UTILIZATION OF FOUR DIMENSIONAL ULTRASOUND IN THE DETECTION OF LIVER MASS

Transcription

1 UTILIZATION OF FOUR DIMENSIONAL ULTRASOUND IN THE DETECTION OF LIVER MASS Ismael A Lapus, Diana A Payawal, Rodney L Fernan, RR Reyes Section of Gastroenterology, Cardinal Santos Medical Center San Juan, Metro Manila, Philippines Background: Significant advances have been recently introduced in various fields of medical technology, one of these is the 4-Dimensional Ultrasound or 4-D US. The 4-D US is an imaging machine used in the detection of liver mass. Objective: To determine the sensitivity and specificity as well the likelihood of 4-D US in comparison to Two- Dimensional US (2-D US) and Magnetic Resonance Imaging (MRI) in the detection of neoplastic liver mass. Methodology: This is a preliminary cross-sectional study conducted at the Cardinal Santos Medical Center. Patients with liver mass or suspicious liver mass on 2-D US underwent MRI and 4-D US. Tumor morphology and intralesional vascular involvement were noted. Histopathologic correlation was done to all patients. Results: Seventeen patients were identified to have liver mass or suspicious liver mass on 2-D US. Of these, thirteen (13) were found to have neoplastic and non-neoplastic mass on 2-D US and 4-D US, while fourteen (14) patients were seen to have neoplastic mass based on MRI findings. When correlated histopathologically, sensitivity and specificity of 4-D US, results showed 92% and 75% with a likelihood ratio of 3.7 and 0.11 for positive and negative test. 2D-US had 85% and 50% sensitivity and specificity with a likelihood ratio of 1.7 and 0.3 for positive and negative test. MRI showed 100% sensitivity and 75% specificity with a likelihood ratio of 4 and 0 for a positive and negative test. Intralesional involvement was only seen on 4D-US. Conclusion: Preliminary results showed that 4-D US is more superior compared to 2-D US and is almost comparable with MRI in the detection of neoplastic liver mass. INTRODUCTION Significant advances in medical technology have been introduced recently. In the field of imaging techniques, several non-invasive modalities are being utilized for the detection of liver mass. Recent progression of noninvasive imaging technology includes various techniques of harmonic ultrasound (US) imaging with several kinds of US contrast agents, multi-slice helical computed tomography (CT) and rapid high-quality magnetic resonance (MR) technique with new tissue specific contrast agents. These techniques seem to have a strong potential to improve detection and characterization of hepatocellular carcinoma (1). At present, the use of stateof-the art hepatic Magnetic Resonant Imaging techniques is being used as the standard in diagnostic imaging in the detection of hepatic mass with a reported sensitivity of 75%-100% (2). The use of technologically advanced ultrasound machines, like the Three- Dimensional Ultrasound (3-D US) permits volume imaging methods to be incorporated in interactive manipulation of volume by means of rendering, rotation and zooming in on localized features. Integration of views obtained over a region of a patient with 3-D US may permit better visualization in these situations and allow a more accurate diagnosis (3). Several investigations concluded the superiority of 3-D US as compared to Two-Dimensional Ultrasound (2-D US) in volume rendering of the liver (4), usefulness in procedures of ablation for liver cancers (5) and in imaging vascularity in hepatocellular carcinoma (6). The Four-Dimensional Ultrasound (4-D US) is the latest in ultrasonographic imaging, it is a 3-D US with an element of real-time that is exclusive to General Electric (GE). One of its advantages is that it allows the operator to visualize internal anatomy moving in real-time and increases the accuracy in US-guided biopsy. 47

2 Compared with the standard two-dimensional ultrasonographically guided biopsy of hepatic masses, fourdimensional ultrasonography provides an improved visualization of biopsy devices and more perceptible information on the spatial relationships between the biopsy needle and the target lesion (7). Currently however, there is limited study on the usefulness of this new imaging technology in the detection of liver mass and majority of its utilization is on images of fetus of an expectant mother. Could 4-D US be utilized as a diagnostic tool in the detection of neoplastic liver mass? This study aims to determine the sensitivity and specificity as well the likelihood of 4-D US in comparison to 2- D US and MRI in the detection of neoplastic liver mass. METHODOLOGY A preliminary cross-sectional study was done at Cardinal Santos Medical Center for one year (2004). Patients with liver mass or suspicious liver mass on 2-D US were selected to become part of the study. Computed sample size for the study was 80. Two-dimensional ultrasounds of different patients were conducted in different hospitals or diagnostic clinics and findings were read by different ultrasonologists. Consent forms signifying the patients intention to be included in the study were also provided. Patients were then asked to undergo imaging of the liver using Siemens high field 1.5 Tesla Magnetic Resonance Imaging (MRI) with Gadolinium enhancement and General Electric Voluson 730 Expert Diamond 4-D US. The MRI served as the standard of diagnostic imaging of the study. A radiologist from each diagnostic imaging modality was tapped to interpret the findings. The radiologists were unaware of the patient s clinical diagnosis and the findings of the other diagnostic imaging modalities. The tumor size, number and intralesional vascular involvement were noted. The tumor size was classified as to greater or less the 5 cm which was based on Barcelona Clinic Liver Cancer Classification (BCLC). Tumor frequency was classified as uninodular or multinodular ( 2) and this was based on CLIP (Cancer of the Liver Italian Program) classification. On ultrasound findings, the following descriptions were given to hepatocellular carcinoma: focal solid lesions (8), homogeneous or heterogeneous hypoechoic lesion, heterogeneous hyperechoic lesion, target lesion (a solid lesion with a hypoechoic halo), and lesion with internal color Doppler flow (9). The tumor descriptions for malignancy on MRI were as follows: a mass demonstrating homogenous, heterogeneous, or ring enhancement during the hepatic arterial phase (HAP) and hyperintensity on T2 weighted images (10). A bulls-eye lesion which is a common description of metastasis was classified as neoplastic. Non-enhancing lesions after contrast material administration were considered as non-neoplastic. To confirm the findings on images, all patients underwent ultrasound guided needle core biopsy (U.S. Biopsy gauge 20, 19 cm) except for one patient who underwent diagnostic paracentesis. The sensitivity and specificity with likelihood ratio for a positive and a negative test of the different imaging modalities for the detection of neoplastic liver mass were done on a per-patient basis. RESULTS Seventeen (17) patients with liver mass or suspicious mass on bimodal ultrasound were included in the study. Fifteen (15) patients were male and the mean age is 63.8 years with the youngest at 42 years old and oldest being at 76 years. Of the seventeen (17) patients, seven (7) were known to have hepatitis B, five (5) were known alcoholics, one (1) with ascites, one (1) with suspicious intra-abdominal malignancy and three (3) underwent routine 2-D US as part of check-up. Thirteen (13) patients were considered to have neoplastic mass on 2-D US. The neoplastic mass findings were described as solid mass, heterogeneous hypoechoic nodule and heterogeneous hyperechoic nodule. The other four (4) patients have non-neoplastic mass findings on 2-D US findings. The findings were described as regenerating nodule, nodular echo pattern and liver cirrhosis. The size of neoplastic mass lesion ranges from 48

3 1.3 cm to 15.4 cm and the frequency ranges from 1 to 2 lesions. No intralesional vascular finding was noted. Table 1 showed the summary of 2-D US findings. Table 1. Summary of 2-D US findings 2-DIMENSIONAL ULTRASOUND FINDING NUMBER OF PATIENTS Liver cirrhosis/nodular pattern 4 Solid Mass/Nodule Hyperechoic/Hypoechoic Nodule 13 Number of Solid Mass Solitary Size of Mass 5 cm 8 5 cm 5 Intralesional vascular involvement 0 Histopathologic correlation showed that of the thirteen (13) patients with neoplastic masses, eleven (11) had hepatocellular carcinoma, one (1) with chronic hepatitis and one (1) with abscess. Two (2) patients with nonneoplastic consideration had adenocarcinoma on biopsy. Two (2) patients with liver cirrhosis had the same histopathologic finding. Sensitivity and specificity on per-patient basis of 2-D US was 85% and 50% respectively. Likelihood ratio for a positive test was 1.7 and for a negative test was 0.3. Appendix A shows the summary of histopathologic findings in correlation to the 2-D US findings. On 4-D US, thirteen (13) patients had consideration of neoplastic mass while three (3) had liver cirrhosis and one (1) had nodular echopattern with focal hypoechogenecity. Neoplastic mass was described as solid mass and irregular heterogeneous mass with intralesional vascular involvement. The size of mass lesion ranges from 1.6 cm to 16 cm and frequency ranges from 1 to 2 lesions. Seven (7) of these patients had intralesional vascular lesion. Table 2 shows the summary of 4-D US findings. Table 2. Summary of 4-D US findings. 4-D ULTRASOUND FINDING NUMBER OF PATIENTS Fatty liver with focal sparing 1 Nodular echopattern 1 Liver cirrhosis 2 Solid mass Irregular heterogeneous mass Number of Solid Mass Solitary Size of Mass 5 cm 8 5 cm 5 Intralesional vascular involvement 7 On histopathologic correlation, one (1) patient with solid mass turned out to be an abscess. The other ten (10) patients had hepatocellular carcinoma and two (2) with adenocarcinoma. Liver cirrhosis on imaging correlated 13 49

4 well with histopathology. Chronic hepatitis was the histopathologic finding for fatty liver with focal sparing. The per-patient sensitivity and specificity for 4-D US were 92% and 75% respectively with a likelihood ratio of 3.68 for positive test and 0.11 for negative test. Appendix A shows the summary of histopathologic findings in correlation to the 4-D US findings. Fourteen (14) patients had neoplastic liver mass findings on magnetic resonance imaging. Neoplastic masses were described as enhancing mass, bull s eye lesion, hyperintense mass on T1 and T2 W and multiple contrasts enhancing peritoneal surface nodule. Three (3) patients had findings of non-neoplastic lesions. Non-neoplastic findings were abscess, liver cirrhosis and conglomeration of poorly enhancing masses. The mass lesion size ranges from 1 cm to 18 cm and frequency ranges from 1 to 3 lesions per patient. Summary of MRI findings isshown in Table 3. Table 3. Summary of MRI findings. MRI FINDINGS NUMBER OF PATIENTS Liver cirrhosis 1 Large complex mass/abscess 1 Conglomerate of small poorly enhancing masses 1 Enhancing Mass/Hepatocellular Ca/Bull s eye lesion Peritoneal carcinomatosis/neoplasm 14 Number of Mass Size of Mass Solitary cm 8 5 cm 6 On histopathologic correlation, one (1) patient with solid hepatic mass on 4-D US and 2-D US was read as an abscess according to the MRI, and this was later confirmed by a biopsy. Two other non-neoplastic findings on MRI showed histologic findings of liver cirrhosis and chronic hepatitis. On patients with ascites, the lesions seen on MRI can be described as an abnormal enhancement of the peritoneal lining and multiple contrasts enhancing peritoneal surface nodule on right subdiapraghmatic region. MRI considerations were peritoneal carcinomatosis, primary peritoneal neoplasm and inflammatory process. As such, the patient underwent diagnostic paracentesis instead of liver biopsy which showed malignant cells in the ascitic cell block. Further investigations revealed an ovarian carcinoma. In another patient, the mass found on MRI can be described as a subtle mildly enhancing lesion measuring 2.0 x 3.0 cm and considerations were a focus of inflammation, benign or malignant neoplasm. On histology, the lesion was read as chronic hepatitis. One patient whose lesions were considered on MRI as regenerating or dysplastic nodules or hepatocellular carcinoma had a histopathologic finding of hepatocellular carcinoma. The MRI description of the lesions were diffuse micronodular liver cirrhosis and a 3.2 cm mass at the posterior segment with hyperintensity on both T1 and T2 W examinations and non-visualized vascularity. The same mass showed intralesional vascular involvement on 4-D US and was interpreted as hepatocellular carcinoma. On 2-D US, finding on the same patient was liver parenchymal disease with hepatic nodule. Summary of histologic correlation with MRI findings are shown in Appendix A. Overall, per-patient s sensitivity and specificity for MRI were 100% and 75% respectively with a likelihood ratio of 4 for positive test and 0 for negative test. 50

5 DISCUSSION Over the past few years, ultrasound imaging has made tremendous progress in obtaining important diagnostic information for patients in a rapid and non-invasive manner. The inherent flexibility of ultrasound imaging, its moderate cost, non-radiation, real time imaging and no known bioeffects gave ultrasound a vital role in the diagnostic process and great advantages compared with MRI and CT scan. In the clinical field, hepatocellular carcinoma (HCC) is the fifth (5 th ) among the most common cancer in the world and third (3 rd ) in terms of cause of cancer-related death (11). However, local data shows that HCC is the third (3 rd ) most common cancer in the Philippines and it ranks second (2 nd ) among males. The early detection of hepatocellular carcinoma is critical to patient treatment and survival in this era of surgical techniques for resection and transplantation and new alternative therapeutic options such as transchatheter chemoembolization or radiofrequency ablation (12). Therefore, imaging and the accuracy of sonography in the detection of hepatocellular carcinoma is very important. However, reported sensitivities of ultrasonography ranges from 33% to 96 % (13). With the advent of a revolutionary technology in ultrasonography, which is the four-dimensional ultrasound, we investigated its accuracy in detecting mass lesions of the liver as compared to the two-dimensional ultrasound and magnetic resonance imaging. Result showed that 4-D US had a slightly better per-patient s sensitivity (92%) than 2-D US (85%). Two malignant lesions were missed on 2-D ultrasound while one lesion which was a peritoneal neoplasm was missed on 4-D US. The only neoplastic lesion missed by 4-D US was the small peritoneal neoplasm with ascites. The per-patient s sensitivity to MRI in our study is 100%. The specificity based on our results was similar with 4-D US and MRI (75%) unlike with 2-D US (50%). With a near difference in sensitivity and similar specificity, and considering the cost of undergoing an MRI, 4-D US can be an option as an imaging technique in the detection of liver mass. However, a larger sample collection is needed to affirm this predicament. One limitation of our study that could have affected the sensitivity of the 2-D US was that the radiologist was not blinded to the patient s clinical condition unlike the radiologist of 4-D US and MRI. Another aspect for determining the accuracy of 4-D US was the assessment of per-lesion sensitivity. This was not done since its measurement of accuracy would need the gross liver specimen for a lesion per lesion analysis. Despite of the better sensitivity of 4-D US to 2-D US, the added feature of real-time imaging is not useful in detecting liver mass since the lesion is just stationary. Eliminating the real-time feature of 4-D US makes it a 3-D US and will yield the same sensitivity result. The machine s real time feature can find its value in ultrasound guided biopsy of hepatic mass (7). In conclusion to this preliminary study, 4-D US is superior to 2-D US and is almost comparable with MRI in the detection of neoplastic liver mass. The real-time feature of the machine is not useful in detecting liver mass. Appendix A. Summary of histopathologic findings in correlation to the imaging findings 2-D US 4-D US MRI Histopathology 1 Solid mass Solid mass / Avascular Abscess Abscess 2 Solid mass Solid mass / Vascular Enhancing mass Hepatocellular Ca 3 Solid mass Solid mass / Avascular Enhancing mass Hepatocellular Ca 4 Solid masses Solid masses / Vascular Cluster of enhancing mass Adenocarcinoma 5 Solid Irregular Solid nodule / Vascular Enhancing mass Hepatocellular Ca Nodule 6 Solid mass Solid mass / Vascular Enhancing mass Hepatocellular Ca 7 Solid mass Solid mass / Vascular Enhancing mass Hepatocellular Ca 8 Solid mass Solid mass / Avascular Multiple enhancing mass Hepatocellular Ca 9 Liver cirrhosis Liver Cirrhosis Nodular liver margins Liver cirrhosis 10 Nodular echo Nodular echo pattern Multiple contrast Peritoneal malignancy 51

6 pattern Avascular enhancing nodule 11 Liver cirrhosis Liver Cirrhosis Conglomeration of small masses Chronic hepatitis 12 Solid mass Solid mass / Avascular Small enhancing nodules Hepatocellular Ca 13 Heterogeneously Hypoechoic illdefined mass Heterogeneous mass Vascular Hypoenhancing liver lesion and hyperenhancing liver lesion Hepatocellular Ca 14 Heterogeneously Hyperechoic nodule Solid mass Vascular Regenerating or dysplastic nodule or hepatocellular Ca Hepatocellular Ca 15 Regenerating nodule Solid mass Vascular Bulls-eye lesion Adenocarcinoma 16 Solid mass Focal hypoechogenicity Ill-defined mildly Chronic hepatitis Avascular enhancing lesion 17 Solid mass Solid mass (vascular) Large enhancing mass Hepatocellular Ca REFERENCES 1. Choi, BI, The Current Status of Imaging Diagnosis of HCC. Department of Radiology, Seoul National University Hospital. 2. Peterson MS, et al. Hepatic malignancies: usefulness of acquisition of multiple arterial venous phase images at gadolinium-enhanced MR imaging, Radiology 1996; 201: Pretorius DH, et al. 3-Dimensional ultrasound imaging in patient diagnosis and management: the future. Ultrasound Obstet Gynecol 1991; 1(6): Hui-Xiong Xu, et al. Three dimensional Power Doppler Imaging in depicting vascularity in HCC. J Ultrasound Med 22: Hui-Xiong Xu. Et al. Three-dimensional Gray scale volume rendering of the liver prelimenary clinical experience. J Ultrasound Med 21: Hui-Xiong Xu. Et al. Usefulness of three-dimensional sonography in procedures of ablation of liver cancers. J Ultrasound Med. 7. Hyung Jim Won, et al. Value of Four dimensional Ultrasonography in Ultrasonographically Guided Biopsy of Hepatic masses. 8. Genevieve, L. et al. Sonographic Detection of HCC and dysplastic nodules in cirrhosis; Correlation pf Pretransplantation Sonography and Liver Explant Pathology in 2000 Patients. American Journal of Radiology. 2002; 179: Ihab R., et al. Imaging Evaluation of HCC. Journal of Vascular and Interventional Radiology 2002;13: S173-S Earls J, Rofsky NR, DeCorato D, Krinsky G, Weinreb JC. Arterial-phase dynamic gadoliniumenhanced MR imaging: optimization with a test examination and a power injector. Radiology 1997; 202: Parkin DM, et al, Estimating the world cancer burden: GLOBOCAN Int J Cancer 2001; 94: Trinchet JC, Beaugrand M. Treatment of hepatocellular carcinoma in patients with cirrhosis. J Hepatol 1997; 25: Tanaka S, Kitamura T, Nakanishi K, et al. Effectiveness of periodic checkup by ultrasonography for the early diagnosis of hepatocellular carcinoma. Cancer 1990;66: [Medline] 52