Detection and staging of malignant liver tumors by CT and MRI.
Prof. Dr. W. Schima.
Department of Radiology, University of Vienna, Austria.
(Радиология в медицинской диагностике [современные технологии] 2003: 54-60)

Abstract
Liver metastases are much more common than primary malignant tumors of the liver and diagnosis is largely based on imaging findings. Thus, the radiologist plays a key role in the management of patients with suspected liver metastases by accurately detecting or excluding liver metastases. In patients with HCC, treatment strategies (e.g., resection, liver transplantation, ethanol injection or radiofrequency ablation) largely depend on the number and size of lesions and the clinical stage of cirrhosis. Thus, imaging guides further therapy by assessment of tumor burden. Different imaging techniques can be used for this purpose.
Sonography is a widely available and low-cost examination technique, but it has limitations in the detection of metastases, with an overall sensitivity of approximately 50%. Preliminary experience with contrast-enhanced sonography is promising. This technique is able to depict lesions smaller than 1 cm more accurately. The standard imaging technique for staging in patients with extrahepatic neoplasms at high risk for developing liver metastases is contrast-enhanced helical CT, with a sensitivity of 58–85%. The development of multi-detector array CT has recently expanded CT to the “third dimension” by providing excellent 3D imaging capability. CT during arterial portography, which has been the “gold standard” of preoperative liver imaging, has largely been abandoned because of its invasiveness and its high rate of false-positive diagnoses (13–31%) due to tumur-like perfusion defects. MR imaging with liver-specific contrast agents has been shown to be at least as accurate as CTAP, with a sensitivity of 85-90%. For assessment of HCC, gadolinium-enhanced MRI is at least as useful as MRI with liver-specific contrast agents. In many institutions, contrast-enhanced MR imaging has become the procedure of choice for preoperative assessment. The diagnostic strategy for assessing liver metastases must take into account (1) the histology of the neoplasm (i.e., primary versus metastasis), (2) the clinical situation, and (3) the planned therapeutic procedure.

Introduction.
Liver metastases are one of the most common problems in the evaluation of oncologic patients. Liver metastases are more than 10 times more common than primary malignant tumors of the liver. A variety of imaging techniques, including sonography, computed tomography (CT), magnetic resonance (MR) imaging, angiography, and nuclear medicine are used for the assessment of metastatic disease to the liver. Metastases from colorectal cancer have been most extensively studied in the radiologic literature because of the possible surgical consequences.

Prevalence of Metastatic Disease.
The liver is the second most common site for metastatic spread of extrahepatic malignancies, second only to lymph node metastases. The prevalence of liver metastases varies considerably. Primary tumors in the gallbladder, pancreas, colon and rectum, breast, stomach, and lung, in particular, tend to metastasize to the liver.

The incidence of colorectal cancer ranges from 20 to 47 cases per 100,000 per year in the Western world. In up to 20% of patients, synchronous liver metastases are present at the time of diagnosis of the primary cancer. Another 20-30% of patients develop liver metastases during the course of their disease. In autopsy series of colon cancer patients, the liver was found to be involved in 65%.

Diagnostic Challenge for Imaging: Detection and Characterization of Lesions.
In a radiology department serving a surgical and internal medicine service, focal liver lesions are found at transabdominal ultrasound (US), CT, and MR imaging every day. These focal lesions are often seen at follow-up studies of patients with known metastatic malignant disease. Solid liver lesions may also be found in patients with a known malignant tumor without prior evidence of generalization. Focal liver lesions of unknown origin may be found as an incidental finding in patients who undergo imaging for suspected biliary disease or blunt abdominal trauma or in patients with abnormal liver function tests.
Autopsy studies of Karhunen et al. have shown that up to 50% of the population have benign liver lesions, such as hemangiomas, focal nodular hyperplasia, and hamartomas . These lesions are of no clinical relevance, unless they are incidentally found by imaging. In this case, characterization of focal lesions is of utmost importance. Thus, the diagnostic accuracy of imaging techniques in the detection of focal liver lesions should be supplemented by reliable and, if possible, noninvasive characterization.

In clinical practice, characterization of focal liver lesions is achieved with sonography, contrast-enhanced CT, or MR imaging. Sonography is very accurate in the differentiation between hemangioma and non-hemangiomatous lesions. However, sonographic visualization of focal lesions is often limited in the presence of diffuse liver disease. In a study on MR imaging of patients with newly diagnosed breast cancer, benign lesions were common in patients suspected clinically of having liver metastases. In 32% of patients, only benign lesions were found, and the diagnosis was made by MR imaging and corroborated by biopsy or follow-up.

Therefore, in patients with a history of malignant disease, sonography may be insufficient to make a reliable diagnosis of benign focal lesions. Some authors strongly recommend a corroborative study, such as contrast-enhanced CT, MR imaging, or biopsy, to confirm a diagnosis of benign disease.

Contrast-enhanced multi-phasic CT allows differentiation between metastases and hemangiomas with a high degree of certainty. Hemangiomas typically show peripheral globular enhancement with progressive centripetal fill-in of the lesion with contrast material, whereas metastases show irregular rim enhancement, early arterial enhancement (e.g., hypervascular metastases or HCC), or only minimal enhancement (e.g., hypovascular metastases from colon cancer). However, contrast-enhanced CT has some limitations for the characterization of lesions smaller than 2 cm and for lesions in a fatty liver.
In recent years, contrast–enhanced MR imaging has been shown to be the method of choice for differentiation between hemangioma and metastases. The superior tissue contrast available with MR allows reliable differentiation between small hemangiomas and metastases. Soyer et al. have shown that gadolinium chelate-enhanced MR imaging even allows reliable differentiation between hemangiomas and hypervascular metastases from neuroendocrine tumors.

Helical CT and Multidetector Array CT (MDCT).
In clinical practice, CT is the mainstay of oncologic imaging. Advantages include excellent availability, good patient tolerance, and reproducibility of examinations. With the introduction of high-speed helical CT scanners in 1988, CT became capable of imaging the entire liver during one breath-hold of 25 sec. The dramatic shortening of acquisition time compared to dynamic CT scanners increased patient comfort and diagnostic value. The decrease in acquisition time also allowed to use higher injection rates of contrast material with better enhancement of the liver parenchyma and improved delineation of focal masses. With dynamic CT scanners, acquisition of a single scan begins in a portal-venous phase and ends in a late equilibrium phase, when only minimal reidual tumor-liver contrast is present. With rapid helical CT scanning, acquisition of scans in the arterial-dominant phase and the portal-venous phase have become possible, which is important for optimal detection of hypervascular and hypovascular liver metastases. The appearance of metastases varies based on the vascularity of the tumor and the timing of image acquisition. Many metastases (e.g., colon, gastric, and pancreatic cancer) are hypovascular compared to the background parenchyma and tend to be hypodense on contrast-enhanced scans. In contrast, renal cell cancer and neuroendocrine tumors tend to seed hypervascular tumors, which show strong enhancement in the arterial-dominant phase of contrast enhancement. Sometimes, both hypervascular and hypovascular metastases are seen in one patient. Likewise, HCC nodules are predominantly herpervascular, but hypo- and hypervascular lesions may be seen in the same patient. Therefore, a dual-phasic CT scan is always recommended as the initial study to optimize lesion detection and characterization. With helical CT scanning, the sensitivity for detection of liver metastases and HCC ranges from 53 to 89% overall, with a still modest sensitivity of 25–56% for lesions smaller than 1 cm.

However, multi-phasic helical CT scanning helps in the characterization of focal liver lesions, as many benign lesions, such as hemangiomas and FNH, have distinctly different contrast accumulation features.

In 1998, multi-detector array helical CT (MDCT) has been introduced in clinical practice. It is a technological advance that allows the simultaneous acquisition of multiple images (in contrast to single slice helical CT) during a single rotation of the X-ray tube. Recently, 16-row scanners have become available. A clear advantage of this technology is its ability to scan large volumes, such as the chest and abdomen, within one breath-hold. Moreover, thinner collimation with a slice thickness of 0.5-1 mm and the acquisition of isotropic voxels has become possible. This allows performance of multi-planar 3D-reformations in the coronal, sagittal, and oblique planes without loss of spatial resolution. Variations of arterial blood supply to the liver and of the hepatic or portal venous system are reliably depicted by MDCT. However, a limitation of the technique is the large number of images produced per study. With reconstruction of thin overlapping images and multi-planar reformations, the total number of images can easily reach 1000 per examination. For review and interpretation of these large data sets, a workstation is mandatory.

MR Imaging.
In the past few years, MR imaging has evolved as a very powerful tool in abdominal imaging. This success was driven by the development of high-field strength MR units (>=1.0T) with high-performance gradients (>=20 mT/m) and the availability of phased-array body coils. With these achievements, a breath-hold MR imaging technique is possible to minimize motion artifacts from breathing. The main advantage of MR imaging over CT is the better intrinsic tissue contrast for improved delineation of lesions. MR imaging has some limitations: the examination time is, even with breath-hold imaging, still longer than that of helical CT; availability is limited in many institutions; and examinations must be focused on a particular anatomic area.

Acquisition of MR pulse sequences with different weighting (i.e., T1-weighted, T2-weighted, STIR, etc.) allows the detection of diffuse alterations in the liver parenchyma in patients with fatty infiltration or iron overload. Moreover, characterization of some focal liver lesions, such as cysts, hemangiomas, and focal fatty infiltration, is often possible. Administration of gadolinium-chelate MR contrast agents may further improve the diagnostic yield. These non-specific MR contrast agents help to assess lesion vascularity during dynamic imaging, similar to that in multi-phasic helical CT scanning. In general, gadolinium-enhanced MRI does not improve detection of liver metastases, except in the case of primary tumors known to seed hypervascular metastases, but it helps in the differentiation between metastases and benign lesions, such as hemangioma or FNH.

Liver-specific MR contrast agents have been developed that are selectively taken up into the liver parenchyma but not into non-hepatocellular lesions such as metastases, thus increasing tumor conspicuity. Two classes of liver-specific MR contrast agents are available. Reticulo-endothelial agents are taken up by the Kupffer-cells in the liver parenchyma. This group comprises the small paramagnetic iron oxide particles (ferumoxide, Guerbet, France; and SHU 555 A, Schering, Germany). After IV administration of these agents, uptake into the parenchyma results in signal drop (“blackening”) of the parenchyma with improved conspicuity of metastases and HCC. The second group comprises the hepato-biliary agents, which are T1-enhancing agents (mangafodipir trisodium [formerly known as Mn-DPDP], Amersham Health, Norway; Gd-BOPTA, Bracco, Italy; and Gd-EOB-DTPA, Schering [not yet available]). After IV administration, these agents are selectively taken up into the hepatocytes and excreted in the bile. These agents increase the signal intensity of normal liver, whereas metastases appear dark compared with liver tissue. Thus, this results in a considerably improved delineation of liver metastases on post-contrast images. These agents can also be used to reliably differentiate between metastases and incidentally found benign lesions, such as FNH.
Ferumoxide, mangafodipir, SHU 555A, and Gd-BOPTA are already approved for clinical use in most Western European countries and the USA. Recent studies have shown an improvement in the diagnostic yield of MRI after administration of liver-specific agents. Several studies have compared the diagnostic performance of contrast-enhanced MRI and helical CT for detection of colo-rectal liver metastases in surgical patients (Table 1). In most studies, MRI proved to be superior to helical CT in the detection of liver metastases. MRI is particularly helpful in the detection of small lesions, whereas helical CT has certain limitations. However, the economic impact of the use of liver-specific MR contrast agents with improved delineation of small metastases remains to be studied. At present, intraoperative US is still the most sensitive method. Even in patients, who had undergone state-of-the art imaging preoperatively, intraoperative US may alter the surgical approach in a substantial number of patients.

Table 1. Sensitivity of Contrast-enhanced Helical CT, CTAP, and MR Imaging for Detection of Liver Metastases.

Conclusions:
In patients with extrahepatic malignancies at high risk for developing liver metastases, contrast-enhanced helical CT is the standard technique for preoperative evaluation and follow-up. Multi-detector array CT is superior to single-detector helical CT because of the improved visualization of small metastases and HCC with thinner collimation and 3D-imaging of liver anatomy. MR imaging with liver-specific contrast agents is more accurate than helical CT for detection of liver metastases and HCC. If available, it should be used for preoperative evaluation of all surgical candidates.

References:
(1) Ba-Ssalamah A, Heinz-Peer G, Schima W, Schibany N, Schick S, Noebauer I, Lechner G: Detection and characterization of hepatic lesions: unenhanced and Resovist-enhanced MR imaging versus biphasic helical CTAP. J Magn Reson Imaging 2000;11:665-672.
(2) Baker ME, Pelley R: Hepatic metastases: basic principles and implications for Radiologists. Radiology 1995;197:329-337.
(3) Bennett GL, Petersein A, Mayo-Smith WW, Hahn PF, Schima W, Saini S: Addition of gadolinium chelates to heavily T2-weighted MR imaging: Limited role in differentiating hepatic hemangiomas from metastases. AJR 2000;174:477-485.
(4) Bluemke DA, Paulson EK, Choti MA, DeSena S, Clavien PA: Detection of hepatic lesions in candidates for surgery: comparison of ferumoxides-enhanced MR imaging and dual-phase helical CT. AJR 2000;175:1653-1658.
(5) Caudana R, Morana G, Pirovano GP, Nicoli N, Portuese A, Spinazzi A, Di Rito R, Pistolesi GF: Focal malignant hepatic lesions: MR imaging enhanced with gadolinium benzyloxypropionictetra-acetate (BOPTA)- preliminary results of phase II clinical application. Radiology 1996;199:513-520.
(6) Heiken JP, Weyman PJ, Lee JKT, Balfe DM, Picus D, Brunt EM, Flye MW: Detection of focal hepatic masses: prospective evaluation with CT, delayed CT, CT during arterial portography, and MR imaging. Radiology 1989;171:47-51.
(7) Helmberger T, Gregor M, Holzknecht N, Gauger J, Rau H, Reiser MF: Detektion und Charakterisierung fokaler Leberlasionen. Vergleich von biphasicher kontrastmittelverstarkter Spiral-CT, nativer und SPIO-verstarkter MRT. Radiologe 1999;39:678-684.
(8) Karhunen PJ: Benign hepatic tumours and tumour like conditions in men. J Clin Pathol 1986;39:183-188.
(9) Kulinna C, Helmberger T, Kessler M, Reiser M: Verbesserung der Diagnostik von Lebermetastasen mit der Multi-Detektor-CT. Radiologe 2001;41:16-23.
(10) Kuszyk BS, Bluemke DA, Urban BA, Choti MA, Hruban MA, Sitzmann JV, Fishman EK: Portal-phase contrast-enhanced helical CT for the detection of malignant hepatic tumors: sensitivity based on comparison with intraoperative and pathologic findings. Am J Roentgenol 1996;166:91-95.
(11) Lencioni R, Donati F, Cioni D, Paolicchi A, Cicorelli A, Bartolozzi C: Detection of colorectal liver metastases: prospective comparison of unenhanced and ferumoxides-enhanced magnetic resonance imaging at 1.5 T, dual-phase spiral CT, and spiral CT during arterial portography. MAGMA 1998;7:76-87.
(12) Noone TC, Semelka RC, Balci NC, Graham ML: Common occurrence of benign liver lesions in patients with newly diagnosed breast cancer investigated by MRI for suspected liver metastases. J Magn Reson Imaging 1999;10:165-169.
(13) Reimer P, Jahnke N, Fiebich M, Schima W, Deckers F, Marx C, Holzknecht N, Saini S: Hepatic lesion detection and characterization: Value of nonenhanced MR imaging, superparamagnetic iron oxide-enhanced MR imaging, and spiral CT-ROC analysis. Radiology 2000;217:152-158.
(14) Said B, McCart JA, Libutti SK, Choyke PL: Ferumoxide-enhanced MRI in patients with colorectal cancer and rising CEA: surgical correlation in early recurrence. Magn Reson Imaging 2000;18:305-309.
(15) Schima W, Saini S, Echeverri JA, Hahn PF, Harisinghani M, Mueller PR: T2-weighted MR imaging for characterization of focal liver lesions: conventional spin-echo vs fast spin-echo. Radiology 1997;202:389-393.
(16) Schmidt J, Strotzer M, Fraunhofer S, Boedeker H, Zirngibl H: Intraoperative ultrasonography versus helical computed tomography and computed tomography with arterioportography in diagnosisng colorectal liver metastases: lesion-by-lesion analysis. World J Surg 2000;24:43-48.
(17) Scott DJ, Guthrie JA, Arnold P, Ward J, Atchley J, Wilson D, Robinson PJ: Dual phase helical CT versus portal venous phase CT for the detection of colorectal liver metastases: correlation with intra-operative sonography, surgical and pathological findings. Clin Radiol 2001;56:235-242.
(18) Seneterre E, Taourel P, Bouvier Y, Pradel J, Van Beers B, Daures J-P, Pringot J, Mathieu D, Bruel J-M: Detection of hepatic metastases: ferumoxides-enhanced MR imaging versus unenhanced MR imaging and CT during arterial portography. Radiology 1996;200:785-792.
(19) Soyer P, Bluemke DA, Hruban RH, Sitzmann JV, Fishman EK: Hepatic metastases from colorectal cancer: detection and false-positive findings with helical CT during arterial portography. Radiology 1994;193:71-74.
(20) Valls C, Andia E, Sanchez A, Guma A, Figueras J, Torras J, Serrano T: Hepatic metastases from colorectal cancer: preoperative detection and assessment of resectability with helical CT. Radiology 2001;218:55-60.
(21) Ward J, Naik KS, Guthrie JA, Wilson D, Robinson PJ: Hepatic lesion detection: comparison of MR imaging after administration of superparamagnetic iron oxide with dual-phase CT by using alternative-free response receiver operating characteristic analysis. Radiology 1999;210:459-466.
(22) Wong K, Paulson EK, Nelson RC: Breath-hold three-dimensional CT of the liver with multi-detector row helical CT. Radiology 2001;219:75-79.