From the New England Society for Vascular Surgery
Infrapopliteal angioplasty for critical limb
ischemia: Relation of TransAtlantic InterSociety
Consensus class to outcome in 176 limbs
Kristina A. Giles, MD, Frank B. Pomposelli, MD, Allen D. Hamdan, MD, Seth B. Blattman, MD,
Haig Panossian, BS, and Marc L. Schermerhorn, MD, Boston, Mass
Objective: Recent data suggest that percutaneous transluminal angioplasty (PTA) may be appropriate primary therapy for
critical limb ischemia (CLI). However, little data are available regarding infrapopliteal angioplasty outcomes based on
TransAtlantic InterSociety Consensus (TASC) classification. We report our experience with infrapopliteal angioplasty
stratified by TASC lesion classification.
Methods: From February 2004 to March 2007, 176 consecutive limbs (163 patients) underwent infrapopliteal angioplasty
for CLI. Stents were placed for lesions refractory to PTA or flow-limiting dissections. Patients were stratified by TASC
classification and suitability for bypass grafting. Primary outcome was freedom from restenosis, reintervention, or
amputation. Primary patency, freedom from secondary restenosis, limb salvage, reintervention by repeat angioplasty or
bypass, and survival were determined.
Results: Median age was 73 years (range, 39-94 years). Technical success was 93%. Average follow-up was 10 months
(range, 1-41 months). At 1 and 2 years, freedom from restenosis, reintervention, or amputation was 39% and 35%,
conventional primary patency was 53% and 51%, and freedom from secondary restenosis and reintervention were 63% and
61%, respectively. Limb salvage was 84% at 1, 2, and 3 years. Within 2 years, 15% underwent bypass and 18% underwent
repeat infrapopliteal PTA. Postoperative complications occurred in 9% and intraprocedural complications in 10%. The
30-day mortality was 5% (9 of 181). Overall survival was 81%, 65%, and 54% at 1, 2, and 3 years. TASC D classification
predicted diminished technical success (75% D vs 100% A, B, and C; P < .001), primary restenosis, reintervention, or
amputation (hazard ratio [HR], 3.4; 95% confidence interval [CI], 2.1-5.5, P < .001), primary patency (HR, 2.2; 95%
CI, 1.3-3.9, P < .004), secondary restenosis (HR, 3.2; 95% CI, 1.6-6.4, P ? .001), and limb salvage (HR, 2.6; 95% CI,
1.1-6.3, P < .05). Unsuitability for surgical bypass also predicted restenosis, reintervention, or amputation, secondary
restenosis, need for repeated angioplasty, and inferior primary patency and limb salvage rates.
Conclusion: Infrapopliteal angioplasty is a reasonable primary treatment for CLI patients with TASC A, B, or C lesions.
Restenosis, reintervention, or amputation was higher in patients who were unsuitable candidates for bypass; however, an
attempt at PTA may be indicated as an alternative to primary amputation. Although restenosis, reintervention, or
amputation is high after tibial angioplasty for CLI, excellent limb salvage rates may be obtained with careful follow-up
and reinterventions when necessary, including bypass in 15%. (J Vasc Surg 2008;48:128-36.)
Good technical and clinical results have been obtained
with pedal bypass for the treatment of tibial occlusive
disease causing limb ischemia.1A combination of excellent
durability and low mortality make this procedure an attrac-
tive option for patients with a threatened extremity. How-
ever, significant morbidity can be associated with distal
bypass surgery. A mortality rate approximating 2% is typi-
cally cited, but some studies have found rates as high as 5%
for distal bypass surgery. Patients may lack conduit or
target, be nonambulatory, have a limited life expectancy,
have an extensive soft tissue infection overlying a bypass
target, or infrequently have comorbidities that make them
an unacceptable risk. In these patients, percutaneous trans-
luminal angioplasty (PTA) may constitute a feasible revas-
cularization method rather than primary amputation.
Recently, the Bypass Versus Angioplasty in Severe Isch-
emia of the Leg (BASIL) study suggested that if the anatomy
priate first therapy even if the patient is a good candidate for
bypass. Ideal anatomy was not well defined in BASIL, how-
ever, and outcomes were not stratified by the distal extent of
disease (superficial femoral/popliteal/tibial).7
Outcomes of tibial PTA are difficult to predict from the
existing literature owing to a lack of details regarding
indications for intervention and lesion characteristics.6-17
The TransAtlantic InterSociety Consensus (TASC) criteria
represents a standardized definition for lesion characteris-
tics (Table I).8,9Our objective was to review our results of
infrapopliteal angioplasty stratifying patients by anatomic
characteristics according to the TASC classification.
We performed a retrospective study analyzing periop-
erative and follow-up outcomes of infrapopliteal angio-
plasty for patients with critical limb ischemia. The study
From Beth Israel Deaconess Medical Center.
Competition of interest: none.
Presented at the Thirty-fourth Annual Meeting of the New England Society
for Vascular Surgery, Ledyard, Conn, Oct 5-7, 2007.
Reprint requests: Marc Schermerhorn, MD, Beth Israel Deaconess Medical
Center, 110 Francis St, 5B, Boston, MA 02215 (e-mail: mscherm@
Copyright © 2008 by The Society for Vascular Surgery.
protocol was approved by the Beth Israel Deaconess Med-
ical Center (BIDMC) Institutional Review Board.
Subjects and setting. All patients who underwent an
endovascular procedure performed by a member of the
Division of Vascular and Endovascular Surgery at the
BIDMC consecutively from February 2004 to March 2007
were recorded in a computerized vascular registry. Demo-
graphics, procedural details, and in-hospital outcomes are
prospectively recorded in this registry. From this database,
we retrospectively reviewed all patients who underwent an
attempt at a percutaneous, infrapopliteal angioplasty pro-
cedure on a native vessel.
Interventions were performed for critical limb ischemia
(CLI), defined as tissue loss, rest pain, or a stenosis in the
outflow vessel of a tibial bypass combined with low graft
flow velocities that threatened the graft viability. Not in-
cluded in this review are five interventions that were
performed during this time period for disabling claudi-
cation in patients who also had concurrent femoropop-
liteal disease. Preoperative segmental pressures and
Doppler waveforms were obtained in all elective cases. In
our patients with diabetes, we routinely obtain forefoot
pulse volume recordings because of the well-known un-
reliability of ankle-brachial index (ABI) measurements in
these patients. Palpable distal pulses were absent in all
patients, and all patients with preoperative vascular stud-
ies had dampened forefoot pulse volume recordings.
Most cases were performed in a dedicated angiography
suite, initially in a cardiac catheterization laboratory and
eventually in an operating room endovascular suite. A few
C-arm. Most procedures were performed under conscious
for patients unable to adequately lie still.
All patients were anticoagulated with heparin during
the procedure to an activating clotting time of 250 to 300
seconds. After the procedure, patients were given a loading
dose of clopidogrel and maintained on a 75-mg daily dose
for at least 30 days, along with aspirin and statin therapy
Cases were performed using 5F or 6F sheaths preferen-
tially through retrograde contralateral and occasionally
through antegrade ipsilateral access. No brachial or other
upper extremity access was used. Preference was given to
using 0.014- or 0.018-inch guidewire and catheter sys-
tems. Angioplasty was performed with noncompliant low-
profile balloons. Repeated 2- to 3-minute inflations or
cutting balloon angioplasty were performed if significant
residual stenosis or flow-limiting dissection was present
after a first angioplasty attempt. Subintimal angioplasty was
performed for complete occlusions that could not be
crossed within the lumen. Stents were placed only for
flow-limiting dissections or residual stenosis ?30% after
primary angioplasty. The type of stent used depended on
is a low-profile self-expanding nitinol stent.
Measurements. We recorded patient demographics,
TASC classification, indication, and bypass candidacy sta-
tus. The primary outcome variable was freedom from reste-
nosis, reintervention, or amputation. Secondary outcomes
were technical success, procedural and postoperative com-
plications, conventional primary patency, secondary reste-
nosis, tissue healing, limb salvage, reintervention, and pa-
TASC classification was assessed for the individual ves-
sel that underwent intervention. If more than one vessel
had a successful intervention, the limb was assigned the
worst of the TASC classes for lesions in series (eg, tibio-
peroneal trunk and posterior tibial) and the lesser class for
lesions in parallel (eg, anterior tibial and posterior tibial).
The indication for the procedure was classified as tissue
loss (gangrene or nonhealing ulcer) or rest pain. A few
patients were treated for the presence of tibial outflow
stenosis in the native tibial artery distal to an existing vein
graft with low graft velocities. All bypass grafts had initially
been placed for either tissue loss or rest pain.
Patients were deemed unsuitable candidates for a by-
pass procedure if they lacked a bypass target or an adequate
vein conduit. A small number of patients had underlying
conditions precluding surgery, including severe dementia,
significant medical comorbidities, nonambulatory status,
or open wounds overlying the only potential bypass target.
Outcome variables. Technical success was defined as
a residual stenosis of ?30% as assessed on single-view
completion angiography. Adjunctive procedures included
stent placement, mechanical atherectomy, mechanical
thrombectomy, thrombolysis, or intra-arterial nitroglyc-
erin infusion to treat flow-limiting spasm. Concomitant
procedures included superficial femoral artery, popliteal, or
vein graft angioplasty or stent placement, or both.
Patients were seen for follow-up typically at 2 weeks,
then every 3 months for 1 year, and every 6 months
thereafter, or more frequently if stenoses were detected or
to monitor wound healing. Restenosis and patency were
assessed with duplex ultrasound (DUS) analysis of the
treated vessel, Doppler waveforms, segmental pressures,
pulse volume recordings, and when indicated, angiogra-
phy. All noninvasive vascular studies were performed in one
of two vascular laboratories by a dedicated vascular techni-
cian and interpreted by a staff vascular surgeon.
The primary outcome was freedom from tibial resteno-
sis, reintervention, or major amputation. Restenosis was
Table I. TransAtlantic Inter-Society Consensus
classification for infrapopliteal lesions9
Classification Lesion characteristics
Single stenosis ?1 cm long
Multiple focal stenoses ?1 cm long or 1 or
2 stenoses ?1 cm involving the
Stenoses 1 to 4 cm long, occlusion 1 to 2
cm long, or extensive stenosis involving
Occlusion ?2 cm long or diffusely diseased
TASC, TransAtlantic Inter-Society Consensus.
JOURNAL OF VASCULAR SURGERY
Volume 48, Number 1
Giles et al 129
defined as occlusion or significant stenosis as demonstrated
by peak systolic velocity ratio increase of ?3.0 on the DUS
examination. In our vascular laboratory, this value repre-
sents 60% to 80% vessel stenosis and is consistent with
reporting standards for prior endovascular series.10
Measurement of primary patency followed Society for
Vascular Surgery (SVS) reporting standards criteria for
patency to allow a more accurate comparison of this series
with bypass outcomes.11The minimum standard of docu-
mented flow by arteriography or DUS was required for a
vessel to be considered patent. Treatment failure was de-
fined as any patient who required reinterventions, whether
for restenoses or occlusions.
Freedom from secondary restenosis was determined by
the last available vascular laboratory study inclusive of pa-
tients who had undergone infrapopliteal reintervention by
endovascular as well as bypass to maintain or restore flow.
The same DUS examination threshold was used for this
Limb salvage was defined as freedom from major am-
putation (below or above knee). Toe, ray, or transmetatar-
sal amputations were considered minor amputations.
Wound healing or symptom resolution was documented as
complete, improved, stable, or worse.
Reintervention included repeat infrapopliteal PTA or
bypass graft procedure. Repeat PTA was any attempt at a
intervened on. Reintervention by bypass graft was any
bypass performed in the ipsilateral limb to an infrapopliteal
target encompassing the previously treated area. Survival
was assessed using computerized hospital medical records
and verified by referencing the Social Security Death Index
that increased length of stay or required blood transfusion,
operation, or additional therapy. Intraprocedural compli-
cations were those that occurred while in the fluoroscopy
suite or operative room and included flow-limiting spasm
or thromboembolus. No vessel ruptures occurred in this
Statistical analysis. All analyses were performed on a
per-limb basis. Preoperative characteristics and outcomes
were reported as percentages of the sample. Categoric
variables were analyzed by Pearson ?2and the Fisher exact
test. Median length of stay was compared using the Wil-
coxon rank sum test. Treatment outcomes during the
course of follow-up were analyzed using Kaplan-Meier
the log-rank test. Univariate and multivariate Cox regres-
sion models were used to assess predictor variables for
time-dependent outcomes. Primary restenosis, reinterven-
tion, or amputation was assessed both with inclusion and
exclusion of technical failures; however, the final reported
tables are representative of the entire series as an intention
to treat model. Statistical significance was defined as P ?
.05. All statistical tests were done using STATA 8 software
(StataCorp, College Station, Tex).
Demographics. Tibial PTA was used to treat 176
limbs in 163 patients. Demographics are summarized in
Table II. Median age was 73 years (range, 39-94 years).
Most patients were men with hypertension, diabetes, hy-
perlipidemia, and coronary artery disease. Half of the pa-
tients had a history of smoking. Of the 163 patients, 47
(27%) had undergone a prior infrainguinal bypass in the
limb intervened on, and 40 (23%) were considered unsuit-
able candidates for bypass. The lesions that underwent
intervention were evenly distributed among the TASC
Indication for intervention. Most patients were
treated for tissue loss, and a lesser number were treated for
More TASC A lesions (8 of 12) were performed for vein
graft outflow stenosis than other TASC classifications.
Procedural details. Stents were placed in infrapopli-
teal vessels in 8% of limbs. Four were placed for flow-
Table II. Demographic data of patients undergoing
Variables No. or median% or range
Coronary artery disease
Dialysis-dependant renal failure
Creatinine ?2.0 mg/dL
Prior myocardial infarction
Congestive heart failure
Prior infrapopliteal bypass
Not bypass candidate
No bypass target
No bypass conduit
Wound over potential targets
COPD, Chronic obstructive pulmonary disease; TASC, TransAtlantic
aMissing values excluded from denominator.
JOURNAL OF VASCULAR SURGERY
130 Giles et al
limiting dissections, and the 11 remaining stents were
placed for residual stenosis. The stents included six self-
expanding stents, six stainless steel balloon expandable
stents, and three drug-eluting balloon expandable stents.
No significant differences existed among the demographics
of patients receiving stents or for outcomes of restenosis,
reintervention, or amputation after stenting vs angioplasty
Antegrade access was used in 13% of cases. The primary
method of intervention was PTA; however, three arthrec-
tomies were performed. Two patients underwent 24-hour
thrombolysis of an infrapopliteal vessel, followed by PTA
after a stenotic lesion was discovered. A total of 102 pa-
tients (58%) had concomitant femoropopliteal angioplasty
or stenting, or both. No significant differences were found
between groups undergoing single vs multilevel interven-
tions among procedural indications, prior distal bypass
grafts, technical success, or restenosis.
Technical success. Technical success was obtained in
163 of 176 limbs (93%). TASC class predicted technical
success, with the only technical failures occurring in TASC
D lesions (100% for TASC A to C vs 75% for TASC D; P ?
.0001). Of the 13 technical failures, eight had secondary
interventions, one had immediate amputation, and four
had no further interventions. Of the eight secondary inter-
ventions, five underwent immediate bypass and four had
tibial PTA (1 after bypass failure). Amputation was ulti-
mately required in one of the bypass patients and in two of
the repeat tibial PTA patients. The remaining reinterven-
tions remained patent at last follow-up. Of the four patients
in whom no further procedure was performed, one died at
home ?1 month and two continue to receive wound care
for stable wounds. The remaining patient had a concurrent
angioplasty in the superficial femoral and popliteal arteries
and a toe amputation healed. Technical outcome was inde-
pendent of stent placement, multilevel interventions, and
Intraprocedural complications. Procedural compli-
perforation (n ? 1), and thromboembolus (n ? 8). The
difference in the rate of complications among TASC classes
was not significant (P ? .284). In addition, no significant
difference was found in the number of intraprocedural
complications based on the preoperative use of aspirin (P ?
.602) or clopidogrel (P ? .282). All complications were
successfully treated using repeated PTA alone in 4, intra-
arterial vasodilators in 6, or rheolytic thrombectomy in 8,
or thrombolysis in 9, or both. No patient required emer-
gency surgery, and no potential infrapopliteal bypass tar-
gets were compromised.
Perioperative mortality and postoperative compli-
cations. Five patients died for an in-hospital mortality of
3%, and four other patients died after discharge ?30 days
from their procedure, for a 30-day mortality of 5%. Two
an uncontrolled retroperitoneal hemorrhage. The remain-
ing in-hospital deaths occurred ?3 weeks in patients who
to septic complications, and one patient sustained a myo-
cardial infarction with acute renal failure. The four out-of-
hospital perioperative deaths were from unknown causes.
The postoperative complication rate was 9%. Two
pseudoaneurysms were treated with thrombin injection.
Five hematomas occurred, one retroperitoneal, one rectus,
and three groin (two were surgically evacuated). Conges-
tive heart failure complicated the course of two patients,
and myocardial infarctions occurred in two others. Two
patients had temporary dysrhythmia, and transient contrast
nephropathy developed in four patients.
Length of stay. The median total length of stay was 4
days (range, 0-41 days), whereas the postprocedural length
of stay was 2 days (range, 0-38 days). The median length of
stay was longer for those intervened on for tissue loss
compared with other indications (5 vs 2 days; P ? .0005).
Patients with postoperative complications also had median
longer hospitalizations than those without complications
(8 vs 3 days; P ? .005). Discharge destination was home in
75% and to a rehab facility in 25%.
Freedom from restenosis, reintervention, or ampu-
tation. Mean follow-up was 10 months (range, 1-41
months). Freedom from restenosis, reintervention, or am-
putation was 39% at 1 year and 36% at 2 years (Fig 1, A).
Freedom from restenosis, reintervention, or amputation at
respectively (P ? .0001; Fig 1, B). TASC D lesions had a
higher restenosis, reintervention, or amputation rate than
all others (Fig 1, C), even when technical failures are
excluded (P ? .005). Because all technical failures were
TASC D, this confirms the significant effect of TASC D
regardless of the initial technical success. Freedom from
restenosis, reintervention, or amputation at 1 and 2 years
for bypass graft candidates (46%, 41%) was higher than for
unsuitable bypass candidates (19%, 19%).
TASC D lesions predicted restenosis, reintervention, or
amputation whether technical failures were included or
excluded: including failures had a hazard ratio (HR) of 3.5
(95% CI, 2.2-5.6, P ? .001) and excluding failures had a
HR of 2.3 (95% CI, 1.3-4.1, P ? .005). Univariate predic-
tors of restenosis, reintervention, or amputation were
TASC D classification and patients who were not candi-
dates for bypass (Table III). The strongest predictor within
the bypass candidacy group was a lack of a bypass target
vessel (HR, 2.9; 95% CI, 1.6-5.5, P ? .001). Patients
intervened on for graft outflow stenosis had a lower risk.
Age, sex, other comorbidities, procedure indication, and
multilevel disease were not significant factors. On multivar-
remained significant predictors.
A subset analysis excluded all procedures in which
multilevel interventions were performed to ensure that this
was not a significant factor influencing restenosis. In the 74
patients who underwent intervention for isolated infrapop-
liteal lesions, freedom from restenosis, reintervention, or
amputation at 1 and 2 years was 37%, with a significant
JOURNAL OF VASCULAR SURGERY
Volume 48, Number 1
Giles et al 131
difference among TASC classes (P ? .007). This was not
significantly different than that of the entire series (P ?
.538). TASC D class and lack of bypass target similarly
tivariate analysis (TASC D: HR, 2.5 [95% CI, 1.2-5.2, P ?
.015]; no target: HR, 3.8 [95% CI, 1.4-10.3; P ? .007]).
Primary patency. Primary patency was 53% and 51%
at 1 and 2 years (Fig 2, A). Patency at 1 year for TASC A
through D was 53%, 58%, 67%, and 37%, respectively (P ?
.001). Multivariate predictors of loss of primary patency
were TASC D classification and unsuitable bypass candi-
dacy (Table IV).
Secondary restenosis. Freedom from secondary re-
stenosis was 63% at 1 year and 61% at 2 and 3 years, with a
significant difference among TASC classification as well
(P ? .001; Fig 3). Freedom from secondary restenosis for
TASC A through D at 1 year was 79%, 82%, 57%, and 43%,
respectively. On univariate analysis, secondary restenosis
was predicted by TASC D classification, whereas TASC A
and B lesions resulted in lower rates. On multivariate
analysis, TASC D and lack of bypass target predicted sec-
tive (Table V).
Limb salvage and wound healing. Mean follow-up
assessing wound healing was 12 months (range, 1-43
months). At the last follow-up for those with an indication
of tissue loss, wounds were completely healed or improved
Fig 1. A, The primary freedom from restenosis, reintervention,
or amputation at 1 year rises to 42% when technical failures are
excluded. The standard error (SE) remains ?10% throughout. B,
Freedom from restenosis, reintervention, or amputation by Trans-
Atlantic InterSociety Consensus (TASC) classification. A carat
indicates where SE ?10%. The SE remains ?10% throughout for
the TASC D curve. C, Freedom from restenosis, reintervention, or
amputation by TASC D, technical success only. The SE remains
?10% throughout for both curves.
Table III. Predictors of restenosis, reintervention, or
amputation after infrapopliteal angioplasty
PredictorsHR 95% CI
Coronary artery disease
Dialysis-dependant renal failure
Creatinine ?2.0 mg/dL
Prior distal bypass
Not bypass candidate—all
No bypass target
No bypass conduit
Graft outflow stenosis
No bypass target
JOURNAL OF VASCULAR SURGERY
132 Giles et al