Abstract: Aims. Vascular closure device (VCD)-based venous closure has been anecdotally reported, but systematic evaluation of the reparative response of the vessel wall to venous closure is lacking. The need to control groin complications, and minimize risks associated with postponed sheath removal under conditions of persistent anticoagulation, has generated interest in the role of VCDs for venous access closure. We sought to characterize the vessel wall response to venous closure, both acutely and in delayed fashion at 30 days using angiography, ultrasound, and histology. Methods. Ten venous 7 Fr sheaths were deployed in the femoral veins of swine. Bilateral venous access sites were subsequently closed utilizing manual compression (MC; control arm: n = 4) or a closure device utilizing an extravascular polyethylene glycol sealant (MynxGrip treatment arm: n = 6). Acute (post closure), 3-day, and 30-day vascular ultrasound, as well as venography (internal jugular approach) were used to assess outcomes. Gross pathology and histology were obtained at the 30-day endpoint for all femoral venous closure sites. Results. Hemostasis was successfully achieved in all cases without access-site complications. Venography and ultrasound confirmed normal ilio-femoral anatomy and flow at all study time points. Gross pathology and histopathology revealed no evidence of deep vein thrombosis, and no abnormalities were seen in the vena cava, heart, or lungs. Histology at 30 days showed complete healing of the vein wall access site, with a small focus of chronic inflammation and fibrosis in the perivascular adventitial tissue of the access tract. There was no microscopic evidence of the sealant. The tissue tract showed mild discrete inflammation (foamy macrophages, lymphocytes, plasma cells) with microgranulomas centered on residual red cells in both treatment and control groups. Conclusions. This study characterizes the angiographic, ultrasound, and histopathology outcomes of femoral vein closure, and provides insight into the healing mechanisms following venotomy. The bio-resorptive role of MynxGrip extravascular sealant in achieving effective venous closure and preserved long-term vessel patency without venous thromboembolism is demonstrated.
J INVASIVE CARDIOL 2015;27(2):121-127
Key words: venous closure, venous thromboembolism
Hemostasis after venous sheath extraction is usually achieved using manual compression (MC). The need to control groin complications and minimize risks associated with deferred sheath removal, especially where there is persistent requirement for systemic anticoagulation, has generated clinical interest in the role of vascular closure devices for venous access closure.
Much attention has been given to arterial vascular closure devices (VCD), but less so venous closure techniques. Manual compression is cited as the “gold” standard for venotomy closure, but suffers certain limitations: (1) discomfort and pain for the patient — which is a particularly difficult problem in the pediatric population; (2) postponement of sheath extraction post procedure until after dissipation of anticoagulant effect, thereby: (a) prolonging patient discomfort; (b) increasing bleed risk from an indwelling sheath; and (c) necessitating discontinuation of anticoagulation post procedure, which may be clinically detrimental; (3) risk of deep venous thrombosis, and delayed bleeding after compression, especially if anticoagulants are restarted after compression; and (4) risk of failed closure due to inaccurate location of MC site. For all these reasons, attention has been focused on developing venous applications for various VCDs, whether suture based,1 collagen plug based,2 or extravascular polyethylene glycol (PEG) sealant based.3 We report the first detailed imaging and histology analysis of the vein wall response to an extravascular water-soluble hydrogel PEG sealant for vascular closure (MynxGrip device; AccessClosure).4
This study followed good laboratory practices under Food and Drug Administration CFR 21/I/A/Pt.58 (good laboratory practice for non-clinical laboratory studies) and was also approved by the Institutional Animal Care and Use Committee. Accordingly, the present study was conducted under supervision of an independent study director, quality assurance monitoring, and pathologist. All animals received standard care pursuant to human catheterization laboratory protocols following the act of animal welfare and the “Principle of Care of Laboratory Animals.”
Experimental design. A total of 6 domestic Yorkshire pigs weighing more than 90 kg (range, 90.9 kg-98.1 kg) were obtained. All animals were uniquely identified and quarantined at an approved facility, and then underwent a 7-day acclimation period where animal care occurred in accordance with the United States Department of Agriculture Animal Welfare Act (9CFR parts 1-3). All animals were deemed in good health prior to study participation.
Medication and anesthesia. All animals received preoperative intravenous (IV) cefazolin 22 mg/kg and intramuscular (IM) buprenex (buprenorphine) 0.01 mg/kg. Anesthesia was induced with telazol 3 mg/kg. Glycopyrrolate 0.008 mg/kg was administered IM to decrease respiratory secretions and prevent bradycardia. Animals were then intubated after 0%-3% isoflurane administration. A mean initial dose of 5000 U IV heparin (range, 4750-5600 U) was administered to achieve a target activated clotting time (ACT) between 250-450 seconds. Periodic blood aliquots were taken to monitor ACT intraprocedurally every 15-30 minutes. Heparin was administered as necessary to maintain ACT in the protocol-specified target range between 250-450 seconds. Post procedure, analgesia was provided with carprofen 4 mg/kg IM and buprenorphine 0.01 mg/kg IM. Heart rate, respiratory rate, temperature, and animal position were monitored every 15-30 minutes until the animal was sufficiently alert, after which it was returned to housing. Prophylactic antibiotics were administered post procedure. Daily aspirin 162 mg was administered orally until study termination.
Venographic and Doppler ultrasound image techniques. A 6 Fr introducer sheath was placed percutaneously in the right internal jugular vein and a guiding catheter (5 Fr Vista Brite Tip; Cordis Corporation) was advanced over a standard 0.035˝ guidewire from the internal jugular vein retrograde into the ilio-femoral vein for retrograde venography. This technique was used at both index procedure and at follow-up to obtain venograms of the treated femoral vein. Standard transcutaneous ultrasound was used to assess venous flow and tissue tract abnormalities at day 0 (immediately after closure), day 3, and day 30 (final follow-up [FUP]).
Study design and methodology.
Femoral vein access technique. Percutaneous access to the femoral vein was performed under ultrasound guidance (Siemens Sequoia Model c512) using a 5 Fr micropuncture kit (Cook, Inc). Following micropuncture and modified Seldinger cannulation, the access sheath was upsized to 7 Fr standard introducer sheath (Terumo).
Femoral vein closure. After cannulation with 7 Fr sheath, the femoral veins were closed either by applying standard MC (5 minutes initially,5 with additional MC on an as-needed basis or using the MynxGrip vascular closure device).
Device description. The MynxGrip VCD is a novel closure device that provides a polyethylene-glycol mediated, extravascular sealing of the puncture site.6 The MynxGrip VCD was used according to the closure protocol provided by the manufacturer. After sealant deployment, operators laid the device down for 60 seconds; the balloon was then deflated, and the device was removed followed by 2 minutes of finger-tip compression.
Follow-up study. All animals remained in veterinarian care throughout the 30-day study. At day 3, animals underwent clinical assessment of the groin and Doppler ultrasound to assess durability of the femoral vein closure. At day 30, all animals underwent final clinical examination, groin assessment, Doppler ultrasound, and ilio-femoral venography to assess long-term venous patency.
Gross pathology and histopathology. Post mortem gross necropsy examinations were performed by an independent veterinary pathologist. The ilio-femoral veins were carefully dissected so as not to disturb the vascular access test site. Once the test site was localized, the vein was flushed with heparinized saline followed by 10% neutral buffered formalin (NBF). The vascular access test site was then removed en bloc and immersion fixed in NBF. After the vascular access test sites were explanted, a full post mortem gross necropsy examination (excluding the brain) was performed. A specific goal of the study was to evaluate each animal for venous thrombosis or thromboembolism to downstream tissues (ie, vena cava, heart, and lungs). The presence or absence of thrombosis or thromboembolism along the vena cava and in the heart and lungs was specifically evaluated. Representative sections of the tissues were collected and immersion fixed in NBF for histologic examination. After adequate tissue fixation (alcohol/xylene, paraffin), the ilio-femoral veins were trimmed in serial sections through the access site (4-6 µm thickness, at 250-500 µm intervals), for identification of the specific venotomy location including the proximal and distal reference segments. In addition, representative sections of vena cava, right heart, and lungs were also obtained. All tissue sections were then stained with hematoxylin and eosin (H&E) for light microscopy.
A semiquantitative score was developed to grade histological/microscopy findings as follows: 1 = absent; 2 = minimal; 3 = mild; and 4 = severe. The score was applied during analysis of the vascular sections for the following parameters:
- hydrogel sealant
- intimal hyperplasia
- venous stenosis
- adventitial fibrosis
- perivascular hemorrhage
- luminal fibrin/venous thrombosis.
Other data analyses. Serial, all-time-point venograms and ultrasound images were analyzed offline for evidence of bleeding or flow abnormalities. Quantitative procedural data, such as ACT time and time-to-hemostasis, were analyzed using SigmaStat software. A P-value <.05 was considered statistically significant.
A total of 10 femoral venous access sites were established in 6 pigs. Of the study access sites, 6/10 were closed using MynxGrip VCD (treatment group) vs 4/10 using MC. At the time of closure, ACTs were 336.83 ± 50.28 seconds in the treatment group compared with 368.75 ± 49.97 seconds in the MC group (P>.05). Hemostasis was successfully achieved in all cases, without groin complications in either group, and without device failures in the treatment group.
Angiographic and ultrasound assessment. Wide patency and normal flow were noted at all time points after MynxGrip VCD deployment. No filling defects, lumen loss, or flow abnormalities were seen immediately post procedure, or at 30-day time points by either angiography or ultrasound (Figures 1, 2, and 5). One mild non-actionable extravasation was noticed in the post-closure venogram of the MC group, which self-resolved in the subsequent venogram (Figure 3).
Gross necropsy. All treated vascular segments appeared grossly patent with no evidence of thrombosis. Occasionally, mild areas of fibrous adhesions (fibrous scarring, indicative of normal healing response) were present between the treated vein and the surrounding muscle and fascia. In addition, a small focus of nodular scar tissue was occasionally identified in the fascial tissue of the muscle overlying the treatment site. No embolism of either thrombus or sealant was identified in the vena cava, heart, or lungs in either study group.
Histology. Semiquantitative analysis of the microscopic sections (Table 1) showed both treatment and control groups were closely comparable. There was no evidence of relevant hyperplasia, mineralization, hemorrhage, or luminal thrombosis in either group. Both treatment and control groups demonstrated small (1-2 mm) focally discrete areas of adventitial fibrosis (score: 1.67 vs 1.25, respectively; P>.05). This was associated with modest cellular (mostly macrophages) inflammation (score: 1.83 vs 1.50, respectively; P>.05) (Figures 1 and 2C). In a subset of cases (3/6 treatment sites, and 3/4 control sites, collectively in 4 animals), inflammation consisted mostly of monocyte-macrophages and multinucleated giant cells organized into microgranulomas (≤2 mm), oriented around residual red blood cells (RBCs). The discrete tissue tract was preserved in some of the histology specimen, and extended from the venipuncture site through the perivascular fat and overlying skeletal muscle to the skin. In the treatment group, the tissue tract was infiltrated with foamy macrophages and lymphocyte clusters (Figure 4).
We report the first study of serial venous blood flow and chronic pathology responses of porcine femoral veins undergoing 7 Fr catheterization and vascular closure using MC or an extravascular closure device. This is relevant to clinical venous catheterization procedures, where there is a need to avoid postponement of sheath removal and to control groin complications. In the present study, all attempts to close the femoral vein after catheterization were successful in both treatment and control groups, with the benefits of quick (2-minute) and complete (free of extravasation by venography) closure in the MynxGrip treatment arm. During follow-up studies, the veins were patent and free of abnormalities, including stenosis, thrombi, and downstream embolizations. Necropsy and pathology analysis of central organs (heart, lungs) confirmed freedom from central and pulmonary emboli.
Histological specimens of treatment and control groups appeared closely comparable and were characterized by the presence of modest inflammation and focally discrete adventitial fibrosis. In both groups, the cellular infiltrates consisted predominantly of microgranulomas oriented around residual RBCs. In the treatment specimens, there was no evidence of the PEG hydrogel sealant at 30 days. However, the region of the tissue tract (formed by previous sheath insertion) was variably infiltrated with foamy macrophages, which was unique to the treatment group specimens. This is in line with a previously published report for the MynxGrip device in the porcine arterial closure tissue tract.6 Moreover, tissue repair and healing reactions after neurosurgical procedures involving PEG hydrogel sealant of canine dura mater have also been shown to be associated with similar foamy macrophages.7
The chronic vessel wall response to the vascular closure implant may be driven by the nature and position of the implant in the tissue tract. For example, Thai et al reported similar histological response after use of the MynxGrip in femoral artery closure.6 Since both the femoral artery and vein share the same tissue tract where the sealant is implanted, we suggest that the host’s inflammatory response to an extravascular implant can be extrapolated from arterial to venous applications. On the other hand, the vessel wall reaction to an intravascular component of the closure device may trigger spasm8 and thrombosis,9 and may be chronically obstructive, especially if the bioresorptive process is prolonged.10 Tellez et al reported serial histology and imaging findings after femoral artery access closure using Angio-Seal and showed that significant portions of the intravascular component continue to exist within the vessel wall at 30 days post implantation.10
Furthermore, the vessel wall reaction to the intravascular component may be markedly different between arteries and veins, as the responses evoked are likely to be specific to the arterial or venous vascular biology and the type of intravascular implant.11 Sanghi et al reported higher inflammatory scores with Angio-Seal closure (intravascular anchor and extravascular component) compared with StarClose closure (extravascular implant).12 The healing response and cellular inflammatory reaction of the venous wall to the chronic presence of an intravascular implant remains unknown. To date, there are no vascular pathology data on the use of closure devices with intravascular implants in veins, where blood flow is slow, and chronic mechanical venous wall injury is shown to lead to occlusive thrombosis.9,10 We suggest an extravascular approach to vein access closure, given the available evidence.
Limitations. This study was limited to 30 days, which is comparable to previously presented arterial data for the same device. However, in order to exclude delayed venous vessel wall reaction or clinically adverse events, a further longer-term study with larger study cohorts remains desirable.
Venous closure using the MynxGrip device is a safe and reliable procedure. There were no differences in the histological responses of the MC and MynxGrip VCD sites. Further long-term histologic and imaging studies of venous closure in animal models and human use are suggested.
- Hamid T, Rajagopal R, Pius C, Clarke B, Mahadevan VS. Preclosure of large-sized venous access sites in adults undergoing transcatheter structural interventions. Catheter Cardiovasc Interv. 2013;81(4):586-590.
- Coto HA. Closure of the femoral vein puncture site after transcatheter procedures using Angio-Seal. Catheter Cardiovasc Interv. 2002;55(1):16-19.
- Verma D, Tandar A, Dranow E, et al. Venous access closure using extravascular closure device: a propensity score matched analysis. Abstract #184: Presented at EuroPCR, 2013.
- Scheinert D, Sievert H, Turco MA, et al. The safety and efficacy of an extravascular, water-soluble sealant for vascular closure: initial clinical results for Mynx. Catheter Cardiovasc Interv. 2007;70(5):627-633.
- Wallace MJ, Ahrar K, Wright KC. Validation of US-guided percutaneous venous access and manual compression for studies in swine. J Vasc Interv Radiol. 2003;14(4):481-483.
- Thai HM, Weinstock BS. The Mynx Grip vasular closure device. Endovasc Today. 2012;11(4):28-33.
- Hutchinson RW, Mendenhall V, Abutin RM, Muench T, Hart J. Evaluation of fibrin sealants for central nervous system sealing in the mongrel dog durotomy model. Neurosurgery. 2011;69(4):921-928; discussion 929.
- Cooper RM, Krishnan U, Pyatt JR. Central venous spasm during pacemaker insertion. Heart. 2010;96(18):1484.
- Kohler TR, Kirkman TR. Central venous catheter failure is induced by injury and can be prevented by stabilizing the catheter tip. J Vasc Surg. 1998;28(1):59-65; discussion 65-56.
- Tellez A, Cheng Y, Yi GH, et al. In vivo intravascular ultrasound analysis of the absorption rate of the Angio-Seal vascular closure device in the porcine femoral artery. EuroIntervention. 2010;5(6):731-736.
- Lüscher TF. Vascular biology of coronary bypass grafts. Curr Opin Cardiol. 1991;6(6):868-876.
- Sanghi P, Virmani R, Do D, et al. A comparative evaluation of arterial blood flow and the healing response after femoral artery closure using Angio-Seal STS Plus and StarClose in a porcine model. J Interv Cardiol. 2008;21(4):329-336.
From the 1Heart Artery and Vein Center of Fresno, Fresno, California; and 2VDx – Preclinical Pathology Services, Davis, California.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.
Manuscript submitted April 2, 2014, provisional acceptance given June 25, 2014, final version accepted August 1, 2014.
Address for correspondence: S. Sanjay Srivatsa, MBBChir, Heart Artery and Vein Center of Fresno, 7210 N. Milburn Ave, Suite #101, Fresno, CA 93722. Email: firstname.lastname@example.org