Differentiation Between Benign and Malignant Hepatic Lesions

Transcription

1 Article Differentiation Between Benign and Malignant Hepatic Lesions Utility of Color Stimulated Acoustic Emission With the Microbubble Contrast Agent Levovist Alexandra von Herbay, MD, Christoph Vogt, MD, Dieter Häussinger, MD Objective. This study was undertaken to determine whether the examination of color stimulated acoustic emission in the late phase of Levovist (SH U 508A; Schering AG, Berlin, Germany) enhancement is helpful in the discrimination between benign and malignant liver lesions. Methods. Fifty-six patients with focal hepatic lesions were examined. Diagnosis of the lesions was confirmed by liver biopsy, computed tomography, or scintigraphy. Thirty-one patients had malignant liver lesions: hepatocellular carcinoma (n = 14), cholangiocellular carcinoma (n = 1), metastasis (n = 14), and lymphoma (n = 2). Twenty-five patients had benign lesions: focal nodular hyperplasia (n = 8), hepatic adenoma (n = 1), focal hyposteatosis or hypersteatosis (n = 6), hemangioma (n = 7), and regenerative cirrhotic nodules (n = 3). After a delay of 5 to 10 minutes without scanning, the liver was examined by color stimulated acoustic emission with a fast sweep of 1 to 3 seconds. Results. All patients with homogeneous color stimulated acoustic emission in the late phase of Levovist enhancement had benign liver lesions (P <.001; specificity, 100%; sensitivity, 68%; positive predictive value, 100%; and negative predictive value, 79%). Eighty-one percent of the patients with nonenhancing lesions in the late phase surrounded by enhanced liver parenchyma had malignant liver lesions (P <.001; specificity, 72%; sensitivity, 94%; positive predictive value, 81%; and negative predictive value, 90%). Interobserver agreement (weighted κ value) improved from ± for baseline sonography to ± for color stimulated acoustic emission sonography. The area under the receiver operating characteristic curves for color stimulated acoustic emission sonography (0.927) was significantly higher than for baseline sonography (0.739; P <.05). Conclusions. Color stimulated acoustic emission in the late phase of Levovist enhancement has a high specificity and sensitivity for differentiation between benign and malignant focal liver lesions. Key words: adenoma; contrast agent; focal nodular hyperplasia; hepatocellular carcinoma; liver; oncology; sonography; stimulated acoustic emission. Abbreviations A z, area under the fitted curve; CT, computed tomography; FNH, focal nodular hyperplasia; HCC, hepatocellular carcinoma; LABMRMC, Laboratory Multiple Readers Multiple Cases; MI, mechanical index; ROC, receiver operating characteristic; SAE, stimulated acoustic emission Received December 2, 2002, from the Department of Medicine and Department of Hepatology, Gastroenterology, and Infectiology, University of Düsseldorf, Düsseldorf, Germany. Revision requested January 21, Revised manuscript accepted for publication August 26, We thank Reinhard Willers, MD, for excellent statistical assistance. This study was supported by the Freunde und Förderer der Heinrich-Heine- Universität Düsseldorf and by Deutsche Krebshilfe. Address correspondence and reprint requests to Alexandra von Herbay, MD, Department of Medicine and Department of Hepatology, Gastroenterology, and Infectiology, University of Düsseldorf, Moorenstrasse 5, D Düsseldorf, Germany. Levovist (SH U 508A; Schering AG, Berlin, Germany) is a microbubble contrast agent that causes systemic Doppler enhancement after intravenous injection. 1 7 Levovist has 2 different effects. In the first 2 minutes after bolus injection (early phase), enhancement is seen within blood vessels. In the late phase (starting 2 3 minutes after bolus injection and persisting for >30 minutes), Levovist accumulates in the hepatic and splenic parenchyma, probably because of phagocytosis by reticuloendothelial cells. 1,2,8 12 The use of color Doppler sonography with a high acoustic intensity results in rupture of microbubbles in the liver parenchyma. This causes a short intensive flash of color pixels, called stimulated acoustic emission (SAE), visualized as a color mosaic pattern. 2,8,13,14 Color SAE does not show blood flowing in vessels; it results from rupture of station by the American Institute of Ultrasound in Medicine J Ultrasound Med 23: , /04/$3.50

2 Color Stimulated Acoustic Emissions in Hepatic Lesions ary phagocytized microbubbles Because metastases lack the color SAE, absence of color SAE might be diagnostic for liver metastases. 8,18,19 We evaluated the diagnostic sensitivity and specificity of delayed color SAE imaging with the combined use of the color Doppler mode and a microbubble contrast agent (Levovist) in differentiation of benign from malignant focal hepatic lesions. Materials and Methods The protocol was approved by the Institutional Review Board at our hospital. Oral informed consent according to the ethical guidelines of the Declaration of Helsinki was obtained from all patients after they were informed about the purpose of the study before the sonographic examination was started. Patients In a prospective study, 56 patients (28 men and 28 women; mean age ± SD, 57 ± 14 years) were examined. The study patients were selected from among those patients undergoing abdominal sonographic examinations. For practical reasons, we could not include every patient with focal liver lesions from the large number of patients undergoing sonography in our department. Selection criteria were patients with a final diagnosis based on histologic examination, computed tomography (CT), magnetic resonance imaging, or scintigraphy and patients with a similar appearance of all lesions to minimize the possibility of different etiologies of lesions in the examined livers. The final diagnosis for each patient was confirmed by liver biopsy (36 lesions), CT (16 lesions: 5 hemangiomas, 6 hypersteatosis or hyposteatosis, and 5 metastases), or scintigraphy (4 lesions: all focal nodular hyperplasia [FNH]), performed as part of the clinical workup of the patients. The mean size of the individual lesions was 42 ± 25 mm. All patients had focal liver lesions detectable on conventional sonography. In 31 patients, malignant liver lesions were diagnosed: hepatocellular carcinoma (HCC; n = 14), cholangiocellular carcinoma (n = 1), metastatic liver disease (n = 14), and lymphoma (n = 2). In the remaining 25 patients, benign lesions were diagnosed: FNH (n = 8), hepatic adenoma (n = 1), focal fat deposition or focal spared lesion in fatty liver (n = 6), hemangioma (n = 7), and regenerative cirrhotic nodules (n = 3). The sonographers were unaware of the final diagnosis. Contrast Agent Levovist is a sonographic contrast agent consisting of galactose microparticles (99.9%), air bubbles, and palmitic acid (0.1%). The Levovist suspension has to be prepared by mixing with water, agitation for 10 seconds, and then standing for 2 minutes. During this procedure, multiple small stabilized air microbubbles with a mean diameter of 2 to 5 µm are produced (97% <7 µm), which are sufficiently small and stable to pass through capillary beds. 3,15 Levovist injection was performed intravenously through a needle having a diameter of at least 21 gauge. Smaller canules or injections through the ventile of the needle were not used to avoid destruction of the Levovist microbubbles under these conditions. 20 Sonographic Technique All sonographic examinations were performed with a 3.5-MHz curved transducer (Sonoline Elegra; Siemens AG, Erlangen, Germany). First, the liver was screened for focal lesions in the native B-mode and the native tissue harmonic imaging mode. The number and location of focal lesions were documented. Power Doppler sonography was performed to check for intralesional vessels. Only patients with focal liver lesions with a distance of less than 8 cm between the transducer and the lesions were included in the study because the acquired acoustic intensity to produce color SAE is limited to locations not deeper than 8 cm. Second, the equipment settings for color SAE were adjusted in the color Doppler mode with the preset contrast burst using a frequency of 2.5 to 3.5 MHz, parallel processing turned on, maximal pulse repetition frequency, maximal color acoustic power, maximal frame rate, persistence off, and a high mechanical index (MI > 1.4). The color gain was set below the noise floor of the system so that no flow signals from blood vessels were displayed. Because the color SAE mosaiclike pattern shows a tendency to cluster at the focal zone, where the power is maximal, a color box was used to focus the color SAE effect in the region of interest. 2 The focal zone was localized on the color box. After preparation of the color SAE settings, 2 g of Levovist (300 mg/ml, 1 ml/s) was injected intravenously as a bolus, followed by 10 ml of 0.9% saline. After a 208 J Ultrasound Med 23: , 2004

3 von Herbay et al delay of 5 minutes without scanning, the liver was examined. 2,16,17 The scans were performed according to the following schedule: the focal liver lesion was localized in the native gray scale B-mode with only 1% to 10% transit power (MI < 0.4) to protect early microbubble destruction. Thereafter, the insonation was interrupted, and the color box was localized in projection of the visualized lesion. The patient was asked to perform breath holding after deep inspiration. Then the transmit power was set to 100% (MI > 1.4), and the scan was started cranial to the liver lesion, followed by a fast sweep of 1 to 3 seconds through the liver. Analysis of each scan was performed on a digital stored frame-by-frame cine loop. All examinations were stored on super VHS videotapes and magneto-optical disks. The analysis was performed by the consent of 2 physicians experienced in sonography. In all patients with intralesional vessels found on power Doppler sonography associated with a homogeneous color mosaic pattern 5 minutes after Levovist injection, the color SAE sweep was repeated again 10 minutes after Levovist injection to rule out the possibility that hypervascularization of focal liver lesions could be the reason for prolongation of the blood pool phase of the Levovist microbubbles, making the color SAE effect false-positive. Review of the examinations was performed according to the following parameters: a homogeneous color pixel mosaic in the focal lesion and the surrounding liver or clear demarcation of the lesion without color pixels surrounded by a homogeneous color pixel mosaic of the surrounding liver and an inhomogeneous color pixel mosaic in the liver, which makes classification impossible. Image Analysis The analysis of the baseline sonograms and the contrast-enhanced studies was performed by 2 independent observers. Both observers evaluated the videotapes and the digital stored images. The videotapes were coded by consecutive numbers so that the reviewers did not know the patients names and patients diagnoses at the time of the analysis. To interpret the results, a 5-point grading system was used for baseline and contrast-enhanced images as follows: 1, definitely benign; 2, probably benign; 3, indeterminate; 4, probably malignant; and 5, definitely malignant. On the basis of prior experience, lesions with homogeneous color pixel enhancement were given benign ratings, whereas lesions without color pixel enhancement or with inhomogeneous color pixel enhancement were given higher ratings for malignancy. Lesions that could not be classified as benign or malignant were classified as score 3. The sonographer was 1 of the 2 physicians experienced in sonography who performed the review of the examination. In addition to the 5-point-scale, contrastenhanced-images were classified according to the following features: A, homogeneous color pixel enhancement; B, dark nodules without color pixel enhancement surrounded by bright enhancement of the liver; and C, inhomogeneous color pixel enhancement. The 2 readers performed this classification by consensus reading after they had made their independent blinded ratings on the 5-point scale. Statistical Analysis Analysis of our data was made on a by-patient basis. In patients with multiple lesions in whom 1 lesion was diagnosed by biopsy of the liver (n = 36), the diagnostic lesion was chosen for analysis. In the remaining patients, the observer chose a lesion that was visible without difficulty at baseline to guarantee that the contrast enhancement of the lesion could be evaluated. Patients with multiple lesions that appeared different were excluded from the study. A mean rating score was computed for each lesion as the average of the ratings provided by the 2 independent readers. For the calculation of sensitivity and specificity, a score of 3 or higher was counted as a positive test result. Our rationale for treating a score of 3 (indeterminate) as malignant was to avoid missing any malignant lesions. In addition, unweighted and weighted κ statistics were calculated on the basis of a 5 5 table of observer ratings for the comparison between the independently performed reviews of baseline sonography and contrast-enhanced sonography. 15,16 The κ analysis was performed with SAS 8.2/FREQ software (SAS Institute Inc, Cary, NC). The linear weighted κ statistic was weighted according to the interval between the rating scores (5-point system) of both reviewers. Receiver operating characteristic (ROC) curves were obtained for baseline and contrastenhanced sonography according to the 5-point system described above. Comparison of ROC curves was performed with Laboratory Multiple Readers Multiple Cases (LABMRMC) software J Ultrasound Med 23: ,

4 Color Stimulated Acoustic Emissions in Hepatic Lesions (University of Chicago, Chicago, IL). 17,18 The LABMRMC algorithm uses jackknifing and analysis of variance techniques to test the difference between cases and readers. P <.05 was considered significant. Results Analysis of Power Doppler Sonography Fifty-eight percent of the patients with malignant lesions and 48% of the patients with benign lesions had intralesional vascularization found on power Doppler sonography. There was no significant difference in the prevalence of intralesional vascularization between both groups. Analysis of Color SAE Enhancement A consensus classification by both reviewers of the enhancement patterns in our lesions is summarized in Table 1. Most lesions (53 of 56) were classified as clearly demarcated without enhancement (group B, 36 lesions) or homogeneously enhancing with the surrounding liver (group A, 17 lesions). Among the 3 patients who had lesions with inhomogeneous enhancement (group C), 2 had hepatoma and 1 had hemangioma. All lesions that showed homogeneous enhancement were benign (Figures 1 and 2; P <.001). However, in 4 other patients with hypervascularized malignant liver lesions, Levovist enhancement was dependent on the period between Levovist injection and SAE examination. Although these 4 patients had Levovist enhancement 5 minutes after injection, we observed no Levovist enhancement when the color SAE examination was performed again 10 minutes after injection. Because of this very striking observation, all patients with hypervascularized lesions and the finding of homogeneous enhancement 5 minutes after Levovist injection were reexamined by color SAE 10 minutes after Levovist injection. In all 9 of these benign hypervascularized lesions (FNH, n = 8; and adenoma, n = 1), the homogeneous Levovist enhancement was still present 10 minutes after injection. The finding of homogeneous Levovist enhancement as a sign of benign lesions, when used in the procedure described, had a specificity of 100%, a sensitivity of 68%, a positive predictive value of 100%, and a negative predictive value of 79% (Table 1 and Figures 1 and 2). Table 1. Color SAE in the Late Phase of Levovist Enhancement SAE Homogeneous Inhomogeneous Power Clear Demarcation Color Pixel Enhancement Doppler Without Color Pixels Enhancement In the Liver Intralesional (5 10 min After (5 10 min After (No Classification Diagnosis n Vessels Levovist Injection)* Levovist Injection)* Possible) Malignant focal liver lesions HCC * 0 2 Cholangiocellular carcinoma Metastases * 0 0 Lymphoma Benign focal liver lesions FNH * 0 Adenoma * 0 Regenerative cirrhotic nodule Focal hypo/hypersteatosis Hemangioma Malignant lesions Benign lesions *Patients with hypervascularized lesions associated with an intralesional color pixel mosaic at 5 minutes after Levovist injection were examined again at 10 minutes after Levovist injection (HCC, n = 3; metastasis, n = 1; FNH, n = 8; and adenoma, n = 1). Histologic examination of this regenerative nodule found dysplasia. Hemangioma with intralesional vessels had a diameter of 8 to 20 cm (giant hemangioma). P <.001, significantly different compared with patients with benign focal liver lesions. P <.001, significantly different compared with patients with malignant focal liver lesions. 210 J Ultrasound Med 23: , 2004

5 von Herbay et al A B Figure 1. Focal nodular hyperplasia. A, Native B-mode image showing a hypoechoic inhomogeneous lesion with a hypoechoic rim (arrows). B, Color SAE examination in the late phase of Levovist enhancement showing homogeneous Levovist enhancement in both the focal lesion (arrows) and the surrounding liver parenchyma. Additionally, a central scar became visible without color pixel enhancement. Among lesions that were clearly demarcated by their lack of enhancement, 81% were malignant (Figures 3 and 4; P <.001). The finding of clear demarcation without contrast enhancement surrounded by enhanced liver parenchyma as a sign of malignant lesions had a specificity of 72%, a sensitivity of 94%, a positive predictive value of 81%, and a negative predictive value of 90% (Table 1 and Figures 3 and 4). Comparison Between B-Mode Sonography and Color SAE Enhancement Baseline B-mode sonography detected 87% of malignant lesions (27 of 31 lesions) on the basis of a mean score of 3 or higher, with a specificity of 44% (11 of 25). Ten patients had lesions rated as indeterminate (mean score of 3). Contrastenhanced sonography detected 100% of malignant lesions (31 of 31 lesions) on the basis of a mean score of 3 or higher, with a specificity of 68% (17 of 25). Figure 2. Benign adenoma of the liver. A, Native B-mode image showing a hypoechoic lesion (arrows) in the right lobe of the liver. B, Color SAE examination in the late phase of Levovist enhancement showing homogeneous Levovist enhancement of the lesion. No difference could be seen between Levovist enhancement of the lesion and the surrounding liver parenchyma. A B J Ultrasound Med 23: ,

6 Color Stimulated Acoustic Emissions in Hepatic Lesions A B Figure 3. Metastatic liver disease. A, Native B-mode image showing a hypoechoic lesion in the left lobe of the liver (arrow). B, Color SAE examination in the late phase of Levovist enhancement showing clear demarcation of the hypoechoic lesion (arrow) in the left lobe of the liver without color pixels surrounded by homogeneous color pixel enhancement of the liver. A Interobserver agreement during baseline imaging was poor, with a weighted κ value of ± and a simple κ value of ± Interobserver agreement during contrastenhanced imaging improved markedly, with a weighted κ value of ± and a simple κ value of ± Receiver operating characteristic curves are presented in Figure 5. The area under the fitted curve (A z ) for contrastenhanced sonography (A z = 0.927) was significantly higher than the A z for baseline sonography (A z = 0.739; P <.05). Fourteen percent of the patients with metastases had additional lesions found on color SAE. Discussion The finding of a color pixel mosaic in a color SAE examination during the late phase of Levovist enhancement in benign hepatic lesions, such as FNH, adenoma, regenerative nodules, and focal steatosis, is probably due to the reticuloendothelial cells of these lesions. In contrast, metastases, HCC, and cholangiocellular carcinoma contain no reticuloendothelial cells. However, the exact mechanism of Levovist uptake in the liver and spleen is not completely understood and is a subject of ongoing research. 2,19 Figure 4. Hepatocellular carcinoma. A, Native B-mode image showing a hypoechoic lesion in the right lobe of the liver. B, Color SAE examination in the late phase of Levovist enhancement showing clear demarcation of the focal lesion without Levovist enhancement surrounded by a homogeneous color pixel mosaic of the liver parenchyma. B 212 J Ultrasound Med 23: , 2004

7 von Herbay et al Clear demarcation of focal liver lesions surrounded by a homogeneous color pixel mosaic of the surrounding liver parenchyma was found to be associated with malignancy of the lesion in 81% of the patients. All patients with metastases had clear demarcation in color SAE. These findings correspond to reports of other groups. 12,18,19 Albrecht et al 19 and Blomley et al 2 found demarcation of liver metastases in all of 15 and 18 patients, respectively, with known metastases. Furthermore, Albrecht et al 19 and Blomley et al 2 found demarcation of additional liver metastases in 33% and 17% of patients, respectively, which could not be seen on B-mode sonography. In all the studies, demarcation of liver metastases was independent of tumor etiology. In our study, demarcation of the lesions in color SAE was also found in 86% of the patients with HCC. Demarcation of HCC during color SAE examination was found in patients with both less and more differentiated HCC. The inhomogeneous Levovist enhancement in the remaining 14% of the patients with HCC could be due to an inhomogeneous enhancement of Levovist in cirrhosis. Figure 5. Receiver operating characteristic curves obtained for baseline and color SAE sonography according to the 5-point system (1, definitely benign; 2, probably benign; 3, indeterminate; 4, probably malignant; and 5, definitely malignant). The cutoff value for discrimination between benign and malignant hepatic lesions was set at 3. The A z for color SAE sonography (solid line; A z = 0.927) was significantly higher than that for baseline sonography (dashed line; A z = 0.739; P <.05). Comparison of ROC curves was performed with the LABMRMC software. An important finding in our study was that intralesional hypervascularization seemed to prolong the intralesional vascular phase of Levovist, which makes it necessary to study the color SAE late phase of Levovist not earlier than 10 minutes after Levovist injection in patients with hypervascularized lesions. As far as we know, this is the first study showing that fact. In our study, 13% of the patients with liver malignancy (3 patients with HCC and 1 patient with metastatic liver disease) had associated intralesional hypervascularization on power Doppler sonography and color pixel enhancement in the color SAE examination when examined 5 minutes after Levovist injection. However, when the color SAE examination was not performed until 10 minutes after Levovist injection, clear demarcation of the malignant lesion was observed in all these patients. In contrast, in all the patients with hypervascularized benign focal liver lesions, the homogeneous color pixel mosaic was still sustained when the color SAE examination was performed 10 minutes after Levovist injection. The main differential diagnosis for the finding of clear demarcation was hemangioma. However, in 86% of the patients with hemangiomas, clear demarcation of the lesion was found, whereas in the remaining patient with hemangioma, an inhomogeneous color pixel mosaic was found. The clear demarcation of hemangiomas in color SAE might be due to the fact that hemangiomas are free of liver tissue and reticuloendothelial cells. However, our results are contrary to those of Beissert et al, 18 who found SAE signals in patients with hemangioma as well. Albrecht et al 19 reported color pixel enhancement in 3 of 3 hemangiomas; however, this enhancement had a lower intensity compared with the surrounding liver parenchyma. These differences could be due to different sizes of the lesions. However, it could be hypothesized that the intravascular Levovist accumulation in hemangiomas could be prolonged because of the vessel architecture. Hemangiomas contain very small venous vessels surrounded by endothelial cells, which could be the reason for prolongation of the blood pool phase of Levovist, resulting in Levovist accumulation even at 3 to 5 minutes after injection, whereas regular vessels should be free of Levovist at that time. This hypothesis would fit the results of Chou et al, 21 who found that Levovist made the slow flow in J Ultrasound Med 23: ,

8 Color Stimulated Acoustic Emissions in Hepatic Lesions the venous spaces of a large-vessel hemangioma visible for about 3 minutes, whereas normally the postvascular phase started 3 minutes after Levovist injection. 16 One patient with histologic proof of a regenerative cirrhotic nodule with dysplasia had clear demarcation of the lesion on the color SAE examination, whereas the remaining 2 patients with histologic proof of a regenerative nodule had a homogeneous color pixel mosaic. It could be speculated that the clear demarcation in that particular patient could be due to the dysplasia. Dysplastic nodules of the cirrhotic liver are currently considered preneoplastic in hepatocellular carcinogenesis. 22,23 The finding of a homogeneous color pixel mosaic in the SAE examination in patients with FNH is comparable with the findings described by Beissert et al 18 and Blomley et al. 24 The homogeneous color pixel mosaic in the SAE during the late phase of Levovist enhancement in patients with hypersteatosis or hyposteatosis might be due to the intact liver architecture in these lesions, with regular content of reticuloendothelial cells. In conclusion, contrast-enhanced sonography has a high sensitivity and specificity for discrimination between benign and malignant liver lesions. However, in patients with hypervascularized lesions, color SAE should be performed no earlier than 10 minutes after Levovist injection to rule out artifacts caused by intralesional vessels. Color SAE in the late phase of Levovist enhancement is helpful as a screening method to decide which sonographically detected liver lesions need further investigation. According to the results found in this study, all patients with homogeneous contrast enhancement had benign focal lesions. From this point of view, the new technique could be helpful for reducing unnecessary invasive examinations, such as liver biopsy, CT, and magnetic resonance imaging, which could lower patients risk and could reduce diagnostic costs. Conversely, in patients with clear demarcation or an inhomogeneous color pixel mosaic in the SAE examination, further characterization of the lesion is necessary. However, the findings of this study need to be confirmed in a larger series of patients with liver lesions of various histologic origins. References 1. Schlief R. Developments in echo-enhancing agents. Clin Radiol 1996; 51(suppl 1): Blomley MJ, Albrecht T, Cosgrove DO, et al. Improved imaging of liver metastases with stimulated acoustic emission in the late phase of enhancement with the US contrast agent SH U 508A: early experience. Radiology 1999; 210: Goldberg BB, Liu JB, Burns PN, Merton DA, Forsberg F. Galactose-based intravenous sonographic contrast agent: experimental studies. J Ultrasound Med 1993; 12: Strobel D, Krodel U, Martus P, Hahn EG, Becker D. Clinical evaluation of contrast-enhanced color Doppler sonography in the differential diagnosis of liver tumors. J Clin Ultrasound 2000; 28: Hosten N, Puls R, Lemke AJ, et al. Contrastenhanced power Doppler sonography: improved detection of characteristic flow patterns in focal liver lesions. J Clin Ultrasound 1999; 27: Hedrick WR, Hykes DL. An overview of thermal and mechanical acoustic indices. J Diagn Med Sonography 1993; 9: Uhlendorf V, Scholle FD, Reinhardt M. Acoustic behaviour of current ultrasound contrast agents. Ultrasonics 2000; 38: Blomley MJ, Albrecht T, Cosgrove DO, Jayaram V, Butler-Barnes J, Eckersley RJ. Stimulated acoustic emission in liver parenchyma with Levovist. Lancet 1998; 351: Forsberg F, Goldberg BB, Liu JB, Merton DA, Rawool NM, Shi WT. Tissue-specific US contrast agent for evaluation of hepatic and splenic parenchyma. Radiology 1999; 210: Schlief R. Galactose-based echo-enhancing agents. In: Goldberg BB (ed). Ultrasound Contrast Agents. London, England: Martin Dunitz Ltd; 1997: Blomley MJ, Albrecht T, Cosgrove DO, et al. Stimulated acoustic emission to image a late liver and spleen-specific phase of Levovist in normal volunteers and patients with and without liver disease. Ultrasound Med Biol 1999; 25: Fritsch T, Heldmann D, Reinhard M. The potential of a novel ultrasound contrast medium. In: Goldberg BB (ed). Ultrasound Contrast Agents. London, England: Martin Dunitz Ltd; 1997: J Ultrasound Med 23: , 2004

9 von Herbay et al 13. Shi WT, Forsberg F, Raichlen JS, Needlman L, Goldberg BB. Pressure dependence of subharmonic signals from contrast microbubbles. Ultrasound Med Biol 1999; 25: Walker KW, Pantely GA, Sahn DJ. Ultrasound-mediated destruction of contrast agents. Invest Radiol 1997; 32: de Jong N. Physics of microbubble scattering. In: Nanda N, Schlief R, Goldberg BB (eds). Advances in Echo Imaging Using Contrast Enhancement. 2nd ed. Lancaster, England: Kluwer Academic; 1997: Kim CK, Lim JH, Lee WJ. Detection of hepatocellular carcinomas and dysplastic nodules in cirrhotic liver. J Ultrasound Med 2001; 20: Blomley MJ, Sidhu PS, Cosgrove DO, et al. Do different types of liver lesions differ in their uptake of the microbubble contrast agent SHU 508 in the late liver phase? Early experience. Radiology 2001; 220: Blomley M, Albrecht T, Cosgrove D, et al. Stimulated acoustic emission imaging ( sono-scintigraphy ) with the ultrasound contrast agent Levovist: a reproducible Doppler ultrasound effect with potential clinical utility. Acad Radiol 1998; 5(suppl 1):S236 S Burns PN, Fritsch T, Weitschies W, Uhlendorf V, Hope-Simpson D, Powers JE. Pseudo-Doppler shifts from stationary tissue due to the stimulated emission of ultrasound from a new microsphere contrast agent [abstract]. Radiology 1995; 197(suppl): Beissert M, Jenett M, Keberle M, et al. Comparison of contrast harmonic imaging in B-mode with stimulated acoustic emission, conventional B-mode US and spiral CT in the detection of focal liver lesions. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 2000; 172: Albrecht T, Blomley MJ, Heckemann RA, et al. Stimulated acoustic emission with the ultrasound contrast agent Levovist: a clinically useful contrast effect with liver-specific properties. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 2000; 172: Kratzer W, Kachele V, Merkle E, et al. Contrast enhanced Power Doppler sonography: comparison of various administration forms of the ultrasound contrast agent. Rofo Fortschr Geb Rontgenstr Neuen Bildgeb Verfahr 2000; 172: Chou YH, Tiu CM, Chiou HJ, Lai CR, Hsu CC, Yu C. Echo-enhancing sonography of a large-vessel hemangioma of the neck. J Clin Ultrasound 1999; 27: Theise ND, Schwarz M, Miller C, Thung SN. Macroregenerative nodules and hepatocellular carcinoma in 44 sequential adult liver explants with cirrhosis. Hepatology 1992; 16: J Ultrasound Med 23: ,