Diagnostic Arteriography

Vascular and Interventional Radiology is in large measure an outgrowth of angiography. Therefore, it is not surprising that a large number of the procedures done in the interventional laboratory are related to the vascular system. As a diagnostic modality, angiography continues to play an essential role in the work up of peripheral vascular disease, mesenteric ischemia, gastrointestinal bleeding, and trauma. Cross sectional imaging, particularly CT, has surpassed angiography as the primary modality in the work up of abdominal malignancy. Nevertheless, angiography is undertaken in many cases where surgical resection of a solid organ is contemplated, and in conjunction with interventional procedures such as embolization and chemo-embolization for palliative treatment of unresectable tumors. In addition, new techniques combining angiography and CT (CT Arteriography, CT Arterial Portography) appear to significantly increase the detection of small hepatic metastases.

Peripheral Vascular Disease

Approximately 2.5 million Americans have symptomatic peripheral vascular disease (PVD), with an additional 350,000 new cases per year. PVD is a progressive disease in about 25% of patients, and about 25% of initial presentations are for acute limb ischemia. Risk factors for PVD are similar to those for coronary and carotid artery disease, conditions that coexist with PVD in a large number of cases. The most important risk factors are tobacco use, diabetes, hypertension, hyperlipidemia, chronic renal failure, and advanced age (> 60). Males are affected more commonly than females, and a sedentary life style appears to contribute to the disease. Currently, less than 40% of symptomatic patients receive any treatment for this condition. PVD results in 110,000 bypass surgeries, 70,000 angioplasties and 100,000 amputations per year.

The diagnosis of PVD is straight forward in the majority of cases, based on history and physical examination. Noninvasive studies such as pulse volume recording and ankle-brachial indices serve to confirm the diagnosis and establish a baseline for follow up evaluation. However, noninvasive studies give relatively gross information and lack the anatomic detail necessary for planning surgical or other invasive interventions.

Peripheral angiography is the definitive examination for surgical planning in PVD. A well performed angiogram not only defines the precise location of significant lesions, it also gives important information regarding the quality of inflow and runoff, and the character and abundance of the collateral arterial supply to the extremity.

An arteriogram for lower extremity PVD should begin with a complete study of the abdominal aorta to assess both the aorta and its major visceral branches. Coexistent renal and mesenteric artery stenosis is relatively frequent in PVD, and significant plaques are common in the infrarenal aorta. Incidental discovery of an unsuspected abdominal aortic aneurysm is not unusual in these cases. Study of the major pelvic vessels is best performed in oblique projections to evaluate the aortic, iliac and common femoral bifurcations, as significant lesions are common at these sites. A multistation runoff technique using a distal aortic injection is usually satisfactory for study of the lower extremities. However, this should be supplemented with additional views of any segments that are not optimally seen on the runoff study. Patients with diabetes or chronic renal failure often have occlusion of the trifurcation vessels. In such patients, additional views of the ankle and foot should be done with digital subtraction technique to look for a suitable vessel for distal bypass.

Although peripheral angiography as described above is sufficient for preoperative work up in most patients, this technique does underestimate the extent and severity of disease when compared to more invasive methods such as multiplane selective angiography, angioscopy and intravascular ultrasound. This is because atherosclerosis tends to be most severe along the posterior walls of arteries. Consequently, the standard angiographic views (PA) for a peripheral study show the lesions en face rather than in profile. Furthermore, there is wide interobserver variability in measuring stenoses. While concentric stenoses can be measured accurately on a single projection, eccentric stenoses, which are far more common, will be underestimated or overestimated depending on the position of the plaque relative to the view. Features such as post stenotic dilatation, density of the contrast column at the site of the lesion, and slowing of flow through the lesion are useful for refining these estimates. However, interpretation of these features is quite subjective and measurements incorporating them are prone to significant error. The problematical nature of angiographic measurement is most relevant in moderate stenoses where there is a question of hemodynamic significance.

Despite these shortcomings, angiography is the safest and most reliable diagnostic evaluation for surgical planning in PVD. Moreover, angiography often reveals lesions that are most appropriately treated by angioplasty. This is particularly true in patients with concurrent iliac and superficial femoral artery disease. Many of these patients are best served by iliac angioplasty followed by femoropopliteal bypass grafting. Other patients with isolated iliac or femoral lesions appropriate for angioplasty can be diagnosed and treated at one sitting.

Cerebrovascular Disease

Results of the North American Symptomatic Carotid Endarterectomy Trial show that carotid endarterectomy is highly beneficial in patients with recent hemispheric and retinal transient ischemic attacks or nondisabling stroke, and ipsilateral 70-99% stenosis of the internal carotid artery. When compared to optimal medical therapy alone, endarterectomy reduced the 2 year cumulative risk of any ipsilateral stroke from 26% to 9%, and, for major and fatal strokes, the risk was reduced from 13.1% to 2.5%.

The initial imaging work up of cerebrovascular disease may include cranial CT and MRI to evaluate the brain parenchyma. Magnetic resonance angiography (MRA) techniques have improved greatly in the last few years and may eventually play a significant role in the work-up the cerebrovascular disease. However, in current practice, the initial imaging evaluation of the carotid arteries in the neck is by color flow Doppler. Although some surgeons will rely on color flow Doppler for surgical planning, most will request angiography prior to operation.

In many respects, color flow Doppler and angiography can be considered complementary to one another. For example, a very high grade stenosis can be quite difficult to differentiate from occlusion on Doppler studies, but can be shown quite clearly by angiography. In addition, Doppler studies give little or no information regarding the intracranial vessels or the origins of the brachiocephalic arteries, whereas angiography demonstrates pathology in those areas quite clearly. Finally, color flow Doppler is very dependent both on the skill of the operator and the anatomy of the patient. Thus, angiography serves to confirm the Doppler results prior to surgery.

Angiography for carotid stenosis should begin with a study of the aortic arch to evaluate the brachiocephalic artery origins. However, brachiocephalic studies from arch injections are generally inadequate for study of the carotid bifurcations and intracranial circulation. Therefore, selective injections should be made into the common carotid arteries with filming of both the cervical and cranial circulation in at least two projections. Tortuosity of the cervical carotid arteries should be further studied by rotating the patients head to rule out the presence of a carotid kink. The study should also demonstrate the vertebral arteries, including their origins, and the posterior cerebral circulation should be studied from at least one vertebral or subclavian injection. In patients with posterior fossa symptoms, selective injections should be made into both vertebral arteries with filming in at least two projections.

A complete angiographic study as described above defines all surgically accessible lesions, and allows identification of surgically inaccessible carotid and branch vessel disease. Patterns of collateral circulation, and incidental pathology such as cerebral aneurysms are well demonstrated. A thorough angiographic study as described here can be completed in less than 45 minutes using modern equipment, with significant CNS complications occurring in less than 0.1% of patients.

Acute Mesenteric Ischemia

Acute bowel ischemia may be due to arterial occlusion (thrombotic, embolic or vasculitic), mesenteric venous obstruction, or "nonocclusive ischemia". These etiologies will all produce similar clinical symptoms that vary with the duration of ischemia. This disease has a very high mortality without prompt identification and intervention. Irreversible bowel necrosis may occur within 12 hours of onset, although good results have been obtained when patients have had angiography and surgery within 24 hours. Since the therapeutic approach is contingent on whether or not an arterial occlusion is present, the specific cause of the ischemia should be defined as quickly and as precisely as possible.

Nonocclusive mesenteric ischemia occurs in patients with low flow states such as myocardial insufficiency and hypotension, and in patients receiving digitalis. This is probably the most common etiology of acute mesenteric ischemia. The pathophysiological basis of nonocclusive ischemia is persistent splanchnic vasoconstriction in response to the low flow state or cardiac glycoside. Patients with nonocclusive ischemia usually have significant underlying disease. Consequently, the prognosis in nonocclusive ischemia is extremely poor. Nevertheless, some patients have benefited from intraarterial infusion of vasodilators such as papaverine.

Thrombotic and embolic occlusions in the proximal 10 cm of the superior mesenteric artery (SMA) are amenable to surgery. The prognosis in these patients is somewhat better than in nonocclusive ischemia if prompt diagnosis and therapy are undertaken. However, in mesenteric embolization, multiple small branches may be obstructed rendering surgical intervention ineffective.

Mesenteric venous obstruction may occur in hypercoagulable states, low flow syndromes, and as a result of torsion, intussusception and strangulating bowel obstruction. Encasement of the superior mesenteric or portal vein by malignant tumor or inflammatory mass may also produce mesenteric venous obstruction. Clinical symptoms mimic acute arterial obstruction, and angiographic findings mimic nonocclusive ischemia with constricted SMA branches, delayed arterial emptying, and absent visualization of the venous phase. Pharmacologically enhanced angiography using the vasodilator priscoline may improve venous phase opacification and disclose the thrombus. However, absence of the venous phase following priscoline may occur in either nonocclusive ischemia or venous obstruction. In cases of mechanical venous obstruction, other radiographic abnormalities should point to the correct diagnosis.

The angiographic evaluation of mesenteric ischemia should start with a biplane abdominal aortogram to define the extent of disease in the aorta and to assess the mesenteric stem vessels and proximal branches. If the main stem of the SMA is patent, a selective superior mesenteric arteriogram should be performed in a frontal projection with late filming to evaluate the mesenteric veins. If the venous phase is absent or weak, the angiogram should be repeated with priscoline enhancement.

Chronic Mesenteric Ischemia

Chronic obstruction of the SMA is usually due to atherosclerotic plaques involving the proximal 2 cm of the artery, although more distal portions of the artery are commonly involved. There is often a well developed collateral circulation from the celiac axis (pancreaticoduodenal arcades) or inferior mesenteric artery (Arc of Riolan, marginal artery) if these vessels are patent. Symptomatic chronic ischemia is characterized by postprandial pain and malabsorption with fatty stools and weight loss. Occasionally, the main SMA will be widely patent but peripheral branches cannot be observed in detail. In some cases, ischemia will be caused by a vascular steal, as may occur in occlusion of the celiac axis or inferior mesenteric artery or in cases of arteriovenous fistulae. Nonatheromatous narrowing of the SMA may be due to tumor or inflammatory encasement, and stenoses of the distal SMA and its branches are relatively common in carcinoma infiltrating the mesentery. The main stem of the SMA and the SMA branches may also be narrowed by fibromuscular hyperplasia, pseudoxanthoma elasticum, and various forms of arteritis.

Angiography for chronic mesenteric ischemia is similar to that for acute ischemia. The study should begin with biplane abdominal aortography followed by selective injection into the SMA if its origin is patent. Occasionally a critical stenosis will be discovered adjacent to the origin of the SMA; such lesions can be considered for balloon angioplasty. If the SMA is occluded, selective celiac and/or inferior mesenteric injections will disclose the pattern of collateral blood flow and define the distal extent of the SMA lesion.

Acute Gastrointestinal Bleeding

The initial work up for acute gastrointestinal (GI) bleeding should be by endoscopy and nuclear medicine studies. Barium studies should be avoided as they are of very limited value and the retained barium will preclude subsequent angiography. When radionuclide and endoscopic studies fail to disclose the site of bleeding, or in cases where transcatheter intervention is being considered, angiography is an appropriate next step. However, in order to utilize angiography effectively in the emergency management of acute GI hemorrhage, the procedure must be performed rapidly and efficiently, and the indications as well as the limitations of angiography must be well understood by both the angiographer and the attending clinician. About 75% of patients hospitalized for GI bleeding can be managed conservatively with bed rest, sedation and transfusion. These patients are not candidates for angiography.

Successful angiographic visualization of GI hemorrhage requires active bleeding at a rate of 0.5 to 1.0 ml per minute. The major limitation of angiography relates to the intermittent nature of GI bleeding. If bleeding has temporarily stopped at the time of the injection, the angiogram will be negative. Another major limitation of angiography is its inability to demonstrate venous bleeding.

Hematemesis or aspiration of blood via a nasogastric tube usually indicates a bleeding site proximal to the ligament of Treitz, while passage of bright red blood per rectum may occur with bleeding anywhere in the GI tract. This information, together with the findings at endoscopy and on scintigraphy, helps to direct the angiographic work-up. In patients with upper GI tract bleeding, selective angiography of the celiac branches with good opacification of left gastric, gastroduodenal, pancreaticoduodenal, and splenic arteries is done first. If these are negative, the SMA is selectively studied. With presumed lower tract bleeding, the inferior mesenteric artery is studied first to minimize interference from an opacified urinary bladder. This is followed by the SMA and celiac branches as required. In general, aortography is not required in GI bleeding except in those cases where there is difficulty in localizing the mesenteric artery origins.

Once a bleeding site is identified, transcatheter methods can be employed to halt the hemorrhage. Intraarterial infusion of vasopressin is quite successful in both upper and lower GI bleeding. In upper tract hemorrhage, transcatheter embolization techniques are also very successful and have little risk of bowel infarction. Embolization is generally not advised for lower GI bleeding because of the lack of an extensive collateral blood supply.

The presence of small bowel bleeding can be diagnosed accurately from the arteriogram. However, precise localization for surgery can be very difficult because of the complex anatomy of the small bowel. This difficulty can be overcome by placing a microcatheter into the artery supplying the bleeding site prior to surgery. Methylene blue can then be injected intraoperatively to identify the small bowel segment for resection.

Chronic Gastrointestinal Bleeding

Angiography is of limited value in cases of chronic GI bleeding. The most common cause of chronic bleeding is tumor, which is most appropriately worked up with endoscopy and/or barium studies. Unusual lesions such as angiodysplasia, hemangioma and vascular malformations may be diagnosed by angiography. However, an exhaustive work up using other modalities should precede the angiogram since arteriographic studies have limited sensitivity for the common offending lesions.

Angiographic localization in cases of chronic GI bleeding often requires distention of the bowel with air and highly selective injections into the various vascular territories. These procedures are very time consuming and are of relatively low yield.

Abdominal Malignancy

The role of diagnostic angiography in the work up of abdominal malignancies has changed considerably with the advent of high resolution CT and sonography. Angiography now serves primarily to confirm or clarify a diagnosis made from imaging studies, and to rule out resectability of a lesion.

In the kidney, a mass seen by ultrasound or CT may be difficult to distinguish from a complicated cyst. In such cases, angiography can usually make the diagnosis based on the vascularity of the lesion and its effect on neighboring vessels. In the past, angiography was often used successfully to differentiate masses in the liver. However, image guided needle biopsy is generally more accurate and less traumatic to the patient. Nevertheless, angiography will often be requested in these cases to plan a surgical approach to the lesion.

In cases of a known malignancy with metastases to the liver, angiography is very useful if intraarterial chemotherapy via an hepatic artery porta-cath is planned. The arteriogram may show aberrant arterial anatomy (present in >20% of patients) that would preclude this form of therapy. In addition, the arteriogram will demonstrate the origin of the right gastric artery. That artery has a highly variable origin and may arise distal to the infusion catheter. In those cases the right gastric artery must be ligated or embolized prior to intraarterial chemotherapy in order to prevent erosive gastritis and its complications. A further use of angiography in metastatic disease to the liver is in those cases where only a few metastases are present and a segmental resection or lobectomy may be curative. In this situation, angiography is combined with CT in a procedure called CT Arterial Portography. This procedure is based on the fact that hepatic metastases take their blood supply solely from the hepatic arteries while normal liver receives both arterial and portal venous blood. This difference in blood supply is exploited by scanning the liver during injection of contrast medium into either the superior mesenteric or splenic artery. Since no contrast is injected into the hepatic artery, enhancement of the liver is solely from portal venous blood. Consequently, metastases do not opacify and appear hypodense against the background of opacified normal liver parenchyma. This technique can detect very small lesions in the range of 3-5mm.

Angiography and arterial portography without CT has proven useful in the surgical work up of pancreatic carcinomas. Usually, pancreatic cancer is already unresectable at the time of diagnosis. However, occasional tumors are detected early with minimal signs of local invasion on imaging studies. In these cases angiography may show encasement or invasion in the portal vein or superior mesenteric artery and spare the patient a futile abdominal surgery.