The effect of everolimus on renal angiomyolipoma in patients with tuberous sclerosis complex being treated for subependymal giant cell astrocytoma: subgroup results from the randomized, placebo-controlled, Phase 3 trial EXIST-1

Abstract

Background

Tuberous sclerosis complex (TSC) is characterized by benign tumours in multiple organs, including the brain, kidneys, skin, lungs and heart. Our objective was to evaluate everolimus, an mTOR inhibitor, in the treatment of angiomyolipoma in patients with subependymal giant cell astrocytoma (SEGA) associated with TSC.

Methods

EXamining everolimus In a Study of Tuberous Sclerosis Complex-1 (NCT00789828), a prospective, double-blind, randomized, placebo-controlled, Phase 3 study, examined everolimus in treating SEGA associated with TSC. Patients with serial SEGA growth from pre-baseline to baseline scans were randomly assigned (2:1) to receive 4.5 mg/m²/day everolimus (target blood trough: 5–15 ng/mL; n = 78) or placebo (n = 39). Angiomyolipoma response rates were analysed in patients (n = 44) with target baseline angiomyolipoma lesions (≥1 angiomyolipoma; longest diameter ≥1.0 cm). An angiomyolipoma response rate, defined as the proportion of patients with confirmed angiomyolipoma response, was assessed by kidney CT or MRI screening at baseline, at 12, 24 and 48 weeks and annually.

Results

Angiomyolipoma response rates were 53.3% (16/30) and 0% (0/14) for everolimus- and placebo-treated patients, respectively. Angiomyolipoma reductions ≥50% in the sum of volumes of all target lesions were seen only in everolimus-treated patients (56.5, 78.3 and 80.0%) compared with placebo-treated patients (0% at each time point) at Weeks 12, 24 and 48, respectively. Greater percentages of everolimus-treated patients had angiomyolipoma reductions ≥30% at these same time points (82.6, 100 and 100% versus 8.3, 18.2 and 16.7% for everolimus versus placebo, respectively).

Conclusions

Everolimus showed efficacy in reducing angiomyolipoma lesion volume in patients with SEGA associated with TSC.

The trial is registered with ClinicalTrials.gov, number NCT00789828; http://clinicaltrials.gov/ct2/show/NCT00789828?term=EXIST-1&rank=1.

INTRODUCTION

Tuberous sclerosis complex (TSC) is a multi-system disorder characterized by the growth of non-malignant tumours (hamartomas) in various organs throughout the body. The estimated birth incidence is 1 in 6000, and ∼1 million individuals worldwide are affected [1, 2]. In the kidney, angiomyolipomata (also known as AMLs, but not to be confused with acute myelogenous leukaemia) occur in up to 75% of patients and are typically bilateral and multiple [3]. Histologically, the angiomyolipoma is characterized as a non-malignant mesenchymal tumour composed of abnormal blood vessels, immature smooth muscle cells and fat cells [3]. Proportions of each component may vary between lesions within the same kidney [3, 4]. As angiomyolipomata enlarge, they may develop aneurysms. Yamakado et al. [5] retrospectively evaluated the relationship between tumour size, aneurysm formation and spontaneous rupture of angiomyolipomata. Significant differences in tumour size (P < 0.02) and aneurysm size (P < 0.02) were noted in angiomyolipomata that ruptured compared with unruptured angiomyolipomata [5]. All ruptured angiomyolipomata had a tumour size >4 cm and an aneurysm size ≥5 mm [5]. In a separate retrospective analysis, all tumours in patients with TSC who experienced severe haemorrhage requiring transfusion measured >4 cm [6]. Other manifestations of TSC include subependymal giant cell astrocytomas (SEGAs), hypomelanotic macules, shagreen patch, periungual or subungual fibromas, facial angiofibromas and/or forehead plaques [7].

The pathogenesis of TSC is thought to result from mutations in either the TSC1 or TSC2 gene, which encodes the proteins hamartin and tuberin, respectively. Together, hamartin and tuberin form a tumour suppressor complex that negatively regulates the activity of mammalian target of rapamycin (mTOR) complex 1—a critical regulator of cell growth and proliferation [1, 2]. Everolimus, an oral mTOR inhibitor, has demonstrated efficacy in the treatment of SEGA and renal angiomyolipomata in patients with TSC [8–10]. In the Phase 3 EXamining everolimus In a Study of Tuberous Sclerosis Complex (EXIST-1) trial (NCT00789828), patients with serial SEGA growth from pre-baseline to baseline scans were randomly assigned 2:1 to everolimus, initiated at 4.5 mg/m²/day and titrated to blood trough levels of 5–15 ng/mL on the basis of tolerability, or to placebo [8]. Analysis of the primary end point demonstrated a clinically meaningful and statistically significant difference in the overall SEGA response rate in favour of everolimus [8]. Thirty-five per cent (27/78) of patients in the everolimus group had a ≥50% reduction in the volume of SEGAs versus none (0/39) in the placebo group (difference, 35; 95% confidence interval [CI], 15–52; one-sided exact Cochran–Mantel–Haenszel test, P < 0.0001) [8]. A predefined exploratory end point in the EXIST-1 trial was to assess the effect of everolimus on renal angiomyolipoma growth in the subset of patients with one or more target renal angiomyolipomata present at baseline. The data presented herein represent the subgroup analysis on patients with renal angiomyolipoma.

MATERIALS AND METHODS

The methods used in EXIST-1 have been previously published [8]. In brief, patients with serial SEGA growth from pre-baseline to baseline scans were randomly assigned 2:1 to everolimus, initiated at 4.5 mg/m²/day and titrated to blood trough levels of 5–15 ng/mL on the basis of tolerability, or to placebo [8]. Randomization was stratified by the use of enzyme-inducing anti-epileptic drugs. The primary end point of the trial, SEGA response rate, was defined as the proportion of patients with a ≥50% reduction in SEGA volume relative to baseline (where the SEGA volume was the sum of all target SEGA lesion volumes identified at baseline), no unequivocal worsening of non-target SEGA lesions, no new SEGA lesions (≥1 cm in longest diameter) and no new or worsening hydrocephalus. The protocol was approved by an ethics committee at each centre, before the first patient was enrolled. The study was done in accordance with the principles of Good Clinical Practice, Declaration of Helsinki and all local regulations. An independent data monitoring committee reviewed the safety every 6 months. All patients (or their legal representatives) provided written informed consent before enrolment. Eligible patients had a definitive diagnosis of TSC and one or more SEGA lesions ≥1 cm in diameter and serial growth of a SEGA, a new SEGA lesion ≥1 cm in diameter and/or new or worsening hydrocephalus.

A predefined exploratory end point of the trial was to analyse angiomyolipoma response rates in patients with one or more target baseline angiomyolipomata lesions. All measurable angiomyolipomata with a longest diameter ≥1.0 cm were identified as target lesions, and only patients with target angiomyolipomata were followed up for angiomyolipoma response during the trial. Up to five of the largest measurable lesions on each kidney were to be reported at baseline. The volume of these lesions was measured at each computed tomography (CT) or magnetic resonance imaging (MRI) assessment of the kidney during the trial, and the same imaging modality was to be used throughout.

Study end points

An angiomyolipoma response rate was defined as the proportion of patients with confirmed angiomyolipoma response. An angiomyolipoma response was defined as reduction in the sum of volumes of all target lesions ≥50% relative to baseline, with no new lesions ≥1 cm in longest diameter, no increase in kidney volume ≥20% from the lowest value obtained from the patient for each kidney separately and no angiomyolipoma-related bleeding of Grade ≥2 as defined by the National Cancer Institute Common Terminology Criteria for Adverse Events, version 3.0 (NCI-CTCAE v. 3). Kidney CT or MRI was to be repeated at 12, 24 and 48 weeks after the start of treatment and annually thereafter, and required confirmation ∼12 weeks and no sooner than 8 weeks after the initial response was observed. For each patient, the same imaging modality was to be used throughout the trial. All radiological evaluations were performed first by the local radiologist; however, the designation of response and progression was based solely on evaluations performed by independent central radiology review in patients with measurable disease.

The pharmacokinetics of everolimus in terms of pre-dose (Cmin) and 2 h post-dose (C2h) exposure was examined for patients with ≥1 target angiomyolipoma lesion at baseline who received everolimus. Pharmacokinetic pre-dose trough blood samples for determination of the everolimus concentration (Cmin) were collected immediately before dosing, starting at Visit 3 (Week 2) and at every visit thereafter until discontinuation of study drug. Samples for trough (24 h after the last dose) and C2h (2.0 h ± 30 min after dosing) were taken 2 weeks after an everolimus dose increase to a higher level than was previously taken, reduction in the dose of a CYP3A4 or PgP inducer (e.g. reduction in anticonvulsant dose) or initiation of, or increased dose of, a CYP3A4 or PgP inhibitor.

Adverse events (AEs) were assessed at every visit using the NCI-CTCAE v. 3 for all patients who received one or more doses of the double-blind study drug and had received a valid post-baseline assessment. Patients were analysed according to the treatment they actually received in the double-blind period.

Statistical analysis

All analyses were performed on patients who had one or more target angiomyolipomata ≥1 cm in longest diameter as assessed at baseline. Efficacy analyses included all patients who were randomly assigned to treatment. Safety and pharmacokinetic analyses included all patients who were given one or more doses of the study drug and had undergone at least one post-baseline assessment. Exact 95% CIs for response rates were obtained through the Clopper–Pearson method. Cmin values were included in the pharmacokinetic analysis if they were collected before dose administration on the same study day and at 20–28 h after the previous dose at steady state following 5 days of consistent dosing, and if the patient did not vomit within 4 h of the previous dose. C2h values were included in the pharmacokinetic analysis if they were collected at 1–3 h after dose administration on the same study day, at steady state, and if the patient did not vomit between taking the previous dose and the time of blood collection. Cmin and C2h were summarized by time window by descriptive statistics and graphically via box plots. Statistical analyses were performed with SAS software, v. 9.2. The data cut-off date for all analyses was 6 months after the last patient was randomly assigned to treatment.

RESULTS

Overall study findings

The study enrolled 117 patients from 24 centres and 10 countries between August 2009 and September 2010. Of these 117 patients, 78 were randomly assigned to treatment with everolimus and 39 to treatment with placebo. Baseline demographics and disease characteristics were considered representative of the larger TSC population, and the median age of patients in this trial was 9.5 years (range, 0.8–26.6 years). After a median follow-up duration of 9.7 months, treatment was ongoing for 97.4% (n = 76) of patients in the everolimus arm and for 79.5% (n = 31) of patients in the placebo arm. Median study drug exposure was longer by 5.8 weeks in the everolimus arm (41.9 weeks) compared with the placebo arm (36.1 weeks).

Subgroup findings

A total of 44 patients (44/117; 37.6%) identified at baseline had one or more angiomyolipomata that were ≥1.0 cm in longest diameter. Within this subgroup of patients, 30 received everolimus and 14 received placebo. At the data cut-off (2 March 2011), treatment was on-going for 93.3% (n = 28/30) of patients in the everolimus arm and for 92.9% (n = 13/14) of patients in the placebo arm. Two patients in the everolimus arm were discontinued—one withdrew consent and the other was lost to follow-up. One patient in the placebo arm was discontinued as the result of withdrawal of consent (Figure 1). The median dose intensity was similar between the two study arms (everolimus: 5.5 mg/m²/day; placebo: 5.7 mg/m²/day); the mean dose intensity was 6.1 mg/m²/day in the everolimus arm and 5.6 mg/m²/day in the placebo arm. The median exposure duration was 38 weeks for everolimus and 44 weeks for placebo. Baseline demographic and disease characteristics were generally well balanced between the two treatment arms with the exception of gender (Table 1). Male patients constituted 60.0% (n = 18) of patients in the everolimus treatment arm and 35.7% (n = 5) of patients in the placebo treatment arm; this difference was considered to be not clinically relevant. Most patients were <18 years of age (everolimus: 76.7%; placebo: 100.0%), and the median age of patients in the subgroup was 12.5 years (range, 4.5–23.9 years). A majority of patients were Caucasian, reflecting the countries that participated in the study. Bilateral angiomyolipomata were present in 53.3 and 64.3% of patients in the everolimus and placebo arms, respectively. The median sum of the volumes of target angiomyolipomata at baseline was 10.9 cm³ (range, 0.5–198.1 cm³) in the everolimus arm and 8.8 cm³ (range, 0.7–68.5 cm³) in the placebo arm.

FIGURE 1:

CONSORT flow diagram of patient disposition in the double-blind period up to the cut-off date (2 March 2011).

Table 1.

Baseline patient demographic and disease characteristics

Characteristic	Everolimus (n = 30)	Placebo (n = 14)
Age, years, median (range)	12.49 (5.9–23.9)	12.26 (4.5–17.8)
Age, n (%)
<3 years	0 (0.0)	0 (0.0)
3–<18 years	23 (76.7)	14 (100.0)
≥18 years	7 (23.3)	0 (0.0)
Sex, n (%)
Male	18 (60.0)	5 (35.7)
Female	12 (40.0)	9 (64.3)
Race, n (%)
Caucasian	27 (90.0)	14 (100.0)
Black	2 (6.7)	0 (0.0)
Other	1 (3.3)	0 (0.0)
Body surface area, m2, median (range)	1.38 (0.75–2.16)	1.51 (0.7–1.78)
Diagnosis of TSC, n (%)	30 (100.0)	14 (100.0)
Bilateral angiomyolipoma, n (%)	16 (53.3)	9 (64.3)
Target angiomyolipoma lesions, n (%)
0	0 (0.0)	0 (0.0)
1–5	23 (76.7)	11 (78.6)
6–10	7 (23.3)	3 (21.4)
Sum of volumes of target angiomyolipoma lesions
Patients with ≥1 target angiomyolipoma, na	27	14
Median (range), cm3	10.89 (0.49–198.08)	8.77 (0.74–68.53)

^aMissing values are due to poor quality of the scan.

Efficacy

The angiomyolipoma response rate was 53.3 (95% CI, 34.3–71.7%) in the everolimus arm (16/30) versus 0.0 (95 CI, 0.0–23.2%) in the placebo arm (0/14). Angiomyolipoma reductions of ≥50% occurred only in patients in the everolimus arm. After 12, 24 and 48 weeks of treatment, the proportions of patients in the everolimus arm with ≥50% reduction in the sum of volumes of target angiomyolipomata were 56.5, 78.3 and 80.0%, respectively. After 12, 24 and 48 weeks of treatment, the proportions of patients in the everolimus arm with a ≥30% reduction in the sum of volumes of target angiomyolipomata were 82.6, 100.0 and 100.0%, respectively, versus 8.3, 18.2 and 16.7%, respectively, in the placebo arm (Figure 2). Spontaneous shrinkage of angiomyolipomata has not been recognized previously in clinical care but was noted in the EXIST-2 study (NCT00790400) [10]. Some of these changes may be due to measurement error. However, this may be a real but temporary phenomenon that can be detected if multiple measurements are made over a short time period. At Week 24, the median percentage change from baseline in the sum of volumes of target angiomyolipoma lesions in the everolimus arm was −58.88% (range, −78.62 to −32.54%) compared with 8.08% (range, −49.19 to +79.03%) in the placebo arm (Figure 3). At this same time point, the median change from baseline in the sum of volumes of target angiomyolipoma lesions in the everolimus arm was −7.89 cm³ (range, −104.54 to −0.23 cm³) compared with 0.65 cm³ (range, −4.10 to9.01 cm³) in the placebo arm (Figure 4). The angiomyolipoma response rate within the everolimus arm was similar regardless of age or gender; response rates in male and female patients were 50.0% (9/18) and 58.3% (7/12), respectively, and 52.2% (12/23) in patients <18 years of age and 57.1% (4/7) in patients ≥18 years of age.

FIGURE 2:

Percentage change from baseline of sum of volumes (cm³) of target angiomyolipoma lesions by time window.

FIGURE 3:

Box plot of percentage change from baseline in sum of volumes of target angiomyolipoma lesions by time window. Fifth percentile, first quartile, median, third quartile, 95th percentile and mean (•) are displayed in the box plot, as are extreme values (*).

FIGURE 4:

Box plot of change from baseline in sum of volumes of target angiomyolipoma lesions by time window. Fifth percentile, first quartile, median, third quartile, 95th percentile and mean (•) are displayed in the box plot, as are extreme values (*). ^#Values that lie outside the range of the y-axis.

Pharmacokinetics

Patients in this study received a 4.5 mg/m²/day starting dose; the dose was subsequently to be adjusted to maintain pre-dose everolimus trough concentrations (Cmin) between 5 and 15 ng/mL. As expected with this starting dose, the median Cmin was 3.6 and 2.9 ng/mL at Week 2 and Week 4 and then maintained at a relatively constant level between 4.5 and 7.7 ng/mL from Weeks 4 to 48, which probably reflected the dose adjustments made during the first month of treatment (Figure 5). Interpretation of the median Cmin beyond Week 48 should be made with caution because of the lower patient numbers after this time point. The median C2h was ∼24.3 and 15.0 ng/mL for Week 2 and Week 4 and ranged between 22.8 and 29.3 ng/mL from Weeks 6 to 12 (Figure 6). The median C2h beyond Week 12 should be interpreted with caution because of differences in the numbers of patients for whom data were available at each time point.

FIGURE 5:

Box plot of concentration–time profiles for everolimus at pre-dose (trough) by time window. Median is displayed as (−), mean as (•), boxes are drawn from 25th percentiles to 75th percentiles and whiskers extend from 10th percentiles to 90th percentiles. ^#Values that lie outside the range of 10th percentiles and 90th percentiles.

FIGURE 6:

Box plot of concentration–time profiles for everolimus at 2 h post-dose by time window. Medians are displayed as (−) and means as (•), boxes are drawn from 25th percentiles to 75th percentiles and whiskers extend from 10th percentiles to 90th percentiles. ^#Values that lie outside the range of 10th percentiles and 90th percentiles.

Safety

AEs were consistent with the known safety profile of everolimus in TSC. Mouth ulceration, convulsion, stomatitis, fatigue and rash were the most common AEs with everolimus therapy (each reported in ≥20% of patients) (Table 2). Of note, convulsion was not more common in everolimus-treated patients but was reported in 30.0% of patients in the everolimus arm and in 28.6% of patients in the placebo arm. AEs suspected to be drug related were reported in 93.3% (28/30) of patients in the everolimus arm and in 35.7% (5/14) of patients in the placebo arm. Grade 3–4 AEs occurred in 20.0% (6/30) of patients in the everolimus arm and in 7.1% (1/14) of patients in the placebo arm. Only 6.7% (2/30) of Grade 3–4 AEs in the everolimus arm were suspected to be related to drug treatment.

Table 2.

AEs, regardless of relationship to study drug, experienced by ≥10% of patients in the everolimus arm

AEs	Everolimus (n = 30), n (%)	Placebo (n = 14), n (%)
Mouth ulceration	13 (43.3)	2 (14.3)
Convulsion	9 (30.0)	4 (28.6)
Stomatitis	8 (26.7)	1 (7.1)
Fatigue	6 (20.0)	0 (0.0)
Rash	6 (20.0)	1 (7.1)
Nasopharyngitis	5 (16.7)	4 (28.6)
Acne	4 (13.3)	0 (0.0)
Dizziness	4 (13.3)	1 (7.1)
Pyrexia	4 (13.3)	1 (7.1)
Upper abdominal pain	3 (10.0)	1 (7.1)
Aggression	3 (10.0)	1 (7.1)
Amenorrhoea	3 (10.0)	0 (0.0)
Increased blood lactate dehydrogenase	3 (10.0)	0 (0.0)
Constipation	3 (10.0)	0 (0.0)
Cough	3 (10.0)	1 (7.1)
Gastroenteritis	3 (10.0)	0 (0.0)
Headache	3 (10.0)	1 (7.1)
Nausea	3 (10.0)	0 (0.0)
Otitis media	3 (10.0)	0 (0.0)
Pharyngitis streptococcal	3 (10.0)	0 (0.0)
Respiratory tract infection	3 (10.0)	0 (0.0)
Upper respiratory tract infection	3 (10.0)	1 (7.1)
Vomiting	3 (10.0)	2 (14.3)

DISCUSSION

EXIST-1, a Phase 3, placebo-controlled trial, demonstrated the efficacy of everolimus for the treatment of SEGA associated with TSC. Analysis of the primary end point showed a clinically meaningful and statistically significant difference (P < 0.0001) in overall SEGA response rate in favour of everolimus [8]. Among the subset of patients within the EXIST-1 trial who had one or more angiomyolipomata ≥1.0 cm in longest diameter, angiomyolipoma reductions ≥50% were observed only in the everolimus arm (Weeks 12, 24 and 48), and reductions ≥30% were more frequently reported in the everolimus arm than in the placebo arm [8].

TSC-associated angiomyolipomata are more likely to grow than remain stable, making these patients susceptible to complications such as spontaneous life-threatening haemorrhage [3, 11, 12]. A retrospective analysis of patients with TSC noted that among children <2 years of age, angiomyolipomata were detected in 17%, which increased to 92% among children aged 14–18 years [13]. The risk of haemorrhage is increased as angiomyolipomata enlarge, and a significant relationship has been noted between lesion size, aneurysm formation and haemorrhagic rupture [5]. In a retrospective UK renal registry of 52 patients (2000–2012), growth of angiomyolipoma was associated with haemorrhage; among 27 patients with serial angiomyolipoma growth, 41% (11/27 or 21% = 11/52) experienced haemorrhage and 59% (16/27 or 31% = 16/52) did not. Among 25 subjects with no serial angiomyolipoma growth, 8% (2/25 or 11% = 2/52) had a bleeding event and 92% (23/25 or 44% = 23/52) did not [14]. Therefore, serial growth is a risk factor for haemorrhagic events. To determine the relationship between growth rate and size of angiomyolipoma, Kingswood et al. [14] retrospectively analysed a database of patients with TSC seen at St. George's Clinic in the UK from May 1998 to December 2012 and found no correlation between rate of growth of angiomyolipoma and the size of the angiomyolipoma [14]. Furthermore, the rate of angiomyolipoma growth was not correlated with patient age, and growth tended to be greater in younger patients [14]. Angiomyolipomata exhibit continued growth throughout the life of a patient with TSC, and as angiomyolipomata increase in size, so too does the risk of complications; therefore, patients should be continuously monitored.

The main goals in treating patients with angiomyolipomata are preserving renal function and limiting complications. Symptomatic lesions or lesions ≥4 cm in diameter are typically managed through surgical procedures and/or embolization [12, 15]. Other indications for surgical intervention include significant haemorrhage, suspicion of malignancy and lesions with local tissue or vascular invasion [16]. Nephrectomy may not be an optimal therapeutic modality for patients with TSC, and medical therapy with mTOR inhibitors may be preferred to embolization as pre-emptive treatment for enlarging angiomyolipomata because they are often multiple, bilateral and progressive and regrow after embolization [17]. A recent expert review has suggested pre-emptive treatment with mTOR inhibitors as a means of preventing further growth and minimizing kidney damage with fast-growing angiomyolipomata ≥3 cm [18]. Everolimus, an oral mTOR inhibitor, has demonstrated efficacy in treating patients with TSC-associated angiomyolipoma and has a safety profile similar to that previously reported for everolimus in the TSC population. A larger (n = 118) Phase 3, randomized, double-blind study in patients with TSC ≥18 years of age with one or more angiomyolipomata ≥3.0 cm in longest diameter has produced similar results (EXIST-2) [10]. Everolimus treatment demonstrated a clinically and statistically significant reduction in angiomyolipoma volume compared with placebo (P < 0.0001) [10]. The safety profile of everolimus was consistent with previous reports. The most common AEs in the everolimus and placebo arms were stomatitis (48 versus 8%, respectively), nasopharyngitis (24 versus 31%, respectively), acne-like skin lesions (22 versus 5%, respectively), headache (22 versus 18%, respectively), cough (20 versus 13%, respectively) and hypercholesterolaemia (20 versus 3%, respectively) [10]. Everolimus is the first pharmacological treatment option to demonstrate benefit in this patient population in a controlled clinical trial. The overall clinical benefit/risk assessment supports the use of everolimus as a viable treatment option for patients with TSC.

This trial has shown that everolimus is effective in reducing angiomyolipoma lesion volume in patients with SEGA associated with TSC who also presented with angiomyolipoma. The safety profile of everolimus was consistent with that previously observed in TSC; most AEs were of Grade 1 or 2 severity. Everolimus represents a pharmacological treatment option for patients with TSC who have SEGA and concomitant angiomyolipoma.

Our hypothesis is that the main goal in the treatment of TSC-associated renal angiomyolipomata is to prevent them from enlarging, thereby ameliorating the risk of future bleeding and chronic kidney disease. Although EXIST-2 showed that progression of renal angiomyolipomata could be halted in adults with TSC, this subgroup analysis of EXIST-1 raises the possibility that in younger patients, renal disease might be preventable. However, this preliminary finding of efficacy and tolerability must be confirmed in longer-term studies and/or by clinical observation.

CONFLICT OF INTEREST

J. Chris Kingswood, John J. Bissler, David N. Franz, Steven Sparagana, E. Martina Bebin, Michael Frost, Michael H. Kohrman, Joyce Y. Wu and Serguisz Jozwiak are consultants for Novartis Pharmaceuticals (including advisory boards) and have received travel payments, research funding or speaker honoraria from Novartis. David N. Franz has received compensation from various attorneys for legal work reviewing medical malpractice cases and occasionally gives expert testimony. Steven Sparagana and Joyce Y. Wu have received honoraria from Lundbeck Pharmaceuticals. Michael Frost is a member of speaker bureaus and advisory boards for Lundbeck Pharmaceuticals, Novartis Pharmaceuticals and UCB. Joyce Y. Wu serves on a professional advisory board for and receives research support from the Tuberous Sclerosis Alliance. Thomas Brechenmacher, Karen Stein and Noah Berkowitz are employees of Novartis Pharmaceuticals. The other authors declare that they have no conflicts of interest.

The results presented in this manuscript have not been previously published except in abstract form.