How to create and preserve distal fistulas in a large number of patients: the experience of a single centre

Introduction

International guidelines such as K-DOQI support the use of distal native fistula (AVF) as the best first-line vascular access (VA) [1,2] that guarantees the least amount of complications (stenosis, overflow syndrome, thrombosis and distal ischemia) [1–4]. Despite these indications, all over the world except in Japan the number of AVFs in the upper arm, and related complications, is constantly increasing [5]. As suggested by Bourquelot, the prevalence of distal AVFs may be increased through additional procedures capable of incentivizing the maturation process or establishing a more reliable inflow or outflow [3]: early referral to the surgeon, precise preoperative Ultrasound Vascular Mapping (UVM) of vein and arteries, post-operative clinical and Duplex evaluation of maturation, Percutaneous Transluminal Angioplasty (PTA) of artery and vein stenosis, or surgical procedure in case of frequent relapses or failure of the PTA. To increase the incidence of distal AVFs our team has specialized in vascular access surgery from 2004 onwards, and has progressively introduced Intraoperative Transluminal Angioplasty (ITA) [7–10] and Vessel Pre-Dilatation (VPD) in clinical practice [8].

The purpose of this study is to highlight the possibility to create and preserve distal fistulas in a large number of patients. We report our experience with ITA and VPD and claim that the two methods may be able to increase the incidence of distal AVFs.

Procedures to increase the incidence and prevalence of distal AVF

VPD [8]. It is a simple procedure consisting in cannulating the vessel with a 26 Gauge needle after having clamped it downstream and upstream. A small quantity (0.1-0.2 ml) of saline solution is then injected at an adequate pressure to expand significantly the arterial and especially the venous vessels, thus allowing the surgeon to also exploit very small vessels (Fig. 1).

ITA [7–10]. This technique has been introduced in our interventional practice in 2004. It uses the incision already created for the anastomosis to introduce the angioplasty devices [7]. As far as veins are concerned, the purpose of the ITA is to make available vessels that would otherwise be discarded because of resistance to intraoperative irrigation (due to stenosis, fibrosis or their small calibre) [8–10]. The small size of the vein (less than 2 mm) is not an absolute indication for the ITA, but the resistance to intraoperative vessel irrigation is considered fundamental. As far as arteries are concerned, the purpose of the ITA is to make available vessels that show a low blood flow after the incision (due to atherosclerotic lesions or their small calibre) [8]. Generally, the arteriosclerotic lesions and the thickening of the venous walls showed by UVM represent an indication to plan an ITA, but the final decision is based on intraoperative observation of low blood flow for the arteries and resistance to irrigation for the veins. This could generate a bias due to “confounding by indication”. The operator, in fact, must quickly decide whether to perform the ITA. Since 2004, 240 procedures have been carried out in our Department. We use balloons with a diameter 1 mm larger than the vessel to be treated (from 2.5 to 4 mm). The whole vessel is treated, from the elbow to the incision site, with one or two insufflations exerting a pressure up to 6-7 atm for the veins and 6-10 atm for the arteries. In case of re-operations on already arterialized veins, high pressure and larger balloons are used. To pursue more complex ITAs, angiography has been used for two years (Fig. 2). During the ITA, when the guide wire encounters resistance in the vascular lumen and the stenosis is not negotiated, switching from ultrasound to fluoroscopic guide can allow for the procedure to be concluded successfully.

Materials and methods

Study design

A retrospective case note review was performed on patients identified from our Department vascular access database of as having undergone the creation of an autogenous AVF over a five-year period (from January 2014 to October 2018). Data from all consecutive patients directly admitted at our Hospital to create or repair an AVF or to create a prosthetic fistula (AVG) was collected. All fistulas were performed by nephrologists trained in vascular access surgery. The same nephrologists, also trained in ultrasound examination, performed the UVM and the ultrasound diagnosis of the AVF failure. Fistula thrombosis was treated with surgical procedure by the same nephrologists. Patients with severe radial atherosclerosis, absence of colour signal in the UVM and no distal pulse were not included as candidates for an AVF at the forearm. Small diameter vessels (<2 mm) were not considered unacceptable, but were potential candidate for ITA. Generally, veins with size less than 1.5 mm after tourniquet were excluded. During the procedures, local anaesthesia was used, whilst regional block or general anaesthesia was rarely performed. The administration of low molecular weight heparin and antiplatelet drugs to prevent thrombosis was not systematic but individualized [11]. In presence of risk factors such as reduced thrill, arteriosclerotic lesions, hypotension and in the case of intraoperative endovascular manoeuvres, an immediate dose of Enoxaparin (2000-4000 UI) was administered and Clopidogrel 75 mg/day was started at the same time. As recommended by the Vascular Access Study Group of Italian Society of Nephrology [12], we considered separately the AVFs set up in the four parts of the superior limb: arm (AM), proximal forearm (PF), middle forearm (MF) and distal forearm (DF).

Data on the patients’ age, gender, history of cardiovascular disease, hypotension and diabetes was collected retrospectively. Primary and secondary patency, early failure (EF), failure to mature (FTM) and late failure (LF) were evaluated. Primary patency was defined as the time lapse between AVF creation, its first thrombosis and any surgical intervention in case of stenosis or FTM [13,14]. Secondary patency was considered as the time lapse between AVF creation and its abandonment with the creation of a fistula in a new vascular district or of a new vascular access (CVC or AVG), regardless of any rescue intervention [13,14]. EF was defined as the thrombosis of the AVF within 3 months of its creation. A fistula was defined mature when it was used for six consecutive dialysis sessions with two needles and with adequate blood flow [15]. FTM was the inadequate fistula maturation within 3 months of its creation. LF was defined as the thrombosis of an AVF after more than 3 months of its creation. The incident surgical operations performed over that period were compared with a previous period (2000-2004) in which ITA and VPD were not routinely carried out. The operators, the surgical techniques and the materials were the same over the two periods. Although the experience of the operators was likely greater in the second period, they were already expert during the first. The prevalence of vascular accesses in our dialytic population at September 2019 was analysed to define whether the choice of AVF type exerted an impact also on prevalence itself.

The study protocol was approved by the institutional ethics review board of the Azienda Sanitaria Locale (ASL) in Lecce. This study was conducted in adherence with the Declaration of Helsinki.

Statistical methods

We obtained descriptive statistics, including means and standard deviations (SD) for continuous measures; frequencies and proportions for categorical ones. In order to compare prevalence and incidence, we used the chi-square test for independence. Primary and secondary patency was assessed through the Kaplan-Meier curve (SPSS, Inc, Chicago, Il) by separately analysing the ITA group vs the non-ITA group.

 

Results

Patients characteristics

During the period considered in this study 660 consecutive vascular accesses were performed in 497 patients: 647 native AVFs e 13 prosthetic AVFs. The AVGs have been excluded from the present analysis because they were too few. In Table I the demographic data and comorbidities of the patients are reported. Mean age was 69±14 years; 325 patients were over 65 years old. The patients were 64.2% male, 35.8% female. Black patients were 15 (3%). Ninety-four (19%) patients were affected by cardiac disease, 132 (26.7%) by vascular disease and 35 (7%) showed hypotension. Diabetics were 108 (21.5%) and obese patients were 32(6.4%).
 

	n.	%
Total number of pts	497
Age (years±SD)	69±14
Over 65-year-old	325	65.4
Male	319	64.2
Female	178	35.8
White	482	97.0
Black	15	3.0
Diabetics	108	21.7
Heart Disease	94	18.9
Peripheral Vasculopathy	132	26.6
Hypotension	35	7.0
Obese	32	6.4
Post-transplant	18	3.6

Table I: Demographics and comorbidities of the patients

Types and sites of VA

Concerning the placement of fistulas, see Table II. Of 647 AVFs, 11 (1.7%) were located on the AM and, 98.3% on the forearm. Of these, 472 on the DF, 150 on the MF and 14 on the PF. The anastomosis in the 472 AVFs of DF was side (artery)-to-end (vein) in 38 cases, side-to-side with the ligation of distal vein in 434, end-to-end in 11. Among the 647 AVFs performed, 298 represented the first operation for the patient, 218 the second one, 85 the third one and 46 the fourth one or more (Tab. III). Out of 298 AVFs set up as first access, 30% involved late referral patients that had started dialysis with a CVC. There were no differences in either the timing of early failure or the site of the AVF when compared to early referral patients. The late referral ones started dialysis with a CVC but received the fistula earlier.

 

	Distal Forearm	Mean Forearm	Proximal Forearm	Arm
Radio-Cephalic	468	139	8
Radio-Median antebrachial	4	10	3
Radio-Basilic	1 (transposition)	1 (transposition)	1
Radio-Perforating			2
Brachio-Cephalic				4
Brachio-Basilic				2 (transposition)
Brachio-Perforating				1
Brachio-Median antecubital				4
Total	472	150	14	11

Table II: Site of the AVFs (n=647) and vessels involved

	DF	MF	PF	Arm	Total
First Intervention	278	17	1	2	298
Second Intervention	155	58	4	1	218
Third Intervention	28	47	5	5	85
Fourth intervention or beyond	11	28	4	3	46
Total	472	150	14	11	647

Table III: Site of the AVF and number of interventions received by the patient

The interventions following the first are reported in Table IV. Primary and secondary patency rate at 1, 2, 3 and 4 years was respectively 80%, 74%, 68%, 64% and 94%, 91%, 89%, 88% (Fig. 3). EF and LF are reported in Table V. FTM occurred in 57 cases, 31 of which were treated with PTA, whilst 26 were resolved performing a new anastomosis upstream from the previous one (Tab. V). The procedures/patient/year required to guarantee the patency of AVFs was 0.19. The analysis of the prevalence of VA at December 2013, December 2017 and September 2019 of patients referring to our center for management of VA did not show any difference in time (Figs. 4-5). From the comparison of the incident interventions in the years 2000-2004 and the period 2014-2018, a significant decrease of the AVF located in the upper arm was evident (from 18% to 1.7%) (p <0.0001) as well as a significant increase in the AVFs located on the forearm (79% to 96.3%) (p <0.0001) [6].

 

	n.	DF	MF	PF	Arm
Surgical- rescued AFV for stenosis	119	80	38	1
Surgical- rescued AVF for thrombosis	94	49	44	1
New AVF	136	65	51	11	9

Table IV: AVFs performed after the failure of a previous fistula creating a new upstream the previous one or a new AVF in another vascular district

	n.	Abandoned	Surgical Rescue	PTA
Early Failure	83	16	67
Late Failure	100	32	68
Failure To Mature	57		26	31
Late Stenosis	75		29	46

Table V: Early and late complications

Procedures to increase distal AVFs

VPD: it was carried out systematically on all patients. Out of 298 first interventions, 63 were performed exploiting veins less than 2 mm in diameter. In these patients we recorded a higher rate of FTM (22% vs 10%, p <0.01). In all cases, FTM was resolved creating a new fistula upstream from the previous or performing PTA. Regarding arteria, VPD was also routinely performed, although arteries showed a more relevant resistance to VPD. No side effects due to VPD were found.

ITA: it was performed during AVF creation when required. In the last 5 years, 128 (20% of 647 AVFs) ITAs have been performed, in 84 cases on venous vessels and in 30 cases on arteries. In 14 cases, angioplasty was performed on both the artery and the vein. No complications occurred due to ITAs. The pain caused by the procedure regressed at the end of the insufflation and never required drug treatment. In 9 cases, ITA was carried out under fluoroscopic guidance in the angiographic phase (Fig. 2); in the remaining cases, we carried out the US control. Technical failures occurred in 6 cases of US-guided ITA (4 venous and 2 arterial), leading to the creation of an AVF in another site. In the ITA group, the incidence of diabetes (39% vs 14%, p <0.001), heart disease (22% vs 13%, p <0.05) and peripheral artery disease (43% vs 15%, p <0.001) was higher than in the non-ITA group. By dividing patients into an ITA group and a control group, we did not find any difference in primary and secondary patency (Fig. 3). The procedures/patient/year required to guarantee the patency of AVFs was higher for patients who underwent ITA (0.21 vs 0.16, p <0.05).

 

Discussion

Our study shows that distal AVF can be set up in a large number of patients. Of the 647 AVFs, 636 (98.3%) were located on the forearm (472 at the DF) and 11 on the upper arm. Of the 11 AVFs located on the arm, only in 2 cases the intervention was the first one for that patient. In case of re-intervention for failure of previous distal AVF (juxta-anastomotic stenosis or thrombosis), the new accesses were located at the DF, and then at the MF; a small number was located at the PF or AM. In our policy the AVFs on the upper arm is considered the last choice, to be used only after the failure all the attempts required to set up a distal AVF. There were no differences in either the EF or the site of the AVF when late referral patients were compared to early referrals. The late referral patients started dialysis with CVC, but received fistulas early, avoiding prolonged use of CVC.

It is challenging to establish how much the procedures introduced in the last 10 years have influenced our case series. In the event of maturation delay, the ultrasound evaluation is central for the diagnosis of stenosis and to direct the patient to PTA or to surgical revision with a clinical selection [16]. PTA or surgical approach for frequent relapses or for failure of angioplasty are procedures frequently performed. In our case series, out of 132 cases of stenosis, 77 were treated with PTA and 55 with surgical revision. Therefore, we consider the follow-up of the AVF a key phase to recognize the stenosis and treat it on time. Regarding the interventions subsequent to the first, in 94 cases it was possible to retrieve a thrombosed AVF through a surgical revision and the creation of a new anastomosis upstream from the previous one. We prefer the surgical approach, instead of the interventional radiology, for the treatment of thrombosed AVFs because of the lack of experience of radiologists in this field. In this respect, the early referral of the thrombosed AVF to the surgeon may allow for the recovery of the anastomosis.

As mentioned, venous VPD is systematically performed in all patients. Rarely, in case of new interventions, veins already have a large diameter and VPD is not performed. This technique allows to increase the size of the vessel, particularly the vein, considerably facilitating the suture also for small veins (diameter <2 mm). We believe that the dilatation of the immediately post-anastomotic part of the vein and of the pre-anastomotic part of the artery may facilitate the outflow and the inflow, thus reducing the incidence of intra or post-operative thromboses. This technique could induce some vessel wall injury by over-distention and could increase the risk of EF but, in our opinion, the advantage of creating a suture in a larger vessel can make this risk acceptable.

The ITA was performed in 128 cases in this series, mainly on venous vessels. Initially, the angioplasties were carried out under US guidance, whilst recently, in complex cases, fluoroscopy with intraoperative angiography is also used (Fig. 2). The ITA group reached a comparable patency in respect to the non-ITA group, which had less risk factors. This is relevant, especially considering that the unavailability of such a procedure would have meant the selection of a more proximal site of AVF. The patients in Fig. 2 received a distal fistula only thanks to the availability of angiography and subsequent angioplasty; without it, a proximal fistula would have been created in the patient of Fig. A2. In the patient of Fig. B2 no fistula with native vessels would have been possible without the recanalization of the cephalic vein of the arm, the only outflow of the forearm superficial venous system. Nevertheless, the procedures/patient/year required to guarantee the patency of AVFs were higher for patients who underwent ITA (0.21 vs 0.16, p <0.05). The use of suboptimal vessels could increase the risk of EF, however the 81% of EF was rescued by creating a new anastomosis in the forearm, upstream from the previous one. The comparison of the incident interventions in the years 2000-2104 and 2014-2018 showed a huge reduction in the AVF performed at the arm, from 18 to 1.7% (Fig. 6). Furthermore, ITAs and VPD were not carried out in the period 2000-2004. It is likely that these procedures have been crucial in bringing down the number of AVFs on the arm and increasing the AVFs on the forearm. Our prevalence data (Fig. 4) suggest that our clinical approach to fistula creation would be able to maintain a high prevalence of distal AVFs over time.

 

Conclusions

VPD and ITA could be useful to increase the incidence and the prevalence of distal AVF. Our data on incidence and prevalence of VA show that it is possible to create and maintain distal AVFs in a large part of the dialysis population, although further studies are certainly required to prove it conclusively.

 

References

1. Lok CE, Huber TS, Lee T, Shenoy S, et al.; National Kidney Foundation. KDOQI Clinical Practice Guideline for Vascular Access: 2019 Update. Am J Kidney Dis 2020 Apr; 75(4 Suppl 2):S1-S164. https://doi.org/10.1053/j.ajkd.2019.12.001
2. Gallieni M, Hollenbeck M, Inston N, Kumwenda M, et al. Clinical practice guideline on peri- and postoperative care of arteriovenous fistulas and grafts for haemodialysis in adults. Nephrol Dial Transplant 2019; 1(34)(2):ii1-ii42.
3. Bourquelot P. Why we should insist on making distal autologous AV fistulas. [VAS abstract] J Vasc Access 2011; 12(2):137-81.
4. Scheltinga MR, van Hoek F, Bruijninckx CMA. Time of onset in haemodialysis access-induced distal ischaemia (HAIDI) is related to the access type. Nephrol Dial Transplant 2009; 24(10):3198-204.
5. Pisoni RL, Zepel L, Fluck R, et al. International Differences in the Location and Use of Arteriovenous Accesses Created for Hemodialysis: Results from the Dialysis Outcomes and Practice Patterns Study (DOPPS). Am J Kidney Dis 2018; 71(4):469-78.
6. Lefons ML, Accoto S, Caccetta F, et al. A new register of the vascular accesses for hemodialyisis. G Ital Nefrol 2019; 36(6):9. https://giornaleitalianodinefrologia.it/2019/12/36-06-2019-9/
7. Napoli M, Montinaro A, Russo F, et al. Ultrasound guided brachial arterial angioplasty during the creation of a radio-cephalic arteriovenous fistula: a case report. J Vasc Access 2006; 7(1):38-42.
8. Napoli M, Lefons ML, Mangione D, et al. Primary intraoperative transluminal angioplasty: a new approach to reduce the early failure of distal arteriovenous fistulas. J Vasc Access 2015; 16(3):250-54.
9. Veroux P, Giaquinta A, Tallarita T, et al. Primary balloon angioplasty of small (≤2 mm) cephalic veins improves primary patency of arteriovenous fistulae and decreases reintervention rates. J Vasc Surg 2013; 57(1):131-36.
10. De Marco Garcia LP, Davila-Santini LR, Feng Q, et al. Primary balloon angioplasty plus balloon angioplasty maturation to upgrade small-caliber veins (<3 mm) for arteriovenous fistulas. J Vasc Surg 2010; 52(1):139-44.
11. Schmidli J, Widmer MK, Basile C, et al. Vascular access: 2018 clinical practice guidelines of the European Society for Vascular Surgery. Eur J Vasc Endovasc Surg 2018; 55:757-818.
12. Napoli M, Tazza L, Postorino M, et al. The Italian Registry of Vascular Access. G Ital Nefrol 2013; 30(6). https://giornaleitalianodinefrologia.it/wp-content/uploads/sites/3/pdf/GIN_A30V6_00101_15.pdf
13. Mokrzycki M, Moist L, et al. Standardized definitions for hemodialysis vascular access. Semin Dial 2011; 24:515-24.
14. Sidawy AN, Gray R, Besarab A, et al. Recommended standards for reports dealing with arteriovenous hemodialysis accesses. J Vasc Surg 2002; 35:603-10.
15. Wilmink T, Hollingworth L, Stevenson T, Powers S. Is early cannulation of an arteriovenous fistula associated with early failure of the fistula? J Vasc Access 2017; 18(S1):S92-S97. https://doi.org/10.5301/jva.5000674
16. Napoli M, Prudenzano R, Russo F, et al. Juxta-anastomotic stenosis of native arteriovenous fistulas: surgical treatment versus percutaneous transluminal. J Vasc Access 2010; 11(4):346-51.

table td:first-child {font-weight: bold;}h4 { margin-bottom: 4px !important;}