Ultrasound assessment of upper limb arteriovenous fistulas in hemodialysis patients at the Yaounde University Teaching Hospital

Data was collected via interviews and Doppler ultrasound assessments of the arteriovenous fistulas. The blood vessels were scanned in their longitudinal axis for the measurement of velocity and flow, while the short axis was used to measure diameter. All Doppler measurements were done at an insonation angle of less than or equal to 60°, with the Doppler window parallel to the vessel wall and occupying most of the vessel diameter. Vascular definitions used were those of the American Institute of Ultrasound in Medicine (AUIM).

Data entry and analysis was done using Statistical Package for Social Sciences (SPSS) version 28 software. Quantitative variables were expressed as means ± standard deviation (SD) when the distribution was considered normal. Pearson’s chi-squared test or Fischer’s exact test was used for the comparison of proportions, and the ANOVA test was used for the comparison of means. Univariate analysis was used to investigate independent associations to the outcome. A p-value <0.05 was considered statistically significant.

This study was carried out after obtaining ethical clearance (Reference number: 288/2022) from the Institutional Ethical Review Board at the Faculty of Medicine and Biomedical Sciences of the University of Yaoundé I. Informed oral consent was obtained from the participants after a detailed oral explanation of the aim of the study and the procedure involved. Patient data was kept strictly confidential.

A total number of 41 participants were enrolled, the majority being male patients, representing 56.1% of the study population. The mean (standard deviation) age of the study population was 48.39 ± 16.56 years. In 78.04% of cases, the fistula scanned was the first on the upper limb and in 21.96% it was the second. We did not have any higher order AVFs.

The main comorbidity found in our study population was hypertension, found in 34 participants, which represented 83% of the study population. Eleven participants suffered from diabetes, and five were persons living with HIV (Tab. 1).

The mean (SD) BMI of the participants was 24.6 (4.21) kg/m², with median (IQR) and mean (SD) dialysis vintage of 16.00 (8–32) months and 23.60 (22.40) months. The mean (SD) AVF age of the AVFs evaluated in our study population was 20.25 (18.62) years. Two types of AVFs, radiocephalic and brachiocephalic, were identified in our study population.

Evaluating the AVFs, we identified six fistulas which were stenosed with PSV ratio >3, and 18 using the criteria of absolute PSV >375 cm/s (Tab. 2). In the study population, a total of 56.1% of the AVFs were stenosed by either one or more of the definitions. In total, 78.05% of the AVFs had at least one vascular modification.

Reverse flow in the artery segment distal to anastomosis, aneurysms (Fig. 1) in the efferent vein, and venous luminal flaps were the most common complications identified in the scanned AVFs. We also had seven cases of non-stenosing thrombi within the efferent veins (Tab. 2).

Fig. 1.

Image showing a transverse section through an aneurysm

The mean flow volume in the AVFs as seen in Tab. 3 was 680.47 ml/min (Fig. 2), with 34.1% of the fistulas having a low flow volume. Also, the resistivity index in the afferent artery was lower than or equal to 0.5 in 37 out of the 41 fistulas studied.

Fig. 2.

Measurement of flow volume within the efferent vein

Table 4 shows that among the factors evaluated diabetes, aneurysms, and venous luminal flaps were found to be the determinants of flow volume. Stenosis of the AVF was not associated with the flow volume.

Aneurysms with a prevalence of 53.7% were the most commonly encountered vascular abnormality. This finding concurred with that reported by Pietura et al. who had a rate of 54%, while contrasting with the 27.65% obtained by Fokou et al. It is possible that the shorter follow-up time of their cohort might explain the difference. The prevalence of thrombosis in our study was 17.1%, which was not very different from the prevalence reported in the studies by Fokou et al. and Pietura et al., at 10.63% and 7% respectively^(5,7).

Venous luminal flaps were frequent in our study, with a prevalence of 22%. Pietura et al. described venous wall irregularities in 20% of the AVFs⁽⁵⁾.

We had a high prevalence of stenosis at 56.1%, which is similar to that obtained by Pietura et al. at 64%, though different from that reported by Fokou et al. (10.63%)^(5,7). The main reason for the differences in prevalence could be the difference in the method of AVF evaluation, which was mainly clinical in their study and sonographic ours. As concerns the site of stenosis, the venous juxta-anastomotic segment (75%) of the efferent vein was shown to be the main site for stenosis, contrary to the study by Pietura et al.⁽⁵⁾, who observed the anastomosis (57%) as the main site.

Our mean flow volume of 680.47 (365.98) ml/min, obtained by measuring flow in the efferent vein, was similar to that reported by Farrington et al. who, by performing measurements at the same site, obtained an average volume flow of 665 ml/min in unassisted mature, assisted mature, or failed-to-mature fistulas⁽⁸⁾.

Our mean flow volume was, however, lower than those obtained by Pietura et al. and Colombo et al. who had the flow volumes of 1204.1 ml/min and 1437 ml/min respectively^(5,9). Both authors measured their flow volume in the brachial artery, which could explain the difference observed. Importantly, the American Institute of Ultrasound in Medicine recommends scanning the efferent vein for flow volume⁽⁴⁾. Also, the divergent results with those in the study by Pietura et al. might be explained by the fact that the authors included only well-functioning fistulas in their study.

We carried out a univariate analysis to determine the factors associated with low flow rates in AVFs, and identified diabetes (p = 0.003), aneurysms (p = 0.026), and luminal venous flaps (p = 0.013) to be determinants of flow volume. Based on a greater proportion of diabetic patients within the low flow volume group, it could be inferred that diabetes is a factor associated with low flow volumes. This conclusion is similar to that obtained by Tonelli et al. who found diabetes to be independently associated with flow volume⁽¹⁰⁾. On the contrary, we had more aneurysms and luminal venous flaps with appropriate flow volumes, suggesting that their absence is linked to low flow volumes.