| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Perspectives |
1 From the Department of Radiology, Harvard Medical School, Boston, Mass, and the Department of Radiology, GRB 290A, Massachusetts General Hospital, 32 Fruit St, Boston, MA 02114. Received September 25, 2000; revision requested October 9; revision received October 12; accepted October 16. Address correspondence to the author (e-mail: athanasoulis.christos@mgh.harvard.edu).
Index terms: Angiography, 9*.1212, 9*.1222 , 9*.1242 • Interventional procedures, 9*.1262 • Perspectives, 9*.122 • Radiology and radiologists, history, 9*.122
At the end of 1 century and the beginning of a new century, three statements apply to the state of vascular radiology. First, the specialty has changed. Second, the change is for the worse, and the discipline is on a decline. Third, the change is for the better; the field is flourishing and will continue to grow. Having spent a lifetime as a vascular radiologist, I can safely say that the first statement is valid. The other two assertions contradict one another. The nature of change causes gloom in the pessimist and brings excitement to the optimist. Who is right? Which view is more realistic? In this article, I will consider the major changes that shaped vascular radiology in the past. I will reflect on whether these changes were good or not. Moreover, I will speculate on what the future may hold for this field of radiology.
The Past: Changes in Vascular Radiology
Vascular Radiology on a Roller Coaster
The modern era of vascular radiology began in 1953, when Seldinger (1) introduced the technique of percutaneous vessel catheterization. Advances in computer technology and medical imaging caused major changes in radiology overall. Vascular radiology kept pace through the adoption of new imaging modalities and the introduction of new percutaneous endovascular interventions. New techniques were not always successful. The road to progress was not always smooth; rather, it resembled a roller coaster.
The golden era of selective and superselective angiography lasted through the end of the 1970s. Selective angiography was the cornerstone in the diagnostic work-up of patients suspected of having renal, adrenal, hepatic, pancreatic, and splenic pathologic conditions. Superselective venous sampling from the adrenal, the pancreatic, and the parathyroid veins yielded impressive results. On the cutting edge of vascular radiology were (a) intraarterial infusions of vasopressin for the control of gastrointestinal bleeding and (b) transcatheter vessel embolizations in the management of hemorrhage and neoplasms. The Dotter method of vessel recanalization (2) and Dotter’s (3) germinal work on intravascular stents opened new horizons and created excitement. These procedures laid a strong foundation for the emerging broader field of interventional radiology.
However, danger was looming. Threatening signals came from the introduction of intravenous digital subtraction angiography, ultrasonography (US), computed tomography (CT), and advances in fiberoptic endoscopy. The threat from intravenous digital subtraction angiography soon faded away, but the use of US and CT made it abundantly clear that the days of the supremacy of diagnostic angiography were numbered. A sense of gloom prevailed among vascular radiologists.
Soon thereafter, balloon angioplasty came along to disperse the threatening cloud (4). Vascular radiology reinvented itself. However, restenosis became an issue that tempered the initial enthusiasm. Laser angioplasty did not solve the problem. Atherectomy had limited success. Intravascular stents became available and dispelled all anxieties. Stents led to the realization of a dream long held by vascular radiologists, the creation of a percutaneous portosystemic shunt. On another front, regional thrombolysis became fashionable. In an attempt to revive the glory of selective angiography, male impotence was addressed. Penile arteriography and venography were declared the last frontier. The attempt failed. Little did we know that the frontier was elsewhere in the pelvis. The application of vessel embolization in the management of uterine fibroleiomyomas was about to take off.
It was a great ride. However, something was missing. Missing was some kind of revenue-enhancing "low-tech" procedure that could be performed in large numbers within the existing infrastructure. The solution was to enter new territories. In the early 1990s, vascular radiologists obliged by aggressively pursuing the business of venous access and management of dialysis fistulas. In the meantime, the "high-tech" intravascular stents became bigger and were covered. Endovascular stent placement for aortic aneurysms started in earnest. Initial experience proved a little too promising for the comfort of vascular surgeons, who started to move aggressively into what was assumed to be the domain of vascular interventional radiology.
Looking at the Numbers
There is another way to ride the vascular radiology roller coaster. Perusal of the numbers and types of vascular studies performed during the last 3 decades in one hospital gives us the opportunity to examine the changes in the field from this viewpoint. All data presented in the following sections are from the Massachusetts General Hospital, Boston.
Vascular radiology versus other radiologic subspecialties.—The numbers of vascular radiologic studies increased over time; 1,660 vascular studies were performed in 1973, compared with 10,284 studies performed in 1999, an increase of 519.5%. Figure 1 illustrates the rising trend on the basis of 26 years of observation. This rise was consistent with the increasing numbers of radiologic examinations overall.
|
|
|
|
|
|
|
Catheter-based studies: selective versus nonselective.—Over the course of 26 years, the numbers of selective and superselective catheterization procedures decreased. In 1973, selective catheterization studies, including those performed for diagnosis and others associated with interventions, constituted 57.4% (686 of 1,196) of the catheter-based procedures and 41.3% (686 of 1,660) of all vascular studies. In 1999, selective work dropped to 27.9% (1,317 of 4,718) of the catheter-based studies and to 12.8% (1,317 of 10,284) of all vascular radiologic studies. The decreasing opportunity for the performance of selective vessel catheterizations has raised concerns about the ability of teachers to teach these methods to young vascular radiologists (5).
Noncatheter vascular imaging studies.—Figure 5 illustrates changes in the relative percentages of various types of noncatheter examinations over time. Leg venography dropped from 27.7% (538 of 1,940) of all vascular radiologic studies in 1976 to 0.2% (20 of 10,284) in 1999. It was replaced by US, which increased from 6.7% (165 of 2,472 studies) in 1988 to 42.8% (4,397 of 10,284) in 1999. CT angiography increased from 0.3% (eight of 2,472) in 1988 to 7.7% (791 of 10,284) in 1999. MR angiography increased from 2.6% (65 of 2,472) in 1988 to 3.3% (340 of 10,284) in 1999. Lymphangiography dropped from 6.3% (105 of 1,660) of all vascular studies in 1973 to 0.2% (18 of 10,284) in 1999.
|
Abdominal aortic aneurysms and aortic dissections.—In the past, catheter-based aortography was the only imaging method used to establish the diagnosis and to assess the anatomy before surgery. The translumbar approach was abandoned in favor of the retrograde femoral catheter-based approach. Later, US became the method of choice for screening, and contrast medium– enhanced CT provided all detail necessary for diagnosis and surgical planning. Catheter-based aortography was no longer necessary. However, with the arrival of endovascular stent placement for aneurysm repair, aortography was again considered necessary. Ironically, even the translumbar approach was called back from retirement. The method is now used to enter the aneurysm sac for the management of endoleaks occurring after grafting.
For aortic dissections, the changes have been equally remarkable. We went through a period when we would not dare to pass a catheter into the aorta of a patient suspected of having an aortic dissection. The best we could do was to evaluate the aorta during the late phase of a right atrial injection of contrast medium. Later, we learned that it was all right to insert catheters and to cross from the true to the false lumen and vice versa. Now, CT and MR imaging are the optimal imaging methods for diagnosis. The catheter-based approach is used for endoluminal fenestration or stent placement.
Atherosclerotic peripheral vascular disease.—Some things never change. This is true for angiography of peripheral vascular disease. There have been a few changes, such as digital image acquisition, moving tabletops, and others, but the classic runoff arteriogram has remained essentially unchanged. MR angiography promises to replace the time-honored method.
Venous thromboembolism.—Ascending leg venography was the principal radiologic test performed early in our practice for the diagnosis of deep venous thrombosis of the leg. The procedure was painful. No sedatives were administered because, after all, the test was not invasive. With the use of full-strength ionic contrast media (diatrizoates), extravasation in the dorsum of the foot could cause extensive necrosis of the skin and tissue. The incidence of proved postvenographic fresh thrombosis was 2.7% (six of 220) (6). Many venograms were obtained in patients undergoing hip surgery. Patients interviewed at discharge mentioned that the venographic procedure was painful, an experience worse than the operation. Things improved with the use of half-strength iothalamate contrast medium. A major change and a dramatic improvement came with the introduction of compression US as an alternative test (circa 1986). The shift from venography to the new noninvasive test was gradual, but the trend was definitive. During the period of 1973–1990, we obtained a mean of 460 venograms per year. Subsequently, the mean dropped to 29 venograms per year.
No major changes occurred in the numbers of pulmonary angiograms obtained for the diagnosis of clinically suspected pulmonary embolism. The mean number of pulmonary angiograms obtained was 183 per year (median, 187; range, 89–255; SD, 46). The first hint of a major change in the field came in 1999, when angiographers and clinicians began to adopt pulmonary CT angiography.
Improvements in technology have had a marked effect on the application of inferior vena cava filters. In the 1970s, filter insertion was a complex procedure. Surgeons exposed the internal jugular vein, and radiologists managed the imaging aspects of filter deployment. Now, the procedure is percutaneous, faster, and safer. As a result, there has been a considerable increase in the use of these devices. The number rose from 20 vena cava filter insertions in 1979 to 185 in 1999 (7).
Imaging of the extracranial carotid arteries.—The clinical acceptance of US and MR angiography as definitive tests for assessment of the extracranial carotid arteries has drastically decreased the need for conventional catheter-based carotid angiography. The number of carotid angiograms increased from 31 in 1974 to 177 in 1993. It decreased thereafter to 55 in 1999. More informative is the trend in the number of carotid endarterectomies performed without prior catheter-based angiography. In 1990, angiography was performed in 85.8% (151 of 176) of the patients prior to carotid endarterectomy. The proportion decreased to 16.0% (54 of 337) in 1998 (P < .05, 
2 test). Conversely, the proportion of carotid endarterectomies performed that were based solely on US increased from 6.0% (six of 100) in 1990 to 56.0% (56 of 100) in 1998 (P < .05, 2 test) (8).
Gastrointestinal bleeding and portal hypertension.—During the 1970s, angiographers used to spend long nights trying to control gastrointestinal bleeding with intraarterial infusions of vasopressin. The clinical need for this intervention has decreased. Improved prevention in the intensive care units and better endoscopy have proved effective in the management of bleeding erosive gastritis and gastroduodenal ulcers. Colonic diverticula seem not to bleed as often as in the past. Colonic angiodysplasia, an entity that I suspect was overdiagnosed in the 1970s, is now in the domain of endoscopy. The intraarterial infusion of vasopressin in the superior mesenteric artery for the control of variceal bleeding and the transhepatic approach to embolization of gastroesophageal varices are procedures of historical interest. They were replaced by endoscopic sclerosis and banding. These procedures had limitations, too. The transjugular intrahepatic portosystemic shunt, otherwise known as the TIPS procedure, came along as the miracle intervention.
Arteriovenous malformations.—I met Dr Robert Linton in 1972. He was a skilled and highly respected vascular surgeon. He established vascular surgery as a distinct specialty and organized the vascular surgical clinic in our hospital. He had seen and taken care of many patients with extensive congenital arteriovenous malformations of the trunk and the extremities. He had operated on many by ligating feeding arteries. He knew that artery ligation was not the solution and was frustrated by the results. We discussed transcatheter embolization of the nidus and of the feeding arteries as a new approach. He was willing to explore. We performed several embolizations. The results were excellent, we thought. Later, we realized that, with the exception of a moderate degree of short-lived palliation, we had accomplished little.
Since that time, I have observed at least two generations of younger vascular radiologists attacking the same problem with new enthusiasm. Catheters have improved, different and new embolic materials came along, and direct approaches to the nidus were tried but to no avail. As one generation concedes defeat, a new one takes over. In the treatment of patients with arteriovenous malformations, I have seen little progress. Transcatheter embolization is not the way to go. The solution may be in gene-based technology and the angiostatins.
The Present: Assessing the Changes
The pessimist considers all that has changed in vascular radiology and reasons as follows:
The eternal optimist has a different outlook:
The Future
What will vascular radiology be like 5 years from now? We cannot be certain about the future, but we can speculate. For organizational and politicoeconomic reasons, vascular radiology has been incorporated into the wider field of interventional radiology. Consequently, the future of vascular radiology is tied to the fortunes of the broader specialty. Some things we can predict with confidence. I am certain, for example, that the debate will continue about pulling the vascular-interventional specialty away from departments of radiology.
Moreover, arguments about changing the name of the discipline will intensify. Will the new name be endosurgery, image-guided surgery, or minimally invasive surgery? I went back to the pages of our textbook on interventional radiology that was published in 1982 (9). In the book, we defined radiologic interventions "as procedures that offer a diagnostic and/or therapeutic alternative to surgery" (9). At that time, the slogan was to make surgery unnecessary. We offered less invasive alternatives. Now, the cliché is that we want to be surgeons. However, why the debate? The press already has made the decision for us. On a recent morning, I opened my local newspaper, and there it was. In a front-page article on the subject of fading lines between medical fields, the reporter included these definitions: The old radiologist "read x-rays, CT scans and MRIs." The new radiologist "uses imaging tests to perform operations" (10).
I asked our young fellows in training, "What will vascular radiology be like in 5 years?" After all, they will be creating the future! Most of them were of the opinion that catheter-based angiography would be eclipsed by noninvasive cross-sectional imaging. Percutaneous catheter-based procedures would be performed only for the purpose of intervention. The number of interventional procedures would increase. New interventions would be introduced, including catheter-directed drug delivery and regional organ-specific cell infusions. The prophecy of one trainee impressed me as particularly astute. He predicted that in the future, we would have two distinct kinds of specialists. One would be a combination of vascular surgeon– vascular radiologist performing catheter-based procedures. The other kind would include the imagers, those radiologists who would continue to interpret cross-sectional imaging studies and would stay away from invasive procedures.
We learned from the past that vascular radiology reinvents itself. Past successes ensure a promising future. Applications of vascular radiology in the fields of molecular medicine, gene therapy, angiogenesis, the statins, the vulnerable atherosclerotic plaque, and others are waiting to be explored. Angiographic techniques will be tools for the delivery of drugs, intravascular brachytherapy, and plaque ablation. Stents will become coated with drugs aimed at preventing restenosis. The trend toward the application of noninvasive diagnostic methods will continue. Electron-beam CT and optical coherence tomography will open new avenues. Some of the noninvasive methods may become semi-invasive. The intraarterial injection of ultrasound or MR signal-enhancing agents and endovascular MR imaging with coil-tipped catheters will be explored. Therefore, I am optimistic about the future. However, progress will be realized only through basic research and development.
In this regard, I submit that the assimilation of vascular radiology into the broad discipline of interventional radiology may not be such a good deal. We need physician-scientists whose interests would primarily be in the domains of vascular imaging and vascular intervention. I know of no distinguished vascular surgeon who, in addition to vascular surgery, performs a little biliary, urologic, biopsy, or abscess drainage work on the side. In the ideal world, vascular radiologists would practice, teach, and conduct research pertinent to the vascular tree. I concede that in the real world (ie, in the community hospital), we could not afford the luxury of having interventionalists dedicated to the vascular system. However, in the university hospital and academic centers, we should make it possible for vascular radiologists to concentrate on the vascular tree to further research. Advances in vascular radiology will be realized only if we can attract the ablest and brightest young minds and if we endow them with the funds, time, and resources needed to conduct research. If we succeed, vascular radiology will flourish.
FOOTNOTES
2 9*. Vascular system, location unspecified. 
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
= absolute numbers (scale at left), 
= vascular studies as percentage of all radiology department studies (scale at right). Data are from Massachusetts General Hospital.
= CT, * = US, 
= mammograms, 
= vascular studies. Data are from Massachusetts General Hospital.
) and noncatheter (
