Objectives: Creation of upper extremity arteriovenous access ipsilateral to central venous lesions can lead to severe arm swelling and pain in addition to thrombosis of the arteriovenous access. Although historically associated with poor outcomes, femoral arteriovenous graft (fAVG) are an option in these cases, as is the Hemodialysis Reliable Outflow (HeRO) device. This study aims to provide a contemporary comparison between HeRO and fAVG.
Methods: A single center, retrospective analysis of consecutive fAVG and HeRO was conducted evaluating index arteriovenous accesses placed between 2014 and 2023. Cases were identified using the local Vascular Quality Initiative (VQI) database and supplemented by review of surgeon case logs. Medical history, demographics, and operative details were obtained utilizing a combination of VQI data and electronic medical record. Data were analyzed using standard statistical tests, Kaplan-Meier survival techniques, and multivariate Cox proportional hazards (PH) and logistic regressions.
Results: Data for 99 patients (57 HeRO, 42 fAVG) were collected. Median follow-up was 464 days. Between the groups, there were no differences (P=NS) in female sex at birth (HeRO 53%, fAVG 57%) and being non-white (HeRO 49%, fAVG 67%). The HeRO patients were older (HeRO 67.3, fAVG 57.6 years, P=.002), had higher BMI (HeRO 31.3, fAVG 27.8, P=.03), had fewer previous AV accesses (HeRO 1.4, fAVG 3.6, P<.001), and were more likely to be diabetic (HeRO 73%, fAVG 50%, P=.015). The ability to use the arteriovenous access for hemodialysis was similar in both groups (HeRO 87.7%, fAVG 78.6%, P=NS). Mortality at 30 days occurred more frequently in the fAVG group (HeRO 3.5%, fAVG 11.9%); multivariate logistic regression also suggested that fAVG was associated with increased 30-day mortality (OR 13.3, P=.025). One year after being successfully used for hemodialysis, primary patency (HeRO 35%, fAVG 53%; Figure 1), primary assisted patency (HeRO 44%, fAVG 59%), and secondary patency (HeRO 62%, fAVG 81%) were better in the fAVG group. Multivariate PH analysis suggested that primary patency and primary assisted patency were similar in both groups (P=NS), while fAVG was associated with improved secondary patency (HR 0.25, P=.014, 95% CI 0.08-0.76). Graft removal due to infection occurred more frequently in the fAVG group (HeRO 7%, fAVG 21%, P=.036).
Conclusions: In this contemporary series comparing fAVG and HeRO, there was no difference in primary and primary assisted patency, though there was improved secondary patency among the fAVG patients. fAVG were associated with higher rates of perioperative mortality and graft infection.
