharmaco kinetic study, a signifi-cant fraction of the healthy volunteers developed adverse (although nonsevere) events [50].

CYP2C8nThe human CYP2C8 gene is one of the four members of the CYP2C cluster, together with CYP2C9, CYP2C18 and CYP2C19. The CYP2C8 enzyme is mainly expressed in the liver [51,52]. It is the main enzyme in the hepatic metabolism of AQ [53]. In total, only 15 alleles were described for CYP2C8, reflect-ing a relatively low number of nonsynonymous SNPs documented for this gene (Figure 5), as compared with other polymorphic P450s [106]. The main polymorphisms, CYP2C8*2 (har-boring 805 A>T), CYP2C8*3 (carrying both 416 G>A and 1196 A>G) and CYP2C8*4 (792 C>G) lead to the amino acid changes I269F, R139K, K399R and I264M, respectively. The prevalence of the three major alleles has been suggested to differ significantly between ethnic groups [54].

The *2 and *3 variants have been reported to be associated with significant in vitro differences in the kinetic parameters of AQ as compared with the CYP2C8 ‘wild-type’ (*1) allele [55]. The

Exons Introns

101 bp

5´1

~8.65 Kb 1175 bp

111 T>C(none)

190 C>T(R64W)

191 G>A(R64Q)

364 G>A(D122N)

481 C>T(none)

499 G>A(E167K)

590 G>A(R197Q)

518 A>G(R197Q)

683 C>T(G286E)

766 A>G(G286E)

857 G>A(G286E)

859 T>C(I287T)

859de1T(frameshift)

3´

845 A>G(none)838 G>A(none)

803 A>G(none)

759 C>T(none)

609 G>T(none)

282 C>T(none)

341 T>C(I114T)

403 C>G(L137F)

411 A>T(L137F)

472 A>C(Q145P)434 A>G(Q145P)

8p22

Figure 3. NAT2 gene structure and presently documented sNPs [108]. The main SNPs associated to decreased INH acetylation are denoted in blue [20,28–30]. The respective alleles: I114T (*5), R197Q (*6), G286E (*7) and R64Q (*14) [108]. To note that although the influence of NAT2*14 on in vivo INH metabolism is not formally documented, this variant (frequent in Native African population [75,76]) shows a marked reduced sulfamethazine N-acetyltransferase activity in vitro, compared with NAT2*4 [77]. INH: Isoniazid.

Personalized Medicine (2008) 5(6)632 future science group

Review Ferreira, Cavaco & Gil

DE

AQ

DE

AQ

-qui

none

imin

e

NC

l

N

O

NH

H

H

NC

l

N

O

NH

diA

C-H

Z

AQ

AQ

-qui

none

imin

e

NC

l

N

O

N

H

H

NC

l

N

O

N

INH

Ac-

INH

HZ

INA

N

NO

N

HHH

N

OOH

N

NO

N

H

HO

NN

O

H

H

H

NNH

H

H

H

NN

OH

HO

Ac-

HZ

diA

C-H

Z

Fig

ure

4. r

elev

ant

chem

ical

str

uct

ure

s. A

c-H

Z: A

cety

l hyd

razi

ne;

Ac-

INH

: Ace

tyl i

soni

azid

; AQ

: Am

odi

aqui

ne;

DEA

Q: D

eset

hyla

mo

diaq

uin

e; d

iAc-

HZ:

Dia

cety

l-hy

draz

ine;

H

Z: H

ydra

zin

e; IN

A: I

soni

cotin

ic a

cid

; IN

H: I

soni

azid

.

www.futuremedicine.com 633future science group

Pharmacogenetic tools for malaria & TB in the Developing World Review

enzyme coded by CYP2C8*2 was reported to present a lower V

max and intrinsic clearance, and

a higher Km when compared with CYP2C8*1,

while the *3 enzyme did not show a convinc-ingly detectable metabolism of AQ in this microsomal-based system [55]. The *2 allele was not associated to a significant difference in AQ monotherapy efficacy [55], although a significant increase in mild side effects was detected among the *2 carriers (self-reported rate of abdominal pain). CYP2C8*3 was not found in this popu-lation, despite having been detected in other native African populations [56]. Its carriers, most probably low metabolizers for AQ, are expected to show an increased risk of serious effects of the drug. There is no published data on the AQ metabolism capacity of the *4 allele, a variant found in East Africa [56]; however, its low activ-ity in the in vitro metabolism of the CYP2C8 probe drug paclitaxel [57] points towards *4/*4 carriers being AQ low metabolizers.

From the present stage of knowledge, CYP2C8 shows some promising characteris-tics as a pharmacogenetic marker of practical application: the enzyme is highly specific for AQ, to the extent of this drug having been pro-posed as a probe drug for this P450 isoform [58]; CYP2C8 has low complexity, with a low num-ber of SNPs at over 1% prevalence and no major gene rearrangements. In addition, as *3 is a very low activity protein, induced increases in CYP2C8 expression are not expected to have significant phenotypic effects – that is, it will have negligible influence on the specificity of the test. Although the sequence similarities of CYP2C8 with other members of the CYP2C gene cluster pose challenges for SNP typing, these have already been optimized for PCR-restriction fragment length polymorphism [56], Pyrosequencing [59] and real-time PCR-based methods [60].

Even though the CYP2C8*3 allele shows potential as a pharmacogenetic tool, it is impor-tant to stress the urgent necessity of in vivo data, in order to better understand the impact of this polymorphism in the treatment efficacy/side effects of AQ-based therapies. The avail-able studies do not show a significant influence of the CYP2C8 polymorphism in the efficacy of AQ based therapies, although it is of note that the very low metabolism *3 allele was not present in the East African populations stud-ied [55,61]. Besides, the available in vitro data is sufficiently clear to support the view that *3/*3 (and the rarer *4/*4) carriers are significantly more exposed to AQ.

Is a personalized medicine approach realistic?In the previous sections we have proposed that at least in the case of two genes, CYP2C8 and NAT2, the analysis of a relatively small number of SNPs are likely to bring advantages for the patient, through potential drug dose adjust-ments. This would be expected to decrease the number of adverse effects associated with the use of these drugs, as well as its efficacy in the case of INH. However, is it achievable in the near future to translate this pharmacogenetic knowledge as a set of applicable patient testing procedures in the typical malaria and TB settings?

Artesunate-amodiaquine & malarianA full ACT treatment in Africa costs less than US$5.0 (three daily doses) [62]. The drug itself, taking ASAQ as a reference associated with CYP2C8, will cost US$0.5–1 per treatment (Coarsucam/Winthorp fixed formulation). The analysis of CYP2C8*2, *3 and *4 through PCR-restriction fragment length polymorphism will locally cost an estimated minimum of US$70–100 (approximately two working days at US$10/day/patient, assuming the parallel analysis of at least five patients, plus US$20/SNP for consum-ables). This represents an investment of at least one order of magnitude larger than the treatment itself. In other words, it represents the further treatment of approximately 20 other patients. Even if we limit the analysis to *3 (and to the sole analysis of the K399R SNP), the value will still reach a minimum of US$40–50.

Faced with these costs, it is expected that proj-ects of local CYP2C8 testing will only be appli-cable after total ASAQ coverage of the popula-tion, associated with a significant decrease in the transmission of the disease, the related decreased number of patients allowing the logistic conditions for a more personalized medicine, including the possible introduction of pharmacogenetics tools.

This situation has been achieved in the Zanzibar islands [63]. Through a combined intervention of ACT massive distribution (ASAQ as first-line chemo therapy) and impregnated antimosquito bed nets, the incidence of the disease has dropped from over 106 clinical cases annually, towards approximately 500–1000/year. This reflects a pres-ent incidence of asymptomatic parasite carriers of only approximately 4%. However, a new aspect can be raised: taking in account that severe effects of AQ have been considered to have an incidence of 1:2000 [39], only an average of one patient with adverse events is expected at least every 2 years. Is the allocation of routine funds (>US$25,000/year)

Personalized Medicine (2008) 5(6)634 future science group

Review Ferreira, Cavaco & Gil

and the investment in molecular biology training and facility maintenance worthwhile? Viewed in a harsh realistic way, it would be hard to convince the funding agencies that are a backbone of the massive treatment of these diseases (e.g., WHO, Global Fund, Bill and Melinda Gates Foundation), to be able (and be willing) to significantly increase their investment in order to include these genetic tests. Each test, although representing one event per patient lifetime, is financially equivalent to a large number of drug treatments. In the context of the African communities suffering from several concurrent infectious diseases, the investment of these extra funds in other conditions or issues (e.g., better nutrition) will probably be a more effective strategy.

Isoniazid & TBnIn the case of NAT2, recent reports have stated that the analysis of four SNPs (defining *5, *6, *7 and *14; see Figure 3), would be sufficient for covering over 95% of the slow metabolizer sub-jects. Alternatively, a comprehensive commercial outsourced ‘first pass’ DNA sequencing covering the full gene in just three PCR amplicons is pos-sible [64]. Both procedures would cost approxi-mately US$100–120, depending on the cost of the local human resources.

Treatments in malaria are typically 3-day course programs, with relatively low human resources costs. The situation with TB is

markedly different: the ‘rather bizarrely’ [11] denominated ‘short course’ WHO DOTS reference curative treatment regimens have a minimum duration of 4–6 months, involv-ing significantly higher costs. These routinely overpass values of more than US$200, even in very poor settings [65,66], while we calculate that the analysis of four NAT2 SNPs would cost approximately US$100 (see previous calcula-tions for CYP2C8 analysis). In this case, the application of NAT2 polymorphisms for an optimization of the DOTS strategy might be advantageous. This possibility takes in account the following:

The fact that the pharmacogentic test would n

be a one event in a lifetime and would cost significantly less than a treatment program;

That the frequency of TB drug serious adverse n

events (partially INH-driven) can reach near 30% [3], contributing to low compliance and health costs [67];

That the association of NAT2 slow acetyla-n

tor status with the adverse effects of INH is well established;

The circumstance that TB also affects mid-n

dle-/high-income countries with more devel-oped public health systems and know-how to apply these technologies (e.g., India, China, The Russian Federation and South Africa).

5´

Exons

4472 G>A(G171S)

1 2 3 4 5 6 7 8 9

10q24.1

475delA(159frameshift)

2130 G>A(R139K)

4517 C>T/G(R186X/G)

10979 A>G(I244V) 10989 A>G(K247R)

11054 A>T(I269F)

11041 C>G(I264M)

30411 A>G(K399R)

26513 G>T(K383N)

167 bp 163 bp 150 bp 161 bp 177 bp 142 bp 188 bp 142 bp 537 bp

~1.54 Kb

~0.17 Kb

~2.25 Kb ~6.29 Kb ~12.38 Kb ~2.73 Kb ~3.85 Kb ~1.59 Kb

Introns

32184_32186 delTTG(461delV)

10918 T>G(I223M)

10961 G>C(A238P)

3´

Figure 5. CYP2C8 gene structure and main documented sNPs [106]. The main SNPs associated with documented in vitro and/or in vivo alterations in drug metabolism are denoted in blue [54]. The respective alleles are: I269F (*2), R139K + K399R (*3) and M264I (*4).

www.futuremedicine.com 635future science group

Pharmacogenetic tools for malaria & TB in the Developing World Review

In these settings, the application of a NAT2 pharmacogenetic test and dose adjustment [20,37] is likely to lead not only to a reduction in adverse events-associated costs, but also to an improvement of patient compliance, a pivotal factor against drug resistance development.

Similarly to the previously discussed malaria scenario, in the economically most unfavorable settings – where the average health expenditure is less than US$25/year – the available funds are expected to be invested on the full access to TB DOTS treatment, before any commitment to applied pharmacogenetics.

Pharmacogenetics & the poorest of nthe Developing WorldPharmacogenetics presently has an impor-tant function, but from a population rather than from a personalized point of view. Well-designed studies intended to determine the prevalence of relevant pharmacogenetic markers in a particular population – particularly when taking into account detailed ethnicity informa-tion – can contribute to an optimized massive distribution of drugs, a typical intervention against these diseases. This can contribute to the prediction of the expected frequency of severe adverse effects or a possible decreased efficacy of the drug in specific populations and regions. As an example, CYP2C8*3/*3 carriers, as detailed previously, are expected to have a significant risk of developing agranulocytosis and hepatic reactions upon the administration of AQ. Although such individuals have not been detected in West Africa, our studies have shown that they are present in 0.5–1% of the specific population of the Zanzibar islands, rep-resenting 5000–10,000 individuals. This infor-mation was transmitted to the local malaria control program, as AQ-based ACT is presently the first-line drug in these islands [56].

As for NAT2, this gene has long been known to have highly variable frequencies world-wide [18,68]. For example, the NAT2*14 allele seems specific of African populations, being further present in significantly dissimilar preva-lence in different ethnic groups [69], a result sup-ported by recent studies showing high NAT2 diversity between different sub-Saharan ethnic groups [70,71]. In addition, as an example, the *13 is the main NAT2 allele among the Vietnamese Khin, contrary to that observed in other Asian populations [64].

In fact, it should be noted that although important projects, like the HapMap consor-tium output [107], are pivotal for the knowledge

of human molecular genetic diversity, they tend to over-simplify the multiethnic variation both in qualitative (by missing population-specific polymorphisms) and quantitative terms (by ignoring the large differences in allele frequen-cies between specific populations). This gap is expected to be filled by smaller and more focused studies, as in the previous exemplified works.

An exceptional possibility of real application of personal pharmacogenetics in the Developing World is the circumstance of a limited group of patients associated with specific therapies, where a favorable cost–benefit could be inferred by taking in account potential drug–drug inter-actions. An obvious example is the genotyping of patients simultaneously under AQ-based ACT and the anti-HIV drug efavirenz. A clear drug–drug interaction has been recently iden-tified, leading to significant increases in C

max,

T1/2

and AUC0–96h

of AQ [72]. This set of effects, might be exacerbated if present in a subject with low CYP2C8 activity – for example, a relatively common *2/*2 carrier [55,56] might experience an increase in AQ exposure on par with what is expected to be seen in the situation of the rare *3/*3 genotype.

At any rate, the wide use of anti-infection drugs in these populations implies a demand for the translation of pharmacogenetic research in these areas, not as an ideal personal medicine in every case, but as an aid for pharmacovigilance efforts at a population level.

ConclusionPharmacogenetic testing is still not used as a routine tool for the optimization of chemo-therapy in infectious diseases. The first fully applied example is the recent successful use of HLA-B*5701 detection for the identification of Caucasian patients with immune-mediated hypersensitivity to the HIV nucleoside reverse transcriptase inhibitor abacavir [73]. This suc-cess has raised the confidence in the drug, prompting its increased use in the UK [74]. Similarly, the analysis of NAT2 and CYP2C8 alleles can potentially be a valuable aid for opti-mization of INH and AQ dosing, leading to better treatment compliance.

In our view, NAT2 alleles presently repre-sent the only realistically applicable phramaco-genetic marker, probably confined to medium-income countries. In the Developing World, genotype-driven individualized drug dosing for AQ and INH has to wait for full access of the population to these drugs. However, the pharmacogenetic analysis of populations under

Personalized Medicine (2008) 5(6)636 future science group

Review Ferreira, Cavaco & Gil

drug exposure can be an important aid for deci-sion making in AQ and INH mass distribution programs on these settings.

The application of pharmacogenetics in the scenario of the massive numbers involved in the treatment of infectious diseases have never been addressed in detailed studies for the clear definition of the operational cost–benefit issues of personalized medicine. These urgent studies have to take into account the relation between decision-making factors with relevance for these typical economically unprivileged regions – for example, the cost of the intended test, the avail-able budgets, the number of target patients, the incidence and economical impact of side effects, its health impact and associated economical bur-dens, the size and costs of technology transfer or the prevalence of the marker in question in the specific target region/population.

The sound applicability of pharmaco genetics in TB and malaria is strongly dependent on more research and its integration in the reality of the disease settings of the Developing World.

Future perspectiveNAT2 and CYP2C8 SNPs are promising tools to aid a better application of the anti-TB INH and the antimalarial amodiaquine, respectively. According to the high costs of a full TB DOTS therapeutic program, the analysis of NAT2 for the preview of adverse events/drug efficacy is expected to be immediately applicable in rising economies where this disease still constitutes a public health concern (e.g., China, India and Russian Federation).

CYP2C8 SNPs show potential for the surveil-lance of adverse events related to amodiaquine, but need further research for the establishment

of improved genotype–phenotype associations. The first routine translational towards personal-ized medicine will probably not take place before the next 5–10 years.

In general, pharmacoeconomic studies focused in the particular scenarios of these large infectious diseases are urgently needed, as deci-sional tools are not generally available for the implementation of pharmacogenetic strategies. These are expected to be performed in the near future, as otherwise the pharmacogenetic appli-cation will take significantly more time to occur in these settings.

Importantly, the implementation of pharmaco-gentic markers is only acceptable from the point that the drug access coverage of the population is essentially complete. This is still to be achieved in the poorest regions. The progressive economical evolution of the Developing World, associated with a better control of these diseases, will most likely open up a large number of opportunities for pharmaco genetics in the next two decades, both medically and commercially.

AcknowledgementsWe thank Proffessor Akira Kaneko for several valuable discussions on the manuscript.

Financial & competing interests disclosurePF and IC are supported by Fundação para a Ciência e Tecnologia, Ministério da Ciência, Tecnologia e Ensino Superior, Portugal. The authors have no other relevant affiliations or financial involvement with any organiza-tion or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

executive summary

The translation of pharmacogenetics towards the chemotherapies against malaria and TB, two large global infectious diseases affecting �populations of low resources, is a challenging endeavor.

N � -acetyltransferase 2 (NAT2) SNPs are valuable markers of slow/rapid isoniazid acetylation. Their implementation has high potential for personal drug adjustment, leading to better efficacy among the fast acetylator subjects, while significantly decreasing the risk of isoniazid-associated adverse events among the slow acetylators.

Isoniazid dose adjustment has the potential to significantly decrease the medical/financial costs associated with adverse events, while �increasing compliance. The latter is an important factor against the development of pathogen resistance.

Application of � NAT2 pharmacogenetics in rising economies with significant TB burden and large WHO directly observed treatment short course strategy coverage, including China, India and the Russian Federation, is presently achievable. The cost of analysis (~US$100) is significantly inferior to the cost of treatment in these countries.

CYP2C8 � analysis is presently not worthwhile, as it costs an order of magnitude more than the full amodiaquine artemisinin derivative-based combination therapy treatment (US$1 vs US$50–100).

The determination of the prevalence of the � CYP2C8 and NAT2 main alleles in the isoniazid and amodiaquine target population can be of value in control programs involved in the mass distribution of these drugs.

The full treatment coverage of the target populations has to be guaranteed before effective plans of pharmacogenetic-aided �personalized medicine can be put in place.

www.futuremedicine.com 637future science group

Pharmacogenetic tools for malaria and TB in the Developing World Review

BibliographyPapers of special note have been highlighted as:n of interestnn of considerable interest

Nosten F, White NJ: Artemisinin-based 1

combination treatment of falciparum malaria. Am. J. Trop. Med. Hyg. 77(Suppl. 6), 181–192 (2007).

Cox HS, Morrow M, Deutschmann PW: 2

Long term efficacy of DOTS regimens for tuberculosis: systematic review. Br. Med. J. 336(7642), 484–487 (2008).

Tostmann A, Boeree MJ, Aarnoutse RE, 3

de Lange WC, van der Ven AJ, Dekhuijzen R: Antituberculosis drug-induced hepatotoxicity: concise up-to-date review. J. Gastroenterol. Hepatol. 23(2), 192–202 (2008).

Compact but comprehensive review n

on the most important adverse events of anti-TB therapies, including isoniazid-based regimens.

Taylor WR, White NJ: Antimalarial drug 4

toxicity: a review. Drug Saf. 27(1), 25–61 (2004).

White NJ: Cardiotoxicity of antimalarial 5

drugs. Lancet Infect. Dis. (8), 549–558 (2007).

Dye C, Scheele S, Dolin P, Pathania V, 6

Raviglione MC: Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project. JAMA 282(7), 677–686 (1999).

White NJ: Antimalarial drug resistance.7 J. Clin. Invest. 113(8), 1084–1092 (2004).

Sachs JD, Hotez PJ: Fighting tropical diseases. 8

Science 311(5767), 1521 (2006).

Wu AC, Fuhlbrigge AL: Economic evaluation 9

of pharmacogenetic tests. Clin. Pharmacol. Ther. 84(2), 272–274 (2008).

Grossman I: Routine pharmacogenetic testing 10

in clinical practice: dream or reality? Pharmacogenomics 8(10), 1449–1459 (2007).

Valuable set of comments on the demands n

and difficulties of establishing pharmacogenetic tools.

Harries AD, Dye C: Tuberculosis. 11 Ann. Trop. Med. Parasitol. 100(5–6), 415–431 (2006).

Selikoff IJ, Robitzek EH, Ornstein GG: 12

Treatment of pulmonary tuberculosis with hydrazide derivatives of isonicotinic acid. JAMA 150, 973–980 (1952).

Jindani A, Doré CJ, Mitchison DA: 13

Bactericidal and sterilizing activities of antituberculosis drugs during the first 14 days. Am. J. Respir. Crit. Care Med. 167(10), 1348–1354 (2003).

Mitchison DA: Role of individual drugs in the 14

chemotherapy of tuberculosis. Int. J. Tuberc. Lung Dis. 4(9), 796–806 (2000).

Mitchison DA: Basic mechanisms of 15

chemotherapy. Chest 76(Suppl. 6), 771–781 (1979).

Donald PR, Sirgel FA, Botha FJ 16 et al.: The early bactericidal activity of isoniazid related to its dose size in pulmonary tuberculosis. Am. J. Respir. Crit. Care Med. 156(3 Pt 1), 895–900 (1997).

Donald PR, Sirgel FA, Venter A 17 et al.: The influence of human N acetyltransferase genotype on the early bactericidal activity of isoniazid. Clin. Infect. Dis. 39, 1425–1430 (2004).

Patient nn NAT2 genotype status influences INH early bactericidal activity (EBA), a pivotal factor for its efficacy.

Weber WW: 18 The acetylation genes and drug response. Oxford University Press, NY, USA (1987).

Peloquin CA, Jaresko GS, Yong CL, 19

Keung AC, Bulpitt AE, Jelliffe RW: Population pharmacokinetic modeling of isoniazid, rifampin, and pyrazinamide. Antimicrob. Agents Chemother. 41(12), 2670–2679 (1997).

Kinzig-Schippers M, Tomalik-Scharte D, 20

Jetter A et al.: Should we use N acetyltransferase type 2 genotyping to personalize isoniazid doses? Antimicrob. Agents Chemother. 49(5), 1733–1738 (2005).

Parkin DP, Vandenplas S, Botha FJ 21 et al.: Trimodality of isoniazid elimination, phenotype and genotype in patients with tuberculosis. Am. J. Respir. Crit. Care Med. 155, 1717–1722 (1997).

Weiner MW, Burman A, Vernon D 22 et al.: Low isoniazid concentrations and outcome of tuberculosis treatment with once-weekly isoniazid and rifapentine. Am. J. Respir. Crit. Care Med. 167, 1341–1347 (2003).

Hughes HB: On the metabolic fate of 23

isoniazid. J. Pharmacol. Exp. Therap. 109, 444–452 (1953).

Blum M, Grant DM, McBride W, Heim M, 24

Meyer UA: Human arylamine N-acetyltransferase genes: isolation, chromosomal localization, and functional expression. DNA Cell Biol. 9(3), 193–203 (1990).

Zang Y, Doll MA, Zhao S, States JC, 25

Hein DW: Functional characterization of single-nucleotide polymorphisms and haplotypes of human N-acetyltransferase 2. Carcinogenesis 28(8), 1665–1671 (2007).

Leff MA, Fretland AJ, Doll MA, Hein DW: 26

Novel human N-acetyltransferase 2 alleles that differ in mechanism for slow acetylator phenotype. J. Biol. Chem. 274, 34519–34522 (1999).

Deguchi T, Mashimo M, Suzuki T: 27

Correlation between acetylator phenotypes and genotypes of polymorphic arylamine N-acetyltransferase in human liver. J. Biol. Chem. 265(22), 12757–12760 (1990).

Kita T, Tanigawara Y, Chikazawa S 28 et al.: N-acetyltransferase2 genotype correlated with isoniazid acetylation in Japanese tuberculous patients. Biol. Pharm. Bull. 24(5), 544–549 (2001).

Chen B, Li JH, Xu YM, Wang J, Cao XM: 29

The influence of NAT2 genotypes on the plasma concentration of isoniazid and acetylisoniazid in Chinese pulmonary tuberculosis patients. Clin. Chim. Acta 365(1–2), 104–108 (2006).

Hiratsuka M, Kishikawa Y, Takekuma Y 30

et al.: Genotyping of the N acetyltransferase 2 polymorphism in the prediction of adverse drug reactions to isoniazid in Japanese patients. Drug Metab. Pharmacokinet. 17(4), 357–362 (2002).

Weber WW, Hein DW: 31 N-acetylation pharmacogenetics. Pharmacol. Rev. 37, 25–79 (1985).

Cho HJ, Koh WJ, Ryu YJ 32 et al.: Genetic polymorphisms of NAT2 and CYP2E1 associated with antituberculosis drug-induced hepatotoxicity in Korean patients with pulmonary tuberculosis. Tuberculosis 87(6), 551–556 (2007).

Huang YS, Chern HD, Su WJ 33 et al.: Polymorphism of the N-acetyltransferase 2 gene as a susceptibility risk factor for antituberculosis drug-induced hepatitis. Hepatology 35(4), 883–889 (2002).

Possuelo LG, Castelan JA, de Brito TC 34 et al.: Association of slow N-acetyltransferase 2 profile and anti-TB drug-induced hepatotoxicity in patients from Southern Brazil. Eur. J. Clin. Pharmacol. 64(7), 673–681 (2008).

Ohno M, Yamaguchi I, Yamamoto I 35 et al.: Slow N-acetyltransferase 2 genotype affects the incidence of isoniazid and rifampicin-induced hepatotoxicity. Int. J. Tuberc. Lung Dis. 4(3), 256–261 (2000).

Huang YS, Chern HD, Su WJ 36 et al.: Cytochrome P450 2E1 genotype and the susceptibility to antituberculosis drug-induced hepatitis. Hepatology 37(4), 924–930 (2003).

Kubota R, Ohno M, Hasunuma T, Iijima H, 37

Azuma J: Dose-escalation study of isoniazid in healthy volunteers with the rapid acetylator genotype of arylamine N-acetyltransferase 2. Eur. J. Clin. Pharmacol. 63(10), 927–933 (2007).

Deen JL, von Seidlein L, Dondorp A: 38

Therapy of uncomplicated malaria in children: a review of treatment principles,

Personalized Medicine (2008) 5(6)638 future science group

Review Ferreira, Cavaco & Gil

essential drugs and current recommendations. Trop. Med. Int. Health 13 (9), 1111–1130 (2008).

Hatton CS, Peto TE, Bunch C 39 et al.: Frequency of severe neutropenia associated with amodiaquine prophylaxis against malaria. Lancet 1(8478), 411–414 (1986).

Neftel KA, Woodtly W, Schmid M, 40

Frick PG, Fehr J: Amodiaquine induced agranulocytosis and liver damage. Br. Med. J. 292 (6522), 721–723 (1986).

Larrey D, Castot A, Pëssayre D 41 et al.: Amodiaquine-induced hepatitis. A report of seven cases. Ann. Intern. Med. 104(6), 801–803 (1986).

Jewell H, Maggs JL, Harrison AC, 42

O’Neill PM, Ruscoe JE, Park BK: Role of hepatic metabolism in the bioactivation and detoxification of amodiaquine. Xenobiotica 25, 199–217 (1995).

Clarke JB, Maggs JL, Kitteringham NR, 43

Park BK: Immunogenicity of amodiaquine in the rat. Int. Arch. Allergy Appl. Immunol. 91(4), 335–342 (1990).

Tingle MD, Jewell H, Maggs JL, 44

O’Neill PM, Park BK: The bioactivation of amodiaquine by human polymorphonuclear leucocytes in vitro: chemical mechanisms and the effects of fluorine substitution. Biochem. Pharmacol. 50, 1113–1119 (1995).

Seminal work pointing to n

desethylamodiaquine formation as a detoxification path.

Olliaro P, Nevill C, LeBras J 45 et al.: Systematic review of amodiaquine treatment in uncomplicated malaria. Lancet 348(9036), 1196–1201 (1996).

Bloland PB, Ruebush TK: Amodiaquine. 46

Lancet 348(9042), 1659–1660 (1996).

Olliaro P, Mussano P: Amodiaquine for 47

treating malaria. Cochrane Database Syst. Rev. (2), CD000016 (2003).

Adjuik M, Agnamey P, Babiker A 48 et al.: Amodiaquine-artesunate versus amodiaquine for uncomplicated Plasmodium falciparum malaria in African children: a randomized, multicentre trial. Lancet 359(9315), 1365–1372 (2002).

Orrell C, Taylor WR, Olliaro P: 49

Acute asymptomatic hepatitis in a healthy normal volunteer exposed to 2 oral doses of amodiaquine and artesunate. Trans. R. Soc. Trop. Med. Hyg. 95(5), 517–518 (2001).

Orrell C, Little F, Smith P 50 et al.: Pharmacokinetics and tolerability of artesunate and amodiaquine alone and in combination in healthy volunteers. Eur. J. Clin. Pharmacol. 64(7), 683–690 (2008).

Enayetallah AE, French RA, Thibodeau MS, 51

Grant DF: Distribution of soluble epoxide hydrolase and of cytochrome P450 2C8, 2C9, and 2J2 in human tissues. J. Histochem. Cytochem. 52(4), 447–454 (2004).

Klose TS, Blaisdell JA, Goldstein JA: 52

Gene structure of CYP2C8 and extrahepatic distribution of the human CYP2Cs. J. Biochem. Mol. Toxicol. 13(6), 289–295 (1999).

Li XQ, Bjorkman A, Andersson TB, 53

Ridderstrom M, Masimirembwa CM: Amodiaquine clearance and its metabolism to N-desethylamodiaquine is mediated by CYP2C8: a new high affinity and turnover enzyme-specific probe substrate. J. Pharmacol. Exp. Ther. 300(2), 399–407 (2002).

Discovery of CYP2C8 as the main nn

metabolizer of amodiaquine.

Gil J, Gil Berglund E: CYP2C8 and 54

antimalaria drug efficacy. Pharmacogenomics, 8(2), 187–198 (2007).

Parikh S, Ouedraogo JB, Goldstein JA, 55

Rosenthal PJ, Kroetz DL: Amodiaquine metabolism is impaired by common polymorphisms in CYP2C8: implications for malaria treatment in Africa. Clin. Pharmacol. Ther. 82(2), 197–203 (2007).

First report associating the nn CYP2C8 polymorphism with adverse events of amodiaquine therapy in malaria patients.

Cavaco I, Strömberg-Nörklit J, Kaneko A 56

et al.: CYP2C8 polymorphism frequencies among malaria patients in Zanzibar. Eur. J. Clin. Pharmacol. 61(1), 15–18 (2005).

First report of n CYP2C8 allele frequencies among African native populations, alerting for the presence of the potentially low amodiaquine metabolism alleles *3 and *4.

Dai D, Zeldin DC, Blaisdell JA 57 et al.: Polymorphisms in human CYP2C8 decrease metabolism of the anticancer drug paclitaxel and arachidonic acid. Pharmacogenetics 11(7), 597–607 (2001).

Walsky RL, Obach RS: Validated assays for 58

human cytochrome P450 activities. Drug Metab. Dispos. 32(6), 647–660 (2004).

Hruska MW, Frye RF, Langaee TY: 59

Pyrosequencing method for genotyping cytochrome P450 CYP2C8 and CYP2C9 enzymes. Clin. Chem. 50(12), 2392–2395 (2004).

Weise A, Grundler S, Zaumsegel D 60 et al.: Development and evaluation of a rapid and reliable method for cytochrome P450 2C8 genotyping. Clin. Lab. 50(3–4), 141–148 (2004).

Adjei GO, Kristensen K, Goka BQ 61

et al.: Effect of concomitant artesunate administration and CYP2C8 polymorphisms

on the pharmacokinetics of amodiaquine in Ghanaian children with uncomplicated malaria. Antimicrob. Agents Chemother. PMID: 18779360 (2008) (Epub ahead of print).

Shillcutt S, Morel C, Goodman C 62 et al.: Cost–effectiveness of malaria diagnostic methods in sub-Saharan Africa in an era of combination therapy. Bull. World Health Organ. 86(2), 101–110 (2008).

Bhattarai A, Ali AS, Kachur SP 63 et al.: Impact of artemisinin-based combination therapy and insecticide-treated nets on malaria burden in Zanzibar. PLoS Med. 4(11), E309 (2007).

Cavaco I, Asimus S, Peyrard-Janvid M 64 et al.: The Vietnamese Khin population harbours particular N acetyltransferase 2 allele frequencies. Clin. Chem. 53(11), 1977–1979 (2007).

Okello D, Floyd K, Adatu F, Odeke R, 65

Gargioni G: Cost and cost–effectiveness of community-based care for tuberculosis patients in rural Uganda. Int. J. Tuberc. Lung Dis. 7(9 Suppl. 1), S72–S79 (2003).

Floyd K, Skeva J, Nyirenda T, Gausi F, 66

Salaniponi F: Cost and cost–effectiveness of increased community and primary care facility involvement in tuberculosis care in Lilongwe District, Malawi. Int. J. Tuberc. Lung Dis. 7(9 Suppl. 1), S29–S37 (2003).

Roy PD, Majumder M, Roy B: 67

Pharmacogenomics of anti-TB drugs-related hepatotoxicity. Pharmacogenomics 9(3), 311–321 (2008).

Patin E, Barreiro LB, Sabeti PC 68 et al.: Deciphering the ancient and complex evolutionary history of human arylamine N-acetyltransferase genes. Am. J. Hum. Genet. 78(3), 423–436 (2006).

Cavaco I, Reis R, Gil JP, Ribeiro V: 69

CYP3A4*1B and NAT2*14 alleles in a native African population. Clin. Chem. Lab. Med. 41(4), 606–609 (2003).

Patin E, Harmant C, Kidd KK 70 et al.: Sub-Saharan African coding sequence variation and haplotype diversity at the NAT2 gene. Hum. Mutat. 27(7), 720 (2006).

Sabbagh A, Langaney A, Darlu P, Gérard N, 71

Krishnamoorthy R, Poloni ES: Worldwide distribution of NAT2 diversity: implications for NAT2 evolutionary history. BMC Genet. 9, 21 (2008).

German P, Greenhouse B, Coates C 72 et al.: Hepatotoxicity due to a drug interaction between amodiaquine plus artesunate and efavirenz. Clin. Infect. Dis. 44(6), 889–891 (2007).

First report on drug–drug interactions nn

between amodiaquine and an anti-HIV therapy drug.

www.futuremedicine.com 639future science group

Pharmacogenetic tools for malaria and TB in the Developing World Review

Mallal S, Phillips E, Carosi G 73 et al.: PREDICT-1 Study Team. HLA-B*5701 screening for hypersensitivity to abacavir. N. Engl. J. Med. 358(6), 568–579 (2008).

Ingelman-Sundberg M: Pharmacogenomic 74

biomarkers for prediction of severe adverse drug reactions. N. Engl. J. Med. 358(6), 637–639 (2008).

Cavaco I, Gil JP, Gil-Berglund E, Ribeiro V: 75

CYP3A4 and MDR1 alleles in a Portuguese population. Clin. Chem. Lab. Med. 41(10), 1345–1350 (2003).

Sabbagh A, Langaney A, Darlu P, Gérard N, 76

Krishnamoorthy R, Poloni ES: Worldwide distribution of NAT2 diversity: implications for NAT2 evolutionary history. BMC Genet. 9, 21 (2008).

Leff MA, Fretland AJ, Doll MA, Hein DW: 77

Novel human N acetyltransferase 2 alleles that differ in mechanism for slow acetylator phenotype. J. Biol. Chem. 274 (49), 34519–34522 (1999).

Mount DL, Patchen LC, Nguyen-Dinh P, 78

Barber AM, Schwartz IK, Churchill FC: Sensitive analysis of blood for amodiaquine and three metabolites by high-performance liquid chromatography with electrochemical detection. J. Chromatogr. 383(2), 375–386 (1986).

Harrison AC, Kitteringham NR, 79

Clarke JB, Park BK: The mechanism of bioactivation and antigen formation of amodiaquine in the rat. Biochem. Pharmacol. 43(7), 1421–1430 (1992).

websitesWorld Health Organization. The WHO 101

Global Tuberculosis Program www.who.int/gtb

World Health Organization (2006). WHO 102

guidelines for the treatment of malaria www.who.int/malaria/docs/TreatmentGuidelines2006.pdf

World Health Organization (2006). 103

Facts on ACT (artemisinin based combinantions) www.who.int/malaria/wmr2008/malaria2008.pdf

Sanofi Aventis: Coarsucam™ 104

(artesunate/amodiaquine) first fixed-dose antimalarial combination to receive WHO Prequalification www.webdisclosure.com/news/45008.html

WHO treatment policies 105

www.who.int/malaria/treatmentpolicies.html

Home Page of the Human Cytochrome P450 106

(CYP) Allele Nomenclature Committee www.cypalleles.ki.se

International HapMap Project 107

www.hapmap.org

Consensus arylamine N-acetyltransferase 108

(NAT) gene nomenclature www.louisville.edu/medschool/pharmacology/NAT.html