The prevalence of aPL and APS in our hemodialysis population with native AVF was 18.4%, which is similar to the prevalence reported by others (11–37%) (10). The prevalence of AVFMF was of 45.8%, which is also consistent with the literature. In fact, up to 60% of AVF will fail maturation after they are created (14).
The pathophysiology of AVFMF is complex and multifactorial. It mainly involves intimal hyperplasia, endothelial dysfunction, oxidative stress, and inflammation. Uremia itself promotes inflammation, endothelial dysfunction as well as a hypercoagulability state and platelet dysfunction. In addition, arterial or venous stenosis and hypoxemia are factors influencing the maturation process. Uremia, atherosclerosis, and hyperlipidemia on the one hand, and on the other hand vascular damage induced by surgical interventions, favor AVFMF (15–17). Finally, patients’ age, chronic low blood pressure, smoking, female gender, diabetes mellitus, and the presence of a thrombophilia are additional risk factors for AVFMF (16, 17). Figure 4 summarize the pathogenesis of AVFMF.
To our knowledge, we report for the first time a statistically significant association between AVFMF and aPL (p = 0.04) or APS (p = 0.01). We also found this association with LA positivity alone (p < 0.001). In 2013, Birgitta Salmela et al. performed a prospective observational study, following 219 patients with underlying thrombophilia and assessing primary and functional primary patency defined respectively as the time of AVF creation or first AVF cannulation, until the need for the first vascular access intervention. Eleven percent of these patients (n = 23) had aPL, but the latter was not associated with patency failure. AVF maturation was not analyzed in this specific group (17).
The high prevalence of aPL and APS in the hemodialysis population is currently well established, however its association with AVF thrombosis and stenosis remains uncertain (18, 19). Thrombosis and stenosis both may influence AVF maturation (14), but they cannot always be distinguished. As expected, our study found an association between AVFMF and stenosis or thrombosis (p = 0.007), but independently of thrombosis, only stenosis was associated with AVFMF. In the multivariate analysis, aPL and stenosis were independent risk factors for AVFMF (adjusted OR = 6.8, 95% CI: 1.80, 20.51; p-value 0.004). This suggests that there is a non-thrombotic or non-stenotic mechanism of AVFMF.
A common cause of stenosis is intimal hyperplasia (IH), which is well-described in AVFMF (14, 20). IH is more likely to occur at the anastomotic level and involves endothelial cells activation (21). IH has also been reported in aPL-associated disorders such as aPL-associated nephropathy and is referred to as aPL-associated Vasculopathy and is a non-thrombotic manifestation of APS. Its pathophysiology involves the activation of the mammalian Target of Rapamycin (mTOR) pathway (18, 22). One of the pathophysiological hypotheses that may explain AVFMF in aPL patients, is that endothelial cells activation and intimal proliferation under the influence of aPL will lead to IH, stenosis and/or thrombosis and thus maturation failure. Whether this phenomenon is mediated by mTOR pathway is not yet studied. Interestingly, when looking for aPL, APS and UnK aPL all together, anastomotic stenosis was found in 10.4% versus 4.2% in aPL negative patients. This was even more predominant when considering only APS and aPL subgroups (16.7% vs 4.2%). It would be interesting to explore more specifically IH by Doppler ultrasonography or histology and given our low prevalence of anastomotic stenosis, to confirm this on a larger cohort.
Furthermore, aPL has been associated with accelerated atherosclerosis, arterial vascular disease such as cardiovascular disease and peripheral artery disease (23, 24). Thus, an atherogenic hypothesis has been proposed by some authors and may explain the link between aPL and fistula occlusion (19). Furthermore, atherosclerosis with thickened vessels, and vascular calcification could also lead to an impaired remodeling process, to stenosis and/or thrombosis and therefore to AVFMF. We did not find an association between the aPL, APS and thrombosis or stenosis in our study, as it has been previously described in some studies (18).
Endothelial dysfunction has been reported in APS patients (25). Nitric oxide (NO), which is generated by endothelial NO synthase, has a vasodilatation effect, as well as anti-inflammatory and antiplatelet properties and has been shown to be crucial in AVF maturation (26). Impaired NO formation seems to be a major aspect of the aPL-induced endothelial dysfunction (27, 28). This may lead to insufficient vascular remodeling process of the outflow vein (14, 29). Figure 4 summarize our pathophysiological hypotheses of AVF maturation failure considering the involvement of aPL.
Interestingly, we studied a group of unknown significance (UnK aPL group), which was also associated with AVFMF and stenosis when compared to the negative group (p = 0.002 an p = 0.04 respectively). This group merged patients with one positive antiphospholipid antibody assay without confirmation (either absence of confirmation assay (11.7%) or negative (20.4%)). These patients do not fulfill the diagnosis criteria for APS of aPL, however, this group represents 32% of our cohort and therefore is of interest in terms of clinical outcomes. Antiphospholipid antibody negativation or fluctuation have been described in about 10% of APS patients, with uncertain clinical impact (30, 31). In a cohort of 472 patients of APS ACTION, 11% of patients with clinically meaningful aPL at baseline were unstable at a median follow-up of 5 years (32). Follow up of these patients is crucial, and confirmation of their antiphospholipid antibody status should be performed.
The literature also suggests a role for platelet activation in the maturation of AVF (33). The 2019 ERA-EDTA Clinical practice guidelines on peri- and postoperative care of AVF and grafts for hemodialysis in adults suggest the administration of antiplatelet therapy during the first two months after AVF creation, in order to favor its maturation (3). Our study did not find an association between AVF maturation and antiplatelet therapy. Moreover, AVFMF was not associated with statins, ACE inhibitors or phosphate binder use. Age was associated with AVFMF in the literature (16) but we did not find such association in our cohort. This can be explained by the relatively young age of our HD population (34). In our cohort, we did not find an association between AVFMF and URR, Kt/V, BCM overhydration status. However, aPL was associated with lower URR (p = 0.02).
This study has several limitations: first, it a retrospective monocentric study involving a limited number of patients. Also, the definition of AVFMF by Doppler ultrasound is operator dependent and this may have influenced our results. Finally, not all clinical and anamnestic data were systematically reported, and other data such as utilization of erythropoietin (10), ultrasound mapping before AVF creation (35), the AVF location (36) and AVF geometric parameters (37) could not be collected in this study.