Catheterization and Cardiovascular Interventions 73:350–360 (2009)

Basic Science

A Comparison of the Mechanical Performance

Characteristics of Seven Drug-Eluting Stent Systems
W. Schmidt,1* DR-ING, P. Lanzer,2 MD, P. Behrens,1 DIPL-ING, L.D.T. Topoleski,3 PhD,
and K.-P. Schmitz,1 DR.-ING
Objectives: Mechanical properties of drug eluting stents (DES) will be measured to pro-
vide comparable numerical data to assess deliverability, and thus clinical performance.
Background: DES are routinely used in coronary interventions to reduce the rates of
restenosis and target vessel revascularizations. Current research is primarily con-
cerned with issues related to late stent thrombosis. However, mechanical properties of
DES are a critical determinant of deliverability, and consequently the ultimate arbiter of
their clinical performance. Methods: Mechanical properties (pushability, trackability,
crossability) were measured under standardized in-vitro conditions. The vessel models
were derived from typical vessel anatomy but adapted to the individual tests. Addition-
ally, profile and bending forces of the stent segment of the delivery system were meas-
ured. Seven different commercially available balloon-expandable coronary DES sys-
tems were included. All stents were 3.0 mm diameter with a stent length from 14 to 18
mm. Results: The pushability expressed as the ratio of distal force at a specific proxi-
mal push force (4N) ranged between 38.66 and 18.53%. The trackability as the mean
track-forces ranged from 0.551 N to 1.137 N. One stent system could not pass this
test. The mean crossing forces at a 1.4 mm stenosis model ranged from 0.038 N up to
0.103 N. The mean crimped stent profiles ranged from 1.055 mm to 1.198 mm and the
bending stiffness of the crimped stent was 17.22 to 47.20 Nmm2. Conclusion: Better
understanding of mechanical properties of DES shall improve tactile skills of the inter-
ventionists during PCI and to improve criteria for DES selection in specific clinical set-
tings. ' 2008 Wiley-Liss, Inc.

Key words: invasive cardiology; drug-eluting stents; mechanical performance; clinical

impact

INTRODUCTION 1
Institute for Biomedical Engineering, University of Rostock,
Drug-eluting stents (DES) are routinely used in cor- Rostock, Germany
2
onary interventions to reduce the rates of restenosis Department for Cardiology and Angiology, Hospitals and Clin-
ics Bitterfeld/Wolfen, Bitterfeld, Germany
and target vessel revascularizations [1–3]. Current 3
Department of Mechanical Engineering, University of Mary-
research is concerned primarily with safety issues land, Baltimore County, Baltimore, Maryland
related to late stent thrombosis [4–6]. However, the
mechanical properties of DES are a critical determi- Conflict of interest: Nothing to report.
nant of deliverability and therefore the ultimate arbiter
*Correspondence to: Wolfram Schmidt, DR-ING, Institute for Bio-
of their clinical utility. It is essential for the clinician medical Engineering, University of Rostock, Ernst-Heydemann-
to know and understand the differences in model-spe- Straße 6, 18055 Rostock, Germany. E-mail: wolfram.schmidt@uni-
cific DES behavior for optimum clinical results. Never- rostock.de
theless, surprisingly little data is available on this im-
portant topic. Received 8 April 2008; Revision accepted 5 September 2008
The mechanical response of coronary stent systems DOI 10.1002/ccd.21832
and other endovascular devices has been described Published online 9 December 2008 in Wiley InterScience (www.
using terms such as pushability, trackability, crossabil- interscience.wiley.com).

' 2008 Wiley-Liss, Inc.

 Mechanical Performance of Drug-Eluting Stent Systems 351

Fig. 1. 3D structure of the crimped stents. 1: Biosensors Biomatrix; 2: Conor CoStar; 3: Cor-
dis Cypher select 1; 4: Medtronic Endeavor; 5: Boston Scientific Taxus Liberté; 6: Guidant
Xience V; 7: B. Braun Coroflex Please.

ity, flexibility, conformability, and torque control [7– All measurements were performed in a 378C heated
16]; yet, their definitions have remained vague or water basin at a travel speed of 7.5 mm/s delivered by
inconsistent. a guide (maximum travel distance 500 mm, Cleveland
To characterize the mechanical responses of DES, Präzisionssysteme GmbH, Germany), which is driven
we have defined pushability, trackablity, and crossabil- by a DC servo motor (type tendo PM 41, Chr. Mayr
ity by objective measurements of the externally applied GmbH, Germany). The details of tubing material and
force and the resulting force measured at the catheter dimensions are given in the following sections and fig-
tip and the relations between them. We conducted ures. The guiding catheter and the vessel model were
these measurements under well-established, standar- flushed prior to each testing series. The vessel models
dized, and validated in vitro experimental conditions. were derived from typical vessel anatomy but adapted
In addition, diameter and bending forces required to to the individual tests.
deform the stent segment of the delivery system were All measurements were performed on a single stent
measured. The mechanical performances of seven com- system of each DES brand. Single measurements are
mercially available DES systems were studied. adequate because quality standards required for com-
mercially distributed medical devices do not allow any
relevant variation in product properties. These stand-
ards guarantee reliable characterization of a brand by
MATERIALS AND METHODS
measuring one specimen.

Measurements were performed on seven different Measurement of Pushability

commercially available balloon-expandable coronary In this study, pushability was defined as the ability
DES systems (see Fig. 1). To allow comparisons, all of the stent delivery system (SDS) to transfer the prox-
stents were 3.0-mm design diameter with the stent imal external force exerted by the investigator (prox-
length ranging from 14 to 18 mm depending on the imal force, Fprox) to the force measured at the tip of
manufacturer’s specifications. For each DES brand, the system (distal force, Fdist). The pushability is
one randomly selected specimen was studied. essentially the amount of force lost in the system.
To perform the measurements, each DES was Improving pushability was considered with increasing
removed from the original wrapping and introduced ratio of the two forces.
into a guiding catheter (Cordis Vista brite tip 6F To measure pushability, a model of coronary circula-
JL3.5, I.D. 0.067@) over a 0.014@ guide wire (Guidant tion was modified by simulating a total occlusion at the
HI-Torque Balance MiddleWeight 0.014@) to complete end of a selected ‘‘vascular’’ pathway. Figure 2 shows
the interventional system. Neither the guiding catheter the experimental setup developed for the pushability
nor the guide wire was changed during testing to tests. The load cell (range 0–5 N, type 8432, burster
ensure a constant test environment. Prior validation GmbH, Germany) at the distal end of the vascular path-
testing showed that there is good reproducibility [10], way includes a model of a totally occluded lesion. The
in particular no measurable change in physical proper- load cell measures the axial component of the distal
ties due to abrasion or contamination processes. force, Fdist, as the stent system advances toward the total
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
352 Schmidt et al.

Fig. 3. Experimental model of the left coronary artery circula-

Fig. 2. Vessel model and total occlusion for measurement of tion. Test path 5 was used for the presented tests.
the pushability of the DES systems.

10 mm), with a milled path for the guiding catheter

occlusion. At the beginning of each measurement, the which leads to the model of the coronary arteries. The
tip of the stent system was placed immediately in front two PMMA plates are pinned and bolted together. The
of the simulated occlusion (‘‘start point of test’’). The diameter of the path for the guiding catheter is 2 mm,
proximal shaft was attached to the automatic linear drive and the milled path diameter for the coronary arteries
via a second load cell (measurement range 0–5 N, type is 2.5 mm (Fig. 3).
8432, burster GmbH, Germany) to measure the proximal Validation measurements have shown that number
push force Fprox during DES delivery system advance- and radii of the curvatures are responsible for the diffi-
ment. The stent systems were advanced at a travel speed culty level of the relating model path [13]. Out of the
of 7.5 mm/s until a load-controlled test stop at a proxi- total of eight vessel paths, all trackability tests were
mal push force of >4 N was reached. Both forces, Fdist performed using the paths marked in the drawing. It
and Fprox, were measured simultaneously. was chosen as a medium difficulty path intended to
The pushability parameter was calculated from the allow complete data collection for all DES systems
maximum forces of Fprox (Fprox
4 N) and the corre- and to differentiate characteristic handling properties
sponding distal force Fdist as follows: of the stent systems. A commercially available guide
wire provided the required access to the chosen vessel.
Pushability ¼ ðFdist =Fprox Þ 3 100% ð1Þ The investigated stent systems are advanced over the
test path from point A, which is still inside of the
Only a single measurement of pushability with lim- guiding catheter, until the tip of the stent system
ited push force was performed per stent system to has reached point E. The total travel distance was
exclude damage of the distal tip of the SDS during the 220 mm. All measurements are performed in a 378C
push against the total occlusion. heated water environment.
The measurements of trackability were performed
Measurement of Trackability three times in each DES delivery system to reduce ran-
In this study, trackability was considered the ability dom errors. To obtain a measure of the SDS’s track-
of SDS to track or move easily through a curved vas- ability, the proximal track forces Fprox of each single
cular pathway. To characterize the trackability in this measurement were averaged from all three measure-
experimental setup, the proximal push force required ments, and a mean track force was calculated from the
to advance the SDS through the tubing system was curves as follows:
measured. The trackability is essentially a measure of
1X n
the resistance against the SDS moving through a Trackability ¼ Fprox ðiÞ ð2Þ
curved vessel pathway. The lower the push forces, the n i¼1
better are the trackability.
All measurements of forces were performed using a where n is the number of force measurements along
standardized experimental model of the left coronary the entire travel path (n 5 440). The test procedure as
circulation developed in our laboratory [10,13,17]. The well as the data analysis is in accordance with the in-
model consists of two PMMA plates (thickness: ternal standard of the test laboratory [17].
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
 Mechanical Performance of Drug-Eluting Stent Systems 353

Fig. 4. Vascular and stenosis model for crossability tests.

Fig. 5. Laser scanner employed for DES-diameter measure-

Measurement of Crossability ments.
In this study, crossability was considered as the abil-
ity of SDS to pass through the target lesion. To char- urements: the stent profile and the bending stiffness.
acterize the crossability in this experimental setup, the These parameters are likely to influence the complex
distal reactive force developed during the crossing ma- functional parameters.
neuver was measured. The lower the reactive force
developed during crossing of the lesion, the better is
the crossability. Measurement of Stent Profile
To measure the crossability, the experimental setup The mechanical SDS behavior is potentially influ-
was modified by integrating a stenotic lesion model at enced by the stent profile. To fully characterize the
the end of the test path (Fig. 4). This lesion model is contour of the entire stent-carrying SDS segment, the
characterized by an eccentric narrowing of the vessel mean stent profile over the entire length of the stent
lumen from 2.5 to 1.4 mm with a conical transition. segment was measured.
Thus a reduction by 44% diameter (68% cross-sec- The stent profile, defined as the outside diameter of
tional lumen area) was achieved. The lesion model the stent system with the stent crimped on the balloon
allows narrowing of the lumen at either the inner ra- catheter, was determined using a noncontact special
dius (lesion model type 1) or at the outer radius (lesion laser measuring head ODAC 32XY (Zumbach Elec-
model type 2) of the test curvature. For this study, the tronic GmbH) designed to make measurements in a
more challenging lesion model type 2 with obstruction water bath. The operation principle of the measurement
at the outer radius was used. system is shown in Fig. 5. A temperature-controlled
A load cell was mounted to the lesion model to pro- water filled test chamber is situated in the center of
vide measurements of the distal reactive force Fdist in the two-axis laser measurement head. The linear drive
addition to the proximal push force Fprox. The proxi- motion section consists of a DC motor with encoder
mal travel distance (driven by the linear actuator) was and gears as well as provisions for clamping the cathe-
60 mm beginning at the end of the vessel simulation ter [13,18].
(point A) until the entire stent/balloon has passed the For the measurement, the balloon catheter with a
stenosis model (point B). crimped on stent is slipped over a guide wire. The
Three measurements of SDS crossability were per- guide wire directs the motion within the measurement
formed for each of the SDS studied. The distal reac- chamber such that the laser beam strikes the distal end
tive forces were averaged over the three measurements. of the balloon. Profiles were measured in two perpen-
The mean distal force Fdist was calculated as the mea- dicular projections along the stent axis z at every
sure of crossability. 0.2 mm [dx(z), dy(z)], thus allowing measurements of
Because trackability, crossability, and pushability are both stent diameters along the entire stent length. All
complex, mutually interacting properties of DES, not diameter values are the mean values calculated over
fully defined by measurements of the proximal, distal, the effective stent length and are stored in a data table.
and reactive forces, two additional properties of DES The complete stent profile is obtained from the indi-
were determined using straightforward physical meas- vidual measurements performed at the orthogonal x
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
354 Schmidt et al.

TABLE I. Summary of All Numerical Test Results

Mean Mean Crimped Bending
track cross stent stiffness of
Pushability force force profile crimped stent
Stent system (%) (N) (N) (mm) (Nmm2)
Biomatrix 35.89 0.55 0.09 1.130 30.06
Costar 38.66 0.76 0.04 1.088 22.44
Cypher Select 1 27.34 1.14 0.08 1.198 25.90
Endeavor 32.20 0.69 n.a.a 1.130 47.20
Taxus Liberté 20.91 n.a.b 0.09 1.124 17.24
Xience V 37.60 0.87 0.04 1.055 25.78
Coroflex Please 18.53 0.67 0.10 1.145 20.16
a
Medtronic Endeavor was excluded because of stent loss.
b
Taxus Liberté was excluded because of force overload after 146-mm
travel distance.

Fig. 6. Test setup for the measurement of the bending stiff-

ness of DES employed in the study. [Color figure can be
cell (Q11/10 g) has a range of 6100 mN at an accu-
viewed in the online issue, which is available at www. racy of 60.5% FS.
interscience.wiley.com.] The displacement and force data are recorded using
a computer, which is connected to the testing device
using a serial interface. The force–distance curves
and y axes. This is accomplished by generating the describe the spring modulus of the test objects for
root mean square value (RMS) of dx(z) and dy(z): bending. The bending stiffness EI is calculated taking
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the mean value of F/f calculated by linear regre-
dx2 ðzÞ þ dy2 ðzÞ ssion from the whole force–distance curve using the
dRMS ðzÞ ¼ ð3Þ equation
2

Here z denotes the measurement position at the F  l3

EI ¼ ð5Þ
length axis. All the values of dRMS(z) of the scan 3f
along the z axis are recorded as the stent profile.
For comparison purposes, the mean value of the based on the elastic beam theory, where E is the mod-
stent profile is calculated taking all profile values in ulus of elasticity (or Young’s modulus, a property of
the region of the crimped stent. the material), and I is the moment of inertia (based on
the geometry of the beam). Taking into account possi-
1 X ble asymmetric structures of the test samples, bending
dðpÞ ¼ dðzi Þ ð4Þ
n i¼1...n stiffness is measured in five directions around the cir-
cumference and subsequently averaged. Figure 6 shows
where n specifies the number of diameter values in the the experimental setup to measure bending stiffness.
stent region. The inflation pressure, p, is 0 bar for the Validation of the experimental setups and all meas-
initial state of the crimped stent. ured variables are provided in the literature
[10,11,13,18,19].
Measurement of Bending Stiffness
Bending stiffness is a measure of the structure’s re- RESULTS
sistance to bending deformation and is the reciprocal
All numerical test results are summarized in Table I.
of flexibility. Bending stiffness is a function of the ra-
tio of the bending force and the deflection of the SDS
due to this force. Thus, the larger the ratio, the higher Pushability
is the bending stiffness. The relationships between the measured proximal
The experimental setup to determine the bending and distal forces are shown in Fig. 7. The distal force
stiffness of a crimped stent with the free bending increases steadily with increasing proximal force after
length l creates a deflection f of the stent, which is the initial system friction (i.e., Fprox is not zero when
applied automatically using a linear actuator. The Fdist begins to increase). The pushability at the maxi-
resulting force F is measured by a load cell. The load mum Fprox ranged from 18.53% to 38.66%. The high-
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
 Mechanical Performance of Drug-Eluting Stent Systems 355

Fig. 7. Pushability. Fprox vs. Fdist developed while crushing the stent systems in the vessel
model at the total occlusion. [Color figure can be viewed in the online issue, which is avail-
able at www.interscience.wiley.com.]

est pushability was measured for the Conor Costar Bending Stiffness
DES and the lowest for the Braun Coroflex Please. The bending stiffnesses of the stent carrying regions
of the DES are also shown in Table I. The highest
Trackability bending stiffness of the crimped stent was measured
for the Medtronic Endeavor (47.20 Nmm2). The lowest
The resulting force–distance graphs derived from the
bending stiffness was measured for the Taxus Liberté
test path 5 are given in Fig. 8. The initially low-track
(17.22 Nmm2).
forces begin to rise as the distance of DES advance-
ment (s) increases, with marked differences between
the individual DES past the 130-mm distance. The DISCUSSION
Taxus Liberté system could not be pushed through the The mechanical behaviors of a stent may be impor-
entire track model since the force exceeded the maxi- tant for four broad categories of the stent’s life: (1)
mum of 4 N after 146 mm (66.4% of the total tracking manufacturing and packaging, (2) delivery, (3) deploy-
path’s length) of tracking. The mean track forces ment, and (4) in vivo service. The mechanical proper-
ranged from 0.551 N of Biosensors Biomatrix to 1.137 ties during both deployment and in vivo service have
N of Cordis Cypher select 1. been studied to some extent. A broad survey is given
in Ref. 20. However, the mechanical behavior during
Crossability delivery has received little attention.
Figure 9 shows the measured distal reactive forces Mechanical performance of SDS can be described
Fdist of the stent systems across the total distance of by the stent dimensions, physical properties, and com-
60 mm (A to B) required to pass through the stenotic binations of forces required for safe and easy stent
lesion (model 2). The mean crossing forces for all delivery to target destinations. Typical dimensions
seven DES ranged from 0.038 N (Abbott Xience V) employed to describe SDS include their entry and the
up to 0.103 N (Braun Coroflex Please). crossing profiles. Typical systems of forces that define
their mechanical behavior have been combined and
described by their pushing, tracking, and crossing abil-
Crimped Profile ities. Ultimately, atraumatic delivery and precise posi-
The mean stent profiles of the investigated DES are tioning of an intact stent across the target lesion are
summarized in Table I. They ranged from 1.055 mm the most basic prerequisites for a successful interven-
(Abbott Xience V) to 1.198 mm (Cordis Cypher select tion. Other than commercial product descriptions avail-
1). able to the interventional cardiologists, virtually no
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
356 Schmidt et al.

Fig. 8. Trackability. Force–distance curves from the trackability measurements (test path 5)
of the investigated stent systems are shown. [Color figure can be viewed in the online issue,
which is available at www.interscience.wiley.com.]

Fig. 9. Crossability of the DES. Fdist as a function of the travel distances (measurement with
lesion type 2). Medtronic Endeavor was excluded because of stent loss. [Color figure can be
viewed in the online issue, which is available at www.interscience.wiley.com.]

technical data on individual products have been pub- In this study, we undertook a pioneering effort to
lished, leaving the vast majority of critical mechanical define and measure objective mechanical parameters in
performance descriptors to the proprietary sector and vitro to compare mechanical performance of seven
cardiologist’s best guess. commercially available SDS products. There is no
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
 Mechanical Performance of Drug-Eluting Stent Systems 357

doubt that clinical scenarios are more complex than an profiles along the stent carrying segment of the balloon
in vitro simulation; however, to understand the physi- catheter was selected to characterize relevant SDS
cal properties of stent systems, and to provide the cli- dimensions.
nician with the best possible guidance, we believe that Our measurements showed that differences in push-
in vitro simulations are invaluable. Good performance ability are on the order of 50% and in trackability on
in vitro will not guarantee safe and easy target access the order of 100%. Differences in crossability and in
in vivo; however, poor in vitro performance very likely bending stiffness are in the range of 60%, whereas the
indicates a risk of difficulty or complication, at least in variation in average profile of the seven SDS studied
case of complex lesions. is in the range of 12%. The best pushability was meas-
Pushability, defined as the ratio of the distally and ured for Conor CoStar-SDS (38.66%) and the lowest
proximally exerted forces, trackability, defined as the for Coroflex Please-SDS (18.53%).
sum of proximal forces required to pass a curved vas- In all systems, the proximal pushing forces were
cular pathway, and crossability, defined as distal reac- low, meaning good trackability at the proximal seg-
tive force required to pass a defined narrowing, were ment of the pathway. Proximal push forces increased
selected to represent deliverability of SDS. steadily with progression through the model. In this
Although the terms of pushability, trackability, and study, the best trackability was measured for the Bio-
crossability are commonly used in the field of cardio- Matrix SDS (0.551 N) and the lowest for the Cypher
vascular interventions, there is no unique definition. select1-SDS (1.137 N), i.e., a 100% increase in force
Product standards for vascular stents [21,22] use the required. For our experimental setup, Taxus Liberte
trackability and pushability to describe the ability of a had to be excluded from comparison because at 146-
stent system to be advanced to the lesion without SDS mm travel distance the upper force limit (4 N) had
damage or harm to the vasculature. This has to be been exceeded.
shown by the manufacturer to the authorities for product Pushability represents the efficiency by which the
approval. For this demonstration, suitable in vitro mod- applied force is transmitted to the catheter tip, and
els may be used but no measurements are required. hence the accuracy of the feedback, or ‘‘feel’’ to the
Comparative design evaluation of coronary stents interventionalist. A pushability of 0% can be repre-
was provided using among others the term of ‘‘deliver- sented by a completely flexible rope, where pushing
ability,’’ which includes tracking and lesion crossing on the proximal end will only result in the rope’s
[23]. Evaluation was given without further definition crimping, with no force transmitted to the distal end.
or provision of measured data. To our knowledge, the A pushability of 0% could also be achieved by the
only other approach leading to objectively measured case where the shear resistance of the vessel wall
forces was using 908 and 1358 bent tortuosities and along the length of the SDS is greater than or equal to
measuring the required proximal forces needed to pass the proximal force, thus preventing any transmission to
the curvatures with different stent systems [9]. This the distal end. Conversely, a pushability of 100% can
method is suitable for the assessment of trackability if be represented by a perfectly rigid rod, where all of
the pure technical design of vessel geometry can be the proximal force is transmitted to the distal end. In
accepted. We improved acceptance by introducing a the experimental system, this could be achieved only if
technically well-defined coronary vessel model with there were no losses along the SDS length. Thus, push-
more realistic dimensions. ability represents the loss of force from interactions
Each of the three delivery parameters (pushability, between the vessel walls and the SDS. For a given
trackability, and crossability) are based on the resisting vessel wall architecture, differences in pushability must
forces that can be quantified along some part of the arise from differences in the SDSs; such differences
delivery path. The definitions also include a measure most likely originate from differences in SDS surfaces,
of the force felt or applied by the clinician, Fprox. The such as surface roughness, composition, or actual con-
parameters will therefore potentially give the clinician tact area among others, giving rise to frictional resist-
an indication of how a specific stent is progressing as ance. The Conor CoStar (Fig. 1) shows a smooth metal
they push on the proximal end. As the arterial models surface with many undulations and small structural ele-
are improved and become more sophisticated, the defi- ments, which may allow for better lubrication, and
nitions may also evolve to reflect new understanding hence easier passage. The Braun Coroflex Please, in
of stent delivery. contrast, appears to have a rougher polymer coating
In addition, the bending stiffness, a function of the and larger structural elements, both of which may lead
ratio of the bending force and resulting deflection at to greater friction between the SDS and the vessel
the point of load application for the mounted stents, wall. The bending stiffness of the stent system will
was also determined. The average of the sum of all also contribute to the pushability. A stiffer stent system
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
358 Schmidt et al.

may transmit force more efficiently because it will act the crimped stent. The stent system with good track-
more like the rigid rod example; however, a stiffer ability (Braun Coroflex Please) fared the worst in
system will also not conform to a curved segment of a crossability. Another stent system with poorer track-
vessel, and hence potentially result in greater contact ability (Abbott Xience) fared the best in crossability.
forces, which will reduce the force transmission and This suggests that trackability is not the governing pa-
hence pushability. rameter for crossability.
Trackability depends on the facility of the SDS to In contrast, the Xience V and the Conor CoStar,
negotiate curved portions of the model test track. In which have the smallest profile, also have the best
each of the SDS tested (Fig. 1), the devices are seg- crossability in this test, even though both stents have
mented, that is, they are designed with individual sec- small and tortuous structural elements. These results
tions connected by structural ‘‘struts.’’ The ability of suggest that it is definitely important and an advantage
the SDS to bend and twist through tortuous vessels to have a small, smooth profile of the balloon with the
will most likely depend on the ability of the ‘‘strutted’’ crimped stent. The crossability will also depend, in
connections, rather than the more rigid segments, to part, on the stiffness of the lesion. The stiffness, or
bend. The ability of the struts to bend will depend on mechanical responses, of a lesion (that is, how easy is
both their geometry and material. It would seem that it to push the lesion out of the way as the stent pro-
trackability and bending stiffness, where a lower bend- gresses through) is lesion dependent [24,25]. This
ing stiffness means that the SDS is more deformable, effect was not simulated in our model.
should be related; however, two of the three SDSs The measurements of pushability, trackability, and
with the best trackability (e.g., #1-Biosensors BioMa- crossability as defined in this study establish the first
trix and #3-Medtronic Endeavor) had the greatest systematic, quantitative framework for rigorous com-
bending stiffness in the stent region. The second best parisons of mechanical properties of DES systems.
trackability (Braun Coroflex) also had the second low- Safe and atraumatic passage of DES to the target sites
est bending stiffness. The SDS with the lowest bending and their precise positioning across the target lesions
stiffness (Taxus Liberte) had such a high trackability depend on the facility of delivery in the endovascular
force and poor trackability performance that it could environment. For example, in the majority of anatomi-
not complete the experimental task. This suggests three cally straightforward cases, specification of standard or
explanations. First, there is an optimum system bend- average mechanical properties of the SDS shall suffice.
ing stiffness that results in good trackability. The data With increasing hostility of the upstream vascular seg-
suggest that a stiffer device is required to overcome ment and tightness and complexity of the target lesion,
the resisting forces generated in a curved vessel. Sec- the mechanical properties of SDS become increasingly
ond, the bending stiffness is a combination of the stent critical. Ultimately, an ideal combination of individual
material, through the material property ‘‘E’’, the mechanical properties shall characterize a perfect SDS
Young’s modulus, or elastic stiffness—and the stent product; however, in the real world, SDS products typ-
geometry—through the moment of inertia term, ‘‘I’’ ically feature different strengths and weaknesses in
[from Eq. (5)]. The equation used to determine the their clinical mechanical performance. For the inter-
bending stiffness [Eq. (5)] assumes that the beam is a ventionist, it appears important to select the most
continuous solid, and it is evident that the stents are appropriate instrument for a given vascular territory.
not. The third possibility is that the boundary condi- Appropriate selection, however, depends on under-
tions in the experimental model, and in vivo, are more standing the mechanical behavior and differences
complex than our single-fixed-end cantilever beam between individual products.
model (which also assumes a homogeneous solid struc- In challenging endovascular anatomy, usually more
ture), and that a different bending model may provide than one important mechanical property of SDS may
different results. It should be evident that a very stiff come in to play. Clinical scenarios exist that have par-
stent requires a high force to conform and pass ticularly demanding, specific performance parameters.
through small radii of curvature of this model. Cathe- As an example, greater SDS-pushability may become
ters on which the stents are mounted play an equally particularly important in patients with narrow proximal
important role. In our studies, the stiffness of the cath- access, such as seen in patients with small vessels and
eters was not precisely characterized. Nevertheless, our diffuse disease. Similar demands are expected in
test method allows a simple, easy, and precise compar- patients with diffuse or severe focal calcifications of
ative measurement of trackability, which is important the upstream segments of the target vessel. Likewise,
for clinical application. excellent trackability is required in patients with
Crossability depends, in addition to the same param- highly tortuous proximal vessels, particularly if com-
eters that control trackability, mainly on the profile of bined with diffuse and/or calcific disease and distal
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
 Mechanical Performance of Drug-Eluting Stent Systems 359

target lesions. High demands on crossability can be reproducible measurements of well-defined physical
expected in cases with tight and complex target variables.
lesions. Low bending stiffness appears to be particu- Pushing, tracking, and crossing abilities are compos-
larly beneficial in cases with narrow take offs of tar- ite parameters of the DES mechanical properties that
get vessels and tight bends in tortuous vascular path- do not lend themselves to straightforward measure-
ways. Low-profile SDS devices appear to be prerequi- ments. Their expressions in terms of active and reac-
sites in cases with extremely tight and complex target tive forces may represent simplifications allowing only
lesions. limited insights into their complex nature. Neverthe-
If manufacturers define and quantify mechanical less, exact measurements of the forces developed
behaviors of their stents during delivery, the data are during insertion allow objective comparisons of DES
not often available to the clinical community. The devices.
behavior definitions and measurement methods may
also vary from manufacturer to manufacturer. In the
absence of logical and consistent definitions for quanti- SUMMARY
fying mechanical behaviors during stent delivery, it is This study reports, for the first time, objective meas-
impossible to compare and contrast the behaviors of urements of DES performance and compares mechani-
different SDS. At ASTM Committee F04 on Medical cal responses of seven representative commercially
and Surgical Materials and Devices, subcommittee available DES systems. Differences in pushability,
F04.30 on Cardiovascular Standards, a task group is crossability, and trackability were found. The data sug-
currently developing a ‘‘Guide for endovascular de- gest that a small profile is necessary for good cross-
vice trackability and pushability’’ that may help to ability. Trackability and pushability are complex pa-
address the consistency of such measurements [26]. rameters influenced by many variables such as the
By creating and using consistent and meaningful defi- stiffness of the stent system and the friction between
nitions for delivery behaviors, facilities independent the system and the vasculature.
of manufacturers may be able to provide the clinician Systematic comparison as described here and a bet-
specific information they need to insure that the best ter understanding of the mechanical properties of DES
possible intervention is applied to each specific clini- systems may allow a more precise selection of the
cal case. DES system appropriate for the known and expected
Developing a better systematic understanding of characteristics of the target sites in individual patients
mechanical properties of SDS will improve the selec- and in other clinical settings.
tion of optimal instrumentation and markedly improve
the technical operator’s performance, particularly in
challenging cases. This work provides an initial REFERENCES
incentive to develop this important technical frame- 1. König A, Schiele TM, Rieber J, Theisen K, Mudra H, Klauss V.
work. Influence of stent design and deployment technique on neoin-
tima formation and vascular remodeling. Z Kardiol 2002;
91(Suppl 3):III/98–III/102.
LIMITATIONS OF THE STUDY 2. Nordmann AJ, Briel M, Bucher HC. Mortality in randomized
controlled trials comparing drug-eluting vs. bare metal stents in
In these initial tests, only one single specimen of coronary artery disease: A meta-analysis. E Heart J 2006;27:
each brand was tested. Although the stringent regula- 2784–2814.
tions imposed on production of medical devices should 3. Serruys PW, Kutryk MJB, Ong ATL. Coronary-artery stents. N
Engl J Med 2006;354:483–495.
assure a uniform mechanical behavior between all 4. Daemen J, Wenaweser P, Tsuchida K, et al. Early and late
stents of the same brand, variability of mechanical coronary stent thrombosis of sirolimus-eluting and paclitaxel-
properties of individual products has not yet been stud- eluting stents in routine clinical practice: Data from a large
ied for the DES systems of this study. two-institutional cohort study. Lancet 2007;369:667–678.
The in vitro experimental setting represents another 5. Camenzind E, Steg PG, Wijns W. Stent thrombosis late after
implantation of first-generation drug-eluting stents: A cause for
limitation of this study. Our stationary experimental in concern. Circulation 2007;115:1440–1455.
vitro vessel model consists of a plastic construct with 6. Sick PB, Thiele H, Brosteanu O, Hentschel B, Lauer B, Nieba-
smooth internal surface. It represents only an approxi- uer J, Hambrecht R, Schuler G. Prospective randomized compar-
mation of the complex in vivo endovascular ‘‘tunnels’’ ison of early and late results of 4 different stent designs. Am
with highly corrugated endoluminal interfaces and Heart J 2003;146:134–141.
7. Dyet JF, Watts WG, Ettles DF, Nicholson AA. Mechanical
neglects the complex boundary interactions between properties of metallic stents: How do these properties influence
the vessel wall and the SDS. However, our stand- the choice of stent for specific lesions? Cardiovasc Interv Radiol
ardized and validated experimental setup allows 2000;23:47–54.
Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.
Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).
360 Schmidt et al.

8. Ormiston JA, Dixon SR, Webster MWI, Ruygrok PN, stents: An experimental comparison. J Vasc Interv Radiol 2000;
Stewart JT, Minchington I, West T. Stent longitudinal flex- 11:645–654.
ibility: A comparison of 13 stent designs before and after 17. SOP026: Trackability, Internal Standard, IIB e.V., Rostock-
balloon expansion. Catheter Cardiovasc Interv 2000;50:120– Warnemünde, 2005.
124. 18. Schmitz K-P, Schmidt W, Behrens P, Behrend D, Lootz D, Graf
9. Rieu R, Barragan P, Garitey V, Roquebert PO, Fuseri J, Com- B. In-vitro examination of clinically relevant stent parameters.
meau P, Sainsous J. Assessment of the trackability, flexibility, Prog Biomed Res 2000;5:197–203.
and conformability of coronary stents: A comparative analysis. 19. Schmidt W, Andresen R, Behrens P, Schmitz K-P. Characteristic
Catheter Cardiovasc Interv 2003;59:496–503. mechanical properties of balloonexpandable peripheral stent sys-
10. Schmidt W, Grabow N, Behrens P, Schmitz K-P. Track- tems. Fortschr Röntgenstr 2002;174:1430–1437.
ability, crossability, and pushability of coronary stent sys- 20. Lewis G. Materials, fluid dynamics, and solid mechanics aspects
tems—An experimental approach. Biomed Tech 2002;47: of coronary artery stents: A state-of-the-art review. J Biomed
124–126. Mater Res B: Appl Biomater 2008;86:569–590.
11. Schmitz K-P, Behrens P, Schmidt W, Behrend D, Urbaszek W. 21. EN 14299. Non-active surgical implants—Particular require-
Quality determining parameters of balloon angioplasty catheters. ments for cardiac and vascular implants—Specific requirements
J Invasive Cardiol 1996;8:144–152. for arterial stents and endovascular prostheses. European Com-
12. Timmis AD. Percutaneous transluminal coronary angioplasty: mittee for Standardization, 2004.
Catheter technology and procedural guidelines. Br Heart J 1990; 22. ISO 25539-2. Cardiovascular implants—Endovascular devices,
64:32–35. Part 2: Vascular stents. International Organization for Standardi-
13. Schmidt W, Schmitz K-P. Devices. In: Lanzer P, editor. Mas- zation, 2008.
tering of Endovascular Techniques—A Guide to Excellence. 23. Kandzari DE, Tcheng JE, Zidar JP. Coronary artery stents: Eval-
Philadelphia: Lippincott William and Wilkins; 2006. pp 114– uating new designs for contemporary percutaneous intervention.
135. Catheter Cardiovasc Interv 2002;56:562– 576.
14. Wiskirchen J, Pusich B, Kramer U, Koenig C, Truebenbach J, 24. Topoleski LDT, Salunke NV, Humphrey JD, Mergner W. Com-
Tepe G, Claussen C, Duda S. Stent struts and articulations: position- and history-dependent radial compressive behavior of
Their impact on balloon-expandable stents’ hoop strength, push- human atherosclerotic plaque. J Biomed Mater Res 1997;35:
ability, and radiopacity in an experimental setting. Invest Radiol 117–127.
2002;37:356–362. 25. Salunke NV, Topoleski LDT, Humphrey JD, Mergner WJ. Com-
15. Zoarski GH, Mathis JM, Hebel JR. Performance characteristics pressive stress-relaxation of human atherosclerotic plaque.
of microcatheter systems in a standardized tortuous pathway. J Biomed Mater Res 2001;55:236–241.
AJNR Am J Neuroradiol 1998;19:1571–1576. 26. ASTM WK14950. New guide for endovascular device trackabil-
16. Duda SH, Wiskirchen J, Tepe G, Bitzer M, Kaulich TW, ity and pushability. Available at: www.astm.org/database.cart/
Stoeckel D, Claussen CD. Physical properties of endovascular workitems/WK14950.htm.

Catheterization and Cardiovascular Interventions DOI 10.1002/ccd.

Published on behalf of The Society for Cardiovascular Angiography and Interventions (SCAI).