Predictors of In-hospital Mortality in Patients with Acute Limb Ischemia – Lesson Learnt from A Tertiary Hospital in Jakarta


 Background

This study aims to identify risk factors associated with in-hospital and 30-days mortality in patients with acute limb ischemia (ALI).
Methods

This study was a single-centered cohort enrolling a total of 160 consecutive patients with a diagnosis of ALI. The ALI diagnosis was based on clinical history, physical examination, and Doppler studies of the extremities. The main outcome of this study is in-hospital and 30-days mortality.
Results

There were a total of 170 patients involving 192 limbs with the diagnosis of ALI. Intra-aortic balloon pump (IABP) insertion (HR 3.4; 95% CI 1.0-11.3, p = 0.042), no vitamin E treatment (HR 5.6; CI 1.7–18.3, 0.004), arrhythmia (HR 12.00; CI 3.8–37.7, p < 0.001), and acute renal failure (HR 6.70; CI 1.88–24.3, p = 0.003) were an independent predictor of intra-hospital mortality. For 30-days mortality, the independent predictors were menopause (HR 3.2; CI 1.16–8.85, p = 0.02); IABP insertion (HR 4.51; CI 1.14–17.92, p = 0.03); arrhythmia (HR 0.11; CI 0.04–0.32, p < 0.001); bleeding requiring transfusion (HR 3.77; CI 0.10-14.28, p = 0.05); and acute renal failure (HR 5.5; CI 1.79–16.95, p = 0.003).
Conclusion

In-hospital mortality in patients with ALI remains high in our center. Several factors contributing to mortality were arrhythmia, renal failure, no vitamin E supplementation, and a history of recent cardiac operation.


Introduction
Acute limb ischemia (ALI) is one of the major vascular emergencies. Acute limb ischemia is de ned as a sudden decrease in limb perfusion, threatening the limb's viability. 1 Even though the incidence is not as high as acute coronary syndrome, it potentially causes disability and death if inadequately treated. The incidence of this condition is approximately 1.5 cases per 10,000 persons per year. 2 Thirty-day mortality and amputation rates of ALI are up to 15% and 25%, respectively. 3 Acute limb ischemia is caused by thrombosis in situ, embolism from the heart or diseased arteries, aortic dissection, and trauma. Clinically, ALI is classi ed as (I) viable, (II) threatened, and (III) non-viable tissue. This classi cation helps to guide therapeutic decisions regarding the urgency of intervention, appropriate pre-intervention evaluation, and intervention modalities. 1,4 The management of ALI aims to restore arterial ow and viability of the limb. Treatment for non-viable ALI is amputation, while the treatment options for viable and threatened ALI involves endovascular or surgical revascularization. Current advances of endovascular therapeutic approach in ALI management have improved overall amputation rates, but mortality rates remain high. 5 Epidemiological studies on ALI are currently lacking. The incidence and outcome were estimated based on hospital registries, interventional trials, and autopsy studies. Increasing age, level of occlusion, recent myocardial infarction, pre-existing peripheral arterial disease, and cardiopulmonary class are touted to increases morbidity and mortality in patients with ALI. 6 In the modern era of advanced surgical and endovascular revascularization, the prognostic factors may differ from that of prior decades. Factors affecting mortality and morbidity in patients with ALI in the modern era remains inadequately explored.
This study aims to identify risk factors associated with in-hospital and 30-days mortality in patients with ALI.

Study design and data collection
This study was a single-centered cohort enrolling a total of 160 consecutive patients admitted to the National Cardiovascular Center Harapan Kita with a diagnosis of ALI between January 1, 2015 and December 31, 2018. This study is a part of the vascular registry database in the National Cardiovascular Center Harapan Kita which is the national referral center for cardiovascular diseases in Indonesia. The database includes baseline characteristics, diagnosis, comorbidities, procedures performed, and outcome. The ALI diagnosis was based on clinical history, physical examination, and Doppler studies of the extremities. The main outcome of this study is in-hospital and 30-days mortality. All patients were treated based on standard treatment of ALI based on the European Society of Cardiology Guideline on Peripheral Arterial Disease. Additional drugs are left to the discretion of the treating physician. Additional drugs are as follow: Allopurinol 1 × 300 mg, Vitamin E 2 × 400 mg, Pentoxifylline 1200 mg and Sodium bicarbonate 3 × 500 mg. Exclusion criteria were incomplete data (5%). The study has been approved by the institutional review board of the National Cardiovascular Center Harapan Kita

Statistics Analysis
For the baseline characteristics, we classi ed the patients into survivor and non-survivor; Chi-Square and independent t-test analyses were performed to compare the two groups. We performed a Cox-regression analysis for univariate and multivariate analysis to identify factors associated with in-hospital and 30days mortality. Variables with p-value < 0.25 in the univariate analysis were included in a multivariate coxregression analysis to identify the most in uential factors related to in-hospital and 30-days mortality. We considered a two-sided p-value of less than 0.05 as statistically signi cant. All analyses were conducted using SPSS version 23.0.

Results
A total of 170 patients involving 192 limbs were admitted to our hospital with the diagnosis of ALI from January 1st 2015 to December 31st 2018. Ten patients were excluded due to incomplete data. Among 160 patients, 63.1% were male and the mean age was 56 ± 13 years old. In-hospital and 30-days mortality were 28.1% and 36.9%, respectively (Fig. 1). The baseline characteristics of our study population were shown in Table 1. There are no differences between survivor and non-survivor groups in terms of age, gender, onset of ALI, and limb ischemia severity. History of recent operation and aorta involvement were higher in the non-survivor group. Ischemic sign on ECG, low hematocrit, higher creatinine level was more frequent in the non-survivor group. Survival rate was higher in patients receiving Additional treatment such as vitamin E, pentoxifylline, sodium bicarbonate. Complications such as sepsis, acute renal injury, bleeding were more frequent in the non-survivors.   Table 3 showed the result of univariate and multivariate analysis of variables related to 30-days mortality.
In the univariate analysis, risk factors of menopause in women, acute renal failure, arrhythmia, and bleeding requiring transfusion were related to 30-days mortality. In the multivariate analysis, the predictors for 30-days mortality were menopause (HR 3.

Discussion
Despite recent advances in ALI management, it is still a challenging issue associated with high mortality.
The in-hospital mortality rate in our institution was 28.1% in the past 3 years (2015-2018). This number was larger than other studies, which is likely due to the late presentation of patients in our cohort. Clason et al. reported that the 30-days mortality rate among ALI patients was around 26%. 6 Baril et al. showed that the in-hospital mortality for ALI declines from 12.1% in 1998 to 9.0% in 2009 (P < .001) in the United States Medicare population. 5 Endovascular therapy for the treatment of ALI is increasingly used and contributes to mortality reduction in the Medicare population. In our institution, the percentage of patients undergoing endovascular therapy and surgical embolectomy was only 0% and 21.3%, respectively. The remainder of the patients was only managed by anticoagulation without primary reperfusion during hospitalization. The choice of therapeutic modalities was not associated with mortality in our patients. A meta-analysis demonstrated no signi cant difference in short-term and 12-months mortality, limb amputation, and recurrent ischemia in endovascular or surgical approach in ALI. 7 There were differences in the subjects' baseline characteristics regarding history of cardiac operation, aortic lesion, and ischemic sign on ECG between survivor and non survivor groups (P-value: 0.006, < 0.001, and 0.037 respectively). These ndings describe the degree of atherosclerotic burden and vascular damage in a patient's body. Patients who presented with any of these history have higher mortality risk. According to several studies, cardiopulmonary complications account for the majority cause of death. Approximately 15 to 20 percent of patients die within one year of presentation of their limb ischemia, usually from the medical illnesses that predispose them to acute limb ischemia. 1,8,9 Most of the studies reported the risk factors associated with 30-days and 1-year mortality. The data on factors that contribute to in-hospital mortality is currently lacking. This study showed that in-hospital development of cardiac arrhythmias and acute renal failure were associated with mortality. Previously, 30-days mortality and amputation were attributed to several factors, including increasing age, level of occlusion, recent myocardial infarction, pre-existing peripheral arterial disease, and cardiopulmonary functional class. 6 United States Medicare registries indicate that patients with advanced age, chronic renal failure, dementia, cancer, and atrial brillation have higher 30-days mortality. 5 In our registry, all of these factors did not contribute to in-hospital mortality.
History of recent cardiac operation is one of the factors contributing to in-hospital mortality. In this registry, 30 from 160 patients had ALI after CABG or valve surgery. Most of these patients have a history of IABP insertion prior to or after the procedure. Approximately 32.6% of patients with ALI who underwent IABP insertion died in-hospital. Allen et al. found that IABP insertion to cardiac surgery patients is a strong risk factor for acute leg ischemia. Morbidity and mortality of ALI after cardiac surgery were 92% and 46%, respectively. 10 Folkert et al. showed that IABP insertion leads to a ve-fold increase in ALI. 11 It is interesting to note that vascular complications, including acute limb ischemia remains a signi cant risk associated with IABP, occurring in 6-25% of cases. IABP insertion could cause leg ischemia from several mechanisms, insertion to femoral branches rather than the common femoral artery could cause limb ischemia, improper insertion/needle puncture, which is too low is the most common cause of ischemic complication during IABP. 12,13 Lethality in ALI is attributed to the reperfusion injury. After occlusion and tissue ischemia, blood ow restoration may paradoxically exacerbate prior ischemic injury by an overt in ammatory response that promotes local tissue destruction and remote organ dysfunction. 14,15 Reperfusion to severely ischemic muscles may induce the release of toxic metabolites such as potassium, free radicals, and myoglobin, culminating in a life-threatening systemic complications including renal, cardiac and pulmonary failure. 16,17 Reperfusion to rhabdomyolytic muscle induces the release of myoglobin, causing acute tubular necrosis. Reperfusion also induces over ow of intracellular potassium and hydrogen ions due to the destroyed potassium-sodium pump causing metabolic acidosis and cardiac arrhythmia. 16 Acute renal injury and life-threatening arrhythmia are amongst the fatal ALI complications. In our registry, acute renal injury and fatal arrhythmia contributed signi cantly to in-hospital mortality.
Vitamin E has been known for its antioxidant effect and reducing ischemic reperfusion injury. Vitamin E can inactivate reactive oxygen species (ROS), where its production increased after reperfusion occurred. Reperfusion will increase ROS release, in ltration of in ammatory cells and humoral mediators which further potentiates cellular damage. 22 Initiation and propagation of ischemic-reperfusion injury depend upon transcription factor activation, which is responsible for the induction of in ammatory genes required for rapid production of some proteins such as cytokine, adhesion molecules, complement factors, and NO synthase. 23,24 Vitamin E also affects the regulator of signal transduction and modi cation of NF-κB as predominant transcription factor activation during reperfusion. 25 A vitamin E pre-treatment in a model of ischemic-reperfusion of lower limb muscle of patients undergoing aortic cross-clamp during surgical repair of abdominal aortic aneurysm was shown to prevent the accumulation of neutrophils within ischemic and reperfused muscle by reducing expression of endothelial adhesion proteins such as E-selectin and ICAM-1. 26 In patients undergoing aortic aneurysm resection, vitamin E administration was associated with reduced oxidative skeletal muscle damage secondary to ischemiareperfusion injury. 27 In our institution, we administered 200 mg of Vitamin E per day in patients with ALI to reduce ischemiareperfusion injury; however, we cannot provide it for all patients due to insurance reimbursement cap. Surprisingly, our study showed that a no vitamin E treatment in the acute phase of ALI causes a six-fold increased risk for in-hospital mortality. Arato et al. demonstrate that 200 mg vitamin E administration in patients undergoing lower limb vascular surgery, starting from preoperative day until seven days postoperative reduced the level of oxidative stress (lipid peroxidation, antioxidant enzymes) generated after ischemia-reperfusion insult. Vitamin E administration could also reduce white blood cell activation (MPO activation, free-radicals production, expression of adhesion molecules), reverse prooxidant-antioxidant balance and the consecutive local in ammatory process during early reperfusion compare to placebo. 28 To the best of the authors' knowledge, this is the rst study to demonstrate the bene t of Vitamin E supplementation in ALI patients. Further randomized controlled trials should be performed to generate more concrete evidence.
In-hospital morbidity of ALI found in our hospital mainly consists of sepsis, arrhythmia (tachy/bradyarrythmia), bleeding (gastrointestinal and cerebral), acute renal injury, and their proportion was signi cantly higher in non-survivor group. Primary amputation rate during hospitalization is about 20%. We also observed that the number of bleeding correlates with the number sepsis in both survivor and non-survivor groups of in-hospital mortality analysis. Sepsis is one of the most common causes of disseminated intravascular coagulation, which would most likely give way to major bleeding requiring blood transfusion. This explains why bleeding was an independent predictor of mortality in the multivariate analysis for in-hospital and 30-days mortality. 29,30 Limitation The registry was single-center, and the data was collected from a National Referral Hospital. There is a potential for selection bias which may explain the high mortality. The data were collected both prospective and retrospectively from medical records. Some of them have incomplete data or were lost to follow up. The treatments were based on physician discretion; nevertheless, it re ects real-world practice.

Conclusion
In-hospital mortality in patients with ALI remains high in our center. Several factors contributing to mortality were arrhythmia, renal failure, no vitamin E supplementation, and a history of recent cardiac operation.  Figure 1 In-hospital and 30-days mortality