Review

Hemostasis in the Era of the Chronic Anticoagulated Patient

Bonnie Weiner, MD, *Thomas Fischer, PhD, †Sergio Waxman, MD
Bonnie Weiner, MD, *Thomas Fischer, PhD, †Sergio Waxman, MD
Mechanisms for Hemostasis and Comparative Efficacy The technology to produce poly-N-acetylglucosamine (pGlcNAc) polymer is based on a biomaterial that is derived in a fiber form from aseptic cultures of a marine microalgae diatom. Once isolated and purified, the high quality, pure material is subject to rigorous quality control and quality assurance. The end product can be formulated as patches, lyophilized patches, gels, microspheres, and foams. The isolation of pGlcNAc fibers provides the source of the material for the manufacture of the Syvek Patch® (Marine Polymer Technologies, Danvers, Massachusetts). The ability of this material to provide hemostatic action is a very specific property of the way the polymer is organized in the pGlcNAc fibers (Figure 1). Hemostasis is defined as the process through which the body controls vascular injury and bleeding from it. Hemostasis has three inter-related components: vasoconstriction, red blood cell functions and platelet/coagulation factor activation. Examples of red blood cell function include rheology, during which red blood cells direct platelets toward the endothelium, and also release of certain paracrine-like receptors from red blood cells. Manual compression alone takes approximately 20 to 30 minutes to control bleeding from a diagnostic procedure and up to 60 minutes for an interventional procedure. These times are dramatically reduced to 5 and 10 minutes with a pGlcNAc patch. pGlcNAc provides hemostasis through mechanisms that involve vasoconstriction, red blood cell agglutination and platelet activation. Vasoconstriction. A series of studies conducted by Ikeda et al.1 utilized an aortic ring model where one of the metrics was the ability of these in-vitro rings to generate force. They titrated these rings with increasing concentrations of pGlcNAc and achieved a dose-dependent generation of force corresponding to a vasoconstrictive effect. They found that when the endothelium was denuded, this vasoconstrictive effect was ablated. Thus, it is an endothelium-dependent vasoconstriction. They sought to determine if the pGlcNAc affected the nitric oxide (NO) signal by looking at NO generation from these rings. They found that the presence or absence of the pGlcNAc did not affect the generation of that messenger. Then they examined the role of endothelin-1 in this system by looking at an inhibitor of the endothelin-1 receptor. They found that if this inhibitor was added, the result was partial ablation of the pGlcNAc-induced vasoconstriction. The sum of what is known today in this area is essentially that the vasoconstrictive process is in part mediated by the messenger endothelin-1. Red blood cell agglutination. Figure 2 shows a histological section from a hemorrhaging spleen where the pGlcNAc patch was placed on a wound. Agglutination of red blood cells is visible on the patch. At a scanning electron micrographic level, the red blood cells appear in an altered morphology. A fiber nest forms over the red blood cells. Although the exact mechanism of this altered morphological agglutination is unknown, it could be a change in the membrane potential in the cells and/or an osmotic effect on the cells. Platelet activation. In an experiment where platelets were allowed to contact pGlcNAc fibers, a morphological change of the platelet that is characteristic of a full-blown activation response was seen (Figure 3). As Figure 3 demonstrates, the pseudopodia make a robust contact with the pGlcNAc. In another experiment, the polymer was placed on a microscopic slide on a fluorescent stage and then stained to see if contact with the fiber induced expression of the activation marker P-selectin (Figure 4). P-selectin is a protein that normally lives in the alpha granules, and when the cells are activated, it is expressed on the surface of the protein. A red fluorescent counter-stain was used for P-selectin, and expression of P-selectin was achieved when the platelets activated on the fiber. Similarly, counter-stain with an antibody was used to activate integrin complex IIbIIIa, which is an activation-dependent antibody (Figure 5). The platelets that contact this fiber also have activated integrin complex. Outside end signaling of the integrin complex drives the activation response of platelets to contact with a fiber. Inter-cellular calcium signal draws platelet activation in a general sense, and contact with the fiber generates a calcium signal. To demonstrate this, platelets were loaded with two chromofluorophoric dyes, a red dye and a green dye, which are calcium indicators (Figure 6). The green dye is bright at high calcium and the red dye is bright at low calcium. In this manner, a metric can be obtained for inter-cellular calcium concentration (Figure 6). This is in the micromolar range of inter-cellular calcium, which is what typically has been known to drive a full activation response. In contrast, inhibition of the integrins with Integrilin keep the calcium concentration in the range of an unactivated platelet. The calcium signal drives the surface presentation of phosphatidyl serine (PS) on platelets in a very tightly coupled manner. In another study, platelets were allowed to settle on the pGlcNAc and were then counter-stained with Annexin V to see if PS was exposed, which, indeed, it was (Figure 7). Consistent with the last result, inhibition of the cells with Integrilin shows no PS exposure. This is important because the PS exposure drives the turnover of the coagulation cascade for thrombin generation and thus fibrin polymerization. Fibrin gel formation. If platelet-free plasma is mixed, and clot time is measured, the presence or absence of the pGlcNAc does not measurably affect it. In these conditions, it takes approximately 30 minutes for a physical macroscopic clot formation to occur. When platelets are added, clot time is reduced to approximately 20 minutes. When platelets plus pGlcNAc are added, the time is reduced in half again to approximately 10 minutes. Having determined that the platelet pGlcNAc complexes accelerated coagulation, contact activation and turnover of the factor XII driven intrinsic coagulation was pursued. To answer this question, mixtures of pGlcNAc and platelets were titrated with corn trypsin inhibitor, a specific competitive inhibitor of factor XII, the Hageman factor. A dose-dependent prolongation or inhibition of the clot formation process was attained, which indicates that the PS generated on the surface of the platelet is driving the turnover of the intrinsic coagulation cascade for accelerated clot formation. Physical structure of the pGlcNAc fiber. In a series of experiments, plasma containing platelets was incubated with three different materials and a control (no material added). The materials tested included pGlcNAc, Chito-Seal topical hemostasis pad (Abbott Vascular Devices, Redwood City, California) and Clo-Sur pad (Scion Cardio-Vascular, Miami, Florida). Chito-Seal, Clo-Sur pad and control all produce a clot in 20 minutes. Only the purified pGlcNAc fibers were able to reduce clotting time to 10 minutes. The physical structure of the pGlcNAc fibers is important. Figure 8 shows activated platelets on a mesh of pGlcNAc fibers. Everything is sized so the platelets can intercalate with the fibers in an intimate fashion. Once fibrin polymerization begins, nucleated at these platelets, the result is a tertiary mesh because the fibrin can polymerize within that mesh to interlace with it. In contrast, the Clo-Sur product presents a plainer sheath to the platelets. It contact activates them, but there is no intimate physical interaction with the material. Similarly, the Chito-Seal product produces long tubular structures. They contact activate the platelets, but again, there is no close interaction. In summary, when blood contacts the pGlcNAc matrix, the first event to occur on a time scale of nanoseconds to microseconds is a chemical-physical absorption of serum proteins to the matrix. These serum proteins are in an altered conformation, particularly fibrinogen. The platelets recognize the altered conformation and stimulate an activation response. The coagulation cascade activates, the red blood cells agglutinate and generate thrombin and a fibrin mesh is synthesized in the microenvironment of the pGlcNAc patch. At this point, the platelets generate force through the clot retraction process, and the vasoconstrictive effects augment a mechanical constriction. In terms of what is occurring in the clinic, the platelet/pGlcNAc/red blood cell/fibrin mesh forms a hemostatic plug that creates stagnant blood in the track. Three conditions, the triad of conditions needed for thrombus formation, are now operant: injured tissue, static blood, and a procoagulative state of blood that is driven by the presence of the activation complex on the patch. Clinical Experience with pGlcNAc The rationale for the need for rapid and safe hemostasis occurs at three levels. The first level is the patient. Patients want to have the bleeding stopped, but they also want to minimize discomfort associated with sheath dwell time, compression time and bed rest. The second level is the physician. Physicians want adequate hemostasis to decrease vascular complications, reduce procedural times and time to hemostasis and gain expertise with new devices and techniques. On the third level, hospitals and catheterization laboratories want to achieve earlier ambulation and discharges, which in turn will free beds in observations units and hospital rooms and will lead to better utilization of resources and lower costs. Thus, there is a justifiable need at all three levels to improve or to facilitate hemostasis. The first requisite of an ideal hemostatic device is the ability to achieve complete hemostasis regardless of the level of anticoagulation and other clinical factors, such as in patients with the presence of peripheral vascular disease, bioprosthetic devices or materials in their bodies and different body habitus. Ideally, the hemostatic device should have low complication rates and will not compromise the arterial lumen by either embolization or narrowing. In addition, it will produce minimal sequelae in the surrounding tissues of the arteries so the site can be accessed repeatedly if necessary. Added bonuses include low cost and the ability to ambulate early. All hemostatic devices are compared with the time-tested standard of care, which is manual compression. Historically, manual compression consisted of 15-minute compression cycles at the arteriotomy sites with 6 to 8 hours of bed rest. This recommendation was based on empiric knowledge; there are no randomized trials to support this. There are, however, observational studies and registries of patients looking at shorter compression times or bedrest, depending on the sheath size. The complication rates for diagnostic procedures range from 0% to 1.1%, and for interventional procedures from 1.3% to 5.9%, depending on the level of anticoagulation and sheath sizes. Hoffer and Bloch2 reviewed 31 studies reported in the literature that measured efficacy rates of arterial closure devices (VasoSeal, Datascope Corporation, Mahwah, New Jersey; Angio-Seal, St. Jude Medical, Minnetonka, Minnesota; Duett, Vascular Solutions, Inc., Minneapolis, Minnesota; and Perclose, Abbott Vascular Devices, Redwood City, California). They reported success rates above 90%, hemostasis rates between 90% and 95% and failure rates around 5%. The major complication rates of arterial closure devices ranged from 3% to 4%, from 0.5% in the lowest to almost 10% in the highest. The major complications with any of these devices are not trivial; they can be life or limb threatening on occasion. The same report compared arterial closure devices with manual compression and found that there was no difference between the devices and manual compression. Therefore, it is justifiable to evaluate less invasive technologies that utilize the principle of assisted, augmented, or accelerated compression that access the site with a topical patch. Limited clinical experience suggests that assisted compression with pGlcNAc can decrease compression time, can decrease time to ambulation and can be associated with low complication rates. A study conducted by Khuri et al.3 addressed whether assisted compression with pGlcNAc can decrease compression times. They studied 33 patients who underwent diagnostic cardiac catheterizations, randomized to either placebo gauze pad (n = 17) or pGlcNAc (n = 16). They developed a clamp device to measure the pressure that was being applied with either test product. They applied 5-minute cycles of compression, comparing placebo with the pGlcNAc patch. They determined time to hemostasis was reduced with use of the pGlcNAc patch and were able to obtain adequate hemostasis at 10 minutes. There was no difference in hematoma formation. Hirsch et al.4 conducted a randomized trial of 40 patients undergoing both diagnostic and therapeutic neurointerventional procedures. The standard of care was 15-minute cycles with additional 5 minutes as needed if bleeding persisted. The protocol with the pGlcNAc patch was 8-minute cycles with 4-minute repeated cycles as needed. The sheath sizes varied from 5 Fr to 7 Fr. With the pGlcNAc patch, hemostasis was achieved in 8 minutes in 19 out of 20 cases. With the standard of care, hemostasis was achieved in 15 minutes in 18 out of 20 patients. There was no difference in outcome with either 8-minute or 15-minute compressions, with pGlcNAc or with the standard of care. There was one hematoma in the standard of care cohort, and no complications in the pGlcNAc cohort. A study by Shubrooks et al.5 examined early ambulation in 40 consecutive patients after diagnostic cardiac catheterization with sheath sizes from 6 Fr to 8 Fr. Assisted compression with pGlcNAc was utilized, and patients were ambulated at two hours. The groups were divided as follows: 20 patients were discharged home immediately after a 2-hour period and 20 patients were kept in observation for an additional four hours for close follow up in the hospital. There was no difference in either group as far as hemotoma formation or other vascular complications. Palmer et al.6 studied 200 consecutive patients who underwent outpatient cardiac catheterization with 6 Fr sheaths. Heparin was given in about half (53%) of the patients, and assisted compression with pGlcNAc was applied for 15 minutes. The objective was to demonstrate that early ambulation was possible with assisted compression. Patients had one-hour bed rest and ambulated in one hour. One-hundred ninety six patients (98%) were discharged within three hours of the angiogram. There were no major adverse events; however, there were two small bleeds and two hematomas during the observation procedure. None of the events prevented the patients from going home within four hours. Nader et al.7 reported their experience with 1,000 consecutive patients in the largest series of patients that has been reported to date with the use of assisted compression with pGlcNAc. There were 636 diagnostic procedures and 364 interventional procedures observed. Assisted compression was used for 10 minutes in the diagnostic group and for 20 minutes in the interventional group as long as the ACT was less than 300 seconds. Their endpoint was complication rates rather than compression times. In the interventional group, about 76% of patients had ACTs of 200 seconds or more. A significant number of patients were taking platelet inhibitors. The overall major complication rate was 0.1%, and the minor complication rate was 1.3%. These are observational cohorts. Divided by type of procedures, the major complication was a pseudoaneurysm, which occurred in the interventional group; there were no major complications in the diagnostic group. There were 4 minor complications in the interventional group (3 small hemotomas and one nuisance bleed) and 9 minor complications in the diagnostic group (4 small hemotomas, 5 nuisance bleeds), suggesting that assisted compression with the pGlcNAc patch has a benign complication profile based on this cohort of patients. Clinical Overview There is mounting clinical evidence that assisted or accelerated compression with pGlcNAc may be as effective as standard of care, if not superior, but there are no trials yet to prove this. Assisted compression is as effective as arterial closure devices; has low complication profiles; may decrease the sheath dwell times and compression times, which is an outcome that is important for patients; and enables early ambulation, which is important for catheterization laboratory staff. Decreased sheath dwell time, decreased time to hemostasis or early ambulation are the most desirable outcomes when dealing with hemostasis. Assisted compression with pGlcNAc may help achieve these outcomes with the added benefit of lower complication rates and costs when compared with closure devices. The present data invites physicians to take pause when deciding the best and safest ways of achieving hemostasis and calls for additional studies of assisted compression.
References
1. Ikeda Y, Young LH, Vournakis JN, Lefer AM. Vascular effects of poly-N-acetylglucosamine in isolated rat aortic rings. J Surg Res 2002;102:215–220. 2. Hoffer EK, Bloch RD. Percutaneous arterial closure devices. J Vasc Interv Radiol 2003;14:865–885. 3. Najjar SF, Healey N, Healey CM, et al. A blinded randomized trial to evaluate the hemostatic effectiveness of poly-N-acetylglucosamine in patients undergoing cardiac catheterization. Am J Cardiol 2003;92(Suppl):151L. 4. Hirsch, et al., American Society of Neuroradiology, 39th Annual meeting, April 23–27, 2001, Hynes Convention Center, Boston, Massachusetts, "A Clinical Protocol to Test the Use of the SyvekPatch as a Technique to Effect the Closure of the Arteriotomy Site in Interventioal Neuroradiology. ASNR 2001. 5. Shubrooks SJ Jr., RW Nesto, DE Leeman, et al., Beth Israel Deaconess Hospital, Boston, MA, SCA&I 2000 Meeting Abstracts, “Earlier Ambulation at 2 Hours Following Cardiac Catheterization Using Syvek Patch.” 6. Palmer B, Gantt DS, Lawrence ME, et al., Scott & White Hospital and the Texas A&M University System Health Science Center College of Medicine, Temple, Texas, "One Hour Bedrest After Coronary Angiography with 6 F Catheters", Abstract submission, SCAI, 26th Annual Scientific Meeting, May 7–10, 2003, Boston, Massachusetts. 7. Nader RG, Garcia JC, Drushal K, Pesek T. Clinical evaluation of the SyvekPatch® in consecutive patients undergoing interventional, EPS and diagnostic cardiac catheterization procedures. J Invas Cardiol 2002;14:305–307.