Acute kidney injury in peripheral arterial surgery patients: a cohort study

Abstract

Purpose: The aim of the study was to evaluate the prevalence of acute kidney injury (AKI) in a cohort of surgically treated patients with peripheral artery disease (PAD) and its association with the short-term and long-term outcome. Materials and methods: We conducted a retrospective cohort study on all the consecutive PAD inpatients in 2008. Data on the patients’ demographic characteristics, medical history, treatment, outcome and laboratory tests measurements were retrieved from the medical records. Results: We analyzed 166 patients (71.6% males, mean age 63.2 years ± SD 10.7 years) and found an AKI prevalence of 12.7%. The AKI patients’ group had more chronic kidney disease (CKD) (23.8% vs. 6.2%, p = 0.005), diabetes mellitus (DM) (61.9% vs. 33.1%, p = 0.011) and a higher length of hospital stay (19.48 vs. 15.42 days, p = 0.047). At one year, the mortality was 33.3% in the AKI group compared to 1.3% in non-AKI group, with a strong association between AKI and death (OR = 35.7; 95%CI = 6.7 to 189) and AKI and major cardiovascular events (OR = 29.1; 95% CI = 6.8 to 123.4). There was no significant difference in terms of age, cardiovascular disease and medication between the two groups. Conclusions: AKI was associated with a poorer one-year outcome after the surgery of PAD patients. In our study, the presence of previous chronic kidney disease and type 2 diabetes increased the incidence of acute kidney injury after surgery.

• Acute kidney injury
• peripheral artery disease
• outcome

Introduction

Previously known as acute renal failure (ARF) and quantified by the RIFLE criteria (Risk, Injury, Failure, Loss and End-stage) according to the ADQI Group Consensus, acute kidney injury (AKI) represents the latest concept adopted to ensure early diagnosis of acute kidney impairment, based on early detection of kidney damage.1

AKI is a complication that worsens the short-term and long-term prognosis of cardiac and vascular surgery patients; however, its effect on peripheral arterial surgery patient’s prognosis was not extensively analyzed.2–4 AKI is also a severe complication in general surgery, increasing mortality, morbidity and hospital costs.5,6

Studies evaluating kidney function in abdominal aorta surgery demonstrated both glomerular (albuminuria) and tubular (increased urinary NAG-creatinine ratio) injury in up to 96% of the subjects, although only 22% had RIFLE-defined acute kidney injury, and a worse short-term (30 days) and long-term (10 years) survival of the AAA (abdominal aortic aneurysm) surgery patients who experienced a temporary AKI, even if their renal function completely recovered.7,8

The aim of our study was to evaluate the prevalence of AKI, its association with the short-term (within 30-days postoperatively) and long-term (1 year postoperatively) prognosis in a cohort of patients with severe peripheral artery disease (PAD) undergoing surgical procedures of lower extremity revascularization or amputation and the most important comorbidities associated with AKI development.

Methods

Study subjects

We conducted a retrospective cohort study on 166 consecutive PAD inpatients admitted to the Vascular Surgery Ward of the Emergency County Hospital from Timisoara, Romania, from January to December 2008, scheduled for surgical lower extremity revascularization or lower limb amputation. There were no investigations involving contrast substances carried out on the study population during hospital stay.

Data on the patients’ demographic characteristics, medical history, medical treatment and outcome (post-surgical and 1 year after the surgery) were retrieved from the medical records (hospital or general practitioner) and from the surgeon caring for the patient.

This study was reviewed and approved by the ethics committee of the “Victor Babes” University of Medicine and Pharmacy, Timisoara.

All patients underwent either lower limb amputations (at any level) or lower extremity revascularization (aortofemoral or femoropopliteal bypass). The need for surgical re-intervention was noted, but the patient was included as a study subject only at the first surgical procedure.

The one-year survival and the presence of major cardiovascular events (non-fatal myocardial infarction and non-fatal stroke) one year after the surgery were verified for each patient by checking with the surgeon or the general practitioner caring for the patient.

Studied parameters

AKI was diagnosed on admission or during hospital stay, according to AKIN diagnosis criteria: creatinine increase to ≥0.3 mg/dL or to 1.5 to 2-fold from baseline in 48 hours (stage 1); to >2 to 3-fold from baseline (stage 2); to >3-fold from baseline or more than 4 mg/dL with an acute rise of at least 0.5 mg/dL or on renal replacement therapy in less than a week (stage 3), or on the reduction of urinary output: <0.5 mL/kg/h for more than 6 hours (stage 1), <0.5 mL/kg/h for more than 12 hours (stage 2), <0.3 mL/kg/h for 24 hours or anuria for 12 hours (stage 3).9

PAD was diagnosed previously to the hospital admission by computer tomography angiography and was classified into four stages, accordingly: Stage I—Asymptomatic (patients who are for the most part asymptomatic; non-specific symptoms such as paresthesias and cold extremities; lack of normal pulses); Stage II—Intermittent claudication (Stage IIa: Intermittent claudication after more than 200 meters of pain free walking; Stage IIb—Intermittent claudication after less than 200 meters of walking); Stage III—Rest pain; Stave IV—Ischemic ulcers or gangrene.10

Previously diagnosed hypertension (considered as systolic blood pressure higher than 140 mmHg and diastolic blood pressure higher than 90 mmHg), diabetes mellitus (type 1 or 2), dyslipidemia (cholesterol higher than 200 mg/dL or triglycerides level higher than 150 mg/dL), coronary artery disease (defined as angina pectoris, myocardial infarction or silent myocardial ischemia), stroke (defined as brain ischemia or hemorrhage) and chronic kidney disease (CKD, diagnosed according to K/DOQI 2002) were recorded.11 Obesity, diagnosed as body-mass index >30 kg/m² and current smoking status were also assessed.

Blood pressure of inpatients measured one day before surgery (systolic and diastolic values measured using a mechanical sphygmomanometer with aneroid manometer) and some of the blood chemistries (biological characteristics of the patients) performed during hospital stay (serum creatinine and fasting blood glucose one day before surgery and serum creatinine measured again one day after the surgery for those with impaired renal function) were also collected. Estimated GFR (glomerular filtration rate) was calculated using abbreviated MDRD equation.11,12

Regular use of cardiovascular medication (at the time of the surgical procedure) including angiotensin-converting enzyme inhibitors (ACEI), angiotensin II receptor blocker (ARB), calcium channel blockers (CCB) and statins (HMG-CoA reductase inhibitors) was also recorded.

Statistical analysis

Data were analyzed using SPSS version 14 (SPSS, Chicago, IL) and are presented as mean ± standard deviation or percentages. To assess the significances of the differences t-student test was used for means and chi-square for percentages. Continuous variables distributions were tested for normality using D’Agostino-Pearson test and for equality of variances with Levene’s test. For assessing the involvement of multiple confounding factors in dichotomous outcomes, multivariate logistic regression models were built, their goodness of fit being presented using Nagelkerke’s generalized R². A p-value <0.05 was considered the threshold for statistical significance.

Results

Patient’s demographical, clinical and biological characteristics are shown in Table 1. In our studied cohort, the incidence of AKI after surgical intervention for limb ischemia was significantly higher in patients previously diagnosed with type 2 diabetes mellitus (21.3% vs. 7.7% in non-diabetics, OR = 3.25, p = 0.011) and previous chronic kidney disease (35.7% vs. 9.9%, OR = 5.04, p = 0.005). Patients which developed AKI had higher mean SBP with 7.8 mmHg (p = 0.043). No statistically significant associations between the incidence of AKI and gender (p = 0.98) smoking status (p = 0.53), presence of obesity (p = 0.45), presence of hypertension (p = 0.49), use of ACEI/ARB (p = 0.7), use of CCB (p = 0.73) or use of statins (p = 0.52) were found. Using age- and sex-adjusted multivariate logistic regression (Nagelkerke R² = 0.248), type 2 diabetes (p = 0.038) and chronic kidney disease (p = 0.047) were confirmed as statistically significant cofactors for AKI having in the same time no significant involvement from other analyzed confounders (hypertension, smoking or obesity).

Table 1. Studied group main characteristics.

Parameter	Entire group (n = 166)	Without AKI (n = 145)	AKI present (n = 21)	p (AKI vs. without AKI)
Age (years)	63.2 ± 10.8	63.5 ± 10.9	61.5 ± 9.3	0.45
Males (%)	71.6	71.7	71.4	0.98
Current smokers (%)	63.3	64.1	57.1	0.53
Obese (%)	15.1	15.9	9.5	0.45
eGFR before surgery (mL/min/1.73m^2)	75.7 ± 27.5	76.5 ± 26.9	70.1 ± 31.8	0.32
eGFR before discharge (mL/min/1.73m^2)	75.8 ± 38.1	82.6 ± 28.7	67.8 ± 46.4	0.21
SBP (mmHg)	133.9 ± 16.5	132.9 ± 16.6	140.7 ± 14.3	0.043*
DBP (mmHg)	74.4 ± 8.7	74.3 ± 8.9	75.5 ± 7.2	0.55
Fasting blood glucose (mg/dL)	139.7 ± 63.5	138 ± 62.6	150.8 ± 69.9	0.39
CKD (%)	8.43	6.2	23.8	0.005*
Type 2 Diabetes (%)	36.7	33.1	61.9	0.011*
Hypertension (%)	72.9	73.8	66.7	0.49
Dyslipidemia (%)	34.9	35.2	33.3	0.87
Coronary artery disease (%)	47	48.3	38.1	0.38
History of stroke (%)	9	8.3	14.3	0.37
Use of ACEI/ARB (%)	42.2	41.4	47.6	0.7
Use of CCB (%)	30.1	29.7	33.3	0.73
Use of statins (%)	81.3	82.1	76.2	0.52

*Differences are significant.

Data are presented as mean ± SD respectively percentages.

p was assessed with student t-test (means)or chi-square test (percentages). Continuous variables distributions were tested for normality using D’Agostino-Pearson test and for equality of variances with Levene’s test.

BMI, body mass index; eGFR, estimated glomerular filtration rate; SBP, systolic blood pressure; DBP, diastolic blood pressure; AKI, acute kidney injury; CKD, chronic kidney disease; DM, diabetes mellitus; ACEI, angiotensin-converting enzyme inhibitors; ARB, angiotensin II receptor blocker; CCB, calcium channel blockers.

Hospital mortality was absent in the study population. A number of 53 patients (31.9%) needed a surgical re-intervention, and this re-intervention was not significantly associated with the development of AKI during hospitalization (31% in AKI absent vs. 38.1% in AKI present group, p = 0.52).

All of the patients who developed AKI (12.7%) were at stage 1 or 2 and none of them required renal replacement therapy. The patients with previously diagnosed renal impairment were in stage 3 CKD (eGFR = 30–59 mL/min for more than 3 months) according to K/DOQI 2002 definitions.

We found a significant association between AKI and one-year post-surgery outcome: death and major cardiovascular events (Table 2).

Table 2. Differences between main outcomes and risk factors in the two groups (patients with and without AKI).

	AKI group (n = 21)	Without AKI (n = 145)	OR [95% CI]	p
Decease at 1 year (n = 9)	33.3%	1.3%	35.7 [6.7 to 189]	<0.001
CVE at 1 year (n = 11)	38.1%	2.1%	29.1 [6.9 to 123.4]	<0.001

OR, odds ratio; p was assessed using chi-square test.

CVE, cardiovascular events.

The presence of AKI during hospitalization was associated with an increase in one-year mortality (33.3% vs. 1.4%, OR = 35.7, p < 0.001), one-year major CVE (38.1% vs. 2.1%, OR = 29.1, p < 0.001) and hospitalization (average 19.48 days vs. 15.42 days, p = 0.047).

Having as outcome the decease at one year, in the gender-adjusted multivariate logistic regression model (Nagelkerke R² = 0.519), we found that AKI during hospitalization (p < 0.001) and type 2 diabetes (p = 0.038) were significant cofactors while smoking (p = 0.069), hypertension (p = 0.121) and age (p = 0.431) were not significant.

For the major CVE at one year, the gender-adjusted multivariate logistic regression analysis (Nagelkerke R² = 0.461) revealed that the presence of AKI during hospitalization (p < 0.001) and type 2 diabetes (p = 0.019) is a significant factor while smoking (p = 0.147), hypertension (p = 0.324) and age (p = 0.761) are not influencing significantly this outcome in our cohort.

Discussions

The wide range of definitions available for acute renal impairment makes comparisons of different studies difficult. Nevertheless, the 12.6% prevalence of AKI in our study population is consistent with the reported prevalence of AKI in surgical patients. Depending on the surgical type and the AKI definitions, AKI prevalence has been reported as: 7.7% to 11.4% after cardiac surgery, 8.5% after gastric bypass surgery, 15% to 46% after aortic aneurysm repair surgery and 22% after elective abdominal aorta surgery.13,14

Studies have shown that the risk factors for developing AKI include clinical-setting related (cardiac or vascular surgery, sepsis, intensive care need, trauma, burns, etc.) or patient-specific factors (advanced age, CKD, DM, impaired cardiac function, volume depletion and nephrotoxic medication and agents use) and patients are, usually, exposed to more than one insult. Once the patients develop AKI, they have a longer hospital stay and intensive care requirement.15 Our findings of positive correlations between DM, CKD and hospital stay with AKI are consistent with the literature.

Patients who underwent revascularization surgery (aortofemoral or femoropopliteal bypass) were more prone to develop AKI than those who underwent amputation, possibly because the surgical procedure was more complicated, required more time and involved a higher risk of ischemia-reperfusion injury due to the release of mediators after re-establishing normal blood flow in the lower limb. The high number of patients needing amputation (42%) or re-intervention (32%) and the high prevalence of multiple artery involvement (cardiac, cerebral) show the advanced level of PAD and the extensive impairment of vessels, mostly due to atherosclerosis.

Although many patients were diabetic or hypertensive, fasting blood glucose and blood pressure measured at hospital admission showed an acceptable control of these two factors, consistent with the fact that the patients in this study were scheduled for surgery after thorough investigations and proper treatment of co-morbidities and were not admitted as emergencies. With this selection, we intended to analyze perioperative AKI strictly related to vascular surgery procedures (revascularization or amputation surgery).

The overall one-year survival of 94% and no in-hospital deaths indicate a good prognosis; nevertheless, mortality was 35-fold higher in our study in the AKI group than in the non-AKI patients at one year. Correlations between mortality and AKI are reported in the literature in different clinical situations, the risk of hospital mortality being 3 to 25 times higher in the surgical patients who develop AKI.16

Given the powerful association between AKI and one-year CVE (non-fatal myocardial infarction and non-fatal stroke) and the significantly higher one-year mortality in this group of patients, we could conclude that patients in the AKI group had a more severe and extensive vascular involvement (possibly atherosclerosis) that progressed more rapidly leading to a worse outcome, although most of the patients were treated with statins.

The novelty of our study consists of the analysis of AKI in patients undergoing peripheral vascular surgery, which represents a less frequently analyzed type of patients, because most of the literature has assessed AKI in cardiac surgery or general surgery patients. Also, one-year cardiovascular outcome and survival are less reported in the literature than in-hospital mortality.

The limitations of our study begin with its retrospective design with no possibility of adding new data apart from those collected from the medical records. We could not use other blood chemistries or urinalysis because they had been determined only in a small number of the study patients. The most important disadvantage of our study was the impossibility of evaluating the long-term renal outcome. The exact cause or time of death could not be determined and the mortality data were based on all-cause mortality.

Conclusions

Acute kidney injury can occur frequently in vascular surgery patients. In our study, the presence of previous chronic kidney disease and type 2 diabetes increased the incidence of acute kidney injury after surgery.

Acute kidney injury was associated with a poorer one-year outcome after the surgery, in terms of mortality or major cardiovascular event. Surgeons, intensive care doctors and all healthcare professionals involved in the treatment of PAD patients should be aware of the risk of perioperative AKI, prevent it and treat it properly.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Acknowledgments

Adelina Mihaescu is a PhD student of the “Victor Babes” University of Medicine and Pharmacy Timisoara Doctoral School.