Abstract
Objective:
Liver disease increases mortality after abdominal surgery, including endovascular repair of aortic aneurysms. However, its effect on mortality and morbidity after endovascular and open management of peripheral vascular disease has not been widely evaluated.
Methods:
The National Surgical Quality Improvement Program was used to evaluate patients undergoing infra-inguinal bypass and endovascular intervention (2005 – 2016). Aspartate aminotransferase to platelet ratio (APRI score) is a non-invasive tool recommended by the World Health Organization to identify liver disease and was calculated for all patients. A ratio of >0.5 was used to identify patients with liver fibrosis. Demographics, comorbidities, and 30 day outcomes were evaluated for patients with and without liver fibrosis. A subgroup analysis was completed in patients with APRI scores > 0.5, to evaluate the effect of increasing Model for End-Stage Liver Disease (MELD) scores on outcomes. Multivariable regression was used to account for differences in baseline factors.
Results:
In total, 17 603 patients underwent infra-inguinal bypasses. Fibrosis was associated with higher mortality (3.8% vs 2.4%; p < .001), major complications (23% vs 20%; p = .020), pulmonary (5.1% vs 2.9%; p < .001), and renal complications (1.9% vs 1.1%; p = .007) after bypass. These differences persisted following multivariable adjustment. Altogether, 7 830 patients underwent endovascular intervention. Fibrosis was also associated with higher mortality (4.7% vs 2.2%; p < .001), pulmonary (3.9% vs 2.5%; p = .022), and renal complications (1.9% vs 0.8%; p = .003) after endovascular intervention. After adjustment, only renal complications persisted. In a subgroup analysis of patients with liver fibrosis, morbidity (31% vs 17%; p < .001) and mortality (7.2% vs 1.8%; p < .001) increased after bypass among those with MELD scores > 15 but not after endovascular intervention.
Conclusion:
Liver fibrosis was associated with higher 30 day mortality and major complications after infra-inguinal bypass, with outcomes worsening as MELD scores increased. Surgeons may consider an endovascular first approach in managing peripheral arterial disease among those with liver fibrosis.
Keywords: Bypass, Cirrhosis, Liver fibrosis, Peripheral arterial disease
INTRODUCTION
Chronic liver disease is associated with increased morbidity and mortality after major surgical procedures.1,2 However, its effect on outcomes in vascular patients, particularly those receiving less invasive endovascular interventions, have not been widely quantified. Recently, studies have suggested that despite being the preferred procedure for high risk populations, percutaneous operations such as endovascular repair of aortic aneurysms are associated with significant risks of morbidity and mortality in patients with liver disease.3,4
Peripheral artery disease (PAD) is more common in patients with chronic liver disease than in those without liver disease.5 However, the effect of liver disease on outcomes after the treatment of PAD has not been well defined. In a study of patients undergoing lower extremity bypasses, researchers found an increased risk of morbidity and mortality, with elevated Model of End-Staged Liver Disease (MELD) scores.6 Similar findings have also been shown after carotid endarterectomy.7 However, the effect of liver disease on outcomes after less invasive endovascular procedures is unknown.
The aspartate aminotransferase (AST) to platelet ratio index (APRI) and MELD scores are non-invasive measures used to assess liver disease based on laboratory testing alone. Importantly, the APRI score is recommended by the World Health Organization (WHO) as a screening mechanism to assess for the presence of liver fibrosis and cirrhosis.8 Alternatively, the MELD score is used to evaluate severity in patients with known liver disease. This score is widely used for the stratification of patients undergoing transplant evaluation, as well as pre-operative and mortality risk stratification in patients with cirrhosis.9 Therefore, in the current study the APRI score was used to identify patients with liver fibrosis, in accordance with WHO recommendations. The MELD score was then used to assess for the effect of worsening disease in patients with liver fibrosis, as defined by their APRI score. Both APRI and MELD scores have been used previously for risk assessment in both vascular and general surgical populations.3,10
Given the lack of data, this study aimed to evaluate the effect of liver fibrosis on outcomes in patients undergoing lower extremity bypass and endovascular intervention. Furthermore, this study assessed whether the risks of surgical intervention are altered by disease severity, and determined if a preferred method of treatment exists for patients undergoing surgical management of PAD.
MATERIALS AND METHODS
The American College of Surgeon’s National Surgical Quality Improvement Program (NSQIP) was used to identify all patients undergoing an infra-inguinal bypass or infra-inguinal endovascular intervention from 2005 to 2016. Endovascular procedures included angioplasty, stent, or atherectomy. Patients who underwent an isolated femoral endarterectomy or hybrid procedure were excluded to allow for more accurate and uniform comparison between operative approaches. Bypasses were included if they originated below the inguinal ligament. Patients who underwent emergent procedures (n = 1 742) and those with unknown liver function, for whom APRI and MELD scores could not be calculated, were also excluded from analysis (n = 30 480). All other variables had < 5% missing data. NSQIP is a validated database that prospectively collects data on patient demographics, comorbidities, operative details, and 30 day outcomes from more than 700 academic, community, and military hospitals in North America, Asia, Australia, and Europe. Trained clinical reviewers collect data in a prospective manner and data are independently audited for accuracy and reliability. These efforts have led to confirmation of the quality and effectiveness for accurate research.11 Further information can be found at https://www.facs.org/quality-programs/acs-nsqip.
APRI and MELD score calculation
The APRI score is a low cost screening tool recommended by the WHO for its ability to assess the spectrum of liver disease liver fibrosis to cirrhosis in a non-invasive manner for patients with hepatitis.8 The APRI score is calculated by the ratio of AST to platelets, and is a validated scoring method recommended for the screening of liver disease owing to its ease of calculation and low cost.12,13 Additionally, while initially used to assess hepatitis related liver disease, it has evolved to be used as a marker for both fibrosis and cirrhosis in a wide range of patients, including those undergoing chemotherapy with hepatotoxic agents, those undergoing liver resection, and general surgical patients.10 APRI scores of 0.5 – 1.0 have been used as a cut-off for identifying liver disease, and have been shown to be as effective or superior for identifying liver disease versus other laboratory based studies.13 For the current study, a value of 0.5 was used to identify liver fibrosis. This is consistent with the WHO recommendations and is more sensitive in identifying liver fibrosis than a value of 1.0 or 1.5, which have also been proposed to identify more severe liver cirhosis.8
The MELD score is a commonly used and widely validated metric used for the assessment of disease severity and mortality risk in patients with chronic liver disease. It is calculated using a patient’s international normalised ratio (INR), creatinine, and bilirubin levels, with scores ranging from 6 to 40.1,14 Because the MELD score is not recommended as a screening mechanism owing to false elevations from chronic kidney disease (CKD) or anticoagulation alone, MELD score was used for a subgroup analysis in those with liver fibrosis as identified by an APRI score > 0.5. Groups were divided into those with scores > 15 and ≤ 15 because a MELD of 15 is generally considered of severe enough to be considered for transplantation. Moreover, this cut-off point allowed for comparison to other studies, which used a similar value. This was then applied to further evaluate outcomes in patients with liver disease for both open and endovascular procedures, as well as to assess the effects of worsening liver disease on 30 day outcomes.
Variables
NSQIP defined variables were used for all demographics and comorbidities. Smoking was defined as current tobacco use. Pre-operative transfusion was defined as one or more units of packed red blood cells administered within 72 hours of surgery. All laboratory values used to calculate MELD and APRI scores were collected in the pre-operative period. All outcomes occurred within 30 days of the index operation. A stroke was defined as a haemorrhagic, embolic, or thrombotic event, with symptoms that persisted beyond 24 hours. Renal complications were defined by an increase in creatinine of at least 2 mg/dL from baseline, or new dialysis. Thrombo-embolic complications included pulmonary embolism or deep vein thrombosis. Pulmonary complications included pneumonia, re-intubation, pulmonary embolism, or ventilator use for more than 48 hours. A major complication was defined as one or more of the following complications: myocardial infarction (MI); pulmonary complication; renal complication; stroke; or unplanned return to the operating room.
Statistical analysis
For the current study, APRI score was used to evaluate the effect of liver fibrosis on outcomes after open and endovascular surgery. An adjusted subgroup analysis of those with APRI scores > 0.5 was performed to evaluate the effect of on increasing MELD scores on 30 day outcomes. Univariate analysis was completed using chi square and Fisher exact tests. The Student’s t test and Wilcoxon rank sum test were selected for use based on the normalcy of distribution of continuous data. Multivariable logistic regression was performed to adjust for patient demographic and comorbidities. Variables for inclusion were selected by purposeful selection, which includes those variables found to be significant on univariate analysis, as well as those shown to be predictive of the outcome of interest in prior studies.15 Statistical analysis was performed using the STATA statistical package version 15.1 (StataCorp, College Station, TX, USA). The Institutional Review Board of Beth Israel Deaconess Medical Center approved the current study and waived informed consent owing to the de-identified nature of the NSQIP database.
RESULTS
Patient cohort
In total, 57 655 patients who underwent open or endovascular infrainguinal revascularisation were identified. After excluding emergent procedures and those patients with incomplete laboratory data, 25 433 patients were identified who underwent open or endovascular intervention. This included 17 603 patients who underwent infrainguinal bypass and 7 830 patients who underwent endovascular intervention. Of those patients who underwent an infrainguinal bypass 1 671 (9.5%) had liver fibrosis as identified by an APRI score > 0.5. Of patients who underwent endovascular intervention, 836 (10.7%) had liver fibrosis.
When baseline demographics were compared between patients who underwent an infrainguinal bypass, patients with liver fibrosis were less commonly female (27% vs 37%; p < .001) had lower rates of diabetes (41% vs 46%; p < .001) and hypertension (84% vs 86%; p = .042) but higher rates of congestive heart failure (CHF; 4.5% vs 3.4%; p =.023) (Table 1). Additionally, while creatinine was similar, other laboratory values such as INR (1.1 vs 1.0; p < .001), bilirubin (0.6 vs 0.5; p < .001), platelets (165 vs 247; p < .001), and AST (46 vs 20; p < .001) differed.
Table 1.
Baseline and pre-operative characteristics in patients with and without liver fibrosis (aspartate aminotransferase to platelet ratio [APRI] > 0.5) undergoing infra-inguinal bypass for peripheral arterial disease.
| No liver fibrosis APRI < 0.5 (n = 15 932) | Liver fibrosis APRI > 0.5 (n = 1 671) | p value | |
|---|---|---|---|
| Age - y | 67.73 ± 11 | 67.38 ± 11 | .22 |
| Female sex | 5 885 (36.9) | 447 (26.8) | < .001 |
| White | 10 365 (77) | 1 073 (76) | .42 |
| BMI | .97 | ||
| Underweight | 568 (3.6) | 63 (3.8) | |
| Normal | 5 205 (32.7) | 556 (33.3) | |
| Overweight | 5 414 (34.0) | 561 (33.6) | |
| Obese | 4 014 (25.2) | 417 (25.0) | |
| Morbidly obese | 731 (4.6) | 74 (4.4) | |
| Current smoking | 6 856 (43.0) | 715 (42.8) | .85 |
| Diabetes | 6 895 (46) | 646 (41) | < .001 |
| Hypertension | 13 660 (85.7) | 1 402 (83.9) | .042 |
| COPD | 2 232 (14.0) | 246 (14.7) | .43 |
| CHF | 544 (3.4) | 75 (4.5) | .023 |
| eGFR - mL/min/1.73m2 | < .001 | ||
| >60 | 9 206 (57.8) | 1 016 (60.8) | |
| 30–60 | 4 673 (29.3) | 410 (24.5) | |
| <30 | 2 006 (12.6) | 243 (14.5) | |
| Pre-operative transfusion | 283 (1.8) | 48 (2.9) | .002 |
| Operative time - min | 212 (156–285) | 221 (162–289) | .013 |
| INR | 1.0 (1.0–1.1) | 1.1 (1.0–1.2) | < .001 |
| Creatinine - mg/dL | 1.0 (0.8–1.4) | 1.0 (0.8–1.4) | .49 |
| Bilirubin - mg/dL | 0.5 (0.3–0.6) | 0.6 (0.4–0.9) | < .001 |
| Platelets - ×109/L | 247 (201–309) | 165 (123–208) | < .001 |
| AST - U/L | 20 (16–25) | 46 (35–68) | < .001 |
Data are presented as n (%), mean ± standard deviation, or median (interquartile range). BMI = body mass index; COPD = chronic obstructive pulmonary disease; CHF = congestive heart failure; eGFR = estimated glomerular filtration rate; INR = international normalised ratio; AST = aspartate aminotransferase.
Of patients who underwent endovascular intervention, patients with liver fibrosis were also less likely to be female (32% vs 42%; p < .001) and had higher rates of CHF (8.1% vs 4.4%; p < .001) (Table 2). Laboratory values also differed including creatinine (1.2 vs 1.1; p < .001), INR (1.2 vs 1.1; p < .001), bilirubin (0.6 vs 0.4; p < .001), platelet (157 vs 242; p < .001), and AST (45 vs 20; p < .001).
Table 2.
Baseline and pre-operative characteristics of patients with and without liver fibrosis (aspartate aminotransferase to platelet ratio [APRI] > 0.5) undergoing infra-inguinal endovascular intervention for peripheral arterial disease.
| No liver fibrosis APRI < 0.5 (n = 6 994) | Liver fibrosis APRI > 0.5 (n = 836) | p value | |
|---|---|---|---|
| Age - y | 69 ± 11 | 69 ± 11 | .94 |
| Female sex | 2 926 (41.8) | 265 (32) | < .001 |
| White | 4 854 (76) | 559 (73) | .072 |
| BMI | .11 | ||
| Underweight | 206 (2.9) | 21 (2.5) | |
| Normal | 2 083 (29.8) | 271 (32.4) | |
| Overweight | 2 352 (33.6) | 299 (35.8) | |
| Obese | 1 959 (28.0) | 206 (24.6) | |
| Morbidly obese | 394 (5.6) | 39 (4.7) | |
| Current smoking | 2 148 (30.7) | 251 (30.0) | .68 |
| Diabetes | 3 763 (53.8) | 471 (56.3) | .19 |
| Hypertension | 6 018 (86.0) | 726 (86.8) | .53 |
| COPD | 898 (12.8) | 116 (13.9) | .40 |
| CHF | 306 (4.4) | 68 (8.1) | < .001 |
| eGFR - mL/min/1.73m2 | < .001 | ||
| >60 | 3 611 (51.6) | 403 (48.2) | |
| 30–60 | 2 183 (31.2) | 244 (29.2) | |
| <30 | 1 186 (17.0) | 187 (22.4) | |
| Pre-operative transfusion | 127 (1.8) | 19 (2.3) | .36 |
| Operative time - min | 96 (65–145) | 97 (68–149) | .32 |
| INR | 1.0 (1.0–1.2) | 1.1 (1.0–1.3) | < .001 |
| Creatinine - mg/dL | 1.1 (0.9–1.5) | 1.2 (0.9–1.8) | < .001 |
| Bilirubin - mg/dL | 0.4 (0.3–0.6) | 0.6 (0.4–0.9) | < .001 |
| Platelets - ×109/L | 242 (195–301) | 157 (117–205) | < .001 |
| AST - U/L | 20 (16–25) | 45 (34–67) | < .001 |
Data are presented as n (%), mean ± standard deviation, or median (interquartile range). MBI = body mass index; COPD = chronic obstructive pulmonary disease; CHF = congestive heart failure; eGFR = estimated glomerular filtration rate; INR = international normalised ratio; AST = aspartate aminotransferase.
Thirty day outcomes
Of patients who underwent an infra-inguinal bypass, those with liver fibrosis (APRI > 0.5) had higher 30 day mortality (3.8% vs 2.4%; p < .001), major complications (23% vs 20%; p = .020), pulmonary complications (5.1% vs 2.9%; p < .001), and renal complications (1.9% vs 1.1%; p = .007) (Table 3). These results persisted after multivariable analysis (Table 4).
Table 3.
Outcomes after infra-inguinal bypass among patients with and without liver fibrosis (aspartate aminotransferase to platelet ratio [APRI] > 0.5).
| No liver fibrosis APRI < 0.5 (n = 15 932) | Liver fibrosis APRI > 0.5 (n = 1 671) | p value | |
|---|---|---|---|
| 30 day mortality | 377 (2.4) | 63 (3.8) | < .001 |
| Major complication | 3 202 (20.1) | 376 (22.5) | .020 |
| Return to OR | 2 586 (16.2) | 274 (16.4) | .86 |
| Pulmonary complication | 469 (2.9) | 86 (5.1) | .010 |
| Renal complication | 169 (1.1) | 31 (1.9) | .007 |
| MI | 322 (2.0) | 41 (2.5) | .24 |
| Ventilation > 48 h | 176 (1.1) | 38 (2.3) | < .001 |
| Re-intubation | 296 (1.9) | 51 (3.1) | .002 |
| Pneumonia | 156 (1.0) | 20 (1.2) | .37 |
| Thrombo-embolic event | 147 (0.9) | 16 (1.0) | .89 |
| Stroke | 98 (0.6) | 10 (0.6) | 1.0 |
| Wound infection | 379 (2.4) | 52 (3.1) | .067 |
| Transfusion | 3 034 (10) | 346 (20.7) | .10 |
| Not discharged home | 2 946 (29) | 340 (32) | .061 |
Data are presented as n (%). OR = operating room; MI = myocardial infarction.
Table 4.
Odds of complication among patients with and without liver fibrosis (aspartate aminotransferase to platelet ratio > 0.5) after endovascular intervention and infra-inguinal bypass for peripheral arterial disease*
| Endovascular | Bypass | |
|---|---|---|
| 30 day mortality | 1.8 (0.9–3.4) | 3.2 (2.1–4.8) |
| Major complication | 1.4 (0.9–1.9) | 1.4 (1.1–1.8) |
| Pulmonary complication | 1.6 (0.8–3.0) | 2.5 (1.7–3.8) |
| Renal complication | 3.5 (1.6–7.6) | 3.6 (2.1–6.1) |
| Ventilation > 48 h | - | 4.1 (2.5–6.8) |
| Re-intubation | - | 2.9 (1.8–4.5) |
| MI | 0.7 (0.2–2.4) | 1.8 (1.2–2.8) |
Data are presented as odds ratio (95% confidence interval). MI = myocardial infarction.
Analysis adjusted for age, sex, diabetes, glomerular filtration rate, and congestive heart failure.
Of those undergoing endovascular treatment, patients with liver fibrosis had higher 30 day mortality (4.7% vs 2.2%; p < .001), major complications (18% vs 14%; p < .001), pulmonary complications (3.9% vs 2.5%; p = .022), renal complications (1.9% vs 0.8%; p = .003), and MI (2.6% vs 1.5%; p = .020) (Table 5). However, following adjustment only renal complications persisted (odds ratio [OR] 3.5, 95% confidence interval [CI] 1.6 – 7.6) (Table 4). When further factors associated with renal complications were evaluated, pre-existing CHF (OR 3.7, 95% CI 2.0 – 6.9) and decreased pre-operative glomerular filtration rate (GFR) (GFR 30 – 60 mL/min/1.73 m2 OR 1.9 [95% CI 1.1 – 3.4]; GFR < 30 mL/min/1.73 m2 OR 3.0 [95% CI 1.6 – 5.4]) were also predictive of renal complications after endovascular intervention.
Table 5.
Outcomes after infrainguinal endovascular intervention for peripheral arterial disease in patients with and without liver fibrosis (aspartate aminotransferase to platelet ratio [APRI] > 0.5).
| No liver fibrosis APRI < 0.5 (n = 6 994) | Liver fibrosis APRI > 0.5 (n = 836) | p value | |
|---|---|---|---|
| 30 day mortality | 154 (2.2) | 39 (4.7) | < .001 |
| Major complication | 960 (13.7) | 153 (18.3) | < .001 |
| Return to OR | 757 (10.8) | 106 (12.7) | .11 |
| Pulmonary complication | 175 (2.5) | 33 (3.9) | .022 |
| Renal complication | 55 (0.8) | 16 (1.9) | .003 |
| MI | 105 (1.5) | 22 (2.6) | .020 |
| Pneumonia | 94 (1.3) | 15 (1.8) | .28 |
| Thrombo-embolic event | 35 (0.5) | 5 (0.6) | .61 |
| Stroke | 23 (0.3) | 4 (0.5) | .52 |
| Wound infection | 46 (0.7) | 6 (0.7) | .82 |
| Transfusion | 536 (7.7) | 66 (7.9) | .81 |
| Not discharged home | 1 059 (16) | 121 (14.5) | .57 |
Data are presented as n (%). OR = operating room; MI = myocardial infarction.
Next, a subgroup analysis was performed to evaluate the effect of increasing MELD score on outcomes. Of patients with liver fibrosis on screening (APRI > 0.5), patients with MELD scores > 15 were compared to those ≤ 15. Of patients undergoing infra-inguinal bypass, 30 day mortality was increased (7.2% vs 1.8%; p < .001). Additionally, major complications (31% vs 17%; p < .001), pulmonary complications (9.2% vs 3.0%; p < .001), renal complications (3.2% vs 1.4%; p = .032), MI (4.0% vs 1.4%; p = .003), return to the operating room (22% vs 13%; p < .001), and discharge to a facility (43% vs 29%; p < .001) were all more common in those with a MELD score > 15 (Table 6). Following adjustment, 30 day mortality (OR 3.1, 95% CI 1.4 – 6.9), major complications (OR 1.6, 95% CI 1.1 – 2.3), pulmonary complications (OR 2.5, 95% CI 1.3 – 4.8), intubation beyond 48 hours (OR 4.1, 95% CI 2.5 – 6.8), and discharge to a nursing facility (OR 2.1, 95% CI 1.2 – 3.8) remained significantly higher in patients with increasing MELD scores (Table S1; see Supplementary Material).
Table 6.
Outcomes after infra-inguinal bypass in patients with increasing Model for End-Stage Liver Disease (MELD) scores.
| MELD ≤ 15 (n = 855) | MELD > 15 (n = 531) | p value | |
|---|---|---|---|
| 30 day mortality | 15 (1.8) | 38 (7.2) | < .001 |
| Major complication | 146 (17.1) | 166 (31.3) | < .001 |
| Return to OR | 108 (12.6) | 114 (21.5) | < .001 |
| Pulmonary complication | 26 (3.0) | 49 (9.2) | < .001 |
| Renal complication | 12 (1.4) | 17 (3.2) | .032 |
| MI | 12 (1.4) | 21 (4.0) | .003 |
| Ventilation > 48 h | 11 (1.3) | 23 (4.3) | < .001 |
| Re-intubation | 15 (1.8) | 31 (5.8) | < .001 |
| Pneumonia | 6 (0.7) | 11 (2.1) | .041 |
| Thrombo-embolic event | 8 (0.9) | 6 (1.1) | .79 |
| Stroke | 5 (0.6) | 3 (0.6) | 1.0 |
| Wound infection | 27 (3.2) | 17 (3.2) | 1.0 |
| Transfusion | 154 (18.0) | 144 (27.1) | < .001 |
| Not discharged home | 161 (29) | 137 (43) | < .001 |
Data are presented as n (%). OR = operating room; MI = myocardial infarction.
Of patients undergoing endovascular intervention, those with a MELD score > 15 had higher 30 day mortality (7.1% vs 2.0%; p = .003), major complications (24% vs 14%; p = .002), return to the operating room (17% vs 9.5%; p = .005), and discharge to a facility (21% vs 8.1%; p < .001) (Table 7). However, after adjustment only discharge to a nursing facility (OR 2.3, 95% CI 1.2 – 4.3) persisted (Table S1).
Table 7.
Outcomes after infra-inguinal endovascular intervention for peripheral arterial disease in patients with increasing Model for End-Stage Liver Disease (MELD) scores.
| MELD ≤ 15 (n = 306) | MELD > 15 (n = 353) | p value | |
|---|---|---|---|
| 30 day mortality | 6 (2.0) | 25 (7.1) | .003 |
| Major complication | 44 (14.4) | 84 (23.8) | .002 |
| Return to OR | 29 (9.5) | 60 (17.0) | .005 |
| Pulmonary complication | 9 (2.9) | 19 (5.4) | .17 |
| Renal complication | 5 (1.6) | 10 (2.8) | .43 |
| MI | 8 (2.6) | 11 (3.1) | .82 |
| Pneumonia | 5 (1.6) | 9 (2.5) | .59 |
| Thrombo-embolic event | 1 (0.3) | 3 (0.8) | .63 |
| Stroke | 1 (0.3) | 2 (0.6) | 1.0 |
| Wound infection | 1 (0.3) | 4 (1.1) | .38 |
| Transfusion | 25 (8.2) | 33 (9.3) | .59 |
| Not discharged home | 24 (8.1) | 71 (21) | < .001 |
Data are presented as n (%). OR = operating room; MI = myocardial infarction.
Finally, multivariable analysis was performed comparing open and endovascular treatment. In this adjusted analysis, open surgery was associated with increased 30 day mortality (OR 2.7, 95% CI 1.2 – 6.1), major complication (OR 1.9, 95% CI 1.2 – 3.0), and pulmonary complications (OR 3.1, 95% CI 1.3 – 7.4) when compared to endovascular intervention. After adjustment for demographics and comorbidities, there were no differences in renal complications of postoperative MI.
DISCUSSION
This study found that liver fibrosis, as defined by an APRI score > 0.5, is associated with increased 30 day morbidity and mortality in patients undergoing infra-inguinal bypass. However, endovascular intervention does not increase these risks in patients with liver fibrosis. Moreover, the risk of adverse outcomes increases with worsening MELD scores in patients with liver fibrosis.
Few studies have evaluated the effect of liver disease on outcomes in patients with PAD. In a study of approximately 6 000 patients who underwent an infra-inguinal bypass, Krafcik et al. concluded that an elevated MELD score increases the risk of morbidity and mortality; however, they did not evaluate endovascular intervention. The authors found mortality rates of 2.9% overall, including a rate of 4.5% in those with MELD scores > 15. This led them to conclude that while rates of morbidity and mortality are elevated, these rates may be acceptable. This result differs significantly from the current study, which identified a mortality rate of 7.5% in those with MELD scores > 15. This substantial difference between results is likely due to differences in methodology. In the current study MELD scores were only used for patients with documented liver fibrosis (APRI > 0.5), while Krafcik et al. used MELD scores for all patients. Importantly, compared to the APRI score, which is indicated for screening, MELD scores are not intended for the screening of liver disease rather for the assessment of disease severity and prognosis of patient with known liver disease. This likely led to their results reflecting a substantial number of patients who did not have liver fibrosis, but rather had elevated MELD scores due to creatinine elevations from CKD or elevated INR due to anticoagulation, and may have improved outcomes as compared to those with known liver disease and elevated MELD scores. A key example of this can be seen in patients on anticoagulation for atrial fibrillation or a mechanical heart valve with an INR of 3.0 but normal creatinine and bilirubin. This patient would have a high MELD score of 20, but would be not have the same high three month mortality rate of 20% seen in patients with liver disease who have a MELD score of 20.
This study provides important data to guide clinical practice. A large study of more than 57 000 patients found that patients with chronic liver disease have higher rates of PAD, as well as acute coronary disease than those without chronic liver disease owing to the increase risk for endothelial injury and decreased arterial compliance found in cirrhosis.5 However, despite this finding, the risks of intervention and optimal treatment approaches for these patients have not been well defined. Our study suggests that outcomes in patients with liver fibrosis are significantly worse in patients with PAD after infra-inguinal bypass; however, endovascular intervention does not carry the same risk. Moreover, the risk of infra-inguinal bypass increases with MELD scores > 15. This suggests that an endovascular approach to patients with liver fibrosis in all patients (including those with MELD scores > 15), as defined by an APRI score > 0.5, may be warranted to improve patient safety. Moreover, those patients with liver fibrosis would benefit from the assessment of their MELD score for further risk stratification if the decision is made to proceed with an infra-inguinal bypass. In those with MELD scores > 15, the risk of operative mortality should be strongly weighed against the risk of no intervention with consideration of an endovascular only intervention.
While endovascular intervention carries a lower morbidity and mortality than infra-inguinal bypass, it should be mentioned that in those with liver fibrosis even endovascular treatment carries a risk of renal complications, particularly in those with baseline congestive heart failure or CKD. Previous studies have found patients with cirrhosis to be at increased risk of acute kidney injury (AKI), compared to those without cirrhosis, and the contrast used for endovascular intervention may increase this risk; however, this cannot be determined from the data currently available in NSQIP. This renal complication is not insignificant, as previous studies have found AKI to be associated with increased mortality in patients with cirrhosis.16,17 While our current study did not identify increased mortality associated with endovascular intervention, these data suggest that endovascular intervention may be best reserved for those patients with severe limb threat. Moreover, surgeons should have significant consideration of need for intervention in those with additional predictors of AKI, including CHF and CKD.
Limitations
There are several limitations to this study, which should be addressed. This is a retrospective analysis of a large multi-institutional database and is therefore subject to the limitations of all large database studies, including errors in coding, missing data, the availability of variables provided, lack randomisation, and selection bias. NSQIP, however, is not an administrative database, but rather a quality improvement registry that uses trained nurse abstractors to obtain variable data directly from the clinic documentation and patient contact, and is widely audited for accuracy. Additional challenges of large observational studies include the potential for findings to be explained by factors not captured in the dataset. In this study these variables included an inability to identify the cause of liver disease, which varies by region or the progression of liver disease in the postoperative period. Other data points, which were not available in NSQIP include contrast dose, lesion complexity/extent of reconstruction, anaesthetic agent, prior infections, and indication for intervention. Finally, outcomes beyond 30 days, including cause of death, could not be assessed in the current study.
Conclusions
Liver fibrosis is associated with significantly higher 30 day mortality and major complications after infra-inguinal bypass. While the endovascular approach is associated with increased risk of AKI, it does not appear to increase other morbidity or mortality. Moreover, in patients with liver disease the risk of adverse events increases with worsening MELD scores. Given the high rates of adverse events, surgeons may consider an endovascular first approach for the treatment of PAD in suitable patients with liver fibrosis.
Supplementary Material
WHAT THIS PAPER ADDS.
Liver disease has been shown to increase morbidity and mortality after open and endovascular aortic surgery, with several studies suggesting an increased size threshold for repair to adjust for these risks. However, the effects of liver disease on open and endovascular treatment of peripheral arterial disease and the optimal treatment modalities have not been established. This paper assesses the morbidity and mortality of liver disease and its severity on outcomes after infrainguinal bypass and endovascular intervention and found that an endovascular first approach may be optimal to reduce morbidity and mortality in this patient population.
FUNDING
This work was supported by the Harvard-Longwood Research Training in Vascular Surgery National Institutes of Health T32 Grant 5T32HL007734-22.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
CONFLICTS OF INTEREST
M.L.S. is a consultant for Abbott, Medtronic and Cook Medical. All other authors have no conflicts.
REFERENCES
- 1.Havens JM, Columbus AB, Olufajo OA, Askari R, Salim A, Christopher KB. Association of model for end-stage liver disease score with mortality in emergency general surgery patients. JAMA Surg 2016;151:e160789. [DOI] [PubMed] [Google Scholar]
- 2.Millwala F, Nguyen GC, Thuluvath PJ. Outcomes of patients with cirrhosis undergoing non-hepatic surgery: risk assessment and management. World J Gastroenterol 2007;13:4056–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Zettervall SL, Dansey K, Swerdlow NJ, Soden P, Evenson A, Schermerhorn ML. Aspartate transaminase to platelet ratio index and Model for End-Stage Liver Disease scores are associated with morbidity and mortality after endovascular aneurysm repair among patients with liver dysfunction. J Vasc Surg 2020;72:904–9. [DOI] [PubMed] [Google Scholar]
- 4.Chou AH, Chen CC, Lin YS, Lin MS, Wu VC, Ting PC, et al. A population-based analysis of endovascular aortic stent graft therapy in patients with liver cirrhosis. J Vasc Surg 2018;69:1395–404. [DOI] [PubMed] [Google Scholar]
- 5.Lin SY, Lin CL, Lin CC, Wang IK, Hsu WH, Kao CH. Risk of acute coronary syndrome and peripheral arterial disease in chronic liver disease and cirrhosis: a nationwide population-based study. Atherosclerosis 2018;270:154–9. [DOI] [PubMed] [Google Scholar]
- 6.Krafcik BM, Farber A, Eslami MH, Kalish JA, Rybin D, Doros G, et al. The role of Model for End-Stage Liver Disease (MELD) score in predicting outcomes for lower extremity bypass. J Vasc Surg 2016;64:124–30. [DOI] [PubMed] [Google Scholar]
- 7.Krafcik BM, Farber A, Eslami MH, Kalish JA, Rybin D, Doros G, et al. The role of the Model of End-Stage Liver Disease score in predicting outcomes of carotid endarterectomy. Vasc Endovascular Surg 2016;50:380–4. [DOI] [PubMed] [Google Scholar]
- 8.World Health Organization. Guidelines for the Prevention, care and treatment of persons with chronic hepatitis B infection. Available at: https://www.who.int/hepatitis/publications/hepatitis-b-guidelines/en/ [Accessed 10 February 2020]. [PubMed]
- 9.Murray KF, Carithers RL, Jr. AASLD practice guidelines: evaluation of the patient for liver transplantation. Hepatology 2005;41:1407–32. [DOI] [PubMed] [Google Scholar]
- 10.Pereyra D, Rumpf B, Ammann M, Perrodin SF, Tamandl D, Haselmann C, et al. The combination of APRI and ALBI facilitates preoperative risk stratification for patients undergoing liver surgery after neoadjuvant chemotherapy. Ann Surg Oncol 2019;26:791–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Shiloach M, Frencher SK Jr., Steeger JE, Rowell KS, Bartzokis K, Tomeh MG, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg 2010;210:6–16. [DOI] [PubMed] [Google Scholar]
- 12.Xiao G, Yang J, Yan L. Comparison of diagnostic accuracy of aspartate aminotransferase to platelet ratio index and fibrosis-4 index for detecting liver fibrosis in adult patients with chronic hepatitis B virus infection: a systemic review and meta-analysis. Hepatology 2015;61:292–302. [DOI] [PubMed] [Google Scholar]
- 13.Lin ZH, Xin YN, Dong QJ, Wang Q, Jiang XJ, Zhan SH, et al. Performance of the aspartate aminotransferase-to-platelet ratio index for the staging of hepatitis C-related fibrosis: an updated meta-analysis. Hepatology 2011;53:726–36. [DOI] [PubMed] [Google Scholar]
- 14.Roth JA, Chrobak C, Schadelin S, Hug BL. MELD score as a predictor of mortality, length of hospital stay, and disease burden: A single-center retrospective study in 39,323 inpatients. Medicine (Baltimore) 2017;96:e7155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Bursac Z, Gauss CH, Williams DK, Hosmer DW. Purposeful selection of variables in logistic regression. Source Code Biol Med 2008;3:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Tandon P, Garcia-Tsao G. Renal dysfunction is the most important independent predictor of mortality in cirrhotic patients with spontaneous bacterial peritonitis. Clin Gastroenterol Hepatol 2011;9:260–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fagundes C, Barreto R, Guevara M, Garcia E, Sola E, Rodriguez E, et al. A modified acute kidney injury classification for diagnosis and risk stratification of impairment of kidney function in cirrhosis. J Hepatol 2013;59:474–81. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.