Gadolinium-Enhanced MRI for Tumor Surveillance Before Liver Transplantation: Center-Based Experience

Transcription

1 Lauenstein et al. MRI Before Liver Transplantation Hepatobiliary Imaging Original Research 09_07_2171_Lauenstein.fm 7/27/07 Thomas C. Lauenstein 1 Khalil Salman 1 Roger Morreira 2 Thomas Heffron 3 James R. Spivey 4 Enrique Martinez 5 Puneet Sharma 1 Diego R. Martin 1 Lauenstein TC, Salman K, Morreira R, et al. Keywords: gadolinium, hepatocellular carcinoma, liver, MRI, transplantation DOI: /AJR Received January 26, 2007; accepted after revision April 20, Department of Radiology, Emory University Hospital, Emory Clinic, Bldg. A, Ste. AT-627, 1365 Clifton Rd., Atlanta, GA Address correspondence to D. R. Martin. (dmartin@emory.edu). 2 Department of Pathology, Emory University Hospital, Atlanta, GA. 3 Department of Surgery, Division of Transplantation, Emory University Hospital, Atlanta, GA. 4 Division of Digestive Diseases, Liver Transplant Center, Emory University Hospital, Atlanta, GA. 5 Department Gastroenterology, Emory University Hospital, Atlanta, GA. AJR 2007; 189: X/07/ American Roentgen Ray Society Gadolinium-Enhanced MRI for Tumor Surveillance Before Liver Transplantation: Center-Based Experience OBJECTIVE. The purpose of this study was to evaluate prospectively acquired institutional results to determine the accuracy of gadolinium-enhanced MRI in liver tumor surveillance before transplantation. SUBJECTS AND METHODS. One hundred fifteen patients underwent MRI of the abdomen within 90 days before liver transplantation. Images were acquired with gadolinium-enhanced 3D gradient-echo sequences in the arterial, venous, and delayed phases. Detection of hepatocellular carcinoma (HCC) was based on the imaging criteria arterial phase enhancement, delayed phase hypointensity, and development of an enhancing outer margin capsule. Imaging findings were compared with findings at histopathologic evaluation of the explanted liver. RESULTS. Thirty-six HCCs in 27 patients were detected at histopathologic evaluation. Patient-based analysis showed the sensitivity of MRI was 88.9% (24/27); specificity, 97.7% (falsepositive findings in two patients); and accuracy, 95.7%. MRI depicted 28 of 36 HCCs, resulting in a lesion-based sensitivity of 77.8%. Although all 18 HCCs 2 cm or larger were depicted with MRI, only 10 of 18 HCCs smaller than 2 cm were correctly diagnosed. However, two HCCs measuring smaller than 2 cm at pathologic examination were rated as dysplastic nodules on MRI. CONCLUSION. Contrast-enhanced MRI can be used as a primary diagnostic method for accurate detection and characterization of HCC 2 cm or larger as required by the criteria of the Model for End-Stage Liver Disease used by the United Network for Organ Sharing. MRI can be considered a standard tool for surveillance before liver transplantation. Reduction in cost and risk may be derived from the diminished need for other diagnostic imaging studies and biopsy and the avoidance of use of iodinated contrast agents in imaging of patients with cirrhosis, many of whom have impaired renal function. ccurate detection of hepatocellular A carcinoma (HCC) in the background of advanced cirrhosis is crucial because of the importance in triage of patients on liver transplantation waiting lists, the need for optimal use of a limited number of donor livers, and the objective of curative treatment by transplantation in optimally selected patients. The number and size of lesions are the major criteria in scoring methods such as the Model for End-Stage Liver Disease used by the United Network for Organ Sharing in the United States [1]. HCC is a common tumor in patients with chronic liver disease and cirrhosis [2] and causes approximately 500,000 deaths per year worldwide [3]. Diffuse chronic liver disease and cirrhosis are a massive world health problem. In the United States, rates of cirrhosis and HCC have been increasing, HCC being the most common malignant tumor of the liver [4]. Another consideration is that although liver lesions are common and usually benign, the specificity and accuracy of diagnostic imaging are critical in the evaluation of patients at risk of neoplasia. In this setting, MRI has been found [5, 6] more accurate than CT in the depiction of hepatic lesions, with emphasis on superior diagnostic specificity for lesions. MRI also has been found [7 10] to be an effective imaging method for HCC in particular. In the largest series to date [9], to our knowledge, findings on MRI of 555 patients older than 11 years and 279 liver transplantation patients correlated with the histopathologic findings in the explant. In that study, only five HCC lesions in four patients were missed on MRI, and all the missed lesions were 2.2 cm in diameter or smaller. In most cases of missed lesions, the images were degraded by motion, or the small tumor was found to be isointense on arterial phase imaging. The range of overall sensitivity AJR:189, September

2 Lauenstein et al. for detection of HCC with MRI has been reported to be 68 91% [7, 11]. However, those studies involved only a small number of patients (< 40) with small (< 2 cm) liver lesions, potentially influencing the lower apparent sensitivity. Limitations in detection and characterization of small HCC remain an area for further investigation [8, 12]. On the basis of our experience and the body of current published evidence, we implemented a diagnostic imaging strategy in which gadolinium-enhanced MRI was used for HCC surveillance of patients with cirrhosis awaiting liver transplantation. The aim of this study was to evaluate the accuracy of gadolinium-enhanced liver MRI in tumor surveillance using our center s data for the assessment, explanted liver for histopathologic correlation, and the Model for End-Stage Liver Disease scoring criteria for determining the accuracy of triage. Subjects and Methods Patients Prospective informed consent and permission for subsequent retrospective analysis were obtained from all patients undergoing MRI for liver tumor surveillance. The trial was approved by our institutional review board and was compliant with the Health Insurance Portability and Accountability Act. Between January 2004 and March 2006, 210 patients received a liver transplant at our institution. One hundred thirty of the 210 patients underwent MRI of the liver within 90 days (range, 3 89 days; mean, 46.5 days) before transplantation. The other 80 patients underwent MRI more than 90 days before transplantation, had contraindications to MRI, or underwent CT. Fifteen patients were excluded because they had undergone liver chemoembolization therapy for tumors (n = 13) or were unable to complete MRI (n = 2). Thus the final study cohort consisted of 115 patients: 38 women and 77 men (average age, 53.9 years; range, years). The underlying causes of liver cirrhosis were hepatitis C (n = 38), hepatitis B (n = 10), hepatitis B and C (n = 2), ethanol abuse (n = 24), ethanol abuse and hepatitis C (n = 6), cryptogenic cirrhosis (n = 10), nonalcoholic steatohepatitis (n = 4), primary sclerosing cholangitis (n =6), autoimmune hepatitis (n = 4), primary biliary cirrhosis (n = 2), and other diseases (n =9). MRI All MRI examinations were performed with a current-generation 1.5-T MRI unit (Magnetom Avanto, Siemens Medical Solutions; Gyroscan Intera, Philips Medical Systems; or Twin-EXCITE, GE Healthcare). A torso phased-array surface coil was used for signal reception. Routine imaging consisted of unenhanced MRI in the axial plane. T1-weighted images were acquired with a breathhold spoiled gradient-echo dual-echo in-phase and out-of-phase sequence (TR/TE, / ; flip angle, ). T2-weighted images were acquired with a HASTE sequence (effective TR, infinite; effective TE, milliseconds). Section thickness was 7 8 mm, and the matrix size was (phase frequency encoding) for all sequences. Gadolinium was administered to all patients with a power injector in a bolus of 0.1 mmol/kg of gadolinium chelate (Magnevist, Berlex Laboratories; or Omniscan, Amersham Health) at 2 ml/s followed by a saline flush. Four sets of breathhold serial axial T1-weighted 3D spoiled gradientecho fat-suppressed images were acquired before contrast administration and during the contrast-enhanced hepatic arterial dominant phase, venous phase, and delayed late vascular and interstitial phases. Acquisition parameters were TR/TE, / ; flip angle, 12 ; sensitivity encoding factor, 2; section thickness, 2 3 mm; matrix size, ; field of view, 400 mm (rectangular field of view, 75%); scan percentage, 80%; partial Fourier imaging in phase and slice directions. Arterial phase images were acquired with a real-timebolus-tracking method triggered 8 seconds after the arrival of contrast medium in the celiac arterial axis, during which time the patient was given breath-hold instructions. The venous and later delayed phase images were acquired 70 and 180 seconds, respectively, after contrast administration. Acquisition time ranged from 12 to 15 seconds because of the varying fields of view and because the number of sections (96 130) was individually modified according to patient stature. MR Image Analysis All MRI data sets were transferred to and evaluated with a postprocessing workstation (Advantage Windows, GE Healthcare). Two radiologists with 12 and 5 years of experience in liver MRI assessed the images in consensus. Both reviewers were aware the subjects were patients at high risk before liver transplantation, but they did not have access to the pathology reports. Underlying criteria for the diagnosis of HCC included increased enhancement of the lesion compared with normal liver tissue in the arterial contrast phase; washout of lesions during the later contrast phases with isointensity or hypointensity compared with adjacent liver tissue; and development of peripheral rim enhancement, previously referred to as pseudocapsule enhancement [13], on delayed phase images. Lesions were rated as HCC if at least two of the three features were present. Lesions with focal hepatic enhancement without washout or rim enhancement were not considered HCC and were considered hypervascular nonspecific lesions or dysplastic nodules on the basis of characteristics described by Krinsky et al. [10, 14]. Lesion size and location were precisely described according to the segmental anatomy of the liver. Histopathology Explanted livers were serially sectioned into 5- to 10-mm contiguous slices. On gross examination, suspicious nodules were identified as those distinct from surrounding regenerative nodules in terms of size, texture, color, and degree of bulging beyond the cut surface of the liver. All lesions other than regenerative nodules were sampled for histologic examination. At microscopic examination, malignant nodules were characterized on the basis of histologic type, size, location, number, and stage (American Joint Committee on Cancer TNM staging system) [15]. Lesion size and location were described in the manner used for the MRI evaluation and were compared with the findings in the detailed MRI report. Statistical Analysis The pathologic findings served as the standard of reference. Subgroups were formed for HCC 2 cm or larger and HCC smaller than 2 cm. This size categorization was based on the Model for End-Stage Liver Disease score criteria [1] in which significance is attributed only to HCC tumors larger than 2 cm in diameter. The diagnostic accuracy for MRI was at first assessed in a patient-based analysis. Point estimates for sensitivity, specificity, and positive and negative predictive values were calculated. Sensitivity also was determined in a lesion-based evaluation. To account for discrepancies between MRI and pathologic findings, two additional retrospective analyses were performed. First, a possible overlap of hypervascular HCC and dysplasia on MRI was evaluated in a second lesion-based analysis in which the MRI findings on hypervascular lesions that did not exhibit washout or rim enhancement were compared with the pathologic findings. Second, a retrospective analysis was performed to examine the MR images in instances in which HCC was found at pathologic examination but had not been visualized on MRI. Specific attention was paid to the location of the tumor described in the pathology report. These images also were assessed for technical factors, such as artifacts interfering with image quality and timing of the arterial phase images. Results The histopathologic finding showed the presence of HCC in 27 of 115 patients. Tumors larger than 2 cm were found in 18 of the 27 patients; the other nine patients had HCC measuring 2 cm or smaller. A total of 36 HCC tumors were found at histopathologic exami- 664 AJR:189, September 2007

3 MRI Before Liver Transplantation TABLE 1: Patient-Based Results of MRI for Detection of Hepatocellular Carcinoma (n =115) Tumors Smaller Than 2 cm (n) Tumors 2 cm or Larger (n) All Tumors Result (n) True-positive True-negative False-positive False-negative nation; 18 of these 36 lesions were larger than 2 cm, and 18 measured 2 cm or smaller. A single HCC was depicted in 20 patients, and seven patients had more than one lesion (five patients with two lesions and two patients with three lesions). At patient-based analysis, HCC was detected with MRI in 24 of 27 patients, resulting in an overall sensitivity of 88.9%. The specificity was 97.7% because two patients had false-positive results. In all 18 patients with HCC larger than 2 cm, the tumors were correctly identified with MRI. In only six of the nine patients with HCC 2 cm or smaller found at pathologic examination was HCC described in the MRI report. In the two cases of apparently false-positive findings, one patient found to have HCC larger than 2 cm and one patient found to have HCC 2 cm or smaller on MRI were found not to have HCC at histopathologic examination of the explant. The results of the patient-based analysis are displayed in Tables 1 and 2. Lesion-based analysis revealed an overall sensitivity of 77.8% for MRI. MRI showed 28 lesions had features of HCC, and the pathologic finding was that 36 tumors were HCC (Figs. 1 and 2). There were two false-positive results. Retrospective evaluation of the MR images showed one of these lesions to have typical imaging features of HCC (Fig. 3). At TABLE 2: Patient-Based Statistics on MRI for Detection of Hepatocellular Carcinoma retrospective second interpretation, the other false-positive lesion was believed to be more consistent with a perfusion abnormality. Six of 36 lesions described as HCC at pathologic examination were not found to have a corresponding enhancing liver lesion on prospective MRI. Retrospective second interpretation of these MRI studies showed no evidence of a tumor corresponding to the lesions described in the pathology report. In addition, imaging review of missed lesions showed no significant artifacts, and the arterial phase contrast enhancement was found to be within the usual parameters, as in all cases in this study. For two other of the 36 lesions found at pathologic examination, corresponding lesions were identified on MRI and had imaging features interpreted as nonspecific hypervascular foci possibly including dysplastic nodules. Accounting for a potential overlap of HCC and dysplasia at pathologic evaluation and for the possibility of disease progression during the interval between MRI and transplantation increased the lesion-based sensitivity of MRI to 30 (83.3%) of 36 lesions. Although all 18 HCCs larger than 2 cm were prospectively identified on MRI (sensitivity, 100%), 10 of 18 HCCs 2 cm or smaller were described as HCC on the basis of MRI criteria (sensitivity, 55.6%). If the lesions 2 cm or smaller depicted as nonspecific hypervascular foci or dysplastic nodules on MRI were to be reinterpreted as HCC, the sensitivity for depiction of HCC 2 cm or smaller would increase to 66.7%. The lesion-based results are shown in Table 3. Discussion This study was a center-specific evaluation of the utility of MRI in liver tumor surveillance in the care of patients with chronic liver disease awaiting liver transplantation. The critical findings from this study are that tumors 2 cm or larger in diameter, considered significant according to Model for End-Stage Liver Disease Characteristic All Tumors Tumors Smaller Than 2 cm Tumors 2 cm or Larger Sensitivity Specificity Positive predictive value Negative predictive value Accuracy Note Values are percentages. scoring criteria, can be identified on MRI with high sensitivity (100%) and specificity (99%). In the group of explants studied, there were no cases of inappropriate transplantation with false-positive or false-negative results for HCC that would have retrospectively altered the Model for End-Stage Liver Disease score or the priority ranking of any study patient on the transplantation waiting list. Clinical Significance The current practice standard in the United States established by United Network for Organ Sharing for triage of liver transplantation patients entails a combination of Model for End-Stage Liver Disease score [1] and exclusion behavioral and social criteria, such as smoking and ethanol consumption. The Model for End-Stage Liver Disease score is a mortality risk predictor and is calculated with a formula that relies on serum creatinine concentration, total bilirubin level, and international normalized ratio. Detection of HCC in the liver has been incorporated into priority scoring in triage for the waiting list and can be used to provide additional priority points to a patient. A solitary lesion 2 5 cm in diameter or multiple lesions numbering three or fewer with a total combined diameter of 5 cm are considered curable disease. This criterion is meant to reduce time to transplantation and improve chances for cure because HCC is known to have high potential for rapid growth [1, 16]. Any lesion larger than 5 cm is considered incurable and therefore is a contraindication to transplantation. None of the patients in this study who underwent transplantation had explant pathologic findings of tumors larger than 5 cm, more than three tumors, or diffuse infiltrative disease. Thus no patient had inappropriately received a transplant because of incurable disease. The ability to detect tumors larger than 2 cm with high sensitivity is significant because this size is the threshold for scoring additional priority points. Alternative Diagnostic Strategies Although the use of sonography for liver tumor surveillance has been described [17], the use of sonography in this role has been seriously questioned [18]. Results of comparative studies of sonography, CT, and MRI support the hypothesis that because of lack of sensitivity, sonography should not be used for tumor surveillance in the care of patients with cirrhosis [19, 20]. Although results of individual studies have suggested that CT and MRI may have similar sensitivity in detection of HCC AJR:189, September

4 Lauenstein et al. A C E Fig year-old man with hepatitis C, cirrhosis, and 1.9-cm hepatocellular carcinoma in segment VI of liver. A, Unenhanced T1-weighted MR image shows lesion of low signal intensity (arrow). B, Arterial phase MR image shows lesion with marked enhancement (arrow). C, Venous phase image shows contrast washout with rim enhancement (arrow). D, Delayed phase MR image shows contrast washout with rim enhancement (arrow). E, Histopathologic photograph shows malignant hepatocytes growing in solid pattern. (H and E) B D [21] and that MRI may have higher relative specificity [22], authors [20] of a recent review have argued that on average MRI has been shown to have better sensitivity and specificity than CT. MRI strategies based on the use of various IV contrast compounds have been evaluated. Paramagnetic gadolinium agents have been found better than superparamagnetic iron oxide compounds and mangafodipir 666 AJR:189, September 2007

5 MRI Before Liver Transplantation trisodium alone in the detection of HCC [11, 23, 24]. Radiologists should be aware, however, of the risks of use of gadolinium contrast agents in imaging of patients with reduced renal function. Gadodiamide has been found to induce nephrogenic systemic fibrosis in 3 5% of cases of severe renal insufficiency [25]. Repeated use of contrast-enhanced CT has raised A C E safety concerns regarding the as yet undocumented risk of renal exposure to iodinated contrast agents in the setting of hepatorenal disease [26 29] and the cumulative radiation exposure from repeated scans required for surveillance [30 33]. Evaluation of the financial effects of adoption of MRI as a primary surveillance imaging strategy was beyond the Fig year-old man with hepatitis B, cirrhosis, and 2.3-cm hepatocellular carcinoma at border of segments V and VIII. A, Unenhanced T1-weighted MR image shows high signal intensity of lesion (arrow). B, Arterial phase MR image shows high signal intensity of lesion (arrow). Contrast enhancement is difficult to appreciate because of intrinsic high T1-weighted signal intensity within lesion. C and D, Venous (C) and delayed (D) phase contrast-enhanced MR images show contrast washout and rim enhancement (arrow), which are features of hepatocellular carcinoma. E, Photograph of gross pathologic specimen shows round bile-stained subcapsular nodule (arrow) and surrounding cirrhotic liver parenchyma. scope of our study. Cost-effectiveness would have to be evaluated in terms of factors such as relative reimbursement and accuracy of patient selection for transplantation. None of the patients in this study who underwent transplantation had undergone needle-guided biopsy before transplantation with complete reliance on imaging findings. This observation is B D AJR:189, September

6 Lauenstein et al. particularly important because of concern about increased risk of complications after needle biopsy among patients with cirrhosis [34]. Lesions smaller than 2 cm are technically more challenging to biopsy. In addition, the diagnostic yield from percutaneous biopsy may be relatively low, approximately 60% [35]. Limitations A total of 23% of the patients in our series who underwent transplantation had HCC. Although relatively high and raising the possibility of bias, this percentage is in keeping with the finding of a 19% incidence of HCC in a large recent trial [36]. Although our study was biased A C Fig year-old man with hepatitis C, cirrhosis, and 1.3-cm-diameter lesion resembling hepatocellular carcinoma on border of segments IVa and VIII. No hepatocellular carcinoma or other tumor was identified at pathologic examination. A, Unenhanced T1-weighted image shows lesion is inconspicuous. B, Arterial phase MR image shows enhancement (arrow). C, Venous phase MR image shows contrast washout and rim enhancement (arrow). D, Delayed phase MR image shows contrast washout and rim enhancement (arrow). toward patients awaiting transplantation who were at high risk of HCC, this bias was inherent to our study design of contemporaneous MRI and whole-organ pathologic examination. The finding of lower sensitivity for tumors smaller than 2 cm has been described previously [7 10] and introduces the potential for inappropriate transplantation in the treatment of patients with multifocal small HCC who have four or more tumors. None of our patients was in this category, and we conclude that this event must be rare. Furthermore, regular tumor surveillance with routine imaging at intervals of no more than 6 12 months has been suggested. HCC is presumed to grow during imaging intervals and should become differentiated as either a new or a growing nodule. This potential limitation should be the focus of future investigation. We found that one lesion had typical features of HCC on MRI but was not described in the pathology report (Fig. 3). This finding may be a limitation of gross analysis of liver specimens whereby it is not always easy to ensure uniform thin sections and to exactly coordinate anatomic relations between imaging and pathologic specimens, as has been discussed previously [7]. We believe that the lesion in this case probably was HCC that was not identified on gross analysis of the specimen. B D 668 AJR:189, September 2007

7 MRI Before Liver Transplantation TABLE 3: Lesion-Based Analysis of MRI in Detection of Hepatocellular Carcinoma (n = 36) Characteristic All Tumors Tumors Smaller Than 2 cm Tumors 2 cm or Larger No. detected at pathologic examination No. confirmed on MRI No. rated as dysplastic nodule on MRI No. missed on MRI No. of false-positive findings on MRI Sensitivity (%) Corrected sensitivity (%) a a Sensitivity based on pooling all lesions rated as hepatocellular carcinoma or dysplastic nodule on MRI. Another limitation of MRI pathologic correlation is that differentiation of a regenerative nodule, dysplasia, and HCC can be challenging. These lesions are part of a spectrum from benign reactive to malignant transformation. It is difficult to understand the precise relations between the imaging and histologic features given that these methods are examinations of fundamentally different components and properties of the tissues. Histologic examination predominantly involves the cellular and intracellular structures; gadolinium-enhanced MRI involves the amount of vascular supply derived from the hepatic arterial branches. We identified two of 36 pathologically defined HCCs smaller than 2 cm that exhibited arterial enhancement with isointensity in later contrast phases and no evidence of rim enhancement. This finding may represent a variable appearance of HCC on MRI. An alternative explanation is that these lesions may have continued to progress from dysplasia to malignancy during the interval between imaging and transplantation. The histopathologic findings may have been overinterpreted, or imaging may have been impaired as the result of technical limitations. Regarding the latter possibility, retrospective review of the images showed no technical cause of failure to identify tumor features. Enhancing foci in cirrhotic liver have been described as being common, most commonly not associated with malignancy [10, 14]. Observations from our study, however, suggest that enhancing foci in association with cirrhosis at least should be monitored closely. Six HCCs identified at pathologic examination were not seen on MRI, and all were tumors smaller than 2 cm. On retrospective image review, no technical factors were identified to account for this discrepancy. There was no evidence of artifacts over the liver, and timing in the arterial phase was optimal. The possibility of interval tumor progression and histologic overinterpretation are alternative causes of MRI pathologic discrepancies. Conclusion Our experience with MRI for liver tumor surveillance before transplantation showed that this diagnostic method is accurate for triage according to the accepted Model for End- Stage Liver Disease, may be a safe alternative that avoids radiation, avoids the risk associated with use of iodinated contrast medium, and minimizes the need for invasive biopsy. References 1. Sachdev M, Hernandez JL, Sharma P, et al. Liver transplantation in the MELD era: a single-center experience. Dig Dis Sci 2006; 51: Sherman M. Hepatocellular carcinoma: epidemiology, risk factors, and screening. Semin Liver Dis 2005; 25: Parkin DM, Bray F, Ferlay J, Pisani P. Estimating the world cancer burden: Globocan Int J Cancer 2001; 94: Rocken C, Carl-McGrath S. Pathology and pathogenesis of hepatocellular carcinoma. Dig Dis 2001; 19: Semelka RC, Martin DR, Balci C, Lance T. Focal liver lesions: comparison of dual-phase CT and multisequence multiplanar MR imaging including dynamic gadolinium enhancement. J Magn Reson Imaging 2001; 13: Ward J, Robinson PJ, Guthrie JA, et al. Liver metastases in candidates for hepatic resection: comparison of helical CT and gadolinium- and SPIO-enhanced MR imaging. Radiology 2005; 237: Hecht EM, Holland AE, Israel GM, et al. Hepatocellular carcinoma in the cirrhotic liver: gadolinium-enhanced 3D T1-weighted MR imaging as a stand-alone sequence for diagnosis. Radiology 2006; 239: Holland AE, Hecht EM, Hahn WY, et al. Impor- tance of small (< or = 20-mm) enhancing lesions seen only during the hepatic arterial phase at MR imaging of the cirrhotic liver: evaluation and comparison with whole explanted liver. Radiology 2005; 237: Karadeniz-Bilgili MY, Braga L, Birchard KR, et al. Hepatocellular carcinoma missed on gadolinium enhanced MR imaging, discovered in liver explants: retrospective evaluation. J Magn Reson Imaging 2006; 23: Krinsky GA, Lee VS, Theise ND, et al. Hepatocellular carcinoma and dysplastic nodules in patients with cirrhosis: prospective diagnosis with MR imaging and explantation correlation. Radiology 2001; 219: Kim YK, Kim CS, Kwak HS, Lee JM. Three-dimensional dynamic liver MR imaging using sensitivity encoding for detection of hepatocellular carcinomas: comparison with superparamagnetic iron oxide-enhanced MR imaging. J Magn Reson Imaging 2004; 20: Shah TU, Semelka RC, Pamuklar E, et al. The risk of hepatocellular carcinoma in cirrhotic patients with small liver nodules on MRI. Am J Gastroenterol 2006; 101: Grazioli L, Olivetti L, Fugazzola C, et al. The pseudocapsule in hepatocellular carcinoma: correlation between dynamic MR imaging and pathology. Eur Radiol 1999; 9: Krinsky GA, Israel G. Nondysplastic nodules that are hyperintense on T1-weighted gradient-echo MR imaging: frequency in cirrhotic patients undergoing transplantation. AJR 2003; 180: Varotti G, Ramacciato G, Ercolani G, et al. Comparison between the fifth and sixth editions of the AJCC/UICC TNM staging systems for hepatocellular carcinoma: multicentric study on 393 cirrhotic resected patients. Eur J Surg Oncol 2005; 31: Carr BI. Hepatocellular carcinoma: current management and future trends. Gastroenterology 2004; 127: Sangiovanni A, Del Ninno E, Fasani P, et al. Increased survival of cirrhotic patients with a hepatocellular carcinoma detected during surveillance. Gastroenterology 2004; 126: Di Bisceglie AM. Issues in screening and surveillance for hepatocellular carcinoma. Gastroenterology 2004; 127: Hann LE, Winston CB, Brown KT, Akhurst T. Diagnostic imaging approaches and relationship to hepatobiliary cancer staging and therapy. Semin Surg Oncol 2000; 19: Colli A, Fraquelli M, Casazza G, et al. Accuracy of ultrasonography, spiral CT, magnetic resonance, and alpha-fetoprotein in diagnosing hepatocellular carcinoma: a systematic review. Am J Gastroenterol 2006; 101: AJR:189, September

8 Lauenstein et al. 21. Bartolozzi C, Donati F, Cioni D, Crocetti L, Lencioni R. MnDPDP-enhanced MRI vs dual-phase spiral CT in the detection of hepatocellular carcinoma in cirrhosis. Eur Radiol 2000; 10: Marrero JA, Hussain HK, Nghiem HV, Umar R, Fontana RJ, Lok AS. Improving the prediction of hepatocellular carcinoma in cirrhotic patients with an arterially-enhancing liver mass. Liver Transpl 2005; 11: Youk JH, Lee JM, Kim CS. MRI for detection of hepatocellular carcinoma: comparison of mangafodipir trisodium and gadopentetate dimeglumine contrast agents. AJR 2004; 183: Kim YK, Kim CS, Lee YH, Kwak HS, Lee JM. Comparison of superparamagnetic iron oxide-enhanced and gadobenate dimeglumine enhanced dynamic MRI for detection of small hepatocellular carcinomas. AJR 2004; 182: Broome DR, Girguis MS, Baron PW, Cottrell AC, Kjellin I, Kirk GA. Gadodiamide-associated nephrogenic systemic fibrosis: why radiologists should be concerned. AJR 2007; 188: FOR YOUR INFORMATION 26. Yun AJ, Doux JD, Lee PY. Contrast nephropathy may be partly mediated by autonomic dysfunction: renal failure considered as a modern maladaptation of the prehistoric trauma response. Med Hypotheses 2006; 66: Gonwa TA, McBride MA, Anderson K, Mai ML, Wadei H, Ahsan N. Continued influence of preoperative renal function on outcome of orthotopic liver transplant (OLTX) in the US: where will MELD lead us? Am J Transplant 2006; 6: Barrett BJ. Contrast nephrotoxicity. J Am Soc Nephrol 1994; 5: Woitas RP, Heller J, Stoffel-Wagner B, Spengler U, Sauerbruch T. Renal functional reserve and nitric oxide in patients with compensated liver cirrhosis. Hepatology 1997; 26: Martin DR, Semelka RC. Health effects of ionising radiation from diagnostic CT. Lancet 2006; 367: Tsapaki V, Aldrich JE, Sharma R, et al. Dose reduction in CT while maintaining diagnostic confidence: diagnostic reference levels at routine head, chest, and abdominal CT IAEA-coordinated research project. Radiology 2006; 240: International Commission on Radiological Protection. ICRP publication 60: 1990 recommendations of the International Commission On Radiological Protection, 60. New York, NY: Elsevier, Board on Radiation Effects Research. Health risks from exposure to low levels of ionizing radiation: BEIR VII. Washington, DC: National Academies Press, Terjung B, Lemnitzer I, Dumoulin FL, et al. Bleeding complications after percutaneous liver biopsy: an analysis of risk factors. Digestion 2003; 67: Stewart CJ, Coldewey J, Stewart IS. Comparison of fine needle aspiration cytology and needle core biopsy in the diagnosis of radiologically detected abdominal lesions. J Clin Pathol 2002; 55: Santoyo J, Suarez MA, Fernandez-Aguilar JL, et al. True impact of the indication of cirrhosis and the MELD on the results of liver transplantation. Transplant Proc 2006; 38: Mark your calendar for the following ARRS annual meetings: April 13 18, 2008 Marriott Wardman Park Hotel, Washington, DC. April 26 May 1, 2009 John B. Hynes Veterans Memorial Convention Center, Boston, MA. May 2 7, 2010 Grand Hyatt San Diego, San Diego, CA May 1 6, 2011 Hyatt Regency Chicago, Chicago, IL April 29 May 4, 2012 Vancouver Convention Center, Vancouver, BC, Canada 670 AJR:189, September 2007