Evaluation of a new model of hind limb ischemia in rabbit

Costantino Del Giudice, MD,a,b Gabriel Ifergan, MD,b Guillaume Goudot, MD,b,c Valerie Bellamy, MD,b
Emmanuel Messas, MD, PhD,b,c Olivier Clement, MD, PhD,b,d Patrick Bruneval, MD, PhD,b,e
Philippe Menasche, MD, PhD,b,f and Marc Sapoval, MD, PhD,a,b Paris, France

ABSTRACT
Objective: Various animal models of critical limb ischemia have been developed in the past. However, there is no animal
model that can undergo endovascular treatment, while providing reproducible true critical limb ischemia with arterial
ulcers and rest pain. We evaluated the efficacy of a new model of rabbit hindlimb ischemia created through a percu-
taneous approach using embolization with calibrated particles.
Methods: Through a percutaneous transauricular artery approach and selective catheterization of the superficial femoral
artery, embolization of distal limb vessels was performed using a mixture of 300- to 500-mm calibrated microparticles
(Embosphere, Merit Medical, Salt Lake City, Utah), saline solution, and iodine contrast. Clinical and ultrasound imaging-
based blood flow evaluation was performed before embolization and during follow-up. Histologic evaluation was per-
formed at humane killing 14 days after the procedure.
Results: The model was successfully created in 10 rabbits (10 limbs). One rabbit died of sudden death at 8 days after the
procedure. The nine surviving rabbits developed hind ulcers. All rabbits had a higher pain score in the follow-up
compared to baseline value (P < .0001). Blood flow in the saphenous artery decreased significantly after the proced-
ure and later at 14 days follow-up (baseline value 63.4 6 31.3 mL per cardiac cycle vs 32.0 6 28.4 mL per cardiac cycle
postprocedure [P ¼ .0013] and 32.0 6 28.4 mL per cardiac cycle at 14 days [P ¼ .0015]). Pathology showed signs of severe
limb ischemia in all rabbits with subacute and chronic injury patterns.
Conclusions: A rabbit hind limb ischemia model created by percutaneous transauricular distal femoral artery emboli-
zation with calibrated particles may overcome some of the limitations of existing animal models. As such, this model
could prove useful for assessing therapies designed to improve arterial perfusion and collateral growth. (J Vasc Surg
2017;-:1-9.)
Clinical Relevance: An in vivo animal model of critical limb ischemia, consisting of surgical excision of the superficial
femoral artery, presents limitations to prove the efficacy of therapies designed to improve arterial perfusion and
collateral growth, such as stem cell treatment. The natural propensity of collaterals to overcome the proximal arterial
occlusion usually precludes the onset of significant critical limb ischemia. We evaluated the efficacy of a new model of
rabbit hindlimb ischemia created through a percutaneous approach using embolization of distal hind vessels by
calibrated particles that may overcome some of limitations of existing animal models and prove useful for future
studies.

Critical limb ischemia (CLI) is a major public health prob-
lem worldwide with an estimated incidence of 500 to
1000 patients per million per year.1 The prevalence of this
From the Vascular and Oncological Interventional Radiology, Université Paris
Descartes, Hôpital Européen George Pompidou,a the Institut National de la
Santé et de la Recherche Médicale (INSERM), UMRS-970, Centre de
Recherche Cardiovasculaire, Université Paris Descartes, Sorbonne Paris Cité,b
the Angiology, Université Paris Descartes,c the Department of Radiology, Uni-
versité Paris Descartes,d the Department of Pathology, Université Paris Des-
cartes, Hôpital Européen George Pompidou,e and the Department of
Cardiovascular Surgery, Assistance Publique-Hôpitaux de Paris, Hôpital
Européen Georges Pompidou.f
This study was partly funded by the “Fondation Coeur et Artères.”
Author conflict of interest: none.
Correspondence: Costantino Del Giudice, MD, Vascular and Oncological Inter-
ventional Radiology, Université Paris Descartes, 20 Rue Leblanc, Île-de-France,
Paris 75015, France (e-mail: costantino.delgiudice@gmail.com).
The editors and reviewers of this article have no relevant financial relationships to
disclose per the JVS policy that requires reviewers to decline review of any
manuscript for which they may have a conflict of interest.
0741-5214
Copyright Ó 2017 by the Society for Vascular Surgery. Published by Elsevier Inc.
https://doi.org/10.1016/j.jvs.2017.07.140
disease increases rapidly owing to the aging population
and the growing prevalence of diabetes mellitus.2
Surgical and endovascular revascularizations are the
first-line approaches to improve blood flow to the
ischemic foot.3 However, particularly in advanced dis-
ease, revascularization is not always feasible, which leads
to a higher risk of major amputations. Recently, various
biotherapeutic angiogenic approaches, including cell
therapy, gene therapy, and recombinant proteins, have
been evaluated to improve revascularization in animal
models of limb ischemia and in clinical trials.4,5 These
therapies aim at improving distal arterial perfusion by
stimulating angiogenesis, with the growth of capillaries,
and arteriogenesis, with remodeling of existing arteriolar
connections into large caliber collateral vessels, all of
which represent a physiologic response to tissue
ischemia.6,7 Numerous efforts have been conducted in
the past to develop an appropriate preclinical animal
model mimicking a diabetic pattern with tibial and
pedal vessel involvement, which is less amenable to
endovascular or open surgical revascularization and
would most benefit from therapeutic arteriogenesis

1
2 Del Giudice et al
or angiogenesis. Rodents models have been used exten-
sively and have generated extremely useful data, but
without obtaining an ideal model of diabetic CLI.8,9
Although rodents models entailing a simple superficial
femoral artery (SFA) ligature, leave most of the collateral
circulation, with a recovery of distal blood flow within
7 days,10 surgical excision of the SFA associated with
that of the iliac artery could cause a rapid necrosis with
acute limb ischemia.11
Because microembolization is able to occlude small
arteries distally,12 we anticipated that the direct intra-
arterial injection of calibrated microparticles could
overcome this limitation and allow the induction of sig-
nificant CLI. In this study, a new model of rabbit hind
limb ischemia was created through a percutaneous
approach using embolization with calibrated particles.
METHODS
Thirty New Zealand White rabbits weighing 4.5 to 5.5 kg
were kept in separate cages in an environmentally
controlled animal research facility. Food and water
were provided ad libitum. The local Hospital’s Scientific
and Ethics Committee (Project no. 16003) approved this
investigation. A veterinarian dedicated to our laboratory
supervised the study and evaluated the daily pain and
discomfort of rabbits.
Overall description of the stepwise approach. The
development of the model proceeded through two
distinct stages:
1. At first, 10 consecutive normal rabbits were treated
with different protocols, increasing and modifying
particle sizes and testing different sites of emboli-
zation. Using this doseeresponse method, we
could optimize the various parameters and pro-
ceeded to the next phase that we report in this
study.
2. In this phase, global morphological and functional
assessment of the embolization-induced rabbit
hind ischemia model was performed in 10 rabbits
with the optimal protocol defined at step 1.
Procedures. All procedures were performed under gen-
eral anesthesia with a mixture of ketamine (50 mg/kg)
and xylazine (5 mg/kg) given intramuscularly. Cardiovas-
cular monitoring was used during the whole procedure
to control heart rate and arterial oxygen saturation. Rab-
bits were ventilated through a laryngeal mask at a rate of
20 breaths per minute and a tidal volume of 45 to 50 mL
using room air enriched with oxygen. A 72-hour trans-
dermal fentanyl patch 12 mg/h was used after induction
to obtain a pain control during and after the procedure.
Hind skin was shaved bilaterally.
Hind limb ischemia model. The central auricular artery
was punctured with a 21-gauge, 4-cm needle (Cook
Journal of Vascular Surgery
--- 2017
ARTICLE HIGHLIGHTS
d
Type of Research: Experimental rabbit model of
limb ischemia using percutaneous endovascular
embolization with blood flow measurements and
histologic evaluation
d
Take Home Message: In nine surviving rabbits of 10
animals, selective embolization of the superficial
femoral artery resulted in reproducible subacute
ischemic changes with reduced blood flow, limb
pain ulceration, and histologic changes similar to
human ischemic injuries at 14 days.
d
Recommendation: The authors propose a reproduc-
ible ischemic rabbit model to study effective thera-
pies of subacute limb ischemia.
Medical, Bloomington, Ind)13 (Fig 1, A). A 0.014-inch hydro-
philic guide wire (Advantage, Terumo, Tokyo, Japan) was
carefully advanced into the external carotid artery and
navigated down to the aortic arch. A standard 18-guage
catheter (Angiocath, BD Medical, Salt Lake City, Utah) was
then advanced at the site puncture to increase guidewire
support and serve as an arterial introducer (Fig 1, B). A 2.7
microcatheter (Progreat, Terumo, Japan) was finally
advanced in the aortic arch and, selectively, down into the
left SFA (Fig 1, C and D). A pre-embolization angiography (by
injecting 3 mL of contrast medium [Iopamidol 370, Bracco,
Italy]) was then performed (Fig 2, A). The right SFA was not
embolized and served as a control limb in each rabbit.
For embolization, we used the following mixture: a 1-mL
vial of 300- to 500-mm calibrated particles (Embo-
spheres, Merit, Salt Lake City, Utah) was diluted in 7 mL
of saline solution and 7 mL of iopamidol, to obtain a fluid
radiopaque mixture. Embolization of both anterior tibial
and saphenous arteries was performed using 6 mL of the
mixture infused through the microcatheter. A final
angiographic control was performed to confirm success-
ful embolization (Fig 2, B). Hemostasis was obtained by
manual compression of the puncture site and the animal
was awakened.
Clinical evaluation and follow-up. Apart from general
daily care, a specific per protocol clinical assessment
was performed before (T0) and after the procedure (T1),
at 7 days (T2) and at 14 days before humane killing (T3),
evaluating the animal’s general behavior, position at
rest, oral intake, movements under solicitation, and mo-
tion of the embolized limb zone. Because pain is a major
symptom in CLI, we used a dedicated pain score previ-
ously validated in rabbits at each clinical control
(Table I).14
In addition, at each clinical control point (T0-T3), oxygen
saturation (SaO2) was measured on each hindlimb. An
angiographic control was performed with an auricular
access at T3 (Fig 2, C).
Journal of Vascular Surgery Del Giudice et al 3
Volume -, Number -

Fig 1. Embolization of the hind limb. A, Puncture of the auricular artery with a 22-gauge needle. B, Advance-
ment of 0.014-inch guidewire and insertion of an 18-gauge catheter. C, Preparation of the calibrated particles
with contrast medium and saline solution. D, Advancement of the microcatheter to perform a selective femoral
artery catheterization.

Fig 2. Preprocedural angiography (A), postprocedural angiography after embolization of the superficial femoral
artery (SFA) at the bifurcation (B), and control (before human killing) (C).

Duplex scan color Doppler evaluation. An echo-color
Doppler evaluation was performed at T0, T1, and T3 in
group 2. A high-frequency, high-resolution digital imag-
ing platform (linear array MS400, 18-38-MHz frequency;
Vevo 2100 Imaging system VisualSonics, Amsterdam, the
Netherlands) was used. In each rabbit, evaluation of the
vessel area (p*[arterial diameter/2]2) and the velocity
time integral were measured into the SFA, approxi-
mately 1 cm distal to the bifurcation of common femoral
artery, and saphenous artery, approximatively 1 cm distal
to the bifurcation of the SFA, in the treated limb and the
contralateral limb. Evaluation of the flow per cardiac
cycle (CC) was evaluated according to the equation:
flow ¼ vessel area  velocity time integral and expressed
as mL/CC, as previously described.15,16
Histologic evaluation. A histologic analysis of the
embolized and control limbs was performed in all rab-
bits. Immediately after humane killing, ischemic muscle
samples were obtained and fixed in 10% formaldehyde
for histologic analysis, processed following standard
procedures, embedded in paraffin, and sectioned. Each
4 Del Giudice et al Journal of Vascular Surgery
--- 2017

Table I. Rabbit Pain score Statistics. All data were analyzed by SPSS 20.0 (SPSS
Variables Score Inc., Chicago, Ill). Categorical variables were expressed
as absolute values and percentages and compared
Activity level
with the c2 test. Numerical variables were expressed as
Very active 0
mean value 6 standard deviation and compared with
Attentive without activity 1
paired Student t test. Data analysis for repeated mea-
Awake but prostrate 2
sures was performed using one- or two-way analysis
Apathetic/unconscious 3 of variance followed by Bonferroni tests. A two-sided
Behavior P < .05 was considered significant in all analyses.
Normal, grooming activities 0
Agitation, nervousness, apathy 1
RESULTS
Absence of characteristic behaviors 2
The model was successfully created in all 10 rabbits.
Self-harm 3 One rabbit died of sudden cardiac death at 8 days after
Rest posture the procedure. The nine surviving rabbits developed sig-
Comfortable, sphinx 0 nificant hind ulcers (Fig 3): seven small ulcers (<1 cm2)
Agitated, positioning difficulties 1 were localized in the toe pads and two large ulcers
Analgesic posture/hunched 2 (>1 cm2) extended in the hindfoot and leg. One rabbit
Aspect developed a large blister in the hindfoot with edema
Well-kept fur, shiny 0 before humane killing (Table II).
Dive fur, nasal secretions 1 All rabbits were symptomatic with a stable pain score
Defiled fur, spasms 2
during the observational period. In fact, the pain score
evaluated during the follow-up after the procedure was
Feed
significantly higher compared to value at T0 (P < .001),
Normal 0
without significant difference between values observed
Less 1
at T1, T2, and T3 (respectively pain score 0 at T0,
Very little 2 4.1 6 1.2 at T1, 4.4 6 2.1 at T2, and 5.0 6 3.4 at T3).
Not at all 3 The SaO2 was similar in the preprocedural period (100%
Soliciting movement in both limbs). Conversely, at the T1 and T3 time points, a
Fluid movements, ease 0 significant difference between the embolized hind and
Hesitations but normal locomotion 1 the control was observed: 79.8 6 9.8% in the embolized
Abnormal locomotion (lameness) 2 hind vs 100% in the contralateral limb at T1 (P < .0001)
Refusal movement, defense 3 and 81.1 6 14.1% vs 100% in the contralateral limb at T3.
Pain The difference between SaO2 values measured at T0
No response 0
and T1 and at T0 and T3 was statistically significant (P ¼
.0032 and P ¼ .0039, respectively).
Increased response (muscular tone) 1
Resting blood flow evaluated in the treated hind limb
Jumps, escape attempt 2
and the contralateral limb at T0 was similar in the SFA
Violent reaction (to debate, shouts) 3
(238.5 6 46.03 mL/CC and 241.0 6 91.01 mL/CC, respectively;
P ¼ NS) and saphenous artery (68.4 6 31.3 mL/CC and
77.3 6 15.4 mL/CC, respectively; P ¼ NS). After emboliza-
tion, an immediate reduction of resting blood flow in
section was stained with hematoxylin and eosin. One the SFA was observed at T1 (68.4 6 31.3 mL/CC; P <
independent pathologist blindly analyzed the samples .0001) with return to normal at T3 (261.5 6 201.5 mL/CC;
and characterized the lesions. Gastrocnemius muscles P ¼ NS) (Fig 4). In contrast, the resting blood flow into
were all evaluated for necrosis, inflammatory, fibrosis, the saphenous artery decreased at T1 (32.8 6 19.4 mL/CC;
and atrophic changes. P ¼ .0013) but remained significantly lower than before
the procedure at T3 (32.0 6 28.4 mL/CC; P ¼ .0015).
Endpoints of the study. The severity of ischemia was
The comparative limb showed that resting blood
assessed on the presence of pain and/or skin ulcerations
flow was stable during the follow-up in the SFA
for more than 2 weeks, according to the TASC classifica-
(241.04 6 91.8 mL/CC at T0, 191.7 6 53 mL/CC at T1, and
tion (Inter-Society Consensus for the Management of
388.6 6 247.3 mL/CC at T2; P ¼ NS) and in the saphenous
Peripheral Arterial Disease).17 Secondary endpoints were
artery (77.3 6 15.4 mL/CC at T0, 71.0 6 35.1 mL/CC at T1 and
pain scale (Table I), oxygen saturation, and arterial flow
81.6 6 72.0 mL/CC at T3; P ¼ NS; Fig 4).
into the SFA and the saphenous artery. All these pa-
rameters were compared between the ischemic limb Histologic analysis. Histologic findings are described
and the contralateral control. in Table III. Histologic analysis was performed in
Journal of Vascular Surgery Del Giudice et al 5
Volume -, Number -

Fig 3. Hind ulcers developed after embolization in the toe pads (A-D) and leg (C).

Table II. Clinical evaluation of hind limb ischemia necrosis, progressively surrounded by inflammation with
Variables n (%) numerous eosinophil and replacement fibrosis. An in-
flammatory response, characterized by eosinophil infil-
Immediate success (n, %) 10/10 (100)
tration followed by skeletal myocyte replacement with
Death
scar tissue and necrosis, was observed in all treated
Periprocedural 0/10
limbs. Only in one rabbit was an arterial thrombosis
Early death 1/10 (10) observed (Fig 5). In the contralateral control limbs, no
Ulcers 9/10 (90) signs of ischemia were observed.
Small (<1 cm2) 7/10 (70)
Large (>1 cm2) 2/10 (20)
Blister 1/10 (10) DISCUSSION
Edema 1/10 (10) In this study, we describe a model of a sustainable and
Pain score under morphine treatment stable hind limb ischemia lasting at least 2 weeks. The
Mild pain (<3 points) 3/10 (30) animals demonstrated both clinical signs and objectives
Moderate pain (>3 points) 7/10 (70) measures of CLI. This model might, thus, serve as a clini-
cally relevant preclinical tool to evaluate various treat-
ments, including growth factors and stem cells,
targeted at promoting neovascularization of ischemic
nine rabbits; the rabbit that suddenly died during limbs and increased tissue perfusion to prevent necrosis.
follow-up was excluded from analysis. Histologic sam- Several animal models of CLI have previously been
ples showed lesions related to ischemia induced by described, but they are fraught with limitations that
particle embolization. These lesions followed the hamper their clinical relevance18-20 and might explain
clinical kinetics of ischemic injury9 with central muscle the failure of the human trials that have relied on
6 Del Giudice et al
Fig 4. Blood flow of the superficial femoral artery (SFA)
and saphenous femoral artery (SA) in the treated hindlimb
(continuous line) and controls (dashed line) at preproce-
dural (T1), postprocedural (T2), and before human killing
(T3) ultrasound assessments.
Table III. Histologic findings
Treated Control
Variables limb, n/N (%) limb, n/N
Necrosis 5/9 (55.6) 0/9 .02
Scar tissue 8/9 (77.8) 0/9 .002
Inflammatory response 5/9 (55.6) 0/9 .02
Arterial thrombosis 1/9 (11) 0/9 .3
them.21 These hind limb ischemia models have used
both small and large animals.22-24
The use of small animals, such as rodents, has the
advantage of easy management, unlimited availability
and low cost, whereas large animals like the pig and
sheep have a size and physiology that more closely
mimic humans,8 but they are logistically more complex
to handle and their higher cost usually prevents large
sample sizes, thereby reducing the statistical power of
subsequent analyses. The rabbit may represent an attrac-
tive trade-off because of its anatomy that mimics the
human vascularization with hind perfusion supported
by a femoral artery that bifurcates into a saphenous ar-
tery, which corresponds with the posterior tibial artery,
and a popliteal artery, which vascularizes the front of
the paw. Furthermore, the commonly used surgical
models are created by surgical dissection and ligation
of the SFA, which may not provide the expected out-
comes for various reasons. First, surgery is an invasive
approach that triggers an inflammatory response associ-
ated with wound healing. This inflammatory response
leads to the release of several cytokines and growth fac-
tors, which may themselves stimulate angiogenesis and
Journal of Vascular Surgery
--- 2017
arteriogenesis, thereby potentially contributing to the
confounding of any therapeutic intervention.25 Liddell
et al26 have demonstrated that a hind limb ischemia
created in rabbit by coiling the SFA through an endovas-
cular approach resulted in a reduced inflammatory
response and hind limb reperfusion compared with the
surgical group. In our study, ischemia was created by a
totally percutaneous approach by a transauricular arte-
rial access, as previously described by Karnabatidis
et al.13 This access allowed us to limit the inflammatory
response related to surgical dissection and ligation.
Second, the total percutaneous approach is less inva-
sive than open access, as demonstrated by the high post-
operative survival rate of our animals. A final major
limitation of the commonly used surgical hindlimb
ischemia models is the progressive blood flow restora-
tion from 7 days after surgery because of the fast-
growing development of collaterals.27 Several procedural
changes have thus been developed to overcome this
roadblock, including single ligation of the superficial
and iliac arteries,28 a double ligation of the SFA,29 and
the combined ligation of vein and nerve.30 Seifert
et al28 evaluated the ligation of iliac branches and lum-
bar artery, from which collaterals potentially develop,
associated with a left femoral artery ligation. However,
P
limb ischemia did not persist beyond 5 days after the
procedure. Hendricks et al31 have investigated the associ-
ated ligation of the iliac and femoral arteries without
obtaining a satisfactory result regarding the sustainability
of ischemia. Indeed, surgical ligation leaves the collateral
circulation intact and a total surgical excision of the SFA
leads to a stimulation of the arteriogenesis and angio-
genesis processes, which results in the creation of a
collateral network to the distal vasculature, with a resto-
ration of distal flow within 7 days. If a single ligation of
the SFA results only in a mild ischemia, that mimics an
intermittent claudication rather than a CLI, proximal
and multilevel arterial ligation results in more severe
ischemia with greater distal disease, but also a rapid ne-
crosis with autoamputation of the limb.8 The associated
ligation of the vein, even if it reduces the reperfusion phe-
nomenon, inhibits collateral vessel formation and thus
represents a confounding factor which could prevent
the proper assessment of an angiogenic therapy.8 CLI is
a complex disease characterized by a multilevel injury
with macroangiopathy and microangiopathy.
In our model, the slow infusion of 300- to 500-mm cali-
brated particles reliably restricted flow in small distal ves-
sels with a good distribution throughout the limb. These
particles, routinely used for embolization procedures in
humans, are able, once injected from a large vessel, to
navigate through the flow and cause the occlusion of
distal capillary vessel of the same range, as previously
demonstrated.12 This approach allowed the reduction
of collateral recruitment and mimicked a microangiop-
athy process similar to that observed in CLI. All rabbits
Journal of Vascular Surgery Del Giudice et al 7
Volume -, Number -

Fig 5. Histologic patterns of gastrocnemius muscle. A, Low magnification showing the organization of the le-
sions in the ischemic muscle at day 14. Between the black lines, areas of necrotic skeletal muscle cells (N).
Between the black lines and the red lines and in an islet surrounded by a black line, granulation tissue (G).
Between the red lines and the blue lines, replacement fibrosis (F). The rest of the skeletal muscle on the left is
normal. Stain: hematoxylin and eosin; original magnification, 2. B, Typical ischemic necrosis of the skeletal
muscle cells in the center of the injured muscle. This area is the remnant of the initial ischemic focus at day 0,
the rest having been replaced by granulation tissue and scar fibrosis during the 14-day evolution. Stain: he-
matoxylin and eosin; original magnification, 20. C, Granulation tissue replacing and surrounding the necrotic
muscle cells. It is made of polymorphic inflammatory cells and new vessels. Stain: hematoxylin and eosin;
original magnification, 20. D, Fibrosis area at the periphery of the granulation tissue. It is made of fibroblasts
and abundant extracellular matrix. It may include some skeletal muscle cells. Stain: hematoxylin and eosin;
original magnification, 20. E, Artery thrombosis induced by the interventional procedure in the gastrocnemius
muscle. Stain: hematoxylin and eosin; original magnification, 20. F, The contralateral nonischemic gastroc-
nemius muscle shows a normal histologic pattern. Stain: hematoxylin and eosin; original magnification, 20.

treated in our study had pain after the procedure, which ulcers that were similar to those observed in diabetics
was stable throughout the 2-week follow-up, as assessed with foot skin wounds. Ultrasound evaluation reported
by a dedicated pain score. Nine rabbits developed distal a significant reduction of blood flow in the superficial
8 Del Giudice et al
and saphenous arteries in the immediate postemboliza-
tion period. Moreover, at the end of the follow-up, a par-
tial recovery of the SFA flow was observed as a result of
the collateral flow recruited from the profunda femoral
artery, but a stable reduction of the flow in the saphe-
nous artery confirmed the ability of our model to mimic
the microangiopathy state associated with CLI. This state
was also confirmed by the stably reduced values of SO2
evaluated on the distal hind. Moreover, clinical and
instrumental observations were confirmed by histologic
findings that described lesions evolving to a chronic
ischemia pattern, characterized by partial or complete
replacement of myocyte necrosis by inflammation and
fibrosis.
Another advantage of the rabbit model is that geneti-
cally modified species can develop atherosclerotic
lesions under an appropriate saturated fat regimen.32
These models are characterized by the development of
proximal atherosclerotic plaques that are not related to
ischemia. The potential combination of an associated
saturated fat regimen in these species could further
strengthen the potential clinical relevance of this model.
Despite the pathophysiologic similarity with human
setting, this model has some limitations. First, human
CLI is a pathologic condition that develops gradually. Le-
sions cause progressive vessel occlusions and not acute
embolization like in our model. However, like in CLI, distal
embolization allows to obtain a homogenous distribu-
tion of vessel lesions, which may induce neoangiogenesis
and arteriogenesis akin to what can be observed in
humans. Another limitation of the study is the duration
of the follow-up, which was limited to 14 days to comply
with the Ethical Committee recommendations and
those of the laboratory veterinarian, for both reducing
rabbit discomfort and matching the TASC definition of
CLI. Nevertheless, it remains to be seen how long signifi-
cant ischemia persists in our model, and will be required
to have a more prolonged ischemic period for gener-
ating clinically relevant efficacy outcomes.
In conclusion, the present study shows that a rabbit
hindlimb ischemia model created by the minimally inva-
sive percutaneous transauricular embolization of the
distal vessels with calibrated particles may overcome
some of the limitations of existing animal models. As
such, this model could prove useful for assessing thera-
pies designed to improve arterial perfusion and collateral
growth.
Thanks to Dr. Chachques JC (Laboratory of Biosurgical
Research “Carpentier Foundation”, Pompidou Hospital,
University Paris Descartes, France) for the creation of
the research protocol. We acknowledge the major tech-
nical contribution of Lalot A. et Piquet J. (Laboratory of
Biosurgical Research “Carpentier Foundation”, Pompidou
Hospital, University Paris Descartes, France) for the care
and welfare of laboratory animals.
Journal of Vascular Surgery
--- 2017
AUTHOR CONTRIBUTIONS
Conception and design: CDG, GI, VB, EM, CO, PB, PM, MS
Analysis and interpretation: CDG, CO, PM, MS
Data collection: CDG, GI, GG
Writing the article: CDG, PB, PM, MS
Critical revision of the article: CDG, GI, GG, VB, EM, CO, PB,
PM, MS
Final approval of the article: CDG, GI, GG, VB, EM, CO, PB,
PM, MS
Statistical analysis: CDG
Obtained funding: PM, MS
Overall responsibility: CDG
REFERENCES
1. Ouriel K. Peripheral arterial diseases. Lancet Lond Engl
2001;358:1257-64.
2. Clair D, Shah S, Weber J. Current state of diagnosis and
management of critical limb ischemia. Curr Cardiol Rep
2012;14:160-70.
3. Adam DJ, Beard JD, Cleveland T, Bell J, Bradbury AW,
Forbes JF, et al. Bypass versus angioplasty in severe
ischaemia of the leg (BASIL): multicentre, randomised
controlled trial. Lancet Lond Engl 2005;366:1925-34.
4. Olea FD, Locatelli P, Hnatiuk A, De Lorenzi A, Valdivieso L,
Rocha E, et al. Vascular endothelial growth factor over-
expression does not enhance adipose stromal cell-induced
protection on muscle damage in critical limb ischemia.
Arterioscler Thromb Vasc Biol 2015;35:184-8.
5. Peeters Weem SMO, Teraa M, de Borst GJ, Verhaar MC,
Moll FL. Bone marrow derived cell therapy in critical limb
ischemia: a meta-analysis of randomized placebo controlled
trials. Eur J Vasc Endovasc Surg Off J Eur Soc Vasc Surg
2015;50:775-83.
6. Heil M, Eitenmüller I, Schmitz-Rixen T, Schaper W. Arterio-
genesis versus angiogenesis: similarities and differences.
J Cell Mol Med 2006;10:45-55.
7. Lawall H, Bramlage P, Amann B. Treatment of peripheral
arterial disease using stem and progenitor cell therapy.
J Vasc Surg 2011;53:445-53.
8. Lotfi S, Patel AS, Mattock K, Egginton S, Smith A, Modarai B.
Towards a more relevant hind limb model of muscle
ischaemia. Atherosclerosis 2013;227:1-8.
9. Long CA, Timmins LH, Koutakis P, Goodchild TT, Lefer DJ,
Pipinos II, et al. An endovascular model of ischemic myop-
athy from peripheral arterial disease. J Vasc Surg 2017;66:
891-901.
10. Hellingman AA, Bastiaansen AJNM, de Vries MR, Seghers L,
Lijkwan MA, Löwik CW, et al. Variations in surgical proced-
ures for hind limb ischaemia mouse models result in dif-
ferences in collateral formation. Eur J Vasc Endovasc Surg
Off J Eur Soc Vasc Surg 2010;40:796-803.
11. Masaki I, Yonemitsu Y, Yamashita A, Sata S, Tanii M,
Komori K, et al. Angiogenic gene therapy for experimental
critical limb ischemia: acceleration of limb loss by over-
expression of vascular endothelial growth factor 165 but not
of fibroblast growth factor-2. Circ Res 2002;90:966-73.
12. Stampfl S, Bellemann N, Stampfl U, Sommer CM,
Thierjung H, Lopez-Benitez R, et al. Arterial distribution
characteristics of Embozene particles and comparison with
other spherical embolic agents in the porcine acute
embolization model. J Vasc Interv Radiol JVIR 2009;20:
1597-607.
Journal of Vascular Surgery
Volume -, Number -
13. Karnabatidis D, Katsanos K, Diamantopoulos A, Kagadis GC,
Siablis D. Transauricular arterial or venous access for car-
diovascular experimental protocols in animals. J Vasc Interv
Radiol JVIR 2006;17:1803-11.
14. Hawkins P. Recognizing and assessing pain, suffering and
distress in laboratory animals: a survey of current practice in
the UK with recommendations. Lab Anim 2002;36:378-95.
15. Fernández del Palacio MJ, Luis Fuentes V, Bonagura JD,
Schober KE, Hatfield DG, Laughlin MH. Evaluation of trans-
cutaneous Doppler ultrasonography for the measurement
of blood flow in the femoral artery of pigs. Am J Vet Res
2003;64:43-50.
16. dos Reis GFM, Nogueira RB, Silva AC, Oberlender G,
Muzzi RAL, Mantovani MM. Spectral analysis of femoral ar-
tery blood flow waveforms of conscious domestic cats.
J Feline Med Surg 2014;16:972-8.
17. Norgren L, Hiatt WR, Dormandy JA, Nehler MR, Harris KA,
Fowkes FGR, et al. Inter-Society Consensus for the Manage-
ment of Peripheral Arterial Disease (TASC II). Eur J Vasc
Endovasc Surg Off J Eur Soc Vasc Surg 2007;33(Suppl 1):S1-75.
18. Attanasio S, Snell J. Therapeutic angiogenesis in the man-
agement of critical limb ischemia: current concepts and
review. Cardiol Rev 2009;17:115-20.
19. Aranguren XL, Verfaillie CM, Luttun A. Emerging hurdles in
stem cell therapy for peripheral vascular disease. J Mol Med
Berl Ger 2009;87:3-16.
20. Emanueli C, Madeddu P. Therapeutic angiogenesis: trans-
lating experimental concepts to medically relevant goals.
Vascul Pharmacol 2006;45:334-9.
21. Leeper NJ, Hunter AL, Cooke JP. Stem cell therapy for
vascular regeneration: adult, embryonic, and induced
pluripotent stem cells. Circulation 2010;122:517-26.
22. Arras M, Ito WD, Scholz D, Winkler B, Schaper J, Schaper W.
Monocyte activation in angiogenesis and collateral growth
in the rabbit hindlimb. J Clin Invest 1998;101:40-50.
23. Burkhardt GE, Spencer JR, Gifford SM, Propper B, Jones L,
Sumner N, et al. A large animal survival model (Sus scrofa)
of extremity ischemia/reperfusion and neuromuscular
Del Giudice et al 9
outcomes assessment: a pilot study. J Trauma 2010;69(Suppl
1):S146-53.
24. Nakada MT, Montgomery MO, Nedelman MA, Guerrero JL,
Cohen SA, Barnathan ES, et al. Clot lysis in a primate model
of peripheral arterial occlusive disease with use of systemic
or intraarterial reteplase: addition of abciximab results in
improved vessel reperfusion. J Vasc Interv Radiol JVIR
2004;15:169-76.
25. Buschmann I, Heil M, Jost M, Schaper W. Influence of in-
flammatory cytokines on arteriogenesis. Microcirc N Y N
1994 2003;10:371-9.
26. Liddell RP, Patel TH, Weiss CR, Lee DS, Matsuhashi T,
Brown PR, et al. Endovascular model of rabbit hindlimb
ischemia: a platform to evaluate therapeutic angiogenesis.
J Vasc Interv Radiol JVIR 2005;16:991-8.
27. Hoefer IE, van Royen N, Buschmann IR, Piek JJ, Schaper W.
Time course of arteriogenesis following femoral artery oc-
clusion in the rabbit. Cardiovasc Res 2001;49:609-17.
28. Seifert FC, Banker M, Lane B, Bagge U, Anagnostopoulos CE.
An evaluation of resting arterial ischemia models in the rat
hind limb. J Cardiovasc Surg (Torino) 1985;26:502-8.
29. Couffinhal T, Silver M, Zheng LP, Kearney M, Witzenbichler B,
Isner JM. Mouse model of angiogenesis. Am J Pathol
1998;152:1667-79.
30. Westvik TS, Fitzgerald TN, Muto A, Maloney SP, Pimiento JM,
Fancher TT, et al. Limb ischemia after iliac ligation in aged
mice stimulates angiogenesis without arteriogenesis. J Vasc
Surg 2009;49:464-73.
31. Hendricks DL, Pevec WC, Shestak KC, Rosenthal MC,
Webster MW, Steed DL. A model of persistent partial hin-
dlimb ischemia in the rabbit. J Surg Res 1990;49:453-7.
32. Waqar AB, Koike T, Yu Y, Inoue T, Aoki T, Liu E, et al. High-fat
diet without excess calories induces metabolic disorders
and enhances atherosclerosis in rabbits. Atherosclerosis
2010;213:148-55.
Submitted Mar 23, 2017; accepted Jul 28, 2017.