The Impact of Atrial Fibrillation on One-Year Mortality in Patients with Severe Lower Extremity Arterial Disease

Atrial fibrillation (Afib) is associated with the presence of lower extremity arterial disease (LEAD), but its effect on a severe LEAD prognosis remains unclear. We investigated the association between Afib and clinical outcomes. We retrospectively enrolled consecutive severe LEAD patients undergoing percutaneous transluminal angioplasty between 1 January 2013 and 31 December 2018. Patients were divided according to the history of any type of Afib and followed for at least one year. The primary outcome was all-cause mortality. Secondary outcomes were cardiac-related mortality and major adverse cardiovascular events (MACEs). The study included 222 patients aged 74 ± 11 years (54% male), and 12.6% had acute limb ischemia. The Afib group had significantly higher rates of all-cause mortality (42.9% vs. 20.1%, p = 0.014) and MACEs (32.1% vs. 14.4%, p = 0.028) than the non-Afib group. Afib was independently associated with all-cause mortality (adjusted HR: 2.153, 95% CI: 1.084–4.276, p = 0.029) and MACEs (adjusted HR: 2.338, 95% CI: 1.054–2.188, p = 0.037). The other factors associated with all-cause mortality included acute limb ischemia (adjusted HR: 2.898, 95% CI: 1.504–5.586, p = 0.001), Rutherford classification, and heart rate. Afib was significantly associated with increased risks of one-year all-cause mortality and MACEs in patients with severe LEAD. Future studies should investigate whether oral anticoagulants benefit these patients.


Introduction
Atrial fibrillation and peripheral artery disease (PAD) are common atherosclerotic consequences that interact with each other. Patients with atrial fibrillation have a higher incidence of PAD than those without atrial fibrillation [1]. Conversely, PAD is also associated with an increased risk of developing atrial fibrillation [2]. A nationwide cohort study showed that PAD was independently associated with new-onset atrial fibrillation [3]. One type of severe complications of atrial fibrillation are thromboembolic events, which are frequently diagnosed in PAD patients. These episodes are not limited to cerebrovascular events and include peripheral artery thromboembolic complications, such as splenic, renal, and mesenteric artery infarction, and acute limb ischemia [4]. Overall mortality and cardiac-related mortality were also found to be elevated in patients with atrial fibrillation and symptomatic PAD, independent of age and comorbidities [5]. Lower extremity arterial disease (LEAD), defined by an ankle-brachial index ≤ 0.9, was found to be associated with two-fold higher risks of myocardial infarction and vascular death in patients with atrial fibrillation [6]. Baseline atrial fibrillation was reported to be associated with an increased risk of an incident of acute limb ischemia in a large-scale case-control study [7]. Patients with LEAD without a medical history of atrial fibrillation at the baseline had an increased incidence of atrial fibrillation in a population-based cohort study [8]. Regarding atrial fibrillation and LEAD, it remains unclear which comes first [3,9], but it is known that patients with concomitant atrial fibrillation and PAD have a synergistically elevated risk of cardiac mortality [10]. To date, the effect of atrial fibrillation on prognosis has rarely been reported in patients with severe LEAD. Therefore, we conducted the present study to comprehensively investigate mortality and cardiovascular outcomes in patients with severe LEAD and atrial fibrillation.

Materials and Methods
We retrospectively enrolled consecutive patients with severe LEAD who received percutaneous transluminal angioplasty at our hospital between 1 January 2013 and 31 December 2018. The present study enrolled all-comers: we excluded three patients with critical limb ischemia and unsalvageable limbs who refused surgical amputation. The patients' baseline characteristics, laboratory data, procedural details, and outcomes were collected from the medical records. We followed the patients until 31 December 2019. The study was approved by the Mackay Memorial Hospital under Institutional Review Board number 20MMHIS034e. The informed consent was waived by the Mackay Memorial Hospital institutional review board because the present study was a low-risk retrospective cohort study. We conducted the present study in accordance with the Declaration of Helsinki and performed the analyses on anonymized data.
We performed emergency and urgent percutaneous transluminal angioplasty on patients with acute and chronic limb ischemia, respectively; elective percutaneous transluminal angioplasty was reserved for those patients with Rutherford stage IV disease. Physicians could decide whether to use antegrade or retrograde vascular access based on the type of lesion. During the procedure, we routinely administered heparin to maintain an active clotting time between 250 and 300 s and dual antiplatelet agents, namely, aspirin and clopidogrel, according to the current guidelines. Generally, we treated iliac and femoropopliteal lesions with balloon angioplasty, a drug-coated balloon or a stent, and we treated below-knee lesions with only balloon angioplasty. The decision to use a drug-coated balloon or stent was mainly influenced by the patient's economic status and left to the physicians' discretion. After the procedure was performed, we continued to administer dual antiplatelet agents for at least one month except in the cases of clinically relevant bleeding complications. Because cilostazol has minimal effects on limb salvage and mortality, we did not routinely prescribe it to severe LEAD patients.
We defined severe LEAD based on the Rutherford classification system as follows: resting pain (stage IV), tissue loss (stage V), and gangrene (stage VI) [11]. The primary study outcome was all-cause mortality at the one-year follow-up. The secondary outcomes were cardiac-related mortality, major adverse cardiac events (MACEs) and major adverse limb events (MALEs) at the one-year follow-up and in-hospital mortality. We defined MACEs as the composite of cardiac-related mortality, nonfatal myocardial infarction, and nonfatal stroke and MALEs as amputation due to a vascular event above the forefoot, acute limb ischemia and clinically driven target vessel revascularization.
We present continuous variables as numbers and standard deviations and binary variables as numbers and percentages. We used independent t-tests and chi-squared tests to evaluate differences in continuous variables and binary variables, respectively.
We first used univariate logistic regression analyses to identify potential confounders. Then, we adjusted for these confounders in multivariate logistic regression analyses of the relationships between the presence of atrial fibrillation and the study outcomes. In the statistical models, we adjusted for the presence of acute limb ischemia and the Rutherford classification because these factors obviously contributed to the study outcomes. In the sensitivity analyses, we additionally adjusted for kidney function and body mass index if they were not associated with the study outcomes in multivariate logistic regression. Kidney function was represented by the serum creatinine level, creatinine clearance as determined with the Cockcroft-Gault formula, or the estimated glomerular filtration rate as determined by the Modification of Diet in Renal Disease study equation, as in our previous study [12]. Body mass index was divided by the thresholds of 20 and 25 into groups as follows: <20, ≥20 and <25, and ≥25 [13,14]. We calculated two-tailed p-values, and we considered values of 0.05 or lower to be significant. We performed all statistical analyses with the Statistical Package for the Social Sciences software (SPSS software, version 20.0, International Business Machines, Chicago, IL, USA).

Discussion
In the present study, we showed that in addition to acute limb ischemia and Rutherford stage, atrial fibrillation was an independent predictor of all-cause mortality. The patients with atrial fibrillation had lower creatinine levels but a higher incidence of allcause mortality than the patients without atrial fibrillation. In the various statistical models, atrial fibrillation doubled the risk of all-cause mortality, and the significant associations were fairly consistent, although only a tendency toward significance was found after we adjusted for body mass index in addition to the other cofounders. In our study, we did not find a significant difference in body mass index between the patients with and without atrial fibrillation, and we therefore did not initially adjust for it as a confounder in our reduced statistical models. The effect of body mass index on mortality remains controversial in patients with LEAD [13,15]. Body mass index was found to have a complex inverse association with the three-year survival rate in patients with critical limb ischemia undergoing endovascular treatment for infrapopliteal lesions in a previous study: overweight patients had the highest one-year survival rate (86.1%), followed by normal-weight patients (78.6%), and underweight patients had the lowest mortality rate (68.4%) [15]. In contrast, the Atherosclerosis Risk in Communities (ARIC) study showed that body mass index was positively associated with the incidence of peripheral artery disease requiring hospitalization, and the risk was relatively higher in patients with critical limb ischemia [13]. We believe that a low body mass index value (<20 kg/m 2 ) indirectly reflects a poor nutritional status and compromised wound healing in patients with severe LEAD, whereas a high body mass index value (>25 kg/m 2 ) in this population is generally associated with a greater prevalence of cardiovascular comorbidities, such as hypertension, diabetes, and dyslipidemia. Therefore, we suggest that it is beneficial for patients with severe LEAD to maintain a normal weight [13]. The percentage of statin treatment is remarkably low in both groups, particularly in patients with atrial fibrillation (7.1%), whereas treatment with oral anticoagulants was prescribed in only 25% of the latter. Because our hospital is located in a remote area, and most of the patients did not receive regular primary prevention for hypertension, diabetes mellitus, or hyperlipidemia, these patients with the first medical contact with our hospital presented with severe LEAD. Though we tried to increase the prescription rate of statin at discharge, it was not provided by our national health insurance at that time. We believe that guideline-direct medical therapy can improve prognoses in these patients with severe LEAD.
In this study, we enrolled patients with severe LEAD, and all of our patients with atrial fibrillation had CHA2DS2Vsc scores ≥3. According to the current guidelines, oral anticoagulants are strongly recommended for patients with nonvalvular atrial fibrillation and a CHA2DS2Vsc score ≥2 [16,17]. To date, no randomized controlled trial has been published to determine which antithrombotic regimen is best in patients with simultaneous atrial fibrillation and severe LEAD. The ROCKET-AF trial showed that patients with atrial fibrillation and LEAD did not have an elevated risk of stroke or systemic embolization, and a higher risk of major bleeding or nonmajor clinically relevant bleeding was found in patients treated with rivaroxaban than in those treated with warfarin [18]. The ARISTOTLE trial showed that patients with or without LEAD did not have a significantly different risk of stroke or systemic embolization, and the benefits of apixaban and warfarin were similar [19]; however, the two studies enrolled patients with atrial fibrillation and performed a subgroup analysis of patients with LEAD [18,19], which is quite different from a study that enrolled LEAD patients with or without atrial fibrillation, based on the different natures of atrial fibrillation and LEAD. Although these two diseases have concomitant risk factors and share the characteristics of an increased inflammatory burden and a prothrombotic state, endothelial dysfunction and atherosclerosis play major roles in LEAD but not atrial fibrillation [20]. The VOYAGER trial showed that very-low-dose rivaroxaban in addition to aspirin compared with aspirin alone significantly reduced the incidence of acute limb ischemia in patients with peripheral artery disease after revascularization, but all-cause mortality did not differ significantly between the two groups. Notably, the VOYAGER trial excluded patients with atrial fibrillation and patients with major tissue loss [21], but we did not exclude patients with chronic ulcers or major tissue loss of the lower extremities; instead, we enrolled all-comers. The four major randomized-controlled trials included the Afib patients receiving the percutaneous intervention but did not report the medical histories of LEAD in these studies. [22][23][24][25]. A gap existed in patients with concomitant Afib and LEAD, and we provided limited data to complement the studies. Based on the similar pathophysiology of coronary artery disease and LEAD, patients with atrial fibrillation who received percutaneous coronary stenting could be included in our study. The current evidence from four major randomized controlled trials showed that direct oral anticoagulants (DOACs) plus P2Y12 inhibitors can prevent bleeding complications in patients with atrial fibrillation who receive coronary stent(s) without increasing the risk of stent thromboses. Although the selection of the proper antithrombotic regimen for patients with atrial fibrillation and severe LEAD is far beyond the scope of this study, our results showed that atrial fibrillation increased the risk of mortality, and it is worth investigating the effect of antithrombotic treatment in this population in the future.
This study had several limitations. First, we enrolled patients with severe LEAD, so the number of patients in our study population was relatively small. Therefore, we lacked the adequate power to identify a significant association between atrial fibrillation and the other study outcomes after we adjusted for confounders. We did not find a significant association between atrial fibrillation and MACEs, including cardiac-related mortality, nonfatal myocardial infarction, and nonfatal stroke; however, a systematic review and meta-analysis showed that atrial fibrillation was associated with an increased risk of MACEs [26]. Second, our study enrolled all-comers. We noted that a relatively small proportion of the patients had received secondary prevention for dyslipidemia before they presented at our hospital. In addition, the percentage of the patients who had been prescribed oral anticoagulants seemed low in the present study. The percentage of smokers was relatively low in our study, but it is quite unusual in patients with LEAD. We thought some biases might result from a retrospective cohorts study design, such as recall and record biases. Although these biases were non-differential in the two groups, and the statistical results may not change. In our study, 53% of the patients with atrial fibrillation had a creatinine clearance value less than 30 mL/min, and the net clinical benefits should be balanced between antithrombotic treatment and bleeding complications in these patients. We will try to enroll additional patients with atrial fibrillation and prospectively conduct a study to investigate the effect of antithrombotic treatment on the study outcomes.

Conclusions
Our retrospective cohort study showed that patients with severe LEAD and atrial fibrillation had higher incidences of all-cause and MACEs than those without atrial fibrillation. We showed the adverse effect of atrial fibrillation on all-cause mortality while acute limb ischemia and the Rutherford stage were also associated with all-cause mortality. Future studies should investigate the effect of antithrombotic treatment on mortality in this study population.  Informed Consent Statement: The informed consent was waived by the Mackay Memorial Hospital institutional review board because the present study is a low-risk retrospective cohort study.

Data Availability Statement:
The study data can be provided if the correspondence judges a request is reasonable.

Conflicts of Interest:
The authors declare no conflict of interest.