

* Massachusetts General Hospital, Boston, Massachusetts
Childrens Developmental Center, Maitland, Florida
Clinical Study Centers, Little Rock, Arkansas
|| North Carolina Neuropsychiatry, Chapel Hill, North Carolina on behalf of the SLI381.301 Study Group
| ABSTRACT |
|---|
|
|
|---|
Methods. A multicenter, randomized, double-blind, parallel-group, placebo-controlled trial was conducted at 47 sites. After a 1-week washout of any previous stimulant medication, patients were randomized to receive single-daily morning doses of placebo or SLI381 10 mg, 20 mg, or 30 mg for 3 weeks. Participants aged 6 to 12 years inclusive who satisfied Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria diagnosis of ADHD were included. The primary efficacy parameter was the Conners Global Index Scale for Teachers. Secondary efficacy parameters included the Conners Global Index Scale for Parents, the Clinical Global Impressions Scale for improvement, and the Parent Global Assessment for improvement. Safety was assessed by recording adverse events, laboratory tests, and vital signs at each visit during the study. Physical examinations and electrocardiograms were performed at the screening and the end of the study.
Results. Five hundred eighty-four children were randomized, 563 were included in the intent-to-treat population, and 509 completed the entire study. Intention-to-treat analysis of Conners Global Index Scale for Teachers and Conners Global Index Scale for Parents scores revealed significant improvement in morning, afternoon, and late afternoon behavior for all active treatment groups versus placebo. All active treatment groups showed significant dose-related improvement in behavior from baseline. Both the Clinical Global Impressions Scale for improvement and Parent Global Assessment for improvement showed all doses of SLI381 to be superior to placebo at treatment end and both confirmed the dose-response relationship between improvement and the SLI381 dose. The incidence of spontaneously reported adverse events was low and similar for active treatments and placebo.
Conclusions. SLI381 produced consistent, dose-related improvements on all measures of efficacy. The extended-release nature of the SLI381 formulation was shown by continued, significant improvement in afternoon assessments by teachers and afternoon and late afternoon assessments by parents. The time course and therapeutic effects of SLI381 suggests that this medication is an efficacious once-daily treatment for children with ADHD.
Key Words: attention-deficit/hyperactivity disorder amphetamine Adderall XR randomized controlled trial
Abbreviations: ADHD, attention-deficit/hyperactivity disorder ECG, electrocardiogram DISC, Diagnostic Interview Schedule for Children CGIS-T, Conners Global Index Scale for Teachers CGIS-P, Conners Global Index Scale for Parents CGI, Clinical Global Impressions Scale for improvement PGA, Parent Global Assessment for Improvement ITT, intention-to-treat ANCOVA, analysis of covariance
| INTRODUCTION |
|---|
|
|
|---|
Numerous randomized, controlled trials have shown stimulant medications to be highly effective in ameliorating the symptoms of ADHD.810 However, their use can be problematic because of the need for multiple daily dosing in most individuals. In-school dosing in pediatric patients may lead to ridicule by peers and further negative impact on self-esteem, compliance, and treatment satisfaction. In addition, many patients may require medication coverage that extends into homework time and organized recreational activities. This denotes a compelling need for more effective once-daily dosage forms of stimulant medications that last throughout the school day and into the evening.
Adderall (Shire US Inc, Valley Stream, NY), a single-entity amphetamine drug product consisting of a mixture of neutral salts of dextroamphetamine sulfate, amphetamine sulfate, the dextro isomer of amphetamine saccharate, and d,l-amphetamine aspartate, is effective in treating the symptoms of ADHD.1116 To address the need for a once-daily dosing option for more patients, a 2-component extended-release formulation (SLI381 capsules [Catalytica Pharmaceuticals, Inc, Greenville, NC]) of Adderall has been developed that is designed to produce pulsed-release of amphetamine salts yielding a therapeutic effect that lasts throughout the day and into the evening with 1 morning dose.
The SLI381 capsule formulation is composed of 2 types of Microtrol beads combined in a 50:50 ratio within 1 capsule. The immediate release beads are designed to release drug content in a time course similar to Adderall. The delayed release beads are designed to release drug content 4 to 6 hours after oral administration of the capsule. With the inclusion of the delayed-release component, the 2-unit formulation, given once a day, is expected to produce similar pharmacokinetic and pharmacodynamic effects as immediate release Adderall given twice daily.
A pilot study compared the bioavailability of 3 composite 20-mg experimental extended release formulations with a single-dose administration against the reference, Adderall 10 mg bid with a 4-hour interval.17 The selected SLI381 formulation was bioequivalent to Adderall in terms of the d- and l-amphetamine extent and rate of absorption, and the time to maximum concentration values for d- and l-amphetamine were not different from those observed for Adderall bid with a 4-hour interval.
This multicenter study was designed to assess the safety and efficacy of SLI381 in a naturalistic school and home setting. This naturalistic environment allows efficacy and safety results to be extrapolated to the clinical pediatric population most likely to receive stimulant treatment for ADHD. A randomized, double-blind, parallel group design with up to 3 weeks of exposure permits between-group differences in clinical manifestations of childhood ADHD and in acute safety and tolerability to be evaluated.
The primary objective of this study was to assess the efficacy and safety of SLI381 compared with placebo in the treatment of ADHD in children. The secondary objective was to assess diurnal variation in responses to SLI381 based on morning and afternoon assessments. This study was conducted in compliance with institutional review board and informed consent regulations.
| METHODS |
|---|
|
|
|---|
Children were either known to be responsive to stimulants or naïve to stimulant treatment. Participants incapable of understanding or following the instructions given in the study, known nonresponders to stimulant medication, and those with a comorbid psychiatric diagnosis (psychosis, bipolar illness, pervasive developmental disorder, severe obsessive compulsive disorder, severe depressive or severe anxiety disorder) were excluded from the study. Participants with a history of seizure (exclusive of febrile seizure), tic disorder, or a family history of Tourettes disorder, those with a documented allergy or intolerance to Adderall, and participants taking clonidine, anticonvulsant drugs, pemoline (within 30 days), medications that affect blood pressure or heart rate, steroids, or other medications that have central nervous system effects or affect performance (such as sedating antihistamines and decongestant sympathomimetics, either oral or topical) also were excluded. Other exclusion criteria included a concurrent chronic or acute illness or condition that might confound results or increase risk to the participant, history of suspected substance abuse disorder, or living with someone with a current diagnosed substance abuse disorder.
Study Design
This was a multicenter, randomized, double-blind, placebo-controlled, parallel group, study of 3 doses of SLI381. After screening (Visit 1), participants returned for a washout visit (Visit 2) in which vital signs were measured and any current stimulant treatment (if any) was discontinued. All participants receive placebo treatment in a single-blind fashion for a 1-week washout phase. At Visit 3 (baseline visit), participants were randomly allocated in a 2:2:2:3 ratio to 1 of 4 treatment groups: SLI381 10 mg/d, SLI381 20 mg/d, SLI381 30 mg/d, or placebo.
The study design used a dose-escalation regimen for the SLI381 treatment groups, in which all children received SLI381 10 mg for the first week of treatment, those randomized to the 20-mg and 30-mg groups received 20 mg for the second week of treatment, and those randomized to the 30-mg group received 30 mg for the third week of treatment (Fig 1). Patients returned at weekly intervals for safety and efficacy assessments. During each of the study weeks, patients were also assessed by their teachers at school and by their parents at home for efficacy using standard ADHD rating scales (see "Measurements").
|
Measurements
During the screening visit (Visit 1), ADHD was established by 1) a psychiatric examination which reviewed Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition and criteria, and 2) the National Institutes of Mental Health Diagnostic Interview Schedule for Children Version 4.0 (DISC-4.0).18 One interviewer from each center was trained to administer the DISC-4.0 before the initiation of the study.
Beginning with the washout week, teachers completed the Conners Global Index Scale (teachers version; CGIS-T) on 3 specified days (Monday, Wednesday, and Friday), in the morning and afternoon of each day (
10:00 AM and 2:00 PM). Parents completed the Conners Global Index Scale (parents version; CGIS-P) once per week on either Saturday or Sunday. They rated the child 3 times during the day (at
10:00 AM, 1:00 PM, and after 4:00 PM). Parents were also given a diary to record observations of their childs behavior during the week to aid in completing the Parents Global Assessment, which was completed at baseline (washout week) and the final visit. Adverse event information was recorded, and drug compliance was assessed by capsule counts at each weekly study visit.
The primary efficacy parameter was the average of morning and afternoon CGIS-T total scores at study endpoint (or last treatment week). This included a total of up to 6 assessments, 3 for morning and 3 for afternoon. In addition, the averages of the CGIS-T scores were calculated separately for both morning and afternoon assessments to evaluate the drug effect versus placebo for morning and afternoon separately.
Secondary efficacy parameters included CGIS-P and 2 global assessments of overall improvement by the investigator (the Clinical Global Impressions Scale for improvement [CGI]), and by parents (the Parent Global Assessment for improvement [PGA]). Average total scores were calculated separately for morning, afternoon, and late afternoon CGIS-P; only the improvement score of the CGI and PGA were examined in the study.
Adverse events were recorded by COSTART term and rated as mild, moderate, or severe. Compliance was measured by capsule counts at the end of each visit.
Statistical Analysis
To detect a standardized treatment difference of 0.47 between a single SLI381 treatment group and the placebo group at 90% power (2-tailed) and an
level of 0.05, it was necessary to enroll at least 120 participants in the placebo group and 80 participants in each of the 3 treatment groups (10, 20, and 30 mg/d). Assuming a 20% attrition rate over the course of the study, a total of 450 participants were targeted for inclusion in the study.
The primary efficacy analysis was conducted using the intention-to-treat (ITT) population and a 2-way analysis of covariance (ANCOVA) model with a general linear approach. This ANCOVA model used the average of CGIS-T total scores over the last treatment week assessed (ie, study endpoint), as the dependent variable, and the treatment (four levels) and study site as the independent variable. The corresponding baseline score (obtained during the single-blind washout week) of the primary endpoint was included as a covariate in the statistical model. SAS PROC GLM (Windows version 6.12) was used for this analysis. If statistically significant treatment effect was disclosed (P <.05) by the 2-way ANCOVA, Dunnetts test for multiple mean comparisons was used to compare the differences between treatment groups and the placebo group.
A similar analytic approach to the CGIS-T total scores was adopted for each individual treatment week and for secondary efficacy measures and subgroup population comparisons. For the CGIS-T, those assessments conducted before noon, exclusive, were operationally categorized as morning assessments; those assessments that were conducted after noon, inclusive, were operationally categorized as afternoon assessments. Similarly, for the CGIS-P, morning assessments were those conducted before noon, afternoon assessments were those conducted from noon to 4:00 PM, and late afternoon assessments were operationally categorized as those conducted on or after 4:00 PM. Because participants assigned to SLI381 20 mg and 30 mg received their doses in a dose escalation fashion, a 3-way analysis of variance with split-plot design was used to analyze weekly averages of the CGIS-T total scores without data imputation for participants who completed the study to evaluate the dose titration effect. Safety-related information was analyzed comparatively for active versus placebo groups, using 1-way analysis of variance, to examine differences in changes from baseline to end of study.
| RESULTS |
|---|
|
|
|---|
|
|
Efficacy
Primary Efficacy: Teacher Ratings
The primary efficacy measure was the average of the CGIS-T total scores obtained at study endpoint, including both the morning and afternoon assessments, and the baseline score obtained during the placebo washout week was the covariate in the analytical model (Fig 2). Two-way analysis of variance showed a trend toward a difference across study sites in the baseline values (P = .053), but no difference in baseline scores across treatment groups (P = .402). All active treatment groups showed significant improvements in CGIS-T scores from baseline to study endpoint. The magnitude of changes from baseline to study endpoint was
6 to 1 in unit point for SLI381 groups compared with placebo, a roughly fivefold relative improvement of ADHD behavior at school in participants receiving active treatment. The 2-way ANCOVA at endpoint showed a highly significant treatment effect (P < .0001), a significant effect of baseline covariate (P < .0001), and site effect was not significant (P = .2778). Dunnetts test, which adjusted the P value for multiplicity, disclosed further that at the study endpoint, differences in CGIS-T total score were negative for each SLI381 dose group in comparison to placebo and highly significant (P < .001).
|
|
|
|
Parent Ratings
The results of the parent evaluation, using the parent version of the CGIS, revealed similar efficacy (Fig 5; Table 4). The mean magnitude of changes from baseline to study endpoint was
2.5 to 1 in unit point for SLI381 groups compared with placebo, a 1.5-fold relative improvement of ADHD behavior at home for participants on active treatment. The changes from baseline in the morning and late afternoon assessments were almost identical, indicating that the SLI381 was still effective in the early evening (defined as on and from 4:00 PM).
|
|
|
|
The occurrence of abnormal laboratory values was low and similar in both placebo- and SLI381-treatment groups. No apparent trends were seen and most were considered to be clinically insignificant. No clinically significant changes in group mean ECG parameters were observed over the duration of the study; in all cases where abnormalities occurred, findings were deemed harmless, clinically insignificant, or normal by pediatric cardiologists. In regard to vital signs, no significant changes were seen between treatment groups and no significant changes occurred over time.
| DISCUSSION |
|---|
|
|
|---|
The findings that one early morning dose of SLI 381 was effective throughout the morning and afternoon as reported by parents and teachers confirm, in the naturalistic setting of home and school, previous results in a laboratory setting.19,20 The previous study examined pharmacokinetic and pharmacodynamic parameters of three fixed doses of SLI381 compared with placebo and Adderall 10 mg in an analog classroom setting and found SLI381 20 mg and 30 mg to produce rapid improvements in behavior and cognition by 1.5 hours postdose. These improvements were sustained for up to 12 hours, strongly supporting a duration of action that covers the school day.
The long acting clinical effects of SLI381 clearly eliminate the need for in school administration. Considering the inherent problems associated with in-school dosing of stimulant medications for ADHD (ie, many schools do not have nursing staff available to administer these medicines; many parents and children do not want to receive their treatment in school because of the potential stigmatizing effects of such procedures; and the medicolegal concerns associated with the storage and administration of scheduled medicines in school) the availability of a safe and effective long acting stimulant such as SLI381 is likely to greatly facilitate the treatment of the large number of afflicted youth with ADHD.
Such benefits of SLI381 not only will increase patient privacy and compliance, but are also important to secure a steady clinical effect throughout the school day. Despite the well-documented efficacy of stimulant drugs in the treatment of ADHD, the short duration of action of the immediate release preparations has been associated with therapeutic peaks and valleys with their potential for breakthrough symptoms in between doses and likely adverse impact on treatment satisfaction.
Of equal clinical relevance is the finding documenting that the clinical benefits of SLI381 extend beyond the school day to include time-points after 4:00 PM. Such results indicate that the duration of action of SLI381 should allow for pharmacological support of homework activities, after school athletic and social activities, as well as family lifecritical components of the childs life.
A clear dose-response relationship was evident across all efficacy assessments, with increased improvement at successive increases in dose. In fact, the main efficacy endpoint, the CGIS-T total score average, reached normalized values for the SLI381 30-mg group. In the absence of dose-limiting adverse effects when treating patients with ADHD, much thought should be given to increasing the dose to an optimal level. Although lower doses may not be subtherapeutic, this study demonstrates that optimal results were seen at the 30-mg dose of SLI381.
Study medication was very well tolerated and the incidence of adverse events compares favorably with that of immediate-release compounds. There was little evidence of a dose-response relationship in safety parameters across the SLI381 treatment groups. Thus, increasing the dose had no apparent influence on tolerability. The exception to this may be in the incidence of reported anorexia. However, posthoc analysis of reported anorexia showed that treatment naïve participants had significantly more problems with this (33.6% reporting) than did those participants with prior experience with ADHD medications (16.2% reporting). Because this was a shot-term trial of 3 weeks duration, it can not be determined from these data whether the reported incidence of this adverse event would abate over time. A 24-month, open-label trial is ongoing and may better define the long-term adverse event profile of this medication.
Noteworthy is the clean cardiovascular profile of SLI381 as evidenced by repeated ECG measurements. This is important to emphasize in light of recent concerns regarding potential cardiovascular effects of stimulant medications.
Subgroup analyses showed the same, positive improvements for boys and girls with ADHD. In fact, considering the sizeable representation of girls, this study represents one of the largest clinical trials of girls with ADHD ever conducted. Although almost all of the available literature on ADHD is limited to boys with this disorder, recent work documents that ADHD is equally disabling in boys and girls with ADHD.21 Thus, the findings of similar benefits associated with SLI381 treatment in boys and girls with ADHD are very encouraging and extend to females the well-documented therapeutic benefits of psychostimulants in the treatment of ADHD.
Although the majority of patients in this study had previous exposure to stimulant medication, a substantial minority (
34%) were treatment naïve. The subgroup analyses showed the same, positive improvements for patients with a history of stimulant treatment as for those with no previous exposure to stimulant medication. This demonstrates the efficacy of SLI381 across both patient groups.
Because our sample included a narrow developmental window of children 6 to 12 years, it is unclear whether our results will generalize to older age groups. Considering the wide range of benefits associated with the long duration of action of SLI381, it is likely that other age groups will also benefit from this pharmacologic profile. Clearly, additional clinical trials are needed to document the efficacy and safety of SLI381 in adolescents and adults with ADHD.
Only a very small number of participants had ADHD inattentive type, making it difficult to draw any conclusions regarding this subset of patients. Although stimulant medications are commonly used and effective in these patients, additional studies are needed to accurately assess the effects of SLI381 in this patient group.
| CONCLUSION |
|---|
|
|
|---|
| APPENDIX |
|---|
|
|
|---|
Paul J. Ambrosini, MD, Medical College of Pennsylvania, Philadelphia, Pennsylvania
L. Eugene Arnold, MD, Ohio State University, Columbus, Ohio
Joseph Biederman, MD, Massachusetts General Hospital, Boston, Massachusetts
Jeffery L. Blumer, MD, University Hospitals of Cleveland, Cleveland, Ohio
Samuel W. Boellner, MD, Clinical Study Centers, L.L.C., Little Rock, Arkansas
Larry Brown, MD, Childrens Hospital of Philadelphia, Philadelphia, Pennsylvania
Ronald T. Brown, PhD, Medical University of South Carolina, Charleston, South Carolina
Charles D. Casat, MD, Behavioral Health Center, Charlotte, North Carolina
Mark C. Chandler, MD, North Carolina Neuropsychiatry, PA, Chapel Hill, North Carolina
Edward, Cherlin, MD, Valley Clinical Research, El Centro, California
C. Keith Conners, PhD, Duke University Medical Center, Durham, North Carolina
Daniel Connor, MD, University of Massachusetts Medical Center, Worcester, Massachusetts
Daniel Coury, MD, Pediatric Clinical Studies Center, Columbus, Ohio
Andrew J. Cutler, MD, Coordinated Research of Florida, Winter Park, Florida
W. Burleson Daviss, MD, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire
Carlos Figueroa, MD, Advanced Psychiatric Group, Rosemead, California
L. Matthew Frank, MD, Dominion Psychiatric Associates, Virginia Beach, Virginia
Laurence L. Greenhill, MD, New York State Psychiatric Institute, New York, New York
Saul Helfing, MD, Oregon Center for Clinical Investigations, Inc, Lake Oswego, Oregon
Sharon L. Hirsch, MD, Childrens Memorial Hospital, Chicago, Illinois
Joseph P. Horrigan, MD, University of North Carolina, Chapel Hill, North Carolina
James Hudziak, MD, University of Vermont/ Fletcher Allen Healthcare, Burlington, Vermont
Martin Kremenitzer, MD, Associated Neurologists, PC, Danbury, Connecticut
Kara Lewis, MD, Barrow Neurologic Group, Phoenix, Arizona
Robert S. Lipetz, DO, Encompass Clinical Research, Spring Valley, California
Thomas M. Lock, MD, State University of New York, Buffalo, New York
Peter D. Londborg, MD, Seattle Clinical Research Center, Seattle, Washington
Frank A. Lopez, MD, Childrens Development Center, Maitland, Florida
Keith McBurnett, PhD, University of Chicago, Chicago, Illinois
James T. McCracken, MD, UCLA Neuropsychiatric Institute, Los Angeles, California
Denis Mee-Lee, MD, Hawaii Clinical Research Center, Honolulu, Hawaii
Jeffery Newcorn, MD, Mt. Sinai Medical Center, New York, New York
Donna Palumbo, PhD, University of Rochester School of Medicine, Rochester, New York
Anil Patel, MD, The Damluji Research Center, Vista, California
Steven R. Pliszka, MD, University of Texas Health Science Center, San Antonio, Texas
George Realmuto, MD, University of Minnesota Medical School, Minneapolis, Minnesota
Murray Rosenthal, DO, Behavioral and Medical Research, LLC, San Diego, California
Richard L. Rubin, MD, Miami Research Associates, Miami, Florida
Keith Saylor, PhD, Neuroscience, Inc, Bethesda, Maryland
Ward T. Smith, MD, Pacific Northwest Clinical Research, Portland, Oregon
Michael Sternberg, MD, Advanced Research Center, Woodbridge, Virginia
Harvey Tilker, PhD, Four Rivers Clinical Research, Inc, Paducah, Kentucky
Sharon Wigal, PhD, University of California at Irvine, Irvine, California
Mark Wolraich, MD, Vanderbilt University, Nashville, Tennessee
Daniel R. Wynn, MD, Consultants in Neurology, LTD, Northbrook, Illinois
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Reprint requests to (J.B.) Pediatric Psychopharmacology Unit, Massachusetts General Hospital, 15 Parkman StACC 725, Boston, MA 02114. E-mail: biederma{at}helix.mgh.harvard.edu
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
V. L. Vetter, J. Elia, C. Erickson, S. Berger, N. Blum, K. Uzark, and C. L. Webb Cardiovascular Monitoring of Children and Adolescents With Heart Disease Receiving Medications for Attention Deficit/Hyperactivity Disorder: A Scientific Statement From the American Heart Association Council on Cardiovascular Disease in the Young Congenital Cardiac Defects Committee and the Council on Cardiovascular Nursing Circulation, May 6, 2008; 117(18): 2407 - 2423. [Full Text] [PDF] |
||||
![]() |
S. V. Faraone, S. B. Wigal, and P. Hodgkins Forecasting Three-Month Outcomes in a Laboratory School Comparison of Mixed Amphetamine Salts Extended Release (Adderall XR) and Atomoxetine (Strattera) in School-Aged Children With ADHD J Atten Disord, July 1, 2007; 11(1): 74 - 82. [Abstract] [PDF] |
||||
![]() |
T. E. Wilens, J. B. Prince, T. J. Spencer, and J. Biederman Stimulants and Sudden Death: What Is a Physician to Do? Pediatrics, September 1, 2006; 118(3): 1215 - 1219. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. E. Wilens, K. McBurnett, O. Bukstein, J. McGough, L. Greenhill, M. Lerner, M. A. Stein, C. K. Conners, J. Duby, J. Newcorn, et al. Multisite Controlled Study of OROS Methylphenidate in the Treatment of Adolescents With Attention-Deficit/Hyperactivity Disorder Arch Pediatr Adolesc Med, January 1, 2006; 160(1): 82 - 90. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. B. Wigal, J. J. McGough, J. T. McCracken, J. Biederman, T. J. Spencer, K. L. Posner, T. L. Wigal, S. H. Kollins, T. M. Clark, D. A. Mays, et al. A Laboratory School Comparison of Mixed Amphetamine Salts Extended Release (Adderall XR(R)) and Atomoxetine (Strattera(R)) in School-Aged Children With Attention Deficit/Hyperactivity Disorder J Atten Disord, August 1, 2005; 9(1): 275 - 289. [Abstract] [PDF] |
||||
![]() |
S. V. Faraone, T. J. Spencer, C. B. Montano, and J. Biederman Attention-Deficit/Hyperactivity Disorder in Adults: A Survey of Current Practice in Psychiatry and Primary Care Arch Intern Med, June 14, 2004; 164(11): 1221 - 1226. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Biederman, S. V. Faraone, M. C. Monuteaux, and J. R. Grossbard How Informative Are Parent Reports of Attention-Deficit/Hyperactivity Disorder Symptoms for Assessing Outcome in Clinical Trials of Long-Acting Treatments? A Pooled Analysis of Parents' and Teachers' Reports Pediatrics, June 1, 2004; 113(6): 1667 - 1671. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||