Contrast-induced nephropathy: A dilemma between loss of neurons or nephrons in the setting of endovascular treatment of acute ischemic stroke

Abstract

Purpose

The aim is to report the incidence and risk factors of contrast-induced nephropathy after the use of iodine-based contrast for the endovascular treatment of acute ischemic stroke.

Methods

Data from patients who underwent neuroendovascular procedures in a center over a period of 22 months were analysed retrospectively. Contrast-induced nephropathy was determined by an increase in serum creatinine level of >25% of baseline or an absolute increase in serum creatinine level of at least 44 µmol/L (0.50 mg/dL) occurring after intravascular administration of contrast media without alternative explanation. The primary outcome measure of this study was the presence of contrast-induced nephropathy in these treated patients. Continuous data were presented as mean ± standard deviation, and categorical data as frequencies or percentages. The comparison was made using Student's t-test or Fisher's test. Logistic regression was performed to find independent contrast-induced nephropathy predictors. All statistical analyses were performed using Microsoft Excel. A p value of less than 0.05 was considered statistically significant.

Results

One hundred and eighty-nine patients undergoing endovascular treatment for acute ischemic stroke. Twenty cases of the total cohort (n = 189) presented contrast-induced nephropathy (10.58%). Only diabetes and creatinine levels between 1.3 and 2.5 mg/dL were associated with contrast-induced nephropathy. No patient was treated with dialysis.

Conclusion

Contrast-induced nephropathy is a relatively common complication after endovascular treatment of acute ischemic stroke and is associated with worse outcome in patients with this condition. However, there is no increase in the frequency of hemodialysis after the use of iodinated contrast medium.

Introduction

Contrast-induced nephropathy (CIN) is an acute decrease in kidney function that occurs 48–72 h after injection of intravascular contrast media (CM). The most common definitions in use are an increase in serum creatinine (SCr) level of >25% of baseline or an absolute increase in SCr level of at least 44 µmol/L (0.50 mg/dl) that occurs after intravascular administration of CM without an alternative explanation. SCr usually returns to baseline within 14 days; however, some patients may progress to acute renal injury requiring dialysis.¹ Patients with acute ischemic stroke may experience acute renal failure in several ways: hemodynamic variations in blood pressure, prerenal/acute tubular injury due to volume depletion, CIN, obstructive renal failure, and thromboembolic phenomenon with tissue-type plasminogen activator.² Very few studies have reported rates of CIN after therapeutic neuroendovascular stroke procedures. Although the estimated glomerular filtration rate (eGFR) is widely recognized by nephrologists as more accurate than SCr for the evaluation of renal function in patients with stable chronic kidney disease, SCr-based screening strategies that predict the pre-procedure risk of CIN continue to dominate the clinical picture because it is easy to obtain and it has not been observed that in patients who have undergone intravenous contrast procedures, eGFR measures do not add some to the prediction that with SCr levels one obtains.^3,4 The vast majority of published reports on the incidence of CIN that received an intra-arterial contrast injection are based on patients who underwent cardiac catheterization.⁵ Intra-arterial administration of contrast may involve risk factors for kidney injury that are generally not associated with the injection of intravenous contrast material, including sedation that may lower blood pressure, intra-arterial manipulation of a catheter that may lead to embolization of the renal arteries by cholesterol, gas, or clots, or injection of concentrated contrast material that does not leak or dilute through the pulmonary capillaries.⁵

An incidence of 3–5% has been reported in patients undergoing emergency CT perfusion/CT angiography (CTP/CTA) for acute ischemia. In patients with mild to moderate renal failure and diabetes, the incidence of CIN is between 9 and 50%, and in patients with chronic diabetic azotic nephropathy, CIN may occur in 50% to 90% of patients.⁶

The aim of this study is to assess the incidence and risk factors of CIN following the use of iodinated contrast in patients with acute ischemic stroke who underwent endovascular treatment.

Material and methods

This study was approved by the Institutional Review Board (IRB) and the formal informed consent requirement was not required. A retrospective review of 189 consecutive patients undergoing endovascular treatment for acute ischemic stroke between January 2017 and October 2018 was performed. The primary outcome measure of this study was the presence of CIN in these treated patients.

Procedure

According to our protocol, 120 mL of iodine contrast medium is administered (80 mL for CTA and 40 mL for CTP). The typical agent used is iohexol 300.

All neuroendovascular interventions were performed using non-ionic isosmolar CM (iodixanol, Visipaque 270 Bayer, Switzerland). Predictive factors of obesity (body mass index <30), smoking, use of angiotensin converting enzyme inhibitors, statins, diuretics, metformin, non-steroidal anti-inflammatory drugs, anticoagulants, age, sex were recorded. Premorbid medical conditions of hypertension (systolic blood pressure (BP) ≥130 mm Hg and/or diastolic BP ≥80 mm Hg or treatment for hypertension), hyperlipidemia (serum cholesterol >200 mg/dL), diabetes mellitus (no oral hypoglycemic agents or fasting insulin <126 mg/dL), and chronic renal failure (GFR of <60 mL/min per 1.73 m² estimated from SCr). Information was also collected on the volume (mL) of CM used in the CT scan and the type of anesthesia (sedation vs. intubation). The SCr obtained at the time of admission was considered the baseline for this study. SCr was routinely obtained thereafter during the course of hospitalization in the stroke unit.

All procedures were performed by high-level physicians with experience in endovascular techniques. Endovascular treatments were performed with the patient under general anesthesia or conscious sedation. The type of sedation was chosen individually by the neuroradiologist and the anesthesiologist.

The procedures were performed using a transfemoral transarterial approach. An 80 cm 7 F introducer sheath (Super Arrow- Flex Sheath Introducer, Teleflex, Ireland) was introduced into the distal common carotid artery. The distal catheter of 5 or 6 F used was a Navien (Medtronic, USA) or Sofia (Microvention, Terumo, USA). Trevo (Stryker, USA) or Rapid Transit (Cordis, Corp. USA) catheters were selected. Two types of stentrievers were used: Trevo (Stryker, USA) or Preset (Phenox, Germany). A bolus dose of intravenous heparin was usually not administered. Heparinized pressure saline was used during the intervention. The amount of contrast used in angiography is not known because in stroke scenario we performed our runs with manual injection. A volume of 8 mL per recorded run was estimated.

The incidence of hospital conditions such as pneumonia, urinary tract infection, and contrast allergy was recorded. The rate of dialysis and renal insufficiency prior to treatment were recorded. Renal insufficiency was divided into two types: incipient renal insufficiency between 1.3 and 2.5 mg/dL and definitive type > 2.5 mg/dL.

The primary outcome of this study was the presence of CIN defined as an increase in SCr level of >25% of baseline or an absolute increase in SCr level of at least 44 µmol/L (0.50 mg/dL) occurring after administration of CM.

SCr before and after the procedure were obtained from laboratory records.

A comparison of clinical and pharmacological aspects between patients with and without CIN was performed.

Risk factors and comorbidities/hospital pneumonia and urinary tract infection were assessed after review of their medical history.

All patients were interviewed by telephone to obtain the mRS scale score by a neurologist and compare it with the presence or absence of CIN.

Continuous data were presented as mean ± standard deviation (no biased data), and categorical data as frequencies or percentages. The comparison was made using Student's t test or Fisher's test. Logistic regression was performed to find independent CIN predictors. Multivariate logistic regression analysis for predictors of CIN was performed for all variables significant at p < 0.1 in the univariate analysis. All statistical analyses were performed using Microsoft Excel. A p value of less than 0.05 was considered statistically significant.

Results

Information on the cohort of 189 consecutive patients is provided in Table 1. The mean volume of iohexol used on CT was 97.1 mL (standard deviation (SD) 30.1). The media estimated volume of iodixanol used in angiography was 140 mL (SD 74.1). The basal creatinine average was 0.8 mg/dL (SD 0.3). The mean of the highest creatinine level was 1 mg/dL (SD 0.7).

Table 1.

Demographic aspects.

Age	66.11 (SD 13.36)	%
Females	71/189	37.57
Diabetes	46/189	24.44
Chronic renal disease	4/189	2.12
Hyperlipidemia	82/189	43.39
High blood pressure	132/189	69.84
Obesity	62/189	33.33
Smoker and ex-smoker	86/189	45.50
ACE inhibitor	69/189	36.51
Statins	56/189	29.63
Diuretics	42/189	22.22
Metformine	59/189	31.22
NSAIDs	43/189	22.75
Anticoagulants	20/189	10.58
Basal NIHSS	14.36 (SD 6.56)
General anesthesia	20/189	10.58

ACE: angiotensin-converting enzyme; NSAIDs: nonsteroidal anti-inflammatory drugs; SD: standard deviation; NIHSS: National Institutes of Health Stroke Scale.

Twenty (10.58%) cases of the total cohort (n = 189) presented an increase in SCr level of >25% of the baseline value or an absolute increase in the SCr level by at least 44 µmol/L (0.50 mg/dL). Table 2 summarizes the association of clinical factors and drugs evaluated in this report. Only diabetes (p = 0.042, odds ratio = 2.66, 95% CI 1.04 to 6.79) and creatinine levels between 1.3 and 2.5 mg/dL (114-221 µmol/L) (p = 0.0037, odds ratio = 4.39, 95% CI 1.62 to 11.95) were prognostic factors of CIN. In multivariate logistic regression analyses, only creatinine levels between 1.3 and 2.5 mg/dL (p = 0.007) were independent predictors of CIN, while diabetes did not reach a significant value (p = 0.105).

Table 2.

Factors associated to CIN.

	p
Age	0.5461
Gender	0.8270
Basal NIHSS	0.5690
Basal serum creatinine	0.3836
Obesity	0.7252
Smoker or ex-smoker	0.3550
Diabetes	0.0042*
Chronic renal disease	0.3676
Hyperlipidemia	0.4134
High blood pressure	0.9737
ACE inhibitor	0.8673
Statins	0.6213
Diuretics	0.8359
Metformin	0.5508
NSAIDs	0.3807
Anticoagulants	0.9980
Atherothrombotic origin	0.6136
Cardioembolic origin	0.3325
Dissection	0.1199
Indeterminated origin	0.7993
Volume of contrast used in CT and CTA	0.3353
Number of angiography runs	0.5383
Total of estimated volume used	0.8906
Contrast allergy	0.9983
General anesthesia	0.2790
Creatinine levels between 1.3 and 2.5 mg/dL	0.0037 *

CT: computed tomography; CTA: computed tomography angiography; ACE: angiotensin-converting enzyme; NSAIDs: nonsteroidal anti-inflammatory drugs; NIHSS: National Institutes of Health Stroke Scale.

^*Statistically significant.

In the late follow-up, there were six patients with creatinine values over 2.5 mg/dL, in four a respiratory sepsis was diagnosed, one died by a malignant infarct with therapeutic effort limitation, the last one was a transplant patient with renal insufficiency caused by anti-rejection nephrotoxic drugs.

No patient was treated on dialysis. No significant association was found between CIN and hospital pneumonia (p = 0.5332), urinary tract infection (p = 0.3333), or late renal failure (p = 0.9979). A logistic regression analysis showed that CIN was independently associated with poor evolution rated on the 90-day mRS scale (p = 0.0307). There was a statistically significant relationship between the total estimated contrast volume used and the 90-day mRS scale (p < 0.0001).

Discussion

Only 10.58% of this cohort were found to have CIN assessed by SCr, but no single person was required to undergo dialysis. Therefore, waiting for the level of creatinine in the blood of patients with acute stroke before administering contrast can significantly delay treatment and decrease the chance of a good outcome, since a 45-min delay in reperfusion translates into a 10% decrease in the probability of a successful outcome. Our data suggest that it is generally safe to administer contrast in this setting because the incidence of CIN is extremely low, even in patients with known chronic kidney disease.^7,8

CIN is a well-known complication associated with the use of iodinated contrast agents.⁹ Some researchers provide conflicting evidence about the risk of CIN and even question its existence.⁹ Thus, McDonald et al.,¹⁰ after analyzing patients who underwent CT with and without contrast in a single-center retrospective study, suggest that exposure to intravenous contrast material and CIN was more a coincidental than a causal relationship. This author questions both the usefulness of SCr as a marker of kidney injury as the causal association between exposure to intravenous contrast material exposure and CIN.

Previous studies estimate that the incidence of CIN from intravenous contrast administered in an emergency department for all diagnoses is about 7% to 11%.⁷ In some studies, the incidence of CIN in patients receiving non-ionic iso-osmolar CM was lower than that of iso-osmolar CM,¹¹ although a randomized study showed that iodixanol provides a less cytotoxic endothelial effect that protects renal function.¹²

A meta-analysis by Biondi-Zoccai et al. showed that iodixanol, iomeprol, iopamidol, and ioversol are CM with a similar renal safety profile and can therefore be routinely used for interventional diagnosis and imaging. However, iohexol and ioxaglate would be contraindicated. These data do not match our series.¹³

Very few studies have analyzed intra-arterial CIN rates related to contrast.⁹ A retrospective study of 2015 patients found that proteinuria was an independent risk factor in the development of CIN and one-year mortality in stroke patients undergoing cerebral angiography.³ The definition of CIN varies among different authors. Our criteria are stricter than other studies.⁹ This partly explains our high incidence of CIN (10.58%); another possible explanation is that CIN is undervalued in the literature.¹⁴ However, no patient needed hemodialysis in accordance with recent literature. An association was observed between the contrast volume used and the evolution of patients. The use of higher contrast volume may be related to more complex procedures where the chances of a favorable evolution decrease. This would explain the correlation between CIN and the bad clinical evolution observed in our current research.

The results are very similar to the meta-analyses of the intravenously treated population reported by Zorrilla-Vaca et al.¹⁵ CIN may be more frequent in stroke patients due to factors such as blockage of small arteries due to atherosclerosis, aging, hypertension, and diabetes leading to endothelial damage, tubular inflammation, and activation of intrarenal fibrotic pathways. In addition, many stroke patients need intensive care and multiple medications, including antibiotics that can be nephrotoxic.

It has been speculated that the brain injury may have a harmful effect on the renal system, the so-called brain-renal syndrome.¹⁶ The mechanism is unclear and it is debatable whether nephropathy is a cause or a consequence of impairment.¹⁷

Theoretically, the incidence of CIN will increase in the coming years because the affected patients are usually elderly, hypertensive and diabetic, with unstable organic and hemodynamic functions. In addition, neurointerventionist procedures are often time consuming and require higher doses of CM.¹⁸

Biguanide (metformin) has the potential to worsen CIN, with an associated increased risk of lactic acidosis, but in this series, there is no association of metformin use with CIN. Prevention guidelines on the use of metformin with CM are based on expert consensus on the pharmacokinetics of metformin and the pathophysiology of CIN. However, our data support that the relationship between CIN and diabetes is caused by disease per se and not by this drug.

These results suggest that neuroendovascular procedures can be performed in the setting of an acute stroke without an increased risk of severe renal failure requiring dialysis treatment (in fact, there were no patients who had to be dialysed). No one in this series developed severe renal failure attributed to iodixanol postoperatively or months after endovascular treatment; in total, there were only six patients with creatinine values greater than 2.5 mg/dL, four for respiratory sepsis, one patient with malignant infarction with limited therapeutic effort, and one transplanted patient with renal failure caused by anti-rejection drugs.

However, our findings must be interpreted within the context of several limitations. This is an observational study of a single center without a control group, with the intrinsic deficiency of potentially limited external validity and intrinsic bias. The small sample size represents an additional limitation. The most common renal risk factors were included in the analyses, but other factors may not have been evaluated. Another non-quantified variable was the volume of intra-arterial or venous hydration during the procedure, although all patients received fluids for the duration of the procedure. Manual injection is a procedure that makes it impossible to ensure the exact volume of contrast used in the angiographic series. These results cannot be applied to other CM. It has not been studied whether rtPA IV has an effect on renal function compared to the absence of rtPA IV. Patients may have died before dialysis; however, this is unlikely considering that acute renal failure develops within 24 h in critical patients. SCr measurements were not standardized after hospitalization in the stroke unit and were obtained at different times and intervals for different patients, depending on clinical need. Therefore, the incidence of late renal failure may have been underestimated. However, it is very likely that cases of severe late nephropathy were detected, although some subclinical cases may not have been found.

Conclusions

Physicians managing patients with stroke in the acute setting should not be feeling they are facing a dilemma between losing neurons or nephrons. CIN is a relatively common complication after acute ischemic stroke treated by endovascular route but no significant increase of dialysis treatment is caused in this circumstances.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.