|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Evidence-based Practice |
1 From the School of Health Sciences, University of Ulster, Newtownabbey, Northern Ireland, UK (Z.S.); and Real Statistics, Bangor, Count Down, Northern Ireland (G.S.). Received January 27, 2005; revision requested March 31; revision received June 27; accepted July 20; final version accepted September 1. Address correspondence to Z.S., Department of Medical Imaging Science, Curtin University of Technology, GPO Box U1987, Perth, Western Australia 6845 (e-mail: z.sun{at}curtin.edu.au).
| ABSTRACT |
|---|
|
|
|---|
Materials and Methods: A search of the PubMed, MEDLINE, and EMBASE databases for English-language literature was performed. Studies with at least 10 patients were included for data analysis. Only studies on transrenal fixation of aortic stent-grafts that included follow-up results for renal function were included. A log-linear model was used for meta-analysis to compare transrenal fixation with infrarenal fixation.
Results: Twenty-two studies met the inclusion criteria. Because two studies analyzed the same group of patients, one was excluded, for a total of 21 studies. Comparisons between transrenal fixation and infrarenal fixation were found in seven studies. For transrenal versus infrarenal fixation, the combined odds ratio, 95% confidence interval, and P value were found to be statistically significant with respect to postprocedural renal infarction only (combined odds ratio, 5.189; 95% confidence interval: 3.198, 8.420; P < .001). No significant difference was found between transrenal and infrarenal fixation with respect to renal dysfunction, renal artery occlusion, or endoleaks (P > .05).
Conclusion: Transrenal fixation of aortic stent-grafts seems to be a relatively safe alternative compared with infrarenal fixation in terms of short- to midterm follow-up. Postprocedural renal infarction, however, was significantly higher for transrenal fixation.
© RSNA, 2006
| INTRODUCTION |
|---|
|
|
|---|
The physiologic effects of transrenal fixation on long-term renal function remain unknown. In several reports, investigators describe short- to midterm results for transrenal repair of AAA (1626). These reports provide evidence that demonstrates the safety of placing uncovered suprarenal stents over the renal arteries, as assessed during biochemical examination of renal function and imaging follow-up. There are still concerns, however, about the safety of transrenal fixation because the long-term effects of this technique are still not fully understood (1122). The consequences of suprarenal stent coverage of the renal arteries can manifest in various ways (eg, interference with renal blood flow or renal function, decreased cross-sectional area of the renal ostium, or biologic response of the aorta to the stent) (27,28). To better understand the clinical effect of this technique, it is important to evaluate long-term results in a large cohort of patients.
To our knowledge, no systematic review of transrenal fixation of aortic stent-grafts has been conducted. Therefore, the purpose of our study was to perform a systematic review of the short- to midterm effects of the transrenal fixation of aortic stent-grafts on renal function in patients with AAA.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Data Extraction
Data were extracted by one author (Z.S.) on the basis of study design and procedural techniques. All measurable results for short-term (less than 12 months), midterm (1236 months), and long-term (more than 36 months) follow-up were extracted. Patient characteristics, such as preoperative risk factors including smoking, hypertension, coronary artery disease, diabetes mellitus, chronic obstructive pulmonary disease, and renal failure; the type of stent-graft used; and the study institution involved (location and number of institutions) were evaluated and included in the analysis. Commonly observed complications after stent-graft implantation included renal dysfunction, renal infarction, renal artery stenosis or occlusion, stent-graft migration, and endoleaks. These postprocedural complications were related to the deployment of uncovered stent struts above the renal arteries. Other complications encountered in infrarenal stent-graft implantation, such as wound infection, postprocedural hemorrhage, and graft thrombosis, were excluded from data analysis because these complications were not the main concern of our study.
Follow-up results for renal perfusion or renal function were extracted from each study, and analysis included biochemical examination and imaging assessment. Renal function was evaluated by measuring serum creatinine levels (<2.0 mg/dL or 130 mmol/L), blood urea nitrogen levels (<50 mg/dL), or glomerular filtration rate (<20% decrease compared with baseline value).
Statistical Analysis
We performed a meta-analysis for all studies, which yielded 2 x 2 tables that were generated according to the cross classification of postoperative complications (yes or no) and the type of fixation method used (transrenal or infrarenal). It has been the practice to combine odds ratios in 2 x 2 tables by using the Mantel-Haenszel estimator (30). This method, however, was designed to facilitate manual calculation and is not exact. It has been superseded by modern computing hardware and software that yield exact maximum likelihood solutions (31).
A log-linear response model, also known as a logit model, is appropriate (32). This model facilitates the testing of (a) the homogeneity of odds ratios across studies and (b) the estimation and testing of a combined odds ratio from all studies. In each instance, the test statistic is the likelihood ratio
2 statistic. It is logically equivalent to the traditional Pearson test statistic but is considerably more robust. The two methods of calculation yield values that converge with increasing sample size. All analyses were computed by using a commercially available statistical software program (SPSS, version 12.0.1; SPSS, Chicago, Ill). Two-tailed statistical hypotheses were tested at the 5% level of significance.
| RESULTS |
|---|
|
|
|---|
|
|
Early endoleaks and renal dysfunction were seen in 15 and 12 studies, respectively. There was a significant variation among these studies in terms of the reported occurrence of endoleaks and renal dysfunction (P < .001), with pooled incidence rates of 7.2% (95% confidence interval: 6.0%, 8.4%) for endoleaks and 5.0% (95% confidence interval: 4.0%, 5.9%) for renal dysfunction. Stent migration was noted in four studies, with significant variation observed (P < .05). The pooled incidence rate was calculated as 0.4% (95% confidence interval: 0.12%, 0.73%).
Methods of Monitoring Renal Function and Renal Arteries
The monitoring of renal function was performed by using serum creatinine levels in all studies. Serum creatinine and blood urea nitrogen levels were used in five studies. In addition to serum creatinine, the glomerular filtration rate was included in three studies to evaluate the function of each kidney (8,13,26). In these studies, there was agreement that renal dysfunction was defined as a serum creatinine level of greater than 2.0 mg/dL or 130 mmol/L, an increase in serum creatinine of at least 25%, a blood urea nitrogen level of greater than 50 mg/dL, or a decrease in glomerular filtration rate of at least 20%. In one study, perfusion scintigraphy was performed by using technetium 99m diethylenetriaminepentaacetic acid to estimate the glomerular filtration rate after transrenal fixation (13). Researchers in that study found that a significant reduction in glomerular filtration rate (>20%) was observed in the absence of any relevant changes in serum creatinine level.
Helical computed tomographic (CT) angiography was the most commonly used imaging modality to assess the patency of stent-covered renal arteries at regular periods of follow-up in all studies. Helical CT angiography supplemented by either duplex ultrasonography (US) or angiography was performed in six and four studies, respectively. In one study, magnetic resonance (MR) angiography was used as the sole imaging modality for preoperative and postoperative evaluation of patients with preoperative renal insufficiency in order to reduce adverse events (24).
On the basis of these findings, we have composed a flow chart (Figure) to demonstrate the follow-up procedures that we recommend for patients treated with aortic stent-grafts with a transrenal component.
|
Postprocedural complications.The findings from the meta-analysis are summarized in Table 3. Four types of postprocedural complications (renal dysfunction, renal artery occlusion, renal infarction, and endoleak) tested negative for heterogeneity of odds ratios (transrenal fixation vs infrarenal fixation) among the seven studies that compared both types of fixation, with a mean follow-up ranging from 6 to 37 months. All other outcomes (where published detail permitted) exhibited heterogeneity in this respect and were not evaluated with meta-analysis. Renal infarction was the only complication to occur during midterm follow-up for stent-graft implantation that exhibited statistical significance (P < .001), with an odds ratio of 5.189 (95% confidence interval: 3.198, 8.420).
|
| DISCUSSION |
|---|
|
|
|---|
One of the factors that may have an important influence on renal artery patency is the interference of transrenal stents with the renal arteries. Transrenal stent wires may cross the renal ostium in various configurations and have been reported to reduce the cross-sectional area of the renal ostium, depending on the type of stent-graft used, the diameter of the ostium, and the number of stent wires that cross the ostium (28). Researchers have demonstrated in animal studies that the development of a disorganized acellular matrix that causes partial renal ostial occlusion depends on stent type (27,35). In studies performed on in vitro phantoms, however, researchers have not shown significant interference with the cross-sectional area of the renal ostium and blood flow (28,36).
A variety of stent-grafts are used for proximal fixation in clinical practice, which may account for the difference observed in renal artery patency. Our review showed that, although various stent-grafts were used in these studies, no significant difference in renal artery occlusion was found when transrenal fixation was compared with infrarenal fixation; this finding indicates the safety of deploying suprarenal stent wires across the renal arteries.
Another factor that should be considered when performing transrenal fixation is the coverage of accessory renal arteries, which may affect renal perfusion and lead to renal infarction. Exclusion of these arteries raises the theoretic concern of regional ischemia associated with the loss of parenchyma or worsening of hypertension. Although the results of most of the studies indicated that coverage of accessory renal arteries appeared to be well tolerated and did not significantly compromise renal function (renal infarction, <30%), the results of our meta-analysis showed that there was a significant difference between transrenal and infrarenal fixation with regard to the occurrence of renal infarction at midterm follow-up (P < .001). Therefore, it is necessary to make a quantitative estimate regarding the size of the expected infarction before the placement of a suprarenal stent-graft. Two methods (radionuclide renography and a combination of renal angiography and helical CT) are recommended to assess or calculate the volume of renal parenchyma perfused by each artery before the artery is occluded (13,37). Normally, any vessel that is greater than 3 mm in diameter or that perfuses more than 30% of the kidney parenchyma should be spared from coverage.
Although transrenal fixation was designed to secure proximal sealing of the stent-grafts and was reported to significantly decrease the risk of proximal type I endoleaks (38), our review did not corroborate this finding because there was no significant difference between transrenal and infrarenal fixation with respect to the occurrence of endoleaks. We acknowledge that only a small number of studies in our review were available for comparison; some studies concentrated on renal impairment in patients with transrenal stents and did not provide detailed information about endoleaks (7,11,20). This again emphasizes the lack of completeness in the data we were able to collect.
Stent-graft migration is another potential danger for endovascular repair and was expected to be lower in the transrenal treatment groups because fixation may be improved by stents equipped with hooks and barbs, which are reported to provide an at least 10-fold increase in strength for stent-graft fixation (39). Stent-graft migration in our review was found to be less than 3% in four transrenal groups; however, information about migration was not provided for infrarenal groups, which is another limitation of our reviewed data.
There are some limitations in our study that we would like to address. None of the studies that contributed to the meta-analysis was a randomized controlled trial. However, given that the binary event of interestthat is, postoperative complicationswas subsequently observed within each of the two groups of patients who underwent one of two surgical procedures, these studies could be regarded as independent instances of observational cohort studies.
Publication bias may also have affected our results. There appears to be an increasing occurrence of articles in which the authors perform a meta-analysis to assess the presence of publication bias by means of procedures such as the perceived asymmetry of funnel plots. Such assessments are not presented here because (a) the methods are not robust, as asymmetry may be caused by issues other than publication bias and (b) with a maximum of only four studies in the meta-analysis of any one of the outcome variables presented here, any assessment would be of limited reliability. Studies published in other languages or those with a small number of cases (less than 10 cases) were excluded from the study, which is another limitation. Excluding studies with small numbers may underestimate the rate of renal complications.
Data collection within the individual studies may also have been a source of bias. Patients with preoperative renal insufficiency were excluded from analysis in some studies. In two studies that compared transrenal groups with infrarenal groups, detailed information about renal function was not found. In one of these studies, it was not clear how many patients underwent transrenal or infrarenal treatment, and the number of cases of renal dysfunction was not available. Excluding complications other than renal dysfunction in our study could be questioned because there exists a possible relationship between these complications and adverse renal events.
Another limitation was that only one reviewer was responsible for all data selection and data entry. Two or more reviewers participating in the data screening and selection would be preferable.
Quality assessment of the studies was not performed because these studies had an absence of randomization. We should also acknowledge that not all reports provided complete data. Although we contacted several investigators to obtain clarification or additional data with the aim of enhancing our analysis, it was not successful. Lack of the originally reported data and standardization were the main limitations of our study.
In conclusion, our systematic review has shown that transrenal fixation seems to be a relatively safe technique for the treatment of patients with AAA on the basis of short- to midterm follow-up. Transrenal fixation did not cause renal dysfunction in comparison with infrarenal fixation. Patients who had undergone transrenal fixation exhibited a significantly higher rate of renal infarction compared with those who had undergone infrarenal fixation. Because the long-term effect of stent wires on the renal ostium and renal function is still not fully understood, precautions need to be considered in order to reduce potential adverse effects and guarantee long-term safety. The flow chart that we have provided shows the follow-up procedures that we recommend for patients undergoing transrenal fixation of aortic stent-grafts. We believe that these strategies ensure implementation of the proper follow-up modalities in each particular situation so that long-term safety of this treatment can be monitored accurately.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Abbreviations: AAA = abdominal aortic aneurysm
Authors stated no financial relationship to disclose.
Author contributions: Guarantor of integrity of entire study, Z.S.; study concepts/study design or data acquisition or data analysis/interpretation, Z.S., G.S.; manuscript drafting or manuscript revision for important intellectual content, Z.S., G.S.; approval of final version of submitted manuscript, Z.S., G.S.; literature research, Z.S.; statistical analysis, Z.S., G.S.; and manuscript editing, Z.S., G.S.
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |