| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Commentary |
1 Both authors: Department of Radiology, Mayo Clinic, 200 1st St. SW, Rochester, MN 55901.
Received December 7, 2007; accepted after revision December 11, 2007.
Complication: A difficult factor or issue often appearing unexpectedly
and changing existing plans, methods, or attitudes
[1].
Keywords: cerebral angiography • complications • CT angiography • MR angiography • neuroradiology
Cerebral angiography has for decades been of inestimable value in the diagnosis and evaluation of diseases of the CNS, particularly for vascular lesions such as aneurysms, arteriovenous malformations (AVMs) and fistulas (AVFs), CNS vasculitis, and atherosclerotic vascular disease. It has been nothing less than essential in the evaluation and treatment planning of these diseases, and it has furthered our understanding of the nature of these disease processes. However, catheter angiography remains an invasive procedure, albeit "minimally" so, involving the "unnatural" placement of catheters in very important and sometimes sensitive arteries. There is no way to entirely eliminate risk to these arteries and to the brain from this procedure, and the angiographer ponders primum non nocere when describing risk of stroke or death to the patient during the informed consent process. The focus of this review is the ongoing question: In the setting of constantly improving, utterly noninvasive CT angiography (CTA) and MR angiography (MRA) capabilities, is diagnostic cerebral angiography a brutish test of the past?
Cerebral angiography was born in 1927, when Antonio Caetano de Abreu Freire, an intriguing Portuguese physician and politician who used the name Egas Moniz [2], first described this procedure in humans. Of his first six patients, two developed Horner syndrome from perivascular extravasation of contrast material, one developed transient aphasia, and one died of thromboembolism to the anterior circulation [3]. Apparently, Moniz made no reference to patient consent, and preclinical evaluation of the toxicity of the strontium bromide contrast material initially used was less than thorough [3].
This was the infancy of cerebral angiography, when carotid arteries were
surgically exposed for injection. In 1931, Moniz began using Thorotrast
(colloidal thorium dioxide, a contrast agent no longer used), with its
perpetual 
-particle emission (biologic halflife of 500 years) and
resultant induced malignancies
[3,
4]. In 1936, percutaneous
carotid angiography was introduced
[5]. It was 1953 when Seldinger
[6] made the monumental
introduction of the technique of percutaneous transfemoral catheterization,
and the development of pre shaped catheters for selecting brachiocephalic
vessels soon followed [7,
8].
In the days before cross-sectional imaging, when skull radiography and pneumoencephalography were the only nonangiographic means of imaging the head, cerebral angiography was a wonderful option. Of course, vascular lesions such as aneurysms and AVMs were well evaluated, but angiography also could represent the only chance of evaluating CNS tumors—remember the square and round shifts? There was just that persistent, unavoidable problem that patients could suffer infarcts, or even death, from cerebral angiography.
In two of the 100 most-cited AJR articles from its first century of publication, Mani et al. [9] and Earnest et al. [10], respectively, reported on the complication rates of cerebral angiography on 5,000 angiograms retrospectively and 1,517 angiograms pro spectively. Reporting permanent neurologic deficits of 0.06% and 0.33%, respectively, these articles assured practitioners that although some risk of cerebral angiography was unavoidable, absolute risks were quite low. At least partially as a result of these re assuring low rates of complications, catheter angiography flourished, buoyed by such innovations as biplane angiography, digital subtraction angiography, braided preshaped catheters, nonionic contrast material, and pressurized saline flush systems.
In parallel with these advancements in cerebral angiography, CT and MRI burst onto the scene, producing the marvel of "direct" tumor and other brain lesion imaging. Subsequently, along came CTA and MRA, and the cervical and intracranial vasculature could then be visualized with only the small risk of contrast material-related allergy or nephropathy. Now, barriers to patients receiving MRA or CTA seem nonexistent, and legions of surgeons operate on many cervicocerebrovascular lesions based on MRA or CTA findings alone. MDCT angiography [11, 12] takes only seconds, has sub millimeter spatial resolution, and can even be performed in dynamic fashion [13]. There are now many variants of MRA [12, 14-16], MR venography (MRV) [17, 18], and even time-resolved MRA for the evaluation of shunting vascular lesions [19, 20].
What advantage does cerebral angiography have over these noninvasive techniques? Perhaps a better question is "Is the incremental spatial and time resolution and vessel selectivity of cerebral angiography worth its inherent risk of complications?" We think the answer frequently is "yes." With regard to accuracy and breadth of diagnostic information offered, conventional angiography is un matched. The spatial resolution (0.2 mm) and temporal resolution (0.25 second) of catheter cerebral angiography remain incomparable. CT may be approaching this for some applications, with potentially 0.4-mm spatial resolution and 0.5-second temporal resolution, whereas MR remains a little further away with regard to temporal resolution (e.g., 2 seconds). However, conventional angiography remains the clear winner in these measures of performance.
These factors and other cross-sectional imaging technique-specific limitations, such as artifacts, translate into clinically meaningful differences in accuracy between these noninvasive techniques (CTA and MRA) and conventional angiography. The accuracy of CTA and MRA has been most thoroughly evaluated in diseases such as intracranial aneurysm, where despite some optimistic reports [21-23], it is not yet sufficient for many patients [24-31]. Clearly, a modern CTA or MRA examination will be more than adequate for many patients' needs: for ex ample, in imaging some small, asymptomatic aneurysms that wouldn't be treated or in unequivocally mild carotid atherosclerotic disease. But there remain common situations, for example in subarachnoid hemorrhage, where a 90% negative predictive value for aneurysm with CTA or MRA is not good enough [24].
With regard to risk, from our more recent report on prospectively evaluated risk of cerebral angiography complications in 19,826 patients, permanent stroke risk does remain small (0.14%) and has mildly decreased over the past 20 years [32]. Other serious complications of the procedure are also rare [32]. Ultimately, of course, physicians must estimate the risk-benefit ratio for each individual patient who requires cerebral angiography.
Thankfully, catheter cerebral angiography is needed less than when Mani et al. [9] and Earnest et al. [10] were reporting complication rates of the procedure in the 1970s and 1980s, but it remains the reference standard in the diagnosis and evaluation of neurovascular pathology such as intracranial aneurysms, AVMs, AVFs, CNS vasculitis, and even atherosclerotic occlusive disease. Endovascular inter ventionalists need the artifact-free spatial resolution and avoidance of overlap of adjacent vascular structures in planning coil embolization of intracranial aneurysms and, of course, require a diagnostic angiogram before any endovascular intervention. Large clinical trials such as the North American Symptomatic Carotid Endarterectomy Trial (NASCET), European Carotid Surgery Trial (ECST), and Asymptomatic Carotid Atherosclerosis Study (ACAS) were based on measuring degree of carotidstenosis as determined at catheter angiography and they remain foundational guides for surgeons [33-35]. Gamma knife AVM nidus targeting and judgment of any posttherapy AVM residual frequently remain better evaluated with catheter angiography. CNS vasculitis commonly involves the third- and fourth-order branches of the cerebral circulation, and these diminutive vessels are still best resolved with catheter angiography.
There are other less tangible advantages of the dynamic, highly time-resolved nature of catheter angiography, such as judging the significance of flow limitation from vessel stenoses and evaluating pial collateralization or external carotid artery (ECA) to internal carotid artery (ICA) collateralization. The ability to selectively inject arteries with catheter angiography gives important information on the relative contribution to brain perfusion of any given artery (e.g., selectively injecting the ECA after ECA-MCA [middle cerebral artery] bypass tells us of the functionality of the bypass).
One other large-tusked pachyderm remaining in the room is the issue of training, and whether the technique will continue to be as safe, or as accurate, if performed by relatively inexperienced practitioners. There is little doubt that neuroradiology fellows today fail to gain nearly as much experience performing angiography as did previous generations of neuroradiologists. Having seen (and personally experienced) how new attending neuroradiologists struggle with both skills and confidence in the angiography suite, leads to the question: Is the future of cerebral angiography further threatened by inadequate training? We surmise that neuroradiologists relatively less experienced in catheter angiography will further skew the imaging in the direction of CTA and MRA, irrespective of the accuracy of these noninvasive techniques relative to catheter angiography.
We wholeheartedly welcome the continuous improvement of CTA and MRA and their slow subsumption of indications for diagnostic catheter cerebral angiography. This process will no doubt continue. In our diagnostic cerebral angiography practice, we grow weary of informing patients that, although the risk is small, complications such as permanent disability or even death cannot be excluded. However, even in the year 2008, for many patients with potentially life-threatening neurovascular conditions, we practitioners—and those patients—must still accept these risks to gain the potentially life-saving diagnostic information provided by the high-quality cerebral angiogram.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
