Download Congenital Heart Disease - Circulation: Cardiovascular Interventions

Congenital Heart Disease
Pharmacokinetics of Sirolimus-Eluting Stents Implanted
in the Neonatal Arterial Duct
Kyong-Jin Lee, MD; Winnie Seto, PharmD, MSc; Lee Benson, MD;
Rajiv R. Chaturvedi, MRCP(UK), MD, PhD
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
Background—Sirolimus-eluting stents may have clinical advantages over bare-metal stents in the extremely proliferative
environment of the neonatal arterial duct. However, sirolimus has immunosuppressive actions and little is known regarding
sirolimus pharmacokinetics in the newborn.
Methods and Results—This is a retrospective review of sirolimus pharmacokinetics in neonates who underwent sirolimuseluting stent implantation in the arterial duct for pulmonary blood flow augmentation. Pharmacokinetic parameters were
obtained by noncompartmental analysis and by a Bayesian one-compartment nonlinear mixed model. Nine neonates
received a single sirolimus-eluting stent with a total sirolimus dose of 245 μg (n=1), 194 μg (n=5), or 143 μg (n=3). Peak
sirolimus concentrations were 13.6±4.5 μg/L (24.8 μg/L highest) and clearance was 0.042±0.03 L/hour (noncompartmental
analysis) and 0.051 L/hour (95% credible intervals 0.037–0.069, nonlinear mixed model). Sirolimus remained >5 μg/L,
the trough level used in oral immunosuppressive therapy, for (95% credible interval) 15.9 (11.4, 22.8), 12.9 (7.6, 19.0),
and 8.4 (2.3, 14.5) days for the 245, 194, and 143 μg sirolimus dose stents, respectively. Estimates of the duration of
systemic immunosuppression are provided for combinations of 2 stents.
Conclusions—In neonates after sirolimus-eluting stent implantation, peak sirolimus levels were 20× higher and clearance
30× lower than previously reported in older children and adults. Sirolimus levels were within the immunosuppressive
range for a prolonged period, but with no observable clinically significant adverse outcomes. (Circ Cardiovasc Interv.
2015;8:e002233. DOI: 10.1161/CIRCINTERVENTIONS.114.002233.)
Key Words: congenital heart disease ◼ drug-eluting stent ◼ ductus arteriosus ◼ neonate
◼ pharmacokinetics ◼ sirolimus
S
tent implantation is a recognized management option in
maintaining arterial duct patency in newborns with ductdependent pulmonary blood flow, but with bare metal stents
the reintervention rate (redilation or need for a surgical shunt)
is 17% to 25% at 6 months.1–3 The neointimal proliferative
process associated both with spontaneous ductal closure and
in reaction to stent implantation compromises luminal diameter.3–7 In this regard, sirolimus, an immunosuppressive agent
with anti-inflammatory and antiproliferative effects, has theoretical benefits.8,9 Drug-eluting stent technology enables topical drug delivery with therapeutic drug concentrations locally
within the blood vessel wall, despite substantially lower systemic blood levels.10 Sirolimus-eluting stents (SES) are effective and safe in adult coronary arteries.11 SES implanted in
the porcine neonatal arterial duct have higher patency rates
compared with bare-metal stents and also exhibit an antiproliferative action on arterial duct smooth muscle.12
Pharmacokinetic studies of SES in adult pigs10 and humans13
have shown low peak serum drug levels occurring at 1 and
between 3 and 4 hours, respectively, with minimal detectable
levels by 3 and 7 days, respectively. SES implantation has
been reported in children, but little is known regarding subsequent pharmacokinetics in the pediatric population.14–19 The
youngest patient previously reported was 3 months old at the
time of SES implantation in the right ventricular outflow tract
(peak ≈8 μg/L, undetectable between 12 and 37 days15), and to
our knowledge, there are no reports of neonatal use. Sirolimus
experience in the pediatric population is predominantly in the
form of oral immunosuppressive therapy in transplant recipients. Dosage is titrated by steady-state trough levels (target
5–15 μg/L).20,21
In this study, we report the pharmacokinetics of sirolimus
after single SES implantation in the neonatal arterial duct.
Model-free point estimates were obtained by noncompartmental analysis, and in addition, a one-compartment Bayesian
population pharmacokinetic model was used to provide point
estimates with credible intervals, and this also allowed the
prediction of pharmacokinetics if 2 stents had been implanted.
Received November 27, 2014; accepted April 1, 2015.
From The Labatt Family Heart Centre, Division of Cardiology (K.-J.L., L.B., R.R.C.) and the Department of Pharmacy (W.S.), The Hospital for Sick
Children, University of Toronto School of Medicine, Toronto, Canada.
The Data Supplement is available at http://circinterventions.ahajournals.org/lookup/suppl/doi:10.1161/CIRCINTERVENTIONS.114.002233/-/DC1.
Correspondence to Rajiv R. Chaturvedi, MD, PhD, The Hospital for Sick Children, 555 University Ave, Toronto, Canada M5G 1X8. E-mail rajiv.
[email protected]
© 2015 American Heart Association, Inc.
Circ Cardiovasc Interv is available at http://circinterventions.ahajournals.org
1
DOI: 10.1161/CIRCINTERVENTIONS.114.002233
2 Lee et al Pharmacokinetics of Neonatal Sirolimus Stents
WHAT IS KNOWN
• Bare metal stents implanted in the human neonatal
arterial duct are associated with a 17% to 25% reintervention rate at 6 months.
• Sirolimus-eluting stents implanted in the porcine
neonatal arterial duct have higher patency rates than
bare metal stents.
• Sirolimus-eluting stents in adults result in low systemic sirolimus levels.
WHAT THE STUDY ADDS
• Implantation
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
of sirolimus-eluting stents in human
neonatal arterial ducts resulted in high peak systemic
sirolimus levels and slow clearance.
• Although systemic sirolimus levels remained in the
immunosuppressive range for a prolonged period of
time, there did not seem to be any adverse clinical
sequelae attributable to immunosuppression.
subject using trapezoidal integration (R-3.1.1,22 package PK 1.3–2)23
and (2) a single intravenous dose one-compartment nonlinear mixed
effects model (NLMM)24 to obtain population estimates with 95%
credible intervals (model in the Data Supplement). Wakefield’s
WinBUGS script24 was adapted to run with JAGS 3.4.0,25 and because
the concentration-time data were a ragged array (a different number
of data points for each subject), nested indexing was used to specify
the subject to which each observation belonged (Data Supplement).26
The JAGS output was analyzed with coda-0.16–1.27 Rate constants
are positive, and hence, lognormal priors are often used in pharmacokinetic studies. Model fitting with log normal priors was sensitive
to initial values, and chain adaptation/equilibriation often failed. This
was particularly the case for the sparse data points in the rising phase
of sirolimus concentrations and ka, the associated rate constant for the
rising phase. Gaussian priors were used instead, with convergence of 3
different chains demonstrated by traceplots, and resulted in unimodal
densities for each parameter and fitted data point. Credible intervals
for population pharmacokinetics were obtained from 104 simulations
of the one-compartment model, using the population parameters from
the NLMM for each of the sirolimus doses (245, 194, and 143 μg).
The time for sirolimus concentration to fall to 5 μg/L was estimated
by linear interpolation for both the raw data and the one-compartment
model. Bioavailability of 100% was assumed throughout.
Results
Methods
Sirolimus-eluting Cypher Select (Cordis Inc., Miami, FL) stents
were fully approved for human use by Health Canada during this
period. Pediatric systemic sirolimus levels after SES had not been
reported at the start of this clinical series and were expected to be
low based on the adult literature and estimates by a Pediatric Clinical
Pharmacologist (12). We thought it prudent to measure sirolimus
levels prospectively for the purposes of clinical monitoring/quality
assurance because symptoms consistent with infection or necrotizing enterocolitis occur frequently in these neonates. This enabled
physicians involved in their direct care to interpret clinical events in
the context of whether there was systemic immunosuppression. Each
candidate for ductal stenting was discussed at a joint Cardiology and
Cardiovascular Surgery case conference before referral for catheterization. Parents provided consent for use of a SES and for a clinical
protocol of serial sirolimus measurements that were coordinated with
blood samples taken for other clinical reasons. The hospital Research
Ethics Board approved the pharmacokinetic analysis and dissemination of this clinical data. All stents were 3.5 mm diameter, and the
sirolimus dose was dependent on the stent length. A drug-free polymer layer has been applied on top of the drug-polymer matrix as a
diffusion barrier to prolong the release of the drug with 50% of the sirolimus eluted over the first week, 80% over 30 days, and 100% over
90 days.11 Sirolimus levels were coordinated with blood sampling for
other clinical reasons and drawn at ≈1, 3, 24 hours poststent insertion
followed by every 7 days until the sirolimus level was below the limit
of quantification (<1 μg/mL) where feasible. Sirolimus levels from
whole blood samples (40 μL) of patients were analyzed by a LCMS/MS (liquid chromatography tandem mass spectrometer)—4000
QTrap at the Therapeutic Drug Monitoring Laboratory, Hospital for
Sick Children. The assay has a precision (defined by percentage of
coefficient of variation) of 6.8% for quality control level of 6 μg/L,
6.8% for quality control level of 12.8 μg/L, and 7.4% for quality control level of 23 μg/L, respectively.
Secondary data collection included cardiac diagnoses, review of
the procedural aspects of the patent ductus arteriosus (PDA) stent
implantation and any other subsequent catheter or surgically based
cardiac interventions, and clinical outcomes, including ductal stent
patency, which was determined by echocardiography.
Data Analysis
Two techniques were used for pharmacokinetic modeling of sirolimus levels: (1) point estimates by noncompartmental analysis of each
Nine newborns underwent single SES implantation in the arterial duct. The specifications of the implanted Cypher Select
stents were 3.5×13 mm (143 μg) in 3 patients, 3.5×18 mm
(194 μg) in 5 patients, and 3.5×23 mm (245 μg) in 1 patient.
Patient Characteristics
The median age at implantation was 10 days (range 1–17).
The median weight at time of implantation was 3.2 kg (range
2.4–4.3). Cardiac diagnoses, cardiac procedures, and arterial duct stent follow-up for each patient are summarized in
Table 1. All patients are alive at a median follow-up of 3.2
years (range 172 days to 5.3 years).
Noncardiac comorbidities included one infant born at gestational age 35 weeks after having undergone in-utero laser
ablation for twin to twin transfusion, who developed necrotizing enterocolitus and pneumoperitoneum with ischemic
bowel perforation requiring surgical anastomosis at 11 days (6
days before arterial duct stent implantation). One patient had
dysmorphism, vertebral anomalies, and bilateral renal pelviectasis. One patient had microdeletion 22q11 and a solitary
kidney.
PDA Anatomy
All PDA stents were implanted for the purpose of providing pulmonary blood flow. The stents were implanted into
the 3 types of PDA anatomy: left aortic arch with PDA from
descending aorta (n=5), right aortic arch with PDA from left
innominate artery (n=3), and right aortic arch with PDA from
aorta (n=1; Figure 1).
Stent Implantation Procedure
All infants were anaesthetized, intubated, and mechanically
ventilated during the PDA stent implantation cardiac catheterization. Vascular access was obtained at the discretion of the
operator for optimal stent implantation (femoral artery [n=4],
femoral vein [n=1], femoral artery and vein [n=3], and carotid
artery [n=1]). Angiography delineated arterial duct anatomy.
3 Lee et al Pharmacokinetics of Neonatal Sirolimus Stents
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
A 0.014″ Wizdom wire (Cordis Corporation, Miami, FL) was
placed across the arterial duct either from the pulmonary artery
or aorta. Additional stiffer 0.014″ wires were used as needed.
The stents were placed such that the implants protruded into
the main pulmonary artery, but were flush with the aorta. One
Cypher Select was implanted into each patient. Additional
bare-metal stents were implanted as deemed necessary. Four
additional bare-metal stents were implanted in 3 patients. In
one patient, the Cypher Select stent migrated forward during
removal of the balloon catheter. The stent was manipulated
partially back into the PDA and 2 bare-metal stents were subsequently implanted to cover the entire arterial duct, leaving
the SES in situ. One patient required an additional bare-metal
stent (3.5×15 mm) to allow for full coverage of the PDA. One
patient had bilateral PDA, and the right PDA was stented with
a 3.5×15 mm bare-metal stent.
Two of the SES were not implanted in the desired position.
As previously mentioned, in one patient, the DES migrated
during balloon catheter removal, necessitating additional
bare-metal stent implantations. In the second patient, the SES
embolized into the right pulmonary artery during guidewire
removal. This stent was surgically removed 12 days later.
Sirolimus levels were monitored before surgery.
Three patients required additional procedures on the PDA
stented with an SES (Table 1). An additional bare-metal stent
was implanted within the sirolimus-eluting stented PDA in
one patient for significant neointimal proliferation, and the
sirolimus-eluting stent was redilated in 2 patients.
Anticoagulation/Antiplatelet Regimen
After stent implantation, patients were maintained on ≥1 anticoagulation or antiplatelet drug therapies, which consisted of
low molecular weight heparin (n=7), clopidogrel (n=5), and
acetylsalicylic acid (n=4). Drug therapy was maintained until
the patency of the stented PDA was no longer deemed to be
necessary for clinical well-being or until complete closure of
the stented PDA was documented.
Clinical Outcomes
Seven patients underwent 2 ventricle repair, of which 2 have
residual atrial septal defects and 2 patients underwent a Fontan
procedure with one developing plastic bronchitis (Table 1).
PDA Stent Outcomes
The sirolimus-eluting PDA stent embolized at the time of
implantation in 1 patient with surgical removal 12 days later
Table 1. Patient Characteristics and Clinical Outcomes
Cardiac Diagnoses
1
PA/IVS
Age at Stent
Implantation,
days
Weight at Stent
Implantation, kg
10
3.7
Other Cardiac
Interventions
PV RF perforation
PV 2nd dilation
PDA Stent Follow-Up
Stent dilation resulted in pulmonary
overcirculation and urgent BCPC, 149 days
after implantation
Latest Follow-Up
Well
PDA stent dilation
BCPC, Fontan
2
PA/IVS
17
2.5
3
Mitral atresia, pulmonary
atresia, right aortic arch
4
2.6
PV RF perforation
Additional PDA BMS×2
Additional BMS in PDA
PA/IVS
15
3.0
PV RF perforation
PV 2nd dilation
5
Critical PS
6
2.8
ASD device closed
6
PA/VSD/right aortic arch,
nonconfluent pulmonary
arteries with bilateral PDA
14
3.5
Right PDA BMS
DORV/TGA/subPS/PS
0
3.4
PDA stent dilation
Arterial switch
operation
8
PA/VSD/right aortic arch
6
4.3
PA/VSD repair
9
TOF/right aortic arch,
discontinuous pulmonary
arteries with isolated LPA
13
3.2
TOF repair
10 (0, 17)
3.2 (2.5, 4.3)
Median (range)
Stent patent until BCPC at 126 days after
implantation
Plastic bronchitis.
Stent occluded at 1144 days after
implantation (patent at 426 days)
Small ASD. Well.
Stent patent at 813 days after implantation
Right and left PDA stents patent up to
surgical repair at 110 days post implantation
Well
Well
Left PDA stent 2nd
BMS
PA/VSD repair
7
Small ASD. Well.
ASD dilation
Bilateral BCPC, atrial
septectomy, Fontan
4
Stent occluded at 571 days after
implantation (patent at 359 days)
Stent patent up to surgical repair at 171 days
after implantation
Stent embolization during implantation
Stent patent up to surgical repair at 39 days
after implantation
Well
Well
Well
ASD indicates atrial septal defect; BCPC, bidirectional cavopulmonary connection; BMS, bare-metal stent; DORV, double outlet right ventricle; LPA, left pulmonary
artery; PA/IVS, pulmonary atresia with intact ventricular septum; PA/VSD, pulmonary atresia with ventriculoseptal defect; PDA, patent ductus arteriosus; PS, pulmonary
valve stenosis; PV, pulmonary valve; RF, radiofrequency perforation; TGA, transposition of great arteries; and TOF, tetralogy of Fallot.
4 Lee et al Pharmacokinetics of Neonatal Sirolimus Stents
Figure 1. Angiograms of arterial duct stenting are shown. A, Lateral projection of arterial duct (*) originating from left descending
aorta. B, Stented arterial duct of A. C, Frontal projection of arterial duct (*) originating from the left innominate artery of a right
aortic arch. D, Stented arterial duct of C.
(Table 1). Excluding this patient, the median time of documented stent patency or time to surgery which obviated further
need for ductal patency was 160 days (range 39–813 days).
For the 2 single ventricle patients, the PDA stent remained
patent until the time of the bidirectional cavopulmonary connection at 126 and 149 days of age. In the 3 patients who
1
2
3
4
5
6
25
20
15
10
5
0
25
Concentration (µg/L)
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
underwent biventricular repair, the stented PDA was patent
up to the time of complete repair. In those 3 patients who did
not require surgery, the stented PDA was documented to be
completely occluded at 359, 426, and 1198 days.
Three patients exhibited signs of infection. One patient who
had previously undergone surgery for necrotizing enterocolitis developed fever 9 days after PDA stent implantation with
an elevated white blood cell count (28.0×109/L) and a positive blood culture for coagulase negative staphylococcus and
was treated with antibiotics for a central venous line infection. One patient exhibited an elevated white blood cell count
(29.4×109/L) 1 day after stent implantation and was treated
empirically with broad spectrum antibiotics for 5 days. Blood
cultures were negative. One patient developed fever 34 days
after stent implantation and was treated for a positive urine
culture for Escherichia coli. No patient developed leukopenia
(<5×109/L), while sirolimus levels were detectable.
Two patients exhibited signs of necrotizing enterocolitis.
One patient had occult blood positive stools and a negative
ultrasound, but was treated with total parenteral nutrition and
antibiotics. One patient experienced necrotizing enterocolitis
before ductal stent implantation with no recurrence.
Liver function tests were performed in 8 patients, with 1
patient demonstrating abnormalities, which coincided with
a metabolic acidosis and increased lactate consistent with
decreased cardiac output. In this patient, aspartate aminotransferase and alanine aminotransferase remained mildly elevated
5 months after stent implantation in the setting of sirolimus
levels being nondetectable at 2 months.
Figure 2. Concentration–time curves for sirolimus in 9 neonates. The target minimal trough
level, for patients receiving oral sirolimus for
immunosuppression, is 5 μg/L (horizontal red
line). Note the sparsity of data in the rising
phase of the curve and the second peak in concentration seen in some patients (1, 3, 5, 7).
20
15
10
5
0
7
8
9
25
20
15
10
5
0
0
20
40
60
80
0
20
40
60
Time (days)
80
0
20
40
60
80
5 Lee et al Pharmacokinetics of Neonatal Sirolimus Stents
30
25
20
15
Concentration (µg/L)
10
5
0
0
20
40
60
Time (days)
cases 2, 4, 5, 7, 9 (194 µg)
15
10
0
5
Concentration (µg/L)
20
25
B
0
20
40
60
80
Time (days)
cases 3, 6, 8 (143 µg)
C
Discussion
This is the first report of drug-eluting stent implantation in
the arterial duct of human neonates, a blood vessel in which
a robust neointimal proliferative process is an essential part
of the normal developmental program that results in closure
within a few days after birth.3–7 Higher stent patency and
10
Concentration (µg/L)
5
SES were implanted into 2 older patients. One patient was 2.2
years old, weight 12.5 kg, with a stent 2.5×12 mm (124 μg
sirolimus) implanted into the left main coronary artery whilst
on extracorporeal membrane oxygenation support. During
the 7 days of extracorporeal membrane oxygenation, 8 sirolimus levels were obtained: peak 2.7 μg/L (3 hours), 1.3 μg/L
(3 days), 1.2 μg/L (9 days) and nondetectable at 21 days. In
another patient, 4 years old, weight 17.1 kg, a stent (3.5×18
mm, 194 μg sirolimus) was implanted into a left pulmonary
vein. Drug levels were 3.7 and 3.1 μg/L at 1 and 3 hours,
respectively, after implantation. The stent remains patent 2.7
years post-implantation with a mean gradient 3 mm Hg.
15
Additional Data on 2 Older Patients to Contrast
With Neonatal Pharmacokinetics
0
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
Concentration–time curves showed fast sirolimus release
from the stent, high peak concentrations, and a subsequent
slow fall (Figure 2). The rising phase of sirolimus levels was
captured in 4 subjects (3, 4, 5, 8) with at least one measurement before the peak (patient 3 had 2), but for the other 5,
the first measurement was the highest. The small number of
data points available for modeling the rising phase resulted in
wide credible intervals for the rate constant of the rising phase
(ka). A biphasic release of sirolimus in the first 12 hours was
seen in 4 subjects (1, 3, 5, 7; Figure 2), and these data were
smoothed rather than modeled (Figures 2 and 3). The duration of time that the sirolimus concentration was >5 μg/L was
found by linear interpolation (Table 2) for both the measured
concentrations and the predictions of the population model.
In 3 patients (3, 6, 8), there was a wide time interval between
samples during which the concentration fell <5 μg/L, and
hence linear interpolation of the raw data will overestimate
the time interval that the sirolimus concentration was in the
immunosuppressive range. The estimates from the one-compartment model are likely more accurate in these cases.
Noncompartmental analysis and the Bayesian one-compartment NLMM showed good agreement for all measured parameters, with the point estimates by noncompartmental analysis
falling within the NLMM 95% credible intervals (Table 2). The
majority of the measured concentration–time data lies within
the 95% credible intervals for the population one-compartment
model (Figure 3; individual fits in the Data Supplement).
Complete coverage of many arterial ducts requires >1 stent.
Using the population parameters from the NLMM, predictions were made of the pharmacokinetics for different combinations of 2 stents (Table 3), that is, higher sirolimus doses.
The highest sirolimus dose (490 μg) with 2×23 mm stents had
a peak level of 31.9 μg/L (95% credible intervals, 20.9, 49.5)
and remained within the immunosuppressive range for 22.8
days (95% credible intervals, 19.0, 28.5).
case 1 (245 µg)
A
20
Sirolimus Pharmacokinetics
0
20
40
60
Time (days)
Figure 3. Population concentration–time curves (50th centile, solid
lines) with 95% credible intervals (broken lines) for the 3 doses of sirolimus used in this study: 245 μg (A), 194 μg (B), and 143 μg (C). The
majority of empirical data points lie within the 95% credible intervals,
suggesting the one-compartment model is a reasonable approximation to the empirical data. Model adequacy is also suggested by the
similarity of the one-compartment nonlinear mixed model estimates
with those obtained by noncompartmental analysis (Table 2).
6 Lee et al Pharmacokinetics of Neonatal Sirolimus Stents
Table 2. Estimated Pharmacokinetic Parameters
Observed
Subject
Time to
Dose, μg Cmax, μg/L Cmax, h
Noncompartmental Analysis
Time >5 Cl ×10−2, Half- AUC ×102,
μg/L, days L/h
Life, h μg h/L
One-Compartment Nonlinear Mixed Model
ka, h−1
ke 10−3, h−1 Cl ×10−2, L/h
Half-Life, h
Time >5
AUC ×102, μg/L, Days
μg h/L (Population)
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
1
245
24.8
1.1
18.1
4.3
245.1
56.9
6.42
4.07
5.17
(3.6, 37.3) (3.49, 4.73) (4.56, 5.87)
170.4
47.4
14.3
(146.5, 198.6) (41.7, 53.7) (11.1, 18.9)
2
194
15
7.3
13.2
6.1
201.1
31.6
4.68
3.87
6.48
(1.33, 40.5) (2.89, 5.21) (5.15, 8.23)
179.3
29.9
11.4
(133.0, 240.0) (23.6, 37.7) (8.1, 16.0)
4
11.6
1.2
22.7
2.7
269.9
70.7
2.59
2.17
3.44
(1.7, 4.06) (1.55, 3.05) (2.5, 4.74)
319.0
56.4
(227.1, 448.5) (40.9, 77.7)
5
14.3
3
11.4
4.0
336.5
48.9
1.42
3.24
5.07
(0.91, 4.63) (2.4, 4.56) (3.86, 6.89)
213.7
38.2
(151.9, 289.1) (28.1, 50.3)
7
12
0.9
5.9
8.7
104.9
22.2
6.06
4.4
7.64
(2.57, 65.7) (3.32, 5.93) (5.91, 9.97)
157.4
25.4
(116.9, 208.7) (19.5, 32.8)
9
10.9
1
9.2
5.4
384.9
36.0
5.18
3.34
6.59
(2.29, 39.6) (2.5, 4.53) (5.05, 8.78)
207.4
29.5
(153.1, 277.1) (22.1, 38.4)
10.5
0.3
16.5
2.5
195.7
57.7
3.8
2.9
2.73
(2.78, 5.57) (2.28,3.69) (2.19, 3.41)
238.9
52.3
7.8
(188.0, 303.9) (41.9, 65.3) (3.8, 12.2)
6
11.6
2.2
17.8
2.7
212.3
52.7
4.44
3.66
5.04
(1.67, 33.8) (2.22, 5.85) (3.17, 7.84)
189.6
28.4
(118.4, 312.3) (18.2, 45.1)
8
11.5
3.1
28.4
5.4
207.7
26.3
4.58
3.67
5.42
(1.93, 36.2) (2.39, 5.7) (3.6, 8.33)
189.0
26.4
(121.6, 290.2) (17.2, 39.7)
4.12
3.42
5.08
(2.3, 13.9) (2.65, 4.4) (3.74, 6.88)
202.6
(157.4, 261.7)
3
143
Population
estimate
Data for the one-compartment nonlinear mixed model are presented as 50th (2.5th, 97.5th) quantiles. AUC indicates area under the concentration–time curve; Cmax,
peak concentration; Cl, clearance; ka, rate constant for drug elution from stent; and ke, rate constant for elimination.
decreased neointimal proliferation were found with SES in
comparison to bare-metal stents when implanted in the neonatal porcine duct.12 Our objective was to establish the pharmacokinetics of SES in human neonates because sirolimus is also
an immunosuppressive agent, before more extensive neonatal
use. In this pilot study, we only used a single SES even if multiple stents were required to cover the whole arterial duct. A
comparison of the clinical efficacy of drug-eluting with baremetal stents in the neonatal arterial duct is beyond the scope
of this study (small number of patients and both SES and baremetal stents used in the same duct), but the SES maintained
their patency to the desired clinical duration.
To our knowledge this is the largest series of neonatal patients
with sirolimus pharmacokinetics, with unique observations:
peak sirolimus levels were 20 to 40 times higher and clearance
was 30× lower in the neonates compared with adult reports.
This resulted in sirolimus levels remaining above the targeted
steady-state trough levels for transplant recipients receiving
oral maintenance therapy for a prolonged period (Table 2).21
The pronounced second peak in sirolimus concentration in the
first 12 hours may be related to biphasic sirolimus elution or
an enterohepatic recirculation. Enterohepatic recycling can
occur after systemic administration, as well as oral dosing,
when first-pass uptake by the liver is often seen.28 Sirolimus
is predominantly metabolized in the liver and intestine, with
<3% of oral sirolimus cleared by the kidneys. Liver and intestinal metabolism of sirolimus is by cytochrome P450-3A4,
3A5, and to a lesser extent, 2C8.29 In children, hydroxylation predominates over the adult pattern of O-methylation,
and piperidine-hydroxyl variants are found which are rare in
adults.30 Hence, the low clearance of sirolimus eluted from
stents in neonates compared with older children and adults
is likely related to age-dependent maturation of cytochrome
P450 enzymes. This liver enzyme immaturity has implications
for neonatal use of all current drug-eluting stents, both of the
rapamycin-family (everolimus, zotarolimus, and umirolimus)
and those based on paclitaxel because both compounds are
predominantly metabolized by cytochrome P450 and cleared
by the liver. High peak levels and slow clearance should be
anticipated with these drugs as well.
Pharmacokinetic parameters were not discernibly affected
by simultaneous use of enoxaparin, aspirin, or clopidogrel.
Cytochrome P450 enzymes are not involved in the metabolism of enoxaparin and have minimal/no involvement in aspirin metabolism.31 Clopidogrel is activated by CYP2C19, and
Table 3. Predicted Population Pharmacokinetic Parameters
for Combinations of 2 Stents
Sirolimus Dose, μg
Cmax, μg/L
Time >5 μg/L, Days
286
18.6 (12.2, 28.9)
16.2 (13.0, 21.0)
337
21.9 (14.4, 34.0)
18.2 (14.9, 23.2)
388
25.2 (16.6, 39.2)
19.9 (16.5, 25.2)
439
28.5 (18.8, 44.3)
21.4 (17.8, 26.9)
490
31.9 (20.9, 49.5)
22.8 (19.0, 28.5)
Data are 50th (2.5th, 97.5th) quantiles. Sirolimus doses: 286 μg (143 μg×2),
337 (143 μg, 194 μg), 388 μg (143 μg, 245 μg, or 194 μg×2), 439 μg (194 μg,
245 μg), and 490 μg (245 μg×2). Stent lengths and corresponding doses: 143
μg, 13 mm; 194 μg, 18 mm; 245 μg, 23 mm. Cmax indicates peak concentration.
7 Lee et al Pharmacokinetics of Neonatal Sirolimus Stents
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
some reports suggest that CYP3A4, which participates in sirolimus metabolism, may also be involved.32
Pharmacokinetic studies of SES in adult humans have
shown low peak serum drug levels occurring between 3 and
4 hours (1 stent, 0.57±0.12 μg/L; 2 stents, 1.05±0.39 μg/L)
with minimal detectable levels by 7 days and clearance of
1.46±0.45 L/hour.13 The experience of drug-eluting stents in
children is limited,14,15,17–19,33 with only 2 reports of drug level
monitoring. In a 5.3 kg, 3-month-old infant with a 143 μg
sirolimus stent implanted into an obstructed right ventricular
outflow tract conduit, 5 drug levels were measured over 35
days.15 Serum levels peaked between 7 and 8 μg/L within the
first 2 days, were below 3 μg/L after 7 days, and undetectable at 5 weeks. An 8-month-old, 4.6 kg infant underwent
both an everolimus-eluting stent (113 μg) implantation and
a paclitaxel-coated balloon dilation (coated with 640 μg of
paclitaxel) of a previously implanted bare-metal stent in 2 stenotic pulmonary veins.16 Serum everolimus levels were undetectable after 48 hours, and paclitaxel levels were low (0.59
μg/L, therapeutic range 3–6 μg/L) at 24 hours. Our experience
with the 2 older patients demonstrated similar pharmacokinetic profiles to the published reports.
Infection, necrotizing enterocolitis, and abnormal liver
function tests are not uncommon clinical problems in newborns with duct-dependent pulmonary blood flow, and it is
difficult to attribute the episodes experienced by our patients
as being secondary to systemic sirolimus. This is illustrated
in the one patient who had necrotizing entercolitis before
stent implantation. The one patient with abnormal liver function tests had persistently mildly elevated transaminases even
when sirolimus levels were not detectable. There were no
long-term clinical consequences from any of these complications. Consideration should be given to delaying live vaccine
administration, while sirolimus levels remain in the immunosuppressive range, especially if 2 SES are used (Table 3).
Limitations
The main limitations of this report are the small number of
patients and the incomplete concentration–time drug level
profiles for several patients, in particular the sparse data in the
rapid early elution phase in the first 120 minutes after stent
implantation. Hence, our estimates of rising-phase kinetics
are based on a small number of data points, and we are likely
to have missed the true sirolimus peak levels and underestimated the rate constant for elution. However, estimates of
elimination phase kinetics are robust, in particular clearance
and duration of systemic immunosuppressive sirolimus levels.
Despite these limitations, this is the most detailed pharmacokinetic description of SES in the neonatal population. Given
the developmental increase in the cytochrome P450 family
enzyme activity during infancy, our numeric values do not
extrapolate to older children.
Conclusions
Pharmacokinetics are important for any clinical drug delivery
system. In this first neonatal SES report, sirolimus levels were
higher and clearance lower by an order of magnitude as compared with adults. However, prolonged immunosuppressive
levels of sirolimus were well tolerated. Similar high systemic
drug levels should be anticipated in neonates with all the other
currently available drug-eluting stents.
Acknowledgments
We acknowledge the contribution of Warren Walsh, Resource
Technologist, Therapeutic Drug Monitoring laboratory, Hospital for
Sick Children who assisted in the analysis of the sirolimus samples.
Disclosures
None.
References
1. Alwi M, Choo KK, Latiff HA, Kandavello G, Samion H, Mulyadi MD.
Initial results and medium-term follow-up of stent implantation of patent ductus arteriosus in duct-dependent pulmonary circulation. J Am Coll
Cardiol. 2004;44:438–445. doi: 10.1016/j.jacc.2004.03.066.
2. Santoro G, Gaio G, Palladino MT, Iacono C, Carrozza M, Esposito R,
Russo MG, Caianiello G, Calabrò R. Stenting of the arterial duct in newborns with duct-dependent pulmonary circulation. Heart. 2008;94:925–
929. doi: 10.1136/hrt.2007.123000.
3. Schneider M, Zartner P, Sidiropoulos A, Konertz W, Hausdorf G. Stent
implantation of the arterial duct in newborns with duct-dependent circulation. Eur Heart J. 1998;19:1401–1409.
4.Rabinovitch M. Cell-extracellular matrix interactions in the ductus arteriosus and perinatal pulmonary circulation. Semin Perinatol.
1996;20:531–541.
5. Clyman RI, Tannenbaum J, Chen YQ, Cooper D, Yurchenco PD, Kramer
RH, Waleh NS. Ductus arteriosus smooth muscle cell migration on collagen: dependence on laminin and its receptors. J Cell Sci. 1994;107(pt
4):1007–1018.
6. Newby AC, Zaltsman AB. Molecular mechanisms in intimal hyperplasia.
J Pathol. 2000;190:300–309.
7.Costa MA, Simon DI. Molecular basis of restenosis and drugeluting stents. Circulation. 2005;111:2257–2273. doi: 10.1161/01.
CIR.0000163587.36485.A7.
8. Poon M, Badimon JJ, Fuster V. Overcoming restenosis with sirolimus:
from alphabet soup to clinical reality. Lancet. 2002;359:619–622.
9. Poon M, Marx SO, Gallo R, Badimon JJ, Taubman MB, Marks AR.
Rapamycin inhibits vascular smooth muscle cell migration. J Clin Invest.
1996;98:2277–2283. doi: 10.1172/JCI119038.
10.Suzuki T, Kopia G, Hayashi S, Bailey LR, Llanos G, Wilensky R,
Klugherz BD, Papandreou G, Narayan P, Leon MB, Yeung AC, Tio F,
Tsao PS, Falotico R, Carter AJ. Stent-based delivery of sirolimus reduces neointimal formation in a porcine coronary model. Circulation.
2001;104:1188–1193.
11. Morice MC, Serruys PW, Sousa JE, Fajadet J, Ban Hayashi E, Perin M,
Colombo A, Schuler G, Barragan P, Guagliumi G, Molnàr F, Falotico R;
RAVEL Study Group. Randomized Study with the Sirolimus-Coated Bx
Velocity Balloon-Expandable Stent in the Treatment of Patients with de
Novo Native Coronary Artery Lesions. A randomized comparison of a
sirolimus-eluting stent with a standard stent for coronary revascularization. N Engl J Med. 2002;346:1773–1780. doi: 10.1056/NEJMoa012843.
12. Lee KJ, Hinek A, Chaturvedi RR, Almeida CL, Honjo O, Koren G, Benson
LN. Rapamycin-eluting stents in the arterial duct: experimental observations in the pig model. Circulation. 2009;119:2078–2085. doi: 10.1161/
CIRCULATIONAHA.107.737734.
13.Vetrovec GW, Rizik D, Williard C, Snead D, Piotrovski V, Kopia G.
Sirolimus PK trial: a pharmacokinetic study of the sirolimus-eluting
Bx velocity stent in patients with de novo coronary lesions. Catheter
Cardiovasc Interv. 2006;67:32–37. doi: 10.1002/ccd.20565.
14. Rashid A, Ringewald J, Saucedo J. Placement of a sirolimus-coated stent
in the distal left internal mammary artery of an 8-year-old boy. J Invasive
Cardiol. 2005;17:329–330.
15. Riede FT, Schneider P, Dähnert I. Transient sirolimus serum levels after
implantation of a sirolimus eluting stent in an infant. Clin Res Cardiol.
2007;96:508–510. doi: 10.1007/s00392-007-0516-x.
16. Müller MJ, Krause U, Paul T, Schneider HE. Serum levels after everolimus-stent implantation and paclitaxel-balloon angioplasty in an infant
with recurrent pulmonary vein obstruction after repaired total anomalous
pulmonary venous connection. Pediatr Cardiol. 2011;32:1036–1039. doi:
10.1007/s00246-011-0054-1.
8 Lee et al Pharmacokinetics of Neonatal Sirolimus Stents
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
17. Salloum JG, Dodd DA, Slosky D, Zhao DX. Treatment of unprotected left
main coronary artery stenosis in a 5-year-old heart transplant patient using
a sirolimus-eluting stent. J Heart Lung Transplant. 2007;26:1061–1064.
doi: 10.1016/j.healun.2007.07.020.
18. Kaichi S, Doi H, Tsurumi F, Heike T. Long-term outcome of sirolimuseluting stent implantation for left main coronary artery stenosis in infancy.
Pediatr Cardiol. 2011;32:94–97. doi: 10.1007/s00246-010-9815-5.
19.Sachdeva R, Seib PM, Frazier EA, Sachdeva R. Percutaneous coro
nary intervention using drug-eluting stents in pediatric heart transplant recipients. Pediatr Transplant. 2009;13:1014–1019. doi:
10.1111/j.1399-3046.2008.01118.x.
20. Lobach NE, Pollock-Barziv SM, West LJ, Dipchand AI. Sirolimus immunosuppression in pediatric heart transplant recipients: a single-center
experience. J Heart Lung Transplant. 2005;24:184–189. doi: 10.1016/j.
healun.2004.11.005.
21. Scott JR, Courter JD, Saldaña SN, Widemann BC, Fisher M, Weiss B,
Perentesis J, Vinks AA. Population pharmacokinetics of sirolimus in
pediatric patients with neurofibromatosis type 1. Ther Drug Monit.
2013;35:332–337. doi: 10.1097/FTD.0b013e318286dd3f.
22. R Core Development Team. R: a language and environment for statistical computing, R Foundation for Statistical Computing, Vienna, Austria,
2014. http://www.R-project.org. Accessed October 26, 2014.
23. Jaki T, Wolfsegger MJ. Non-compartmental estimation of pharmacokinetic parameters for flexible sampling designs. Stat Med. 2012;31:1059–
1073. doi: 10.1002/sim.4386.
24. Wakefield J. General regression methods. In: Bayesian and Frequentist
Regression Methods. New York: Springer; 2013:425–501 (chapter 9).
http://faculty.washington.edu/jonno/book/TheophWinBUGS.txt. Accessed
October 26, 2014.
25.Plummer M, 2013. JAGS version 3.4.0 User Manual. http://source
forge.net/projects/mcmc-jags/files/Manuals/3.x/jags_user_manual.pdf.
Accessed October 26, 2014.
26. Lunn D, Jackson C, Best N, Thomas A, Spiegelhalter D. The BUGS Book:
A Practical Introduction to Bayesian Analysis. Boca Raton, FL: Chapman
& Hall; 2012:Chapter 10.3.1.
27. Plummer M, Best N, Cowles K, Vines K. CODA: convergence diagnosis
and output analysis for MCMC. R News. 2006;6:7–11.
28.
Roberts MS, Magnusson BM, Burczynski FJ, Weiss M.
Enterohepatic circulation: physiological, pharmacokinetic and
clinical implications. Clin Pharmacokinet. 2002;41:751–790. doi:
10.2165/00003088-200241100-00005.
29.Emoto C, Fukuda T, Mizuno T, Cox S, Schniedewind B, Christians
U, Widemann BC, Fisher MJ, Weiss B, Perentesis J, Vinks AA. Agedependent changes in sirolimus metabolite formation in neurofibromatosis type 1 [published online ahead of print August 26, 2014]. Ther Drug
Monit. doi: 10.1097/FTD.0000000000000130.
30. Filler G, Bendrick-Peart J, Christians U. Pharmacokinetics of mycophenolate mofetil and sirolimus in children. Ther Drug Monit. 2008;30:138–
142. doi: 10.1097/FTD.0b013e31816ba73a.
31. Miners JO, Day RO. Relationships between the adverse effects of drugs
and genetic polymorphism in genes encoding drug metabolizing enzymes. Br J Clin Pharmacol. 2007;63:380–381, author reply 381. doi:
10.1111/j.1365-2125.2006.02737.x.
32. Sangkuhl K, Klein TE, Altman RB. Clopidogrel pathway. Pharmacogenet
Genomics. 2010;20:463–465. doi: 10.1097/FPC.0b013e3283385420.
33. Wong DTH, Yoo S-J, Lee K-J. Drug-eluting stent implantation of obstructed infra-cardiac total anomalous pulmonary venous drainage in right atrial
isomerism. Cardiol Young. 2008;18:628–630.
Pharmacokinetics of Sirolimus-Eluting Stents Implanted in the Neonatal Arterial Duct
Kyong-Jin Lee, Winnie Seto, Lee Benson and Rajiv R. Chaturvedi
Downloaded from http://circinterventions.ahajournals.org/ by guest on November 2, 2016
Circ Cardiovasc Interv. 2015;8:
doi: 10.1161/CIRCINTERVENTIONS.114.002233
Circulation: Cardiovascular Interventions is published by the American Heart Association, 7272 Greenville
Avenue, Dallas, TX 75231
Copyright © 2015 American Heart Association, Inc. All rights reserved.
Print ISSN: 1941-7640. Online ISSN: 1941-7632
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circinterventions.ahajournals.org/content/8/5/e002233
Data Supplement (unedited) at:
http://circinterventions.ahajournals.org/content/suppl/2015/05/05/CIRCINTERVENTIONS.114.002233.DC1.html
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation: Cardiovascular Interventions can be obtained via RightsLink, a service of the Copyright
Clearance Center, not the Editorial Office. Once the online version of the published article for which
permission is being requested is located, click Request Permissions in the middle column of the Web page
under Services. Further information about this process is available in the Permissions and Rights Question and
Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation: Cardiovascular Interventions is online at:
http://circinterventions.ahajournals.org//subscriptions/
Supplemental Material
Pharmacokinetics of sirolimus-eluting stents implanted in the neonatal arterial
duct.
Supplemental Methods
The single i.v dose 1-compartment model, parameterized by clearance (equation 9.30 from page
478 of [1]) is:
Y = D* ( ka * ke / Cl * [ka – ke] ) * ( exp[ -ke * t] – exp[ -ka * t] )
Y, sirolimus concentration (µg/L)
D, sirolimus dose (µg)
Cl, clearance (L/hr). Cl = V * ke , where V is the volume of distribution (L).
ka, rate constant for rising phase i.e elution from the stent (hr-1)
ke, rate constant for elimination phase (hr-1)
t, time (hr)
Minor changes were made to Wakefield’s WinBUGS script [2] for JAGS 3.4.0 with nested indexing [3] to
specify which child each observation belonged to. This is the model file:
var Y[Nobs], time[Nobs], child[Nobs], mu[Nobs],fitted[Nobs], resid[Nobs],
theta[N,3]
model
{
# loop across all Nobs observations(Y[i] at time[i], with each Y[i] belonging
# to a child[i]).
for(i in 1:Nobs){
Y[i] ~ dnorm(mu[i],eps.tau)
mu[i] <- dose.ug[child[i]]*
exp(theta[child[i],1] + theta[child[i],2] - theta[child[i],3]) *
exp(-exp(theta[child[i],1])* time[i]) –
exp(-exp(theta[child[i],2])*time[i]) ) /
(exp(theta[child[i],2])-exp(theta[child[i],1]))
fitted[i] <-mu[i]
resid[i] <-Y[i] - fitted[i]
}
# loop across N children for child specific values of parameters
# ke, ka, Cl
for(j in 1:N){
theta[j, 1:3] ~ dmnorm(beta[1:3], Dinv[1:3, 1:3])
ke[j] <- exp(theta[j,1])
ka[j] <- exp(theta[j,2])
Cl[j] <- exp(theta[j,3])
}
logtau ~ dunif(0.001,1000)
eps.tau <- exp(logtau)
sigma <- 1 / sqrt(eps.tau)
Dinv[1:3, 1:3] ~ dwish(R[1:3, 1:3], 3)
beta[1:3] ~ dmnorm(mean[1:3], prec[1:3, 1:3])
#population values for ke, ka and Cl are kemed, kamed, Clmed respectively
kemed <- exp(beta[1])
kamed <- exp(beta[2])
Clmed <- exp(beta[3])
}
Individual fits
25
child 1
●
15
●
●
10
●
●
●
5
●
●
●
●
●
●
●
0
Concentration (µg/L)
20
●
0
20
40
Time (days)
60
80
15
10
●
●●
●
5
●
●
●
●
●
0
Concentration (µg/L)
20
25
child 2
0
20
40
Time (days)
60
80
20
25
child 3
●
15
●
●
●
●● ●
5
10
●
●
●
●
0
Concentration (µg/L)
●
0
20
40
Time (days)
60
80
15
10
●
● ●
●
●
5
●
●
●
0
Concentration (µg/L)
20
25
child 4
0
20
40
Time (days)
60
80
●
10
15
●
●
●
●
●
5
●
●
●
●
●
●
0
Concentration (µg/L)
20
25
child 5
0
20
40
Time (days)
60
80
15
10
●
●
●
5
●
●
●
0
Concentration (µg/L)
20
25
child 6
0
20
40
Time (days)
60
●
80
15
●
●
●
●
●
5
10
●
●
●
●
●
0
Concentration (µg/L)
20
25
child 7
0
20
●
40
Time (days)
60
80
15
10
●
●
5
●
●
0
Concentration (µg/L)
20
25
child 8
0
20
40
Time (days)
60
80
15
10
●
●
●
●
5
●
●
●
●
●
0
Concentration (µg/L)
20
25
child 9
0
20
40
Time (days)
60
80
Supplemental References
1) Wakefield J. Bayesian and frequentist regression methods, Springer, NY, 2013.
2) http://faculty.washington.edu/jonno/book/THeophWinBUGS.txt
3) Lunn D, Jackson C, Best N, Thomas A, Spiegelhalter D. The BUGS book: a practical introduction to
Bayesian analysis, Chapman & Hall, Boca Raton, Fl, 2012.