complications stroke

1. Transcatheter Intraarterial Infusion of rt-PA for
Acute Lower Limb Ischemia: Results and
Complications
James L. Swischuk, MD, Paul F. Fox, MD, Kate Young, RT, Syed Hussain, MD, Bob Smouse, MD,
Flavio Castañeda, MD, and Terrence M. Brady, MD
PURPOSE: To determine the success and complication rates of intraarterial recombinant tissue-type plasminogen
activator (rt-PA) infusion for the treatment of acute lower extremity artery and bypass graft occlusions.
MATERIALS AND METHODS: The results of 74 limbs in 70 patients (mean age, 66 y) treated with catheter-directed
rt-PA infusion for the treatment of acute lower extremity ischemia were retrospectively evaluated. The group included
42 bypass grafts and 32 native arteries. All limbs were viable at presentation. The mean duration of symptoms was 11.9
days. rt-PA was infused for a mean of 27.9 hours for a mean total dose of 38.7 mg. Initial infusion rates of 3–6 mg/h
were lowered to a preferred rate of 1.5 mg/h. Thrombolytic success was defined as 95% thrombolysis of an occluded
segment with return of antegrade flow. Major bleeding complications were defined as any hemorrhagic event leading
to surgery, extended or unexpected hospitalization, transfusion, death, intracranial hemorrhage, or a decrease in
hemoglobin of 5 g/dL or in hematocrit of 15%. Thirty-day mortality and amputation rates were calculated. Patient
characteristics and infusion parameters were evaluated as to whether they contributed to thrombolytic success or
major bleeding events.
RESULTS: Thrombolytic success was achieved in 64 limbs (86%). Major bleeding complications occurred in 33 (47%)
patients. In 22 of these patients, bleeding occurred at a vascular puncture site, whereas remote bleeding occurred in
seven patients. Remote bleeding complications included two retroperitoneal hematomas, two rectus sheath hemato-
mas, one lower gastrointestinal hemorrhage, one episode of hemoptysis, and one dehiscence of a femoral-popliteal
bypass graft revision. No parameters were found to be predictive of thrombolytic success, whereas a negative history
of smoking, increasing duration of infusion, and a low preprocedural ankle-brachial index (ABI) were found to be
associated with major hemorrhagic events. Four patients (6%) underwent amputation and one patient (1%) died,
resulting in a 30-day amputation-free survival rate of 93%.
CONCLUSION: Catheter-directed rt-PA infusion is effective in achieving thrombolysis. Despite a significant number
of bleeding complications, 30-day mortality and amputation rates were favorable. Nonetheless, complication rates
related to bleeding were not trivial and further evaluation with use of variable dosing regimens is indicated.
Index terms: Arteries, peripheral • Extremities, blood supply • Thrombolysis • Tissue-type plasminogen activator
J Vasc Interv Radiol 2001; 12:423–430
Abbreviations: ABI ϭ ankle-brachial index, rt-PA ϭ recombinant tissue-type plasminogen activator, SK ϭ streptokinase, TIMI ϭ Thrombolysis in Myocardial
Ischemia (scale), UK ϭ urokinase
DURING the past 15 years, urokinase
(UK) has been the preferred thrombo-
lytic agent because it combined effi-
cacy with reasonable complication
rates. In late 1998, the Food and Drug
Administration called for the with-
drawal of UK from the market,
thereby creating a significant void in
the treatment regimen for acute arte-
rial ischemia. Currently available
thrombolytic alternatives include
streptokinase (SK), anistreplase, rete-
plase, alteplase, and, most recently,
tenecteplase. Although SK is the only
thrombolytic agent with Food and
Drug Administration approval for the
treatment of acute peripheral arterial
occlusions, it is not considered a viable
alternative because of its low efficacy
and poor safety profile (1). Many have
turned to recombinant tissue-type
plasminogen activator (rt-PA) as the
thrombolytic alternative of choice pri-
marily because of its familiar role in
From the Departments of Radiology (J.L.S., P.F.F.,
K.Y., B.S, F.C., T.M.B.) and Surgery (S.H.), Univer-
sity of Illinois College of Medicine at Peoria, 1 Illini
Dr., Peoria, IL 61605. Received August 8, 2000; revi-
sion requested October 3; revision received Decem-
ber 5; accepted December 10. Address correspon-
dence to J.L.S.; E-mail: jswischuk@peorad.com
© SCVIR, 2001
423

2. the thrombolytic management of acute
myocardial infarction and because
well-controlled trials such as the Study
of Thrombolysis for the Ischemic
Lower Extremity (STILE) trial suggest
that UK and rt-PA are equivalent in
efficacy and safety (2).
Alteplase (rt-PA; Activase; Genen-
tech, South San Francisco, CA) is cur-
rently approved for the intravenous
treatment of acute myocardial infarc-
tion, acute ischemic stroke, and pul-
monary embolism, and, like UK, the
intraarterial infusion of rt-PA repre-
sents an off-label use of the drug (3).
When compared to UK, rt-PA has a
shorter half-life (5 minutes [3] versus
20 minutes [4]) and increased specific-
ity and affinity for fibrin (5,6). Earlier,
these differences led some to speculate
that rt-PA may have less systemic lytic
effect, and, therefore, the use of rt-PA
may lead to fewer bleeding complica-
tions than UK (7,8). However, as with
UK, bleeding is clearly a frequent com-
plication associated with catheter-di-
rected rt-PA infusion (9–13). Consid-
ering this, the purpose of this study is
to evaluate our initial use of rt-PA for
catheter-directed thrombolysis of
acute arterial occlusions of the lower
limbs with special interest in the num-
ber and severity of bleeding events.
MATERIALS AND METHODS
From January 1999 to November
1999, the medical records of patients
treated with intraarterial thrombolytic
therapy for angiographically proven
acute native arterial or bypass graft
occlusions of the lower limbs were ret-
rospectively analyzed. This represents
our first 10 months experience of use
of rt-PA and includes 74 lower limbs
in 70 patients. The mean patient age
was 66.4 years (range, 40–100 y). The
mean duration of symptoms was 11.9
days (range, 1–90 d) with 68 patients
(97%) having symptoms for a duration
of 30 days or shorter. Occlusion and
patient characteristics are summarized
in Table 1. Most occlusions occurred
in bypass grafts (57%) and were
caused by thrombosis (84%) rather
than embolic events (16%). A prepro-
cedural ankle-brachial index (ABI)
was available in 46 of 74 limbs with a
mean value of 0.46, whereas a postpro-
cedural ABI, available in 37 limbs, had
a mean value of 0.86. For the 37 pa-
tients with both pre- and postproce-
dural ABIs, the mean increase in ABI
was 0.41. The mean total dose of rt-PA
infused was 38.7 mg (range, 6–120
mg) with a mean duration of infusion
of 27.9 hours (range, 2–69 h). The
mean length of hospital stay was 9.13
days.
Before beginning therapy, patients
underwent clinical assessment of the
ischemic limb. All patients had a his-
tory of sudden onset of ischemic
symptoms for a duration of less than
90 days with 97% of patients having
symptoms for a duration of 30 days or
less. All patients were categorized
with use of the Society of Vascular
Surgery/International Society of Car-
diovascular Surgery (SVS/ISCVS)
acute ischemia criteria (14). Excluded
were patients with a history of cere-
brovascular accident within 3 months,
gastrointestinal hemorrhage within 14
days, surgery within 14 days, severe
uncontrolled hypertension, or irre-
versible ischemia. The presence of co-
morbid conditions such as diabetes
mellitus, hypertension, coronary ar-
tery disease, cerebrovascular occlusive
disease, or chronic obstructive pulmo-
nary disease was documented.
An initial diagnostic arteriogram
was obtained to confirm the presence
of an occlusion and to document distal
runoff. If the limb was not considered
in immediate jeopardy (ie, SVS/ISCVS
category III) the decision was made to
proceed with thrombolytic therapy as
opposed to surgery. The occlusion was
believed to be embolic in nature if a
discrete filling defect was seen at the
proximal extent of the occlusion or if
occlusions were seen in multiple vas-
cular beds. After diagnostic angiogra-
phy, a 5-F catheter was advanced to
the proximal aspect of the occlusion. A
guide wire was then passed through
Table 1
Patient and Occlusion Characteristics
Characteristic n (%)
Mean age (y) 66.4 (range, 40–100)
Male sex 35 (50)
Comorbidities
Smoking 48 (69)
Coronary artery disease 23 (33)
CHF 5 (7)
Arrhythmias 10 (14)
Hypertension 42 (60)
Hypercholesterolemia 27 (37)
Diabetes 18 (26)
TIA 3 (4)
CVA 10 (14)
COPD 12 (17)
Native artery 32 (43)
Bypass graft 42 (57)
Synthetic 32 (43)
Vein 2 (3)
Composite 8 (11)
Thrombus 62 (84)
Embolus 12 (16)
Duration of symptoms (d) 11.94 (range, 0.04–90 d)
SVS/ISCVS class
I 41 (58.6)
II a 16 (22.9)
II b 13 (18.6)
Site of occlusion
Aortoiliac 13 (18)
Femoropopliteal 37 (50)
Tibial 9 (12)
Multilevel 15 (20)
Note.—CHF ϭ congestive heart failure; TIA ϭ transient ischemic attack; CVA ϭ
cerebrovascular accident; COPD ϭ chronic obstructive pulmonary disease;
SVS/ISCVS ϭ Society for Vascular Surgery/International Society for Cardiovascular
Surgery.
424 • Transcatheter Intraarterial Infusion of rt-PA for Ischemia April 2001 JVIR

3. the occlusion and the infusion catheter
was advanced over the wire to the
distal extent of the thrombus. In three
patients, the clot could not be pene-
trated initially, and the infusion was
initiated with an end-hole catheter po-
sitioned just proximal to the occlusion.
Thrombolytic progression was as-
sessed by serial angiography. In gen-
eral, patients whose initial infusion
was started in the morning were
checked in the afternoon, and patients
whose infusion started in the after-
noon were checked the following
morning. Infusion catheters were re-
positioned to maintain proximity to
any remaining thrombus.
Single catheter and coaxial infusion
techniques were used. Because of our
limited experience with rt-PA and the
disparity in the doses reported in the
literature, our initial infusion rates for
the first 10 patients were between 3 and
6 mg/h. The dose was then lowered to a
preferred rate of 1.5 mg/h in all subse-
quent patients because of a perceived
increase in bleeding complications
when compared to our experience with
UK. The thrombolytic infusion was pre-
pared by adding 50 mg of rt-PA recon-
stituted in 50 mL of sterile water into a
total volume of 1 liter of normal saline.
This creates a final concentration of 0.05
mg/mL. Although no information is
available regarding the biochemical sta-
bility of rt-PA at concentrations this low,
we found that this concentration main-
tained clinical efficacy without visual
evidence of drug precipitation. This di-
lution factor was chosen to maintain in-
fusion rates similar to those used during
previous UK infusion protocols, thereby
reducing the potential for dosing errors
during this transition period. Intrave-
nous heparin was administered on a
sliding scale basis to maintain a partial
thromboplastin time between 40 and 60
seconds. The decision to continue anti-
coagulation therapy after the comple-
tion of thrombolysis was made at the
discretion of the attending physician.
The thrombolytic infusion was
maintained until one of the following
occurred: complete (Ͼ95%) lysis, fail-
ure of progression of lysis on serial
arteriograms, or a hemorrhagic
complication. Any underlying graft or
native artery lesion identified as the
cause of the occlusion was promptly
treated by either percutaneous or sur-
gical techniques, if possible. All pa-
tients were monitored in an intensive
care unit during rt-PA infusion. He-
moglobin, hematocrit, platelet count,
prothrombin time, partial thrombo-
plastin time, and fibrinogen levels
were obtained routinely every 6 hours
during the infusion and 24 hours after
the completion of treatment. If the fi-
brinogen level decreased to less than
100 mg/dL, the infusion rate was re-
duced by half. The thrombolytic infu-
sion was stopped only if bleeding
complications occurred. Patients were
followed during hospitalization and
for 30 days after the completion of the
thrombolytic treatment.
The study cohort was evaluated for
thrombolytic success and clinical
success. Also, 30-day amputation,
mortality, and amputation-free sur-
vival rates were calculated. Thrombo-
lytic success was defined as greater
than 95% resolution of the thrombus
with some degree of antegrade flow
(Thrombolysis in Myocardial Ischemia
[TIMI] perfusion grade II or III),
whereas clinical success was defined
as a return to the preischemic state.
Major bleeding complications were
defined as any hemorrhagic event
leading to surgery, extended or unex-
pected hospitalization, transfusion, or
any event satisfying one of the modi-
fied TIMI criteria described elsewhere
(15). The latter includes any hemor-
rhagic event terminating in death, in-
tracranial hemorrhage, or a decrease
in hemoglobin of 5 g/dL or in hemat-
ocrit of 15%, regardless of whether
there was obvious bleeding. Adjunc-
tive surgical or endovascular proce-
dures occurring in the 30-day fol-
low-up period were also recorded.
Multiple risk factors, patient char-
acteristics, and infusion parameters
were evaluated as to whether they
contributed to thrombolytic success or
major bleeding events. Discrete vari-
ables were evaluated with use of the ␹2
test, whereas normally distributed
continuous variables were evaluated
with use of the Student t test. A P
value of less than 0.05 was considered
statistically significant.
RESULTS
Thrombolytic success was achieved
in 64 (86%) of the treated limbs. All 10
thrombolytic failures occurred in sepa-
rate patients. Of the thrombolytic fail-
ures, two patients progressed to ampu-
tation, two patients required a new
femoral-popliteal bypass graft, one pa-
tient was successfully treated with stent
placement, and five patients were left
with persistent non–limb-threatening
ischemia. In the three patients in whom
initial infusions could not be started
within the occlusion, two were treated
successfully and one patient’s treatment
ended as a thrombolytic failure. Clinical
success was achieved in 57 (81%) treated
patients (61 limbs). The 13 clinical fail-
ures included four amputations, three
femoral-popliteal bypass revisions, one
repeat thrombolysis leading to persis-
tent ischemia, and five cases of contin-
ued ischemia. During the 30-day fol-
low-up period, four (6%) patients
underwent amputation, and one patient
(1%) died. The single death was caused
by irreversible bowel ischemia, which
occurred 19 days after the conclusion of
thrombolytic therapy. Overall, the 30-
day amputation-free survival rate was
93%.
Subsequent endovascular and surgi-
cal interventions are summarized in Ta-
ble 2. All endovascular procedures were
performed immediately after the con-
clusion of thrombolytic therapy. A total
of 51 endovascular procedures were
performed in 50 patients, with simple
balloon angioplasty being the most
common endovascular procedure. A to-
tal of 18 surgical procedures were per-
formed in 18 patients; however, only 10
of these were performed to restore
blood flow to the affected limb. These
included one surgical thrombectomy,
Table 2
Endovascular and Surgical Procedures
Performed after Thrombolysis
Procedure n (%)
Endovascular procedures 50
Stent placement 12 (23.5)
Thrombolytic infusion 2 (3.9)
Angioplasty 37 (72.5)
Surgical procedures 18
Surgical thrombectomy 1 (5.5)
New bypass graft 5 (27.8)
Revision bypass graft 3 (16.7)
Amputation 4 (22.2)
Above knee 1 (5.5)
Below knee 2 (11.1)
Trans-metatarsal 1 (5.5)
Evacuation of hematoma 1 (5.5)
Surgical repair of puncture
site
2 (11.1)
Fasciotomy 1 (5.5)
Pseudoaneurysm repair 1 (5.5)
Swischuk et al • 425Volume 12 Number 4

4. five new bypass grafts, three revisions
of previously placed bypass grafts, and
one fasciotomy.
Major hemorrhagic complications
occurred in 33 (47%) patients in the
30-day follow-up period. In 22 pa-
tients, bleeding occurred at arterial or,
less frequently, venous puncture sites.
Remote bleeding was seen in seven
patients and included two retroperito-
neal hematomas, two rectus sheath he-
matomas, one lower gastrointestinal
hemorrhage, one episode of hemopty-
sis, and one dehiscence of the distal
anastomosis of a femoral-popliteal by-
pass graft revision. In four patients, no
bleeding site was identified. Of the 33
patients with a major bleeding compli-
cation, 14 qualified only by virtue of a
15% drop in their hematocrit level. A
total of 15 patients (21%) required a
blood transfusion. There were no in-
tracranial hemorrhages. Of the 33 pa-
tients with a major bleeding complica-
tion, only four required surgical
treatment. These surgical events in-
cluded the revision of a femoral-pop-
liteal bypass graft in one patient and
surgical repair of the arterial puncture
site in three patients. In the initial 10
patients treated with the higher dosing
regimen, three had a major bleeding
event prompting the change to a lower
dose protocol. These bleeding compli-
cations included one large retroperito-
neal hemorrhage and two large punc-
ture site hematomas. All three of these
patients received blood transfusions.
Statistical analysis was used to
evaluate whether or not several limb
and patient parameters were related to
thrombolytic success (Table 3) and
major hemorrhagic complications (Ta-
ble 4). No parameters were found to
be predictive of thrombolytic success.
Conversely, a negative history of
smoking, increasing duration of infu-
sion, and a low preprocedural ABI
were found to be associated with the
occurrence of major hemorrhagic
events.
DISCUSSION
Thrombolytic success was achieved
in 86% of treated limbs, whereas 30-
day clinical success was achieved in
81% of treated patients. This compares
favorably with multiple large series in
which reports of thrombolytic and
clinical success with use of rt-PA
range from 45% to 98% and from 48%
to 90%, respectively (16–24). In addi-
tion, this compares favorably with
thrombolytic success rates reported in
well-controlled trials with UK (2,25–
27). The wide range of success rates
reported is most likely a result of the
varied definitions of success used
throughout the literature. In fact, the
British Thrombolysis Study Group, as
reported by Braithwaite et al (24), has
addressed this very issue. They pro-
vide several definitions of lysis, essen-
tially to distinguish angiographic and
clinical outcomes. Unfortunately, this
is a recent report, and few studies to
date adhere to this reporting standard.
Unlike thrombolytic or clinical suc-
cess rates, the 30-day amputation-free
survival rate, as a measure of success,
relies on the absence of two very dis-
crete events, the amputation of the af-
fected limb and the death of the
patient. This is not to say that all pa-
tients studied who are alive with all
limbs enjoy either thrombolytic or
clinical success; however, the precise
nature of this outcome makes for an
unequivocal benchmark that can be
used for ready comparison among
studies. Our 30-day amputation-free
survival rate of 93% compares favor-
ably to results of several larger series
in which rt-PA was used (11,18–24)
and multiple well-controlled studies
of UK (2,25–27), in which short-term
follow-up amputation-free survival
rates range from 56% to 99% and from
66% to 93%, respectively.
Of greatest interest and concern to
us was whether we would experience
an increasing number and severity of
bleeding complications during this
abrupt transition to an unfamiliar
thrombolytic agent. The first reports
on the use of intraarterial rt-PA oc-
curred in the late 1980s, coincident to
the approval of the drug by the Food
and Drug Administration in 1987 as a
thrombolytic agent for the treatment
of acute myocardial infarction. When
compared to UK, the increased speci-
ficity and affinity of rt-PA for fibrin-
bound plasminogen and a shorter
half-life caused speculation that the
drug may have less systemic lytic ef-
fects, thereby decreasing the chance
for bleeding complications (7,8). How-
ever, with time, it has become appar-
Table 3
Parameters Related to Thrombolytic Success
Parameter
Successful
P ValueYes No
Limb Parameters (n ϭ 64) (n ϭ 10)
Thrombus vs embolus
Thrombus 54 7 NS
Embolus 10 3 NS
Native artery vs graft
Native artery 27 4 NS
Graft 37 6 NS
Runoff
Single vessel 11 4 NS
Multiple vessel 53 6 NS
Acute SVS class
I 37 6 NS
II a & II b 27 4 NS
Patient Parameters (n ϭ 61) (n ϭ 9)
CAD 19 4 NS
DM 14 5 NS
Smoking history 41 7 NS
CVA 9 1 NS
Pulmonary disease 10 3 NS
Infusion duration (h) 27.5 30.5 NS
Total thrombolytic dose (mg) 38.4 41.0 NS
Duration of symptoms (d) 11.6 11.56 NS
Preprocedure ABI 0.45 0.51 NS
Fibrinogen nadir (mg/dL) 259 390 NS
Note.—SVS ϭ Society for Vascular Surgery; CAD ϭ coronary artery disease; DM ϭ
diabetes mellitus; CVA ϭ cerebrovascular accident; NS ϭ not significant.
426 • Transcatheter Intraarterial Infusion of rt-PA for Ischemia April 2001 JVIR

5. ent that, as with UK, bleeding is a
frequent complicating event during
intraarterial rt-PA infusions. Earlier,
the explanation for this was simple.
The effects of rt-PA are indiscriminate,
and with more prolonged infusion, ly-
sis occurs not only at the treatment site
but also at any acute distant hemo-
static thrombus. Although partially
valid, this explanation probably repre-
sents an oversimplification of the lytic
chain of events that occurs during in-
traarterial rt-PA infusion.
The specificity and affinity of rt-PA
for fibrin account for the several-hun-
dred-fold increase in the rt-PA-medi-
ated activation of plasminogen to plas-
min when the former is also bound to
fibrin (5,6). Unfortunately, the affinity
of rt-PA is not limited to fibrin alone.
In vitro data indicate that the soluble
fibrin degradation products produced
during thrombolysis likewise promote
the rt-PA-mediated conversion of
plasminogen to plasmin (28,29). In
fact, fragment (DD)E, a prevalent fi-
brin degradation product produced
during lysis, has been shown to be just
as effective at stimulating rt-PA-medi-
ated plasminogen activation as fibrin
(30). Given the soluble nature of these
products, plasminogen activation can
then occur systemically, during which
freely circulating fibrinogen is con-
sumed (ie, fibrinogenolysis). Systemic
fibrinogenolysis also leads to the pro-
duction of additional degradation
products, most importantly, factor X,
which, when incorporated into evolv-
ing thrombus, further facilitates
thrombolysis (31). Clinical and exper-
imental studies validate that fibrino-
genolysis, and not simply the activa-
tion of fibrin-bound plasminogen at a
distant hemostatic plug, is associated
with and primarily responsible for the
distant bleeding seen with rt-PA
(2,31,32).
In addition, the strong affinity of
rt-PA for fibrin promotes the formation
of plasmin within thrombus. In the clot
matrix, fibrin prevents the inactivation
of plasmin by alpha 2-antiplasmin, the
primary inhibitor of plasmin (5). There-
fore, rt-PA, by virtue of its strong affin-
ity for fibrin, activates plasminogen in
an environment that lends a relatively
durable thrombolytic effect to the result-
ant plasmin. Some have suggested that
this accounts for the clinical and exper-
imental evidence showing that, unlike
indiscriminate plasminogen activators
such as UK and SK, rt-PA causes a more
delayed and prolonged systemic lytic
effect, a finding somewhat paradoxic
considering the relatively short half-life
of rt-PA (33–36). In the end, although
more specific than UK and SK, rt-PA is
not specific enough to avoid systemic
fibrinogenolysis, and the increased af-
finity of rt-PA for fibrin that investiga-
tors once hoped would be protective ac-
tually prolongs the activity of rt-PA,
thereby overcoming any protective ben-
efits of a shorter half-life.
Bleeding proved to be the most
common major complication encoun-
tered in our study population, with a
major bleeding event occurring in 47%
of patients. Although the first 10 pa-
tients were treated at an arguably ex-
cessive dose rate of 3–6 mg/h, this
group did not include a disproportion-
ately high number of major bleeding
complications. The three major bleed-
ing events occurring in this group rep-
resent 9% of the major bleeding com-
plications seen in 14% of the total
study population. Factors that were
found to be statistically significant
contributors to major bleeding events
were an increased duration of infu-
sion, a low preprocedural ABI, and a
negative history of smoking. Other
large series investigating the use of
intraarterial rt-PA (11,13,16–24) and
UK (2,25–27) report major bleeding
rates ranging from 0% to 46% and 2%
to 23%, respectively. As to why our
major bleeding rate was so high, a few
variables must be considered.
First and foremost, we used a rela-
tively broad definition for a major
bleeding complication. Throughout
the literature, varying rates of hemor-
rhage are reported. This is, in part, the
result of equally varying definitions of
major hemorrhage. As an example, if
we look at our own data and simply
alter our definition of a major bleeding
complication so that those patients
who qualified only as a result of a 15%
decrease in hematocrit (a component
of the TIMI criteria for major hemor-
rhage) no longer qualify, our major
hemorrhagic complication rate drops
dramatically from 47% to 27%. This is
not to say that these patients did not
experience a hemorrhagic event, be-
cause most of these patients did have
documentation of some degree of
bleeding; however, the magnitude of
Table 4
Parameters Related to Major Bleeding Complications
Parameter
Major Bleeding
P ValueYes (n ϭ 33) No (n ϭ 37)
Thrombus vs embolus
Thrombus 27 31 NS
Embolus 6 6 NS
Native artery vs graft
Native artery 12 18 NS
Graft 21 19 NS
Runoff
Single vessel 8 7 NS
Multiple vessel 25 30 NS
Acute SVS Class
I 20 21 NS
II a & II b 13 16 NS
CAD 11 12 NS
DM 8 10 NS
Smoking history 18 30 .02
CVA 3 7 NS
Pulmonary disease 7 5 NS
Infusion duration (h) 32.3 23.9 .02
Total thrombolytic dose (mg) 42.4 35.5 NS
Duration of symptoms (d) 8.4 14.4 NS
Preprocedure ABI 0.37 0.53 .02
Fibrinogen nadir (mg/dL) 266 287 NS
Note.—SVS ϭ Society for Vascular Surgery; CAD ϭ coronary artery disease; DM ϭ
diabetes mellitus; CVA ϭ cerebrovascular accident; ABI ϭ ankle-brachial index; NS
ϭ not significant.
Swischuk et al • 427Volume 12 Number 4

6. bleeding that occurred in this subset of
patients would be considered insignif-
icant by most. In addition, when using
this alternate definition for major
bleeding complications, a low prepro-
cedural ABI remained the only statis-
tically significant predictor of a major
bleeding event.
Second, all patients received hepa-
rin with use of a sliding scale protocol
to ensure therapeutic levels of
anticoagulation. Although there is no
clear consensus on the efficacy or
safety of concomitant anticoagulation
during thrombolytic treatment, sev-
eral studies have cited systemic anti-
coagulation as a contributing factor to
major bleeding events (27,37,38).
Third, as mentioned previously, in-
fusion rates were lowered from 3–6
mg/h to 1.5 mg/h because of a per-
ceived increase in bleeding complica-
tions when compared to our experi-
ence with UK. Although no formal
analysis was performed to determine
whether this rate decrease resulted in
longer infusions, clinical evidence
shows that infusion rates and infusion
times are inversely proportional
(11,24). Considering this and the re-
sults of our own data that show a pos-
itive correlation between an increasing
length of infusion and bleeding, the
attempts to thwart bleeding by lower-
ing our infusion rate may have actu-
ally promoted bleeding by lengthen-
ing infusions.
As to whether there is merit in this
argument, experimental evidence
shows that longer infusions at lower
doses are indeed associated with an
increase in fibrinogenolysis when
compared to shorter, higher-dose infu-
sions (35). In turn, fibrinogenolysis can
then be expected to lead to increased
bleeding (31). Unfortunately, the clin-
ical data available are less convincing.
Several large series that document
both the duration of infusion and ma-
jor bleeding complications are sum-
marized in Table 5 (1,8,11,16,21–
24,39,40). In support is the study by
Hess et al (22) that describes a large
population of 288 patients treated at
relatively high dose rates ranging
from 2.5 to 10 mg/h for a mean dura-
tion of 1.3 hours. With use of this pro-
tocol, good clinical success was
achieved with an almost trivial major
bleeding rate of 0.3%. To the contrary
are the reports by Ward et al (11) and
Braithwaite et al (24). Both studies
compare higher front-loaded dosing
regimens to continuous low-dose pro-
tocols and show that the continuous
low-dose protocol leads to signifi-
cantly longer infusions. However, the
results of Braithwaite et al (24) show
no significant difference in bleeding
complications between the two
groups, whereas the results of Ward et
al (11) actually show a significantly
higher bleeding rate in the group re-
ceiving the higher doses of thrombo-
lytic drugs.
The presence of a lower preproce-
dural ABI was also found to be a sig-
nificant predictor of bleeding. These
patients obviously have more pro-
found preprocedural ischemia, which
is often associated with more complex
disease. As a result, these patients will
often require longer thrombolytic in-
fusions with larger total drug doses,
which can reasonably be expected to
lead to increased bleeding.
The surprising result that significantly
fewer bleeding complications were seen
in patients with a history of smoking is
difficult to explain. Perhaps the explana-
tion is linked to the fact that smokers often
have proximal disease involving larger
arteries. The location of acute thrombosis
in these patients will often parallel this
anatomic distribution. This was certainly
the case in our study population, in which
33% of those with a history of smoking
had occlusions involving the iliac vessels
whereas only 14% of those without a his-
tory of smoking had iliac involvement. As
with any endovascular or surgical proce-
dure, proximal disease is easier to treat
and is associated with better technical and
clinical outcomes. As a result, these pa-
tients may require more abbreviated
thrombolytic infusions leading to fewer
bleeding complications.
In general, our experience is like
that of many throughout the country.
The clinical demands brought about
by the widespread acceptance of
thrombolysis over surgery hastened
the transition from UK to new throm-
bolytic agents. Typically, an appropri-
ate dose of a new thrombolytic agent
is selected and simply introduced into
protocols that are designed for the rel-
atively lengthy infusion times that
most are accustomed to with use of
UK. This approach does not appropri-
ately take into account the biochemical
differences that exist between UK and
other thrombolytic agents. For rt-PA,
these differences include more rapid
thrombolysis (2,13,23) with delayed
and prolonged fibrinogenolysis (35).
Given these differences, it is reason-
able to expect that the lengthy infusion
protocols designed for UK may create
the very environment that enhances
the potential for bleeding when rt-PA
is used.
In an effort to reduce the incidence
of bleeding, we currently infuse rt-PA
at 0.5 mg/h with use of only subthera-
peutic doses of heparin ranging from
300 to 500 U/h. Both maneuvers seem
reasonable because infusion rates as
Table 5
Summary of Studies Comparing Dosing Parameter to Major Bleeding Events
Study (Ref. No.)* N
Mean Duration of
Infusion (h)
Major Bleeding
Rate (%)
Hess 1996 (22) 288 1.3 0
Schweizer 1996 (23) 60 2 0
Risius 1986 (8) 25 3.6 4
Braithwaite 1997 (24)† 49 4 6
Graor 1990 (1) 65 4.7 12
Risius 1987 (16) 40 5.5 8
Braithwaite 1995 (21) 43 7§ 12
Ward 1994 (11)‡ 23 14.4 35
Braithwaite 1997 (24)† 44 20 7
Lonsdale 1992 (39) 69 22 7
Ward 1994 (11)‡ 27 26.7 4
Berridge 1990 (40) 28 29 0
* Studies listed in order of increasing infusing lengths.
† Separate limbs of same study.
‡ Separate limbs of same study.
§ Median.
428 • Transcatheter Intraarterial Infusion of rt-PA for Ischemia April 2001 JVIR

7. low as 0.25 mg/h or 0.02 mg/kg/h
have been shown to be clinically effec-
tive (10,41), and the use of therapeutic
heparin does not seem to be critical for
thrombolytic success. Unfortunately,
this approach will clearly not obviate
the need for longer infusion times. At
the time of this report, the number of
patients treated with use of this up-
dated protocol is small, and results so
far are limited and anecdotal at best.
In conclusion, we achieved good
thrombolytic and clinical success dur-
ing our initial experience with use of
intraarterial rt-PA for the treatment of
acute lower limb arterial occlusion.
Unfortunately, this was achieved
seemingly at the expense of increased
bleeding. Despite this, only four of the
bleeding complications resulted in rel-
atively minor surgery, and therefore
thrombolytic therapy remains our
front line treatment for acute arterial
occlusion. Undoubtedly, as new
thrombolytic agents are put to clinical
use, various new dosing protocols will
evolve; however, continued close clin-
ical surveillance will be required to
ensure that efficacy and safety are
maintained.
Acknowledgment: The authors would
like to express their sincere gratitude to
Tammy Lovell for her assistance in prepar-
ing this manuscript.
References
1. Graor RA, Olin J, Bartholomew JR,
Ruschhaupt WF, Young JR. Efficacy
and safety of intraarterial local infusion
of streptokinase, urokinase, or tissue
plasminogen activator for peripheral
arterial occlusion: a retrospective
review. J Vasc Med Biol 1990; 2:310–
315.
2. STILE Investigators. Results of a pro-
spective randomized trial evaluating
surgery versus thrombolysis for isch-
emia of the lower extremity: the STILE
trial. Ann Surg 1994; 220:251–268.
3. Activase (Alteplase recombinant)
package insert. South San Francisco,
CA: Genentech, Inc., 1999.
4. Abbokinase (Urokinase) package
insert. Abbott Park, IL: Abbott Labora-
tories, 1999.
5. Stump DC, Collen D. The fibrinolytic
system: implications for thrombolytic
therapy. In: Califf RM, Mark DB, Wag-
ner GS, eds. Acute coronary care in the
thrombolytic era. Chicago, IL: Year
Book Medical Publishers, 1988:58–74.
6. Holden RW. Plasminogen activators:
pharmacology and therapy. Radiology
1990; 174:993–1001.
7. Graor RA, Risius B, Young JR, et al.
Peripheral artery and bypass graft
thrombolysis with recombinant human
tissue-type plasminogen activator. J
Vasc Surg 1986; 3:115–124.
8. Risius B, Graor RA, Geisinger MA, et
al. Recombinant human tissue-type
plasminogen activator for thromboly-
sis in peripheral arteries and bypass
grafts. Radiology 1986; 160:183–188.
9. Valji K. Evolving strategies for
thrombolytic therapy of peripheral
vascular occlusion. J Vasc Interv Radiol
2000; 11:411–420.
10. Earnshaw JJ, Westby JC, Gregson RHS,
Makin GS, Hopkinson BR. Local
thrombolytic therapy of acute periph-
eral arterial ischaemia with tissue plas-
minogen activator: a dose-ranging
study. Br J Surg 1988; 75:1196–1200.
11. Ward AS, Andaz SK, Bygrave S. Pe-
ripheral thrombolysis with tissue plas-
minogen activator. Results of two
treatment regimens. Arch Surg 1994;
129:861–865.
12. Bero CJ, Cardella JF, Reddy K, et al.
Recombinant tissue plasminogen acti-
vator for the treatment of lower ex-
tremity peripheral vascular occlusive
disease. J Vasc Interv Radiol 1995;
6:571–577.
13. Cina CS, Goh RH, Chan J, et al. In-
traarterial catheter-directed thrombol-
ysis: urokinase versus tissue plasmino-
gen activator. Ann Vasc Surg 1999; 13:
571–575.
14. Rutherford RB, Baker JD, Ernst C, et al.
Recommended standards for reports
dealing with lower extremity ischemia:
revised version. J Vasc Surg 1997; 26:
517–538.
15. Ouriel K, Kandarpa K, Schuerr DM,
Hultquist M, Hodkinson G, Wallin B.
Prourokinase versus urokinase for re-
canalization of peripheral occlusions,
safety and efficacy: the PURPOSE
Trial. J Vasc Interv Radiol 1999; 10:
1083–1091.
16. Risius B, Graor RA, Geisinger MA,
Zelch MG, Lucas FV, Young JR.
Thrombolytic therapy with recombi-
nant human tissue-type plasminogen
activator: a comparison of two doses.
Radiology 1987; 164:465–468.
17. Verstraete M, Hess H, Mahler F, et al.
Femoro-popliteal artery thrombolysis
with intra-arterial infusion of recombi-
nant tissue-type plasminogen activator.
Report of a pilot trial. Eur J Vasc Surg
1988; 2:155–159.
18. Spengel FA, Kuffer G, Stiegler H. Ef-
ficacy and tolerance of recombinant tis-
sue-type plasminogen activator in pa-
tients with thrombotic or embolic
occlusions of leg-arteries. Clin Invest
1993; 71:323–326.
19. Ward AS, Andaz SK, Bygrave S.
Thrombolysis with tissue-plasminogen
activator:resultswithahigh-dosetrans-
thrombus technique. J Vasc Surg 1994;
19:503–508.
20. Braithwaite BD. Thrombolysis with
tissue-plasminogen activator: results
with a high-dose transthrombus tech-
nique [letter]. J Vasc Surg 1995; 21:540.
21. Braithwaite BD, Birch PA, Poskitt KR,
Heather BP, Earnshaw JJ. Accelerated
thrombolysis with high dose bolus
t-PA extends the role of peripheral
thrombolysis but may increase the
risks. Clin Radiol 1995; 50:747–750.
22. Hess H, Mietaschk A, von Bilderling P,
Neller P. Peripheral arterial occlu-
sions: local low-dose thrombolytic
therapy with recombinant tissue-type
plasminogen activator (rt-PA). Eur J
Vasc Endovasc Surg 1996; 12:97–104.
23. Schweizer J, Altmann E, Stosslein F,
Florek HJ, Kaulen R. Comparison of
tissue plasminogen activator and
urokinase in the local infiltration
thrombolysis of peripheral arterial
occlusions. Eur J Radiol 1996; 22:129–
132.
24. Braithwaite BD, Buckenham TM, Gal-
land RB, Heather BP, Earnshaw JJ.
Prospective randomized trial of high-
dose bolus versus low-dose tissue plas-
minogen activator infusion in the man-
agement of acute limb ischaemia. Br J
Surg 1997; 84:646–650.
25. Ouriel K, Shortell CK, DeWeese JA, et
al. A comparison of thrombolytic
therapy with operative revasculariza-
tion in the initial treatment of acute
peripheral arterial ischemia. J Vasc
Surg 1994; 19:1021–1030.
26. Ouriel K, Veith FJ, Sasahara AA.
Thrombolysis or peripheral arterial
surgery: Phase I results. J Vasc Surg
1996; 23:64–75.
27. Ouriel K, Veith FJ, Sasahara AA. A
comparison of recombinant urokinase
with vascular surgery as initial treat-
ment for acute arterial occlusion of the
legs. N Engl J Med 1998; 338:1105–1111.
28. Weitz JI, Leslie B, Ginsberg J. Soluble
fibrin degradation products potentiate
tissue plasminogen activator-induced
fibrinogen proteolysis. J Clin Invest
1991; 87:1082–1090.
29. Weitz JI, Leslie B, Hirsh J, Klement P.
␣2-Antiplasmin supplementation in-
hibits tissue plasminogen activator-in-
duced fibrinogenolysis and bleeding
with little effect on thrombolysis. J Clin
Invest 1993; 91:1343–1350.
30. Weitz JI, Rischke J. The role of (DD)E
in the kinetics of plasminogen activa-
tion by tissue plasminogen activator
[abstract]. Thromb Haemost 1993; 69:
18:554.
31. Weitz JI. Limited fibrin specificity of
tissue-type plasminogen activator and
its potential link to bleeding. J Vasc
Interv Radiol 1996; 6:18S–23S.
Swischuk et al • 429Volume 12 Number 4

8. 32. Rao AK, Pratt C, Berke A, et al.
Thrombolysis in myocardial infarction
(TIMI) trial-phase 1: hemorrhagic man-
ifestations and changes in plasma fi-
brinogen and the fibrinolytic system
in patients treated with recombinant
tissue plasminogen activator and
streptokinase. J Am Coll Cardiol 1988;
11:1–11.
33. Agnelli G, Buchanan MR, Fernandez F,
et al. A comparison of the thrombo-
lytic and hemorrhagic effects of tissue-
type plasminogen activator and strep-
tokinase in rabbits. Circulation 1985;
72:178–182.
34. Agnelli G, Buchanan MR, Fernandez F,
Van Ryn J, Hirsh J. Sustained throm-
bolysis with DNA-recombinant tissue
type plasminogen activator in rabbits.
Blood 1985; 66:399–401.
35. Verstraete M, Bounameaux H, de Cock
F, Van de Werf E, Collen D. Pharma-
cokinetics and systemic fibrinogeno-
lytic effects of recombinant human tis-
sue-type plasminogen activator (rt-PA)
in humans. J Pharmacol Exp Ther 1985;
235:506–512.
36. Agnelli G. Rationale for bolus t-PA
therapy to improve efficacy and safety.
Chest 1990; 97(suppl 4):161S–167S.
37. Decrinis M, Pilger E, Stark G, Lafer M,
Obernosterer A, Lammer J. A simpli-
fied procedure for intra-arterial throm-
bolysis with tissue-type plasminogen
activator in peripheral arterial occlu-
sive disease: primary and long-term
results. Eur Heart J 1993; 14:297–305.
38. Bovill EG, Terrin ML, Stump DC, et al.
Hemorrhagic events during therapy
with recombinant tissue-type plasmin-
ogen activator, heparin, and aspiring
for acute myocardial infarction: results
of the Thrombolysis in Myocardial In-
farction (TIMI), Phase II trial. Ann In-
tern Med 1991; 115:256–265.
39. Lonsdale RJ, Berridge DC, Earnshaw JJ, et
al. Recombinant tissue-type plasmino-
gen activator is superior to streptokinase
for local intra-arterial thrombolysis. Br J
Surg 1992; 79:272–275.
40. Berridge DC, Gregson RHS, Makin GS,
Hopkinson BR. Tissue plasminogen acti-
vator in peripheral arterial thrombolysis.
Br J Surg 1990; 77:179–182.
41. Koppensteinen R, Minar E, Ahmadi R, et
al. Low doses of recombinant human
tissue-type plasminogen activator for lo-
cal thrombolysis in peripheral arteries
[letter]. Radiology 1988; 168:877–878.
430 • Transcatheter Intraarterial Infusion of rt-PA for Ischemia April 2001 JVIR