Full text of DOI:10.1016/S1051-0443(07)61352-9

Percutaneous Fibrin Sheath Stripping
versus Transcatheter Urokinase
Infusion for Malfunctioning Well-
positioned Tunneled Central Venous
Dialysis Catheters: A Prospective,
1
Randomized Trial
Richard J. Gray, MD 2
Abraham Levitin, MD
David Buck, MD
Lisa C. Brown, RN
Yvonne H. Sparling, MS
Kathleen A. Jablonski, PhD
PURPOSE: To compare central dialysis catheter patency rates af-
ter stripping procedures with those after urokinase (UK) infusion.
MATERIALS AND METHODS: Fifty-seven tunneled catheters with
either (i) flow rates less than 250 mL/min and established baseline
flow rates >300 mL/min or (ii) flow rates 50 mL/min less than
higher established baseline flows were prospectively randomized
to undergo stripping procedures (n ؍ 28) or UK infusion (n ؍ 29)
at 30,000 U/h via each port concurrently, for a total 250,000 U. Suc-
cess and patency were determined by dialysis at normal flow rates
3
Amanuel Fessahaye, MD
Atul K. Gupta, MD
(
>300 mL/min) or at the previously established higher baseline
Index terms: Catheters and catheter-
ization, central venous access ● Catheters
and catheterization, complications ● Dial-
ysis, shunts ● Thrombolysis ● Urokinase
rate. Flow rates were monitored weekly. Primary patency ended
with catheter malfunction or removal. Kaplan-Meier survival anal-
ysis was used to construct survival curves.
JVIR 2000; 11:1121–1129
RESULTS: In the stripping group, initial clinical success was 89%
(25 of 28). The 15-, 30-, and 45-day primary patency rates were 75%
Abbreviation: UK ϭ urokinase
(
n ؍ 20), 52% (n ؍ 13), and 35% (n ؍ 8), respectively. The median
duration of additional function was 32 days (95% CI: 18–48 d). In
the UK group, initial clinical success was 97% (28 of 29). The 15-,
30-, and 45-day primary patency rates were 86% (n ؍ 21), 63% (n ؍
1
From the Department of Radiology
(
R.J.G., A.L., D.B., L.C.B., A.F., A.K.G.),
1
3), and 48% (n ؍ 9), respectively. The median duration of addi-
Washington Hospital Center, and the
tional patency was 42 days (95% CI: 22–153 d). The Wilcoxon test
for equality detected no significant difference in the survival
curves for the two treatment groups (P ؍ .236).
Statistical Analysis and Computer Center
(
Y.H.S., K.A.J.), Medstar Research Insti-
tute, Washington, DC. Interim results of
this study were presented at the 1998
SCVIR annual meeting. Funded by an
unrestricted grant from Abbott Laborato-
ries, North Chicago, IL. Received Decem-
ber 1, 1999, revision requested January
CONCLUSION: There is no significant difference in time to pri-
mary patency between the two methods. Both allow temporary
catheter salvage in most patients.
1
1, 2000, revision received May 16, ac-
cepted May 19. Address correspon-
dence to R.J.G., Department of Radiol-
ogy, Washington Hospital Center, 110
Irving Street, NW, Washington, DC
THE first-line therapy for dialysis
catheters with suboptimal flow
rates (Ͻ300 mL/min) (1) unrespon-
sive to simple positional maneuvers
and port reversal has usually been
instillation of a small quantity of
urokinase (UK) in the dialysis unit
for periods as long as 20 minutes
(2–4). This may be attempted sev-
eral times. Patients whose catheters
fail a thrombolytic instillation are
typically referred for transcatheter
venography to confirm satisfactory
catheter position and to evaluate for
the presence of pericatheter fibrin
sheath or thrombus (1). Poorly posi-
tioned or kinked catheters are usu-
ally treated in a straightforward
2
0010. E-mail: rjgdln@erols.com
Current address: Department of Radiol-
2
ogy, The Cleveland Clinic Foundation,
Cleveland, Ohio.
3
Current address: Department of Radiol-
ogy, Howard University Hospital, Wash-
ington, DC.
©
SCVIR, 2000
1
121

1122 ● Stripping versus UK Infusion for Malfunctioning Catheters
October 2000 JVIR
manner with standard interven-
tional techniques (4,5). Pericatheter
fibrin sheaths and/or thrombus
have been treated with use of a va-
riety of methods, including percuta-
neous fibrin sheath stripping
heart failure. There were eight cur-
rent smokers.
Forty-six catheters had been in-
dwelling for a median of 59 days
(range, 11–682 d) at the time of re-
ferral for catheter malfunction. Be-
cause we practice in a tertiary re-
flowing sluggishly away from the
catheter tip was considered diagnos-
tic of a pericatheter fibrin sheath
(Fig 1a). Filling defects were con-
sidered diagnostic of pericatheter
thrombus (Fig 1b). The veno-
graphic study was considered nor-
mal if contrast material flowed im-
mediately away from the catheter
tip (Fig 1c). With use of these cri-
teria, 79% (45 of 57) of the trans-
catheter venograms before treat-
ment were abnormal; 61% (35 of 57)
revealed fibrin sheaths, 7.0% (4 of
(
4,6–9) and thrombolysis through
the dialysis catheter (2,10–12), with ferral center, the exact indwell
return of catheter function for at
least one dialysis session in most
patients. Nevertheless, the subse-
quent patency rates after thrombol-
ysis are unknown and the results
after percutaneous fibrin sheath
stripping have varied widely. These
issues led us to conduct a prospec-
tive, randomized trial of a 4-hour
times were not known for the other
11 catheters. Either vascular access
surgeons or interventional radiolo-
gists had originally inserted the
catheters. Most had failed a trial of
UK instillation (Opencath; Abbott
Laboratories, North Chicago, IL) in
the dialysis clinic. There were 45
Permcaths (Quinton Instrument
5
1
7) fibrin sheath and thrombus and
1% (6 of 57) pericatheter throm-
bus. None showed large clots
UK infusion compared to percutane- Company, Bothel, WA), seven Tesio
around the catheter. The studies
were normal in 15.8% (9 of 57) of
cases and nondiagnostic (because of
respiratory motion) in 5.3% (3 of 57)
of cases. Based on these contrast
studies, the catheter tips were lo-
cated in the right atrium of 22 pa-
tients, bridging the superior vena
cava/right atrial junction in 24 pa-
tients, and in the superior vena
cava just above the right atrium in
11 patients. The superior vena cava
was considered normal in 23 pa-
tients, stenosed in one patient, and
not evaluable in 34 patients. A com-
puter-generated randomization
schedule was used to assign pa-
tients to the UK infusion or strip-
ping group after transcatheter
ous fibrin sheath stripping. Herein
we report the results of this trial
and review the pertinent literature.
catheters (Medcomp, Harleysville,
PA), and five Hickman catheters
(Bard Access Systems, Salt Lake
City, UT) inserted via the right in-
ternal jugular vein in 40 patients,
right subclavian vein in three, left
internal jugular vein in 12, and left
subclavian vein in two. All cathe-
ters had established baseline flow
MATERIALS AND METHODS
Between April 30, 1996 and Oc-
tober 28, 1998, 57 patients who had
5
7 poorly functioning dialysis cathe- rates greater than or equal to 300
ters were enrolled in the study un-
der the auspices of the hospital’s
Institutional Review Board. There
were 34 women and 23 men, with a
median age of 60 years (range,
mL/min (1) for at least three dialy-
sis treatments after catheter inser-
tion. Catheter malfunction was de-
fined as a flow rate through one or
both ports less than 250 mL/min or
a decrease through one or both
ports greater than 50 mL/min if the
established baseline flow rate was
greater than 300 mL/min. Indica-
tions for treatment included blood
flow rates less than 250 mL/min
through both ports (29 cases), blood
flow rates less than 250 mL/min via
the arterial (nine cases) or venous
(one case) port, a decrease in flow
rate greater than 50 mL/min below
established baseline flow through
both ports (three cases), decrease in
flow rate greater than 50 mL/min
2
6–91 y). The study was explained
to all patients and a protocol con-
sent form was signed. Each patient
was enrolled only once, even if a
new catheter was later inserted at
the same or a new puncture site.
During the course of the study, 54
other patients were excluded from
enrollment for the following rea-
sons: refusal to undergo stripping
procedure before (n ϭ 28) or after
venography showed satisfactory
catheter position and no mechanical
problems such as kinking. The me-
dian catheter indwell time at the
time of treatment in the UK group
(
1
3
26 catheters) was 68 days (range,
4–682 d) and 35 days (range 11–
06 d) in the stripping group (20
(
n ϭ 2) randomization, refusal or
inability to consent (n ϭ 17), and
contraindication to UK (n ϭ 7).
catheters); the exact indwell times
for the other 11 catheters were not
known. The median test did not in-
dicate a significant difference in the
known pretreatment indwell times
between the groups (P ϭ .388). The
median test compares the propor-
tions in each sample that are less
than the combined median with use
of a binomial test (ie, Fisher exact
Risk factors for chronic renal failure below established baseline flow
included hypertension (n ϭ 23), dia- through the venous (three cases) or
betes mellitus (n ϭ 18), chronic glo-
merulonephritis (n ϭ 1), systemic
lupus erythematosus (n ϭ 1), and
multiple myeloma (n ϭ 1). The
cause of renal failure was uncertain
in the other 15 patients. Ten pa-
tients had clinically apparent coro-
nary artery disease, seven had a
history of coronary artery disease
and congestive heart failure, and
two had a history of congestive
arterial (one case) port, and com-
plete occlusion of the arterial (seven
cases), venous (two cases), or both
(four cases) ports.
All catheters were observed with
transcatheter digital subtraction
venography during very slow hand
injections of 5–25 mL iodinated con- test). It is sensitive to difference in
trast material through each port
consecutively. Contrast tracking
retrogradely along the catheter or
location. It is the most powerful test
for comparing skewed, asymmetric
distributions.

Gray et al ● 1123
Volume 11 Number 9
Figure 1. (a) Transcatheter contrast injection shows the catheter tip in the right atrium. Contrast material passes retro-
gradely within a fibrin sheath, outlining the catheter before passing through a fenestration in the sheath to opacify the supe-
rior vena cava (arrow). (b) Transcatheter contrast study shows a globular filling defect indicative of pericatheter thrombus (ar-
rowhead). (c) Image obtained during catheter contrast injection through the red port shows contrast material flowing rapidly
away from the red port.
●
Urokinase Infusion
was engaged with the snare, which
was advanced over the catheter as
far as possible, tightened, and
pulled off the catheter tip (Fig 2).
For Tesio catheters (Medcomp), the
two catheters were stripped inde-
pendently. The dialysis catheter
hub was aseptically prepared and a
guide wire was passed through the
transported to their rooms after he-
mostasis was achieved.
A solution of 250,000 U of UK
(Abbott Laboratories) dissolved in
2
●
Follow-Up
50 mL 0.9 normal saline or 5%
dextrose was administered at 30
mL/h concurrently via both ports
Posttreatment transcatheter con-
trast studies were performed at the
operators’ discretion, but the results
were not used for patency determi-
(60,000 U/h total) over the course of
4 hours and 10 minutes. The 4-hour
infusion was chosen to facilitate
treatment for outpatients. Uroki-
nase was administered to outpa-
tients (n ϭ 21) in either the Inter-
ventional Radiology Recovery Room
or a 23-hour admission bed and to
inpatients (n ϭ 8) usually in their
hospital beds. No other procedure
medications were administered in
this treatment group.
catheter to facilitate readvancement nations because clinical function
of the snare over the dialysis cathe-
ter tip for additional stripping
was considered more important. Im-
mediate clinical success was defined
as at least one successful dialysis
session with flow rates higher than
the previously established baseline
flow rate. Follow-up flow rates were
obtained by routine weekly review
of dialysis clinic records. Primary
patency was defined as a flow rate
greater than the established base-
line flow rate without additional
intervention. Secondary patency
was defined as a flow rate greater
than the established higher base-
line flow rate assisted by repeat
treatment with the same treatment
modality, either stripping or UK
infusion. Study endpoints included
restoration of catheter function by
another treatment modality (eg,
crossover to UK or stripping or
catheter exchange), catheter re-
passes (6) (Fig 3). The number of
stripping passes was at the opera-
tor’s discretion, which resulted in
wide variation in the number of
passes performed: 4–6 passes were
made in two patients, 7–9 passes
were made in five, 10–12 passes
were made in nine, and more than
1
2 passes were made in nine. The
●
Percutaneous Fibrin Sheath
Stripping
number of passes was not recorded
for three procedures. Midazolam
was administered to 23 patients; 19
of these patients also received fenta-
nyl. One patient was pretreated with
intravenous cefoxitin. This patient
also received diphenhydramine be-
cause of a history of hives caused by
iodinated contrast material. Outpa-
tients (n ϭ 25) were observed for 2
hours and inpatients (n ϭ 3) were
Other investigators (4,6–8) have
described the technique in detail.
Briefly, all procedures were per-
formed from a right (24 cases) or
left (four cases) common femoral
access with a 25-mm (11 cases) or
3
5-mm (17 cases) Amplatz Nitinol
snare (Microvena, Vadnaise
Heights, MN). The dialysis catheter

1124 ● Stripping versus UK Infusion for Malfunctioning Catheters
October 2000 JVIR
Figure 2. Amplatz snare wrapped
around the catheter during a stripping
pass.
Figure 4. Stripping and UK survival curves. f ϭ stripping; e ϭ urokinase.
maximum likelihood estimator was
used to estimate survivor functions.
Data were right censored. The dif-
ference between survivor functions
was tested with use of the Wilcoxon
statistic.
catheter clot. The other 12 patients
were not evaluated angiographically
after treatment because the pre-
treatment study was normal (n ϭ 4)
or the operator chose not to (n ϭ 8).
The initial clinical success rate
was 97% (28 of 29). The 15-, 30-,
and 45-day primary patency rates
were 86% (n ϭ 21), 63% (n ϭ 13),
and 48% (n ϭ 9), respectively. The
median duration of additional satis-
factory catheter function after infu-
sion was 42 days (95% CI: 22–364
d). As of 28 weeks after treatment,
one catheter is still being used
RESULTS
According to the randomization
schedule, 28 patients were treated
with a stripping procedure and 29
patients were treated with UK infu-
sion. Plots of the product-limit sur-
vival estimates and 95% confidence
bands are presented in Figure 4
and Table 1.
Figure 3. Amplatz snare positioned
around the guide wire near dialysis
catheter tip to facilitate another strip-
ping pass.
without further treatment. End-
points for primary patency included
recurrent catheter malfunction
moval for any reason (eg, catheter
malfunction, permanent access
available, accidental catheter re-
moval), or patient death. All analy-
ses were performed with the Statis-
tical Analysis Software System
treated by repeat UK infusion (n ϭ
●
Urokinase Group
3
), stripping (n ϭ 2), or catheter
Seventy-six percent (13 of 17)
exchange (n ϭ 5). The other end-
transcatheter contrast studies after
points were catheter removal for
(
Stat Version 7.0; SAS, Cary, NC).
UK infusion were normal; three had recurrent malfunction (n ϭ 6), ma-
The Kaplan-Meier nonparametric
fibrin sheaths and one had a peri-
tured permanent access (n ϭ 7), or

Gray et al ● 1125
Volume 11 Number 9
terminal illness (n ϭ 1). One pa-
tient demanded the catheter be re-
moved and three patients died with
functioning catheters. Because only
three patients were retreated with
UK when the catheter malfunc-
tioned again (operator’s discretion),
the secondary patency curve would
not be significantly different from
the primary patency curve.
Table 1
Values for Survival Curves (see Fig 4)
Survival
Distribution
Function
Estimate
Lower
Bound
of CI
Upper
Bound
of CI
Number
at Risk
Time
Censor
Urokinase Group
0
0
3
4
7
9
0
0
0
0
1
1
0
1
1
1
0
1
0
1
0
0
0
1
0
0
1
1
1
0
1
1
0
1
29
28
27
26
25
24
23
22
21
20
19
18
17
15
14
12
11
10
9
1.00000
0.96552
0.93103
0.89655
0.89655
0.89655
0.85920
0.85920
0.85920
0.85920
0.81624
0.81624
0.72554
0.72554
0.62880
0.58043
0.53206
0.53206
0.47886
0.42565
0.42565
0.42565
0.42565
0.34052
0.34052
0.34052
0.17026
1.00000
1.00000
0.89911
1.00000
0.83881
1.00000
0.78571
1.00000
–
–
–
–
●
Stripping Group
1
1
1
1
1
1
1
2
2
3
3
3
4
5
5
7
7
1
2
3
5
6
7
9
0
2
4
5
9
2
1
1
6
6
0.73105
0.98734
Eighty-seven percent (20 of 23)
–
–
transcatheter contrast studies after
stripping procedures were normal;
one had a fibrin sheath and two
had pericatheter clots. The other
five patients were not evaluated an-
giographically after treatment be-
cause the pretreatment study was
normal (n ϭ 2) or the operator
chose not to (n ϭ 3).
–
–
–
–
0.66942
0.96305
–
–
0.54927
0.90182
–
–
0.43153
0.82608
0.37682
0.78404
0.32452
0.73961
–
–
The initial clinical success rate
was 89% (25 of 28). The 15-, 30-,
and 45-day primary patency rates
were 75% (n ϭ 20), 52% (n ϭ 13),
and 35% (n ϭ 8), respectively. The
median duration of additional satis-
factory catheter function after strip-
ping was 32 days (95% CI: 18–48
d). As of 18 weeks after treatment,
one catheter is still being used
without further treatment. End-
points of primary patency included
restoration of catheter function by
repeat stripping (n ϭ 3), UK infu-
sion (n ϭ 4), or catheter exchange
0.26749
0.69023
8
0.21360
0.63771
7
–
–
6
–
–
5
–
–
1
1
1
3
4
12
17
95
64
65
4
0.11458
0.56647
3
–
–
2
–
0.00000
–
–
0.43187
–
1
0
Stripping Group
1.00000
0.85714
0.82143
0.82143
0.74675
0.70942
0.67208
0.63474
0.59740
0.56006
0.56006
0.52006
0.48006
0.44005
0.44005
0.39605
0.35204
0.30804
0.26403
0.26403
0.21122
0.15842
0.15842
0.15842
0.00000
0
0
7
9
0
0
0
1
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
1
0
0
1
1
0
28
24
23
22
20
19
18
17
16
15
14
13
12
11
10
9
1.00000
0.72753
0.67957
–
0.58437
0.53946
0.49601
0.45389
0.41298
0.37321
–
0.33081
0.28982
0.25019
–
0.20659
0.16503
0.12560
0.08845
–
1.00000
0.98676
0.96329
–
0.90914
0.87937
0.84814
0.81559
0.78183
0.74692
–
0.70931
0.67029
0.62991
–
0.58550
0.53905
0.49047
0.43961
–
1
1
1
1
2
2
2
2
3
3
3
4
4
4
5
5
6
9
3
7
8
9
2
6
7
9
2
6
7
1
4
8
1
8
9
8
(
n ϭ 6). Other endpoints included
catheter removal for recurrent mal-
function (n ϭ 5) or positive blood
cultures (n ϭ 1), permanent access
(
n ϭ 5), transplant (n ϭ 2), and ac-
cidental catheter removal (n ϭ 1).
Because only three patients were
re-treated with a stripping proce-
dure when the catheter malfunc-
tioned again (operator’s discretion),
the secondary patency curve would
not be significantly different from
the primary patency curve.
8
7
6
5
4
0.04300
0.00365
–
0.37945
0.31319
–
3
●
Sample Size Calculation
1
1
2
09
29
67
2
1
–
–
A sample of 70 patients (35 in
0
0.00000
0.00000
each group) was initially chosen
based on an expected difference in
immediate success rates between

1126 ● Stripping versus UK Infusion for Malfunctioning Catheters
October 2000 JVIR
Table 2
Dialysis Catheter Removal
Study
Catheters Malfunction* (%) Infection (%) CV Thrombosis (%)
McDowell (14)
Cappello (15)
Moss (2)
Gibson (16)
Schwartz (17)
Lund (11)
172
107
168
94
118
222
250
77
5
5
5
5
0
0
7
8
0
1.6
0
1.2
0
0
9
28
19
11
7
17
28
19
—
Trerotola (10)
Duszak (18)
13
Note.—CV ϭ central vein.
Variable definitions.
*
Table 3
Etiology of Catheter Malfunction*
Study
No. Catheters
Episodes
Mechanical
Sheath/Clot
Crain (6)
24
—
29
63
44
42
31
63
4
4
40
38
24
40
Figure 5. Slow contrast injection
clearly shows the probable pathway that
a subsequent transcatheter thrombolytic
agent would take from the catheter tip
to the nearest fenestration (arrow) in
the fibrin sheath.
Suhocki (4)
Rockall (7)
Trerotola (10)
7
23
*
UK instillation failed in most patients.
time before removal (4). Until its
removal from the market by the
Food and Drug Administration, sim-
ple UK instillation at dialysis clin-
ics restored immediate function to
the treatment groups of 15%, type I
error of 5%, and power of 80%. The
initial clinical success and patency
curves after 57 patients were en-
rolled were compared to the results
obtained during an interim analysis tures 5 weeks after treatment. An-
after 44 patients were enrolled (13). other patient presented with a
Because there was no notable dif-
ference or change in the trends
since the interim analysis and be-
cause the 95% confidence bands
were broad, we concluded that en-
rollment of an additional 13 pa-
tients to reach the initial goal of 70
patients would not significantly
change the final analysis. The study
was therefore terminated after 57
patients were enrolled.
deaths were attributed to underly-
ing disease and considered to be
unrelated to the treatment. in the
stripping group, one patient devel-
oped fever and positive blood cul-
7
4%–81% of dialysis catheters (2,4).
Although this measure was per-
formed blindly and the durability is
unknown, it was effective enough to
be the appropriate first-line therapy
symptomatic pericatheter innomi-
nate thrombosis necessitating cath-
eter removal within days after the
catheter had been unsuccessfully
stripped.
(
1) because it was considered safe,
inexpensive, and expediently al-
lowed dialysis to resume after a
minimal delay during the same
scheduled dialysis session.
DISCUSSION
Catheters that fail postural ma-
neuvers, port reversal, and UK in-
stillation may be malpositioned or
Central venous dialysis catheter
malfunction is a serious problem
necessitating removal of as many as kinked, but mechanical problems
8% of dialysis catheters simply be- are generally much less common
cause they do not work (2,10,11,14– than pericatheter fibrin sheath or
The Wilcoxon test indicated that
the times to primary patency for
the stripping and UK groups were
not significantly different from each
2
1
7). Catheters are removed for mal-
clot formation (4,6,7,10) (Table 3)
unless the catheter has been re-
cently inserted. Catheter malfunc-
tion in well-positioned catheters is
typically caused by the presence of
a pericatheter fibrin sheath or a
small amount of thrombus about
the catheter tip (4,6,7,10). Pericath-
eter fibrin sheath formation has
been shown in a human autopsy
2
function approximately as often as
for infection and far more fre-
other (␹ ϭ 1.41; 1 degree of free-
dom; P ϭ 0.236).
quently than for symptomatic peri-
catheter thrombus (2,10,11,14–17).
(
Table 2) Dialysis catheter mal-
function not requiring removal is
There were two deaths in the UK extremely common, affecting 3%–
●
Complications
group, one at 2 weeks and another
at 1 month after the infusion. Both
10% of all dialysis sessions (2,16),
and 87% of all catheters at some

Gray et al ● 1127
Volume 11 Number 9
for at least one dialysis session,
range from 55% to 97%. One inves-
tigator instilled a total of 250,000 U
of concentrated UK into both ports
and reported that it “nearly always”
worked for restoring immediate
catheter function (12). Our study
confirms the previous studies’ high
rates of immediate functional resto-
ration. These treatments appear to
be very safe; there were no bleeding
complications attributed to the
Table 4
Thrombolytic Infusion
No. of
Catheters
Clinical
Success
Additional
Patency
Study
Agent/Dose
Moss (2)
Uldall (12)
58
103
Streptokinase (12 h)
UK (250,000 U bolus)
97%
“nearly
always”
79.5%
55%
97%
–
–
Lund (11)
Trerotola (10)
Current Study
39
11
29
UK (250,000 U/6 h)
UK (250,000 U/6 h)
UK (250,000 U/4 h)
–
31 d (mean)
48% at 45 d
(
primary)
thrombolytic agent in any of these
studies. Trerotola et al (10) reported
a 31-day mean period of additional
function in their very small group of
patients; otherwise, the durability
of these treatments has not been
previously studied. Our cumulative
patency rates after transcatheter
thrombolytic infusion indicate that
approximately half (48%) of treated
catheters will maintain function for
an additional 45 days after treat-
ment. Although modest, this addi-
tional period of function will allow
permanent access creation and/or
maturation in many patients.
Note.—No bleeding complications in 240 cases.
Table 5
Fibrin Sheath Stripping
No. of
Catheters
Clinical
Success (%)*
Study
Additional Patency
Crain (6)
40
24
38
31
91
28
98
92
95
61
96
89
45% at 3 months (primary)
8% (2/24) at 2 wks
3 mo (mean)
4.25 mo (median)
51% at 3 mo (primary)
35% at 45 d (primary)
Haskal (8)
Suhocki (4)
Rockall (7)
Brady (9)
Current Study
*
At least one successful dialysis using variable criteria for success.
The investigators (4,6–9) listed
in Table 5 have reported the re-
sults of pericatheter fibrin sheath
stripping. The initial success rates
in these series, including ours, are
generally high, and the overall com-
plication rate is 6% (16 of 253). We
found a modest durability with 35%
of catheters maintaining primary
patency for 45 days. To the con-
trary, Crain et al (6) and Brady et
al (9) reported primary patency
rates of 45% and 65%, respectively,
at 3 months; Crain also performed
multiple stripping procedures and
optimistically reported that 81% of
study (19) to occur as early as 24
hours after placement and is
thought to occur in 80%–100% of
central venous catheters 2–7 days
after insertion (19–21). Fibrin
sheaths propagate from the venous
insertion site and from the tip to-
ward the center of the catheter and
can persist for weeks even after the
catheter is removed (19). When the
pericatheter sheath and/or associ-
ated thrombus infringe on the func-
tional endhole(s) of the catheter,
decreased dialysis flow rates result.
Fibrin sheaths can be demon-
venographic technique, we believe
that it is very important to inject
contrast material through the cath-
eter slowly to avoid creating a hole
in the fibrin sheath near the cathe-
ter tip. An iatrogenic fenestration
from rapid contrast material (or sa-
line) injection can hinder angio-
graphic diagnosis of the fibrin
sheath because contrast material
can pass preferentially through the
fenestration in the sheath rather
than retrogradely around the cathe- the treated catheters functioned
ter. In addition, a subsequently ad-
ministered transcatheter thrombo-
satisfactorily for at least 1 year af-
ter the initial catheter insertion.
strated by intravascular ultrasound
22) but are usually diagnosed on
lytic agent will also run through the Similarly, Suhocki et al (4) empha-
(
iatrogenic fenestration instead of
bathing the catheter from the tip
proximally to the nearest naturally
occurring fenestration (Fig 5). This
will decrease the total surface area
of the fibrin sheath that is exposed
to the contrast agent.
sized that most treated catheters
remained functional for their in-
tended duration of use. Rockall and
colleagues (7) were somewhat less
optimistic, reporting a 61% (19 of
31) initial return of function; how-
ever, their experience illustrated
the importance of evaluating for
catheter malposition or kinks before
percutaneous treatment. Contrary
to the promising results in the
transcatheter venographic studies
with (21) or without (4,6,8,9) pull-
ing the catheter back before injec-
tion of contrast material. Because
the sensitivity of transcatheter
venography for detection of fibrin
sheaths is unknown, we included all
The results of studies reporting
malfunctioning well-positioned cath- thrombolysis (2,10–12), including
eters in this study, whether or not a ours, are presented in Table 4. The
fibrin sheath was detected on trans- immediate success rates, defined as
catheter venography. Regarding
restoration of satisfactory function
above four studies, Haskal et al (8)

1128 ● Stripping versus UK Infusion for Malfunctioning Catheters
October 2000 JVIR
experienced dismal patency rates,
the catheter tip through a fenestra-
catheters with tips placed at the
junction of the superior vena cava
and the right atrium and those with
tip placement in the atrium. Al-
finding that 92% (22 of 24) of cathe- tion in the fibrin sheath to a posi-
ters returned to the pretreatment
blood-liter process rate by the fifth
poststripping dialysis session. As a
result, this group completely aban-
doned the stripping procedure.
Our results showed a significant
difference between 4-hour UK infu-
sion and fibrin sheath stripping for
neither immediate restoration of
catheter function nor maintenance
of long-term patency. Although our
tion outside of the sheath. They
compared these catheter exchanges
in a nonrandomized fashion with de though the numbers were small,
novo catheter placement in the
same patient population and found
no significant differences in cathe-
ter patency or complication rates,
including infections. Whether a
catheter exchange or a thrombolytic and the right atrium are more
infusion is preferable remains unde- prone to malfunction. In addition,
termined. As currently reimbursed
their preliminary results agreed
with the DOQI Vascular Access
Workgroup’s opinion (1) by suggest-
ing that catheter tips above the
junction of the superior vena cava
right internal jugular catheters
results did not demonstrate a differ- by Medicare, a single catheter ex-
with the arterial port oriented later-
ally at any level were more predis-
posed to malfunction; the numbers
were again too small to draw firm
conclusions, but are contradictory to
the recommendations of Trerotola et
al (10). Although central catheter
tips have been shown to withdraw
proximally in the upright position
ence, a trend favoring UK is seen
when looking at the Kaplan-Meyer-
derived interval patency rates and
the median time period of addi-
tional catheter function in both
groups. Even if stripping and UK
infusion have similar results, a
thrombolytic infusion is our pre-
ferred therapy for several reasons.
First, the patients prefer it. Our
greatest difficulty enrolling an oth-
erwise eligible patient for the study
was patient refusal to undergo a
stripping procedure after hearing
about the lysis and stripping op-
tions. Second, transcatheter infu-
sion is noninvasive; a stripping pro-
cedure requires a venous puncture.
Third, stripping procedures have
been associated with potentially di-
sastrous complications, whereas
transcatheter administration of
thrombolytics has not (2,10–12):
one group (6) reported asymptom-
atic common femoral puncture site
change is much less costly than an
infusion. Catheter exchange over a
guide wire spares vein (18) and
usually allows a more expedient re-
turn to dialysis than an infusion
does. Nevertheless, it is an invasive
procedure that can be complicated
by prolonged pericatheter tract ooz-
ing (18). In addition, patients would as opposed to the supine position
probably prefer a thrombolytic infu-
sion because of its noninvasive na-
ture.
Currently, the best method to
minimize catheter malfunction is
careful initial catheter insertion
with good positioning of the cathe-
ter tip under fluoroscopy. Unfortu-
(26,27), associated deleterious ef-
fects on dialysis catheter function
have not been demonstrated.
Much remains to be learned
about the prevention and treatment
of dialysis catheter malfunction.
New catheter designs, including
subcutaneous implantable dialysis
nately, the optimal position and ori- ports and nonthrombogenic catheter
entation of the catheter tip is un-
clear. In rats, a fibrin sheath does
not envelop the catheter tip if it is
in the right atrium (24). This is not
the case for humans. Nevertheless,
there is a growing consensus that
the catheter tip should be at the
materials, are on the horizon. Low
dose warfarin has been suggested in
at least one study to decrease the
incidence of catheter malfunction
(12). Regarding therapy, we selected
a 4-hour UK infusion to allow more
expedient treatment than the previ-
thrombus and another (4) later pub- superior vena cava/right atrial junc- ously reported 6-hour infusions
lished a case report of a septic pul-
monary embolus caused by a strip-
ping procedure (23). In addition,
one of our patients presented with
symptomatic pericatheter innomi-
nate vein thrombosis after an at-
tion or in the right atrium (1). For
dual-lumen catheters, Trerotola et
al (10) emphasized the importance
of orienting the arterial port tip to-
ward the lumen and away from the
(10,11). Nevertheless, if an even
shorter thrombolytic infusion or bo-
lus could provide similar results,
(12) patients could be treated and
return for dialysis on the same day.
vessel wall. For catheters positioned The gradual transformation of the
tempted stripping procedure. For all in the atrium, they recommended
pericatheter sheath from fibrin-con-
taining material to organized fi-
brous connective tissue (24) sug-
gests that thrombolytic instillation
or infusion will lose effectiveness
after longer catheter indwell times.
As previously stated, the effective-
of these reasons, then, thrombolysis
is our preferred treatment.
After poor results of stripping
procedures (8) had been demon-
strated at their institution, Duszak
et al (18) began changing catheters
orienting the arterial port medially,
toward the tricuspid valve. For
catheter tips in the superior vena
cava, they recommended that the
arterial port be oriented laterally
unless the tract is parasternal, in
through the same tract over a guide which case medial orientation of the ness of catheter exchange relative
wire while making an attempt to
position the catheter tip beyond or
outside the confines of the fibrin
sheath. This was done either by re-
arterial port was suggested. Oncay
et al (25) examined the position and to be elucidated; some have sug-
orientation of the catheter tip with
respect to the need for subsequent
to a thrombolytic infusion remains
gested balloon dilation (8,28) or me-
chanical disruption (8,10) of the fi-
brin sheath as an adjunct during
catheter exchange; this has not
positioning the tip more centrally or intervention. They found no differ-
by manipulating a guide wire and ence in malfunction rates between

Gray et al ● 1129
Volume 11 Number 9
been studied in a systematic fash-
ion either.
5. Egglin TKP, Rosenblatt M, Dickey
KW, Houston JP, Pollak JS. Re-
placement of accidentally removed
tunneled venous catheters through
existing subcutaneous tracts. JVIR
17. Schwartz RD, Messana JM, Boyer
CJ, et al. Successful use of cuffed
central venous hemodialysis cathe-
ters inserted percutaneously. J Am
Soc Nephrol 1994; 4:1719–1725.
In conclusion, our study demon-
strates that both UK infusion and
sheath stripping allow a reasonable
period of additional function for
well-positioned central dialysis
catheters with poor flow rates. Un-
fortunately, there is little reported
experience with other thrombolytic
agents for catheter clearance. We
are beginning a catheter clearance
feasibility study to determine the
effectiveness and safety of tissue
plasminogen activator. Conceptu-
ally, tissue plasminogen activator
administered in dose-equivalent in-
fusions to UK should be similar in
effectiveness and safety and may be
faster than UK infusions. Although
we did not detect a significant dif-
ference in outcome between UK in-
fusion and sheath stripping, we pre-
fer a thrombolytic infusion because
it is noninvasive, preferred by pa-
tients, and safer. We reserve strip-
ping for rare cases in which throm-
bolytic infusion fails or is contrain-
dicated and catheter exchange or
replacement cannot be performed.
1997; 8:197–202.
1
8. Duszak R Jr, Haskal ZJ, Thomas-
Hawkins C, et al. Replacement of
failing tunneled hemodialysis cathe-
ters through pre-existing subcutane-
ous tunnels: a comparison of cathe-
ter function and infection rates for
de novo placements and over-the-
wire exchanges. JVIR 1998; 9:321–
6
. Crain MR, Mewissen MW, Os-
trowski GJ, Paz-Fumagalli R, Beres
RA, Wertz RA. Fibrin sleeve strip-
ping for salvage of failing hemodial-
ysis catheters: technique and initial
results. Radiology 1996; 198:41–44.
. Rockall AG, Harris A, Wetton CWN,
Taube D, Gedroyc W, Al-Kutoubi
MA. Stripping of failing haemodi-
alysis catheters using the Amplatz
gooseneck snare. Clin Radiol 1997;
7
3
27.
1
2
9. Hoshal VL Jr, Ause RG, Hoskins
PA. Fibrin sleeve formation on in-
dwelling subclavian central venous
catheters. Arch Surg 1971; 102:353–
358.
0. Ahmed N, Payne RF. Thrombosis
after central venous cannulation.
Med J Aust 1976; 1:217–220.
52:616–620.
8. Haskal ZJ, Leen VH, Thomas-
Hawkins C, Shlansky-Goldberg RD,
Baum RA, Soulen MC. Trans-
venous removal of fibrin sheaths
from tunneled hemodialysis cathe-
ters. JVIR 1996; 7:513–517.
21. Brismar BO, Hardstedt C, Jacobson
S. Diagnosis of thrombosis by cath-
eter phlebography after prolonged
central venous catheterization. Ann
Surg 1981; 194:779–783.
9
. Brady PS, Spence LD, Levitin A,
Mickolich CT, Dolmatch BL. Effi-
cacy of percutaneous fibrin sheath
stripping in restoring patency of
tunneled hemodialysis catheters.
AJR 1999; 173:1023–1027.
2
2. Bolz KD, Fjermeros G, Wideroe TE,
Hatlinghus S. Catheter malfunc-
tion and thrombus formation on
double-lumen hemodialysis cathe-
ters: an intravascular ultrasono-
graphic study. Am J Kidney Dis
1
0. Trerotola SO, Johnson MS, Harris
VJ, et al. Outcome of tunneled he-
modialysis catheters placed via the
right internal jugular vein by inter-
ventional radiologists. Radiology
1
995; 25:597–602.
1997; 203:489–495.
Acknowledgments: The authors
thank Mrs. Nancy Carnes for editorial
assistance and Mr. Peter Brandt for
preparation of the photographs.
2
3. Winn MP, McDermott VG, Schwab
SJ, Conlon PJ. Dialysis catheter
1
1
1. Lund GB, Trerotola SO, Scheel PF
Jr, et al. Outcome of tunneled he-
modialysis catheters placed by radi-
ologists. Radiology 1996; 198:467–
‘
fibrin-sheath stripping’: a caution-
ary tale! Nephrol Dial Transplant
1997; 12:1048–1050.
472.
References
2. Uldall R, Besley ME, Thomas A,
Salter S, Nuezca LA, Vas M.
24. O’Farrell L, Griffith JW, Lang CM.
Histologic development of the
sheath that forms around long-term
implanted central venous catheters.
J Parenteral Enteral Nutr 1996; 20:
156–158.
1
. NKF-DOQI Clinical Practice Guide-
lines for Vascular Access. New
York: National Kidney Foundation,
Maintaining the patency of double-
lumen silastic jugular catheters for
haemodialysis. Int J Artif Organs
1997.
2
. Moss AH, Vasilakis C, Holley JL,
Foulks CJ, Pillai K, McDowell DE.
Use of a silicone dual-lumen cathe-
ter with a dacron cuff as a long-
term vascular access for hemodialy-
sis patients. Am J Kidney Dis 1990;
1
993; 16:37–40.
1
3. Gray RJ, Fessahaye A, Gupta A, et
al. Percutaneous fibrin sheath
stripping versus Urokinase for mal-
functioning central dialysis cathe-
ters [abstract]. JVIR 1998; 9:164.
2
2
2
5. Oncay ST, Levitin A, Dolmatch BL,
Lieber M. Tunneled hemodialysis
catheter level and orientation pre-
dicts intervention. Radiology 1998;
2
09(P):453.
26:211–215.
14. McDowell DE, Moss AH, Vasilakis
C, et al. Percutaneously placed du-
al-lumen silicone catheters for long-
term hemodialysis. Am Surg 1993;
59:569–573.
15. Cappello M, DePauw L, Bastin G, et
al. Central venous access for hae-
modialysis using the Hickman cath-
eter. Nephrol Dial Transplant 1989;
4:988–992.
16. Gibson SP, Mosquera D. Five
years experience with the Quinton
Permcath for vascular access. Neph-
rol Dial Transplant 1991; 6:269–274.
6. Kowalski CM, Kaufman JA, Rivitz
SM, Geller SC, Waltman AC. Mi-
gration of central venous catheters:
Implications for initial catheter tip
positioning. JVIR 1997; 8:443–447.
7. Nazarian GK, Bjarnason H, Dietz
CA Jr, Bernadas CA, Hunter DW.
Changes in tunneled catheter tip
position when a patient is upright.
JVIR 1997; 8:437–441.
3
4
. Schwab SJ, Buller GL, McCann RL,
Bollinger RR, Stickel DL. Prospec-
tive evaluation of a dacron cuffed
hemodialysis catheter for prolonged
use. Am J Kidney Dis 1988; 11:166–
169.
. Suhocki PV, Conlon PJ Jr, Knelson
MH, Harland R, Schwab SJ. Silas-
tic cuffed catheters for hemodialysis
vascular access: thrombolytic and
mechanical correction of malfunc-
tion. Am J Kidney Dis 1996;
28. Mauro MA. Delayed complications
of venous access. Tech Vasc Inter-
vent Radiol 1998; 1:158–167.
28:379–386.