PROJECT SUMMARY
Almost 600,000 patients with end-stage kidney disease (ESKD) are on hemodialysis (HD) in the United States.
An arteriovenous (AV) access is the preferable access for HD. Dysfunction of AV access is common and
predominantly occurs due to stenosis in the AV access, which is typically managed by a cross-disciplinary
team of vascular surgeons, interventionists, and nephrologists. Based on Medicare data analysis, ~ 1.0-1.2
million AV access procedures are performed annually in the US. Although angiography is the standard
diagnostic modality for AV access dysfunction, angiography introduces errors in stenosis estimation and
cannot characterize the morphology of stenosis or measure the pressure across the stenosis. These
parameters are integrated in the decision to intervene in other stenotic vascular diseases and predict their
response to an intervention. The latter aspect is important as AV access restenosis occurs in 50% of
intervened AV accesses, and some stenoses may undergo accelerated restenosis. Therefore, a method for
accurately diagnosing AV access stenoses and predicting the response to an intervention is imminent. We
address the current limitations in AV access dysfunction diagnosis using intravascular ultrasound (IVUS) at the
point of care in ESKD patients.
IVUS provides a 3-D view of the stenosis, thereby enhancing the accuracy of estimating it. Lack of
contrast agent or bony interference and the ability to obtain morphological characterization and pressure
gradients are additional advantages. IVUS is used in the management of other stenotic vascular diseases,
where it has altered the management of > 50% of patients examined via conventional angiography. We
recently completed a pilot study evaluating the feasibility, safety and initial sensitivity of IVUS over angiography
in patients presenting with AV access dysfunction. IVUS detected 14 more stenoses, which were missed by
angiography. Angiography underestimated the severity of stenosis in 30% of lesions. Two morphological
features were extracted using a deep neural network. The current grant proposal examines the superiority of
IVUS over angiography. We will also examine the implications of IVUS-based stenosis, morphological indices
and hemodynamic parameters in the intervention of AV access stenosis using radiological response and target
lesion primary patency rate. An interdisciplinary team of vascular surgeons, interventional radiologists,
nephrologist and clinical research will complete this study, which has the potential to establish IVUS as a new
standard-of-care imaging modality for AV access stenosis.
