Liver: Spiral CT of Hepatic Parenchymal Disease

Introduction

The unique advantage of spiral CT as compared to conventional CT is the ability to scan during a single breath hold during peak contrast enhancement (1, 2, 3). This eliminates data misregistration, minimizes motion artifact, and results in a uniformly high level of contrast enhancement. These factors are especially important in the evaluation of hepatic pathology, where the detection of small lesions and optimal vascular opacification are essential. In addition, a wide spectrum of parenchymal disease can also be accurately diagnosed and characterized with spiral CT (4, 5, 6, 7). This chapter provides an overview of the advantages of spiral scanning in the evaluation of hepatic parenchymal disease.

Spiral Technique: Liver

All patients are instructed on breath hold requirements prior to scanning. We routinely try to hyperventilate patients immediately prior to scanning. In our experience, the great majority of patients are able to tolerate the required 24 to 32 second breath hold. Initial localizing scan is performed, with the start of the spiral scan set just above the right hemidiaphragm. 120 cc of non-ionic contrast is peripherally administered via a power injector at a rate of 3 cc/sec. Non-ionic contrast is preferred to minimize side effects (especially nausea and vomiting) which can delay scanning during peak contrast enhancement. Spiral scanning begins within 10 seconds following the termination of contrast injection. The table is incremented at a rate of 8 mm/sec, with 8-mm collimation at 210 mAs and 120 kVp. Current spiral software allows for data acquisition during a 24 or 32 second breath hold. This allows for coverage of approximately 19.2 or 25.6 cm respectively. These parameters provide adequate length to cover the entire liver in the vast majority of patients (as well as the spleen and pancreas). Images reconstructed from the dataset are formatted using 4-mm incrementation, providing approximately 50% data overlap on the adjacent images. This has been shown to be superior to 8-mm incrementation in evaluation of focal liver lesions (6). Reconstructed multiplanar and three-dimensional analysis are performed in selected cases, especially for the evaluation of vascular anatomy.

Diffuse Parenchymal Disease

Cirrhosis

Cirrhosis represents chronic fibrosis due to hepatic injury from a wide spectrum of disease processes (8). The most common etiologies include alcoholic liver disease and chronic viral hepatitis. Cirrhosis is usually a chronic and progressive disorder. Parenchymal nodules develop from cell regeneration and scar formation. Portal hypertension with ascites, portosystemic shunts, and splenomegaly often ensue.

Characteristic imaging features have been well described with conventional CT. Often, there is a decrease in size of the right hepatic lobe and medial segment of the left lobe, with compensatory enlargement of the lateral segment of the left lobe and caudate lobe (9). The liver demonstrates an irregular surface contour. A ratio comparing the sizes of the caudate lobe and the right lobe may be helpful in determining subtle cases of cirrhosis: the caudate lobe is measured from medial to lateral at the level of the portal vein; the right lobe is measured from the lateral liver edge to the lateral edge of the portal vein. If the caudate to right lobe ratio is greater than 0.65, cirrhosis is likely with a 96% confidence (9).

Because it allows for scanning during peak vascular enhancement, spiral CT is excellent in evaluation of the altered blood flow and distribution commonly seen with patients with cirrhosis (10,11,12). Regenerating nodules, in combination with periportal and perisinusoidal fibrosis, can divert splanchnic venous drainage. In addition, microcirculatory shunts which form between the portal venous and hepatic venous systems result in inhomogeneous decreased flow to the liver (11,12,13,14). Together, these factors can produce a mottled enhancement pattern (figures 1, 2). Fatty infiltration and inflammatory hepatitis can also contribute to this inhomogeneous enhancement pattern (15). In some cases, abnormal parenchymal enhancement may be the only spiral CT abnormality demonstrated in the cirrhotic liver. Spiral scanning can also provide reliable demonstration of the hepatic vasculature into the outer third of the normal liver parenchyma. In patients with cirrhosis, however, the peripheral intrahepatic vasculature is often attenuated (figure 3). This results in a "pruned tree" appearance similar to that described in the lung for patients with pulmonary hypertension.

At times, there can be confusion differentiating the abnormal parenchymal enhancement of cirrhosis from that seen with diffuse or multifocal hepatoma, as well as from multiple regenerating nodules. Clues to the diagnosis of tumor include focal contour abnormality, mass effect upon the hepatic vasculature, and vascular invasion or thrombosis (figure 4). A quality spiral CT with bolus contrast enhancement can provide a reliable, and perhaps more importantly, reproducible method for follow-up in these difficult cases.

Portal hypertension is a frequent complication of cirrhosis. It is related to increased resistance to vascular outflow. Spiral scanning can nicely demonstrate varices and portosystemic collateral pathways. These dilated and tortuous vessels appear as brightly enhancing tubular structures. Common locations include the lower esophagus, gastrohepatic ligament, and splenic hilum (figures 5 and 6). Esophageal varices are by far the most important clinically, present in about 65% of patients with advanced cirrhosis. They are a cause of massive hematemesis and death in half of these patients (8). If the diagnosis of cirrhosis is suspected prior to scanning, beginning the spiral scan slightly cephalad to the dome of the liver will help detect distal esophageal varices. Some cases of portal hypertension result in recanalization of the paraumbilical veins. These can be seen coursing in the region of the falciform ligament, before extending medially and inferiorly to the umbilicus (figure 7). Retroperitoneal and mesenteric collaterals can be demonstrated in cases of severe portal hypertension (figure 8). The ability to scan the entire liver during peak contrast enhancement can significantly improve detection of vascular collaterals which may otherwise be mistaken for lymph nodes. This is especially true in the gastrohepatic ligament, mesenteric root and para-aortic regions. Furthermore, multiplanar reconstruction of the spiral dataset results in exquisite display of the portosystemic collaterals, free from respiratory misregistration seen with conventional dynamic CT (figure 9) (12,13,14,15).

Portal vein thrombosis can complicate cirrhosis, infection or trauma. It can also be seen in patients with a hypercoaguable state due to polycythemia, underlying malignant neoplasm, use of birth control pills, etc., or can result from direct extrinsic compression due to adjacent or invading tumor (20). The spiral scan can frequently help detect the underlying cause of the thrombosis. The portal vein demonstrates central lower density tumoral or non-tumoral thrombus (figures 10,11). Uniform, peak contrast opacification makes spiral CT very useful not only in demonstration of main portal venous thrombosis, but also in the detection of small, dilated periportal collateral vessels, the so-called "cavernous transformation" (figures 12,13). Excellent contrast enhancement has also increased our ability to detect previously unsuspected, relatively small thrombi in the portal vein and major mesenteric veins (figures 14,15). Similar capabilities with spiral CT have been demonstrated in the evaluation of pulmonary emboli (21). The significance of these small portal thrombi has yet to be clinically determined. Anecdotally, we have seen one patient with vague abdominal pain and isolated, minimal portal vein thrombosis respond to anticoagulation therapy (figure 15).

Fatty Infiltration

Fatty liver results from the reversible accumulation of triglycerides. Common etiologies include obesity, excessive alcohol consumption, starvation, steroid administration, total parenteral nutrition, pregnancy, chemotheraputic agents, and a multitude of other causes (8). Infiltration can be diffuse or focal. In diffuse infiltration (figure 16), the liver appears lower in density than the nonenhanced hepatic blood vessels (22). Focal fatty infiltration results in a geographic region of low attenuation. Focal deposits can be confused with parenchymal mass lesions. This commonly occurs adjacent to the falciform ligament or near the gallbladder fossa, probably reflecting a separate vascular supply to these areas (figure 17) (23,24,25,26). Fatty change likely occurs here because of regional tissue hypoxia (24). Because scanning is performed during peak vascular enhancement, spiral CT can readily demonstrate normal caliber vessels coursing through a hypodense lesion. This can confidently establish the diagnosis of fatty infiltration, and avert the need to perform additional confirmatory studies with ultrasound or MRI.

Hepatitis

Viral hepatitis remains a common disease. Acute hepatitis has a spectrum of clinical findings ranging from minor symptoms to fulminate hepatic failure. Chronic hepatitis exists when clinical or biochemical evidence of hepatic inflammation is present for at least 3 to 6 months. Chronic active hepatitis (CAH) is a severe form which often leads to cirrhosis and liver failure. In fulminate hepatic necrosis associated with acute hepatitis, collapse of the periportal reticulin framework, as well as bile duct proliferation, result in enlarged periportal spaces. These widened spaces can be seen on CT as diffuse periportal low attenuation (27). This finding can also be seen in patients with hepatic trauma, congestive heart failure, liver transplant, cirrhosis, or malignant periportal or retroperitoneal adenopathy (27,28,29,30,31). In these conditions, periportal hypodensity reflects either the presence of blood, dilated lymphatics, or portal inflammation (27). Spiral CT accentuates the periportal low attenuation due to peak contrast enhancement in the portal venous system (figure 18). Some patients with hepatitis demonstrate subtle, irregular parenchymal enhancement following bolus injection (figure 19). This is different from the more coarse enhancement seen in patients with cirrhosis, and likely reflects diffuse parenchymal swelling resulting from cellular edema. However, a similar inhomogeneous pattern can also be seen in patients with heart failure and hepatic congestion, and correlation with serum titers is recommended in the appropriate clinical setting.

Radiation hepatitis has an onset 2 to 6 weeks after the completion of radiotherapy, and consists of symptoms of hepatomegaly, ascites and jaundice. Threshold dose is in the range of 3,000 to 5,500 rads. Patients typically demonstrate sharply defined, geographic regions of low attenuation which correspond to the radiation port (32). The low attenuation corresponds to increased water content (33). Spiral CT can demonstrate normal caliber vessels coursing through the low attenuation lesion, much like that seen with focal fatty infiltration (figure 20).

Budd-Chiari Syndrome

Budd-Chiari syndrome results from hepatic vein occlusion. Etiologies include neoplasm, hypercoaguable states, medications, and trauma (8). Many cases are idiopathic. Spiral CT demonstrates non-opacification of the hepatic veins, with inhomogeneous parenchymal enhancement, enhanced lobules of the liver, and hepatomegaly (34,35,36,37). The caudate lobe is enlarged in chronic cases, and often demonstrates increased (probably "normal") enhancement compared to the periphery of the liver. The persistence of patchy peripheral enhancement 45 to 60 seconds after injection results from stagnant venous outflow (figure 21) (37). Some caution must be used in making the diagnosis of Budd-Chiari syndrome with spiral CT, as non-opacification of the hepatic veins occasionally occurs if scanning is performed in patients with poor cardiac output or heart failure (figure 22). In these patients, hepatic venous opacification may yet to have reached adequate levels while scanning is performed. Inhomogeneous parenchymal enhancement is an essential finding for the diagnosis of Budd-Chiari syndrome.

Venous Congestion

Passive hepatic congestion is a possible cause of mottled hepatic enhancement. The intrahepatic vascular hemodynamics become altered, resulting in relative stasis of blood in the hepatic sinusoids (38,39,40). Spiral CT, with rapid scan acquisition and peak contrast opacification, can reliably detect this transient inhomogeneity (figure 23). Differential diagnosis includes cirrhosis, Budd-Chiari syndrome, diffuse malignancy, fatty infiltration, and hepatitis. Confusion can be avoided by noting secondary signs of heart failure, such as dilatation of the inferior vena cava and hepatic veins, cardiomegaly, and lack of hepatic venous opacification detected during the spiral scan. These findings reflect decreased cardiac output and/or right heart failure.

Other Diffuse Diseases

A multitude of miscellaneous diffuse parenchymal diseases can affect the liver. These include glycogen storage disease, hemachromatosis, amiodarone toxicity, Wilson disease, and Thorotrast deposition (20). Patients with glycogen storage disease and Thorotrast exposure should be monitored for the development of malignancy. Spiral evaluation in these disorders will likely add little to findings demonstrated on conventional CT, except in cases of coexistent malignancy where tumor detection and vascularity can be reliably assessed (5).

Focal Parenchymal Disease

Abscess

Bacterial disease can affect the liver either through direct extension from the biliary tract in patients with cholangitis or cholecystitis, or through the portal venous system in patients with diverticulitis, appendicitis, or inflammatory bowel disease. Another common predisposing factor is prior surgery. A characteristic CT finding in bacterial abscess is a central area of low attenuation with contrast enhancement of a peripheral rim or capsule (figures 24,25) (41). Many abscesses have sharp external margins. Improved contrast enhancement with spiral CT can help define these characteristics. Furthermore, the virtual elimination of partial volume averaging provided by spiral scanning can aid in the detection of smaller lesions and in the demonstration of gas bubbles, which are seen with conventional CT in approximately 20% of cases. (figure 26) (20).

Fungal microabscesses, usually due to candidiasis, can be seen in immunocompromised hosts. Systemic candidiasis has also been described in patients following hyperalimentation or during pregnancy. The liver and spleen are usually involved due to hematogenous spread from other organs, especially the lungs. CT demonstrates multiple, very small lesions throughout the liver and spleen (42,43,44,45). Because of peak contrast enhancement and contiguous data acquisition, spiral CT can depict lesions as small as 2 mm.

Hepatic Infarcts

Causes of hepatic infarction include hypercoaguable state, neoplasm, trauma and interarterial chemotherapy. Infarction can be focal or diffuse. Focal infarcts classically demonstrate sharp, peripheral wedge-shaped hypodense lesions which reflect their vascular etiology, although rounded or centrally located lesions can be seen (46). Infarcts can also contain gas. Following liver transplantation, partial or complete occlusion of the hepatic artery can occur. This is a rare but catastrophic outcome. Spiral CT can help in evaluation of the liver transplant patient by reliably demonstrating the central and peripheral portions of the hepatic artery in most normal patients (16,17).

Focal Perfusion Defects

The ability to scan during peak contrast enhancement with spiral CT allows for detection of geographic perfusion defects usually resulting from central tumor encroachment upon portal or hepatic veins. These defects demonstrate sharp linear borders and a wedge-shaped configuration much like that seen with hepatic infarction. Demonstration of a central tumor mass as well as patency of the hepatic arterial vessels can differentiate a relative perfusion defect from infarct. Delayed scans are often helpful in discriminating tumor mass from perfusion defect.

We have seen several cases of focal increased attenuation adjacent to recently placed percutaneous biliary catheters (figure 27,28). Similar areas of hepatic hyperattenuation can be seen adjacent to subcapsular fluid collections (figure 29). This phenomenon likely represents a combination of relative hyperattenuation due to focal compressive effects upon the hepatic parenchyma, and post-traumatic or inflammatory hyperemia. The ability of spiral CT to scan the entire liver during maximum contrast enhancement has increased our ability to detect these very subtle regions of abnormal perfusion, and care must be taken in not confusing them with true hepatic pathology. Their geographic and focal nature are the clues to accurate diagnosis.

References

1. Kalender WA, Polacin A. Physical performance characteristics of spiral CT scanning. Med Phys 1991; 18: 910-915.

2. Kalender WA, Seissler W, Klotz E, Vock P. Spiral volumetric CT with single-hold-technique, continuous transport, and continuous scanner rotation. Radiology 1990; 176: 181-183.

3. Vock P, Soucek M, Daepp M, Kalender WA. Lung: spiral volumetric CT with single-breath-hold technique. Radiology 1990; 176:864-867.

4. Zeman RK, Fox SH, Silverman PM, et al. Helical (Spiral) CT of the abdomen. AJR 1993; 160:719-725.

5. Bluemke DA, Fishman EK. Spiral CT of the liver. AJR 1993; 160:787-792.

6. Urban BA, Fishman EK, Kuhlman JE, Kawashima A, Hennessey JG, Siegelman SS. Detection of focal hepatic lesions with spiral CT: comparison of 4- and 8-mm interscan spacing. AJR 1993; 160:783-785.

7. Bluemke DA, Fishman EK. Spiral CT arterial portography of the liver. Radiology 1993; 186:576-579.

8. Robbins SL, Cotran RS, Kumar V. Pathologic basis of disease. 3rd Ed. Philadelphia: W.B. Saunders, 1984; 884-942.

9. Harbin WP, Robert NJ, Ferrucci JT. Diagnosis of cirrhosis based on regional changes in hepatic morphology. Radiology 1980; 135:273-283.

10. Popper H, Elias H, Petty DE. Vascular pattern of cirrhotic liver. Am J Clin Pathol 1952;22:717-729.

11. Popper H. Pathologic aspects of cirrhosis. A review. Am J Pathol 1977;87:228-264.

12. Huet PM, DuReau A, Marleau D. Arterial and portal blood supply in cirrhosis: a functional evaluation. Gut 1979;20:792-796.

13. Huet PM, Marleau D, Lavoie P, Viallet A. Extraction of I-125 albumin microaggregates from portal blood. An index of functional portal blood supply in cirrhotics. Gastroenterology 1976;70:74-81.

14. Groszman RJ, Kravetz D, Parysow O. Intrahepatic arterio-venous shunting in cirrhosis of the liver. Gastroenterology 1977;73:201-204.

15. Mulhern CB, Arger PH, Coleman BG, et al: Nonuniform attenuation in computed tomography study of the cirrhotic liver. Radiology 1979;132:399-402.

16. Rubin GD, Dake MD, Napel SA, McDonnell CH, Jeffrey RB. Three-dimensional spiral CT angiography of the abdomen: initial clinical experience. Radiology 1993; 186:147-152.

17. Ney DR, Fishman EK, Niederhuber JE. Three-dimensional display of hepatic venous anatomy generated from spiral computed tomography data: preliminary results. J Digit Imaging 1992; 5:242-245.

18. Napel S, Marks MP, Rubin GD, et al. CT angiography with spiral CT and maximum intensity projection. Radiology 1992; 185:607-610.

19. Schwartz RB, Jones KM, Chernoff DM, et al. Common carotid artery bifurcation: evaluation with spiral CT. Work in progress. Radiology 1992; 185:513-519.

20. Lee JKT, Sagel SS, Stanley RJ. Computed body tomography with MRI correlation. 2nd Ed. New York: Raven Press, 1989; 627-639.

21. Remy-Jardin M, Remy J, Wattinne L, Giraud F. Central pulmonary thromboembolism: diagnosis with spiral volumetric CT with the single-breath-hold technique--comparison with pulmonary angiography. Radiology 1992; 185:381-387.

22. Halvorsen RA, Korobkin M, Ram PC, et al: CT appearance of focal fatty infiltration of the liver. AJR 1982;139:277-281.

23. Adkins MC, Halvorsen RA, duCret RP. CT evaluation of atypical fatty metamorphosis. J Comput Assist Tomogr 1990;14:1013-1015.

24. Yoshikawa J, Matsui O, Takashima T, et al: Focal fatty change of the liver adjacent to the falciform ligament: CT and sonographic findings in five surgically confirmed cases. AJR 1987;149:491-494.

25. Brawer MK, Austin GE, Lewin KJ. Focal fatty change of the liver, a hithero poorly recognized entity. Gastroenterology 1980;78:247-252.

26. Kawashima A, Suehiro S, Murayama S, Russel WJ. Focal fatty infiltration of the liver mimicking a tumor: sonographic and CT features. J Comput Assist Tomogr 1986;10:329-331.

27. Siegel MJ, Herman TE. Periportal low attenuation at CT in childhood. Radiology 1992; 183:685-688.

28. Marincek B, Barbier PA, Becker CD, et al: Appearance of impaired lymphatic drainage in liver transplants. AJR 1986;147:519-523.

29. Dupuy D, Costello P, Lewis, D, et al: Abdominal CT findings after liver transplantation in 66 patients. AJR 1991;156:1167-1170.

30. Aspetrand F, Schrumpf E, Jacobsen M, Hanssen L, Endresen K. Increased lymphatic flow from the liver in different intra- and extrahepatic diseases demonstrated by CT. J Comput Assist Tomogr 1991;15:550-554.

31. Koslin DB, Stanley RJ, Berland LL, Shin MS, Dalton SC. Hepatic perivascular lymphedema: CT appearance. AJR 1988;150:11-113.

32. Jeffrey RB, Moss A, Quivey JM, et al: CT of radiation-induced hepatic injury. AJR 1980;135:445-448.

33. Unger EC, Lee JKT, Weyman PJ. CT and MR imaging of radiation hepatitis. J Comput Assist Tomogr 1987;11:264-268.

34. Rossi P, Sposito M, Simonetti G, Sposato S, Cusumano G. CT diagnosis of Budd-Chiari Syndrome. J Comput Assist Tomogr 1981;5:366-369.

35. Harter LP, Gross BH, Hilaire JS, Filly RA, Goldberg HI. CT and sonographic appearance of hepatic vein obstruction. AJR 1982;139:176-178.

36. Vogelzang RL, Anschuetz SL, Gore RM. Budd-Chiari syndrome: CT observations. Radiology 1987;163:329-333.

37. Mathieu D, Vasile N, Menu Y, et al: Budd-Chiari syndrome: dynamic CT. Radiology 1987;165:409-413.

38. Mauro MA, Stackhouse DJ, Parker LA, Schiebler ML. Computed tomography of hepatic venous hypertension: the reticulated-mosaic pattern. Gastrointest Radiol 1990;15:35-38.

39. Holley HC, Koslin DB, Berland LL, et al: Inhomogeneous enhancement of liver parenchyma secondary to passive congestion: contrast-enhanced CT. Radiology 1989;170:795-800.

40. Moulton JS, Miller BL, Dodd III, GD. Passive hepatic congestion in heart failure: CT abnormalities. AJR 1988;151:939-942.

41. Terrier F, Becker CD, Triller JK. Morphologic aspects of hepatic abscesses at computed tomography and ultrasound. Acta Radiol Diagn 1983;24:129-137.

42. Pastakia B, Shawker TH, Thaler M, O'Leary T, Pizzo PA. Hepatosplenic candidiasis: wheels within wheels. Radiology 1988;166:417-421.

43. Shirkhoda A. CT findings in hepatosplenic and renal candidiasis. J Comput Assist Tomogr 1987;11:795-798.

44. Ho B, Cooperberg PL, Li DKB, Mack L, Naiman SC, Grossman L. Ultrasonography and computed tomography of hepatic candidiasis in the immunocompromised patient. J Ultrasound Med 1982;1:157-159.

45. Callen PW, Filly RA, Marcus FS. Ultrasonography and computed tomography in the evaluation of hepatic microabscesses in the immunocompromised patient. Radiology 1980;136:433-434.

46. Kev-Toaff AS, Friedman AC, Cohen LM, Radecki PD, Caroline DF. Hepatic infarcts: new observations by CT and sonography. AJR 1987;149: 87-90.