Abstract
Background
Arteriovenous-fistula (AVF) are crucial for hemodialysis access, yet they frequently experience early failure. While studies have identified potential patient and clinical risk factors, these findings remain inconsistent. This inconsistency might be attributed to the varying definitions of “early failure”. Our retrospective cohort study aimed to evaluate how common risk factors predict four frequently early-failure criteria: thrombosis/stenosis, <500 ml/min blood flow, <5 mm vein diameter, and ≥6 mm deep vein. We also assessed how well these risk factors predict early failure defined as meeting at least one of these criteria. Additionally, we examined the predictive ability of vein-distensibility, a previously overlooked factor in AVF failure.
Methods
Consecutive patients with first-time AVF employing standard minimum preoperative artery- and vein-diameters (1.8–2.0 mm) who underwent first Doppler-ultrasound (DUS) at ≤4 months in 2016–2022 were identified. Early AVF failure was defined as the presence of at least one of the following conditions on the first DUS: poor blood flow (Qa), poor vein diameter, poor vein depth, and thrombosis/stenosis. Factors associated with early AVF failure were explored with multivariate analyses.
Results
105 patients were eligible and 63 (60%) had an early AVF failure. The only strong predictor of early failure was low vein-distensibility (Odds ratio = 0.57, 95% confidence intervals [CIs] = 0.38–0.83, p = 0.005). Female sex only predicted too-deep veins (Odds ratio = 14.29, 95% CIs = 2.00–100, p = 0.024).
Conclusions
venous distensibility may be a useful early-failure determinant when minimum preoperative vessel-diameter limits are met. Moreover, the female sex is associated with too-deep AVF veins.
Keywords: Arteriovenous fistula, early failure, vein distensibility, risk factors, female sex
1. Introduction
In the last two decades, the management of dialysis patients was dominated by the Fistula First Initiative, which promoted the primary use of autologous arteriovenous fistula (AVF) [1]. However, 20–60% of AVFs cannot support dialysis 3–6 months after creation due to failure-to-mature or thrombosis/stenosis [2]. This together with growing clinician calls to provide more holistic approaches led the Kidney Disease Outcomes Initiative (KDOQI) to recommend treatment according to Lifeplan considerations, namely, the choice of kidney replacement modality should be guided by the present and anticipated circumstances of each patient. Nonetheless, AVFs remain an important treatment modality and the search to improve its outcomes continues.
A strong predictor of AVF early failure is distal AVF (radiocephalic) because the wrist/forearm has smaller caliber vessels than the proximal arm: this limits the achievement of the high blood flow (Qa) needed for AVF function [3]. Indeed, smaller preoperative vessel diameters, as determined by Doppler-ultrasound (DUS), predict early failure, and ∼2 mm minimum distal-AVF vessel diameters are now generally routinely employed [4–6]. Other risk factors (age, sex, BMI, diabetes, hemoglobin, and central venous catheter) demonstrate highly variable associations and even AVF location is not always a significant predictor. These interstudy variations have greatly complicated the identification of reliable risk factors. Moreover, the relationships between potential early-failure risk factors and the individual early-failure criteria themselves have rarely been assessed.
Another criterion that could be used to define early failure is AVF-vein distensibility: a few studies show that early failure is associated with low vein distensibility. This is true whether vein distensibility is measured preoperatively with a tourniquet, intraoperatively by clamping and infusing saline, or postoperatively by comparing the diameter before and after AVF creation [6–11]. However, the potential of this criterion for predicting early failure has largely been overlooked.
To explore the role of patient factors in early failure further, we retrospectively identified patients with first-time AVFs that bore ≥1.8 diameter artery and ≥2 mm-diameter vein preoperatively and examined the multivariate relationships between commonly reported risk factors and early AVF failure. We also explored the potential of postoperative vein-distensibility to predict early failure and its relationship with the risk factors.
2. Patients and methods
2.1. Study design
This retrospective single-center cohort study was conducted in Metz-Thionville Regional Hospital, Grand Est, France, and all procedures conformed to the principles of the Declaration of Helsinki. All patients were informed that their surgery-related data would be used for this study. All consented to this possibility. The consent procedure was conducted by the reference methodology MR-004 of the French National Commission for Information Technology and Liberties (No. 2208231-v0). The research is registered on clinicaltrials.gov (NCT05552482) and the French Health Data System (https://www.health-data-hub.fr).
2.2. Study subjects
All consecutive ≥18-year-old patients with end-stage renal disease who underwent first-time AVF-creation surgery from 7 April 2016 to 21 December 2022 and then preoperative and postoperative DUS at ≤4 postoperative months were identified from prospectively maintained electronic medical records. Radiocephalic-AVFs (RCAVFs; created from the radial artery and cephalic vein at the wrist and brachiocephalic-AVFs (BCAVFs; created from the brachial artery and cephalic vein in the proximal forearm), were included. Brachiobasilic-AVFs or arteriovenous grafts were not included. All anastomoses employed the end-to-side technique. AVF choice was determined by patient history, physical examination, and preoperative DUS. Preoperative artery- and vein-diameter thresholds were ≥1.8 and ≥2 mm, respectively. Patients with preoperative central-venous catheter (CVC) were included. AVFs used for intermittent plasmapheresis were excluded.
2.3. Doppler-ultrasound
Preoperative and ≤4-month postoperative DUS was conducted with ARIETTA 70 (Dopplers-Hitachi Ltd. Zug, Switzerland) or EPIQ CVXI (Philips Medical Systems, Bothell, WA, USA) machines. Two angiologists measured preoperative artery- and vein diameters and postoperative Qa, vein diameter, and vein depth. For RCAVFs, distal cephalic vein diameter and vein depth were measured approximately 5 cm downstream of the anastomosis at mid-forearm. For BCAVFs, proximal cephalic vein diameter and vein depth were measured approximately 5 cm downstream of the anastomosis at mid-arm. Qa was measured three times on the brachial artery 3-, 5-, and 7-cm proximal to the anastomosis and averaged. All measurements were taken under a tourniquet in a warm room. Postoperative vein-distensibility was defined as the difference between postoperative and preoperative vein diameters.
2.4. Independent variables
Independent variables were age, sex, body mass index (BMI), diabetes, preoperative hemoglobin, CVC use, anesthesia (general/locoregional), AVF laterality, AVF location, preoperative artery diameter, preoperative vein diameter, and postoperative vein-distensibility.
2.5. Definitions
Early failure was defined as postoperative ≤500-mL/min Qa, ≤5-mm vein diameter, ≥6-mm vein-depth, and/or thrombosis/stenosis on first DUS. Stenosis was defined as ≥50% luminal decrease compared to the normal adjacent vein.
BMI was categorized as nonobese (≤30 kg/m2) and obese (>30 kg/m2). Diabetes mellitus was defined as fasting plasma glucose ≥126 g/dL, 2-h postprandial plasma glucose ≥200 mg/dL, glycosylated hemoglobin (HbA1C) ≥6.5%, or taking anti-diabetes medications.
2.6. Study objectives
The primary objective was to determine the patient/clinical factors that predicted early failure. The secondary objectives were to determine the factors that predicted postoperative Qa, vein diameter, vein depth, thrombosis/stenosis, or vein distensibility.
2.7. Statistical analyses
All data were complete, except for hemoglobin and CVC-use data, which were missing for one patient each. RCAVFs and BCAVFs were combined, by statistical practices observed in the literature. Continuous variables were expressed as median (interquartile range [IQR]). Categorical variables were expressed as n (%). For the primary objective, AVFs were categorized according to whether they met the early-failure definition. The two groups were compared in terms of all variables by Student’s t-test and Chi-square/Fisher’s exact test. Logistic multivariate analyses (binomial-regression model) were then conducted with age, sex, obesity, diabetes, hemoglobin, CVC use, preoperative artery diameter, and vein distensibility. AVF location and preoperative vein diameter were not included in the model due to collinearity with respectively preoperative artery diameter and vein distensibility. Up to eight factors could be included in each analysis due to the number of 42 patients in the smallest analyzed group (AVF success), following the rule of thumb of 5 patients per factor. The results were expressed as odds ratios (OR) and 95% confidence intervals (CIs). All univariate and multivariate analyses were repeated with each early-failure criterion. Finally, a linear multivariate analysis was conducted with postoperative vein-distensibility as the dependent variable. These data were expressed as Beta (95% CIs). p < 0.05 was considered significant. All analyses were conducted with R (4.0.4) using the gt_summary package
3. Results
Of the 113 AVFs, eight were excluded for the reasons shown in Figure 1. Thus, 105 first-time AVFs were included.
Figure 1.
Flowchart of patient distribution in the study. AVF, arteriovenous fistula; DUS, Doppler ultrasound.
3.1. Preoperative patient characteristics
Median patient age was 68 (57–74) years, 38% were female, BMI was 27 (24–32) kg/m2, 34% were obese, 54% had diabetes, 48% used a CVC, and 93% received locoregional anesthesia. Two-thirds of AVFs were distal and on the left. Median preoperative artery- and vein diameters for RCAVFs were 3.00 and 3.00 mm, respectively while those for BCAVFs were 5.00 and 3.30 mm, respectively; 2 and 0% of all arteries and veins were respectively below the recommended 2 mm limit [4] (Table 1).
Table 1.
Demographic, clinical, preoperative DUS, and surgical characteristics of the cohort.
| Variablea | n (%) or median (IQR) |
|---|---|
| Age, years | 68 (57–74) |
| Female sex | 38 (36%) |
| BMI, kg/m2 | 27 (24–32) |
| Comorbidities | |
| Diabetes mellitus | 57 (54%) |
| Cardiovascular disease | 84 (80%) |
| Obesity (>30 kg/m2) | 36 (34%) |
| Preoperative hemoglobin, g/dl (n = 104) | 11.2 (9.7–12.4) |
| CVC use (n = 104) | 50 (48%) |
| Locoregional anesthesia | 98 (93%) |
| AVF on the left side of the body | 72 (69%) |
| Type of AVF | |
| Radiocephalic (distal) | 75 (71%) |
| Brachiocephalic (proximal) | 30 (29%) |
| Preoperative AVF vessel diameter, mm | |
| BC AVF Artery n = 30 | 5.00 (4.30–5.92) |
| RC AVF Artery n = 75 | 3.00 (2.55–3.30) |
| Whole cohort n = 105 | 3.30 (2.70–4.10) |
| Frequency of arteries <2 mm | 2% |
| BCAVF Vein n = 30 | 3.30 (3.00–4.45) |
| RCAVF Vein n = 75 | 3.30 (3.00–4.00) |
| Whole cohort n = 105 | 3.30 (3.00–4.10) |
| Frequency of veins <2 mm | 0% |
n = 105 except where indicated otherwise.
AVF, arteriovenous fistula; BC, brachiocephalic; BMI, body mass index; CVC, central venous catheter; IQR, interquartile range; RC, radiocephalic; SD, standard deviation.
3.2. First-Doppler-ultrasound vessel characteristics
The first postoperative DUS occurred 45 (36–56) days after surgery. Postoperative Qa was 800 (500–1000) ml/min overall. Postoperative RCAVF vein diameter and vein depth were 5.60 (5.15–6.35) and 3.00 (2.25–4.00) mm, respectively. Postoperative BCAVF vein diameter and vein depth were 6.20 (5.53–7.00) and 3.25 (2.42–3.88) mm, respectively. Of the 105 AVFs, 60% met ≥1 early-failure criteria. The most common were thrombosis/stenosis (n = 28) and poor Qa (n = 27), followed by poor vein diameter (n = 23) and poor vein depth (n = 9). Of the 63 early-failures, 43, 16, and four met one, two, and three early-failure criteria, respectively (Figure 1).
3.3. Preoperative factors that are associated with early failure
On univariate analysis, small artery diameter and vein distensibility were associated significantly with early failure (3.00 vs. 3.50 mm, p = 0.038 and 2.00 vs. 2.75 mm, p = 0.007 respectively). Distal location (78% vs. 62%, p = 0.078) tended to be associated with early failure (Table 2). On multivariate analysis, only vein distensibility was significantly associated with early AVF failure (OR = 0.57, 95%CIs = 0.38–0.83, p = 0.005): the lower the distensibility, the higher the risk of failure. Thus, venous distensibility may be a good predictor of early failure when minimum preoperative vessel-diameter limits are met (Table 3).
Table 2.
Univariate analysis of the demographic, clinical, DUS, and surgical factors that are associated with early AVF failure at first DUS.
| Variable | AVF failure (n = 63) | AVF success (n = 42) | p-value* |
|---|---|---|---|
| Age, years | 68 (56–74) | 69 (60–74) | 0.66 |
| Sex | 0.4 | ||
| Female | 25 (66%) | 13 (34%) | |
| Male | 38 (57%) | 29 (43%) | |
| BMI, kg/m2 | 27 (24–32) | 26 (23–33) | 0.56 |
| Comorbidities | |||
| Diabetes mellitus | 33 (52%) | 24 (57%) | 0.6 |
| Obesity (>30 kg/m2) | 23 (37%) | 13 (31%) | 0.6 |
| Preoperative hemoglobin, g/dl | 10.9 (9.6–11.8) | 11.7 (10–12.8) | 0.10 |
| Locoregional anesthesia | 57 (90%) | 43 (98%) | 0.2 |
| CVC use | 33 (52%) | 17 (41%) | 0.3 |
| AVF on the left side of the body | 41 (65%) | 31 (74%) | 0.3 |
| Distal-AVF location | 49 (78%) | 26 (62%) | 0.078 |
| Preop AVF vessel diameter, mm | |||
| Artery | 3.00 (2.70–3.70) | 3.50 (2.92–4.70) | 0.038 |
| Vein | 3.30 (2.80–4.00) | 3.30 (3.00–4.30) | 0.4 |
| Postop vein distensibility, mm | 2 (1.33–3) | 2.75 (2–3.6) | 0.007 |
| Time between surgery and DUS (days) | 47 (35–60) | 43 (37–54) | 0.50 |
| Surgeon | 0.39 | ||
| 1 | 19 (30) | 18 (43) | |
| 2 | 14 (22) | 11 (26) | |
| 3 | 13 (21) | 5 (12) | |
| 4 | 17 (27) | 8 (19) | |
The data are shown as n (%), mean ± SD, or median (IQR).
AVF, arteriovenous fistula; BMI, body mass index; CVC, central venous catheter; DUS, Doppler ultrasound; preop, preoperative.
p values were determined by Chi-square or Fisher’s exact test for categorical variables and the Wilcoxon test for continuous variables.
Table 3.
Multiple logistic regression analysis of factors that are associated with early AVF failure at first DUS.
| Variable | OR | 95% Cis | p-value |
|---|---|---|---|
| Age (years) | 1.01 | 0.97–1.05 | 0.54 |
| Female sex | 1.52 | 0.60–4.00 | 0.54 |
| Diabetes mellitus | 0.62 | 0.23–1.58 | 0.36 |
| Obesity | 1.16 | 0.44–3.10 | 0.81 |
| Preoperative hemoglobin (g/dL) | 0.77 | 0.58–1.01 | 0.11 |
| CVC use | 0.91 | 0.35–2.26 | 0.8 |
| Preop artery diameter (mm) | 0.69 | 0.46–1.01 | 0.057 |
| Vein distensibility (mm) | 0.57 | 0.38–0.83 | 0.005 |
| Time between surgery and DUS | 0.99 | 0.96–1.02 | 0.61 |
| Surgeon | 0.44 | ||
| 1 | Ref. | Ref. | |
| 2 | 0.68 | 0.19–2.47 | |
| 3 | 0.87 | 0.22–3.5 | |
| 4 | 2.30 | 0.47–11.3 | |
AVF location was not included in the analysis because it was collinear with AVF artery diameter. Preoperative vein diameter was not included because it was collinear with postoperative vein expansion.
AVF, arteriovenous fistula; BMI, body mass index; CI, confidence interval; CVC, central venous catheter; DUS, Doppler ultrasound; preop, preoperative.
3.4. Factors that predict failure-to-mature criteria and thrombosis/stenosis
Poor postoperative blood flow was predicted by small preoperative artery diameter (OR = 0.56, 95% CIs = 0.31–0.91, p = 0.032). Distal-AVF tended to be a predictor (OR = 2.94, 95% CIs = 0.92–11.1, p = 0.090).
Poor postoperative vein diameter was only predicted by smaller preoperative vein diameter (OR = 0.25, 95% CIs = 0.06–0.68, p = 0.021).
Poor postoperative vein depth was predicted by the female sex (OR = 10.00, 95% CIs = 1.45–100, p = 0.044) and lower hemoglobin (OR = 0.56, 95% CIs = 0.29–0.93, p = 0.041).
Thrombosis/stenosis was not predicted by any variable.
Thus, early failure and poor postoperative Qa were both associated with small preoperative artery diameter; poor postoperative vein diameter was only associated with small preoperative vein diameter; poor postoperative vein depth was associated with female sex and low hemoglobin; and no variables predicted thrombosis/stenosis.
Finally, we asked whether variables could predict vein-distensibility. High vein distensibility was strongly predicted by small preoperative vein diameter (Beta = –0.73, 95%CIs = –1.0 to –0.51, p < 0.001): this reflects the fact that smaller veins undergo more distensibility under a given blood flow than bigger veins. High vein-distensibility was also predicted by large preoperative artery-diameter (Beta = 0.24, 95% CIs = 0.06–0.42, p = 0.010) (Table 4): this was also expected since preoperative artery-diameter predicted postoperative Qa and the Qa from the artery strongly shapes vein-distensibility. Indeed, another multivariate analysis with vein-distensibility as the dependent variable and postoperative Qa plus instead of preoperative artery diameter (due to collinearity with Qa) showed that vein-distensibility was predicted by both postoperative Qa (Beta = 0.001, 95% CIs = 0.0001–0.0010, p = 0.012) and preoperative vein-diameter (Beta = –0.70, 95%CIs = –0.91 to –0.48, p < 0.001).
Table 4.
Multiple linear regression analysis of the patient/clinical factors that associate with AVF vein-distensibility.
| Variable | Beta | 95% CIs | p-value |
|---|---|---|---|
| Age | 0.00 | −0.01 to 0.02 | 0.6 |
| Female sex | −0.03 | −0.46 to 0.39 | 0.9 |
| Diabetes mellitus | −0.27 | −0.70 to 0.15 | 0.2 |
| Obesity | 0.00 | −0.03 to 0.04 | 0.9 |
| Preoperative hemoglobin | 0.05 | −0.07 to 0.18 | 0.4 |
| CVC use | −0.17 | −0.59 to 0.25 | 0.4 |
| Preop artery diameter | 0.24 | 0.06 to 0.42 | 0.010 |
| Preop vein diameter | −0.73 | −1.0 to −0.51 | <0.001 |
AVF, arteriovenous fistula; CI, confidence interval; CVC, central venous catheter; Preop, preoperative.
4. Discussion
Our study explored how potential risk factors predict early failure, four criteria that are commonly used to define early failure, and vein-distensibility, which is a poorly researched potential early failure criterion. Our early failure rate was 60%. This is among the highest in the literature (range, 20–60%) [12,13], but reflects our early-failure definition, namely, ≥1 of four criteria with relatively demanding cutoffs was met on first DUS.
4.1. Postoperative vein-distensibility predicts early-failure
The only strong predictor of early failure in our study was poor postoperative vein-distensibility. This association has also been observed by six studies: specifically, they found that preoperative, intraoperative, and/or postoperative vein-distensibility predicted early failure [6–11].
Thus, our and these studies suggest that venous compliance plays an important role in AVF failure. This is also likely to be true for the AVF artery: AVF-failure associates with poorer resistive indices in the AVF artery after fist clenching [7], and the small artery-elasticity index measured during surgery is lower in failure-to-mature AVFs [14]. The importance of vessel distensibility is also demonstrated by the vast changes that take place after AVF creation: in successful AVFs, vein-luminal size increases by 86% one week after surgery [15]; this allows the Qa in the radial artery to increase from 20 mL/min before surgery to 539 mL/min at one week, with additional increases being observed thereafter [16]. Failure to achieve this responsive venous dilatation is associated with significant neointimal hyperplasia and poor remodeling [17].
We also found that the only patient/clinical predictors of vein expansion itself were preoperative artery/vein diameters. These variables are often associated with early failure in the literature but did not strongly predict early failure in our study.
This is likely due to our routine preoperative use of recommended cutoffs [4, 5] and demonstrates the generalizability of these thresholds. Notably, since preoperative vessel diameters nonetheless strongly predicted vein distensibility in our cohort, their association with early failure in the literature may largely reflect their more direct relationship with vein distensibility. Indeed, the previous studies on vein-distensibility, which did not apply preoperative vessel-diameter cutoffs, showed that vein-distensibility predicted early-failure better than preoperative vein-diameter [6–11].
Thus, our study suggests that when preoperative vessel-diameter cutoffs are used, the primary predictor of early failure becomes vein distensibility. Since previous studies showed that vein-distensibility can predict early failure regardless of whether it is measured before, during, or after surgery [6–11], our study suggests that this variable may be a clinically useful determinant of early failure.
4.2. Small preoperative artery diameter weakly predicts early failure but strongly predicts poor blood flow
Of the preoperative patient/clinical variables, only a smaller preoperative artery-diameter predicted early failure, and only weakly. By contrast, small preoperative artery diameter strongly predicted poor postoperative Qa, but no other early-failure criterion. Thus, preoperative artery diameter also predicts early failure due to its primary relationship with postoperative Qa. This is consistent with studies showing that postoperative Qa correlates with both preoperative [7–18] and postoperative AVF artery diameter [19].
4.3. Female sex predicts poor postoperative vein-depth
Female sex did not predict early failure in our study. Unexpectedly, however, female sex did predict poor postoperative vein depth, but no other early-failure criterion or vein-distensibility. This direct relationship between the female sex and postoperative vein depth may explain why the female sex is only sometimes associated with early failure in the literature.
These findings are interesting because women are not only more prone to early failure than men, they also require more maturation-promoting interventions, achieve maturation more slowly, and demonstrate shorter post-maturation AVF survival [20–22]. This has long puzzled the field. It was thought to reflect smaller vessel diameters in women [23, 24] but this is likely only a partial explanation. First, some studies find similar vessel diameters between women and men [6, 25, 26]. S, mature and failed AVFs in women do not differ in preoperative artery/vein diameters [27]. Third, proximal-AVFs in women fail more often than distal-AVFs in men despite employing larger vessels [28]. Fourth, pediatric patients have low early failure rates (33%) despite having small vessels [29]. Fifth, while preoperative-vessel cutoffs decrease early failure in both sexes, the difference remains [30]. Sixth, multivariate analyses accounting for preoperative vessel diameters continue to find the sex association [27].
It is now increasingly thought that the greater early failure in women may reflect fundamental sex differences in vascular remodeling, which generates the vessel dilatation and wall thickening needed for AVF function. This may reflect sex differences in inflammation, which plays key roles in vascular remodeling [31].
Other explanations for our findings could include obesity, which can be more common in women [32] and can reduce AVF cannulability [33, 34]. However, obesity was accounted for in our multivariate analyses and the female/male patients did not differ in BMI (26.0 vs. 27.1 kg/mm2, p = 0.3) or obesity (29% vs. 37%, p = 0.4). Another possibility is that our female patients had deeper preoperative veins than our male patients. We cannot test this but it seems unlikely since preoperative vein palpability was an essential determinant of vein choice.
4.4. Low preoperative hemoglobin predicts poor postoperative vein-depth
Another predictor of poor postoperative vein depth was low hemoglobin. In the literature, Low hemoglobin also predicts slower AVF maturation and poor post-maturation AVF survival [35, 37]. It may mark excessive inflammation [36], which both induces erythropoietin resistance and is associated with AVF failure [38, 39]. However, further studies seem necessary to clarify the actual role of hemoglobin levels in AVF maturation, and the possible mechanisms involved.
4.5. Relationships between other variables and early failure or its criteria
Age, obesity, diabetes, and CVC use did not predict early failure. This may reflect the possibility that these variables are weak surrogates for more primary relationships between early failure and vascular dysfunction/poor remodeling [40].
As discussed above, small preoperative artery diameter and vein diameter were the only significant predictors of poor postoperative blood flow and vein diameter, respectively; and female sex and low hemoglobin predicted poor postoperative vein-depth. Moreover, no variables predicted thrombosis/stenosis. These findings are of interest because, to our knowledge, the associations between patient/clinical variables and the failure-to-mature criteria have not been studied previously. Regarding thrombosis/stenosis, the studies on the role of patient/clinical factors are also limited. There is a meta-analysis of 27 studies that found that older age, female sex, and distal-AVF, but not diabetes (3–6 studies/factor) predicted thrombosis. However, some of these studies did not focus on thrombosis early after surgery and some included arteriovenous grafts [41]. Thus, our research adds to this small field.
4.6. Study limitations
First, our study was a single-center retrospective study with a relatively small sample. Second, we used a theoretical definition of early-failure; while its criteria are recommended by KDOQI and supported by ample evidence [3, 10, 19], similar analyses with other early-failure definitions could lead to different findings. Third, both distal AVFs and proximal AVFs were combined: patient/clinical risk factors may vary with location (e.g. female sex) [23]. Moreover, we accounted for location in the multivariate analyses. Finally, we did not investigate the impact of other variables that could shape early failure, including hemodynamic variables such as preoperative blood flow, clinical variables such as uremia and albuminemia, treatment variables (e.g. heparin use), and other comorbidities [35].
4.7. Conclusions
We showed that low postoperative vein distensibility strongly predicted early failure when minimum preoperative vessel diameters were used. Since the few studies in this field show that preoperative venous distensibility also predicts early failure [6, 8, 9, 11], our data supports the notion that this variable could be useful for predicting AVF success. Thus, determining minimum preoperative and/or postoperative vein-distensibility values could aid clinical decision-making in the future. Our study also suggested that the puzzling association between female sex and early failure may reflect a sex hormone-related inability to remodel layers overlying the AVF. Moreover, our study suggested that demographic/clinical variables are generally weak surrogates for more primary relationships between early failure and vascular remodeling/vein dysfunction. Indeed, when postoperative blood flow, vein diameter, and vein depth at six-week DUS were recently added as covariates to multivariate analyses, older age, female sex, diabetes, and BMI no longer predicted early failure [19]. These unpredictable, often weak, relationships may explain why AVF-failure prediction models that use demographic/clinical variables have limited generalizability [20, 26]. Prediction models based on direct vascular-function measures may be more effective [42].
Acknowledgments
We are very grateful to Ms. Nadia OUAMARA for her significant contributions to the implementation of this study.
Author contributions
Concepts, Literature search, Data acquisition, Manuscript preparation: ZT; Design: ZT and VD; Definition of intellectual content: ZT and CG; Data analysis and Statistical analysis: VD; Manuscript editing: ZT, CG, and BS; Manuscript review: ZT, VD, CG, BS.
Disclosure statement
No potential conflict of interest was reported by the authors.
Data availability statement
The data are available on reasonable request from the corresponding author.
References
- 1.Lee T. Fistula first initiative: historical impact on vascular access practice patterns and influence on future vascular access care. Cardiovasc Eng Technol. 2017;8(3):244–254. doi: 10.1007/s13239-017-0319-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kucey AS, Engineer A, Albers SS.. Outcomes of patients referred for arteriovenous fistula construction: a systematic review. Int J Med Students. 2019;7(3):73–81. doi: 10.5195/ijms.2019.331. [DOI] [Google Scholar]
- 3.KDOQI clinical practice guidelines and clinical practice recommendations for 2006 updates: hemodialysis adequacy, peritoneal dialysis adequacy and vascular access. Am J Kid Dis. 2006;48:s1–s322.17045862 [Google Scholar]
- 4.Glass C, Johansson M, DiGragio W, et al. A meta-analysis of preoperative duplex ultrasound vessel diameters for successful radiocephalic fistula placement. J Vasc Ultrasound. 2009;33(2):65–68. doi: 10.1177/154431670903300201. [DOI] [Google Scholar]
- 5.Kordzadeh A, Chung J, Panayiotopoulos YP.. Cephalic vein and radial artery diameter in formation of radiocephalic arteriovenous fistula: a systematic review. J Vasc Access. 2015;16(6):506–511. doi: 10.5301/jva.5000413. [DOI] [PubMed] [Google Scholar]
- 6.Jemcov TK. Morphologic and functional vessels characteristics assessed by ultrasonography for prediction of radiocephalic fistula maturation. J Vasc Access. 2013;14(4):356–363. doi: 10.5301/jva.5000163. [DOI] [PubMed] [Google Scholar]
- 7.Malovrh M. Native arteriovenous fistula: preoperative evaluation. Am J Kidney Dis. 2002;39(6):1218–1225. doi: 10.1053/ajkd.2002.33394. [DOI] [PubMed] [Google Scholar]
- 8.van der Linden J, Lameris TW, van den Meiracker AH, et al. Forearm venous distensibility predicts successful arteriovenous fistula. Am J Kidney Dis. 2006;47(6):1013–1019. doi: 10.1053/j.ajkd.2006.01.033. [DOI] [PubMed] [Google Scholar]
- 9.Kim JT, Chang WH, Oh TY, et al. Venous distensibility as a key factor in the success of arteriovenous fistulas at the wrist. Ann Vasc Surg. 2011;25(8):1094–1098. doi: 10.1016/j.avsg.2011.05.014. [DOI] [PubMed] [Google Scholar]
- 10.Kim MH, Kim YK, Jun KW, et al. Clinical importance of intraoperative cephalic vein distensibility as a predictor of radiocephalic arteriovenous fistula maturation. Semin Dial. 2015;28(6):E64–E70. doi: 10.1111/sdi.12452. [DOI] [PubMed] [Google Scholar]
- 11.Zhang F, Li Y, Wang T, et al. Influence of cephalic vein dilation on arteriovenous fistula maturation in patients with small cephalic veins. Medicine (United States). 2020;99(1):e18576. doi: 10.1097/MD.0000000000018576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Leermakers JJ, Bode AS, Vaidya A, et al. Cost-effectiveness of vascular access for haemodialysis: arteriovenous fistulas versus arteriovenous grafts. Eur J Vasc Endovasc Surg. 2013;45(1):84–92. doi: 10.1016/j.ejvs.2012.10.012. [DOI] [PubMed] [Google Scholar]
- 13.Rezapour M, Khavanin Zadeh M, Sepehri MM.. Implementation of predictive data mining techniques for identifying risk factors of early AVF failure in hemodialysis patients. Comput Math Methods Med. 2013;2013:830745–830748. doi: 10.1155/2013/830745. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kheda MF, Brenner LE, Patel MJ, et al. Influence of arterial elasticity and vessel dilatation on arteriovenous fistula maturation: a prospective cohort study. Nephrol Dial Transplant. 2010;25(2):525–531. doi: 10.1093/ndt/gfp462. [DOI] [PubMed] [Google Scholar]
- 15.Corpataux JM, Haesler E, Silacci P, et al. Low-pressure environment and remodelling of the forearm vein in Brescia-Cimino haemodialysis access. Nephrol Dial Transplant. 2002;17(6):1057–1062. doi: 10.1093/ndt/17.6.1057. [DOI] [PubMed] [Google Scholar]
- 16.Konner K, Nonnast-Daniel B, Ritz E.. The arteriovenous fistula. J Am Soc Nephrol. 740. 2003;14(6):1669–1680. doi: 10.1097/01.asn.0000069219.88168.39. [DOI] [PubMed] [Google Scholar]
- 17.Roy-Chaudhury P, Spergel LM, Besarab A, et al. Biology of arteriovenous fistula failure. J Nephrol. 2007;20(2):150–163. [PubMed] [Google Scholar]
- 18.Usta E, Elkrinawi R, Salehi-Gilani S, et al. Risk factors predicting the successful function and use of autogenous arteriovenous fistulae for hemodialysis. Thorac Cardiovasc Surg. 2013;61(5):438–444. doi: 10.1055/s-0032-1321953. [DOI] [PubMed] [Google Scholar]
- 19.Robbin ML, Greene T, Allon M, et al. Prediction of arteriovenous fistula clinical maturation from postoperative ultrasound measurements: findings from the Hemodialysis Fistula Maturation Study. J Am Soc Nephrol. 2018;29(11):2735–2744. doi: 10.1681/ASN.2017111225. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Farrington CA, Robbin ML, Lee T, et al. Early predictors of arteriovenous fistula maturation: a novel perspective on an enduring problem. J Am Soc Nephrol. 2020;31(7):1617–1627. doi: 10.1681/ASN.2019080848. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lee T, Qian J, Zhang Y, et al. Long-term outcomes of arteriovenous fistulas with unassisted versus assisted maturation: a retrospective national hemodialysis cohort study. J Am Soc Nephrol. 2019;30(11):2209–2218. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6830790/?report=classic. doi: 10.1681/ASN.2019030318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Dunn J, Herscu G, Woo K.. Factors influencing maturation time of native arteriovenous fistulas. Ann Vasc Surg. 2015;29(4):704–707. doi: 10.1016/j.avsg.2014.11.026. [DOI] [PubMed] [Google Scholar]
- 23.Feldman HI, Joffe M, Rosas SE, et al. Predictors of successful arteriovenous fistula maturation. Am J Kidney Dis. 2003;42(5):1000–1012. doi: 10.1016/j.ajkd.2003.07.003. [DOI] [PubMed] [Google Scholar]
- 24.Ernandez T, Saudan P, Berney T, et al. Risk factors for early failure of native arteriovenous fistulas. Nephron Clin Pract. 2005;101(1):c39–c44. doi: 10.1159/000085710. [DOI] [PubMed] [Google Scholar]
- 25.Caplin N, Sedlacek M, Teodorescu V, et al. Venous access: women are equal. Am J Kidney Dis. 2003;41(2):429–432. doi: 10.1053/ajkd.2003.50052. [DOI] [PubMed] [Google Scholar]
- 26.Lo ZJ, Tay WM, Lee Q, et al. Predictors of radio-cephalic arteriovenous fistulae patency in an asian population. J Vasc Access. 2016;17(5):411–416. doi: 10.5301/jva.5000591. [DOI] [PubMed] [Google Scholar]
- 27.Wilmink T, Corte-Real Houlihan M.. Diameter criteria have limited value for prediction of functional dialysis use of arteriovenous fistulas. Eur J Vasc Endovasc Surg. 2018;56(4):572–581. doi: 10.1016/j.ejvs.2018.06.066. [DOI] [PubMed] [Google Scholar]
- 28.Peterson WJ, Barker J, Allon M.. Disparities in fistula maturation persist despite preoperative vascular mapping. Clin J Am Soc Nephrol. 2008;3(2):437–441. doi: 10.2215/CJN.03480807. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sheth R, Brandt M, Brewer E, et al. Permanent hemodialysis vascular access survival in children and adolescents with end-stage renal disease. Kidney Int. 2002;62(5):1864–1869. doi: 10.1046/j.1523-1755.2002.00630.x. [DOI] [PubMed] [Google Scholar]
- 30.Allon M, Lockhart M, Lilly R, et al. Effect of preoperative sonographic mapping on vascular access outcomes in hemodialysis patients. Kidney Int. 2001;60(5):2013–2020. doi: 10.1046/j.1523-1755.2001.00031.x. [DOI] [PubMed] [Google Scholar]
- 31.Chan SM, Weininger G, Langford J, et al. Sex differences in inflammation during venous remodeling of arteriovenous fistulae. Front Cardiovasc Med. 2021;8:715114. doi: 10.3389/fcvm.2021.715114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Muhammad T, Boro B, Kumar M, et al. Gender differences in the association of obesity-related measures with multi-morbidity among older adults in India: evidence from LASI, Wave-1. BMC Geriatr. 2022;22(1):171. doi: 10.1186/s12877-022-02869-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Siddiqui MA, Ashraff S, Carline T.. Fogarty maturation of arteriovenous fistula: analysis of key factors. Kidney Res Clin Pract. 2017;36(4):318–328. doi: 10.23876/j.krcp.2017.36.4.318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kats M, Hawxby AM, Barker J, et al. Impact of obesity on arteriovenous fistula outcomes in dialysis patients. Kidney Int. 2007;71(1):39–43. doi: 10.1038/sj.ki.5001904. [DOI] [PubMed] [Google Scholar]
- 35.Zadeh M, Negahi A.. Factors affecting the maturation of arterio-venous fistula in patients with end-stage renal disease. Saudi J Kidney Dis Transpl. 2014;25(1):161–165. doi: 10.4103/1319-2442.124554. [DOI] [PubMed] [Google Scholar]
- 36.Garrancho M, Kirchgessner J, Arranz M, et al. Haemoglobin level and vascular access survival in haemodialysis patients. Nephrol Dial Transplant. 2005;20(11):2453–2457. doi: 10.1093/ndt/gfi027. [DOI] [PubMed] [Google Scholar]
- 37.Wen M, Li Z, Li J, et al. Risk factors for primary arteriovenous fistula dysfunction in hemodialysis patients: a retrospective survival analysis in multiple medical centers. Blood Purif. 2019;48(3):276–282. doi: 10.1159/000500045. [DOI] [PubMed] [Google Scholar]
- 38.Kordzadeh A, Tokidis E, Askari A, et al. The independent association of preoperative serum albumin on the functional maturation of radiocephalic arteriovenous fistulae. J Vasc Access. 2017;18(2):148–152. doi: 10.5301/jva.5000646. [DOI] [PubMed] [Google Scholar]
- 39.Martinez-Mier G, Camargo-Diaz C, Urbina-Velazquez MA, et al. Predictive factors for unsuccessful use of arteriovenous fistula in a population of end-stage renal disease patients in Southeastern Mexico. Ann Vasc Surg. 2020;62:304–309. doi: 10.1016/j.avsg.2019.06.034. [DOI] [PubMed] [Google Scholar]
- 40.Gameiro J, Ibeas J.. Factors affecting arteriovenous fistula dysfunction: a narrative review. J Vasc Access. 2020;21(2):134–147. doi: 10.1177/1129729819845562. [DOI] [PubMed] [Google Scholar]
- 41.Zhang Y, Yi J, Zhang R, et al. Risk factors for arteriovenous fistula thrombus development: a systematic review and meta-analysis. Kidney Blood Press Res. 2022;47(11):643–653. doi: 10.1159/000526768. [DOI] [PubMed] [Google Scholar]
- 42.Ng O, Gunasekera S, Varcoe R, et al. A review of the predictive methods for arteriovenous fistula (AVF) failure identification. Comput Methods Biomech Biomed Eng Imaging Vis. 2022:1–11. doi: 10.1080/21681163.2022.2074545. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data are available on reasonable request from the corresponding author.