311

ReseaRch aRticle

ISSN 1462-241610.2217/14622416.10.2.311 © 2009 Future Medicine Ltd Pharmacogenomics (2009) 10(2), 311–318

ABCB1 polymorphisms and the concentrations of lopinavir and ritonavir in blood, semen and saliva of HIV-infected men under antiretroviral therapy

Protease inhibitors are frequently included in highly active antiretroviral therapy (HAART) regimens used worldwide in individuals infected with HIV. Protease inhibitors are substrates of, and may also inhibit and/or induce the P-glycoprotein (ABCB1) transporter, encoded by the poly morphic ABCB1 gene [1,2]. The impact of ABCB1 polymorphisms on the pharmaco-kinetics of protease inhibitors has been investi-gated in several studies, with conflicting results (reviewed in [3]). Regarding lopinavir, which is the main object of the present study, two studies detected no influence of the ABCB1 2677G>T/A and/or 3435C>T genotype on the trough con-centration of lopinavir in blood plasma during chronic treatment of HIV-infected patients with ritonavir-boosted lopinavir [4,5]. These results were recently confirmed and extended to other exonic polymorphisms [6]. The present study expands these previous observations in two ways: first, to a three- to four-fold larger sample of HIV-infected subjects under stable lopinavir-containing HAART, and second, by measuring the concentrations of both lopina-vir and ritonavir not only in blood plasma, but

also in semen and saliva. Transmission of HIV occurs primarily via the sexual route, and a sub-stantial proportion of individuals continue to shed the virus in the semen despite HAART treatment [7]. Importantly, not only are these patients infectious, but the virus they transmit is potentially drug-resistant. Antiretroviral drug concentration within the genital tract is related to the effectiveness of HAART to reduce viral load in semen [8,9]. The concentrations of lopi-navir and ritonavir in semen display marked interindividual variability [10–15] and it was sug-gested that polymorphisms in ABCB1 may con-tribute to such variability [10]. Quantification of lopinavir and ritonavir in saliva is of potential interest for the purpose of therapeutic drug monitoring, because of the ease, noninvasive-ness and reduced risk of infection associated with saliva collection.

Materials & methodsPatients & sampling��

HIV-infected Brazilian adult males (n = 113; age: median 42.5 years; range: 28–68 years; weight: 72.2 kg, 52–131 kg; height: 172 cm,

Introduction: Lopinavir and ritonavir are frequently included in highly active antiretroviral therapy (HAART) regimens for HIV infection. These drugs are substrates, and may also inhibit and/or induce the P‑glycoprotein (ABCB1) transporter, encoded by the polymorphic ABCB1 gene. We investigated the impact of three common exonic ABCB1 polymorphisms on the concentrations of lopinavir and ritonavir in blood, semen and saliva of HIV‑infected men under stable HAART containing ritonavir‑boosted lopinavir. Materials & methods: Blood, semen and saliva samples were collected from 113 subjects, 30–35 minutes before the scheduled morning dose of lopinavir/ritonavir, and trough drug concentrations were measured using LC/MS/MS. The 1236C>T, 2677G>T/A and 3435C>T polymorphisms were genotyped using the single base extension–termination method and ABCB1 haplotypes were statistically inferred. Results: Median (25th–75th percentile) trough concentrations (ng/ml) of lopinavir in plasma, semen and saliva were 6326 (4070–8617), 286.0 (128.4–475.5) and 72.7 (38.0–119.6), respectively. The corresponding concentrations (ng/ml) for ritonavir were 261.8 (172.2–398.6), 17.7 (9.2–27.6) and 5.3 (3.2–9.0), respectively. Univariate and multivariate regression analysis revealed no influence of ABCB1 genotypes or haplotypes on the concentrations of lopinavir and ritonavir in plasma, semen and saliva of HIV‑infected men under stable HAART treatment. Conclusion: The ABCB1 1236C>T, 2667G>T/A and 3435C>T genotypes and haplotypes are not predictors of lopinavir and ritonavir concentrations in blood plasma, semen or saliva of HIV‑infected men under stable HAART treatment. The concentrations of lopinavir and ritonavir in saliva are not reliable predictors of the concentration of these drugs in semen.

KEYWORDS: ABCB1, antiretroviral therapy, Brazil, HAART, Iopinavir, P‑glycoprotein, protease inhibitors, ritonavir

Rita de Cassia Estrela1, Fabio S Ribeiro1, Paulo F Barroso2, Mari Tuyama2, Sheila P Gregório3, Emmanuel Dias-Neto3, Cláudio J Struchiner4 & Guilherme Suarez-kurtz1†

†Author for correspondence:1Divisão de Farmacologia, Instituto Nacional de Câncer, Rua André Cavalcanti 37, Rio de Janeiro 21230-050, BrazilTel.: +55 21 3233 1310;Fax: +55 21 3233 1340;kurtz@inca.gov.br2Hospital Universitário Clementino Fraga Filho, Universidade Federal do Rio de Janeiro, Brazil3Faculdade de Medicina da Universidade de São Paulo, Brazil4Fundação Oswaldo Cruz, Rio de Janeiro, Brazil

Pharmacogenomics (2009) 10 (2)312 future science group

ReseaRch aRticle Estrela, Ribeiro, Barroso et al.

158–196 cm) on a stable HAART containing lopinavir/ritonavir for at least 4 weeks were recruited at Hospital Universitário Clementino Fraga Filho, School of Medicine from the Universidade Federal do Rio de Janeiro (Rio de Janeiro, Brazil), between July 2005 and June 2006. The study protocol and written consent forms were approved by the institu-tion’s Ethics Committee. Patients self-identified according to the criteria adopted by the Brazilian Census [101] as white (n = 47), ‘Pardo’ (meaning ‘brown’, here denoted as intermediate, n = 48), black (n = 17) and Amerindian (n = 1). Ritonavir-boosted lopinavir soft-gel capsules (Kaletra®, Abbott Laboratórios do Brasil Ltda., São Paulo, Brazil) were administered at standard doses of 400 mg and 100 mg, respectively, twice-daily in association with other antiretroviral agents as part of their HAART regimen. Only four indi-viduals were receiving Kaletra monotherapy. All others were prescribed nucleoside combinations, including zidovudine, lamivudine, didanosine, stavudine, tenofovir and/or abacavir. In addi-tion, seven individuals had saquinavir added to their HAART, and another one was receiving enfurvitide. Patients were excluded if they were prescribed efavirenz and nevirapine, due to the known interactions between these drugs and lopinavir. Following informed consent, enrolled individuals had a baseline interview and a geni-tal exam. Semen, saliva and blood samples were collected from each individual on one occasion, 30–35 min before the scheduled morning dose of lopinavir/ritonavir, for quantification of the trough (predosing) drug concentrations. Semen was collected by masturbation at the study site after at least 48 h of sexual abstinence, and processed in the laboratory within 30 min of collection. After liquefaction, seminal plasma was obtained by centrifugation at 1400 g for 15 min. Unstimulated saliva was collected by spitting into a vial. Blood samples were drawn from a peripheral vein, and cell free plasma was obtained by centrifugation at 900 g for 10 min. Seminal plasma (semen), saliva and blood plasma were stored at -70°C.

ABCB1�� genotypingABCB1 genotypes were determined using a previ-ously described single base extension–termination method on the SNaPshot® multiplex system from Applied Biosystems (CA, USA) [16]. The ABCB1 haplotypes were statistically inferred by the haplo.stats software, version 1.3 [17,102]. This software attributes a posterior probability value for the diplotype configuration for each individual, based

on estimated haplotype frequencies. Diplotypes were inferred with probabilities greater than 0.95 for all individuals.

Quantification of lopinavir & ��ritonavir in blood, semen & salivaDrug quantification was performed by high-performance liquid chromatography with posi-tive ion electrospray ionization tandem mass spectrometry detection (liquid chromatography/mass spectrometry/mass spectrometry [LC/MS/MS]), using procedures previously validated for blood plasma, semen and saliva [18]. The limit of quantitation was 1 ng/ml for both analytes in all matrices.

Statistical analysis��The χ2 test was used for comparing genotype and haplotype frequencies at the polymorphic ABCB1 loci and for assessing deviations from Hardy–Weinberg equilibrium. Because of the very low frequency (~1%) of 2677A among Brazilians [16], alleles 2677A and 2677T were grouped as ‘nonG’ in the statistical analyses. Kruskal–Wallis testing was used to examine differences in lopinavir and ritonavir trough concentrations among the three genotypes at each ABCB1 locus and among individuals having zero, one or two copies of the referent CGC haplotype at the 1236, 2677 and 3435 loci. The Mann–Whitney U test was used to assess pairwise differences in drug concentrations. Pairwise correlations between drug concen-trations were tested using the Spearman rank correlation coefficient. Linear regression analy-sis was used to assess the association between ABCB1 polymorphisms and the concentrations of lopinavir and ritonavir in plasma, semen and saliva. First, we fitted a univariate ordinary least squares regression to assess whether the ABCB1 1236C>T, 2677G>T/A and 3435C>T genotypes or their haplotypes showed a significant effect on the outcome variable. Following that, we fitted a multivariate ordinary least squares regression containing just the main effects for each vari-able – namely ABCB1 genotypes and haplotypes, self-reported ‘race/color’, weight, height, age, saquinavir coadministration and, in the case of semen and saliva, the respective drug concentra-tion (lopinavir or ritonavir) in blood plasma – but without any of the inter action terms, given the large number of parameters relative to the sample size of our observations. Statistical analyses were performed using the R statistical package [19,103].

A p-value smaller than 0.05 was considered to be significant.

www.futuremedicine.com 313future science group

ABCB1 polymorphisms & antiretroviral therapy ReseaRch aRticle

The statistical power to detect differences in the mean trough concentrations of lopinavir and ritonavir in blood plasma and in semen accord-ing to the ABCB1 genotypes, was calculated online [104] using the distribution of the 1236C>T, 2667G>T/A and 3435C>T genotypes in the population of Rio de Janeiro [15] and our previous data [8] on the trough concentrations of lopina-vir and ritonavir in blood and semen of HIV+ Brazilian individuals under stable treatment with Kaletra. For a sample of n = 100, the percent-age differences in lopinavir concentrations that could be detected at a two-sided a level of 0.05 and statistical power of 0.80, between carriers (hetero zygous and homozygous) and noncarriers of variant alleles at each ABCB1 polymorphic site were estimated as 18–23% (plasma) and 68–72% (semen). The corresponding values for ritonavir were 22–27% and 35–45%. The lack of data on the trough concentrations of lopinavir and ritona-vir in saliva during stable treatment with Kaletra precluded prospective calculation of the statistical power of the present study. Based on the concen-trations measured in this study, the estimated per-cent differences in lopinavir and ritonavir concen-trations in saliva detectable at a two-sided a level of 0.05 and statistical power of 0.80 were 39–43% (lopinavir) and 42–48% (ritonavir).

ResultsA total of 113 HIV-infected men were enrolled in the study and the trough concentrations of lopinavir and ritonavir were determined in all blood plasma samples. A total of 14 individu-als with lopinavir plasma concentration below 200 ng/ml were deemed noncompliant [20] and excluded from subsequent analyses, which were carried out on 99 individuals. From these, five

saliva samples and two semen samples were not available for quantification of the drug concen-trations, and two additional semen samples were below the limit of quantitation for ritonavir.

Quantification of lopinavir & ��ritonavir concentrations in blood, semen & salivaThe lopinavir and ritonavir trough concentra-tions in blood plasma, semen and saliva dis-played large interindividual variability (Table 1). Drug concentrations in semen and saliva were several fold lower than in blood plasma, con-sistent with poor penetration of lopinavir and ritonavir in the male genital tract and salivary glands. For both drugs, the median concen-tration in semen was nearly threefold higher than in saliva (p < 0.0001, Mann–Whitney U test). Statistically signif icant, but weak to moderate correlations (r2 = 0.09–0.29) were observed between the concentrations of each protease inhibitor in plasma versus semen or saliva, and between semen and saliva (Table 2). Lopinavir and ritonavir trough concentrations were correlated in blood plasma (r2 = 0.74, p < 0.0001), semen (r2 = 0.87, p < 0.004) and saliva (r2 = 0.89, p < 0.0001).

ABCB1�� polymorphisms & distribution of lopinavir & ritonavir in blood, semen & salivaThe genotype distribution of the ABCB1 poly-morphisms are shown in Table 3. Genotype fre-quencies did not deviate from Hardy–Weinberg proportions and did not differ significantly from our previous data on healthy Brazilians living in the same geographical region [16]. The distribution of the lopinavir and ritonavir

Table 1. Concentrations of lopinavir and ritonavir in blood plasma, saliva and semen of HIV‑infected men.

Lopinavir (ng/ml) Ritonavir (ng/ml)

Plasma Saliva Semen Plasma Saliva Semen

n 99 94 97 99 94 95

25th percentile 3831 38.0 128.4 152.8 3.2 9.2

Median 6326 72.7 286.1 261.8 5.3 17.7

75th percentile 8617 119.6 503.9 422.0 9.1 27.9

Table 2. Pairwise correlation of the concentrations of lopinavir and ritonavir in plasma, saliva and semen.

Matrices correlation Lopinavir Ritonavir

r2 p-value r2 p-value

Plasma:saliva 0.25 <0.001 0.21 <0.001

Plasma:semen 0.27 <0.001 0.29 <0.001

Saliva:semen 0.09 0.004 0.15 <0.001

Pharmacogenomics (2009) 10 (2)314 future science group

ReseaRch aRticle Estrela, Ribeiro, Barroso et al.L

op

inav

ir p

lasm

a (n

g/m

l)L

op

inav

ir p

lasm

a (n

g/m

l)L

op

inav

ir p

lasm

a (n

g/m

l)

CC CT TT CC CT TT CC CT TT

CC CT TT CC CT TT CC CT TT

Lo

pin

avir

sal

iva

(ng

/ml)

Lo

pin

avir

sal

iva

(ng

/ml)

Lo

pin

avir

sal

iva

(ng

/ml)

1236C>T

Lo

pin

avir

sem

en (

ng

/ml)

Lo

pin

avir

sem

en (

ng

/ml)

Lo

pin

avir

sem

en (

ng

/ml)

2677G>T/A

3435C>T

GG GnonG nonGnonG GG GnonG nonGnonG GG GnonG nonGnonG

20,000

15,000

10,000

5000

0

400

300

200

100

0

400

300

200

100

0

400

300

200

100

0

4000

3000

2000

1000

0

4000

3000

2000

1000

0

4000

3000

2000

1000

0

20,000

15,000

10,000

5000

0

20,000

15,000

10,000

5000

0

A

B

C

concentrations in plasma, semen and saliva among individuals having zero, one or two variant alleles at each ABCB1 locus is plotted in Figures 1 & 2. The data reveal extensive overlap of the concentrations of either protease inhibi-tor in each matrix, across the three genotypes at each locus. Univariate analysis of these data revealed no statistically significant association

between lopinavir or ritonavir concentrations in each matrix and ABCB1 genotype at each locus. Univariate analysis also showed no significant relation between the trough concentrations of lopinavir and ritonavir in plasma, semen or saliva and the ABCB1 haplo types (Table 4). Multivariate regression analysis (Materials & methods) failed to reveal significant association

Table 3. Distribution of ABCB1 genotype in HIV‑infected individuals.

Polymorphisms ABCB1 genotype frequency χ2

1236C>T CC CT TT

0.53 (0.43–0.63) 0.35 (0.27–0.45) 0.12 (0.06–0.36) – 2.34*

2677G>T/A GG GT GA TT

0.60 (0.51–0.69) 0.34 (0.25–0.43) 0.02 (0.002–0.06) 0.04 (0.01–0.10) 0.58*

3435C>T CC CT TT

0.40 (0.31–0.50) 0.48 (0.38–0.57) 0.12 (0.07–0.20) – 0.13*

n = 99. Data are expressed as mean (95% confidence interval). The 2677TA and 2677AA genotypes were not detected in the study population.*Nonsignificant χ2 values

Figure 1. Scatter plot of the distribution of trough concentrations of lopinavir in blood plasma (left panels), saliva (center) and semen (right) among 99 HIV‑infected subjects, according to ABCB1 genotypes at the 1236C>T (A), 2677G>T/A (B) and 3435C>T (C) loci. The lines denote the median values for each genotype.

www.futuremedicine.com 315future science group

ABCB1 polymorphisms & antiretroviral therapy ReseaRch aRticle

Rit

on

avir

pla

sma

(ng

/ml)

Rit

on

avir

pla

sma

(ng

/ml)

Rit

on

avir

pla

sma

(ng

/ml)

CC CT TT CC CT TT CC CT TT

CC CT TT CC CT TT CC CT TT

Rit

on

avir

sal

iva

(ng

/ml)

Rit

on

avir

sal

iva

(ng

/ml)

Rit

on

avir

sal

iva

(ng

/ml)

1236C>T

Rit

on

avir

sem

en (

ng

/ml)

Rit

on

avir

sem

en (

ng

/ml)

Rit

on

avir

sem

en (

ng

/ml)

2677G>T/A

3435C>T

GG GnonG nonGnonG GG GnonG nonGnonG GG GnonG nonGnonG

30

20

10

0

30

20

10

0

30

20

10

0

100

75

50

25

0

A

B

C

1250

1000

750

500

0

250

1250

1000

750

500

0

250

1250

1000

750

500

0

250

30

20

10

0

100

75

50

25

0

100

75

50

25

0

between ABCB1 genotypes or haplotypes and lopinavir or ritonavir concentrations in plasma, semen or saliva.

DiscussionThis study confirms and extends previous obser-vations from our group [10] and others [6,12–14] regarding the marked interindividual variability in the trough concentrations of lopinavir and ritonavir in blood and seminal plasma, and the limited penetration of these protease inhibitors into the genital compartment of HIV-infected men. The observed range of concentrations of lopinavir and ritonavir in plasma and semen, as well as the ratio of these concentrations are consistent with previous data for HIV-infected subjects [10,12–14]. In addition, our results pro-vide the first demonstration of the limited dis-tribution of lopinavir and ritonavir into saliva

during stable therapy with Kaletra. Indeed, the concentrations of each protease inhibitor in saliva averaged less than 2.5% of the respec-tive plasma concentrations. For comparison, the median concentrations of lopinavir and riton-avir in semen reached 4.6 and 6.4% of their respective blood concentrations.

Several factors may combine to account for the low concentrations of lopinavir and ritona-vir in saliva and semen, compared with blood plasma. Prominent among these factors is the extensive (>99%) binding of these protease inhibitors to plasma proteins, especially 1-acid glycoprotein [21], which limits drug transport across cell membranes and into extravascular compartments. Another potential modulator of the distribution of lopinavir and ritonavir in saliva and semen is the ABCB1 drug transporter, of which protease inhibitors are substrates, as

Figure 2. Scatter plot of the distribution of the trough concentrations of ritonavir in blood plasma (left panels), saliva (center) and semen (right) among 99 HIV‑infected subjects, according to ABCB1 genotypes at the 1236C>T (A), 2677G>T/A (B) and 3435C>T (C) loci. The lines denote the median values for each genotype.

Pharmacogenomics (2009) 10 (2)316 future science group

ReseaRch aRticle Estrela, Ribeiro, Barroso et al.

well as inhibitors and/or inducers [1,2,22,23]. Our data show no significant association between the ABCB1 1236C>T, 2667G>T/A and 3435C>T genotypes or haplotypes and the trough con-centrations of lopinavir and ritonavir in blood plasma, semen or saliva of HIV-infected men, under stable Kaletra treatment. These negative results were confirmed when examined by lin-ear regression analysis adjusting for potential demographic variables (self-reported ‘race/color’, age, weight and height), saquinavir coadmin-istration plus, in the case of semen and saliva, the respective drug concentration (lopinavir or ritonavir) in blood plasma. A recent article, pub-lished after completion of the present study, also failed to detect significant influence of ABCB1 polymorphisms on the pharmacokinetics of lopi-navir or ritonavir in the blood plasma of HIV-infected subjects, treated with both drugs for at least 14 days [6]. In the latter study, digoxin was used as a phenotyping probe to demonstrate that P-glycoprotein activity was profoundly inhibited during the treatment with lopina-vir/ritonavir, consistent with in vitro evidence of P-glycoprotein inhibition by both lopinavir and ritonavir [2,24]. We suggest that inhibition of P-glycoprotein activity will occlude the con-sequences of ABCB1 polymorphisms on the pharmacokinetics of lopinavir and ritonavir, thus explaining the lack of impact of ABCB1 genotypes and haplotypes on the trough concen-tration of these drugs in blood plasma and their distribution to semen and saliva. A similar expla-nation has been recently proposed in relation to the effect of polymorphisms in CYP3A4/5 on the pharmacokinetics of protease inhibitors in a clinical setting [25,26].

Our study has distinct strengths: first, we had access to complete information on all drugs being dispensed rather than prescribed, since the participants were covered by the Brazilian program of universal access to HAART treat-ment. Second, this study reflects real-life com-munity prescribing and dispensing of antiretro-viral drugs in a natural clinical setting. We also recognize that this study has limitations: first, it was not designed to evaluate the influence of ABCB1 polymorphisms in treatment efficacy or on surrogates of efficacy, for example, viral load in the different matrices. Second, the lack of association between ABCB1 genotypes and lopinavir/ritonavir concentrations in blood, semen and saliva of this Brazilian cohort, may not extend to other populations because of differences in the distribution of ABCB1 poly-morphisms. For example, the 2677A allele is Ta

ble

4. C

on

cen

trat

ion

s o

f lo

pin

avir

an

d r

ito

nav

ir in

pla

sma,

sem

en a

nd

sal

iva

of

99 H

IV‑i

nfe

cted

su

bje

cts

acco

rdin

g t

o A

BC

B1

hap

loty

pes

, fro

m

SNPs

123

6C>T

, 267

7G>

no

nG

, an

d 3

435C

>T.

AB

CB

1 LP

V c

on

cen

trat

ion

RT

V c

on

cen

trat

ion

Hap

loty

pe

nPl

asm

an

Sem

enn

Saliv

an

Plas

ma

nSe

men

nSa

liva

CG

C

Non

carr

iers

2078

03 (

44

87–8

699

)20

30

8.0

(193

.6–5

84

.0)

206

4.3

(4

4.8

–10

0.6

)20

256

.6 (1

79.6

–54

0.6

)20

17.9

(10.

5–3

9.5

)20

4.7

(3.1

–8.3

)

Car

rier

s79

613

4 (3

761–

843

4)

7628

3.0

(120

.0–4

61.0

)74

75.5

(35.

7–13

0.0

)79

261.

8 (1

53–3

93)

7417

.6 (

9.2–

26.8

)74

5.3

(3.5

–9.4

)

0 co

py20

7803

(4

487

–869

9)

203

08

.0 (1

93.6

–58

4.0

)20

64

.3 (

44

.8–1

00.

6)

2025

6.6

(179

.6–5

40.

6)

2017

.9 (1

0.5

–39.

5)

204

.7 (3

.1–8

.3)

1 co

py52

574

6 (3

733

–90

48

)5

024

7.6

(112

.5–4

52.9

)5

076

.0 (3

6.1

–119

.5)

5222

2.8

(15

4–3

64

.5)

48

16.3

(9.

2–26

.7)

50

5.4

(3.9

–9.0

)

2 co

pies

2775

40

(38

49–9

04

8)

2635

1.6

(126

.3–5

62.0

)24

69.3

(33.

8–1

71.6

)27

288

.9 (1

50.

8–5

26.6

)26

18.4

(9.

5–2

8.4

)24

5.2

(3.2

–12.

1)

Tno

nG

T

Non

carr

iers

616

60

0 (3

796

–902

4)

5927

9.1

(113

.0–4

82.0

)56

75.5

(41

.1–1

29.9

)61

286

.1 (1

54

.0–4

15.6

)5

817

.3 (

8.9

–27.

2)56

5.3

(3.5

–9.7

)

Car

rier

s3

859

92 (3

818

–83

41)

3728

6.0

(179

.3–4

93.3

)3

865

.8 (3

6.8

–119

.0)

38

250.

0 (1

46

.3–4

25.0

)3

619

.5 (1

1.3

–28

.0)

38

5.1

(2.9

–8.4

)

CG

T

Non

carr

iers

7459

92 (3

758

–869

5)

7227

8.0

(13

4.3

–44

0.4

)69

62.6

(35.

7–11

9.0

)74

255.

2 (1

51.0

–40

9.0

)71

17.0

(9.

2–25

.9)

694

.9 (3

.1–8

.5)

Car

rier

s25

66

00

(459

1–8

66

4)

2435

3.1

(118

.2–5

99.

2)25

83.

2 (5

1.7–

129.

9)

2528

6.1

(179

.2–4

74.0

)23

22.1

(9.

2–37

.3)

255.

5 (4

.5–9

.9)

TGC

Non

carr

iers

876

414

(392

6–8

617

)8

428

2.6

(122

.4–4

75.5

)82

75.5

(36

.1–1

30.

3)

8727

3.1

(157

.8–4

38

.3)

8217

.6 (

9.9

–27.

6)

825.

4 (3

.3–9

.7)

Car

rier

s12

5219

(36

62–8

576

)12

307

.0 (1

68

.0–4

79.7

)12

60.

85 (

47.0

–88

.1)

1221

8.1

(119

.5–3

12.7

)12

16.7

(9.

1–25

.9)

124

.1 (3

.1–

6.1

)Th

e ha

plo

typ

e fr

equ

enci

es in

the

ove

rall

stu

dy

po

pul

atio

n w

ere:

CG

C (

0.5

4),

Tno

nGT

(0.1

9),

CG

T (0

.15

), T

GC

(0.

07),

Cno

nGC

, Cno

nGT,

Tno

nGC

and

TG

T (a

ll <

0.02

) LP

V: L

op

inav

ir; R

TV: R

ito

navi

r.

www.futuremedicine.com 317future science group

ABCB1 polymorphisms & antiretroviral therapy ReseaRch aRticle

Executive summary

ABCB1 polymorphisms & pharmacokinetics of HIV protease inhibitorsThe HIV protease inhibitors, which are frequently included in highly active antiretroviral therapy (HAART) regimens used worldwide, are �substrates of the P‑glycoprotein (ABCB1) transporter, encoded by the polymorphic ABCB1 gene.The impact of � ABCB1 polymorphisms on the pharmacokinetics of protease inhibitors is controversial. Regarding lopinavir, the main object of the present study, previous reports detected no effect of common exonic polymorphisms in ABCB1 on the drug concentration in blood plasma.The present study is the first to investigate the influence of � ABCB1 polymorphisms on the distribution of lopinavir and ritonavir into semen and saliva of HIV‑infected subjects under stable lopinavir‑containing HAART. Transmission of HIV occurs primarily via the sexual route, and a substantial proportion of individuals continue to shed the virus in the semen despite HAART treatment. Quantification of lopinavir and ritonavir in saliva is of potential interest for the purpose of therapeutic drug monitoring, because of the ease, noninvasiveness and reduced risk of infection associated with saliva collection.

ABCB1 polymorphisms are not associated with the concentrations of lopinavir & ritonavir in plasma, semen & saliva of HIV-infected men

Blood, semen and saliva samples were collected from 113 HIV‑infected men, the trough (predosing) concentrations of lopinavir and �ritonavir were measured by liquid chromatography/mass spectrometry/mass spectrometry, the ABCB1 1236C>T, 2677G>T/A and 3435C>T polymorphisms were genotyped using the single base extension/termination method, and ABCB1 haplotypes were statistically inferred.The lopinavir and ritonavir trough concentrations in blood plasma, semen and saliva displayed large interindividual variability. Drug �concentrations in semen and saliva were several fold lower than in blood plasma, consistent with poor penetration of lopinavir and ritonavir in the male genital tract and salivary glands. Statistically significant, but weak to moderate correlations (r2 = 0.09–0.29) were observed between the concentrations of each protease inhibitor in plasma versus semen or saliva, and between semen and saliva.The allele and genotype frequency distribution of the � ABCB1 polymorphisms did not differ from previous data for healthy volunteers from the same geographical area.Univariate and multivariate regression analysis revealed no influence of � ABCB1 genotypes or haplotypes on the concentrations of lopinavir and ritonavir in plasma, semen and saliva of the HIV‑infected subjects.

ConclusionGenotyping the � ABCB1 1236C>T, 2667G>T/A and 3435C>T polymorphisms does not seem useful to predict lopinavir and ritonavir trough concentrations in blood plasma, semen or saliva of HIV‑infected men under stable HAART treatment.The concentrations of lopinavir or ritonavir in saliva cannot be used to predict the concentration of these drugs in seminal fluid. �

more frequent in certain Asian populations, but could not be investigated in the present study because of its rarity (<1%). Third, the possibil-ity cannot be excluded that ABCB1 polymor-phisms previously reported but not examined in this study, or yet unidentified ABCB1 vari-ants might impact on the pharmacokinetics of lopinavir and ritonavir.

ConclusionIn conclusion, this is the first study to document the lack of effect of ABCB1 polymorphisms on the distribution of lopinavir and ritonavir into semen and saliva in HIV-infected subjects under stable HAART treatment. Our data extends to these subjects, previous observations [4–6] of no influence of ABCB1 genotypes on the through concentrations of lopinavir and ritonavir in blood plasma. Thus, genotyping the ABCB1 1236C>T, 2667G>T/A and 3435C>T poly morphisms does not seem useful to predict lopinavir and ritonavir trough concentrations in blood plasma, semen or saliva of HIV-infected men under stable HAART treatment. Finally, despite the practical advantages of saliva, compared with semen collection for pharmacokinetic studies, the poor to moderate correlation between the

concentrations of either lopinavir or ritonavir in the two matrices, indicates that saliva cannot be used to predict the concentration of these drugs in seminal fluid.

AcknowledgementsWe thank Abbott Laboratórios do Brasil Ltda. for providing lopinavir, ritonavir and internal standard samples for the analytical experiments.

Financial & competing interests disclosureThis research was supported by grants from Conselho Nacional de Desenvolvimento Científico e Tecnológico, Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro and Swiss Bridge Foundation. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Ethical conduct of research The authors state that they have obtained appropriate insti­tutional review board approval or have followed the princi­ples outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investi gations involving human subjects, informed consent has been obtained from the participants involved.

Pharmacogenomics (2009) 10 (2)318 future science group

ReseaRch aRticle Estrela, Ribeiro, Barroso et al.

BibliographyPapers of special note have been highlighted as:� of interest�� of considerable interest

Storch CH, Theile D, Lindenmaier H, 1

Haefeli WE, Weiss J: Comparison of inhibitory activity of anti-HIV drugs on P-glycoprotein. Biochem. Pharmacol. 73, 1573–1581 (2007).

Rodríguez-Nóvoa S, Barreiro P, Jiménez-2

Nácher I, Soriano V: Overview of the pharmacogenetics of HIV therapy. Pharmacogenomics J. 6, 234–245 (2006).

Leschziner GD, Andrew T, Pirmohamed M, 3

Johnson MR: ABCB1 genotype and PGP expression, function and therapeutic drug response: a critical review and recommendations for future research. Pharmacogenomics J. 7, 154–179 (2007).

Review of the impact of P-glycoprotein ��

(Pgp) expression and function on drug response.

Winzer R, Langmann P, Zilly M4 et al.: No influence of the P-glycoprotein genotype (MDR1 C3435T ) on plasma levels of lopinavir and efavirenz during antiretroviral treatment. Eur. J. Med. Res. 8, 531–534 (2003).

Ma Q, Brazeau D, Zingman BS5 et al.: Multidrug resistance 1 polymorphisms and trough concentrations of atazanavir and lopinavir in patients with HIV. Pharmacogenomics 8, 227–235 (2007).

Analysis of the effect of the �� ABCB1 2677G>T/A and 3435C>T SNPs on the plasma concentrations of lopinavir.

Wyen C, Fuhr U, Frank D6 et al.: Effect of an antiretroviral regimen containing ritonavir boosted lopinavir on intestinal and hepatic CYP3A, CYP2D6 and P-glycoprotein in HIV-infected patients. Clin. Pharmacol. Ther. 84, 75–82 (2008).

Study of the association between ��� ABCB1 polymorphisms and the trough concentrations of lopinavir/ritonavir in blood plasma.

Barroso PF, Schechter M, Gupta P 7 et al.: Effect of antiretroviral therapy on HIV shedding in semen. Ann. Intern. Med. 133, 280–284 (2000).

Analysis of the impact of highly active ���

antiretroviral therapy (HAART) on HIV shedding in semen.

Kashuba AD, Dyer JR, Kramer LM, 8

Raasch RH, Eron JJ, Cohen MS: Antiretroviral-drug concentrations in semen: implications for sexual transmission of human immunodeficiency virus Type 1. Antimicrob. Agents Chemother. 43, 1817–1826 (1998).

Taylor S, Pereira AS: Antiretroviral drug 9

concentrations in semen of HIV-1 infected men. Sex. Transm. Infect. 77, 4–11 (2001).

Vergara TR, Estrela R, Suarez-Kurtz G, 10

Schechter M, Cerbino-Neto J, Barroso PF: Limited penetration of lopinavir and ritonavir in the genital tract of men infected with HIV-1 in Brazil. Ther. Drug Monit. 28, 175–179 (2006).

Pilot study of the distribution of lopinavir ��

and ritonavir into semen of HIV-infected Brazilian subjects under stable HAART therapy.

Taylor S, Back DJ, Drake SM11 et al.: Antiretroviral drug concentrations in semen of HIV- infected men: differential penetration of indinavir, ritonavir, and saquinavir. J. Antimicrob. Chemother. 48, 351–354 (2001).

Sankatsing SU, Droste J, Burger D 12 et al.: Limited penetration of lopinavir into seminal plasma of HIV-1-infected men. AIDS 16, 1698–1700 (2002).

Solas C, Lafeuillade A, Halfon P, 13

Chadapaud S, Hittinger G, Lacarelle B: Discrepancies between protease inhibitor concentrations and viral load in reservoirs and sanctuary sites in human immunodeficiency. Antimicrob. Agents Chemother. 47, 238–243 (2003).

Lafeuillade A, Solas C, Chadapaud S, 14

Hittinger G, Poggi C, Lacarelle B: HIV RNA levels, resistance, and drug diffusion in semen versus blood in patients receiving a lopinavir-containing regimen. J. Acquir. Immune Defic. Syndr. 32, 462–464 (2003).

Ghosn J, Chaix ML, Peytavin G 15 et al.: Penetration of enfuvirtide, tenofovir, efavirenz, and protease inhibitors in the genital tract of HIV-1-infected men. AIDS 18, 1958–1961 (2004).

Estrela RCE, Ribeiro FS, Carvalho RS 16 et al.: The distribution of ABCB1 polymorphisms among Brazilians: impact of population admixture. Pharmacogenomics 9, 267–276 (2008).

First description of the impact of population ���

admixture on the distribution of ABCB1 polymorphisms among Brazilians.

Schaid DJ, Rowland CM, Tines DE, 17

Jacobson RM, Poland GA: Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am. J. Hum. Genet. 70, 425–434 (2002).

Estrela RC, Ribeiro FS, Seixas BV, Suarez-18

Kurtz G: Determination of lopinavir and ritonavir in blood plasma, seminal plasma, saliva and plasma ultra-filtrate by liquid chromatography/tandem mass spectrometry detection. Rapid Commun. Mass Spectrom. 22, 657–664 (2008).

Novel method for measuring lopinavir and ��

ritonavir conentrations in plasma, saliva and semen using liquid chromatography/mass spectrometry/mass spectrometry [LC/MS/MS]).

R Development Core Team: R: a language 19

and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria (2007).

van der Leur MR, Burger DM, la Porte CJ, 20

Koopmans PP: A retrospective TDM database analysis of interpatient variability in the pharmacokinetics of lopinavir in HIV-infected adults. Ther. Drug Monit. 28, 650–653 (2006).

Molla A, Vasavanonda S, Kumar G 21 et al.: Human serum attenuates the activity of protease inhibitors toward wild-type and mutant human immunodeficiency virus. Virology 250, 255–262 (1998).

Huisman MT, Smit JW, Schinkel AH: 22

Significance of P-glycoprotein for the pharmacology and clinical use of HIV protease inhibitors. AIDS 14, 237–242 (2000).

Sankatsing SU, Beijnen JH, Schinkel AH, 23

Lange JM, Prins JM: P-glycoprotein in human immunodeficiency virus Type 1 infection and therapy. Antimicrob. Agents Chemother. 48, 1073–1081 (2004).

Perloff MD, von Moltke LL, Greenblatt DJ: 24

Fexofenadine transport in Caco-2 cells: inhibition with verapamil and ritonavir. J. Clin. Pharmacol. 42, 1269–1274 (2002).

Owen A, Khoo S: Pharmacogenetics of 25

antiretroviral agents. Curr. Opin. HIV AIDS 3, 288–295 (2008).

Estrela RC, Santoro AB, Barroso PF, 26

Tuyama M, Suarez-Kurtz G: CYP3A5 genotype has no impact on plasma trough concentrations of lopinavir and ritonavir in HIV-infected subjects. Clin. Pharmacol. Ther. 84, 205–207 (2008).

Websites��The official site for the 2000 Brazilian 101

Census www.ibge.gov.br/home/estatistica/populacao/censo2000/

Access to the haplo.stats program used for 102

inference of ABCB1 haplotypes mayoresearch.mayo.edu/mayo/research/schaid_lab/software.cfm

A strongly functional language and 103

environment to statistically explore data sets www.r-project.org

Calculation of statistical power online 104

www.stat.uiowa.edu/~rlenth/Power/