MO765CENTRAL VEIN STENOSES IN HD PATIENTS

Abstract

Abstract Background and Aims The prevalence of central vein stenosis (CVS) in patients on hemodialysis (HD) is difficult to be assessed directly. This is mainly caused by the variety of clinical signs and the high frequency of asymptomatic CVS. Aim: to assess the frequency of occurrence of various CVS forms in HD patients. Method The retrospective observational study is based on the results of treatment of 1865 HD patients who underwent diagnostic and therapeutic procedures on vascular access in our center. In case of vascular access dysfunction, patients were examined according to a local protocol: ultrasound of the peripheral (to exclude lesion of peripheral AVF segments) and central veins (over the available length), followed with CT-angiography or percutaneous angiography, if necessary. Results AVF/AVG dysfunction was observed in 29.4% of patients (549 of 1865). 211 patients were diagnosed with CVS. The prevalence of CVS was 11.3% (211 of 1865) among all HD patients and 38.4% (211 of 549) in patients with AVF dysfunction. Among patients with CVS, 37% (78 of 211) had vein lesions without clinical symptoms or with minimal manifestations (a tendency to decrease KT/V). The prevalence of asymptomatic CVS was 4.2% (78 of 1865) in the general population of HD patients and 14.2% (78 of 549) in patients with AVF dysfunction. In case of asymptomatic CVS it was detected by an ultrasound examination during CVC implantation (N=38), during unsuccessful attempts to implant CVC (N=29), in the case of recurrent AVF thrombosis without underlying peripheral segments lesion (N=9) or during echocardiography (N=2). The prevalence of asymptomatic CVS among patients without AVF dysfunction was 5.9% (78 of 1316). True prevalence of subclinical CVS among HD patients without obvious signs of AVF dysfunction may vary widely. A total of 48.8% (103 of 211) of all CVS cases were treated. At the same time, in 10.7% (11 of 103) of cases, patients did not present symptoms of CVS, and surgery was performed due to recurrent AVF thrombosis without damage of the peripheral parts of AVF. Patients with clinically manifest CVS who received endovascular interventions had a significantly higher risk of AVF loss compared to patients with asymptomatic CVS: HR=2.566 [95% CI 1.706; 3.86], log rank p<0.0001. However, patients with an asymptomatic CVS had a higher risk of AVF function loss compared to the general HD population (HR=2,051 [95% CI 1,243; 3,384], log rank p= 0.0004) – fig. 1. The use of CVC is a known risk factor of CVS development. We analyzed the relationship of CVS risk with multiply CVC placements and catheter dwell time using the Cox proportional hazards regression model (fig. 2). In the univariate model, a greater No of CVCs as well as longer time in place increased the risk of CVS. In the multivariate model (χ2=105.516, df=2, p<0.0001), catheter dwell time was no longer associated with an increased risk of CVC, while the mean number of inserted catheters remained an important risk factor. Conclusion The prevalence of both symptomatic and asymptomatic forms of CVS in HD patients is high. Patients with vascular access dysfunction should be carefully examined to identify the asymptomatic CVS. The mean No of catheterizations is a more important risk factor of CVS than longer catheter dwell time.