| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Commentary |
1 Both authors: Department of Radiology, Division of Radiologic Sciences, Wake Forest University School of Medicine, Medical Center Blvd., Winston-Salem, NC 27157-1088.
Received June 5, 2005; accepted after revision June 21, 2005.
This article on cerebral infarction was on the list in "AJR:
The 50 Most Frequently Cited Papers in the Past 50 Years" by Felix Chew.
However, it dropped out of the "Top 100 Cited Articles at AJR's
Centennial" by Liem T. Bui-Mansifeld on which this 100 most-cited
article series is based. Because the editors consider cerebral infarction an
important clinical issue that merits commentary, this article has been added
to the series.
Keywords: cerebral infarction • CT • neuroimaging
The AJR's Centennial celebration affords the opportunity to look back at important articles previously published in the journal to reflect on how these articles and the technology underlying them have affected medical practice. The 1975 article by Ken Davis and colleagues, "Cerebral infarction diagnosis by computerized tomography: analysis and evaluation of findings," is certainly noteworthy and may rightly be considered a classic [1]. We will try to put this article in perspective, both in its own time and in ours as well.
Those of us practicing medicine in the 1970s immediately recognized CT as an amazing technologic advance, providing for the first time a detailed noninvasive glimpse into the human body in cross section. Just as MRI would do a decade later, CT revolutionized the field of neuroradiology and the workup of patients with neurologic disorders. At the time, the radiologic diagnosis of stroke was both difficult and inaccurate, relying principally on technetium brain scans to show focal areas of tracer uptake (or lack thereof, depending on the age of the infarct). In equivocal or difficult cases, it was occasionally necessary to subject patients to cerebral angiography. Clearly a noninvasive and more accurate technique such as CT was rapidly welcomed into medical practice. It was therefore a straightforward process for Davis et al. to review their extensive experience with this technique in cerebral infarction because the Massachusetts General Hospital was one of the first medical centers to acquire a first-generation EMI-brand CT scanner and use it to diagnose strokes.
A detailed reassessment of the classic Davis et al. article is an interesting exercise because it provides the opportunity to reflect on the radiologic diagnosis of stroke over a 30-year interval and to consider how much progress, or lack thereof, we have made. Ironically, notwithstanding the development of an array of advanced CT and MRI techniques, including spectroscopy and perfusion and diffusion imaging, CT remains the mainstay for the emergency diagnosis of stroke at most medical centers even today.
Naturally, a number of diagnostic and treatment paradigms for cerebral infarction have changed during this period. Radionuclide brain imaging, featured prominently in the Davis et al. article, is seldom performed for acute stroke today. Likewise, cerebral angiography, liberally used in the 1970s for stroke diagnosis, is now reserved primarily for the setting of thrombolytic therapy. CT, used sparingly in the 1970s, is now used extensively in the emergency department setting, and it is hard to conceive of any patient with significant neurologic symptoms leaving an emergency department today without receiving CT, MRI, or both. Some technical nomenclature has changed since the Davis et al. article appeared: we now refer to dark areas on CT as decreased attenuation rather than diminished absorption and Hounsfield units have now replaced EMI units as measures of X-ray attenuation.
Even though Davis et al. were limited by first-generation CT technology (their scanner used 1-cm-thick sections and a crude 80 x 80 matrix size), it is interesting to note that many of the observations they made are still seen and taught to current generations of radiology residents. For example, these investigators were the first to recognize several fundamental features of the appearance of cerebral infarction on CT, including that acute infarcts are sometimes triangular or wedge-shaped, the swelling associated with acute infarcts occasionally causes mass-effect on the ventricles, early scans in the first few days after the infarction may be negative, and CT is important for diagnosing or excluding hemorrhage. These fundamental imaging observations are still used today in the radiologic diagnosis of acute stroke, even though scanning is performed using 64 simultaneously acquired 1.5-mm slices with a 512 x 512 imaging matrix.
Although they did not recognize it explicitly, Davis et al. did accurately record an interesting phenomenon in their attenuation data, illustrated in their Figure 4. Specifically, they showed that at approximately 2–3 weeks after an infarction, X-ray attenuation of damaged tissue paradoxically increased slightly to that of nearly normal brain before ultimately decreasing again. What Davis et al. had faithfully recorded, but did not explicitly recognize, was what Becker et al. [2] would later name the "CT fogging phenomenon."
In other words, Davis et al. had fully documented the proper time course of X-ray attenuation after infarction that we still teach to our residents today—that acutely infarcted tissue may be isodense with brain initially, it then becomes darker (decreased attenuation) in the acute phase, returns to a more normal attenuation during the next 2–3 weeks (fogging effect), and then markedly decreases in attenuation as it enters the chronic phase. Contrast-enhanced images will often show gyriform enhancement during this fogging period.
With the Davis et al. article behind us, we must now consider a more profound question: How much have we really advanced in our knowledge of stroke imaging in the last 3 decades? In the 1980s, the widespread introduction of MRI captured the attention of the world as the principal technique for neurologic imaging. MRI was quickly shown to be superior to CT for the diagnosis of acute cerebral infarction, especially in the early hours when CT was often falsely negative. Further advances in the 1990s included FLAIR imaging, especially useful for diagnosing cortical surface infarcts abutting the cerebrospinal fluid in the sub-arachnoid space and diffusion imaging, the current gold standard for radiologic determination of cytotoxic injury to brain cells [3].
MRI rightfully received great publicity for its power to diagnose early acute infarction, and both radiologists and hospital administrators rushed to design new emergency departments with expanded MRI capabilities. All the while, however, CT remained humbly in the background, churning out (then and now) images for the diagnosis of cerebral infarction in most patients. Despite the impressive statistics of increased sensitivity of MRI in the diagnosis of acute infarction, CT is clearly here to stay for the evaluation of acute cerebral infarctions.
Three reasons may explain why many radiologists and clinicians continue to use CT as the first-line technique for assessing patients with acute neurologic deficits, while remaining fully aware of CT's inferior sensitivity to MRI. The first reason is speed. In the acute stroke setting, the mantra "time is brain" is accurate, and the rapidity with which CT can be performed can be approached but not matched or surpassed by any current MR device. The second reason is CT's exquisite sensitivity to cerebral hemorrhage. Even though a variety of phase-sensitive methods and blood product–optimized gradient-echo techniques are available on MRI, there is still the reasonable fear that subtle areas of acute hemorrhage may be overlooked on even the best-quality MR scans, with potentially devastating results if the patient subsequently undergoes thrombolytic therapy. The third reason is that CT will likely remain a triage technique of choice for the foreseeable future because of prognostic and therapeutic implications. Most of the large national and international stroke studies testing different stroke therapies have based their outcomes on initial CT findings, not those of MRI.
Recently, CT has even made a bit of a comeback in the acute stroke setting thanks to the development of two new techniques: CT angiography and CT perfusion. Both techniques allow evaluation of the brain's micro- and macrovasculature, with potential therapeutic implications and stratification of patients into different treatment groups. Thus, modern CT offers a one-stop-shopping approach to the management of patients with acute cerebral ischemia and appears well situated to continue this role into the near future.
So where are we now, and where are we going in terms of stroke imaging? Both CT and MRI have strong proponents, with each side claiming its technique will prevail as the preferred method to evaluate patients with strokelike symptoms. Despite myriad articles and symposia on the subject, there is today no consensus nationally or internationally on the optimal diagnostic workup and therapy of stroke patients.
It is interesting to consider what another 30 years will bring in the workup of stroke patients. Will CT and MRI still be vying for supremacy, or will a completely new and better technique that more accurately gives a physiologic picture of the dynamic state of the acute infarction be the technique of choice? Will molecular imaging probes be used to determine therapy and outcome? Or will we find "superCT," the successor to modern MDCT scanners, still at the forefront? Whatever the future may hold, we can be sure that the fundamental observations about the evolutionary changes in stroke made by Davis et al. in 1975 will still have some value in 2035. Perhaps in our retirement we can wander back to the medical center and see what device is being used in the emergency department for the diagnosis of stroke—as visitors, we hope, not as patients.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
