The correlation of early flow disturbances with the development of infrainguinal graft stenosis: a 10-year study of 341 autogenous vein grafts

Initial in vitro and in vivo evaluation of a self-monitoring prosthetic bypass graft

OBJECTIVE

Prosthetic grafts used for lower extremity revascularization and dialysis access fail because of hyperplastic stenosis and thrombosis. Graft surveillance is advocated to monitor function; however, graft failure can occur between episodic examinations. An innovative sensor with wireless, microchip technology allows automated surveillance with assessment of graft function using a "cloud"-based algorithm. We performed proof-of-concept experiments with in vitro and in vivo models to assess the feasibility such a real-time graft surveillance system.

METHODS

A self-monitoring graft system was evaluated consisting of a prosthetic conduit of expanded polytetrafluoroethylene and a sensor unit, and a microsensor, microelectronics, battery, and remote processor with a monitor. The sensor unit was integrated on the extraluminal surface of expanded polytetrafluoroethylene grafts without compromise to the lumen of the conduit. The grafts were tested in vitro in a pulsatile, recirculating flow system under physiologic flow parameters. The hemodynamic parameters were varied to assess the ability to obtain wireless signal acquisition reflecting real-time flow properties in vitro. Segments of custom tubing with reduced diameters were inserted into the model to mimic stenosis proximal and distal to the grafts. After characterization of the initial data, the self-monitoring grafts were implanted in an ovine carotid model to assess proof of concept in vivo with 30-day follow-up of signal acquisition as well as arteriographic and histologic analysis.

RESULTS

In vitro flow data demonstrated the device was able to determine factors related to prosthetic graft function under varied hemodynamic flow conditions. Wireless signal acquisition using Bluetooth technology (Bluetooth SIG, Inc, Kirkland, Wash) allowed remote data analysis reflecting graft flow parameters through changes in microsensor voltage and frequency. Waveform analysis was applied to construct an algorithm using proprietary software and determine a parameter for graft flow characteristics. This algorithm allowed determination of the degree of stenosis and location of stenosis location (proximal or distal) for display on a remote monitor in real time. Subsequent in vivo experiments confirmed the ability of the system to generate signal acquisition through skin and soft tissue under biologic conditions with no arteriographic stenosis and a favorable healing response at 30-day harvest.

CONCLUSIONS

Initial in vitro and in vivo experiments demonstrate the ability for a self-monitoring graft system to remotely monitor hemodynamic parameters reflecting graft function using wireless data transmission. This automated system shows promise to deliver real-time data that can be analyzed by cloud-based algorithms alerting the clinician of a change in graft function or development of stenosis for further diagnostic study or intervention before graft failure.

Proliferative capacity of vein graft smooth muscle cells and fibroblasts in vitro correlates with graft stenosis

OBJECTIVE

About a quarter of peripheral vein grafts fail due in part to intimal hyperplasia. The proliferative capacity and response to growth inhibitors of medial smooth muscle cells and adventitial fibroblasts in vitro were studied to test the hypothesis that intrinsic differences in cells of vein grafts are associated with graft failure.

METHODS

Cells were grown from explants of the medial and adventitial layers of samples of vein grafts obtained at the time of implantation. Vein graft patency and function were monitored over the first 12 months using ankle pressures and Duplex ultrasound to determine vein graft status. Cells were obtained from veins from 11 patients whose grafts remained patent (non-stenotic) and from seven patients whose grafts developed stenosis. Smooth muscle cells (SMCs) derived from media and fibroblasts derived from adventitia were growth arrested in serum-free medium and then stimulated with 1 muM sphingosine-1-phosphate (S1P), 10 nM thrombin, 10 ng/ml epidermal growth factor (EGF), 10 ng/ml platelet-derived growth factor-BB (PDGF-BB), PDGF-BB plus S1P, or PDGF-BB plus thrombin for determination of incorporation of [(3)H]-thymidine into DNA. Cells receiving PDGF-BB or thrombin were also treated with or without 100 microg/ml heparin, which is a growth inhibitor. Cells receiving thrombin were also treated with or without 150 nM AG1478, an EGF receptor kinase inhibitor.

RESULTS

SMCs and fibroblasts from veins of patients that developed stenosis responded more to the growth factors, such as PDGF-BB alone or in combination with thrombin or S1P, than cells from veins of patients that remained patent (P = .012). In addition, while PDGF-BB-mediated proliferation of fibroblasts from grafts that remained patent was inhibited by heparin (P < .03), PDGF-BB-mediated proliferation of fibroblasts from veins that developed stenosis was not (P > .5).

CONCLUSION

Inherent differences in the proliferative response of vein graft cells to PDGF-BB and heparin may explain, in part, the variability among patients regarding long term patency of vein grafts.

Efficacy of duplex ultrasound surveillance after infrainguinal vein bypass may be enhanced by identification of characteristics predictive of graft stenosis development

OBJECTIVE

Controversy regarding the efficacy of duplex ultrasound surveillance after infrainguinal vein bypass led to an analysis of patient and bypass graft characteristics predictive for development of graft stenosis and a decision of secondary intervention.

METHODS

Retrospective analysis of a contemporary, consecutive series of 353 clinically successful infrainguinal vein bypasses performed in 329 patients for critical (n = 284; 80%) or noncritical (n = 69; 20%) limb ischemia enrolled in a surveillance program to identify and repair duplex-detected graft stenosis. Variables correlated with graft stenosis and bypass repair included: procedure indication, conduit type (saphenous vs nonsaphenous vein; reversed vs nonreversed orientation), prior bypass graft failure, postoperative ankle-brachial index (ABI) < 0.85, and interpretation of the first duplex surveillance study as "normal" or "abnormal" based on peak systolic velocity (PSV) and velocity ratio (Vr) criteria.

RESULTS

Overall, 126 (36%) of the 353 infrainguinal bypasses had 174 secondary interventions (endovascular, 100; surgery, 74) based on duplex surveillance; resulting in 3-year Kaplan-Meier primary (46%), assisted-primary (80%), and secondary (81%) patency rates. Characteristics predictive of duplex-detected stenosis leading to intervention (PSV: 443 +/- 94 cm/s; Vr: 8.6 +/- 9) were: "abnormal" initial duplex testing indicating moderate (PSV: 180-300 cm/s, Vr: 2-3.5) stenosis (P < .0001), non-single segment saphenous vein conduit (P < .01), warfarin drug therapy (P < .01), and redo bypass grafting (P < .001). Procedure indication, postoperative ABI level, statin drug therapy, and vein conduit orientation were not predictive of graft revision. The natural history of 141 (40%) bypasses with an abnormal first duplex scan differed from "normal" grafts by more frequent (51% vs 24%, P < .001) and earlier (7 months vs 11 months) graft revision for severe stenosis and a lower 3-year assisted primary patency (68% vs 87%; P < .001). In 52 (15%) limbs, the bypass graft failed and 20 (6%) limbs required amputation.

CONCLUSIONS

The efficacy of duplex surveillance after infrainguinal vein bypass may be enhanced by modifying testing protocols, eg, rigorous surveillance for "higher risk" bypasses, based on the initial duplex scan results and other characteristics (warfarin therapy, non- single segment saphenous vein conduit, redo bypass) predictive for stenosis development.

Mouse strain differential neointimal response in vein grafts and wire-injured arteries

BACKGROUND

Neointimal development is seen clinically after both vein grafting and balloon catheterization, but may not represent the same pathology under these 2 conditions. This study compared the degree of neointimal hyperplasia after vein grafting or arterial-injury grafts in 2 strains of mice: C57Bl/6 and FVB.

METHODS AND RESULTS

Jugular vein branches were interpositioned as grafts in the femoral artery of syngenic-matched mice, with graft harvest at 30 days. Wire-injured carotid arteries were grafted to the carotid arteries of syngenic-matched mice, with graft harvest at 14 days. Histomorphometry revealed no strain differences in vein grafts in the extent of position-dependent neointimal thickening or lumen cross-sectional area. Both strains showed significantly thicker neointima and less lumen area at the proximal graft site (vs the mid-graft; p<0.05). In contrast, a significantly greater neointimal thickness was found in the wire-injured carotid grafts of FVB mice vs those of C57Bl/6 mice (p<0.05).

CONCLUSIONS

Neointimal formation shows a vessel-dependent, strain-dependent difference, with greater arterial neointimal thickening in FVB mice. These data suggest that different mechanisms operate for arterial-injury- vs vein-graft-associated neointimal development and that the difference has a genetic basis.

Model of murine interpositional vein grafting

A new murine model is presented for interpositional vein grafting into an artery. In ICR (outbred) mice, the femoral vein or a branch of the jugular vein was harvested and placed into a femoral artery defect, using standard microvascular techniques to achieve end-to-end anastomoses with 11-0 nylon suture. The technique requires a high level of microsurgical experience and competence. Success rates (vessel patency) were higher with moderate systemic heparinization (200 units/kg body weight, single i.v. bolus). Vein grafts were successfully transplanted between syngenic C57Bl/6J mice, with patency demonstrated from 1 day to 3 months postoperatively. The model is applicable to studies of vein graft stenosis using murine transgenic/knockout models.

Interpretation of the Results of Doppler Ultrasound Flow Volume Measurements of Infrainguinal Vein Bypasses

OBJECTIVES

To evaluate a pattern for the interpretation of the results of intraoperative Doppler ultrasound flow volume measurement of infrainguinal vein bypasses.

DESIGN

Retrospective analysis of prospective data.

MATERIALS

91 consecutively performed infrainguinal non-reversed free vein bypasses.

METHODS

Using preoperative angiograms, the run-off, which can be expected after the reconstruction, was evaluated by means of a point score. A first measurement of the flow volume was taken after the release of the blood flow and a second after administration of 20 mg Alprostadil into the bypass. From these two results, we calculated an average value, which was set in relation to the run-off score. From this we computed a relative flow, i.e. the flow per open crural vessel = per run-off score unit (quotient:flow/score).

RESULTS

The median relative flow of angiographically perfect reconstructions was 86 ml/min. In 14 reconstructions, the control angiogram showed stenoses: median relative flow 59 ml/min, range between 20 and a maximum of 75 ml/min. The practical application of the flow measurement requires a minimum relative flow guideline for stenosis-free reconstructions. A guideline of 80 ml/min would yield a sensitivity of 100% and a specificity of 68%.

CONCLUSIONS

Measurement of flow volume could be used as a screen, in order to filter out reconstructions, which must be further clarified with an angiogram. A further prospective evaluation of the value of volume flow is needed before any conclusive recommendations can be drawn.

Intimal Hyperplasia in Mouse Vein Grafts Is Regulated by Flow

Altered flow conditions are presumed to cause stenosis in vein grafts due to exaggerated neointimal formation. The aim of this study was to establish a mouse model of flow-regulated intimal hyperplasia (IH) in vein grafts. The caval vein was grafted into the common carotid artery of 38 mice, followed by modulation of the blood flow, resulting in vein grafts with high (HF) and low flow (LF). The vessel wall thickening was evaluated after 3, 14 and 42 days by morphometric analyses and immunohistochemistry. There was an immediate significant change in flow, which was persistent throughout the time of observation. After 42 days, flow was increased 2.7 times in HF animals compared to LF animals. The vessel wall was composed of two layers where the inner layer was positive for alpha-actin and considered as IH. The area of neointimal formation was 74% larger in the LF group compared to the HF group. The present study demonstrates that flow regulates IH in vein grafts in mice. This model gives the potential to study the effect of shear stress on vascular biology in genetically modified animals.

Altered expression of vascular endothelial growth factor and its receptors in normal saphenous vein and in arterialized and stenotic vein grafts

BACKGROUND

Myointimal thickening is a major cause saphenous vein graft failure. The prominence of medial and adventitial microvessels in stenotic vein grafts and the known angiogenic effects of vascular endothelial growth factor (VEGF) lead us to investigate the expression of VEGF and its receptors in vein graft arterialization and stenosis.

METHODS

Normal and arterialized vein graft segments were evaluated by reverse transcription-polymerase chain reaction (RT-PCR) for expression of VEGF-R1 (flt), VEGF-R2 (KDR), and neuropilin-1. The cells expressing VEGF, VEGF-R1, VEGF-R2, and neuropilin-1 were identified in normal, stenotic, and arterialized vein graft segments by immunohistochemistry.

RESULTS

Vascular endothelial growth factor, detected in the wall in endothelial cells and adventitial microvessels in normal vein, localized to smooth muscle cells, endothelial cells and adventitial microvessels in arterialized and stenotic vein. VEGF-R1 and VEGF-R2 were expressed infrequently on endothelial cells, macrophages, and smooth muscle cells in arterialized and stenotic vein. Neuropilin-1 was detected in all specimens. RT-PCR demonstrated significantly greater expression of neuropilin-1 in normal vein compared with arterialized vein (P <0.05).

CONCLUSIONS

The differential expression of VEGF and its receptors in normal, arterialized, and stenotic vein grafts suggests that alterations in VEGF/VEGF-R2/neuropilin-1 interactions may be important determinants of the adaptive response of vein grafts to arterialization.

Longitudinal impedance is independent of outflow resistance

BACKGROUND

Many investigators have measured outflow resistance (R) following peripheral bypass procedures, but correlations with graft patency have been weak. This is because the primary determinants of graft patency are the size and quality of the conduit, not its outflow bed. Efforts at separating conduit resistance from outflow resistance have been unsuccessful. Recently, the concept of longitudinal impedance ( integral Z(L)) has been suggested as a measure of conduit resistance independent of outflow resistance. The purpose of this in vitro experiment was to test the hypothesis that integral Z(L) is independent of R within physiologically relevant ranges.

METHODS

Rigid polyethylene tubing of known internal diameter and length (4.3 mm, 375 cm) was perfused with a glycerin/saline mixture mimicking the viscosity of blood (4.1 cp), utilizing a variable pulsatile pump and Windkessel, with outflow into multiply branched tubes of decreasing diameter simulating the hemodynamic conditions of arterial bypass. Flow and pressure were measured using ultrasonic transit time and catheter transduction, respectively, and waveforms digitized at 200 Hz. Flow was varied while maintaining "systemic" pressure and resistance. After Fourier transformation, integral Z(L) was calculated as deltaP/Q at each harmonic and integrated over 4 Hz.

RESULTS

integral Z(L) calculations were remarkably reproducible within the same day with a coefficient of variation (CV) = 4.0% (at 100 dyne. s/cm(5); n = 4) or over 4 successive days (CV = 4.3%). Furthermore, integral Z(L) was largely independent of R over the physiologic range tested, with integral Z(L) remaining relatively constant as R was increased sixfold.

CONCLUSION

integral Z(L) is a consistent and reproducible measure of conduit resistance independent of R over a wide physiologic range. It may be useful for measuring the adequacy of bypass graft conduits.

Ten-year outcome after saphenous vein patch angioplasty in males and females after carotid endarterectomy

Many surgeons use a vein-patch angioplasty after endarterectomy of small (< 4 mm) diameter carotid arteries in males and females. Because evidence suggests that females may have a poorer outcome than males after revascularization, the long-term success of carotid endarterectomy may be different among the genders. The 10-year outcome after a saphenous vein-patch angioplasty in females was studied. Endarterectomies were performed on 708 carotid arteries of 615 patients, who had evidence of symptomatic disease or had asymptomatic greater than 80% stenosis. A segment of proximal greater saphenous vein was harvested for the patch. The 247 females had a mean age of 69.1 +/-9.8 (SD) years and a follow-up time of 4.0 +/-4.0 (SD) years. The 368 males had a mean age of 69.2 +/-8.8 (SD) years and a follow-up time of 4.4 +/-4.0 (SD) years. After endarterectomy, survival, the ipsilateral stroke-free rates and the restenosis-free rates were determined by life table analyses with 73 endarterectomies in 66 patients being at risk at 10 years. The 5- and 10-year survival rates in males were 81.9% and 62.2%, respectively. The 5- and 10-year survival rates in females were 82.6% and 73.0%, respectively. The 5- and 10-year ipsilateral stroke-free rates after carotid endarterectomy were 98.3% and 93.9% in males and 96.7% and 95.6% in females. The respective 5- and 10-year restenosis-free rates were 96.7% and 93.3% in males and 88.6% and 82.8% in females; p < 0.0002, by the Mantel-Cox test. Although survival and the incidence of an ipsilateral stroke were similar in females and males, the hemodynamic restenosis rate was higher in females. It is especially important to use routine duplex scanning to follow the post-endarterectomy health of the carotid artery in females.

Transfer function index of pulse volume recordings: a new method for vein graft surveillance

OBJECTIVE

Color flow duplex scanning is currently the best method available for vein graft surveillance. However, it puts a considerable strain on the workload of a vascular unit and requires a highly trained operator. The aim of this study was to develop and validate a new, noninvasive tool for graft surveillance. The utility of transfer function index (TFI) of pulse volume recordings is tested for this purpose.

METHODS

The design of the study was a blind comparative study that involved 70 testing procedures that were performed on 58 different infrainguinal vein bypass grafts. The TFI was measured with a portable vascular laboratory multi-cuff unit. Ankle/brachial indexes were obtained with the same device. Color flow duplex scanning was used as a diagnostic standard. A graft was defined as at risk, according to duplex scanning, if a local stenosis with a V2/V1 more than 2 was found or if peak systolic velocity remained less than 45 cm/s throughout the graft. The repeatability of the method was tested on 30 grafts.

RESULTS

A total of 63 tests were available for analysis. Seven tests were excluded. Four were excluded because they had unreliable TFI measurement due to cardiac arrhythmias, and in three tests, the whole graft could not be visualized in the duplex scan. Forty normal and 22 at-risk grafts were found. One graft was occluded. The TFI was significantly lower for at-risk grafts (0.89) versus normal grafts (1.09; P =.005). A TFI of 1.02 or less correctly detected 21 of 22 at-risk grafts. The sensitivity, specificity, and accuracy were 96%, 65%, and 76%, respectively. The ability of the ankle/brachial index to detect the at-risk grafts was clearly inferior to the TFI. The repeatability of the method at proximal thigh, distal thigh, and proximal calf was +/- 0.21, +/- 0.07, and +/- 0.14, respectively.

CONCLUSION

The TFI is a sensitive and reliable method to detect an at-risk graft. The examination is noninvasive, simple, quick to perform, and well tolerated by the patients. We suggest that the TFI could be the first-line screening method in vein graft surveillance.

The natural history of intermediate and critical vein graft stenosis: recommendations for continued surveillance or repair

OBJECTIVE

Duplex ultrasound surveillance (DUS) after autogenous lower extremity bypass grafting is controversial. Specific criteria mandating graft revision are not uniform. It has been suggested that grafts harboring critical stenoses undergo revision, whereas those with intermediate stenoses undergo arteriography with selective repair. We sought to define the natural history and determine the risk of graft occlusion associated with unrepaired vein graft stenoses.

METHODS

We analyzed serial vascular laboratory and clinical data of 156 autogenous infrainguinal vein grafts in 142 patients. Grafts were categorized into three groups according to the first DUS-detected (index) lesion: (1) normal (peak systolic velocity [PSV] < 200 cm/s, velocity ratio [V(r)] < 2); (2) intermediate stenosis (200 cm/s < PSV < 300 cm/s, 2 < V(r) < 4); and (3) critical (PSV > 300 cm/s, V(r) > 4). Our policy was to repair grafts with critical lesions and monitor all others. The risks of stenosis progression, graft revision, and graft thrombosis for each group were compared.

RESULTS

Serial DUS was normal in 100 (64%) grafts. The incidence of graft thrombosis in the normal group was 3% per year (mean follow-up, 27.5 months). Intermediate lesions developed in 32 grafts (20%) and were followed. Among these 32 grafts with intermediate stenoses, 63% progressed to critical and were revised, and 32% resolved or stabilized (mean follow-up, 26 months). Only one graft occlusion occurred in grafts with intermediate lesions subjected to serial DUS monitoring (incidence 1.5% per year, P = not significant). In the third group, 16 of 25 grafts with critical lesions were successfully revised and remain patent. In nine cases, critical lesions were not repaired, resulting in seven (78%) occlusions, all within 4 months of DUS detection.

CONCLUSIONS

Serial surveillance is safe and effective for grafts with intermediate stenoses. The graft occlusion rate for such grafts with careful monitoring is no different from grafts without stenosis, and therefore, arteriography is not indicated in the absence of progression to critical stenosis. The short-term risk of graft occlusion in the presence of an unrevised critical stenosis is nearly 80%. These data have important clinical implications concerning the natural history of vein graft lesions.

The hemodynamics of vein grafts: measurement and meaning

The long-term patency of infrainguinal vein grafts appears to depend primarily on the size and quality of the venous conduit. Therefore, those quantities which directly relate to the conduit's ability to act as a transporter of blood, namely internal diameter and longitudinal impedance (Z(L)), have predictive value for patency. Autologous grafts of good quality frequently remain patent even with compromised outflow. Therefore, those quantities that are outflow dependent, including deltaP, flow, velocity, shear stress, and resistance, carry less predictive value for long-term performance.

Abnormal duplex findings at the proximal anastomosis of infrainguinal bypass grafts: does revision enhance patency?

Using color duplex ultrasound (CDU) surveillance of autogenous infrainguinal bypasses, a peak systolic flow velocity (PSFV) ratio of greater than 3 to 1 within the graft relative to adjacent PSFV has been accepted as predicting significant stenosis mandating revision. At the proximal anastomosis, where significant vessel diameter differences and turbulent flow exist, the validity of these criteria is less clear. Our purpose was to review our experience with proximal anastomotic abnormalities in a CDU surveillance protocol. Routine CDU surveillance for all infrainguinal bypass gratis consisted of evaluation in an accredited vascular laboratory at 1 month postoperatively, every 3 months for the first year, every 6 months in the second year, and annually thereafter. Grafts with a PSFV ratio of >3 at the proximal anastomosis on any CDU study were included in this review. From our results we conclude that currently accepted CDU criteria for graft-threatening stenosis may not be valid for abnormalities at the proximal anastomosis of infrainguinal grafts. Regression of these abnormalities is common. Better CDU criteria are needed for predicting not only severity of proximal anastomotic stenosis but also likelihood of graft thrombosis.