Article Text
PDF
Abstract
Objective Exposure to contrast agents may cause nephrotoxicity. The safety of performing CT angiography without having knowledge of the baseline creatinine level in stroke patients treated with tissue plasminogen activator (tPA) has not been established.
Methods This is an observational cohort study, with a historical control group to evaluate the safety of CT angiography performed before tPA treatment given within 3 h of symptom onset. The CT angiography group represents all patients treated with tPA between September/2003 and November/2007 who had CT angiography. The control group consists of all patients treated with tPA between January 1999 and August 2003 when CT angiography was not performed. The primary outcome was a creatinine increase in 24–72 h compared with baseline; the secondary outcome was a creatinine increase by ≥44 μmol/l in 24–72 h and the incidence of symptomatic intracerebral haemorrhage (sICH).
Results Baseline parameters between the CT angiography group (164 patients, age 70±11; 91 male) and the control group (77 patients, age 67±11; 45 male) were similar. In the CT angiography group, the mean creatinine increase was −0.89 mmol/l and in the control group 2.2 mmol/l (p=0.42). A creatinine increase of ≥44 μmol/l occurred in five patients (3%) in the CT angiography group and in three patients (4%) in the control group (p=0.50). Also, in the CT angiography group, eight patients (5%) had sICH as compared with three patients (4%) in the control group (p=0.73).
Conclusion Contrast agents given for CT angiography, performed in patients with normal and abnormal creatinine level, neither caused renal injury nor interfered with the safety of tPA treatment.
- CT angiography
- thrombolysis
- creatinine
- outcome
- renal disease
- cerebrovascular disease
- diabetes mellitus
- fibrinolysis
- stroke
Statistics from Altmetric.com
- CT angiography
- thrombolysis
- creatinine
- outcome
- renal disease
- cerebrovascular disease
- diabetes mellitus
- fibrinolysis
- stroke
CT angiography is a commonly used method to diagnose cerebral artery occlusion in patients with stroke treated with intravenous recombinant tissue plasminogen activator (tPA).1 Knowledge of the presence of cerebral artery occlusion is an important prognostic factor2 3 and is also crucial for indication of rescue intra-arterial interventions such as intra-arterial thrombolysis or mechanical thrombectomy.4 5 Because CT angiography provides the clinician with useful information for patient management, we consider the procedure justified, even though it does result in additional radiation exposure.
The use of contrast agents during CT angiography with or without CT-perfusion can, however, raise the serum creatinine level,6–9 and an increased creatinine level has been shown to increase mortality in patients with stroke.10 11 Moreover, contrast agents can induce nephropathy called ‘contrast induced nephropathy’ (CIN).12–15 CIN is usually defined as an increase in the serum creatinine level by ≥44 μmol/l within 48–72 h of contrast medium administration and the absence of an alternative aetiology.12–15 CIN increases mortality in patients undergoing cardiovascular interventions,16–19 but data on outcome of patients with acute stroke are insufficient.
In many centres, CT angiography is indicated only if increased creatinine is excluded,8 because pre-existing renal disease has been identified as a risk factor for CIN.20 However, avoiding laboratory exams before CT angiography can shorten onset-to-tPA-treatment time and translate into a greater benefit from tPA treatment.21 The safety of performing CT angiography without having knowledge of baseline creatinine levels in patients with acute stroke treated with tPA is unknown.
Therefore, the first aim of our study was to investigate the safety of performing CT angiography on renal functions. The second aim was to analyse whether the use of contrast agents interferes with the safety or efficacy of thrombolytic therapy.
Methods
This is a retrospective observational non-randomised cohort study with a historical control group to evaluate the safety of CT angiography, indicated without knowledge of baseline creatinine levels. CT angiography was performed before treatment with tPA, which was given within 3 h of symptoms onset. Starting in September 2003, we indicated CT angiography for all candidates of tPA-treatment as a standard of care. We divided all consecutive patients treated with tPA in our centre from January 1999 to November 2007 into two groups. The target group (the CT angiography group) represents all of the patients treated with tPA between September 2003 and November 2007 who had emergent CT angiography. The control group consists of all the patients treated with tPA between January 1999 and August 2003—in this period, CT angiography was not performed on tPA-treatment candidates. All of the patients were treated intravenously with tPA at a dose of 0.9 mg/kg (max. 90 mg) according to National Institute of Neurological Disorders and Stroke trial criteria.22 Patients' treatment and management remained the same in both groups throughout the whole study period.
In our hospital, the CT scanner is located in a different building than the emergency department or stroke unit. To avoid delays due to patient transportation, patients are brought directly to the CT facility by the Emergency Medical Service (EMS). Here, a clinical examination is performed, blood samples for laboratory tests are taken, and immediately afterwards a CT is carried out. The patient is then transported to the stroke unit, where the results of the laboratory tests are obtained, and the decision whether to treat with tPA is made.
In all patients treated with tPA since 1999, we prospectively documented demographic data (age, sex), stroke risk factors (history of hypertension, diabetes mellitus, coronary artery disease, atrial fibrillation, congestive heart failure, cigarette smoking, hyperlipidaemia), history of chronic renal disease, previous use of antiplatelets, admission systolic blood pressure (SBP) and symptom onset-to-treatment time.
All patients had a haematological and biochemical examination performed in a single laboratory at baseline and 24–72 h after admission. We recorded the following laboratory parameters: blood glucose, blood nitrogen urea, blood creatinine (at baseline and highest value 24–72 h), C reactive protein (CRP), haemoglobin, platelets, international normalised ratio (INR), activated partial thromboplastin time (aPTT) and fibrinogen.
The severity of neurological deficits before intravenous thrombolysis and at 3 months was assessed using the National Institutes of Health Stroke Scale (NIHSS). Only Stroke Unit physicians who were trained and certified for the application of the NIHSS recorded the score.
The primary outcome was the maximum creatinine increase (ie, if more results were available, the highest value was used) in 24–72 h as compared with the baseline level. The secondary outcome was the maximum creatinine level within 24–72 h, creatinine increase by ≥44 μmol/l in 24–72 h compared with baseline, the incidence of symptomatic ICH, 3-month mortality and favourable 3-month clinical outcome defined as a modified Rankin Scale (mRS) score 0 or 1. Intracerebral haemorrhages were identified on control CT scans performed between 22 and 36 h after admission. Haemorrhagic transformations were classified according to European Cooperative Acute Stroke Study criteria.23 The mRS score was assessed at the 3-month visit.
CT scan and CT angiography
All patients underwent a non-contrast head CT scan within the first 3 h of symptom onset and prior to the initiation of intravenous thrombolysis. Immediately after the baseline native CT scan, CT angiography was performed. CT perfusion was not performed in any patient. In all cases, CT angiography was performed before we knew the baseline laboratory parameters including creatinine levels. The dose of contrast agents was recorded. Low-osmolar, non-ionic contrast agents (iomeprol, iohexol, iopamidol, iopromid, iodixanol, iobitridol) were used. We did not use any specific postangio hydration protocol or specific management.
Non-contrast CT was repeated between 22 and 36 h after admission (or earlier when a rapid neurological deterioration occurred). This protocol was approved by the local ethics committee.
Statistics
Patients were classified into the CT angiography group and the control group. Statistical significance for intergroup differences was assessed by χ2 or two-tailed Fisher exact test for categorical variables, and the two-sample Student t test for continuous variables. The Mann–Whitney U test was used to compare medians of baseline NIHSS. A multiple linear regression analysis was performed to identify independent clinical variables associated with the creatinine increase in 24–72 h compared with baseline. The adjusted R2 was calculated for this model to assess whether the independent variables were good predictors of the creatinine increase in 24–72 h. Only significant associations are reported in the final model. A level of p<0.05 was selected for significance. For the intergroup differences of the main outcome measures, p values were adjusted for age, baseline NIHSS, history of hypertension, diabetes and history of nephropathy.20 Analyses were performed with NCSS 2007 software.
Results
Between January 1999 and November 2007, a total of 280 patients received intravenous tPA.
The ‘CT angiography group’ was 164 patients treated with tPA between September 2003 and November 2007. In this period, a total of 201 patients were treated with tPA, but seven patients were excluded because baseline creatinine values were unavailable, and another 30 were excluded because no CT angiography was performed. CT angiography was not performed for the following reasons: agitation (n=8); at the discretion of the physician for patients with dense middle cerebral artery sign on unenhanced brain CT and severe neurological deficit (n=7); allergy to contrast agent (n=7); referral from other hospital (n=6); baseline MRI instead of CT (n=1); technical reasons (n=1).
The ‘control group’ was 77 patients treated between January 1999 and August 2003 when no CT angiography was performed. In this period, 79 patients were treated with tPA, but two patients were excluded due to missing creatinine values.
The patient characteristics of the CT angiography group and the control group are shown in table 1. In total, 55 (23%) patients had abnormal baseline creatinine levels (53 patients had abnormal creatinine levels below 200 μmol/l and two patients above 200 μmol/l). No patient required acute dialysis. The distribution of baseline parameters was similar in both groups except for the history of hypertension, which was more frequent in the CT angiography group.
Patient characteristics in the CT angiography group and the control group
The median time that the samples of follow-up creatinine were taken was 48 h in both groups (p=0.86). The performance of CT angiography had no influence on the creatinine increase in 24–72 h compared with baseline, on the maximum creatinine level within 24–72 h or on the incidence of creatinine increase by ≥44 μmol/l in 24–72 h compared with baseline, as shown in table 2.
Laboratory and clinical outcome in the CT angiography group and control group
In univariate regression analysis, we identified baseline creatinine, CRP, fibrinogen, platelets, history of congestive heart failure and chronic nephropathy as predictors of creatinine increase in 24–72 h. In a multiple regression model, only the baseline creatinine and baseline CRP levels remained significant (F=16.1, adjusted R2=0.44, p <0.001). These data are shown in table 3. The mortality of patients with creatinine increase by ≤0 μmol/l (n=151), 1–10 μmol/l (n=56), 11–43 μmol/l (n=26) and ≥44 μmol/l (n=8) in 24–72 h compared with the baseline level was 14, 7, 35 and 63%, respectively (p <0.001). CT angiography did not have any influence on the incidence of symptomatic ICH or on patient outcome at 3 months, as shown in table 2.
Univariate and multivariate regression analysis to predict creatinine increase in 24–72 h compared with baseline
Discussion
Our study evaluated the safety of performing CT angiography without having previous knowledge of baseline creatinine levels in patients treated with tPA. In our study, we found that performing CT angiography was not associated with impairment of renal functions within the first 24–72 h. We conclude this because neither a relative creatinine increase nor absolute values of creatinine differed between the CT angiography group and the control group within the first 3 days after admission to hospital, even though more than 20% of our patients had abnormal creatinine level at baseline. Also, adjustment for the most common predictors of creatinine increase did not change the results.
Previous studies showed that a creatinine increase by ≥44 μmol/l occurred in 2–3% of patients with stroke within 2–5 days after CT angiography with or without CT perfusion.6–9 This increase was considered to be due to CIN.6–9 However, other causes may lead to a significant creatinine increase in patients with acute stroke. To control for such alternate causes, we employed the historical group in our study. We showed that a creatinine increase by ≥44 μmol/l occurred with an incidence of 3–4% in both our groups with and without CT angiography. For this reason, we conclude that CIN is not responsible for the majority of creatinine increase observed in patients with acute stroke. Our study thus supports the previous findings that performing emergent CT angiography is safe in terms of renal functions in patients with acute stroke with and without knowledge of baseline creatinine levels.1 24
We found that the only variables predicting creatinine increase independently from CT angiography were baseline creatinine levels and baseline CRP. Thus, our results are consistent with previous studies identifying (1) pre-existing kidney disease as a risk factor for creatinine increase19 25–29 and (2) sepsis, which is accompanied by elevated CRP, as the most frequent cause of acute renal insufficiency in critically ill patients.30 31 Because CRP elevation has been shown to be associated with increased mortality in patients with stroke,32 these data raise the possibility that one of the mechanisms, how CRP increases mortality, could be through impaired renal functions. Notably, a significant creatinine increase during the first 24–72 h after stroke onset was associated with increased mortality in our study, which is in agreement with the results of previous studies that found an association between increased baseline creatinine level and mortality after stroke.11 33
In our study, we also investigated haemorrhage risk associated with the use of CT angiography, because previous experimental studies reported that contrast agents can disrupt the blood–brain barrier.34–36 Blood–brain barrier disruption has been shown to increase the risk of ICH in patients with stroke treated with thrombolysis.37 In our study, we found no increased risk of ICH after CT angiography. Therefore, it is unlikely that contrast use would compromise the safety of tPA treatment.
In our study, we found a tendency towards a higher mortality and a lower rate of favourable clinical outcome in the CT angiography group. The significance diminished, however, after adjustment for predefined baseline variables. This implies that contrast agents probably do not interfere with the efficacy of tPA treatment.
The limitation of our study was sample size, which restricted the power to detect smaller differences in binary outcome measurements. The other possible limitation was the use of the creatinine value as the approximation of renal function. Also, our results are valid only for patients with a moderate increase in baseline creatinine level, because patients with a creatinine level above 200 μmol/l, were under-represented in our study population. Because both CT angiography and CT perfusion could be performed with the average dose of contrast agents used in our study, our results may be generalised for multimodal CT imaging. The main strength of our study is that CT angiography was always performed in all patients without knowledge of the baseline creatinine levels and that we implemented a historical control group. The historical control group consisted of patients treated with tPA under the same conditions as those in place after the introduction of CT angiography to our clinical routine. Moreover, to minimise any potential bias related to the use of a historical group as a control, we adopted a model-based approach to eliminate the effect of baseline variables.
In conclusion, our study showed that performing CT angiography on patients with acute stroke treated with tPA does not impair renal functions, increase the risk of ICH or lead to a worse clinical outcome. This has important implications for clinical practice because, based on these data, the efficacy and safety of tPA treatment are not altered by the administration of contrast agents; delays in tPA treatment can be avoided by not waiting for laboratory results; and information on artery status can be obtained regardless of baseline creatinine values.
References
Footnotes
Competing interests None.
Ethics approval Ethics approval was provided by the St Anne's Hospital Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.