HOMOCYSTEINE AND VASCULAR ACCESS THROMBOSIS IN HEMODIALYSIS PATIENTS

Abstract

Background: Vascular access remains the Achilles’ heel of successful hemodialysis, and thrombosis is the leading cause of vascular access failure. Hyperhomocystinemia is common in hemodialysis patients and is associated with venous and arterial thrombosis in patients without end-stage renal disease. Subjects and methods: In the study, 65 hemodialysis patients with native arteriovenous fistula were included. Two groups of patients were defined: group A including 45 patients with their vascular access either never or only once thrombosed, and group B including 20 patients with two or more thromboses of their vascular access. We determined serum concentrations of total homocysteine (immunoassay, Abbott) in our patients. Results: In 63 (96.9%) patients, hyperhomocystinemia was presented. There was no statistically significant difference between group A and B regarding age, gender and duration of hemodialysis treatment. Total homocysteine concentrations were higher in group A (42.1 ± 18.6 µmol/l) than in group B (36.1 ± 18.1 µmol/l) patients but the difference was small and not statistically significant. Conclusion: We found no significant differences in total homocysteine concentrations between group A (thrombosis non-prone) and group B (thrombosis prone) patients. Our results suggest that thrombosis of native arteriovenous fistulas may not be caused by hyperhomocystinemia in these patients.

• End-stage renal failure
• Hemodialysis
• Homocysteine
• Vascular access thrombosis

INTRODUCTION

Vascular access is necessary for hemodialysis and still remains the Achilles’ heel of successful hemodialysis [[1]]. The preferred form is native arteriovenous fistula that has the longest survival time and low complication rate [[1]], [[2]], [[3]]. Hemodialysis vascular access dysfunction due to thrombosis is the most common cause of hospitalization among maintenance dialysis patients [[2]], [[4]] and the annual cost of this morbidity is approximately $1 billion in the United States [[5]]. Thrombosis usually results from stenotic lesions caused by progressive neointimal hyperplasia in the venous outflow system [[1]], [[5]]. It is probably caused by a turbulent flow and is leading to increased shear stress stimulation of local growth factors or their release from platelets and leukocytes [[1]]. Risk factors for hemodialysis access thrombosis are ill-defined [[5]], [[6]] and currently, the only definitive risk factor is a placement of a synthetic graft rather than native arteriovenous fistula (AVF) [[3]], [[5]], [[6]]. Risk factors that have been suggested in single studies include the presence of diabetes [[7]], location of the graft [[3]], age greater than 65 years [[6]], time until use of graft after surgical creation [[3]], hypoalbuminemia [[8]], elevated lipoprotein(a) [[9]], elevated serum LDL cholesterol and triglycerides [[10]] and anticardiolipin antibodies [[11]], [[12]]. There is higher risk of blood access failure in patients with severe vascular endothelial injury, a useful marker of which is a high plasma thrombomodulin [[13]].

Hyperhomocystinemia has been identified as an independent risk factor of cardiovascular, carotid-artery stenosis and venous thrombosis in patients without end-stage renal disease [[14]], [[15]], [[16]], [[17]], [[18]]. Hyperhomocystinemia is a common find in patients with end-stage renal failure [[19]], [[20]], [[21]], [[22]], [[23]], [[24]] and is a risk factor for cardiovascular events in dialysis populations [[25]], [[26]], [[27]]. The relationship between homocysteine concentrations and thrombosis of another specific vascular site (AVF) is less clear. The aim of our study was to determine whether hyperhomocystinemia is associated with AVF thrombosis in hemodialysis patients.

PATIENTS AND METHODS

In our study, 65 hemodialysis patients with native AVF were included, 24 were women and 41 were men. Two groups of patients were defined:

• Group A includes 45 patients with their vascular access either never or only once thrombosed. Average age was 52.5 years (ranging from 19 to 78 years, S.D. ± 15.9); 17 (37.8%) were women and 28 (62.2%) were men. Average duration of hemodialysis treatment was 57.4 months (ranging from 12 to 217 months, S.D. ± 49.8).

• Group B includes 20 patients with two or more thromboses of their vascular access. Average age of group B patients was 55 years (ranging from 35 to 70 years, S.D. ± 10.6); 7 (35%) were women and 13 (65%) were men. The average duration of hemodialysis treatment was 46.6 months (ranging from 12 to 138 months, S.D. ± 38.9).

The vascular access failure in less than 1 month after construction was considered to be surgical failure and was not included as a thrombotic event in this study. Patients with diabetes were excluded.

We determined serum concentrations of total homocysteine (immunoassay, Abbott) in all patients included in the study. Fasting samples were obtained prior to dialysis in patients who had morning hemodialysis run. Hultberg et al. [[28]] previously demonstrated that levels of total homocysteine do not change after meals in patients with advanced renal failure despite an increase in plasma methionine, and Sirrs et al. [[29]] didn’t find the difference between fasting and nonfasting levels of total homocysteine in hemodialysis patients, so it was felt appropriate to obtain nonfasting specimens for patients who attended dialysis in the afternoon or evening.

RESULTS

Average total homocysteine concentrations for all patients included in the study were 40.2 ± 18.5 µmol/l (ranging from 9.6 to 96.7 µmol/l). In 63 (96.9%) patients, hyperhomocystinemia (>15 µmol/l) was present. Only one patient in group A had normal plasma total homocysteine concentration, in 44 (97.7%), hyperhomocystinemia was observed. Also, only one patient in group B had normal plasma total homocysteine concentration and in 19 (95%) patients, hyperhomocystinemia was observed. There was no statistically significant difference between group A and group B regarding age, gender and duration of hemodialysis treatment.

Average total homocysteine concentrations were 42.1 ± 18.6 µmol/l (ranging from 12.6 to 96.7 µmol/l) in group A patients. In group B patients, average total homocysteine concentrations were 36.1 ± 18.1 µmol/l (ranging from 9.6 to 70.7 µmol/l). Total homocysteine concentrations were higher in group A, in patients with their vascular access either never or only once thrombosed, but the difference was small and not statistically significant.

DISCUSSION

Elevated total homocysteine concentrations (hyperhomocystinemia) are common in patients with end-stage renal disease (ESRD) treated with hemodialysis or peritoneal dialysis [[19]], [[20]], [[21]], [[22]], [[23]], [[24]]. The prevalence of hyperhomocystinemia was 90.8% among hemodialysis patients and 67.4% among patients treated with peritoneal dialysis in the study of Moustapha et al. [[21]]. In our study, the prevalence of hyperhomocystinemia in hemodialysis patients was also very high (96.9%). The mechanism of hyperhomocystinemia in ESRD patients is complex. Dennis and Robinson [[23]] reviewed the pathophysiology of increased homocysteine concentrations in ESRD. The proposed mechanisms are poor hepatic uptake of homocysteine, poor availability of essential cofactors or substrates [[23]] or the presence of circulating enzyme inhibitors [[30]]. Decreased renal excretion of homocysteine is not likely an explanation, because in normal subjects, only small amounts of sulfur amino acids are excreted in urine [[30]]. Folic acid and vitamins B₆ and B₁₂ play important roles in homocysteine metabolism, and chronic renal failure influences plasma concentrations of these vitamins, hyperhomocystinemia seen in ESRD may be related to concentrations of these vitamins [[19]], [[20]], [[21]], [[26]], [[31]], [[32]]. In studies, supplementation with folic acid (supernormal doses) reduces circulating homocysteine concentrations to a certain extent in these patients [[22]], [[32]]. Interestingly the post-transplant reduction of only 14% in total homocysteine concentration was far smaller than expected with respect to renal function [[33]].

Hyperhomocystinemia is now well established as a risk factor for accelerated atherogenesis and coronary heart disease, also in dialysis patients [[24]], [[25]], [[26]], [[27]], [[34]]. Hyperhomocystinemia has been reported to be associated with early-onset venous thrombosis and recurrent venous thrombosis in patients without ESRD [[15]], [[16]], [[17]], [[18]], [[35]]. Less clear is the relationship between hyperhomocystinemia and thrombosis of another specific vascular site—AVF, which still remains the Achilles’ heel of successful hemodialysis and is the most common cause of hospitalization among maintenance dialysis patients [[1]], [[2]], [[4]]. Manns et al. [[36]] found no association between homocysteine concentrations and vascular access thrombosis. Approximately three-quarters of their patients had native AVF and one quarter had synthetic polytetrafluoroethylene graft. The only variable in the study associated with higher risk for graft thrombosis was a type of vascular access placed [[36]]. In a prospective study, Sirrs et al. [[29]] did not demonstrate the relationship between total homocysteine and risk of thrombosis of native AVF (vascular access failure). In the study of Tamura et al. [[37]] where only patients with polytetrafluoroethylene arteriovenous grafts were included, authors also did not find significant difference in plasma homocysteine concentrations between thrombosis prone and control groups. They concluded that thrombosis of arteriovenous grafts might not be caused by hyperhomocystinemia in patients treated with hemodialysis [[37]]. On the contrary, Shemin et al. [[38]] found that elevated plasma total homocysteine concentrations appear to confer a graded, independent risk for vascular access thrombosis in hemodialysis patients. Authors [[38]] also found reduced risk for thrombotic events in those with AVF versus synthetic grafts, regardless of homocysteine level, what is in accord with the most consistent observation from earlier prospective studies of hemodialysis access thrombosis [[5]], [[6]]. In a small study, Ducloux et al. [[39]] revealed significantly higher homocysteine concentrations in patients with recurrent access thrombosis compared with those with one or less episodes of thrombosis. It is unclear what type of vascular access was studied (native AVF or synthetic grafts). In our study we found no statistically significant differences in total homocysteine concentrations between thrombosis non-prone (group A) and thrombosis prone (group B) hemodialysis patients, and so our results are in accord with most other studies [[29]], [[36]], [[37]]. Total homocysteine concentrations were even higher in thrombosis non-prone patients (group A) but the difference was small and not statistically significant. Manns et al. [[36]] also found that total homocysteine concentrations were higher in patients who had no episodes of access thrombosis compared to patients with at least one documented episode of vascular access thrombosis (differences were not statistically significant). Interestingly, also in two other studies, homocysteine concentrations were higher (but not statistically significant) in patients without vascular access failure [[29]], [[37]]. This illustrates the complexity of the relationship between total homocysteine and vascular access failure, if any. As recently reviewed by Selhub and D’Angelo [[40]], at present, there are no data from appropriately designed mechanistic studies that explain how homocysteine may contribute to thrombogenesis. According to studies showing association between homocysteine and vascular access thrombosis, further studies evaluating the hypothesis that lowering homocysteine concentrations may reduce the rate of hemodialysis access thrombosis are needed.

In summary, we found no significant differences in total homocysteine concentrations between thrombosis prone and thrombosis non-prone hemodialysis patients. Therefore, we conclude that thrombosis of native AVF may not be caused by hyperhomocystinemia in these patients. Because of the complexity of the relationship between homocysteine and vascular access failure, long term prospective studies with serial measurements of homocysteine concentrations may better address the direct metabolic implications of elevated total homocysteine in hemodialysis patients.