Author + information
- Received May 2, 2017
- Revision received June 16, 2017
- Accepted June 29, 2017
- Published online August 21, 2017.
- Evan A. Stein, MD, PhDa,∗ (, )
- Eldad J. Dann, MDb,
- Albert Wiegman, MD, PhDc,
- Flemming Skovby, MDd,
- Daniel Gaudet, MD, PhDe,
- Etienne Sokal, MD, PhDf,
- Min-Ji Charng, MD, PhDg,
- Mafauzy Mohamed, MMh,
- Ilse Luirink, MDc,
- Joel S. Raichlen, MDi,
- Mattias Sundén, PhDj,
- Stefan C. Carlsson, MD, PhDj,
- Frederick J. Raal, PhDk and
- John J.P. Kastelein, MD, PhDc
- aMetabolic and Atherosclerosis Research Center, Cincinnati, Ohio
- bRambam Health Care Campus, Haifa, Israel
- cDepartment of Paediatrics and Vascular Medicine, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands
- dDepartment of Clinical Genetics, University Hospital, and Institute of Clinical Medicine, University of Copenhagen, Copenhagen, Denmark
- eECOGENE-21 Clinical and Translational Research Center and Lipidology Unit, Department of Medicine, Université de Montréal, Chicoutimi, Quebec, Canada
- fUniversité Catholique de Louvain, Cliniques St. Luc, Service de Gastroentérologie et Hépatologie Pédiatrique, Brussels, Belgium
- gTaipei Veterans General Hospital, Taipei, Taiwan, R.O.C.
- hHospital Universiti Sains Malaysia (HUSM), Clinical Trial Unit, Level 2, Kelantan, Malaysia
- iAstraZeneca Pharmaceuticals LP, Gaithersburg, Maryland
- jAstraZeneca Gothenburg, Mölndal, Sweden
- kCarbohydrate and Lipid Metabolism Research Unit, Faculty of Health Sciences, University of Witwatersrand, Johannesburg, South Africa
- ↵∗Address for correspondence:
Dr. Evan A. Stein, Metabolic and Atherosclerosis Research Center, 5355 Medpace Way, Cincinnati, Ohio 45227.
Background Homozygous familial hypercholesterolemia (HoFH), a rare genetic disorder, is characterized by extremely elevated levels of low-density lipoprotein cholesterol (LDL-C) and accelerated atherosclerotic cardiovascular disease. Statin treatment starts at diagnosis, but no statin has been formally evaluated in, or approved for, HoFH children.
Objectives The authors sought to assess the LDL-C efficacy of rosuvastatin versus placebo in HoFH children, and the relationship with underlying genetic mutations.
Methods This was a randomized, double-blind, 12-week, crossover study of rosuvastatin 20 mg versus placebo, followed by 12 weeks of open-label rosuvastatin. Patients discontinued all lipid-lowering treatment except ezetimibe and/or apheresis. Clinical and laboratory assessments were performed every 6 weeks. The relationship between LDL-C response and genetic mutations was assessed by adding children and adults from a prior HoFH rosuvastatin trial.
Results Twenty patients were screened, 14 randomized, and 13 completed the study. The mean age was 10.9 years; 8 patients were on ezetimibe and 7 on apheresis. Mean LDL-C was 481 mg/dl (range: 229 to 742 mg/dl) on placebo and 396 mg/dl (range: 130 to 700 mg/dl) on rosuvastatin, producing a mean 85.4 mg/dl (22.3%) difference (p = 0.005). Efficacy was similar regardless of age or use of ezetimibe or apheresis, and was maintained for 12 weeks. Adverse events were few and not serious. Patients with 2 defective versus 2 negative LDL receptor mutations had mean LDL-C reductions of 23.5% (p = 0.0044) and 14% (p = 0.038), respectively.
Conclusions This first-ever pediatric HoFH statin trial demonstrated safe and effective LDL-C reduction with rosuvastatin 20 mg alone or added to ezetimibe and/or apheresis. The LDL-C response in children and adults was related to underlying genetic mutations. (A Study to Evaluate the Efficacy and Safety of Rosuvastatin in Children and Adolescents With Homozygous Familial Hypercholesterolemia [HYDRA]; NCT02226198)
Homozygous familial hypercholesterolemia (HoFH) is a rare, but severe, genetic disorder with an estimated frequency of 1 in 300,000 to 1,000,000 (1). It is caused by mutations that occur in either the low-density lipoprotein receptor (LDLR), apolipoprotein B (apo B), proprotein convertase subtilisin/kexin type 9 (PCSK9), or the LDLR adaptor protein1 (LDLRAP1, also known as ARH) genes, which all significantly impair low-density lipoprotein cholesterol (LDL-C) clearance (2). This in turn results in a phenotype with severely elevated LDL-C, skin and tendon xanthoma, and premature and often fatal cardiovascular disease (CVD) in childhood or early adulthood (1,2). Statins, when used as monotherapy at their highest doses, are modestly effective in HoFH, reducing LDL-C by 25%. The addition of ezetimibe reduces LDL-C an additional 15% to 20% (2). Due to the very high baseline LDL-C levels, the approximately 40% reduction with statin–ezetimibe combination therapy results in large absolute decreases in LDL-C and appears to contribute to the reduction in CVD morbidity and mortality in HoFH patients reported in observational studies (3,4). However, very few, if any, patients achieve anywhere near optimal LDL-C levels with these conventional drug therapies, and LDL apheresis is required; if available, apheresis is usually performed weekly or biweekly (1).
Numerous trials in HoFH, predominantly with adults, have been carried out with various statins over the past 30 years (5–7). Rosuvastatin, the most efficacious statin, has been shown to significantly reduce LDL-C levels in adult HoFH patients, but as with all statins, data in pediatric patients are limited and almost nonexistent in children younger than 12 years (7). No controlled trials of apheresis in young HoFH children have been reported. Among newer agents approved for HoFH, lomitapide is approved in adults, and mipomersen and evolocumab are approved, but not for use in HoFH children younger than 12 years (8,9). Although recent studies with a PCSK9 inhibitor have shown that the LDL-C response is related to residual LDLR activity in HoFH, similar data are lacking with statins, especially in a pediatric population (9). We now report the first HoFH trial specifically in children and adolescents, 6 to <18 years of age, with a statin, rosuvastatin. Because the LDL-C response to statin therapy varies considerably in HoFH, we also examined the association between underlying genotype and rosuvastatin 20 mg treatment in a broader HoFH population of both children and adults.
Study design and patients
The HYDRA study (A Study to Evaluate the Efficacy and Safety of Rosuvastatin in Children and Adolescents with Homozygous Familial Hypercholesterolemia) was a global trial in 7 countries in Asia, Europe, the Middle East, and North America. The study was reviewed and approved by local human ethics committees. Parents or legal guardians and, where applicable, the patient, signed an informed consent form before any study-related procedures being performed.
The trial involved HoFH children and adolescents, 6 to <18 years of age, randomized to a double-blind, 12-week crossover period of rosuvastatin 20 mg daily versus placebo, followed by a 12-week, open-label rosuvastatin 20 mg maintenance phase. The study design (Figure 1) consisted of a lead-in phase in which eligible patients discontinued all lipid-lowering therapy except ezetimibe and/or apheresis, which were continued throughout the study, and received rosuvastatin 10 mg daily in addition to dietary advice for 4 weeks. Patients already stable on rosuvastatin 20 mg proceeded directly to the randomization period. Eligible patients were then randomized 1:1 to rosuvastatin 20 mg daily for 6 weeks either preceded or followed by 6 weeks of placebo in a 12-week crossover phase. Patients who successfully completed the double-blind phase entered the 12-week maintenance phase.
Eligibility criteria included male and female patients with at least 1 of the following: genetic confirmation of 2 mutated alleles of either the LDLR, apo B, or PCSK9 genes; or phenotype with untreated LDL-C >500 mg/dl and triglycerides <400 mg/dl; plus 1 or more of the following: tendon or cutaneous xanthoma before the age of 10 years or heterozygous familial hypercholesterolemia (HeFH) diagnosed by genetic or clinical criteria in both parents.
Fasting blood samples were analyzed for lipid and safety parameters every 6 weeks after randomization by a central laboratory (Medpace Reference Laboratories, Cincinnati, Ohio, or Leuven, Belgium) that maintained Centers for Disease Control and Prevention part III lipid standardization and certification by the College of American Pathologists throughout the trial. Fasting LDL-C was determined by the Friedewald equation (10).
Genotyping was performed either by a central laboratory (Laboratory for Molecular Diagnostics Experimental Vascular Medicine, Academic Medical Center, Amsterdam, the Netherlands) if patients provided consent or obtained from the investigator if it had been performed previously. LDLR functionality was characterized as either defective (2% to 25%) or negative (<2%) if in vitro assessment of residual activity had been reported. Patients with different mutations in both LDLR alleles were classified according to their estimated receptor activity. Patients with an LDLR mutation combined with a mutation in apo B or PCSK9, or with 2 non-LDLR mutations, were analyzed separately.
To more robustly assess the LDL-C response on the basis of underlying genetic mutations, we expanded the patient population by combining data from the 13 children in the HYDRA trial with 7 children ages 8 to 17 years from a prior trial of HoFH patients who were treated with diet alone for 4 weeks followed by 6-weeks of rosuvastatin 20 mg daily (7). To assess the response in both children and adults, we further expanded the analysis to include 33 adult patients from the prior trial (7).
The primary objective was to assess the efficacy of rosuvastatin 20 mg daily on LDL-C compared with placebo. Secondary objectives were to assess the efficacy of rosuvastatin 20 mg on other lipids, lipoproteins, and apolipoproteins compared with placebo; the maintenance of efficacy over an additional 12 weeks; and in a post hoc analysis, to examine the association between HoFH genotypes and the LDL-C response to rosuvastatin. Additional objectives were to assess the short-term safety and tolerability of rosuvastatin compared with placebo.
In the HYDRA study, the efficacy of rosuvastatin was analyzed in the log scale using a linear mixed-effects model with terms for treatment and period as fixed effects, and patient as a random effect. A similar model with an added crossover period and treatment interaction term was used to assess potential carryover effects between the rosuvastatin and placebo treatment periods of the crossover phase (an effect of the treatment sequence). Safety assessments were conducted during the screening period and at 6-week intervals during the crossover and maintenance phases. Analysis of rosuvastatin efficacy with respect to genetic mutations utilized Student t tests applied to the change from baseline data. To assess a potential trend in rosuvastatin efficacy between genetic mutations (LDLR defective/LDLR defective, LDLR defective/LDLR negative, and LDLR negative/LDLR negative), a heteroscedasticity adjusted linear regression was used.
Twenty patients were enrolled in the screening period of the HYDRA study, 14 of whom met eligibility criteria and were randomized; 13 completed the 24-week study (Online Figure 1). Of the 6 patients not randomized, 3 patients withdrew consent, and 3 did not fulfill eligibility criteria. The 1 patient who discontinued therapy during the crossover phase did so due to repeated difficulty with venous access during blood collection. The baseline characteristics (Table 1) included a mean age of 10.9 years (range: 7 to 15 years), one-half were male, 71% were Caucasian, one-half were on apheresis, and 57% were on ezetimibe. Thirteen patients had confirmed genetic mutations consistent with HoFH, whereas the 1 patient without genetic testing met clinical eligibility criteria and had an LDL-C of 768 mg/dl on entry into the study.
Due to pre-randomization therapy that allowed continuation of ezetimibe and/or apheresis, and during which rosuvastatin treatment was instituted or maintained, LDL-C values at the end of the 6-week placebo phase were used as the baseline for comparison to the 6-week rosuvastatin 20 mg phase (Table 2). The mean ± SD LDL-C level on placebo was 481 ± 185 mg/dl. This compared to a mean LDL-C on rosuvastatin 20 mg of 396 ± 196 mg/dl, representing a mean absolute reduction in LDL-C of 85.4 mg/dl and a least-squares mean relative difference versus placebo of −22.3% (95% confidence interval: −33.5% to −9.1%; p = 0.005). No carryover effect was seen on the basis of the treatment sequence (p = 0.869). Efficacy observed in patient subgroups on ezetimibe, on apheresis, by age, and by sex is shown in Figure 2. During the maintenance period, the least-squares mean reduction in LDL-C was 19.3% (p = 0.009) (Figure 3). Reductions in apo B, and other apo B-containing lipoproteins, paralleled those for LDL-C, with mean absolute reductions of 33 mg/dl (17.1%) in apo B (p = 0.024) and 93.2 mg/dl (22.9%) in non–HDL-C (p = 0.003) (Table 2, Figure 3). Mean absolute reductions in triglycerides of 39.6 mg/dl (30.4%; p = 0.004) were seen with rosuvastatin 20 mg compared with placebo (Table 2). Nonstatistically significant increases in HDL-C of 1.8 mg/dl (7.4%; p = 0.314) occurred with rosuvastatin 20 mg versus placebo during the crossover phase together with a nominally significant 11.2% increase during the maintenance phase (p = 0.026).
LDL-C response associated with genetic mutations
The LDL-C response based on genetic mutations was evaluated in 19 of 20 pediatric patients, because 1 patient was not genotyped, and in 33 adults. The LDL-C reductions based on LDLR subgroups and other mutations in children are shown in Table 3 and in the combined 53 pediatric and adult patients in Table 4. Of the pediatric patients, 18 were confirmed as having ≥2 mutant alleles in the LDLR, apo B gene, or combinations thereof, and 1 patient had mutations in both LDLR adaptor protein alleles (autosomal recessive hypercholesterolemia [ARH]). Of the 33 adults, 32 had ≥2 mutant alleles in the LDLR, apolipoprotein B gene, or combinations thereof, and 1 patient had ARH. The remaining adult patient who met eligibility for the HoFH study by clinical criteria had only a single identifiable LDLR mutation and achieved a 43.2% reduction in LDL-C with rosuvastatin 20 mg. In children with 2 negative LDLR mutations, the mean LDL-C on diet alone or placebo was 427.8 ± 81.5 mg/dl, which was lower than anticipated; however, all 5 patients were also on apheresis, which is known to reduce LDL-C by about 30% from pre-treatment levels (1,2). For comparison, the baseline/placebo mean LDL-C was 507.6 ± 184.6 mg/dl and 565.7 ± 211.5 mg/dl for the 9 patients with 2 defective LDLR mutations and the 3 patients with 1 defective and 1 negative LDLR mutation, respectively. Only 3 of these 12 patients were on apheresis (Table 3).
Similar differences in baseline and placebo LDL-C between the LDLR genotypes were seen in the combined group of adults and children (Table 4), where all 8 patients with 2 negative LDLR mutations were on apheresis. The largest mean reductions in LDL-C with rosuvastatin were seen in the subgroup with the most residual LDLR activity; that is, in those with 2 defective LDLR mutations: 23.5% (p = 0.0044) and 21.3% (p < 0.0001) in the children and all patients, respectively (Tables 3 and 4). In the 3 children with 1 LDLR negative mutation, the reductions were less and not statistically significant, but when pooled with the adults to provide more robust numbers, the subgroup with 1 defective and 1 negative mutation in the LDLR demonstrated a mean reduction in LDL-C of 17.0% (p = 0.014). The lowest reductions in LDL-C were seen in the subgroup with 2 negative LDLR mutations: 12.9% (p = 0.022) and 14.0% (p = 0.038) in the children and all patients, respectively. Among the patients with 2 LDLR mutations, a trend of −27.8 mg/dl in absolute LDL-C reduction (p = 0.008) was found as the number of negative LDLR mutations increased from 0 to 1 to 2. Comparisons of the responses in adult and pediatric patients with these LDLR mutation combinations are shown in the Central Illustration. The child and adult with ARH experienced LDL-C reductions of 40.3% and 26.6%, respectively, while the single patient with an apo B and an LDLR defect had an LDL-C reduction of 43.2%.
Rosuvastatin 20 mg daily was well tolerated, with no patient terminating the trial or experiencing drug discontinuations or dose interruptions due to treatment-related adverse events. The only patient terminating prematurely did so due to repeated problems with venous access during blood collection. There were no serious adverse events, and the overall frequency of adverse events was low (Table 5). The only adverse event considered by the investigators to be related to the study drug was a low serum bicarbonate at a single visit, not confirmed on retesting, during the maintenance period. The results of laboratory parameters for hepatic, renal, and muscle tests are shown in Online Table 1. There were no elevations in hepatic transaminases or creatine kinase ≥2 times the upper reference range. In terms of renal function, all quantitative urine protein measures were within normal limits at all visits, with the exception of 1 patient who had a history of proteinuria. Glomerular filtration rate and serum creatinine were within the normal range at all visits for all patients.
The HYDRA study, the first randomized, placebo controlled trial of any statin in pediatric HoFH patients as young as 7 years of age, demonstrated significant and sustained reductions in LDL-C with rosuvastatin 20 mg daily of 85.4 mg/dl (22.3%) compared with placebo. Although no children in this study had clinical evidence of CVD, atherosclerotic cardiovascular events in children under age 10 years are not uncommon (4,11), highlighting the increased risk of death or serious morbidity in children with HoFH. As reduction in CVD risk is related to absolute rather than percent decrease in LDL-C, the mean absolute decrease of 85.4 mg/dl seen with rosuvastatin could be anticipated from trials in non-FH and HeFH patients to result in a clinically important reduction or delay in long-term cardiovascular morbidity and mortality (12,13). This would also be consistent with registry data in HoFH and represents a benefit that may be even greater when LDL-C reduction is started as early as 7 years of age, as in this trial (1,4).
The study also demonstrated that despite a mean reduction of 85.4 mg/dl with rosuvastatin, including in those already receiving ezetimibe, apheresis, or both, children with HoFH were unable to approach anywhere near optimal LDL-C levels. However, rosuvastatin did provide a good foundation on which to add further LDL-C reducing agents such as ezetimibe, apheresis, and the PCSK9 inhibitor evolocumab, which, when added to stable statin therapy, has been shown to be effective, well tolerated, and approved for children with HoFH 13 years of age or older (9). Of the 2 orphan drugs approved for HoFH, the microsomal triglyceride transfer protein inhibitor lomitapide has not been studied or approved in a pediatric population, and mipomersen, an apo B synthesis inhibitor, is not approved for use in children <12 years. In addition, both drugs are not well tolerated, and have significant side effects and substantial monitoring requirements (8).
The combination of the 13 children in the HYDRA study, together with the 7 children from a previous HoFH trial (7), allowed, for the first time, a broader assessment of rosuvastatin response in a pediatric HoFH population as it relates to underlying genetic defects. The larger group of children also enabled a comparison with the response in adults. Overall, it confirmed that HoFH children, like adults, with more residual LDLR activity respond with greater LDL-C reduction. However, unlike PCSK9 inhibition, which results in no LDL-C decrease in receptor-negative patients (9), rosuvastatin produced a nearly 13%, or 50 mg/dl, reduction in these patients, which would still be anticipated to result in a long-term CVD benefit. This supported the beneficial additional mechanism of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibition, not only by up-regulating LDLR activity through increased LDLR synthesis, but also by decreasing hepatic cholesterol and lipoprotein production (6,7).
Rosuvastatin 20 mg daily was well tolerated, and no safety signals were seen in HoFH children as young as 7 years of age. The safety and tolerability were consistent with, and confirmatory of, 2 larger and longer-term rosuvastatin trials in 375 HeFH pediatric patients ages 6 to 17 years (14,15). In both trials, children received rosuvastatin 5, 10, or 20 mg daily for periods up to 2 years. Rosuvastatin at all doses was well tolerated with no hepatic, skeletal muscle, or renal adverse events resulting in permanent treatment discontinuation. In neither of these pediatric trials were there clinically important changes in renal function or signs of adverse effects on growth or sexual maturation.
On the basis of the efficacy and safety from this trial, the U.S. Food and Drug Administration has approved rosuvastatin for the treatment of HoFH pediatric patients 7 to 17 years of age (16).
Limitations of the study included the short duration of rosuvastatin treatment, which might not reflect either longer-term efficacy or safety for a disorder where lifetime therapy is needed. However, the similar response in adults with HoFH indicate that increasing age, diet, or lifestyle does not alter LDL-C efficacy. Second, despite that this pediatric-specific HoFH trial was the largest to evaluate children, it was still relatively small. Lastly, it did not provide any information as to the ultimate impact on cardiovascular events in this very high-risk population.
The HYDRA study, the first-ever pediatric HoFH trial with a statin, demonstrated effective LDL-C reduction with rosuvastatin 20 mg alone or in combination with ezetimibe and/or apheresis, and resulted in U.S. Food and Drug Administration approval of rosuvastatin for the treatment of HoFH pediatric patients 7 to 17 years of age. Combined with additional pediatric HoFH patients from a prior trial, the LDL-C response to rosuvastatin was related to the underlying FH-causing mutations and was consistent with that seen in adults.
COMPETENCY IN MEDICAL KNOWLEDGE: Patients with HoFH have extremely elevated LDL-C levels from conception, and are at very high risk of accelerated atherosclerotic CVD. Treatment with statins should start at diagnosis; however, no statin has been formally evaluated for use in HoFH children.
TRANSLATIONAL OUTLOOK: Rosuvastatin 20 mg alone or in combination with ezetimibe and/or apheresis demonstrated effective LDL-C reduction in children and adolescents with HoFH.
The authors thank Kerren Davenport at Prime-Medica for editorial support.
For a supplemental figure and table, please see the online version of this paper.
The study was funded by AstraZeneca Pharmaceuticals LP. AstraZeneca was responsible for study conduct, data collection, and analysis. The study was designed by AstraZeneca in conjunction with Drs. Stein and Kastelein. The initial drafts of the manuscript were done by Dr. Stein, Prof. Raal, and Dr. Kastelein, who had full access to the study data. All authors reviewed and had input to the final draft. The academic authors vouch for the validity of the data in the final manuscript. Dr. Stein has received consultant and expert witness fees from AstraZeneca regarding statins. Dr. Gaudet has served as a consultant or advisor for AstraZeneca, Regeneron, Sanofi, Amgen, Aegerion, Chiesi, Gemphire, Novartis, Ionis, Cymabay, Uniqure, and Catabasis; and has received research funding from AstraZeneca. Dr. Sokal has received consultant and expert witness fees from AstraZeneca regarding statins and from Promethera Biosciences, Abbvie, Alexion, and Shire not related to statins; and is chief science officer of Promethera Biosciences. Dr. Mohamed has received consultant and advisor fees from AstraZeneca. Drs. Raichlen, Sundén, and Carlsson are employees of AstraZeneca. Prof. Raal has received research grants and consultant or lecture fees from AstraZeneca, Merck, Pfizer, Regeneron, Sanofi, and Amgen. Dr. Kastelein has received grant support from AstraZeneca, Pfizer, Roche, Novartis, Merck, Merck/Schering-Plough, Ionis, Genzyme, and Sanofi; lecture fees from AstraZeneca, GlaxoSmithKline, Pfizer, Novartis, Merck/Schering-Plough, Roche, Ionis, and Boehringer Ingelheim; and consulting fees from AstraZeneca, Abbott, Pfizer, Ionis, Genzyme, Roche, Novartis, Merck, Merck/Schering-Plough, and Sanofi. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- apo B
- apolipoprotein B
- autosomal recessive hypercholesterolemia
- cardiovascular disease
- heterozygous familial hypercholesterolemia
- homozygous familial hypercholesterolemia
- low-density lipoprotein cholesterol
- low-density lipoprotein receptor
- proprotein convertase subtilisin/kexin type 9
- Received May 2, 2017.
- Revision received June 16, 2017.
- Accepted June 29, 2017.
- 2017 American College of Cardiology Foundation
- Cuchel M.,
- Bruckert E.,
- Ginsberg H.N.,
- et al.
- Raal F.J.,
- Pilcher G.J.,
- Panz V.R.,
- et al.
- Rader D.J.,
- Kastelein J.J.
- Friedewald W.T.,
- Levy R.I.,
- Fredrickson D.S.
- Goldstein J.L.,
- Hobbs H.H.,
- Brown M.S.
- Avis H.J.,
- Hutten B.A.,
- Gagne C.,
- Langslet G.,
- et al.
- ↵AstraZeneca. Crestor prescribing information. 2016. Available at: http://www1.astrazeneca-us.com/pi/crestor.pdf. Accessed February 15, 2017.