Prospective randomized trial of carotid endarterectomy with primary closure and patch angioplasty with saphenous vein, jugular vein, and polytetrafluoroethylene: long-term follow-up.
ABSTRACT This study examines the long-term clinical outcome and the incidence of recurrent stenosis (> or = 50%) after carotid endarterectomy (CEA) with primary closure (PC) versus vein patch closure (VPC), saphenous (SVP), and jugular vein (JVP) and polytetrafluoroethylene patch closure (PTFE-P).
A total of 399 CEAs were randomized into the following groups: 135 PC, 134 PTFE-P, and 130 VPC (SVP alternating with JVP). Postoperative duplex ultrasound scans were performed at 1, 6, and 12 months and every year thereafter. The mean follow-up was 30 months with a range of 1 to 62 months, and demographic characteristics were similar in all groups. Kaplan-Meier analysis was used to estimate the risk of restenosis and the stroke-free survival.
The incidence of ipsilateral stroke was 5% (seven of 135) for PC, 1% (one of 134) for PTFE-P, and 0% for VPC (PC vs VPC, p = 0.008; PC vs PTFE-P, p = 0.034). Seven strokes occurred in the perioperative period. All three groups had similar mortality rates. The cumulative stroke-free survival rate at 48 months was 82% for PC, 84% for PTFE-P, and 88% for VPC (p < 0.01 for PC vs PTFE-P or VPC). PC had a higher incidence of recurrent stenosis and occlusion (34%) than PTFE-P (2%) and VPC (9%) (SVP 9%, JVP 8%) (p < 0.001). PTFE-P had a lower recurrent stenosis rate than VPC (p < 0.045). Restenoses necessitating a redo CEA were also higher for PC (11%) than for PTFE-P (1%) and VPC (2%) (p < 0.001). Women with PC had a higher recurrent stenosis rate than men (46% vs 23%, p = 0.008). Kaplan-Meier analysis showed that freedom from recurrent stenosis at 48 months was 47% for PC, 84% for VPC, and 96% for PTFE-P (p < 0.001). The SVP and JVP results were comparable. The mean operative diameter of the internal carotid artery was similar in patients with or without restenosis. Significantly more late internal carotid artery dilatations occurred in the VPC group compared with the PC group.
Patch closure (VPC or PTFE-P) is less likely than PC to cause perioperative stroke. Patching was also superior in lowering the incidence of late recurrent stenoses, especially in women.
Article: 2011 ASA/ACCF/AHA/AANN/AANS/ACR/ASNR/CNS/SAIP/SCAI/SIR/SNIS/SVM/SVS Guideline on the Management of Patients With Extracranial Carotid and Vertebral Artery Disease: Executive summary: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines, and the American Stroke Association, American Association of Neuroscience Nurses, American Association of Neurological Surgeons, American College of Radiology, American Society of Neuroradiology, CCatheterization and Cardiovascular Interventions 01/2013; 81(1):E76-E123. · 2.51 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Background Multiple studies have been conducted that demonstrate the superiority of patch angioplasty over primary closure for carotid endarterectomy (CEA). Patch angioplasty with polytetrafluoroethylene patches (ACUSEAL) have shown results comparable to patch angioplasty with saphenous vein and polyester patches. This is the prospective randomized study to compare the clinical outcomes of CEA using ACUSEAL versus Bovine pericardium patching (Vascu-Guard). Methods Two hundred patients were randomized (1:1) to either ACUSEAL or Vascu-Guard patching. Demographic data/clinical characteristics were collected. Intraoperative hemostasis times and the frequency of re-exploration for neck hematoma were recorded. All patients received immediate and 1-month postoperative duplex ultrasound studies, which were repeated at 6-month intervals. A Kaplan-Meier analysis was used to estimate the risk of restenosis and the stroke-free survival rates. Results The demographics were similar in both groups, except for a higher incidence of current smokers in the ACUSEAL group and more patients with congestive heart failure in the Vascu-Guard group (P=0.02 and 0.03, respectively). The mean operative internal carotid artery diameter and the mean arteriotomy length were similar in both groups. The mean hemostasis time was 4.90 minutes for ACUSEAL patching versus 3.09 minutes for Vascu-Guard (P = 0.027); however, the mean operative times were similar for both groups (ACUSEAL 2.09 hours versus Vascu-Guard 2.16 hours, P=0.669). The incidence of re-exploration for neck hematoma was higher in the Vascu-Guard group; 6.12% versus 1.03% (P=0.1183). The incidence of perioperative ipsilateral neurologic events was 3.09% for ACUSEAL patching versus 1.02% for Vascu-Guard patching (P=0.368). The mean follow-up period was 15 months. The respective freedom from ≥70% carotid restenosis at 1, 2, and 3 years was 100%, 100%, and 100% for ACUSEAL patching versus 100%, 98%, and 98% for Vascu-Guard patching (P = 0.2478). The ipsilateral stroke free rates at 1, 2, and 3 years was 96% for ACUSEAL and 99% for Vascu-Guard patching. Conclusions Although CEA patching with ACUSEAL versus Vascu-Guard differed in hemostasis time, the frequency of re-exploration for neck hematomas was more frequent in the pericardial patch group; however, only one patient had documented suture line bleeding and the surgical reexploration rate likely is not likely to be patch related. There were not any significant differences in perioperative/late neurologic events and late restenosis in the two groups.Annals of Vascular Surgery 01/2014; · 0.99 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Carotid endarterectomy (CEA) with patch angioplasty produces greater results than with primary closure; however, there remains uncertainty on the optimal patch material in CEA. A systematic review of randomized controlled trials (RCTs) was performed to evaluate the effect of angioplasty using venous patch versus synthetic patch material, and Dacron patch versus polytetrafluoroethelene (PTFE) patch material during CEA. A multiple electronic health database screening was performed including the Cochrane library, Pubmed, Ovid, EMBASE and Google Scholar on all randomized controlled trials (RCTs) published before November 2012 that compared the outcomes of patients undergoing CEA with venous patch versus synthetic patch. RCTs were included if they compared carotid patch angioplasty with autologus venous patch versus synthetic patch material, or compared one type of synthetic patch with another. Thirteen RCTs were identified. Ten trials, involving 1946 CEAs, compared venous patch with synthetic patch materials. Two trials, involving 400 CEAs in 380 patients, compared Dacron patch with PTFE patch. The hemostasis time in CEA with PTFE patch was significantly longer than with venous patch (P<0.0001), and longer than with Dacron patch (P<0.0001). There was no significant difference of mortality rate, stroke rate, restenosis, and operative time in CEA with venous patch versus synthetic patch material, or in CEA with Dacron patch versus PTFE patch (all P>0.05). One RCT of 95 CEAs in 92 patients compared bovine pericardium with Dacron patch, and demonstrated a statistically significant decrease in intraoperative suture line bleeding with bovine pericardium compared with Dacron patch (P<0.001). The hemostasis time in CEA with PTFE patch was longer than with venous patch or Dacron patch. The overall perioperative and long-term mortality rate, stroke rate, restenosis, and operative time were similar when using venous patch versus synthetic patch material or Dacron patch versus PTFE patch material during CEA. More data are required to clarify differences between different patch materials.PLoS ONE 01/2013; 8(1):e55050. · 3.73 Impact Factor
Prospective randomized trial of carotid
endarterectomy with primary closure and
patch angioplasty with saphenous vein,
jugular vein, and polytetrafluoroethylene:
Ali F. AbuRahma, MD, Patrick A. Robinson, MD, S. Saiedy, MD, Jamal H.
Khan, MD, and James P. Boland, MD, Charleston, W.Va.
Purpose: This study examines the long-term clinical outcome and the incidence of recur-
rent stenosis (≥50%) after carotid endarterectomy (CEA) with primary closure (PC) ver-
sus vein patch closure (VPC), saphenous (SVP), and jugular vein (JVP) and polytetra-
fluoroethylene patch closure (PTFE-P).
Methods: A total of 399 CEAs were randomized into the following groups: 135 PC, 134
PTFE-P, and 130 VPC (SVP alternating with JVP). Postoperative duplex ultrasound
scans were performed at 1, 6, and 12 months and every year thereafter. The mean fol-
low-up was 30 months with a range of 1 to 62 months, and demographic characteristics
were similar in all groups. Kaplan-Meier analysis was used to estimate the risk of
restenosis and the stroke-free survival.
Results: The incidence of ipsilateral stroke was 5% (seven of 135) for PC, 1% (one of 134)
for PTFE-P, and 0% for VPC (PC vs VPC, p = 0.008; PC vs PTFE-P, p = 0.034). Seven
strokes occurred in the perioperative period. All three groups had similar mortality rates.
The cumulative stroke-free survival rate at 48 months was 82% for PC, 84% for PTFE-
P, and 88% for VPC (p < 0.01 for PC vs PTFE-P or VPC). PC had a higher incidence
of recurrent stenosis and occlusion (34%) than PTFE-P (2%) and VPC (9%) (SVP 9%,
JVP 8%) (p < 0.001). PTFE-P had a lower recurrent stenosis rate than VPC (p < 0.045).
Restenoses necessitating a redo CEA were also higher for PC (11%) than for PTFE-P
(1%) and VPC (2%) (p < 0.001). Women with PC had a higher recurrent stenosis rate
than men (46% vs 23%, p = 0.008). Kaplan-Meier analysis showed that freedom from
recurrent stenosis at 48 months was 47% for PC, 84% for VPC, and 96% for PTFE-P (p
< 0.001). The SVP and JVP results were comparable. The mean operative diameter of
the internal carotid artery was similar in patients with or without restenosis. Signifi-
cantly more late internal carotid artery dilatations occurred in the VPC group compared
with the PC group.
Conclusions: Patch closure (VPC or PTFE-P) is less likely than PC to cause perioperative
stroke. Patching was also superior in lowering the incidence of late recurrent stenoses,
especially in women. (J Vasc Surg 1998;27:222-34.)
The type of closure after a carotid endarterecto-
my (CEA), primary closure (PC) versus patch angio-
plasty, remains controversial.1-14Selection of patch
material is also controversial, with supporters for the
use of both vein patch (saphenous or neck veins) and
synthetic patch materials (polytetrafluoroethylene or
Dacron). Proponents of vein patch angioplasty
(saphenous vein patch [SVP]) state that the theoret-
ic benefits include an increase in luminal size and
From the Department of Surgery, Robert C. Byrd Health Sci-
ences Center of West Virginia University, Charleston Area
Presented at the Fifty-first Annual Meeting of The Society for
Vascular Surgery, Boston, Mass., June 1–2, 1997.
Reprint requests: Ali F. AbuRahma, MD, 3100 MacCorkle Ave.,
SE, Suite 603, Charleston, WV 25304.
Copyright © 1998 by The Society for Vascular Surgery and Inter-
national Society for Cardiovascular Surgery, North American
0741-5214/98/$5.00 + 0
JOURNAL OF VASCULAR SURGERY
Volume 27, Number 2
AbuRahma et al.
provision of endothelialized tissue to the endarterec-
tomy site. Because of this benefit it is believed that
vein patch angioplasty reduces the risk of periopera-
tive stroke, recurrent carotid stenosis, or occlu-
sion.4,6,7Others cite that they prefer an autogenous
vein over prosthetic materials because the luminal
surface is less thrombogenic and more resistant to
infection.13The long saphenous vein is the most
common autogenous source for carotid vein patch
angioplasty, but it is also in demand for coronary
artery bypass grafting and lower extremity revascu-
larization. To save the saphenous vein for these pro-
cedures, the internal jugular vein has been proposed
as an alternative for carotid vein patch angioplasty.13
Although many authorities prefer vein patching, sev-
eral issues have been raised with respect to availabil-
ity, increased operative time, morbidity related to
harvesting, vein patch rupture, and late aneurysmal
dilatation.15,16Opponents to synthetic patches fear
bleeding through the patch material, intraluminal
thrombus formation, and infection.17Many believe
that inclusion of a patch prolongs the operative time
and clamp or shunt time, makes the procedure tech-
nically more demanding, and may be unnecessary in
Several etiologic factors have been deemed
responsible for recurrent carotid stenosis after a
CEA, for example smoking, hypercholesterolemia,
female sex, hypertension, and PC of the endarterec-
Although several studies have compared the
results of PC with SVP closure,2-4,6-8,10,11PC versus
polytetrafluoroethylene patch closure (PTFE-
P),7,9,12and PC versus jugular vein patch (JVP) clo-
sure,13,14no prospective randomized controlled tri-
als have compared the results of CEA with PC ver-
sus PTFE-P, SVP closure, and JVP closure in one
series. The purpose of this randomized prospective
trial was to study the long-term clinical outcome and
the incidence of recurrent stenosis (≥50%) after CEA
with PC versus vein patch closure (SVP and JVP),
PTFE-P, or both. We have previously reported the
early perioperative results (30 days) of this study.19
PATIENTS AND METHODS
Between October 1991 and November 1995,
399 CEAs (357 patients) were entered into this
study. This group included 315 patients with unilat-
eral procedures and 42 patients with bilateral proce-
dures. Patients were asked to participate in this
prospective randomized trial of CEA with PC versus
PTFE-P (W. L. Gore & Associates, Inc., Elkton,
Md.) versus vein patch closure (VPC, SVP alternat-
ing with JVP). Patients scheduled for a repeat CEA,
a CEA with concomitant coronary artery bypass
grafting, or patients with internal carotid artery
diameters of fewer than 4 mm were excluded (12
CEAs, 3%). This study was approved by the Institu-
tional Review Board of Charleston Area Medical
Center/Robert C. Byrd Health Sciences Center of
West Virginia University.
Before surgery all patients underwent carotid
color duplex ultrasound and angiographic studies to
determine preoperative stenoses. They also under-
went tests for baseline blood cholesterol and triglyc-
eride levels. Preoperative risk factors, including
hypertension, diabetes mellitus, coronary artery dis-
ease, and smoking, were determined for each
patient, along with the preoperative use of aspirin or
dipyridamole. Indications for surgery were catego-
rized into hemispheric transient ischemic attacks
(TIAs), amaurosis fugax, hemispheric cerebrovascu-
lar accident, nonhemispheric TIA, and asympto-
matic carotid bruit.
Randomization included 135 PCs, 134 PTFE-
Ps, and 130 VPCs (70 SVP and 60 JVP). In the vein
group seven assigned neck vein cases were not done.
Four of these were changed to PC by the assigned
surgeon for no specific reason, and these were con-
sequently excluded from the study. Three were
changed to SVP closure because of an unsuitable
jugular vein in two cases and a radical neck dissec-
tion in one case, and these were included in the
All patients were administered aspirin therapy
(325 mg daily) within 24 hours after the operation.
Operative technique. All CEAs were performed
with the patients under general anesthesia with sys-
temic heparin and routine shunting with the use of
a carotid Argyle shunt (C. R. Bard, Inc., Billerica,
Mass.). At the time of surgery the normal internal
carotid artery distal to the lesion was measured in
millimeters with calipers. Saphenous vein was har-
vested from the ankle. Other details of the operative
procedure were previously described by us.19
Surveillance protocol. Surviving patients
underwent clinical follow-up and underwent imme-
diate postoperative color duplex ultrasound scan-
ning, which was repeated at 30 days, 6 months, 12
months, and every year thereafter with an ATL
Ultramark 8 and 9 HDI system (Advanced Technol-
ogy Laboratory, Inc., Bellevue, Wash.). Reportable
complications including death, TIA, reversible
ischemic neurologic deficits (RINDs), or stroke
morbidity and asymptomatic occlusive events were
determined in accordance with the North American
JOURNAL OF VASCULAR SURGERY
224 AbuRahma et al.
Chapter of the International Society of Cardiovascu-
lar Surgery/Society for Vascular Surgery Ad Hoc
Committee Suggested Standards for Reports Deal-
ing with Cerebrovascular Disease.20
Duplex scanning was used to assess the presence
of residual or recurrent stenoses and aneurysmal
dilation. Peak systolic frequencies of the internal
carotid artery greater than 4.5 KHz (or a peak sys-
tolic velocity >140 cm/sec) with spectral broaden-
ing throughout systole and an increased diastolic
frequency were consistent with hemodynamically
significant stenosis (≥50% diameter reduction).21
Recurrent stenosis was considered to be present
only if the abnormality detected by duplex ultra-
sound was not detected on the first immediate post-
operative duplex examination and if it persisted for
at least two examinations done within 6 months of
the original duplex examination. Four patients had
residual stenosis of 50% or more (two in the PC
group, one in the PTFE-P group, and one in the
VPC group). These were not included in the recur-
rent stenosis cases. None of these patients pro-
gressed except for one patient with PC who pro-
gressed to 80% or more stenosis at 18 months, and
this patient refused further intervention. Patients
with duplex findings consistent with 80% or more
stenosis or occlusion had their diagnoses confirmed
by magnetic resonance angiography, conventional
arteriography, or carotid exploration. Duplex ultra-
sound was also used to measure the diameter of the
proximal common carotid artery (low in the neck)
and the maximum diameter of the distal common
carotid artery (just before the bifurcation). Similar-
ly, the proximal internal carotid artery diameter
near its origin and the distal internal carotid artery
diameter were measured. Significant postoperative
dilatation was defined as a dilatation of more than
twice the diameter of the adjacent normal artery.
Statistical methods. The primary null hypothe-
sis was that no difference would be seen in effect
(≥50% recurrent stenosis) among each of three com-
parisons: PC versus PTFE-P, PC versus VPC, and
PTFE-P versus VPC. Therefore the nominal proba-
bility for these comparisons was adjusted with the
Bonferroni method for multiple comparisons, and
the statistical significance for each of these primary
comparisons was defined as p < 0.0167. Other com-
parisons were designated as secondary or explorato-
ry, and no adjustment was made for multiple com-
parisons (i.e., p = 0.05).
The analyses were performed on the patients as
randomized to the specific closure except for the fol-
lowing modifications. All patients who underwent
the assigned closure procedure were to be moni-
tored until the end of the trial. If a surgeon with-
drew his patient from the study before starting the
surgery, the patient was not included in the trial.
These patients had no follow-up available to the
investigators and may have violated randomization.
For example, four patients who were randomized to
JVP were removed from the trial and treated with
PC by their surgeons. Three patients who were ran-
domized to JVP closure had to undergo closure with
SVP; these patients were analyzed “as treated” as
patients who underwent SVP closure. An intent-to-
treat analysis was also performed with these patients
analyzed as randomized (as patients who underwent
JVP). No substantial changes in results were
observed. All patients were monitored for clinical
events and death.
The time to the occurrence of events (e.g., time
to ≥50% recurrent stenosis, time to stroke, or death)
was calculated with the method of Kaplan and
Meier. Statistical comparisons were made with the
Wilcoxon rank sum test. Statistical comparisons of
continuous data were examined with the unpaired
Student’s t test, and discrete variables were com-
pared with χ or Fisher’s exact tests. Baseline vari-
ables were compared among treatment groups with
an analysis of variance.
Potential risk factors were examined for their
effects on clinical outcome or stenosis of the carotid
closure procedure. Univariate analyses were exam-
ined for possible association with outcomes. The fac-
tors most associated with clinical or stenosis out-
comes were examined by multivariate analyses with
multiple linear regression.
As in our previous study,19no statistically signif-
icant differences were seen between the demograph-
ic and clinical data in the various groups. The mean
follow-up of 30 months (range, 1 to 62 months) was
also similar in all groups. The perioperative morbid-
ity and mortality rates were reported previously.19
The combined early and late neurologic com-
plications are summarized in Table I. Ipsilateral
strokes were statistically significantly higher in
patients with PC in contrast to those in patients
with VPC (p = 0.008) or PTFE-P (p = 0.034).
Eight patients had ipsilateral strokes, seven of
which occurred during the perioperative period.
Six of the patients with perioperative strokes were
in the PC group, and one was in the PTFE-P
group. One of these patients awoke in the operat-
ing room with a neurologic deficit and underwent
JOURNAL OF VASCULAR SURGERY
Volume 27, Number 2
AbuRahma et al.
an immediate exploration, which showed an artery
with thrombosis. A thrombectomy was performed,
and the artery was closed with PTFE-P with signif-
icant neurologic improvement. Four other patients
had strokes in the recovery room, where an imme-
diate duplex ultrasound examination confirmed
carotid thrombosis in three patients who under-
went thrombectomy and PTFE patch angioplasty
with significant improvement of their neurologic
deficits. The other patient had a normal duplex
ultrasound result and was treated with anticoagula-
tion therapy with some improvement of the neuro-
logic deficit. The remaining two strokes occurred
at 24 and 48 hours after surgery, and both patients
had normal duplex ultrasound scan results; there-
fore they were treated with anticoagulation with
some neurologic improvement in one and none in
Overall, 15 patients had ipsilateral TIAs, nine of
which occurred in the perioperative period (Table I).
Of these nine patients, three had PC, three had
PTFE-P, two had SVP closure, and one had JVP clo-
sure. All nine patients had a normal postoperative
duplex ultrasound except for one patient with PC.
Table I.Perioperative complications and late neurologic events
(n = 135; %)
(n = 134; %)
(n = 130; %)
(n = 70; %)
(n = 60; %)
and late events
*PC versus all patching (vein and PTFE), p = 0.007; PC versus VPC, p = 0.0165; VPC versus PTFE-P, p = 0.51. For strokes and
RIND combined, PC versus all patching, p = 0.04; PC versus VPC, p = 0.037.
†PC versus all patching, p = 0.003; PC versus PTFE-P, p = 0.034; PC versus VPC, p = 0.008. For strokes and RIND combined, PC
versus all patching, p = 0.019; PC versus PTFE-P, p = 0.11; PC versus VPC, p = 0.021.
Table II. Statistical comparisons of proportion of patients with combined neurologic complications* or
other major events
(n = 135; %)
(n = 134; %)
(n = 130; %)
(n = 70; %)
(n = 60; %)Event
events† and death
events‡ and ≥50%
events§ ≥50% recurrent
stenosis and death
eventsll, redo surgery,
32 (24) 18 (13) 18 (14) 8 (11) 10 (17)
52 (39) 9 (7)20 (15) 8 (11) 12 (20)
63 (47) 21 (16)31 (24) 13 (19) 18 (30)
38 (28) 18 (13) 20 (15)10 (14)10 (17)
*Includes TIA, RIND, and stroke (early and late combined).
†PC versus patching, p = 0.017; PC versus PTFE-P, p = 0.046; PC versus VPC, p = 0.058; PTFE-P versus VPC, p = 0.93; SVP versus
JVP, p = 0.54.
‡PC versus patching, p < 0.001; PC versus PTFE-P, p < 0.001; PC versus VPC, p < 0.001; PTFE-P versus VPC, p = 0.039; SVP ver-
sus JVP, p = 0.27.
§PC versus patching, p < 0.001; PC versus PTFE-P, p < 0.001; PC versus VPC, p < 0.001; PTFE-P versus VPC, p = 0.13; SVP versus
JVP, p = 0.188.
llPC versus patching, p = 0.0015; PC versus PTFE-P, p = 0.005; PC versus VPC, p = 0.018; PTFE-P versus VPC, p = 0.78; SVP ver-
sus JVP, p = 0.90.
JOURNAL OF VASCULAR SURGERY
226 AbuRahma et al.
This patient had more than 50% internal carotid
artery stenosis, but he refused further intervention
because he did not have any symptoms after the ini-
tial TIA, which lasted for a few minutes. The other
six patients had late TIAs, which were associated
with significant recurrent carotid stenoses in three
out of four patients with PC and less than 50%
stenosis in two patients with VPC.
Four patients had ipsilateral RIND, and all of
these situations occurred during the perioperative
period. Three of these patients had normal postop-
erative duplex ultrasounds, and one patient (with
PC) had postoperative carotid artery thrombosis.
This patient’s deficit was noticed 2 days after
surgery, and he was treated with anticoagulation
medication. The other three cases of RIND were
thought to be embolic in nature, and these patients
were treated with anticoagulation therapy.
When stroke, RIND, TIA, and death rates were
combined, patching (PTFE-P and VPC) was superi-
or to PC (p < 0.017; Table II). Patching was also
superior to PC in reducing all neurologic events
combined with 50% or more recurrent stenosis and
death (p < 0.001). Both VPC and PTFE-P were
superior to PC in this regard (p < 0.001). Similarly,
when all neurologic events, death, and redo carotid
surgery were combined, patching (PTFE-P, VPC, or
both) was better than PC (p = 0.0015; PTFE-P vs
PC, p = 0.005; VPC vs PC, p = 0.018) (Table II).
The Kaplan-Meier analysis showed that the
Fig. 1.Kaplan-Meier analysis shows cumulative stroke-free survival rates for PTFE-P,
VPC, and PC. Numbers at risk are shown at 6-month intervals.
Fig. 2. Kaplan-Meier analysis shows cumulative stroke-free survival rates for SVP and JVP.