| |
Genetic factors influencing atazanavir/r plasma concentrations and the risk of severe hyperbilirubinemia
|
| |
| |
AIDS: Volume 21(1) 2 January 2007 p 41-46
[BASIC SCIENCE]
Rodriguez-Novoa, Soniaa; Martin-Carbonero, Luzb; Barreiro, Pablob; Gonzalez-Pardo, Gemaa; Jimenez-Nacher, Inmaculadaa; Gonzalez-Lahoz, Juanb; Soriano, Vincentb
From the aPharmacokinetic & Pharmacogenetic Unit, Spain
bDepartment of Infectious Diseases, Hospital Carlos III, Madrid, Spain.
Received 10 June, 2006
"....In a previous study conducted in patients exposed to ATV without ritonavir boosting, lower ATV plasma concentrations were seen in patients with the MDR1 3435C>T change [12]. A similar effect was observed in the present study, which tested patients on ritonavir-boosted ATV, despite P-gp inhibition by ritonavir which could have blunted the influence of distinct MDR1 genotypes..... The results of the linear regression analyses showed that hyperbilirubinemia, although directly influenced by ATV plasma levels, was mainly dependent of the UGT1A1 enzyme activity, particularly in severe cases....
....Univariate and multivariate linear regression analyses including gender, age, concomitant use of tenofovir, and polymorphisms at MDR1 and UGT1A1 genes were carried out. Only the presence of at least one T allele at MDR1-3435 independently predicted lower ATV plasma levels.... Only older age and higher ATV plasma levels independently predicted hyperbilirubinemia... having at least one TA7 allele at UGT1A1 significantly predicted grade 3-4 hyperbilirubinemia..."
Abstract
Background: Hyperbilirubinemia is frequently seen in patients treated with atazanavir (ATV). Polymorphisms at the uridin-glucoronosyl-transferase 1A1 (UGT1A1) and multidrug resistance 1 (MDR1) genes may influence, respectively, bilirubin and ATV plasma concentrations.
Patients and methods: HIV-infected individuals receiving ATV 300 mg daily plus ritonavir 100 mg daily at one clinic were examined. ATV plasma concentrations were measured at steady state. MDR1-3435C>T and UGT1A1 polymorphisms were examined in DNA extracted from blood mononuclear cells.
Results: A total of 118 patients (all Caucasian) were analysed. The median ATV plasma concentration was 465 ng/ml [interquartile range (IQR), 233-958]. MDR1-3435 genotypes were as follows: CC (32%), CT (47%) and TT (21%). CC patients showed higher ATV minimum concentration than those with CT/TT genotypes: 939 ng/ml (IQR, 492-1266) versus 376 ng/ml (IQR, 221-722) (P = 0.001). In multivariate analyses, having at least one T allele at MDR1-3435 was independently associated with lower ATV plasma concentrations (β: -427 [95% confidence interval (CI), -633 to -223]; P < 0.001). The proportion of patients with grade 3-4 hyperbilirubinemia varied with distinct UGT1A1 genotypes: 80% for 7/7, 29% for 6/7 and 18% for 6/6 (P = 0.012). In the multivariate analysis, having at least one 7 allele at UGT1A1 was independently associated with severe hyperbilirubinemia (odds ratio, 2.96; 95% CI, 1.29-6.78; P = 0.01).
Conclusions: Polymorphisms at MDR1-3435 significantly influence ATV plasma concentrations, as does being Caucasian patients with CT/TT genotypes, having lower ATV levels, even using ritonavir boosting. On the other hand, although ATV plasma concentrations directly correlate with bilirubin levels, the risk of severe hyperbilirubinemia is further increased in the presence of the UGT1A1-TA7 allele.
Introduction
Atazanavir (ATV) has a pharmacokinetic profile that makes it suitable for once daily oral administration [1]. The efficacy of ATV is strongly influenced by plasma exposure, as this parameter is usually related to with intracellular drug concentrations. Several host genetic factors can alter the disposition of antiretrovirals, by influencing their absorption, distribution or metabolism. The P-glycoprotein (P-gp), an efflux pump coded by the MDR1 gene, is one of the most important cell transporters. P-gp works by expelling foreign compounds, such as HIV protease inhibitors (PI), outside the cells [2-5]. The influence of genetic polymorphisms at the MDR1 gene on PI plasma levels is a matter of debate [6-10]. A recent report has highlighted that ATV is a substrate of P-gp [11] and we have shown that a 3435C>T polymorphism at the MDR1 gene may influence ATV plasma concentrations in patients receiving 400 mg of ATV once daily [12]. However, to our knowledge there are no data about the influence of polymorphisms at the MDR1 gene on ATV pharmacokinetics when the drug is boosted with ritonavir. As ritonavir is a potent P-gp inhibitor [13-15], any influence of the P-gp phenotype on ATV plasma concentrations could be blunted in patients receiving ritonavir-boosted ATV.
Hyperbilirubinemia is the most common adverse event in patients treated with ATV, and it directly correlates with ATV plasma concentrations [12]. The underlying mechanism of this association is an inhibition by ATV of the uridine-glucuronosyl transferase (UGT) 1A1 enzyme, which is responsible for bilirubin conjugation [16]. Several polymorphisms at the UGT1A1 gene have been described. The wild type allele (UGT1A1*1) contains six TA repeats in the promoter (TA6) whereas the most common variant allele (UGTA1A1*28), which is associated with the mild form of the Gilbert's syndrome, an inherited unconjugated hyperbilirubinemia disorder, contains seven repeats (TA7) [17-20]. It is unclear to what extent polymorphisms at the UGT1A1 gene may influence the risk of hyperbilirubinemia and its relationship with different ATV plasma concentrations.
In the present study, the frequency and predictive value for hyperbilirubinemia of these two genetic polymorphisms, MDR1 3435C>T and UGTA1A1*28, were analysed in HIV-infected patients treated with ritonavir-boosted ATV.
Patients and methods
All HIV-infected patients who initiated therapy with ATV 300 mg daily plus ritonavir 100 mg daily (ATV/r) together with two nucleoside analogues at our institution during year 2005 were identified. All individuals were antiretroviral-experienced. On the basis of pharmacy records and clinical interviews, only good treatment-adherent patients were chosen for further analysis. Patients who received drugs that could potentially interfere with ATV pharmacokinetics, such as non-nucleoside analogs or proton-pump inhibitors, were excluded. The unique exception was tenofovir (TDF), which may cause a 40-60% reduction in ATV minimum concentration (Cmin), particularly when the drug is not boosted with ritonavir. In fact, the effect of TDF on ATV Cmin seems to be minimal when ATV is co-administered with ritonavir [21,22].
Patients had been advised to take ATV during dinner, which was around 12 h before blood was drawn. Total bilirubin and mid-dose ATV plasma concentrations were determined at week 12 of therapy. Plasma concentrations of ATV were measured by a validated high performance liquid chromatography method [23] and are reported as Cmin values, following standard calculations based on linear regression by square minimums [24]. The half-life of 8.6 h previously reported for ATV 300 mg daily plus ritonavir 100 mg daily was considered for all patients [21,25]. Genetic analyses at MDR1 position 3435 were performed on genomic DNA extracted from peripheral blood mononuclear cells (PBMC). The amplification of exon 26 of the MDR1 gene was made using previously described primers and PCR conditions [26]. Likewise, the recognition of TA repetitions at the UGT1A1 promoter gene was assessed by direct sequencing of DNA extracted from PBMC, using previously reported primers and PCR conditions [27].
Statistical analyses
Descriptive results of continuous variables were expressed as median and interquartile range (IQR) values. Continuous variables were compared with parametric (Student's t) or non-parametric (Mann-Whitney U) tests, as required. Proportions were compared using the chi-square test, with Yates or Fisher corrections when needed. Univariate and multivariate logistic regression analyses were performed for the identification of factors associated with ATV plasma levels and with hyperbilirubinemia. Those parameters with P values < 0.2 in the univariate analysis were entered into a step by step multivariate test. Spearman's rho was calculated for binary correlation analyses. All statistics were calculated using the SPSS package version 11.0 (SPSS, Chicago, Illinois, USA), and differences were considered to be significant when P values were < 0.05.
Results
Study population
A total of 118 HIV-infected patients were identified. All were Caucasians; 77% male. The median age was 42 years (IQR, 37-46 years). The median CD4 cell count was 497 cells/μl (IQR, 308-725 cells/μl) and the median plasma HIV-RNA 1.9 log copies/ml (IQR, 1.7-3.4 log copies/ml), showing < 50 copies/ml in 12% of cases. Hepatitis C virus antibodies were present in 37%. The median total bilirubin was 0.77 mg/dl (IQR, 0.6-0.9 mg/dl). Overall, 82% of patients were taking tenofovir concomitantly.
Predictors of atazanavir plasma levels
The median ATV plasma concentration was 465 ng/ml (IQR, 233-958 ng/ml) at week 12 of therapy. The distribution of different MDR1 genotypes at position 3435 was as follows: C/C in 32%, C/T in 47% and T/T in 21%. As shown in Fig. 1, patients with the CC genotype had higher ATV plasma concentrations than those with at least one T allele (C/T and T/T genotypes): 939 (IQR, 492-1266) versus 376 (IQR, 221-722) ng/ml (P < 0.01). The number of patients with ATV plasma levels below the proposed effective concentration threshold (150 ng/ml) [28] was one of 27 (4%) and seven of 56 (11%) for P-gp genotypes CC and CT/TT, respectively. This difference was not statistically significant.
The distribution of UGT1A1 genotypes was as follows: TA6/TA6 in 45%, TA6/TA7 in 48% and TA7/TA7 in 7%. The median ATV plasma level in each genotype group was as follows: 455 ng/ml (IQR, 267-944 ng/ml) in TA6/TA6, 382 ng/ml (IQR, 245-656 ng/ml) in TA6/TA7, and 467 ng/ml (IQR, 306-1185 ng/ml) in TA7/TA7 (P = 0.98).
Univariate and multivariate linear regression analyses including gender, age, concomitant use of tenofovir, and polymorphisms at MDR1 and UGT1A1 genes were carried out. Only the presence of at least one T allele at MDR1-3435 independently predicted lower ATV plasma levels (Table 1).
Predictors of hyperbilirubinemia
The proportion of patients with hyperbilirubinemia of any degree (total bilirubin > 1.3 mg/dl) at week 12 of ATV/r therapy was 88%. A direct correlation between bilirubin levels and ATV plasma concentrations at week 12 was noticed (rho = 0.27; P = 0.005).
The distribution of patients with hyperbilirubinemia according to MDR1 genotypes was as follows: 97% in C/C, 89% in C/T and 76% in T/T (P = 0.06). With respect to UGT1A1 genotypes, all patients with TA7/TA7 and 87% of those with TA6/TA6 or TA6/TA7, developed hyperbilirubinemia (P = NS).
Univariate and multivariate linear regression analyses including gender, concomitant use of tenofovir, and polymorphisms at MDR1 and UGT1A1 genes were carried out. Only older age and higher ATV plasma levels independently predicted hyperbilirubinemia (Table 1).
The incidence of grade 3-4 hyperbilirubinemia (total bilirubin > 3.2 mg/ml) was 30% at week 12 of ATV/r therapy. According to UGT1A1 variants, the proportion of patients with severe hyperbilirubinemia was as follows: 80% in TA7/TA7, 29% in TA6/TA7 and 18% in TA6/TA6 (P = 0.01) (Fig. 1). The proportion of patients who developed grade 3-4 hyperbilirubinemia according to 3435-Pgp genotypes was as follows: 28% for CC, 35% for CT and 19% for TT (P = 0.42).
The univariate and multivariate linear regression analyses including gender, age, concomitant use of tenofovir, and polymorphisms at MDR1 and UGT1A1 showed that having at least one TA7 allele at UGT1A1 significantly predicted grade 3-4 hyperbilirubinemia (Table 1).
Discussion
Cell membrane transporters, such as P-gp, expel foreign molecules outside the cells, including ATV [11]. There is little information about the functional effects of MDR1 3435C>T polymorphisms on ATV pharmacokinetics. In a previous study conducted in patients exposed to ATV without ritonavir boosting, lower ATV plasma concentrations were seen in patients with the MDR1 3435C>T change [12]. A similar effect was observed in the present study, which tested patients on ritonavir-boosted ATV, despite P-gp inhibition by ritonavir which could have blunted the influence of distinct MDR1 genotypes. In the present study, C/C carriers had higher ATV plasma concentrations than patients with C/T or T/T genotypes, suggesting that P-gp plays an important role on ATV disposition.
It is intriguing that the C to T nucleotide change at position 3435 in the MDR1 gene is silent, and does not change the encoded amino acid [29]. Thus, it may be a marker of another polymorphism in the P-gp gene with functional consequences. A 2677G>T polymorphism has recently been suggested as the possible candidate [30]. The association of the MDR1 3435C>T polymorphism with particular genotypes of metabolic enzymes or gene regulators may influence drug metabolism in the liver. Persons homozygous for the MDR1 2677T allele, which is frequently linked to the 3435T allele, show enhanced constitutive CYP3A4 expression in the liver and gut in comparison with subjects homozygous for the 2677G allele [31]. As ATV is mainly metabolized by the CYP3A4 [1], the recognition of lower ATV plasma levels in 3435T allele carriers in the present study is consistent with this hypothesis. These patients might have enhanced ATV clearance by CYP3A4. In this way, the main effect of MDR1 polymorphisms on ATV plasma levels might be due to differences in CYP3A4 isoenzyme activity, being negligible the direct role of P-gp. Figure 1 summarizes the complex interactions between atazanavir, bilirubin and polymorphisms at MDR1 and UGT1A1 genes.
Differences in ATV plasma concentrations in patients with distinct P-gp genotypes might be particularly relevant in antiretroviral-experienced patients, in whom greater ATV plasma concentrations may be needed to inhibit virus replication. If so, patients with the MDR1 genotype CC could be at a pharmacokinetic advantage with respect to subjects with other genotypes.
Hyperbilirubinemia is the most common laboratory abnormality in patients treated with ATV. It is reversible and usually benign, although it may cause scleral icterus or overt jaundice in 8% of patients, which may limit ATV acceptability [32]. This study confirms a direct correlation between ATV plasma levels and the risk for bilirubin elevations [33]. The influence of polymorphisms at the UGT1A1 gene on bilirubin levels in patients treated with ATV is not well known. We found that 80% of patients with the UGT1A1 allele TA7 in homozygosis developed severe hyperbilirubinemia. This high positive predictive value along with a 7% penetrance of this allele in the study population, in agreement with prior reports [20,34], may further support the clinical usefulness of UGT1A1 testing before prescribing ATV, at least in Caucasians.
The results of the linear regression analyses showed that hyperbilirubinemia, although directly influenced by ATV plasma levels, was mainly dependent of the UGT1A1 enzyme activity, particularly in severe cases. The fact that UGT1A1 is responsible for bilirubin conjugation in the liver, and the target for ATV inhibition, explains these findings. These results need to be validated in larger population studies, in which the accuracy of ATV pharmacogenomics should be improved by considering other genotypes and other ethnicities. In this regard, the role of two common polymorphisms at the CYP3A4*1B promoter and in exon 7 of CYP3A4*2 deserves particular attention.
References
1. Goldsmith D, Perry C. Atazanavir. Drugs 2003; 63:1679-1693.
[Fulltext Link] [Medline Link] [CrossRef] [Context Link]
2. Ueda K, Clark D, Chen C, Roninson I, Gottesman M, Pastan I. The human multidrug resistance (mdr1) gene: cDNA cloning and transcription initiation. J Biol Chem 1987; 262:505-508.
[Medline Link] [Context Link]
3. Chen C, Clark D, Ueda K, Pastan I, Gottesman M, Roninson I. Genomic organization of the human multidrug resistance (MDR1) gene and origin of P-glycoproteins. J Biol Chem 1990; 265:506-514.
[Context Link]
4. Gottesman M, Pastan I, Ambudkar S. P-glycoprotein and multidrug resistance. Curr Opin Genet Dev 1996; 6:610-617.
[Medline Link] [CrossRef] [Context Link]
5. Ambudkar S, Dey S, Hrycyna C, Ramachandra M, Pastan I, Gottesman M. Biochemical, cellular, and pharmacological aspects of the multidrug transporter. Annu Rev Pharmacol Toxicol 1999; 39:361-398.
[Medline Link] [CrossRef] [Context Link]
6. Fellay J, Marzolini C, Meaden E, Back D, Buclin T, Chave J, et al. Response to antiretroviral treatment in HIV-1-infected individuals with allelic variants of the multidrug resistance transporter 1: a pharmacogenetic study. Lancet 2002; 359:30-36.
[Medline Link] [CrossRef] [Context Link]
7. Anderson P, Schuetz E, Fletcher C. CYP3A5 and MDR1 (P-gp) polymorphisms in HIV-infected adults: associations with indinavir concentrations and antiviral effects. XI Conference on Retroviruses and Opportunistic Infections, San Francisco, 2004 [abstract 619].
[Context Link]
8. Nasi M, Borghi V, Pinti M, Bellodi C, Lugli E, Maffei S, et al. MDR1 C3435T genetic polymorphism does not influence the response to antiretroviral therapy in drug-naive HIV-positive patients. AIDS 2003; 17:1696-1698.
[Fulltext Link] [Medline Link] [CrossRef] [Context Link]
9. Saitoh A, Singh K, Powell C, Fenton T, Fletcher C, Brundage R, et al. An MDR1-3435 variant is associated with higher plasma nelfinavir levels and more rapid virologic response in HIV-1 infected children. AIDS 2005; 19:371-380.
[Fulltext Link] [Medline Link] [CrossRef] [Context Link]
10. Winzer R, Langmann P, Zilly M, Tollmann F, Schubert J, Klinker H, 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 2003; 8:531-534.
[Medline Link] [Context Link]
11. Lucia M, Golotta C, Rutella S, Rastrelli E, Savarino A, Cauda R. Atazanavir inhibits P-glycoprotein and multidrug resistance-associated protein efflux activity. J Acquir Immune Defic Syndr 2005; 39:635-637.
[Fulltext Link] [Medline Link] [Context Link]
12. Rodriguez-Novoa S, Barreiro P, Rendon A, Barrios A, Corral A, Jimenez-Nacher I, et al. Plasma levels of atazanavir and the risk of hyperbilirubinemia are predicted by the 3435 C->T polymorphism at the multidrug resistance gene 1. Clin Infect Dis 2006; 42:291-295.
[Medline Link] [CrossRef] [Context Link]
13. Drewe J, Gutmann H, Fricker G, Torok M, Beglinger C, Huwyler J. HIV protease inhibitor ritonavir: a more potent inhibitor of P-glycoprotein than the cyclosporine analog SDZ PSC 833. Biochem Pharmacol 1999; 57:1147-1152.
[Medline Link] [CrossRef] [Context Link]
14. Chaillou S, Durant J, Garraffo R, Georgenthum E, Roptin C, Dunais B, et al. Intracellular concentration of protease inhibitors in HIV-1-infected patients: correlation with MDR-1 gene expression and low dose of ritonavir. HIV Clin Trials 2002; 3:493-501.
[Medline Link] [CrossRef] [Context Link]
15. Penzak S, Shen J, Alfaro R, Remaley A, Natarajan V, Falloon J. Ritonavir decreases the non-renal clearance of digoxin in healthy volunteers with known MDR1 genotypes. Ther Drug Monit 2004; 26:322-330.
[Fulltext Link] [Medline Link] [CrossRef] [Context Link]
16. Burchell B, Brierley C, Rance D. Specificity of human UDP-glucuronosyl-transferases and xenobiotic glucoronization. Life Sci 1995; 57:1819-1831.
[Medline Link] [CrossRef] [Context Link]
17. Bosma P, Chowdhury J, Bakker C, Gantla S, de Boer A, Oostra B, et al. The genetic basis of the reduced expression of bilirubin UDP-glucuronosyl-transferase 1 in Gilbert's syndrome. N Engl J Med 1995; 333:1171-1175.
[Medline Link] [CrossRef] [Context Link]
18. Monaghan G, Ryan M, Seddon R, Hume R, Burchell B. Genetic variation in bilirubin UPD-glucuronosyl-transferase gene promoter and Gilbert's syndrome. Lancet 1996; 347:578-581.
[Medline Link] [CrossRef] [Context Link]
19. Mackenzie P, Owens I, Burchell B, Bock K, Bairoch A, Belanger A, et al. The UDP glycosyl-transferase gene superfamily: recommended nomenclature update based on evolutionary divergence. Pharmacogenetics 1997; 7:255-269.
[Fulltext Link] [Medline Link] [CrossRef] [Context Link]
20. Beutler E, Gelbart T, Demina A. Racial variability in the UDP-glucuronosyl-transferase 1 (UGT1A1) promoter: a balanced polymorphism for regulation of bilirubin metabolism? Proc Natl Acad Sci U S A 1998; 95:8170-8174.
[Medline Link] [CrossRef] [Context Link]
21. Bristol-Myers Squib Company. Reyataz (atazanavir sulfate) product information guide. Bristol-Myers Squib Company; 2003.
[Context Link]
22. Taburet A, Piketty C, Chazallon C, Vincent I, Gerard L, Calvez V, et al. Interactions between Atazanavir-ritonavir and tenofovir in heavily pretreated HIV-infected patients. Antimicrob Agents Chemother 2004; 48:2091-2096.
[Medline Link] [CrossRef] [Context Link]
23. Tribut O, Arvieux C, Michelet C, Chapplain J, Allain H, Bentue-Ferrer D. Simultaneous quantitative assay of six HIV protease inhibitors, one metabolite, and two non-nucleoside reverse transcriptase inhibitors in human plasma by isocratic reversed-phase liquid chromatography. Ther Drug Monit 2002; 24:554-562.
[Fulltext Link] [Medline Link] [CrossRef] [Context Link]
24. Gibaldi M, Perrier D. Pharmacokinetics. 2nd edn. New York: Marcel Deker; 1982.
[Context Link]
25. Le Tiec C, Barrail A, Goujard C, Taburet A. Clinical pharmacokinetics and summary of efficacy and tolerability of atazanavir. Clin Pharmacokinet 2005; 44:1035-1050.
[Fulltext Link] [Medline Link] [CrossRef] [Context Link]
26. Nauck M, Stein U, von Karger S, Marz W, Wieland H. Rapid detection of the C3435T polymorphism of multidrug resistance gene 1 using fluorogenic hybridization probes. Clin Chem 2000; 46:1995-1997.
[Medline Link] [Context Link]
27. Zucker S, Qin X, Rouster S, Yu F, Green R, Keshavan P, et al. Mechanism of indinavir-induced hyperbilirubinemia. Proc Natl Acad Sci U S A 2001; 98:12671-12676.
[Medline Link] [CrossRef] [Context Link]
28. Gonzalez de Requena D, Canta F, Marrone R, D'Avolio A, Sciandra M, Milia M, et al. Atazanavir Ctrough is associated with efficacy and safety: definition of therapeutic range. XII Conference on Retroviruses and Opportunistic Infections, Boston, MA, 2005 [abstract 645].
[Context Link]
29. Hoffmeyer S, Burk O, von Richter O, Arnold H, Brockmoller J, Johne A, et al. Functional polymorphisms of the human multidrug-resistance gene: multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. Proc Natl Acad Sci U S A 2000; 97:3473-3478.
[Medline Link] [CrossRef] [Context Link]
30. Kroetz D, Pauli-Magnus C, Hodges L, Huang C, Kawamoto M, Johns S, et al. Sequence diversity and haplotype structure in the human ABCB1 (MDR1, multidrug resistance transporter) gene. Pharmacogenetics 2003; 13:481-494.
[Fulltext Link] [Medline Link] [CrossRef] [Context Link]
31. Lamba J, Strom S, Venkataramanan R, Thummel K, Lin Y, Liu W, et al. MDR1 genotype is associated with hepatic cytochrome P450 3A4 basal and induction phenotype. Clin Pharmacol Ther 2006; 79:325-338.
[Medline Link] [CrossRef] [Context Link]
32. Sulkowski M. Drug-induced liver injury associated with antiretroviral therapy that includes HIV-1 protease inhibitors. Clin Infect Dis 2004; 38(suppl 2):90-97.
[Context Link]
33. Barrios A, Rendon A, Gallego O, Martin-Carbonero L, Valer L, Rios P, et al. Predictors of virological response to atazanavir in protease inhibitor-experienced patients. HIV Clin Trials 2004; 5:201-205.
[Medline Link] [CrossRef] [Context Link]
34. Rotger M, Taffe P, Bleiber G, Gunthard H, Furrer H, Vernazza P, et al. Gilbert syndrome and the development of antiretroviral therapy-associated hyperbilirubinemia. J Infect Dis 2005; 192:1381-1386.
[Medline Link] [CrossRef] [Context Link]
|
|
| |
| |
|
|
|
