Introduction
Of 2.3 million children living with HIV in 2006, almost 90% in sub-Saharan Africa, an estimated 780 000 urgently require antiretroviral therapy (ART) [1]. When liquids are available, they are costly, have short shelf lives, and are difficult to transport and store [2]. As a result of the lack of availability of affordable and appropriate paediatric ART formulations, divided adult fixed-dose combination (FDC) tablets are frequently provided for children. Adult FDCs, however, do not allow easy dose adjustment as a child grows. Furthermore, and more importantly, drug ratios in these adult FDCs are not correct for children. Our previous study in 127 HIV-infected Malawian and Zambian children receiving divided adult FDCs (Triomune; Cipla Pharmaceuticals, India; 200 mg nevirapine, 30 or 40 mg stavudine, 150 mg lamivudine) demonstrated a risk of nevirapine underdosing [3], as a result of children, particularly the youngest, metabolizing nevirapine more rapidly than adults. Underdosing is a major threat to the long-term success of ART, as a lack of potency in suppressing viral replication will result in the development of mutations, particularly against drugs with a low genetic barrier to resistance such as nevirapine and lamivudine, and limit subsequent treatment options [4,5]. This is of particular concern for children facing a life-time requirement for ART.
To address this problem, Cipla Pharmaceuticals have developed small, dispersible, crushable, scored FDCs for HIV-infected children in two sizes (Triomune Baby and Junior) with relatively higher nevirapine versus stavudine and lamivudine dose ratios compared with the adult FDC, in accordance with paediatric dosing recommendations [6]. In our independent pilot bioequivalence study, the pharmacokinetics of Triomune Baby and Junior were similar to branded products in six healthy males [7]. Formal bioequivalence was demonstrated by Cipla Pharmaceuticals (data on file).
Here we studied the pharmacokinetics of nevirapine, stavudine and lamivudine in Zambian HIV-infected children prescribed Triomune Baby or Junior twice a day according to surface area-derived weight bands, to determine whether the dose ratio results in optimal ART exposure in the target population.
Methods
Study population and design
CHAPAS1 is an open, randomized, controlled, phase I/II trial designed to assess the appropriate dosing of, and adherence to, Triomune Baby (50 mg nevirapine, 6 mg stavudine, 30 mg lamivudine) and Junior (double Baby dose). Two hundred HIV-infected children aged 3 months to 14 years weighing less than 30 kg who fulfill WHO criteria for initiating ART have been enrolled at the University Teaching Hospital, Lusaka, Zambia. Children were randomly assigned in a 1: 1 ratio to take Triomune Baby or Junior either immediately at a full dose in a twice-daily schedule, or in a once-daily dose escalation schedule with an additional stavudine/lamivudine tablet (Lamivir-S; Cipla Pharmaceuticals) for the first 14 days, followed by the full dose. Children previously treated with ART, including for the prevention of mother-to-child HIV transmission, were excluded. Further exclusion criteria were severe laboratory abnormalities, active opportunistic infection, and treatment with any medication known to be contraindicated in combination with nevirapine, stavudine and lamivudine.
The first 64 children enrolled in CHAPAS1 were to participate in the pharmacokinetic substudy, with 16 per age group: under 3, 3-6, 7-10 and 11-14 years. These children had to fulfill all enrolment criteria for CHAPAS1 and also not suffer from illnesses that could influence the pharmacokinetics of the antiretroviral drugs such as diarrhoea, vomiting, renal or liver disease, and not be taking concomitant medication with interactions with the ART.
Full nevirapine dosing was chosen to aim for daily doses of 300-400 mg/m2 [6] using estimated body surface area (BSA) for weight converted into weight bands (3-< 6, 6-< 10, 10-< 15, 15-< 20, 20-< 25, 25-< 30 kg) [8]. Daily stavudine and lamivudine doses were targeted to 2 and 8 mg/kg, respectively. Nevirapine dose ranges in mg/m2 and stavudine and lamivudine dose ranges in mg/kg (Table 1) are the result of the same dose in milligrams being given over weight (and BSA) ranges. Of note is the fact that one weight band (15-< 20 kg) has unequal morning and evening doses.
The protocol was reviewed and approved by the Ethics Committees of the University of Zambia, Lusaka and University College London. Written informed consent was obtained from parents or guardians, and children when appropriate.
Blood collection and drug concentration assays
At least 4 weeks after starting Triomune Baby or Junior, a 12-h pharmacokinetic curve was undertaken, following directly observed medication ingestion. Non-breastfed children fasted for ≥3 h before intake, and standardized meals were given 1-2, 4-6 and 8-12 h after intake. Two millilitres of blood was collected just before and 1, 2, 4, 6, 8 and 12 h after intake. Plasma was separated and stored at -80°C until transportation to the Netherlands on dry ice.
Plasma concentrations of nevirapine, stavudine and lamivudine were assayed at the Department of Clinical Pharmacy of the Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands, using two validated high-performance liquid chromatography assays with ultraviolet detection [9,10]. The plasma drug concentration 12 h postdose (C12h) and maximum concentration (Cmax) were determined directly from concentration-time data. The area under the concentration-time curve 0-12 h post dose (AUC12h) was evaluated using the linear-log trapezoidal rule.
Children were excluded if there was evidence of non-adherence, that is if the ratio between nevirapine C12h and the predose concentration was more than 2 and non-adherence was independently suspected by the study team, or if this ratio was more than 4 without suspected non-adherence.
Statistical methods
Weight-for-age, height-for-age and body mass index-for-age z-scores were calculated using the 1990 British Growth Charts [11] in STATA statistical software, version 9 (STATA Corp., College Station, Texas, USA) and CD4 cells-for-age using published age-related reference ranges [12]. Nevirapine, stavudine and lamivudine AUC12h were compared across weight bands and age groups using analysis of variance adjusted for randomization.
Results
Of the 71 children enrolled in the pharmacokinetic substudy, six were excluded because of evidence of non-adherence. These six, aged 10 months to 12 years, had similar weights and CD4 cell percentages to the rest of the cohort, and daily prescribed nevirapine doses of 363-424 mg/m2.
Of the remaining 65 children, 16, 18, 16 and 15 were aged under 3, 3-6, 7-10 and 11-14 years, respectively. The majority were malnourished and moderately to severely immunodeficient (Table 2). The mean (range) daily prescribed nevirapine dose at enrolment was 370 (317, 486) mg/m2. Of note, two children who were dosed at enrolment according to screening weight (10 and 11 kg) had experienced weight loss (both to 9.8 kg) by the time of enrolment, and therefore received doses equating to 427 and 416 mg/m2, respectively, at enrolment. Thirty-one (48%) and 34 (52%) were randomly assigned to the full dose and dose escalation arms of CHAPAS1, respectively.
The median (range) time from ART initiation to pharmacokinetic day was 27 (26, 56) days. Seven children gained and one lost weight after enrolment such that they were in the next higher (three 6-< 10 kg, one 10-< 15 kg, two 15-< 20 kg and one 20-< 25 kg at enrolment) or next lower (one 15-< 20 kg) weight band on the pharmacokinetic day. Doses were adjusted after the end of the pharmacokinetic day.
Nevirapine concentrations were higher than those previously reported in adults (Table 3), although C12h was subtherapeutic (< 3.0 mg/l) [4,13] in four children (6%): one in each of the weight bands 3-< 6 kg (one of two children), 10-< 15 kg (one of nine), 20-< 25 kg (one of 12) and 25-< 30 kg (one of 10); aged 9 months, 6, 10 and 10 years; and receiving daily prescribed nevirapine doses of 326-409 mg/m2. Furthermore, variability in nevirapine C12h (interquartile range 4.1, 7.0 mg/l) was greater than previously reported in adults (3.2, 5.1 mg/l) [14] (EMEA, Viramune; Summary of product characteristics). The pharmacokinetic parameters of stavudine and lamivudine were comparable to those previously reported in adults (Table 3).
Mean plasma concentrations are shown in Fig. 1 by weight band. As there were only two children in the lowest weight band 3-< 6 kg, and one (5.2 kg) had low plasma concentrations of nevirapine but high concentrations of stavudine and lamivudine, compared with the other (3.4 kg), they were excluded from the statistical analyses. There was no evidence of a difference in nevirapine AUC12h across the five remaining weight bands (P = 0.2) or the four age groups (P = 0.1). There was a difference in stavudine and lamivudine AUC12h by weight band (P = 0.0003 and 0.01, respectively), which appeared to be driven by lower concentrations in the single weight band with a lower dose in the morning than the evening (15-< 20 kg, 19 children; Table 3). There was no significant difference in stavudine or lamivudine AUC12h by age group (P = 0.6 and 0.4, respectively).
All but three of the children were on co-trimoxazole prophylaxis. Eleven children (17%) took one, one took two and one took three additional medications on the pharmacokinetic day (five amoxicillin, seven ferrous sulphate, three paracetamol, one pyrimethamine/sulfadoxine).
Equation (Uncited)Image Tools
Considering adverse events probably related to nevirapine by 8 weeks or less after initiating ART, three of the 65 children in the substudy had grade 2 nevirapine rashes within one month, which all resolved after 9-11 days after the temporary discontinuation of nevirapine (subsequently continued). One further child had a grade 1 nevirapine rash at 2 weeks, which resolved after 7 days, followed by transient grade 3 raised liver enzymes on the pharmacokinetic day (single values, then returned to normal); no changes were made to nevirapine. Four other children had transient single measurement grade 3 raised liver enzymes on the pharmacokinetic day, which returned to normal without nevirapine interruption. Nevirapine pharmacokinetic parameters in the eight children with nevirapine-related toxicities were comparable to those in the 57 without [mean (range) Cmax 10.6 (4.1, 18.9) versus 9.9 (3.8, 22.5) mg/l, P = 0.7; AUC12h 102 (40.4, 200) versus 93.3 (32.1, 232) mg/l per hour, P = 0.6]. In addition, transient asymptomatic laboratory abnormalities not judged to be related to any antiretroviral drug by 8 or less weeks were detected in a small number of children [grade 3 anaemia (n = 3), bilirubin (n = 1); grade 4 bilirubin (n = 1), creatinine (n = 1), thrombocytopenia (n = 1); all single measurements]. One further child had grade 4 bilirubin at week four which was not known to have resolved by death from pneumonia (HIV-related) at 9 weeks.
Equation (Uncited)Image Tools
Discussion
Equation (Uncited)Image Tools
In Zambian HIV-infected children receiving the recently developed paediatric FDCs Triomune Baby or Junior, nevirapine plasma concentrations were higher and more variable than historical data in adults, whereas overall pharmacokinetic parameters of stavudine and lamivudine were comparable.
Triomune Baby and Junior are small, dispersible, crushable, scored tablets with ART ratios designed to fit recommended paediatric doses of individual drugs according to WHO guidelines [6]. Importantly, the nevirapine dose in these FDCs has been increased relative to the dose in adult FDCs (Triomune) to compensate for the greater metabolism of nevirapine in children. We developed weight band-based dosing tables for Triomune Baby and Junior to simplify dosing in resource-limited settings. We aimed to achieve daily nevirapine doses of 300 mg/m2 or greater, converted into weight bands, to minimize the risk of subtherapeutic plasma concentrations [3,13] and virological failure [5].
As no paediatric reference drug concentrations are available for these ART doses and ratios, we compared our results with adult reference values. In these 65 children, nevirapine plasma concentrations were higher than in adults. Nevirapine C12h was subtherapeutic (< 3.0 mg/l) [4,13], however, in 6% of children and the interpatient variability was greater than in adults [14]. Therefore, we would not recommend a lower nevirapine dose in children, who are at greater risk of underdosing for several reasons [15]. Furthermore, in spite of higher average exposure, nevirapine-related adverse reactions were transient, with only temporary treatment interruptions, and nevirapine Cmax and AUC12h were comparable between children with and without nevirapine-related toxicities. This reflects the lack of a well-defined upper limit of the therapeutic range. The higher average nevirapine exposure minimized the percentage of children with subtherapeutic concentrations and therefore the related risk of the development of resistance, without any evidence of increased toxicity. There was no evidence of a difference in nevirapine AUC12h by weight band of 6 kg or greater or age group.
We determined stavudine and lamivudine plasma concentrations as surrogates of intracellular concentrations of the pharmacologically active triphosphate metabolites, because obtaining adequate blood volumes for the determination of intracellular concentrations of nucleoside reverse transcriptase inhibitor triphosphates is ethically, logistically and technically difficult in children [16]. Our assumption was that comparable plasma concentrations between children and adults would reflect comparable intracellular concentrations. Unlike nevirapine, the plasma concentration parameters of stavudine and lamivudine were comparable with those reported in adults, with no differences across prespecified age groups. A previous study reported that the recommended daily lamivudine dosage (8 mg/kg) [6] leads to lower AUC12h and Cmax in younger compared with older children [17]. Of note, however, is the fact that in our study younger children received relatively higher lamivudine doses per kilogram compared with older children, as a consequence of optimizing the dosing of three antiretroviral drugs across six weight bands, resulting in adequate plasma exposures across all ages.
There was a difference in stavudine and lamivudine AUC12h by weight band, which appeared to be driven by the single weight band with unequal dosing. The reduced plasma exposure to stavudine and lamivudine after the lower morning dose would, however, not be representative of the average 24-h exposure; in addition, the long intracellular half-life of lamivudine may not be reflected in plasma concentrations. Nevirapine exposure appears to be less affected by unequal dosing, most likely because of its long elimination half-life. Although unequal dosing should be avoided when possible for drugs with a short elimination half-life, this may be preferable to increasing complexity and thus risking lower adherence by administering larger numbers of lower dose tablets equally per dosing interval. We plan to investigate adherence, in particular between equal and unequal dosed weight bands, using detailed adherence data collected in CHAPAS1.
WHO recently recommended Triomune 30 for all adults irrespective of body size because of increasing concerns about stavudine-associated lipodystrophy and data suggesting that the lower dose may be adequate in adults. It is particularly important not to overdose stavudine in young children who will have long-term exposure to ART by virtue of therapy initiation early in life. When using stavudine as part of an FDC, however, it is also crucial not to underdose nevirapine and lamivudine, which have low genetic barriers to resistance. Our previous study in 127 HIV-infected Malawian and Zambian children receiving divided adult Triomune 30 or 40 (30 or 40 mg stavudine, respectively) demonstrated that it is impossible to achieve adequate nevirapine concentrations in young children without risking overdosing with stavudine, especially when using Triomune 40, but even when using Triomune 30 [3]. In that study, 18% of children had a subtherapeutic random nevirapine plasma concentration (< 3.0 mg/l), compared with only 6% with a subtherapeutic trough nevirapine plasma concentration in the current study. Furthermore, the difference in the proportion of subtherapeutic nevirapine concentrations may have been larger if we had evaluated nevirapine C12h in our previous study, although smaller if we had directly observed medication intake in that study. In a Thai study [18], only 3% of children (one of 34) who received divided adult FDCs (GPO-VIR S30; GPO, Thailand; equivalent to Triomune 30) had subtherapeutic nevirapine concentrations (defined as < 3.4 mg/l) and the average nevirapine dose was considerably higher at 328 mg/m2 [18] compared with 265 mg/m2 in our previous study [3]. The average age was, however, higher and none were less than 3 years (compared with 13% in our previous study), nor did any receive quarter or three-quarter tablets (compared with 42%) [3].
To avoid overdosing with stavudine while aiming to achieve therapeutic nevirapine concentrations in children taking adult FDCs, liquid formulations allow for more accurate dosing of the smallest children when there are no alternatives [19]. Liquid formulations are costly, difficult to transport and store, and are complicated for carers to administer (for example, three syrups each of different volume rather than low numbers of whole or divided FDCs). Furthermore, frequent changes in the volumes of separate liquid antiretroviral drugs as a child grows present considerable challenges for both carers and medical providers. Paediatric FDCs in an appropriate ratio for children, such as Triomune Baby and Junior, are attractive alternatives, especially in resource-limited settings. It is hoped that in the future alternative FDCs, such as Atripla, Truvada and Kivexa, which do not contain stavudine, will become available for children. One disadvantage of FDCs is less flexibility of dosing, exemplified by the common practice of dose escalation of the auto-inducing drug nevirapine during the first 2 weeks of treatment to avoid high plasma concentrations and possible toxicity. In the case of Triomune Baby and Junior, this means the provision of an additional stavudine/lamivudine tablet during the first 2 weeks to allow full dosing of these antiretroviral drugs. CHAPAS1 is currently investigating whether increased complexity of dose escalation outweighs the possible reduction in adverse events. Of note is the fact that nevirapine dose escalation in 52% of children in this pharmacokinetic substudy would not have influenced our results as sampling was carried out at steady state [20] and this was confirmed by sensitivity analyses unadjusted for randomization (results not shown). Our pharmacokinetic data suggest no effect of drug concentrations on the rate of adverse events. Co-medications used in our study would not be expected to influence outcomes.
We assessed the pharmacokinetics of the antiretroviral drugs based on observed intake. Six of 71 children (8%) did not meet our strict adherence criteria; this raises a potential cause for concern regarding adherence, which we will investigate further in our planned adherence analyses. Exclusion of these children allowed reliable estimation of the pharmacokinetic parameters.
As there were only two children in the lowest weight band (3-< 6 kg), there are insufficient data to draw conclusions about dosing for children weighing less than 6 kg at this stage. In the light of critical interim results from the CHER trial [21] suggesting the importance of early treatment of infants, this extension to children under 6 kg is crucial. We are currently enrolling into a further substudy to estimate four-sample pharmacokinetic curves of the three antiretroviral drugs in children aged one month and older weighing 3-< 6 kg at enrollment who are receiving Triomune Baby.
The children in our pharmacokinetic substudy will continue to be followed in CHAPAS1. Further analyses are planned to evaluate the possible predictors of ART plasma concentrations, in particular the impact of malnutrition. We previously demonstrated lower nevirapine concentrations in stunted African children and higher concentrations in wasted children for the same dose per BSA [3]. In addition, analyses are planned to investigate the impact of predictors on subsequent changes in CD4 cell percentage, viral load and resistance, in particular with respect to the lower plasma exposure of stavudine and lamivudine observed in the weight band with unequal dosing.
We conclude that the antiretroviral ratio of nevirapine, stavudine and lamivudine in the paediatric FDCs Triomune Baby and Junior is appropriate for children weighing 6 kg and over; additional pharmacokinetic sampling is needed in children under 6 kg. Following successful review of this study in African HIV-infected children, who are a key target population, WHO now recommend the Triomune Baby and Junior ratios and these FDCs have been approved by the United States Food and Drug Administration. It is hoped they will now be made accessible to national ART programmes at low cost, and that future efforts will be made to develop other solid paediatric FDC that do not contain stavudine.
Acknowledgements
The authors would like to thank the families and children, and staff from the University Teaching Hospital and School of Medicine, Lusaka, Zambia. Technicians from the Department of Clinical Pharmacy, Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands are kindly acknowledged for the analysis of the plasma antiretroviral concentrations. Marthe Le Prevost is acknowledged for help in training on pharmacokinetic techniques for the study. The study medication was supplied by Cipla Pharmaceuticals, India.
Sponsorship: The CHAPAS1 study was funded by the European and Developing Countries Clinical Trials Partnership (grant CHINTU 2004.01.H.d2.33011) and sponsored by the Medical Research Council, UK.
Conflicts of interest: None.
References
1. WHO/UNAIDS/UNICEF.
Scaling up priority HIV/AIDS interventions in the health sector: towards universal access. Progress report, April 2007. Available at:
www.who.int. Accessed: November 2007.
2. Havens PL, Gibb DM. Increasing antiretroviral drug access for children with HIV infection. Pediatrics 2007; 119:838-845.
3. Ellis JC, L'homme RFA, Ewings FM, Mulenga V, Bell F, Chileshe R, et al. Nevirapine concentrations in HIV-infected children treated with divided fixed-dose combination antiretroviral tablets in Malawi and Zambia. Antivir Ther 2007; 12:253-260.
4. de Vries-Sluijs TE, Dieleman JP, Arts D, Huitema AD, Beijnen JH, Schutten M, van der Ende ME. Low nevirapine plasma concentrations predict virological failure in an unselected HIV-1-infected population. Clin Pharmacokinet 2003; 42:599-605.
5. Verweel G, Sharland M, Lyall H, Novelli V, Gibb DM, Dumont G, et al. Nevirapine use in HIV-1-infected children. AIDS 2003; 17:1639-1647.
6. World Health Organization.
Preferred antiretroviral medicines for treating and preventing HIV infection in younger children, August 2007. Available at:
www.who.int. Accessed: November 2007
7. L'homme RFA, Dijkema T, Warris A, van der Ven AJAM, Gibb DM, Burger DM. Pharmacokinetics of two generic fixed-dose combinations for HIV-infected children (Pedimune Baby & Pedimune Junior) are similar to the branded products in healthy adults. J Antimicrob Chemother 2007; 59:92-96.
8. Sharkey I, Boddy AV, Wallace H, Mycroft J, Hollis R, Picton S. Body surface area estimation in children using weight alone: application in paediatric oncology. Br J Cancer 2001; 85:23-28.
9. Hollanders RMW, van Ewijk-Beneken Kolmer EWJ, Burger DM, Wuis EW, Koopmans PP, Hekster YA. Determination of nevirapine, an HIV-1 nonnucleoside reverse transcriptase inhibitor, in human plasma by reversed-phase high-performance liquid chromatography. J Chromatogr B Biomed Sci Appl 2000; 744:65-71.
10. Verweij-van Wissen CP, Aarnoutse RE, Burger DM. Simultaneous determination of the HIV nucleoside analogue reverse transcriptase inhibitors lamivudine, didanosine, stavudine, zidovudine and abacavir in human plasma by reversed phase high performance liquid chromatography. J Chromatogr B Analyt Technol Biomed Life Sci 2005; 816:121-129.
11. Cole TJ, Freeman JV, Preece MA. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood. Stat Med 1998; 17:407-429.
12. Wade AM, Ades AE. Age-related reference ranges: significance tests for models and confidence intervals for centiles. Stat Med 1994; 13:2359-2367.
13. La Porte CJL, Back DJ, Blaschke T, Boucher CAB, Fletcher CV, Flexner C, et al. Updated guideline to perform therapeutic drug monitoring for antiretroviral agents. Rev Antivir Ther 2006; 3:4-14.
14. Van Heeswijk RP, Veldkamp AI, Mulder JW, Meenhorst PL, Wit FW, Lange JMA, et al. The steady-state pharmacokinetics of nevirapine during once daily and twice daily dosing in HIV-1-infected individuals. AIDS 2000; 14:F77-F82.
15. Menson EN, Walker AS, Sharland M, Wells C, Tudor-Williams G, Riordan FAI, et al. Underdosing of antiretrovirals in UK and Irish children with HIV as an example of problems in prescribing medicines to children, 1997-2005: cohort study. BMJ 2006; 332:1183-1187.
16. Bergshoeff A, Burger D, Verweij C, Farrelly L, Flynn J, Le Prevost M, et al. Plasma pharmacokinetics of once- versus twice-daily lamivudine and abacavir: simplification of combination treatment in HIV-1-infected children (PENTA-13). Antivir Ther 2005; 10:239-246.
17. Burger DM, Verweel G, Rakhmanina N, Verwey-Van Wissen CP, La Porte CJ, Bergshoeff AS, et al. Age-dependent pharmacokinetics of lamivudine in HIV-infected children. Clin Pharmacol Ther 2007; 81:517-520.
18. Chokephaibulkit K, Plipat N, Cressey TR, Frederix K, Phongsamart W, Capparelli E, et al. Pharmacokinetics of nevirapine in HIV-infected children receiving an adult fixed-dose combination of stavudine, lamivudine and nevirapine. AIDS 2005; 19:1495-1499.
19. Corbett A, Hosseinipour M, Nyirenda J. Pharmacokinetics between trade and generic liquid and split tablet formulations of lamivudine, stavudine and nevirapine in HIV infected Malawian children. In: 45th International Conference on Antimicrobial Agents and Chemotherapy. Washington, DC, USA; 16-19 December 2005. Abstract 1106.
20. Kappelhoff BS, Van Leth F, MacGregor TR, Lange J, Beijnen JH, Huitema AD, 2NN Study Group. Nevirapine and efavirenz pharmacokinetics and covariate analysis in the 2NN study. Antivir Ther 2005; 10:145-155.
21. Violari A, Cotton M, Gibb DM. Antiretroviral therapy initiated before 12 weeks of age reduces early mortality in young HIV-infected infants: evidence from the Children with HIV Early Antiretroviral Therapy (CHER) Study. In: 4th IAS Conference on HIV Pathogenesis, Treatment, and Prevention. Sydney, Australia; 22-25 July 2007. Abstract WESS 103.
© 2008 Lippincott Williams & Wilkins, Inc.