Direct Oral Anticoagulants as the First Choice of Anticoagulation for Patients with Peripheral Artery Disease to Prevent Adverse Vascular Events: A Systematic Review and Meta-Analysis

The best method of anticoagulation for patients with peripheral artery disease (PAD) is still a topic of interest for physicians. We conducted a meta-analysis to compare the effects of direct oral anticoagulants (DOACs) with those of vitamin-K-antagonist (VKA) anticoagulants in patients with peripheral artery disease. Five databases (Medline (via PubMed), EMBASE, Scopus, Web of Science, and CENTRAL) were searched systematically for studies comparing the effects of the two types of anticoagulants in patients with PAD, with an emphasis on lower-limb outcomes, cardiovascular events, and mortality. In PAD patients with concomitant non-valvular atrial fibrillation (NVAF), the use of DOACs significantly reduced the risk of major adverse limb events (HR = 0.58, 95% CI, 0.39–0.86, p < 0.01), stroke/systemic embolism (HR 0.76; 95% CI 0.61–0.95; p < 0.01), and all-cause mortality (HR 0.78; 95% CI 0.66–0.92; p < 0.01) compared with warfarin, but showed similar risks of MI (HR = 0.81, 95% CI, 0.59–1.11, p = 0.2) and cardiovascular mortality (HR = 0.77, 95% CI, 0.58–1.02, p = 0.07). Rivaroxaban at higher doses significantly increased the risk of major bleeding (HR = 1.16, 95% CI, 1.07–1.25, p < 0.01). We found no significant difference in terms of revascularization (OR = 1.49, 95% CI, 0.79–2.79, p = 0.14) in PAD patients in whom a poor distal runoff was the reason for the anticoagulation. DOACs have lower rates of major limb events, stroke, and mortality than VKAs in PAD patients with atrial fibrillation. Rivaroxaban at higher doses increased the risk of major bleeding compared with other DOAC drugs. More high-quality studies are needed to determine the most appropriate anticoagulation regimen for patients with lower-limb atherosclerosis.


Introduction
Peripheral artery disease (PAD) affects more than 200 million people worldwide and is responsible for a significant proportion of limb losses and even fatal cardiovascular events due to the progression of the underlying atherosclerotic process [1]. Recent standard protocols prescribe antiplatelet therapy after surgical revascularization procedures [2][3][4][5]. The anticoagulation of these patients is still a controversial topic. The most common condition indicating anticoagulation is non-valvular atrial fibrillation (NVAF) [6]. The concomitant presence of atrial fibrillation and symptomatic peripheral artery disease is frequently reported, especially in the older population [7]. These two conditions share several common 2 of 16 risk factors, and their concomitant presence increases the risk of cardiovascular mortality and the incidence of major adverse cardiovascular events exponentially [8][9][10][11]. Therefore, it is important to reduce the risk of these complications with the appropriate drug treatment. DOACs have been shown to be superior in efficacy and safety to vitamin-K-antagonist anticoagulants in the prevention of venous thrombotic events [12][13][14][15][16][17][18], but warfarin and coumadin derivatives are still frequently used anticoagulant drugs in NVAF patients. The detrimental effect of long-term vitamin-K-antagonist therapy on the progression of atherosclerosis has been reported, but there are no clear data on whether the DOACs have the same effect [19][20][21][22]. Nowadays, there is still no clear recommendation for anticoagulation in patients with PAD alone and without any other indication for anticoagulation. In our study, we aimed to determine whether there was a difference in the clinical endpoints in patients with PAD taking direct oral anticoagulants compared with patients taking conventional vitamin-K antagonists. We performed a systematic review and meta-analysis to evaluate the efficacy and safety of DOACs compared with VKA therapy in patients with NVAF and PAD, focusing on mortality, major adverse limb events, major adverse cardiovascular events, and major bleeding.

Materials and Methods
This systematic review and meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Network Meta-Analysis (PRISMA-NMA) statement and following a protocol preregistered at the international prospective register of systematic reviews (PROSPERO) with the registration number CRD42021288677.
For comprehensiveness, we also examined the reference lists of eligible studies. The search was performed in September of 2021; therefore, articles published before this date were selected for the further selection process.

Inclusion Criteria
As few randomized controlled trials are available on the subject, we also included cohort studies in our study. A publication was included in our study if it met the following criteria: 1. the study contained a direct comparison of DOACs and VKA anticoagulants; 2. participants were adult patients (>18 years) suffering from PAD; 3. participants received anticoagulant, more frequently for NVAF or after a lower-extremity arterial revascularization procedure (open surgery or endovascular intervention) to improve patency of the bypass graft or due to poor runoff; 4. patients in the intervention group were receiving DOACs (rivaroxaban, apixaban, dabigatran, edoxaban); 5. patients in the control group were treated with VKA anticoagulant drugs (acenocoumarol, warfarin); 6. the study aimed to compare two types of oral anticoagulants; 7. the study contained at least one of the following outcomes: acute limb ischemia, major amputation, myocardial infarction, ischemic stroke, cardiovascular mortality, or all-cause mortality (efficacy indicators); and/or major bleeding (safety indicator).

Exclusion Criteria
We excluded ongoing trials without results, animal studies, conference abstracts, editorials, case reports, and studies with no original data.

Data Extraction
The publications collected from the databases were screened by two independent authors (E.P. and B.K.). After duplicate removal, publications were screened first by title and abstract, then by full text. Disagreements were resolved by consensus. From each article, we extracted for further analysis important data, such as basic information (author, year of publication, study design, and number of centers); study characteristics (population size, follow-up time); demographic characteristics of the patients involved, such as age, gender, relevant comorbidities, and medication; type and dose of DOACs and VKA; and outcomes of interest (major adverse limb events, incidence of revascularization procedures and amputation, incidence of myocardial infarction, ischemic stroke during follow-up time, all-cause, and cardiovascular mortality). Adjusted hazard ratios (HRs) or propensity-scorematched HRs with 95% confidence intervals (CIs) or calculated odds ratios (ORs) were extracted as measures of effect.

Quality Assessment
The Cochrane Collaboration tool (Rob-2) was used to assess the risk of bias for the three post hoc analyses of RCTs and the ROBINS-I tool for the non-randomized studies of interventions [23,24]. The results of the risk-of-bias assessment are summarized in Figure 1.
were treated with VKA anticoagulant drugs (acenocoumarol, warfarin); 6. the study aimed to compare two types of oral anticoagulants; 7. the study contained at least one of the following outcomes: acute limb ischemia, major amputation, myocardial infarction, ischemic stroke, cardiovascular mortality, or all-cause mortality (efficacy indicators); and/or major bleeding (safety indicator).

Exclusion Criteria
We excluded ongoing trials without results, animal studies, conference abstracts, editorials, case reports, and studies with no original data.

Data Extraction
The publications collected from the databases were screened by two independent authors (E.P. and B.K.). After duplicate removal, publications were screened first by title and abstract, then by full text. Disagreements were resolved by consensus. From each article, we extracted for further analysis important data, such as basic information (author, year of publication, study design, and number of centers); study characteristics (population size, follow-up time); demographic characteristics of the patients involved, such as age, gender, relevant comorbidities, and medication; type and dose of DOACs and VKA; and outcomes of interest (major adverse limb events, incidence of revascularization procedures and amputation, incidence of myocardial infarction, ischemic stroke during followup time, all-cause, and cardiovascular mortality). Adjusted hazard ratios (HRs) or propensity-score-matched HRs with 95% confidence intervals (CIs) or calculated odds ratios (ORs) were extracted as measures of effect.

Quality Assessment
The Cochrane Collaboration tool (Rob-2) was used to assess the risk of bias for the three post hoc analyses of RCTs and the ROBINS-I tool for the non-randomized studies of interventions [23,24]. The results of the risk-of-bias assessment are summarized in Figure  1.

Statistical Analysis
For the selected groups, either hazard ratios (HR) or raw patient numbers were extracted and analyzed. For dichotomous outcomes, the odds ratio (OR) with 95% confidence intervals (CI) was used to measure effect. Raw data from the selected studies were pooled using a random-effects model. We estimated the τ 2 using restricted maximum likelihood approach, and the Q-profile method for calculating the confidence interval of τ 2 . Statistical heterogeneity across trials was assessed using the Cochrane Q-test and the I 2 values. Subgroup analyses followed the descriptions of Harrer et al. [25]. Outlier and influence analyses were performed following the recommendations of Harrer et al. [25], and Viechtbauer and Cheung [26].

Results
The literature search yielded a total of 1089 articles. During the selection procedures, 52 full-text articles were screened for eligibility, of which a total of 12 articles (three post hoc analyses from big randomized controlled trials [30][31][32], and nine observational cohorts [33][34][35][36][37][38][39][40][41]) were finally included ( Figure 2). lihood approach, and the Q-profile method for calculating the confidence interval o Statistical heterogeneity across trials was assessed using the Cochrane Q-test and t values. Subgroup analyses followed the descriptions of Harrer et al. [25]. Outlier an fluence analyses were performed following the recommendations of Harrer et al. [25], Viechtbauer and Cheung [26].

First Author
Year  Table 2. Basic characteristics and patient demographics; parameters expressed as mean with standard deviation (*), or median with interquartile range (**); DOAC-direct-acting oral anticoagulant, NI-no information, VKA-vitamin-K antagonist. All results are presented in Table 3 and Figure 3. To assess the strength of recommendations for each outcome, we used the online GRADEpro tool (https://www.gradepro.org), which is based on the GRADE (Grading, Development and Evaluation of Recommendations) method, providing a useful way of assessing the quality of evidence [42], (Table 3). Our primary point of interest is the lower-limb events, which are reported as a composite outcome of major adverse limb events (MALE). It is defined as a summary of lower-limb revascularization and amputation events. We analyzed this outcome separately, according to whether the indication for anticoagulation was the concomitant NVAF or the aim to improve the patency after an open or endovascular lower-extremity arterial procedure. were significantly less likely to experience a MALE during the study period than patients in the VKA group, with moderate heterogeneity (HR = 0.58, 95% CI, 0.39-0.86, p < 0.01, I 2 : 32%).
We found four observational studies on PAD patients without AF, where the need for reoperation was a separate outcome. In this case, the data of 2,323 patients were available, and we found no significant difference between the two anticoagulant groups, without any evidence of heterogeneity (OR = 1.49, 95% CI, 0.79-2.79, p = 0.14, I 2 : 6%). Unfortunately, there were only a few data available about the need for amputation; therefore, no statistical analysis was applicable.
As for the safety of the anticoagulant medication, major bleeding was observed in most of the studies on PAD patients with AF. Comparing all the DOACs with the VKAs, we found similar risks (HR = 0.91, 95% CI, 0.74-1.12, p < 0.01, I 2 :91%) for the occurrence of major bleeding episodes. For this outcome, we analyzed the results separately due to the considerable heterogeneity, and found that rivaroxaban at higher doses significantly increased the risk of bleeding (HR = 1.16, 95% CI, 1.07-1.25, p < 0.01, I 2 : 12%); on the other hand, we found a significantly lower risk of major bleeding in the composite group of the other three DOAC drugs at conventional dosages and rivaroxaban administered at a reduced dose (HR = 0.71, 95% CI, 0.63-0.79, p < 0.01, I 2 : 35%).

Discussion
In patients with atrial fibrillation and PAD, optimal anticoagulation is of key importance, but the progression of atherosclerosis as the underlying disease also needs to be considered. Apart from this study, no other study has attempted to directly compare patients who take anticoagulants after revascularization.
Our results show that, in patients with NVAF and concomitant PAD, the incidence of MALE was significantly lower in DOAC users compared with VKA users. In the three articles reporting this composite outcome, rivaroxaban was compared with warfarin at the same doses. Coleman et al. described [36] that rivaroxaban was associated with a significant reduction in the risk of major thrombotic vascular events (MTVEs), including cardiovascular events, as well as major adverse limb events. Baker et al. and Chan et al. mainly studied NVAF patients with diabetes [34,35]. In all the participants of their studies as well as in the subgroup of PAD patients, rivaroxaban was associated with a significantly lower risk of MALE, and this effect was due to a reduction in the risk of both major limb amputation and endovascular revascularization; however, the risk of surgical revascularization did not differ between the groups. In a nationwide retrospective cohort study from Taiwan [38], the authors analyzed the data of 7802 AF patients with concomitant PAD who were receiving anticoagulant medication. According to their results, all the DO-ACs (dabigatran, rivaroxaban, apixaban, and edoxaban) were also associated with a lower cumulative risk of lower-limb embolization or amputation and revascularization procedures compared with warfarin.
According to our findings, the use of DOACs in AF patients with PAD was associated with a significantly reduced risk of stroke/STE and all-cause mortality, but we did not find a statistically significant difference in terms of CV mortality and MI. Baker et al. reported results similar to ours [34], stating that AF patients who received a reduced dose of rivaroxaban compared with warfarin had significantly decreased the rates of MALE but not of major adverse cardiac events (MACE).
According to Lee et al., PAD patients with AF had similar rates of ischemic stroke with rivaroxaban, but a significantly lower annual incidence of acute myocardial Three articles contained this outcome, which involved 13,561 PAD patients who received anticoagulation therapy because of their concomitant NVAF. All three articles used propensity-score-matched data. Our results showed that patients in the DOAC group were significantly less likely to experience a MALE during the study period than patients in the VKA group, with moderate heterogeneity (HR = 0.58, 95% CI, 0.39-0.86, p < 0.01, I 2 : 32%).
We found four observational studies on PAD patients without AF, where the need for reoperation was a separate outcome. In this case, the data of 2323 patients were available, and we found no significant difference between the two anticoagulant groups, without any evidence of heterogeneity (OR = 1.49, 95% CI, 0.79-2.79, p = 0.14, I 2 : 6%). Unfortunately, there were only a few data available about the need for amputation; therefore, no statistical analysis was applicable.
As for the safety of the anticoagulant medication, major bleeding was observed in most of the studies on PAD patients with AF. Comparing all the DOACs with the VKAs, we found similar risks (HR = 0.91, 95% CI, 0.74-1.12, p < 0.01, I 2 : 91%) for the occurrence of major bleeding episodes. For this outcome, we analyzed the results separately due to the considerable heterogeneity, and found that rivaroxaban at higher doses significantly increased the risk of bleeding (HR = 1.16, 95% CI, 1.07-1.25, p < 0.01, I 2 : 12%); on the other hand, we found a significantly lower risk of major bleeding in the composite group of the other three DOAC drugs at conventional dosages and rivaroxaban administered at a reduced dose (HR = 0.71, 95% CI, 0.63-0.79, p < 0.01, I 2 : 35%).

Discussion
In patients with atrial fibrillation and PAD, optimal anticoagulation is of key importance, but the progression of atherosclerosis as the underlying disease also needs to be considered. Apart from this study, no other study has attempted to directly compare patients who take anticoagulants after revascularization.
Our results show that, in patients with NVAF and concomitant PAD, the incidence of MALE was significantly lower in DOAC users compared with VKA users. In the three articles reporting this composite outcome, rivaroxaban was compared with warfarin at the same doses. Coleman et al. described [36] that rivaroxaban was associated with a significant reduction in the risk of major thrombotic vascular events (MTVEs), including cardiovascular events, as well as major adverse limb events. Baker et al. and Chan et al. mainly studied NVAF patients with diabetes [34,35]. In all the participants of their studies as well as in the subgroup of PAD patients, rivaroxaban was associated with a significantly lower risk of MALE, and this effect was due to a reduction in the risk of both major limb amputation and endovascular revascularization; however, the risk of surgical revascularization did not differ between the groups. In a nationwide retrospective cohort study from Taiwan [38], the authors analyzed the data of 7802 AF patients with concomitant PAD who were receiving anticoagulant medication. According to their results, all the DOACs (dabigatran, rivaroxaban, apixaban, and edoxaban) were also associated with a lower cumulative risk of lower-limb embolization or amputation and revascularization procedures compared with warfarin.
According to our findings, the use of DOACs in AF patients with PAD was associated with a significantly reduced risk of stroke/STE and all-cause mortality, but we did not find a statistically significant difference in terms of CV mortality and MI. Baker et al. reported results similar to ours [34], stating that AF patients who received a reduced dose of rivaroxaban compared with warfarin had significantly decreased the rates of MALE but not of major adverse cardiac events (MACE).
According to Lee et al., PAD patients with AF had similar rates of ischemic stroke with rivaroxaban, but a significantly lower annual incidence of acute myocardial infarction (MI) with warfarin [38]. In the study by Lopes et al. [39], all DOACs were associated with lower stroke/MI/all-cause mortality rates compared with warfarin. The studies by Hu and Jones reported data on MI, finding no significant difference in MI risk and CV mortality between the DOAC and warfarin groups. However, in the Cunningham study, the risk of CV mortality was higher with high-dose edoxaban than with warfarin [30][31][32].
In a meta-analysis examining the relationship between AF and PAD, Zhu et al. stated [43] that the occurrence of PAD in patients with AF could increase the risk and incidence of several adverse clinical events, including all-cause mortality, cardiovascular (CV) death and MACE; however, they did not find a statistically significant difference in the incidence of major bleeding, myocardial infarction (MI), and stroke among AF patients with and without PAD.
Another meta-analysis by Liao et al. found [44] similar thromboembolic and bleeding risks in AF patients with and without PAD, but patients with PAD had an increased risk of death compared with those without it. In their ROCKET AF trial, Pokorney et al. examined patients who were anticoagulated for NVAF [45], and collected factors associated with a higher risk of mortality. They found that PAD, heart failure, and diabetes were most strongly associated with a higher likelihood of CV death.
In PAD patients without atrial fibrillation and no other indication for anticoagulation but to improve graft patency after lower-extremity surgery, Kretschmer et al. already made a comparison between the postoperative use of antiplatelet (aspirin) and VKA anticoagulants in 1992 [46]. Twenty years later, the Dutch BOA study and several other researches confirmed their original conclusion that VKA treatment is associated with improved graft-patency rates when a vein graft was used, while there is no difference with prosthetic grafts. However, patients receiving an artificial graft might profit more from platelet inhibitors [47]. Recent guidelines [2][3][4][5][6] also suggest that, after endovascular revascularization, a period of combination therapy of anticoagulants and antiplatelets should be considered bearing in mind the bleeding and thrombotic risks, but the period of this combination therapy should be as brief as possible.
Regarding the appropriate antiplatelet and/or anticoagulant therapy in PAD patients, two large randomized controlled trials carried out in recent years have emphasized the beneficial effects of aspirin and low-dose rivaroxaban in the prevention of cardiovascular consequences and lower-extremity events compared with the use of aspirin alone or a higher dose of rivaroxaban alone [48][49][50].
Smith et al. highlighted [51] that the overall use of anticoagulants increased to onethird of all below-the-knee bypasses secondary to the greater use of DOACs. This is due to the widespread adaptation after the publication of the COMPASS trial.
Currently, in our study, we have only found four single-center retrospective studies with a small number of cases directly comparing the two anticoagulant groups (DOAC vs. VKA) in terms of their postoperative use. No larger clinical trial or review is available. We found no significant overall difference in the need for reoperation between patients who were prescribed DOAC or VKA postoperatively. Talukadar et al. found the safety and efficacy profile of rivaroxaban to be comparable to that of warfarin when used in patients after peripheral arterial procedures [41]. The results of Aurshina et al. suggest [33] that therapy with DOACs has an excellent graft primary patency rate at one-year follow-up. According to Ferreira et al. [37], rivaroxaban has equivalent efficacy to acenocoumarol after infrainguinal bypass revascularization, with similar rates of occlusion, major amputation, and mortality. In addition, Obi et al. found in patients undergoing lower-extremity surgical bypass that those receiving DOAC postoperatively had a shorter length of stay and were less likely to receive a transfusion in the following 30 days without compromising the graft patency and readmission rates for anticoagulation complications, thrombectomy, or thrombolysis, or affecting the amputation rate compared with those receiving a VKA [40].
On the basis of the results of these smaller-volume retrospective cohorts, DOACs and VKAs are likely to show equivalent or similar patency and amputation rates following revascularization [33,37,40,41].
Another important issue about anticoagulation is its safety. Although there is almost no difference in the efficacy in preventing thromboembolic complications, there are significant differences in the safety profile of the anticoagulants. In our study, we found that rivaroxaban at a daily dose of 15 or 20 mg significantly increased the risk of bleeding compared with VKA. On the other hand, we found a significantly lower risk of major bleeding episodes in users of the other three DOAC drugs, or even with rivaroxaban at lower doses (2.5 or 5 mg per day). Ingason et al. found [52] that the use of rivaroxaban was associated with higher overall rates of gastrointestinal bleeding (GIB) compared with apixaban and dabigatran in patients with AF. Wang et al. reported [53] that AF patients taking apixaban and dabigatran, but not rivaroxaban, experienced fewer bleeding events compared with warfarin. Radadiya et al. performed a network meta-analysis including 28 RCTs and 139,587 patients. In the study, DOACs at a standard dose, rivaroxaban at 20 mg daily, dabigatran at 150 mg twice daily, and edoxaban at 60 mg daily, but not apixaban at 5 mg twice daily, had a higher risk of major GIB compared with warfarin. The comparison of DOACs with each other did not show risk differences [54]. Numerous other authors state that rivaroxaban should be treated with caution, especially at higher doses, which is also supported by our results. Jones described that NVAF patients with PAD had a higher risk of experiencing a major bleeding episode than those without PAD [31]. We believe that the higher risk of bleeding is due to the concomitant antiplatelet drug usage independently of the anticoagulant dose in PAD patients. Chan et al. also suggest in their study [35] that they also found a lower risk of major bleeding for DOACs compared with warfarin, especially in patients who did not take concomitant antiplatelet medication besides the anticoagulation. Moreover, it is also important to highlight that patients at a higher risk of thromboembolic events are also at a higher risk of bleeding. In their systematic review and meta-analysis, Almas et al. assessed the safety and efficacy of DOACs with and without acetylsalicylic acid (ASA). The risk of major bleeding was significantly lower in the DOAC-alone group compared with the DOAC-plus-ASA group [55]. In the current literature, apixaban appears to have the best safety profile for bleeding among the four available DOACs, with similar efficacy to warfarin for stroke/SE [56][57][58][59].
We believe that all our findings provide useful information to help to select the optimal anticoagulant, although efficacy and bleeding risk should be carefully evaluated, especially in the presence of comorbidities such as peripheral artery disease.

Strengths and Limitations
We acknowledge the strengths and limitations of this meta-analysis. The strength of the study is that it involves thousands of patients. In these cohorts, propensity-score-matched data are available. Our study provides useful findings on lowerlimb outcomes and summarizes the latest results and recommendations.
The limitations of the study include the fact that residual confounding results are due to unmeasured factors, such as the lack of an international normalized ratio (INR) for patients treated with warfarin, the body weight, and the accurate renal function data which may have affected the validity of our findings. There could be a misclassification or miscoding of baseline comorbidities as well as potential bias due to different coding systems in different countries. The nomenclature is also not uniform: there is a significant overlap between the meanings of coronary artery disease (CAD), lower-extremity artery disease (LEAD), and PAD. Data reporting in a non-uniform format makes statistical analysis difficult. The use of different drug doses and differences in patient follow-up times lead to difficulties in the proper analysis of the data. Due to the limited data, we were not able to perform a subgroup analysis by age, comorbidities, or medication.

Conclusions
We have pointed out that, based on the present meta-analysis, the use of DOACs versus VKAs in PAD patients with NVAF is associated with significantly better outcomes in terms of major limb events, stroke, and mortality. Rivaroxaban at conventional doses increases the risk of major bleeding compared with other DOAC drugs. On the other hand, DOAC and VKA seem to produce equivalent or similar patency rates following infrainguinal revascularization procedures, but there is an absence of strong evidence.
Using DOACs in patients undergoing lower-extremity arterial procedures may play a more significant role in the future, but further investigations are needed for definitive results and safe decision-making.

Funding:
The authors received no financial support for the research, authorship, and/or publication of this article.
Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: Data are contained within the article.

Conflicts of Interest:
The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Sample Availability: Samples of the compounds are available from the authors.