Precision Banding for High-Flow Hemodialysis AccessdContemporary Outcomes and Failure Predictors

High-flow hemodialysis accesses are increasingly recognized as the source of complications (cardiopulmonary, hand ischemia, access malfunction). Among management strategies (access elongation, revision using distal inflow), precision banding (inflow constriction based on real-time physiologic flow monitoring) offers a technically straightforward approach with potentially low morbidity. However, large contemporary series on banding outcomes (and their predictors) are lacking. Banding was performed with 2-0 doubly looped Prolene over an internal balloon or 3- to 4-mm external mandrel with Transonic thermodilution flow volume monitoring (Transonic Systems, Ithaca, NY). Demographics and clinical characteristics were captured during a 10-year period through prospective databases and some retrospective record review. Multivariable logistic predictive regression models (stepwise selection of clinically relevant variables, P < .10 as entry criterion) were completed for thrombosis within 30 days and rebanding within 1 year as well as during the entire period. In total, 397 patients underwent banding during the study period. One patient had early wound bleeding requiring return to the operating room, and none had discovered complications of band erosion. The infection rate was 0.5% (2/397). Median (interquartile range) follow-up was 157 (52-373) days. Most accesses were upper arm with brachial artery inflow (Table I); half were performed prophylactically on the basis of experienced clinical judgment (mean duplex ultrasound flow 2 L/min), and a third were for steal. Median flow rate reduction was 41%. Among 134 steal patients, only 2 required subsequent ligation to control symptoms. The 30-day thrombosis rate after banding was 13 of 397 (3.3%). Including all thrombotic events (25/397 [6.3%]), median time was 9.5 days (2.0-101.5 days). The rebanding rate within a year was 51 of 397 (13%), and median time from banding to rebanding within 1 year was 133 days; 8 of 397 (2.0%) had to be rebanded within 30 days. During the entire period, 103 of 397 (26%) patients needed rebanding. Multivariate logistic regression analysis showed prosthetic graft as the sole predictor of early thrombosis. After adjustment, only smaller banding diameter was independently associated with any thrombotic event (including those beyond 30 days) and prebanding vein diameter with any need for rebanding (Table II). Precision banding offers an effective, low-morbidity approach for high-flow hemodialysis accesses. Early thrombosis is a rare event after precision banding, although in the long term, one in four patients will require rebanding to control flow.Table IPatient cohort characteristicsVariableTotal cohort (n = 397)Demographic characteristics Median age, years [IQR]59 [48, 73] SexFemale (%)152 (38)Male (%)245 (62) RaceWhite (%)183 (46)Latino (%)132 (33)African-American (%)63 (16)Asian (%)8 (2)Other/unknown (%)13 (3)Comorbidities Hypertension (%)368 (93) Diabetes (%)160 (40) Coronary artery disease (%)129 (32) Dialysis (%)296 (75)Access details SiteForearm (%)50 (13)Upper arm (%)347 (87) TypeAutogenous (%)360 (91)Graft (%)37 (9)Inflow arteryBrachial (%)342 (86)Radial (%)51 (13)Ulnar (%)2 (1) Outflow veinCephalic (%)300 (76)Basilic (%)66 (17)Axillary (%)13 (3)Other (%)16 (4)Access in use at time of banding (%)269 (68) Median preoperative flow rate by Duplex, ml/min [IQR]1975 [1510, 2532]Indication for banding Prophylactic (%)203 (51) Enlarging aneurysm/pseudoaneurysm (%)18 (4) Ischemic steal (%)134 (34) High output heart failure (%)42 (11)Procedural details Time from access creation to banding1234 [534, 2283] Median intraoperative pre-banding flow rate by flow meter, ml/min [IQR]1545 [1210, 2080] Median number of bands [IQR]1 [1, 1] Banding diameter3 [3, 4] Technique of precision bandingIntravascular balloon (%)100 (25)External mandrel (%)297 (75) Median intraoperative post-banding flow rate by flowmeter, ml/min [IQR]640 [445, 865] Median percent flow reduction [IQR]58 [39, 73] Open table in a new tab Table IIUnadjusted univariate and adjusted multivariate logistic regression analysis for association of demographic, access, and procedural characteristics with thrombosis and banding failureVariableThrombosis at 30 daysBanding failure at 1 yearUnadjusted analysisAdjusted analysisUnadjusted analysisAdjusted analysisOdds ratio (95% CI)P valueOdds ratio (95% CI)P valueOdds ratio (95% CI)P valueOdds ratio (95% CI)P valueAge0.97 (0.94-1.0).080.97 (0.94-1.01)0.090.99 (0.97-1.00).080.98 (0.97-1.0).09Male1.4 (0.43-4.66).570.79 (0.44-1.44).45Dialysis at time of banding3.76 (0.48-29.32).210.96 (0.48-1.92).90Access in use at time of banding5.32 (0.68-41.4).111.11 (0.5-2.16).75Indication for banding ProphylacticRefRef Enlarging aneurysm/pseudoaneurysm0.001 (<0.001->999).962.19 (0.58-8.32).72 Ischemic steal0.56 (0.14-2.14).972.39 (1.23-4.64).32 High output heart failure1.22 (0.25-5.95).952.19 (0.84-5.67).63Prosthetic graft5.03 (1.47-17.28).014.96 (1.43-17.25).011.45 (0.57-3.68).44Lower arm access1.30 (0.28-6.05).740.26 (0.06-1.10).070.24 (0.06-1.03).05Inflow artery BrachialRefRef Radial1.23 (0.26-5.71).980.24 (0.06-1.04).98 Ulnar0.001 (<0.001->999).980.001 (<0.001->999).97Outflow vein CephalicRefRef Basilic0.001 (<0.001->999).940.14 (0.66-2.86).20 Axillary2.42 (0.29-20.4).940.57 (0.07-4.54).72 Other4.14 (0.83-20.7).930.46 (0.06-3.58).52Number of bands1.08 (0.32-3.65).911.00 (0.53-1.91).99Pre-banding vein diameter1.00 (0.92-1.09).981.02 (0.98-1.06).31Banding diameter0.47 (0.15-1.51).210.98 (0.56-1.69).93% Flow reduction0.99 (0.97-1.02).631.00 (0.99-1.01).81Intraoperative post-banding flow rate (flowmeter)0.99 (0.99-1.00).571.00 (0.99-1.00).77 Open table in a new tab