Back to Annual Meeting Program

Objective: Past studies have indicated the presence of subclinical pulmonary emboli (PE) after open or percutaneous interventions to achieve or restore patency on arteriovenous fistulas (AVF) or grafts (AVG). Although subclinical in presentation, PE may conceivably diminish cardiac function, which can have long standing implications for patient wellbeing. This study investigates the association between type of intervention on AVF or AVG and changes in left ventricular ejection fraction (LVEF) and qualitative estimates of pulmonary hypertension (PH) by pre- and post-procedure cardiac assessment.
Methods: This retrospective study analyzed a cohort of 113 dialysis patients who underwent at least one AVG or AVF procedure and had both pre- and post-procedure echocardiograms between 2/3/2000 and 5/22/12. Patients were separated by intervention type: all methods of operative or percutaneous thrombectomy (29 patients); angioplasty with no history of thrombectomy (44 patients); and other interventions with no history of thrombectomy or angioplasty (40 patients). The last category was used as a negative control. Assessments of LVEF and PH were recorded, with PH qualitatively rated as: 1=none; 2=minimal; 3=mild; 4=moderate; 5=severe. Median time between echocardiogram and intervention was 142 days.
Results:

Intervention Type	Other	Thrombectomy	Angioplasty
Mean LVEF, Pre-intervention	55.20% ± 12.27%	58.67% ± 16.48%	56.67% ± 15.07%
Mean LVEF, Post-intervention	57.86% ± 9.83%	61.42% ± 11.46%	53.45% ± 13.83%
Mean Change in LVEF from Baseline	2.66%	2.75% ± 9.79%, p = 0.97	-3.22% ± 13.05%, p = 0.03
Mean PH, Pre-intervention	2.82 ± 1.20	3.03 ± 1.20	2.85 ± 1.29
Mean PH, Post-intervention	2.77 ± 1.25	3.47 ± 1.26	3.35 ± 1.14
Mean Change in PH from Baseline	-0.12	0.43 ± 1.25, p = 0.25	0.50 ± 1.35, p = 0.16

In comparison to the Other group, the mean change in LVEF after Angioplasty was statistically significant. No statistically significant difference was seen in comparing mean change in LVEF after Thrombectomy and Other or when considering changes in PH among the cohorts, although a worsening trend in PH after both Thrombectomy and Angioplasty was seen.
Conclusions: This study found a decrease in LVEF associated with angioplasty interventions on AVFs or AVGs. Whether this is an effect from angioplasty or reflects something inherent about the patient that predisposes him to both deterioration in LV function and poorly functioning dialysis access cannot be determined in this small retrospective study. Further studies to elucidate these findings as well as to assess quantitative changes in PH following dialysis access intervention are warranted.