Steven S. Kang

Definition of venous reflux in lower-extremity veins

PURPOSE This prospective study was designed to determine the upper limits of normal for duration and maximum velocity of retrograde flow (RF) in lower extremity veins. METHODS Eighty limbs in 40 healthy subjects and 60 limbs in 45 patients with chronic venous disease were examined with duplex scanning in the standing and supine positions. Each limb was assessed for reflux at 16 venous sites, including the common femoral, deep femoral, and proximal and distal femoral veins; proximal and distal popliteal veins; gastrocnemial vein; anterior and posterior tibial veins; peroneal vein; greater saphenous vein, at the saphenofemoral junction, thigh, upper calf, and lower calf; and lesser saphenous vein, at the saphenopopliteal junction and mid-calf. Perforator veins along the course of these veins were also assessed. In the healthy volunteers, 1553 vein segments were assessed, including 480 superficial vein segments, 800 deep vein segments, and 273 perforator vein segments; and in the patients, 1272 vein segments were assessed, including 360 superficial vein segments, 600 deep vein segments, and 312 perforator vein segments. Detection and measurement of reflux were performed at duplex scanning. Standard pneumatic cuff compression pressure was used to elicit reflux. Duration of RF and peak vein velocity were measured immediately after release of compression. RESULTS Duration of RF in the superficial veins ranged from 0 to 2400 ms (mean, 210 ms), and was less than 500 ms in 96.7% of these veins. In the perforator veins, regardless of location, outward flow ranged from 0 to 760 ms (mean, 170 ms), and was less than 350 ms in 97% of these veins. In the deep veins, RF ranged from 0 to 2600 ms. Mean RF in the deep femoral veins and calf veins was 190 ms, and was less than 500 ms in 97.6% of these veins. In the femoropopliteal veins, mean RF was 390 ms, and ranged from 510 to 2600 ms in 21 of 400 segments; however, RF was less than 990 ms in 99% of these veins. Duration of RF was significantly longer in all three veins systems in patients (P <.0001 for all comparisons). With a cutoff value of more than 1000 ms rather than more than 500 ms, prevalence of abnormal RF in the femoropopliteal veins was significantly reduced, from 29% to 18% (P =.002). Thirty-seven vein segments (2.4%) had RF greater than 500 ms in the supine position, compared with less than 500 ms in 22 of these vein segments (59%) in the standing position. Of the 48 vein segments (3.1%) with RF greater than 500 ms in the standing position, RF was less than 500 ms in 6 of these vein segments (13%) in the supine position. Similar observations were noted in patient veins. There was no association between RF and peak vein velocity. Peak vein velocity had no significance in determining reflux. CONCLUSIONS The cutoff value for reflux in the superficial and deep calf veins is greater than 500 ms. However, the reflux cutoff value for the femoropopliteal veins should be greater than 1000 ms. Outward flow in the perforating veins should be considered abnormal at greater than 350 ms. Reflux testing should be performed with the patient standing.

Percutaneous ultrasound guided thrombin injection: A new method for treating postcatheterization femoral pseudoaneurysms

PURPOSE Since its introduction in 1991, ultrasound guided compression repair of postcatheterization femoral artery pseudoaneurysms has been shown to be effective. Disadvantages of ultrasound guided compression repair include patient discomfort during compression, inability to treat noncompressible pseudoaneurysms, prolonged use of ultrasound equipment and personnel, limited success with patients being treated with anticoagulants, and some early recurrences. We conducted a prospective study to evaluate a new method of treating femoral pseudoaneurysms, percutaneous ultrasound guided thrombin injection. METHODS Under duplex ultrasound guidance, a 22- or 25-gauge needle was percutaneously positioned within the pseudoaneurysm. Without compressing the pseudoaneurysm, 0.5 to 1 ml thrombin solution (1000 U/ml) was injected to induce thrombosis. Early in the study, the procedure was modified to allow more than one injection. After successful thrombosis, the patients were kept at rest in bed for at least 1 hour. Duplex ultrasound examination was repeated in 1 to 4 days. Distal pulses and ankle-brachial indexes were measured before and after the procedure. RESULTS Twenty of 21 consecutive pseudoaneurysms were successfully treated with thrombin injection. Fifteen pseudoaneurysms thrombosed immediately (<20 seconds) after one injection. The other five had partial thrombosis after one injection and complete thrombosis immediately after a second injection. The one failure occurred in a patient who had only one injection and then underwent subsequent ultrasound guided compression repair, which failed. No patient required sedation or analgesia during thrombin injection. There were no procedure-related complications and no recurrences. CONCLUSIONS Percutaneous ultrasound guided thrombin injection appears to be a safe and expeditious method for treating postcatheterization femoral pseudoaneurysms. It has significant advantages with respect to ultrasound guided compression repair or surgical repair.

EXPANDED INDICATIONS FOR ULTRASOUND-GUIDED THROMBIN INJECTION OF PSEUDOANEURYSMS

PURPOSE We previously reported preliminary data on a new procedure that we developed for the treatment of femoral pseudoaneurysms after catheterization. This study presents our current results of percutaneous ultrasound-guided thrombin injection for treating pseudoaneurysms that arise from various locations and causes. METHODS Between February 1996 and May 1999, we performed thrombin injection of 83 pseudoaneurysms in 82 patients. There were 74 femoral pseudoaneurysms: 60 from cardiac catheterization (36 interventional), seven from peripheral arteriography (four interventional), five from intra-aortic balloon pumps, and two from dialysis catheters. There were nine other pseudoaneurysms: five brachial (two cardiac catheterization, two gunshot wounds, one after removal of an infected arteriovenous graft), one subclavian (central venous catheter insertion), one radial (arterial line), and one distal superficial femoral and one posterior tibial (both after blunt trauma). Twenty-nine pseudo-aneurysms were injected while on therapeutic anticoagulation. Patients underwent repeat ultrasound examination within 5 days and after 4 weeks. RESULTS Eighty-two of 83 pseudoaneurysms had initial successful treatment by this technique, including 28 of 29 in patients who were undergoing anticoagulation therapy. The only complication was thrombosis of a distal brachial artery, which resolved spontaneously. There were early recurrences in seven patients: four patients underwent successful reinjection; reinjection failed in two patients, who underwent surgical repair; and one patient had spontaneous thrombosis on follow-up. After 4 weeks, ultrasound examinations were completely normal or showed some residual hematoma, and there were no recurrent pseudoaneurysms. CONCLUSION Ultrasound-guided thrombin injection of pseudoaneurysms has excellent results, which support its widespread use as the primary treatment for this common problem.

Where does venous reflux start

PURPOSE This study was designed to identify the origin of lower limb primary venous reflux in asymptomatic young individuals and to compare patterns of reflux with age-matched subjects with prominent or clinically apparent varicose veins. METHODS Forty age- and sex-matched subjects with no symptoms (age, 15 to 35 years; 80 limbs; group A), 20 subjects (age, 19 to 32 years; 40 limbs) with prominent but nonvaricose veins (n = 26 limbs; group B), and 50 patients (age, 17 to 34 years; 100 limbs) with varicose veins (n = 64; group C) were examined with color flow duplex imaging. All proximal veins (above popliteal skin crease), superficial, perforator, and deep, in the lower limb were examined in the standing position, and all the distal veins in the sitting position. Patients who had a documented episode of superficial or deep vein thrombosis, previous venous surgery, or injection sclerotherapy were excluded from the study. RESULTS The prevalence of reflux in group A was 14% (11 of 80), in group B 77% (31 of 40), and in group C 87% (87 of 100). In more than 80% of limbs in the three groups, reflux was confined to the superficial veins alone. Deep venous reflux or combined patterns of reflux were uncommon even in group C. Reflux was detected in all segments of the saphenous veins and their tributaries. In the 125 limbs that had superficial venous incompetence, the below-knee segment of the greater saphenous vein was the most common site of reflux (85, 68%), followed by the above-knee segment of greater saphenous vein (69, 55%) and the saphenofemoral junction (41, 32%). Nonsaphenous reflux was rare (3, 2.4%). Reflux in the lesser saphenous vein (21, 17%) was seen in all groups, whereas involvement of both greater and lesser saphenous veins (8, 6.4%) was seen in group C alone. The incidence of multisegmental reflux was significantly higher in group C (61 of 64, 95%) than in group A (two of 11, 18%) or group B (14 of 26, 54%). The prevalence of distal reflux was comparable in all groups. CONCLUSIONS Primary venous reflux can occur in any superficial or deep vein of the lower limbs. The below-knee veins are often involved in asymptomatic individuals and in those who have prominent or varicose veins. These data suggest that reflux appears to be a local or multifocal process in addition to or separate from a retrograde process.

Biointeractive polymers and tissue engineered blood vessels

The regulation of endothelial cell (EC) and smooth muscle cell (SMC) proliferation following vascular interventions is critical to clinical efficacy. Our laboratory has developed a method of impregnating biomaterials with suspensions containing bioactive proteins resulting in the capability of differentially modulating EC and SMC growth in vitro and in vivo following implantation. We have previously reported that 60 mu internodal distance ePTFE grafts impregnated with fibrin glue (FG) containing FGF-1 and heparin develop confluent endothelialization with transiently increased EC and SMC proliferation after 4 weeks in dogs. Thoraco-abdominal implants after 20 weeks were developed significantly thicker (139 mu) inner capsules in response to the FGF. To minimize SMC proliferation we studied the effects of FGF-1, heparin, and thrombin concentrations on SMC growth in vitro. FG caused a 182% increase (P < 0.001) in DNA synthesis. Heparin within FG diminished this effect in a dose-dependant manner, with complete inhibition of FG-induced growth at 500 U ml-1 (versus FG alone, P < 0.001). FGF-1 within FG without heparin had no effect, but together, FGF-1 caused a dose-dependant growth increase while increasing heparin concentrations initially increased and then decreased proliferation. FGF-1 and heparin in the medium of quiescent SMCs had similar effects. Only thrombin concentrations > 3.2 U ml-1 stimulated SMC growth and this stimulation was blocked by heparin. A synergism between FGF and heparin on EC proliferation was also found but without EC growth inhibition in response to higher concentrations of heparin. It is thus possible to modulate the relative proliferative activity of ECs versus SMCs by altering the FGF:heparin ratio. This same system may be useful with other proteins to induce other local affects by the applied protein or systemic affects following release of that protein.

Carotid endarterectomy for recurrent stenosis

PURPOSE The purpose of this study was to report our results in the surgical management of recurrent carotid stenosis (RCS) after carotid endarterectomy (CEA). METHODS In a 20-year period, we performed 1209 CEAs; 82 operations (6.8%) were for RCS. There were 33 men and 36 women, with an average age of 66.3 years. Nine patients underwent two redo CEAs and two patients underwent three redo CEAs for either bilateral recurrence or a second recurrence on the same side. Overall, 10 patients were identified with a second recurrence. RESULTS The average time to presentation with RCS was 65 months (range, 3 to 361 months). The majority of patients (66%) were symptomatic, 34% had transient ischemic attacks, 17% had amaurosis fugax, 9% had strokes, and 6% had nonhemispheric symptoms. Before repair, angiograms were obtained. Patch repair was performed in 61 procedures (74%), 41 with vein, 11 with Dacron, and nine with polytetrafluoroethylene. Autogenous or synthetic bypass grafts were used in 20 procedures (24%), vein in eight, Dacron in two, and polytetrafluoroethylene in 10. In one patient, an occluded internal carotid artery was ligated and an endarterectomy of the external carotid artery was performed without a patch. The operative stroke rate was 4.8%. Minor complications included transient or permanent cranial nerve deficits in 7.3% and wound hematomas in 2.4%. CONCLUSION Although repeat endarterectomy to treat RCS is technically more demanding, it can be performed safely. Long-term follow-up examination shows that a second recurrence may develop, and we recommend serial noninvasive testing.

Patterns and distribution of isolated calf deep vein thrombosis

PURPOSE In the search for calf deep vein thrombosis (DVT) with color-flow duplex scanning (CFDS), most vascular laboratories investigate only the posterior tibial and peroneal veins. Few laboratories assess the soleal and gastrocnemial veins. This study was designed to determine the patterns and distribution of isolated calf DVT, including the soleal and gastrocnemial veins. METHODS In the last 3 years, 5250 patients (mean age, 66 +/- 15 years; range, 22 to 93 years) were referred to the vascular laboratory for clinical suspicion of DVT and underwent examination with CFDS. All superficial and deep named veins, excluding the anterior tibial from groin to ankle, were imaged. Of the deep veins in the calf, the peroneal, the posterior tibial, the gastrocnemial, and the soleal veins were examined throughout their length. RESULTS DVT was detected in 14% of the patients. Isolated calf DVT was detected in 282 limbs of 251 patients (4.8%). No significant difference was noted for the sex (114 men vs 137 women; P =.15) or the limb preference (145 left vs 137 right; P =.5). The peroneal veins were most frequently involved, with 115 limbs (41%) affected. The soleal veins were involved in 109 limbs (39%), followed by the posterior tibial in 105 limbs (37%) and the gastrocnemial in 79 limbs (29%). Thrombus in the soleal vein alone was found in 57 limbs (20%), in the gastrocnemial in 48 limbs (17%), in the peroneal in 41 limbs (15%), and in the posterior tibial vein in 35 limbs (12%). Thrombus confined to a single or paired vein was found in 181 limbs (64%). Thrombus involving two different veins (27%) was the second most frequent pattern, and thrombus in three (7%) or four (1.4%) different veins was less prevalent. Isolated thrombosis in veins not routinely investigated was found in 113 limbs (40%; soleal, n = 57; gastrocnemial, n = 48; soleal + gastrocnemial, n = 8). Multifocal origin of thrombosis, defined as thrombi in two different veins that do not anatomically communicate, was identified in 63 limbs (22%). CONCLUSION Forty percent of the patients with acute isolated calf DVT would be judged to have normal CFDS examination results if the muscular veins in the calf were not imaged. Multifocal origin of thrombosis was found in 22% of the involved limbs. The prevalence of thrombosis in any calf vein either alone or in combination is comparable. Accordingly, the soleal and gastrocnemial veins should be examined routinely.

Compensatory arterial enlargement is a common pathobiologic response in early atherosclerosis

BACKGROUND Human arteries are dynamic conduits that respond to different stimuli by remodeling their structure and size. Arterial dilatation has been shown to occur in moderate and advanced atherosclerosis in studies that evaluated only one artery, either coronary, carotid, or superficial femoral artery (SFA). The purpose of this study was to quantify and compare compensatory arterial enlargement throughout the peripheral vascular system in early atherosclerosis. METHODS Seventy-two patients (40 male, 32 female, mean age 67 +/- 12 years) underwent transcutaneous B-mode ultrasound imaging during routine examinations. Thirty-nine carotid, 19 aorta, 19 iliac, 23 common femoral (CFA), 21 SFA, and 23 popliteal arteries were longitudinally imaged. Eight healthy volunteers (6 male, 2 female, mean age 27 +/- 2.2 years) had the same arteries evaluated (n = 48). Internal diameter (ID) and external diameter (ED) were measured in disease-free areas and in paired adjacent areas exhibiting increased intima-media thickening (IMT) and small atherosclerotic plaques. The percent change in ID, ED, IMT, and plaque thickness were calculated. RESULTS There was no observed change in ID or ED in all arteries of the healthy volunteers. When compared with normal vessel segments, all arteries demonstrated a marked decrease in ID and increase in ED in areas of small, hemodynamically insignificant plaque. The aorta had a 6.00% +/- 1.92% increase in ED, which was significantly less than the percent increase in ED observed in carotid (8.14 +/- 4.5%. P = 0.05), CFA (9.73 +/- 3.54%, P = 0.0001), SFA (9.15 +/- 4.25%, P = 0.005), and popliteal arteries (9.67 +/- 4.34, P = 0.002). In all arteries there was a strong correlation between plaque thickness and percent change in ED with the best correlation observed in the popliteal artery (R2 = 0.823, P < 0.0001). IMT was significantly increased in all normal vessel segments of the patients when compared with the healthy volunteers (P < 0.001). CONCLUSION All peripheral arteries dilate in response to intima-media thickening and early atherosclerotic plaque formation. This adaptive response occurs at the site of the lesion to preserve luminal area. The percent change in ED is strongly related to plaque thickness and is greatest in the more distal arteries.

Selective stimulation of endothelial cell proliferation with inhibition of smooth muscle cell proliferation by fibroblast growth factor-1 plus heparin delivered from fibrin glue suspensions

BACKGROUND Pretreatment of expanded polytetrafluoroethylene grafts with fibrin glue (FG) containing fibroblast growth factor-1 (FGF-1) (10 ng/ml) and heparin (50 units/ml) has been shown to induce a transmural angiogenesis with proliferation of both endothelial cells (ECs) and smooth muscle cells (SMCs) in dogs. To induce EC without SMC proliferation, we studied the effects of different FGF-1:heparin ratios within FG in vitro. METHODS First passage human umbilical vein ECs (factor VIII+) or primary canine carotid artery SMCs (alpha-actin +) were seeded onto 96-well plates coated with FG containing 10 ng/ml FGF-1 and 0, 5, 50, or 500 units/ml heparin. Control wells were coated with FG without FGF-1 or heparin. Cells were fed standard growth medium without soluble FGF-1 or heparin. Tritiated thymidine (1 microCi/well) was added after 1, 2, or 3 days, and proliferation was assayed by scintillation counting 48 hours later. RESULTS For both ECs and SMCs, proliferation on FG containing FGF-1 but no heparin was not different from control. EC proliferation on FG containing FGF-1 was significantly increased by addition of 5, 50, or 500 units/ml heparin (+68%, +99%, and +106%, respectively; p (0.0001 for all), reflecting the synergism of FGF-1 by heparin. SMC proliferation was also significantly increased by the addition of 5 or 50 units/ml heparin (+85% and +66%, respectively; p (0.0001 for both). However, SMC proliferation with 500 units/ml heparin was significantly decreased from control (-12%; p = 0.014), reflecting heparins SMC growth inhibitory activity. CONCLUSIONS FG containing 10 ng/ml FGF-1 and 500 units/ml heparin stimulates EC proliferation while inhibiting SMC proliferation in vitro. Application of this modified FG to vascular grafts or to arteries after direct or transcutaneous interventions may promote endothelialization without intimal hyperplasia.

Outcome of moderate carotid artery stenosis in patients who are asymptomatic

PURPOSE The incidence rate of disease progression and stroke after the diagnosis of a moderate (50% to 79%) carotid stenosis was determined by means of color-flow duplex scanning. METHODS During a 4-year period, 344 male veterans with moderate internal carotid artery stenoses, on one or both sides, were examined at regular intervals for a mean period of 25 months. Carotid color-flow scans were obtained semiannually. Clinical follow-up was performed to determine the incidence rate of amaurosis fugax, transient ischemic attacks, nonhemispheric symptoms, and strokes. RESULTS New neurologic symptoms developed in 75 patients (21.8%). Fifty-one (14.8%) had ipsilateral symptoms during follow-up: 18 amaurosis fugax (5.2%), 14 transient ischemic attacks (4%), 5 nonhemispheric symptoms (1.4%), and 14 strokes (4%). Twenty-four patients (6.9%) had contralateral symptoms: 20 strokes (5.8%) and 4 transient ischemic attacks (1.2%). Life-table analysis showed that the annual rate of ipsilateral neurologic events was 8.1%, and the annual rate of stroke was 2.1%. Seventy-five patients (22%) died in the follow-up period. Disease progression to 80% to 99% stenosis or occlusion occurred in 71 of 458 vessels (15.5%). The internal carotid arteries that showed evidence of disease progression had a significantly higher initial peak systolic velocity (251 vs 190 cm/s; P <.0001) and end diastolic velocity (74 vs 52 cm/s; P < 0.0001). Black patients and patients with ischemic heart disease were at a higher risk for disease progression. We could not identify any atherosclerotic risk factors that reliably predicted patients in whom future ipsilateral neurologic symptoms were more likely to develop. However, there was an increased risk of stroke associated with progression of disease. CONCLUSION Patients who are asymptomatic and who have moderate carotid stenoses are at significant risk for neurologic symptoms and death, but have a relatively low incidence rate of ipsilateral events. The initial flow characteristics in the stenotic vessel are predictive of future disease progression, but they are not helpful in identifying patients in whom symptoms will develop.