| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1995;60:1561-1562
© 1995 The Society of Thoracic Surgeons
Vascular Cell Biology and Cardiovascular Research Laboratories, Texas Heart Institute, Houston, Texas
In the early 1970s, the National Heart, Lung and Blood Institute decided to concentrate its efforts in development of short and long-term mechanical circulatory support devices on the left ventricular assist device (LVAD). The article by Rafii and associates [1] in this issue of The Annals is a result of this effort. This important scientific study helps to characterize cells deposited on the textured lining of the HeartMate LVAD, which has been recently approved by the Food and Drug Administration for clinical use as a bridge to transplant.
To improve the biocompatibility of ventricular assist devices, investigators have been working to document the variety of host cell types that populate the surfaces of implanted blood pumps. Ideally, tests to determine cell type should be performed on the pannus that forms on the blood-contacting surfaces immediately after a pump is explanted. Despite the best intentions of investigators to remain unbiased, the procedures used to culture cells from such material are invariably selective for certain phenotypes. Studies have suggested the presence of a variety of phenotypes in these surfaces, including endothelial cells, macrophages [2], myofibroblasts or smooth muscle cells (Johnson AD, personal communication), and a variety of hematopoietic cell types, as Rafii and associates describe. However, the variety of phenotypes identified in material that has been obtained from textured pump surfaces may be dependent on the duration of implantation and the humoral status of the patient.
Investigators have also attempted to improve the biocompatibility of blood-contacting surfaces by studying the seeding of these surfaces with purified cell populations expanded in tissue culture [3]. Although such procedures have provided information on how well different cell types attach and grow on the textured surfaces in a tissue culture, they give little insight into how these processes work in vivo. For example, aortic endothelial cells adhere poorly to the textured polyurethane pulsatile diaphragm of the HeartMate LVAD, which is currently being used clinically with great success and which is the subject of Rafii and associates' study. In contrast, smooth muscle cells attach strongly to these textured surfaces and establish themselves as a multilayer within an extracellular matrix in a short time. These observations are inconsistent, however, with published data suggesting that endothelial cells readily populate the biomaterial surfaces of a number of implanted cardiovascular devices [2, 4, 5]. The presence of smooth muscle cells on the biomaterial surfaces of explanted HeartMate LVADs remains to be clearly demonstrated.
To define the ideal conditions for the establishment of a biocompatible cellular layer on the blood-contacting surfaces of LVADs, it is important to consider the environment within which a specific phenotype resides. This environment is the extracellular matrix, and it is the complexity of composition of this matrix that defines which cell will reside and develop in a specific locus. Despite many years of research, however, we still do not know exactly how the extracellular matrix regulates cell differentiation [6]. Which cells populate the internal surfaces first and how these phenotypes differ from patient to patient are other relevant questions for which we currently have no answer. It is also likely that the composition of the cellular and extracellular matrix differs for the variety of biomaterials used for internal surfaces of LVADs (eg, sintered titanium microspheres and textured polyurethane).
Thus, a number of apparently simple questions remain to be answered regarding the formation of cellular linings in ventricular assist devices. For example, where do the connective tissue cells that produce the collagenous matrix come from? Are they present in the circulation, do they originate from the surgical procedures, or do they slough off from tissues undergoing wound healing at the ventricular implantation site? It has been suggested that fibroblasts and endothelial cells may migrate into LVAD chambers from microarterioles and stromal tissue exposed at the surgical implantation site [2]. The latter appears unlikely because the time scale for the appearance of a collagenous matrix is too rapid for this to occur. More than two decades have passed since Stump and colleagues [4] showed that endothelial cells were present in circulating blood and could attach and propagate on biomaterials such as Dacron. Other studies have suggested the presence of pluripotential cells in blood that can differentiate into endothelial cells [7]. If this differentiation occurred readily, a relatively large number of endothelial cells might be expected to populate the early fibrin pannus, whereas very few such cells are evident at these times. Thus, the source of these endothelial cells remains unclear: they could simply come from circulating endothelial cells that attach to foreign bodies in the bloodstream. Regardless, clinical experience has confirmed that flocked surfaces attract a cellular lining.
Another question that remains unanswered relates to the apparent nonthrombogenic nature of the collagenous pannus that develops soon after implantation of assist devices. Intuitively, the collagen-rich surface would appear ideally suited for the activation of platelets. Investigators have attempted to accelerate the establishment of an extracellular matrix by seeding device surfaces with fetal fibroblasts that have a high capacity to produce collagen and proteoglycans [3]. These promising studies with allogeneic cells have demonstrated the formation of a multilayer consisting of a collagenous extracellular matrix and proteoglycans surrounding the stromal cells. Surprisingly, this pseudointima was not excessively thrombogenic, and there was no apparent evidence of lymphocytic lysis even in cell-seeded devices implanted for prolonged periods. Such observations raise almost more questions than answers.
We believe that immediately after implantation, platelet deposition on the blood-contacting surfaces of LVADs leads to rapid deposition of a fibrin matrix, which acts as the initial scaffold for the attachment of hematopoietic cells. Among this population will be a variety of phenotypes, including monocytes and a few pluripotential hematopoietic cells that can self-generate and proliferate, as demonstrated by Rafii and associates. Shortly thereafter, connective tissue cells (myofibroblasts) associate with the fibrin matrix, possibly as a consequence of chemotactic factors released by monocytes. Once these cells have taken up residence within the fibrin coagulum, they establish a collagenous matrix containing proteoglycans and glycopeptides. They form a scaffold that allows a pseudointima to develop, which is typically observed on the internal surfaces of pumps that have been explanted after as little as 2 weeks. This process is likely to involve continuous matrix remodeling by enzymatic turnover of macromolecular structures.
In addition, the fibroblastic cells may be responsible for the production of cytokines, which promote hematopoietic cell differentiation and, in addition, provide a stimulus for the later attachment of host endothelial cells. Monocytes and myofibroblasts may produce nitric oxide as a consequence of the induction of nitric oxide synthase (the inducible enzyme isoform [8]); this could account for the development of a nonthrombogenic surface despite the prevalence of an extensive collagenous extracellular matrix.
The data obtained by Rafii and associates support this theory: the pseudointima formed on the textured biomaterial surfaces of LVADs contains a variety of cell types that include hematopoietic lineages. Rafii and associates propose that the hematopoietic cells contribute to the development of a nonthrombogenic surface on the textured biomaterials. They give evidence for the presence of hematopoietic stem cells based on CD-34 expression, which suggests that lineages may arise as a consequence of cytokine production by cells populating the fibrin-rich matrix. The differentiation of hematopoietic stem cells is highly dependent on interleukin-3 or peptides that possess similar activity, such as mast cell growth factor, multicolony-stimulating growth factor, and P-cell growth factor [9]. The cellular source of interleukin-3 in the developing pseudointima remains to be identified, but it is becoming evident that there are significant deposits of cellular material 4 to 5 days after implantation.
Many aspects of this theory remain to be tested, but the new insights provided by Rafii and associates suggest that hematopoietic cells may play an important initiating role in the formation of a nonthrombogenic pseudointima in implanted LVADs. Hematopoietic cell interaction with the fibrin coagulum may be an explanation for the excellent results achieved with this device in a large number of patients.
In summary, the source and nature of the cells on the blood-contacting surfaces of LVADs remain controversial, but the beneficial clinical effects are proven. The low incidence of thromboembolic complications in patients with these devices is remarkable (4%, 3/40 in a recent multicenter clinical trial [10]), especially because most of the patients in the trial were given only antiplatelet therapy. In addition, the thrombus was not attributed to the pump in any of these high-risk patients with severe cardiomyopathy. Our understanding of this complicated field has increased through the work of Rafii and associates. Their article in this issue of The Annals is an important addition to the literature on biocompatibility of ventricular assist devices.
Acknowledgments
We acknowledge support for our research (RO1 HL53233) from the National Heart, Lung and Blood Institute, National Institutes of Health.
Footnotes
Address reprint requests to Dr Scott-Burden, Texas Heart Institute, PO Box 20345, MC 2-255, Houston, TX 77225-0345.
References
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
