Surgical management of acute limb ischemia, the first experience from Ethiopia

Background Acute Limb Ischemia is a devastating emergency condition due to a sudden decrease in limb perfusion that threatens life or limb viability. It carries a high morbidity and mortality rate. This is the first experience to report on risk factors, etiologies and surgical outcomes of acute limb ischemia in Ethiopian. Methods A Prospective Cross-sectional Study on patients operated for acute limb ischemia at and from January 1, 2018, to December 31, 2018, Ababa, Ethiopia was done. They were followed for 3 months to determine risk factors, etiologies and surgical outcomes of the disease. Data were collected using a structured questionnaire. Findings were analyzed using SPSS version 20 and compared with current literature. Result A total of 102 patients were operated with a male to female ratio of 2:1 and the mean age at presentation was 54±17 years. Hypertension, 40 (39.2%), and diabetes mellitus, 32 (31.4%), were the most common risk factors followed by combined hypertension & diabetes mellitus, 20 (19.6%). The most common cause of acute limb ischemia was thrombosis, 77 (75.5%) and followed by embolism 14 (13.7%) and trauma 11 (10.8%). The source of emboli in all cases was cardiac disease due to atrial fibrillation, 9 (7.8%), myocardial infarction, 6 (5.9%) & rheumatic valvular heart disease, 4 (3.9%). The overall Presentation was very late with an average duration of 9±4.8 days and all of them arrive after 24 hours of symptom onset. One hundred eight procedures were done in 102 patients. Type of procedure performed were, thrombectomy 51(47.2%), primary amputation 24(22.2%), bypass or interposition vascular grafts 10(9.2%), embolectomy 10(9.2%), primary vascular repair 7(6.4%), and Femoro-femoral graft 6(5.5%). The 30-day amputation & mortality rate was 52.9% those presented within 3 days. Previous Myocardial infarction was associated with 5 folds increase in mortality (p = 0.036) as compared to those without myocardial infarction. Conclusion This review has tried to show the overall perspective of acute limb ischemia specific to the population we serve. Patients generally presented very late with a significant number of them arrived with irreversible ischemia and tissue loss. Age ≥ 60 years; hypertension, previous myocardial infarction and delayed presentation were associated with poor surgical outcomes. An awareness to create the importance of early arrival, optimizing co-morbidities, timely detection and treating immediately on arrival of the patient, play a key role in improving surgical outcomes of acute limb ischemia.

those presented within 3 days. Previous Myocardial infarction was associated with 5 folds increase in mortality (p = 0.036) as compared to those without myocardial infarction.
Conclusion This review has tried to show the overall perspective of acute limb ischemia specific to the population we serve. Patients generally presented very late with a significant number of them arrived with irreversible ischemia and tissue loss. Age ≥ 60 years; hypertension, previous myocardial infarction and delayed presentation were associated with poor surgical outcomes. An awareness to create the importance of early arrival, optimizing co-morbidities, timely detection and treating immediately on arrival of the patient, play a key role in improving surgical outcomes of acute limb ischemia. Keywords: Acute ischemia, Thrombosis, Embolism, Re-vascularization, Limb amputation Background According to the 2007 Transatlantic Intersociety Consensus (TASC II) for the management of PAD, Acute Limb Ischemia (ALI) is defined as a sudden decrease in limb perfusion that causes a potential threat to limb viability in patients who present within two weeks of acute onset of symptoms [1]. ALI is a medical emergency and it continues to have high morbidity and mortality. Despite urgent revascularization with thrombolytic, endovascular or open vascular surgery, the 30 days amputation and mortality rates range between 10-30% and 5-15% respectively [2,3]. This condition affects between 15-26 persons per 100,000 each year and the associated morbidity and mortality is extremely high, with a yearly mortality rate of 40% in the USA [1,3].
The commonest cause of ALI is PAD (85%), where thromboses formed in atherosclerotic vessels or previously patent artery or graft. Embolus dislodged from distant sources accounts for15% of nontraumatic cases (2,3,4). Owing to lack of a valid survey in LMICs, a population-based estimate of patients with ALI is difficult. However, analysis of combined data from the Global Burden of Diseases program (5) and a systematic review of prevalence surveys (6) can describe the burden of PAD at global and regional levels (7). This population burden can be measured with the Disability Adjusted Life Year (DALY), which is a composite of the years of life lost owing to premature death caused by PAD and the years lived with disability owing to the disease. (5) Accordingly, the rates of DALYs were consistently greater in higher-income than in lower-income regions. However, the rate of growth between 1990 and 2010 was found greater in the lower-income regions, occurring in both men and women. These findings are consistent with those observed on the prevalence of PAD using the ABI, in which a marked increase was found between 2000 and 2010, more so in LMIC. (6) Similarly, analysis made on the Ethiopian national mortality burden during 1990-2015 showed that Ethiopia has successfully reduced premature mortality related to a communicable disease, however, diseases like cardiovascular, diabetes, cancer, and other non-communicable diseases have become the leading causes of death in 2015. (8) This all shows that ALI has begun and will continue to be a disease burden to our community in the coming years. Besides, the disease impact on the occupation and quality of life of the individual patient is substantial, often resulting in loss of independence and/or lively hood. In Ethiopia, surgical services to treat ALI have been started recently within a few hospitals. Our experience in surgically treating ALI is limited; hence in order to significantly improve the surgical outcome and functionality of the limb, a proper understanding of the problem and its course of managements are essential.
Besides, the unique demographic, Ethnic, and geographic factors that exist in our community could influence the disease progress and affect the surgical outcome which may require further investigations. To date, literature review showed that the majority of publications made so far were from developed countries, with very few reports from Africa and none from Ethiopia. Accordingly, we report our experience on the risk factors, etiologies, outcome and associated factors influencing the surgical outcome of ALI treated at Tikur Anbessa Specialized Hospital (TASH) and Teklehaimanot General Hospital.

Materials And Methods
Institutional based prospective cross-sectional study was done on all patients operated (limb amputation or revascularization surgery) for acute limb ischemia from January 1, 2018, to December 31, 2018, at TASH and Teklehaimanot General Hospitals, Addis Ababa, Ethiopia. TASH is the governmental tertiary center and Teklehaimanot General Hospital is a private center. Both of them are located in Addis Ababa, the capital city of Ethiopia and these hospitals were selected as the site of our study because most of the vascular surgery cases in Ethiopia are being managed in these institutions.
The study population included all patients who underwent surgery for acute limb ischemia at TASH and Teklehaimanot General Hospitals over a period of one year. Those patients with PAD without ischemia; chronic or critical limb ischemia, who were managed medically; patient disappeared before the intervention or refused surgery, and who had incomplete data set for analysis were excluded from the current study. The measurable variables in this study were defined by reviewing similar literature previously published in acute limb ischemia. Accordingly, age, sex, risk factors of ALI and duration of presentation were used as independent variables and the dependent variables were surgical outcomes of ALI (limb amputation/salvage, local and systemic complications, mortality).

Data collection, Processing, and Analysis
Six personnel, principal investigator and 5 research assistants who were oriented on this study collected the data. They used a predesigned structured questionnaire that was prepared in English and translated into Amharic and Afan Oromo languages and translated back to English. The questionnaire was filled from the interview and medical records of patients prospectively. After data collection completed, cleaning, checking and editing of the content was done.
Then data was entered into the computer, data were analyzed using SPSS software version 20 and the results were expressed in a ratio, frequency and percentage were calculated using SPSS software version 20. And statistical significance test using logistic regression analysis (bivariate and multivariate) was applied to see the association between independent variables and dependent variables. P-value < 0.05 was accepted significant and variables found to be significant on bivariate analysis were included in multivariate logistic regression analysis. Then the result was discussed by comparing analyzed data with other research done regionally and globally.

Ethical Consideration
This study was approved by and Research & Ethical Committee of Addis Ababa University, College of Health Sciences. The permission was taken from archives of TASH and Teklehaimanot General Hospitals to retrieve a medical record of needed patients. Pre-consent oral counseling of participant and written informed consent was taken and confidentiality was protected at all steps.
The most common risk factors identified were Hypertension, 40 (39.2%), diabetes mellitus, 32 (31.4%), followed by combined hypertension & diabetes mellitus, 20 (19.6%) and smoking in 7 (6.9%). (Table 1) Based on the diagnosis made using combination of clinical, imaging (Doppler U/S and/or CT Angiography) and intraoperative findings, the single most common etiology of ALI was found to be thrombosis, 77 (75.5%) patients and in 46 (45.1%) cases, thrombus occurred on the background of atherosclerosis. The other causes of ALI were embolism, 14 (13.7%) and trauma in 11 (10.8%) cases. The source of emboli was the heart in all patients in which 4 (3.9%) patients had previous ischemic heart disease; 3 (2.9%) patients had atrial fibrillation; 3 (2.9%) patients had rheumatic valvular heart disease with atrial fibrillation. (Table 1)   Our patients were followed for a maximum of 3 months and adherence to follow up was good (80% at 3 months). In our series, all patients were surgically managed with revascularization surgery or amputation. The most commonly performed procedure was thrombectomy in 51 (47.2%) cases and out of these, 8 patients underwent thrombectomy with primary major amputation and 3 patients had thrombectomy with the reversed great saphenous vein bypass graft. Subsequently, on 30th day follow up; re-thrombectomy was done in 6 (11.7%) patients. On the other hand, 24 (22.2%) patients were treated with primary major amputation and embolectomy was performed in 10 (9.2%) patients. interposition graft for complete traumatic vascular injury and 5 (4.6%) reversed GSV bypass graft for acute thrombosis occurred in atherosclerotic vessels. Six (5.5%) extra-anatomic crossover femorofemoral bypass graft was performed for unilateral common & external iliac thrombosis. The overall amputation rate following each procedure is described in Table 3 below.     Table 5). Amputation and mortality rates became 55.9% & 10.8% respectively on 3 months follow up. In this study, few patients developed additional systemic and procedural complications beyond 30 postoperative days (see Table 4). So, a total of 44 (43.2%) procedural and 19 (18.6%) systemic complications occurred in 3 months postoperative period.
On logistic regression bivariate analysis, age more than 60 years was found to be significantly associated with increased rate of limb amputation (p = 0.017; OR = 3.8; 95% CI, 1.26 to 11.11) and have found out that ALI is coming to be one of the major NCDs and patients generally presented very late with a significant number of them arrived with irreversible ischemia and tissue loss. Age ≥ 60 years; hypertension, previous myocardial infarction and delayed presentation were found to be significantly associated with poor surgical outcomes.
In this study, we found a predominance of ALI in males (67.6%), which was almost twice as common as in females (32.4%). Even though this finding is slightly higher, it is a consistent with other studies [20,24,27]. The mean age at presentation was 54 ± 17 years, which was similar to data of other reports. [27,32] Patients within age 40-60 years were commonly affected (39.2%) and followed by those older than 60 years (33.3%). Most authors agree that ischemic conditions of lower extremities are more common than those of upper extremities [11,15,16,20]. In research done in India (Tertiary center) in 2008, it was also shown that acute lower extremity ischemia was more common than acute upper extremity ischemia [27]. In these studies, the forwarded explanation for these findings is an for ALI and in contrary to our findings, they found cigarette smoking as the most predisposing factor for acute limb ischemia compared to these two comorbidities [24,27,30]. In support of this evidence, many studies have shown an association of hypertension, diabetes mellitus and smoking with the development of PAD (atherosclerosis). The relative risk of developing PAD is two to three folds for smoking & hypertension. So currently hypertension guidelines support aggressive treatment of blood pressure in a patient with atherosclerotic occlusive diseases, targeting a blood pressure of 140/90 mmHg or 130/80 mmHg should be targeted in patients with diabetes or renal insufficiencies [22,23,25,26].
In the current study, the main etiologies of acute limb ischemia identified were thrombosis (75.5%) followed by embolism (13.7%) and trauma (10.8%). The source of embolism in all cases was from the heart. Previous myocardial infarction (3.9%), atrial fibrillation (2.9%), rheumatic valvular heart diseases with atrial fibrillation (2.9%) and ischemic heart disease with atrial fibrillation (1.9%) were common cardiac causes of embolism identified. Similar findings were reported from India, West Africa and other areas [2,3,4,27,32]. Unlike our report, a research article done in University of Michigan (USA), Kaunas University of Medicine Hospital (Lithuania) and Aims Shams University (Egypt) reviewed their single institutional experiences and found embolism as more frequent cause of ALI than acute thrombosis. Such discrepancy was seen because of higher number of patients with cardiac diseases in their community and the availability of improved diagnostic and therapeutic modalities for cardiac patients in their community that allowed for better survival than ours where such facilities are scarcely available [24,31,33].
As opposed to trends seen in other series, in our study, all patients with ALI presented after 24 hours duration, which was beyond the accepted critical cutoff time for limb ischemia (golden hour of 4-6 hours) [8]. Unlike most of the reports done in other countries, the average duration of symptoms before arrival to the hospital was 9 ± 4.8 days. Because of this late presentations, our patients presented rest extremity pain (82.5%), limb swelling (63.7%) and darkish discoloration of limb (45.9%). On physical examination, pulseless extremity (71.6%), cold extremity (66.7%) and swollen limb (63.7%) were the most common clinical signs. In our analysis of 102 patients, 45.9% of them had gangrenous limb and more than one-half of patients developed neurologic deficit (sensory, motor or both) at presentation, the figure, which was higher than the report from Iran by M. Mozzafar A. et al [28]. This implicates most of our patients had a significant delay in presentation and had a late stage of limb ischemia which was responsible for high morbidity and mortality in our set up.
On subsequent 3 months follow up, these patients showed different outcomes in terms of limb salvage, local and systemic postoperative complications. Accordingly, amputation rates (secondary amputations) following thrombectomy, vascular grafts (of all types) and embolectomy were 17.6%, 7.8% & 3.9% respectively while 8.8% of patients underwent re-thrombectomy for an indication of native vessel re-thrombosis in 3 months progress. The 30-day amputation rate was 52.9% with a limb salvage rate of 37.1%. Three more patients underwent amputation on subsequent 3 months follow up and overall amputation rate became 55.9% with a limb salvage rate of 33.3%. This figure was two to three folds higher than the results in other studies [27][28][29][30][31][32]. The rate of limb loss following revascularization procedure was 32.4%, this figure was higher than data from other studies [29,30,32].
In our analysis, we observed that primary vascular repair and reversed GSV interposition graft done for the partial or complete vascular injury had better limb salvage rate (8.8%) compared to bypass graft (limb salvage rate of 2.9%) performed for other causes of ALI but the association was not statistically significant. The reason for this was attributed to the relatively early presentation of trauma patients and prompt intervention that might have salvaged the limb in this group of patients.
However, reversed GSV and PTFE bypass grafts done for ALI showed almost comparable results in terms of limb salvage rate. The amputation rate after bypass grafts in our set up was much lower than a trend in another study as described by Donald T. Baril et al [29] even though the comparison was difficult as few patients underwent bypass procedure in our series.
The 30-day mortality rate of ALI in our study was 9.8% and only one other patient died in 3 months follow up making an overall mortality rate of 10.8%. We found acute coronary syndrome (5.9%), overwhelming sepsis of surgical wound focus (2.9%) and acute respiratory insufficiency (0.9%) as common causes of death in our set up. The mortality rate in our series was much lower (two folds) than a report from other authors [31, 32] but comparable with review made in Iran and Lithuania [28,33]. We also observed that acute coronary syndrome (5.9%), acute kidney injury (4.9%) and sepsis of surgical wound focus were common systemic complications while re-thrombosis of native vessels Another important observation of this study was that the effect of independent factors on morbidity and mortality in ALI was evaluated. Accordingly, age older than 60 years was associated with almost three folds increase in limb loss and about nine folds increase in mortality compared to those within 20-40 years of age. Similarly, hypertensive patients showed about three times higher risk of amputation than those without hypertension in our analysis. There were ten folds increase in limb loss in patients having combined hypertension and diabetes compared to those who had neither of them.
Jonathan L.E et al described in their single institutional experience that diabetic patients had the fourfolds risk of amputation compared to non-diabetic patients [24] and it was also observed that smokers have twice the risk of limb loss compared to non-smokers in another study [27]. In contrast to this, there was no significant difference found in diabetic or smoker patients in terms of amputation or mortality rate in our analysis.
Among risk factors of ALI, in addition to older age, ischemic heart disease showed a strong association with mortality rate & had five times the risk of mortality compared to those without cardiac ischemia. Surprisingly, there was no statistically significant independent factor associated with postoperative complications in this series. In our final model of analysis (multivariate), it was observed that the duration of symptoms before presentation above nine days was associated with five folds increase in limb loss rate compared to the first three days.
From this evidence, we can conclude that a much higher rate of limb loss in this study was attributed to delayed presentation of patients with ALI and associated late stage of limb ischemia & this finding was asserted by many other authors [27,28,32,34]. So, we strongly suggest that the early arrival of patients, timely detection and early intervention play a vital role in the success of the management of ALI. We did not investigate the pattern of referral in our study and difficult to give a concrete explanation for the delayed presentation of patients. However, our speculation as reasons for delayed presentation could be financial problems, the remoteness of vascular centers and lack of health awareness of patients, attention, and level of awareness of referring physicians about this disease entity. We hope this will be the main area of study for the subsequent researcher.
This study has some limitations. First, Some patient's data were found to be incomplete and few patients also lost from follows up so they were excluded from our analysis. The gap in this database could result in selection bias. Second, the duration of follow up is only 3 months hence the long-term effect is not known. Third, as this study was conducted at only two institutions, it may be difficult to generalize the results to the national level. Hence, the national-based study may address this issue in the subsequent research.

Conclusion
The causes of acute limb ischemia were thrombosis, embolism or trauma in our institutions and had significantly high morbidity and mortality. Among risk factors of ALI, age older than 60 years, hypertension, combined hypertension & diabetes mellitus, myocardial infarction and delayed presentation were associated with poor surgical outcomes (high rate of amputation and mortality).
Therefore, treating and optimizing comorbidities, the early arrival of the patient, timely detection and intervention play a key role in improving surgical outcomes of acute limb ischemia.
In an attempt to decrease this high rate of morbidity and mortality resulting from ALI, we should strengthen the awareness of referring health professionals at all levels on this particular issue. Finally, we recommend that the role of anticoagulants in changing the outcome of ALI and the pattern of referral of this disease should be the area of focus on subsequent research.