Late incidence of chronic venous insufficiency after deep vein harvest
Article Outline
Background
The deep veins (DV) of the thigh have proven to be versatile autogenous conduits for arterial reconstruction. Harvesting DV poses a theoretical risk of compromising venous outflow of the limb, which could predispose to chronic venous morbidity. The purpose of this study was to define the late incidence of chronic venous insufficiency (CVI) and to characterize the long-term alterations in venous physiology after DV harvest.
Methods
Since 1991, 269 patients have undergone arterial reconstructions using DV at our facility. Patients with DV harvest at least 43 months prior to the study (n = 151) were eligible for inclusion. Eighty-nine patients were excluded (deceased = 70; lost to follow-up = 19). Forty-six patients who declined formal testing were queried by phone for signs and symptoms of CVI. The current study presents a case-control series of 16 patients (27 limbs) after DV harvest and six age- and gender-matched control patients (12 limbs) who underwent examination and venous testing.
Results
At a mean follow-up of 70.1 ± 5.6 months, 23 of 27 limbs (85.2%) had no significant CVI (CEAP C0 to C2). Four limbs (14.8%) had significant venous morbidity (C3 to C6), including edema alone (C3; n = 2 limbs), edema with skin changes (C4; n = 1 limb), and a healed venous ulceration (C5; n = 1 limb). APG testing confirmed relative venous outflow obstruction after DV harvest (mean outflow fraction: harvested limbs = 38.4 ± 3.9% vs control limbs = 51.7 ± 4.3%; P = .04). Despite the relative outflow obstruction, the mean VFI was not significantly different between harvested and control limbs (harvested limbs = 1.08 ± 0.15% vs control limbs = 0.77 ± 0.16%; P = .19). DV harvest resulted in no significant changes in calf ejection fraction (harvested limbs = 67.4 ± 6.4% vs control limbs = 86.8 ± 9.5%; P = .09) or residual volume fraction measured (harvested limbs = 32.3 ± 6.4% vs control limbs = 47.7 ± 11.6%; P = .22). Of the 46 patients interviewed by phone, five (10.9%) reported bilateral amputations, seven (15.2%) reported chronic edema in their harvested limbs (C3), and 34 (73.9%) reported no signs of CVI in their harvested limbs (C0).
Conclusions
Deep vein harvest produces few symptoms of chronic venous insufficiency, and venous ulceration is infrequent. Despite relative venous outflow obstruction, noninvasive indices of chronic venous insufficiency on APG are often normal, suggesting that the risk of developing venous ulceration is low in the majority of patients after DV harvest.
Harvesting of the deep veins (DV) of the thigh, including the femoral and popliteal veins, for use as conduits for arterial reconstruction was initially described by Schulman for femoral-popliteal arterial bypass.1, 2, 3, 4 Our group subsequently described the use of DV for creation of a neo-aortoiliac system (NAIS) for in situ aortic reconstruction after excision of infected aortic prosthetic grafts.5, 6, 7 DV has also been employed as an alternative conduit for brachiocephalic and mesenteric arterial reconstruction.8, 9 Although DV has proved to be relatively resistant to infection and offers excellent long-term patency, interruption of the venous drainage of the lower extremity by DV harvest remains a concern to many surgeons.
DV harvest impedes venous outflow from the leg and induces relative venous hypertension in the harvested limb.10 In the immediate postoperative period, this venous hypertension has been associated with compartment syndrome necessitating fasciotomy in 17.8% of limbs after DV harvest.11 The incidence of chronic venous morbidity remains undefined. We previously reported that approximately one-third of patients manifested mild leg swelling at a mean follow-up of 37 ± 3 months after DV harvest.10 Since there may be a lag time of several years between venous injury and the clinical onset of chronic venous insufficiency (CVI), quantifying venous morbidity after short- or intermediate-term follow-up may not accurately represent the long-term risk of developing CVI after DV harvest. The purpose of this study was to define the late incidence of chronic venous sufficiency and to characterize the alterations in venous physiology that persist several years after DV harvest.
Material and methods
Study population
Since 1991, 269 patients have undergone arterial reconstructions using DV at the University of Texas Southwestern Medical Center and its affiliated hospitals. Only patients with DV harvest at least 43 months prior to the study (n = 151) were eligible for inclusion. After exclusion of patients who were deceased (n = 70) or lost to follow-up (n = 19), 62 eligible patients were contacted by phone. The current study represents a case-control series of 16 patients (28 harvested limbs) and six age- and gender-matched control patients (12 limbs) without DV harvest who consented to participate. Patients who declined to participate in the experimental protocol (n = 46) were queried by phone regarding signs and symptoms of CVI. The study was conducted in accordance with a protocol approved by the institutional review board of the University of Texas Southwestern Medical Center.
Assessment for CVI
The standard protocol to assess for CVI included clinical examination, lower extremity venous duplex ultrasound, venous function testing, and ankle-brachial indices (ABI). Clinical symptoms were categorized according to the CEAP system for classification of lower-extremity venous disease.12, 13 Measurements of leg circumference were obtained at the thigh (14 cm above the knee joint), proximal and mid-calf (3 cm and 10 cm below the tibial tuberosity, respectively), and ankle.
Venous duplex ultrasound was performed with the patient in recumbent position using a 5.0 MHz probe (Acuson 128XP, Mountain View, Calif). Duplex exams included the great saphenous vein (GSV), small saphenous vein (SSV), and deep veins proximal and distal to the site of DV harvest. Venous reflux was assessed using the rapid cuff deflation duplex technique described by van Bemmelen et al.14 Testing was performed with the patient standing, but without weight-bearing on the limb. Reversal of flow was detected by duplex ultrasonography in the venous segment proximal to the cuff. Abnormal reflux was defined as reversal of venous flow exceeding 0.5 seconds. In this manner, reflux was assessed in the common femoral vein (CFV), profunda femoris vein (PFV), and posterior tibial veins (PTV). An assessment of reflux in the superficial veins (GSV and SSV) was also performed.
Postoperative venous function was assessed by air plethysmography (APG; Model APG-1000C, ACI Medical, San Marcos, Calif), as described by Christopoulos et al.15 Preoperative APG testing was not performed. Parameters quantified by APG included outflow fraction, venous filling Index (VFI), calf ejection fraction, and residual volume fraction.
Statistical analysis
All variables were analyzed by limb, except as indicated. Continuous data were expressed as mean ± SEM. Normally distributed data were compared between treatment groups using a Student t test. Nonparametric data were compared between groups using a Mann-Whitney test or Wilcoxon signed-ranks test as appropriate. Categorical data were analyzed using the Fisher exact test and Cochran-Armitage trend test. For all statistical analyses, the threshold for significance was .05. Statistical analysis was performed using SAS, version 9.13 (SAS Institute, Inc, Cary, NC).
Results
The current study examined 16 patients with 28 harvested limbs (“study patients”) and six age- and gender-matched control patients (12 limbs) without DV harvest (“control patients”). The cohort of patients with DV harvest included 12 (75%) males and four (25%) females. Control patients included four (66.7%) males and two (33.3%) females. The mean age of study patients was 58.3 ± 3.2 years, while the mean age of control patients was 64.2 ± 2.4 years (P = .29). The mean follow-up was for study patients was 70.1 ± 5.6 months (range 43 to 129 months, median 66.0 months). Indications for DV harvest included infected prosthetic aortic grafts (n = 7), primary aortoiliac reconstruction in young patients with premature atherosclerosis (n = 6), brachiocephalic reconstructions in contaminated operative fields (n = 2), and secondary aortoiliac reconstruction after a failed prosthetic aortofemoral bypass (n = 1). The mean preoperative ABI was 0.79 ± 0.07 for study patients and 0.84 ± 0.7 for control patients (P = .64). Five limbs (17.9%) had a history of prior GSV harvest in a limb in which DV harvest was performed.
The technique for DV harvest has been described in detail elsewhere.5, 16 DV harvest was defined as complete if the entire superficial femoral vein and any portion of the popliteal vein below the adductor hiatus were harvested. Complete DV harvest was performed in 26 of 28 (92.8%) limbs. Among study patients, DV harvest was performed bilaterally in 12 of 16 (75%) patients. Fasciotomy was required at the time of DV harvest in eight of 28 (28.6%) limbs. Concurrent GSV harvest was performed in two limbs (7.2%) undergoing DV harvest, and both limbs required fasciotomy. One limb required a below-knee amputation in the perioperative period, while a second limb had an above-knee amputation 43 months after bilateral DV harvest.
The incidence of venous disease is outlined in Table I. For comparison, Table I includes harvested limbs, unharvested limbs contralateral to a unilateral harvested limb, and control patients with no DV harvest in either extremity. After DV harvest, a minority of harvested limbs (four of 27; 14.8%) had clinical evidence of significant venous disease, defined as CEAP class C3 to C6 venous disease. The remaining harvested limbs had minimal or no venous disease. The most advanced venous disease after DV harvest included two limbs with edema alone (C3), one limb with edema and skin changes (dermatosclerosis and hemosiderin deposition; C4), and one limb with a healed venous ulcer (C5). The patient with a healed ulcer developed a small (1 cm diameter), unilateral venous ulceration 52 months after bilateral DV harvest for creation of a neo-aortoiliac system, and the ulcer healed with local wound care. The mean interval since DV harvest was 80.0 ± 16.4 months (range 65 to 129 months) for limbs with CEAP class C3 to C6 venous disease, which was not significantly different than the interval after DV harvest for limbs with CEAP class C0 to C2 venous disease (66.7 ± 5.5 months; P = .32). Among harvested limbs, there was not a statistically significant trend in the proportion of limbs with a higher class of venous disease (CEAP class C3 to C6), compared with control limbs (P = .11, Fisher exact test). Moreover, there was not an increasing trend across the ordinal values of venous disease among harvested limbs compared with control limbs (P = .36, Cochran-Armitage trend test).
Table I. Clinical classification of venous disease after DV harvest
| CEAP classification | Harvested limbs (n = 27) | Unharvested limbs⁎ (n = 4) | Control limbs† (n = 12) |
|---|---|---|---|
| C0 | 16 (59.3%) | 3 (75%) | 4 (33.3%) |
| C1 | 5(18.5%) | 1 (25%) | 8 (66.6%) |
| C2 | 2(7.4%) | 0 | 0 |
| C3 | 2(7.4%) | 0 | 0 |
| C4 | 1(3.7%) | 0 | 0 |
| C5 | 1(3.7%) | 0 | 0 |
| C6 | 0 | 0 | 0 |
⁎Unharvested limbs had a contralateral DV harvest. |
†Control limbs were from patients with no DV harvest in either lower extremity. |
To further assess the incidence of clinically apparent venous disease after DV harvest, phone interviews were obtained from patients who were unable to return for examination. Of the 46 patients interviewed by phone, 5 (10.9%) reported bilateral amputations, 7 (15.2%) reported chronic edema in their harvested limbs (CEAP class C3), and 34 (73.9%) reported no signs or symptoms of CVI in their harvested limbs (CEAP class C0).
The potential for DV harvest to induce chronic limb swelling was assessed by measuring limb circumference using a standard technique. For this analysis, unharvested limbs from patients with unilateral DV harvest were compared with harvested limbs in the same patient. There were no significant differences in the mean limb circumferences between harvested and unharvested limbs at the thigh, proximal calf, mid-calf, or ankle (Table II).
Table II. Limb circumference after DV harvest
| Measurement site | Harvested limbs | Unharvested limbs | P value |
|---|---|---|---|
| Thigh circumference (cm) | 58.9±3.7 | 56.6±2.9 | .25 |
| Proximal calf circumference (cm) | 36.6±1.8 | 35.6±2.0 | .25 |
| Mid-calf circumference (cm) | 38.1±2.2 | 36.5±2.0 | .13 |
| Ankle circumference (cm) | 22.6±0.9 | 22.0±0.9 | .13 |
Duplex ultrasound was used to assess for subclinical venous reflux using the rapid cuff deflation duplex technique described by van Bemmelen et al.14 Five limbs were excluded from this analysis due to a prior amputation (n = 2), inability to stand for the exam (n = 2), and patient time constraints (n = 1). By this technique, venous reflux was noted in 11 of 23 (47.8%) tested limbs after DV harvest (Table III). Among the limbs with reflux detected by this technique, the deep system (CFV, PFV, or PTV) was most commonly affected (n = 6). The superficial system alone (n = 4) or in combination with deep venous reflux (n = 1) demonstrated reflux in several limbs. Reflux was observed in one of four unharvested limbs in patients with a contralateral DV harvest. The reflux test was not performed on control patients.
Table III. Venous reflux after DV harvest⁎
| Site of venous reflux | Harvested limbs (n = 23) | Unharvested limbs† (n = 4) |
|---|---|---|
| No reflux | 12(52.2%) | 3(75%) |
| Deep venous reflux | 6(26.1%) | 0 |
| Superficial venous reflux | 4(17.4%) | 1(25%) |
| Deep and superficial reflux | 1(4.3%) | 0 |
⁎Venous reflux assessed by rapid cuff deflation duplex technique described by van Bemmelen. |
†Unharvested limbs include only unharvested limbs from patients with a unilateral DV harvest. |
To characterize the adaptation of the limb to DV harvest and to assess the potential for future development of venous disease, limbs were examined further using APG. The results of APG testing for harvested and unharvested limbs are outlined in Table IV. Three limbs were excluded from this analysis due to prior amputation (n = 2) or inability to comply with the protocol for the exam (n = 1). Relative venous obstruction was confirmed by the presence of a significantly lower outflow fraction in harvested legs, compared with unharvested and control limbs (Table IV; P = .04). The mean values for VFI, residual volume fraction, and calf ejection fraction were not significantly different between harvested and unharvested limbs (Table IV). More careful scrutiny of APG results noted that two of 25 harvested limbs had an abnormal VFI (1.8 and 3.6 mL/sec). In addition, a minority of harvested limbs had an abnormal residual volume fraction (n = 4) or abnormal calf ejection fraction (n = 6). There was no significant difference in the mean follow-up for limbs with abnormal APG results, compared with limbs with normal results for each APG parameter.
Table IV. APG results after DV harvest
| Parameter | Harvested limbs | Unharvested limbs⁎ | P value |
|---|---|---|---|
| Outflow fraction [nl >38%] | 38.4±3.9% | 51.7±4.3% | .04 |
| Venous filling index [nl = 0.5-1.7 ml/sec] | 1.08±0.15ml/sec | 0.77±0.16ml/sec | .19 |
| Residual volume fraction [nl = 2-35%] | 32.3±6.4% | 47.7±11.6% | .22 |
| Ejection fraction [nl = 60-90%] | 67.4±6.4% | 86.8±9.5% | .09 |
⁎Unharvested limbs included control limbs and unharvested limbs from patients with a contralateral DV harvest. |
Patients with unilateral DV harvest (n = 4) were assessed for differences in venous function between the harvested and unharvested limbs within the same patient. No significant differences in the mean VFI or calf ejection fraction between harvested and unharvested limbs were identified (data not shown).
The patient with a healed ulcer (CEAP class C5) was carefully examined. On APG, the limb with a healed ulcer had an outflow fraction that was on the lower margin of normal (34.5%) normal VFI (1.2 mL/sec), and a modestly decreased calf ejection fraction (46.9%). The contralateral limb was normal in appearance, but yielded plethysmographic data that were comparable to the limb with prior ulceration (data not shown).
Several patient and operative factors were examined for possible association with the late development of chronic venous disease, including concurrent DV and GSV harvest, history of prior ipsilateral GSV harvest, diminished preoperative ABI, indication for surgery, and extent of DV harvest (complete vs subtotal). We found that concurrent GSV harvest was associated with an increased incidence of significant postoperative CVI (C3 to C6; P = .02) since both limbs with concurrent GSV harvest developed persistent postoperative edema (CEAP class C3). In contrast, a history of prior GSV harvest in a limb undergoing DV harvest was not associated with late venous morbidity (P = .56), as three limbs with prior GSV harvest had no venous disease (CEAP class C0) and two limbs with prior GSV harvest had telangectasias alone (CEAP class C1) after DV harvest. There was no significant difference in ABIs among limbs with minimal venous disease (C0 to C2) after DV harvest, compared with those limbs with class C3 to C6 venous disease (C0 to C2 = 0.80 ± 0.07; C3 to C6 = 0.67 ± 0.22; P = .52). The indication for surgery also had no effect on the incidence of CVI (P = .49). Subtotal DV harvest produced no chronic venous disease, although there are relatively few patients (n = 2) who underwent subtotal DV harvest in this series. Thus, concurrent GSV harvest was the only factor that is clearly associated with significant late venous morbidity.
Discussion
DV harvest impedes venous outflow from the leg and induces relative venous hypertension in the harvested limb.10 The immediate implication of venous hypertension is a risk of compartment syndrome requiring fasciotomy in a minority (17.8%) of harvested limbs.11 The delayed effects of DV harvest have not been well defined. At intermediate-term follow-up (37 ± 3 months), less than one-third of patients experienced chronic mild edema without skin changes (CEAP class C3). However, we were concerned that there may be a lag time of several years after DV harvest before the full extent of any venous morbidity would be apparent clinically. The purpose of the current study was to define the late incidence of CVI after DV harvest by focusing exclusively on patients with DV harvest at least 43 months prior to the study. An additional goal of the study was to characterize the alterations in the venous physiology of the leg after DV harvest to estimate the perceived risk of future CVI.
Clinically significant venous disease, defined as CEAP class C3 to C6, occurred in 14.8% of harvested limbs during a mean follow-up of 70.1 ± 5.6 months after DV harvest (Table I). Only one limb developed venous ulceration, which healed with local wound care (CEAP class C5). Clinically apparent venous disease after DV harvest was sufficiently infrequent that we were unable to demonstrable a statistical association between DV harvest and the incidence of late postoperative venous disease. Moreover, mean limb circumference was no different between harvested and unharvested limbs when measured at the thigh, proximal calf, mid-calf, or ankle (Table II).
The only clear risk factor for developing significant venous disease (CEAP class C3 to C6) after DV harvest was concurrent GSV harvest at the time of DV harvest. Two limbs required concurrent GSV and DV harvest, and both limbs developed postoperative edema without skin changes (CEAP class C3). Despite the infrequency of concurrent GSV harvest, this factor was strongly associated with significant postoperative venous disease (P = .02). It is noteworthy that prior GSV harvest in an extremity undergoing DV harvest was not associated with an increased likelihood of developing significant venous disease postoperatively. No additional factors, including the indication for surgery, preoperative ABI, history of fasciotomy, and extent of DV harvest, were associated with an increased risk of postoperative venous disease.
Although clinically apparent, venous disease after DV harvest was uncommon at late follow-up, we found that nearly half (47.8%) of the harvested legs developed sonographic evidence of venous reflux (Table III). Interestingly, reflux was not always confined to the deep system, as the superficial system showed evidence of reflux in five of 11 limbs with reflux. We hypothesize that reflux in the deep and superficial systems alike is related to the relative venous outflow obstruction created by DV harvest, which induces pooling of blood and venous hypertension distal to the site of harvest.10 A potential consequence of this increased venous volume and hypertension may be poor apposition of the venous valve leaflets, causing valvular incompetence and venous reflux. This hypothesis is supported by our prior observation that there appears to be compensatory enlargement of the GSV after DV harvest,10 allowing the GSV to serve as a collateral outflow pathway in the harvested limb. This pathophysiology contrasts with the combination of both outflow obstruction and valvular injury that characterizes deep venous thrombosis (DVT),17, 18, 19 which may explain why DV harvest is tolerated with rare sequelae in the majority of patients.
Further insight into the adaptation of the leg to DV harvest was provided by our characterization of the venous physiology by APG (Table IV). Venous outflow obstruction induced by DV harvest was reflected in a relatively decreased outflow fraction (38.4 ± 3.9%), compared with unharvested limbs (51.7 ± 4.3%; P = .04). However, the ability of the limb to adapt to loss of the deep veins of the thigh by recruiting venous collateral tracts, as described by Raju20 and Masuda,21 was underscored by our observation that the outflow fraction in harvested limbs was less than published normal values (≥38%) in fewer than half of the test limbs. Despite the relative outflow obstruction induced by DV harvest, the mean VFI was not different between harvested and unharvested limbs. In fact, only two of 25 harvested limbs had abnormal venous filling indices, suggesting that reflux was uncommon after DV harvest. This conclusion appears to contradict our finding of venous reflux in nearly half of limbs assessed by the rapid cuff deflation duplex technique (Table III). This apparent discrepancy may be reconciled by recognizing that the two parameters represent different approaches to the assessment of reflux. The rapid cuff deflation technique quantifies reflux in a specific vein upon rapid deflation of a cuff. Only one vein may be assessed with each cuff deflation, so the veins of the limb are tested individually in sequence. A potential pitfall of the rapid cuff deflation technique is that it fails to account for compensation by the veins not being interrogated at the time of cuff deflation. By comparison, the VFI provides a more global assessment of reflux in the calf. The authors believe that the latter is more representative of the net effects of DV harvest on the limb. Moreover, VFI has been correlated with an increased risk of developing venous ulceration.15, 22 It should be noted that only one limb had a VFI >2 ml/sec, which is a threshold associated with a modest increase in risk of venous ulceration.22 With a VFI of 3.6 ml/sec, this limb would not fall into the group of limbs with the highest risk of venous ulceration (VFI >5 ml/sec).22
Calf muscle pump function has been implicated by some investigators as etiologic in the genesis of venous ulceration.23, 24 Although the mean ejection fraction for harvested limbs was less than unharvested limbs, the difference was not significant (Table IV). It has also been suggested that the combination of an increased VFI coupled with a decreased calf ejection fraction is particularly predictive of a risk of venous ulceration.23 In the current study, no limb met both criteria. Taken together, these APG results demonstrate the ability of the harvested limb to compensate for the loss of venous outflow, which would explain the lower rate of significant venous morbidity at late follow-up (14.8%) compared with intermediate follow-up (32%).10 The APG data failed to reveal any alarming trends in the APG parameters that have been most frequently associated with a future risk of developing venous ulceration.15, 22, 23
There are several shortcomings to the present study that are worthy of discussion. First, we assumed that any edema in a harvested limb was caused by the DV harvest. This assumption ignores other potential etiologies for limb edema, including postrevascularization edema unrelated to DV harvest, idiopathic CVI, and congestive heart failure. As such, we may be over-estimating the true incidence of significant CVI after DV harvest. In addition, the current study represents a small sampling of our entire series. Focusing exclusively on patients who underwent DV harvest at least 43 months prior to the study sharply reduced the pool of potential subjects. Further attrition was related to poor long-term survival and challenges in recruitment associated with geographic distance and patient social circumstances. Low enrollment raises the possibility of a risk of type II statistical error. The test most likely to be impacted by this confounding issue was the mean ejection fraction since the differences between harvested and unharvested limbs approached significance (P = .09). Despite this concern, any difference would not be viewed as clinically significant since the mean ejection fraction of harvested limbs (67 ± 6.4%) remained within normal limits (60% to 90%). A further concern relates to the time course for development of the most serious manifestations of CVI, such as venous ulceration. The mean follow-up for the study patients was 70.1 ± 5.6 months. The authors cannot exclude the possibility that a fraction of limbs that are currently without significant venous disease may progress and develop venous ulceration in the future. However, this possibility is unlikely based on the time course of CVI after deep venous thrombosis. In patients with deep venous thrombosis, CVI usually develops within 1 to 2 years after the initial venous insult, and it is rare to develop venous complications more than 5 years later.25, 26
In summary, DV harvest produces few symptoms of CVI, and venous ulceration is infrequent at late follow-up. Despite relative venous outflow obstruction, noninvasive indices of chronic venous insufficiency on APG are usually normal, suggesting that the risk of developing venous ulceration is low in the vast majority of patients after DV harvest. DV remains a viable option as an alternative conduit with few long-term sequelae for the limb.
Author Contributions
References
- . An 11-year experience with deep leg veins as femoropopliteal bypass grafts. Arch Surg. 1986;121:1010;-5.
- . Superficial femoral-popliteal veins and reverse saphenous veins as primary femoropopliteal bypass grafts: a randomized comparative study. J Vasc Surg. 1987;6:1–10
- . Unusual autogenous vein grafts. Vasc Surg. 1992;26:257;-64.
- . Superficial femoral vein: a useful autogenous harvest site. J Vasc Surg. 1994;20:947;-52.
- Creation of a neo-aortoiliac system from lower extremity deep and superficial veins. Ann Surg. 1993;218:239–249
- . Autogenous aortoiliac/femoral reconstruction from superficial femoral-popliteal veins: feasibility and durability. J Vasc Surg. 1997;25:255;-70.
- . Complex aortofemoral prosthetic infections: The role of autogenous superficial femoropopliteal vein reconstruction. Arch Surg. 1999;134:615;-21.
- . Superficial femoral-popliteal vein as a conduit for brachiocephalic arterial reconstructions. Ann Vasc Surg. 2002;16:17–23
- Comparison of superficial femoral vein and saphenous vein as conduits for mesenteric arterial bypass. J Vasc Surg. 2003;37:362;-6.
- Venous morbidity after superficial femoral-popliteal vein harvest. J Vasc Surg. 1998;29:282;-91.
- Deep vein harvest: predicting need for fasciotomy. J Vasc Surg. 2004;39:387;-94.
- . Reporting standards in venous disease: an update. J Vasc Surg. 1995;21:635;-45.
- Revision of the CEAP classification for chronic venous disorders: Consensus statement. J Vasc Surg. 2004;40:1248;-52.
- . Quantitative segmental evaluation of venous valvular reflux with duplex ultrasound scanning. J Vasc Surg. 1989;10:425;-35.
- . Venous reflux: quantitation and correlation with the clinical severity of chronic venous disease. Br J Surg. 1988;75:352–356
- . Harvesting the superficial femoral vein as an autograft. Sem Vasc Surg. 2000;13:257;-64.
- . The investigation and surgery of the post-thrombotic syndrome. J Cardiovasc Surg. 1986;27:5–16
- . The post-thrombotic syndrome. Ann R Coll Surg Engl. 1970;47:92–105
- . ABC of valvular diseases: deep venous insufficiency and occlusion. Br Med J. 1992;304:107;-10.
- . Venous obstruction: an analysis of one hundred thirty-seven cases with hemodynamic, venographic, and clinical correlations. J Vasc Surg. 1991;14:305;-13.
- . Long-term effects of superficial femoral vein ligation: thirteen-year follow-up. J Vasc Surg. 1992;16:741;-9.
- . The role of air plethysmography in the diagnosis of chronic venous insufficiency. J Vasc Surg. 1988;27:660;-70.
- . Investigation of patients with deep vein thrombosis and chronic venous insufficiency. London: Med-Orion; 1991;. p. 47-9.
- The significance of calf muscle function in venous ulceration. J Vasc Surg. 1994;20:872;-7.
- Prevention and treatment of postphlebitic syndrome and venous ulcer: results of a 3-part study. Arch Intern Med. 2001;161:2105;-9.
- The long-term clinical course of acute deep venous thrombosis. Ann Intern Med. 1996;125:1;-7.
Competition of interest: none
PII: S0741-5214(07)00750-1
doi:10.1016/j.jvs.2007.04.061
© 2007 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
