Fate of infected and eroded hemodialysis grafts and autogenous fistulas.

Between 1970 and 1991, we placed 1,090 grafts (bovine and polytetrafluoroethylene) in 1,041 patients and created 1,034 autogenous fistulas in 856 patients for hemodialysis. Subsequent revisions for complications resulted in a total of 3,944 operations performed in patients with grafts and 1,633 operations in patients with autogenous fistulas. A total of 255 infections developed in 158 of the patients with grafts, whereas 8 infections developed in 7 patients with autogenous fistulas. The puncture infection rate was 5%/yr (12%/yr for a second puncture infection). The clean wound infection rate was 3% for grafts and 0.4% for autogenous fistulas. We made an attempt to salvage the graft, usually with a segmental bypass, in 75% of patients with a graft infection. Grafts were salvaged in 80% of patients in whom salvage was attempted (60% of all patients with an infection). The results in the few patients with infected autogenous fistulas were relatively poor.

Intimal fibromuscular hyperplasia at the venous anastomosis of PTFE grafts in hemodialysis patients. Clinical, immunocytochemical, light and electron microscopic assessment.

Failure of arteriovenous communications used for chronic hemodialysis was studied during sequential 5-year periods after placement of either endogenous Brescia-Cimino (B-C) fistulas (50 patients) or polytetrafluoroethylene (PTFE, Gore-Tex) grafts (66 patients). Venous stenosis near the anastomosis was the reason for failure in 45% of PTFE grafts compared with 16% of B-C fistulas (p less than 0.001). Failure occurred, on average, 16 months after PTFE graft placement compared with 22 for B-C fistulas (p = NS). Proximal vein segments removed from five failed and two functioning PTFE graft communications were studied using light and electron microscopy and immunocytochemical techniques. All venous segments removed during surgical shunt repair exhibited a marked intimal hyperplasia. The intimal cellular component was almost exclusively smooth muscle. Accumulation of intracellular lipid droplets was not seen. Foam cells as well as extracellular lipid deposits were absent; macrophages and lymphocytes were absent from the zone of proliferation. Ultrastructural examination revealed a large proportion of extracellular matrix surrounding smooth muscle cells in the neointima. Collagen and elastin were present in the extracellular matrix, in greatest concentration deeper in the intima. Closer to the lumen, most of the extracellular volume consisted of proteoglycan. Hemosiderin was absent from the lesions as were consistent signs of luminal and intimal fibrin. Uniform intimal gradients of actin, collagen, and proteoglycan suggest that this is a steadily progressive, rather than episodic, proliferative response. These clinical and histologic observations and an analysis of hemodynamic stresses support the postulate that upstream release of platelet-derived growth factor, and possibly, shear-induced intimal injury stimulate this response. This myointimal proliferative process provides a readily accessible model of fibromuscular hyperplasia in humans; its understanding may lead to effective methods for its prevention and may provide clues to the pathogenesis of arteriosclerosis.

Transluminal angioplasty versus surgical repair for stenosis of hemodialysis grafts. A randomized study.

A randomized study of transluminal angioplasty versus surgical revision for stenosis at the venous end of hemodialysis grafts was carried out. Median patency for the surgical group was 12 months versus only 4 months for the transluminal angioplasty group. This difference was statistically significant (p less than 0.01). In addition, the average cost for surgical revision was slightly less than that for transluminal angioplasty. Therefore, for long-term patency, surgical treatment is the method of choice, although under special circumstances there may still be a role for transluminal angioplasty.

Scribner shunt conversion to arteriovenous fistula.

Scribner shunts were converted to arteriovenous fistulas in twenty-six patients, with long-term success in twenty patients. Infection with loss of the fistula was a problem in two patients. The fistulas were used for home as well as center dialysis and as a primary blood access site as well as a backup site.