William J. von Liebig Transplantation Center, Division of Gastroenterology and Hepatology, Mayo Clinic College of Medicine, Rochester, Minnesota The diagnosis of hepatocellular carcinoma (HCC) includes detection of the index lesion, staging of the lesion within the liver, and assessment for extrahepatic metastasis. HCC is a highly vascular neoplasm usually arising in a cirrhotic liver. Based on this concept, consensus criteria have been developed for the radiographic diagnosis of HCC. These include: (1) identication of a mass >2 cm in diameter in a cirrhotic liver in 2 imaging modalities, and (2) contrast enhancement on computed tomography, magnetic reso- nance, or angiography. A mass lesion within a cirrhotic liver in the presence of a serum -fetoprotein level >400 ng/mL also is diagnostic. For lesions <2 cm in diameter, histological conrmation is required. Serum markers for the diagnosis of early HCC (<2 cm in diameter) have not been established. Staging HCC for metastases is insensi- tive and is based on conventional criteria (eg, pulmonary nodules, skeletal metastases, and lymphadenopathy). Ad- ditional diagnostic techniques based on cytological ad- vances, genomics, and proteomics are needed for the diagnosis and staging of this highly malignant neoplasm. H epatocellular carcinoma (HCC) is the third leading cause of cancer-related death worldwide. Recent ob- servations indicate that the incidence of HCC is increasing in western countries, including the United States. In the majority of cases, cirrhosis is the major underlying risk factor. Despite advances in surgical and ablative therapies for HCC, a number of controversial issues associated with the accurate diagnosis and staging of HCC remain. This article reviews current standards for the clinical diagnosis of HCC. In particular, attention focuses on the 3 questions of most interest to the clinician: (1) Does the lesion represent HCC and is a biopsy needed to conrm the diagnosis? (2) Are there other lesions in the liver and is there vascular invasion? (3) Are extrahepatic metastases present? Investi- gative opportunities for more accurate diagnosis and staging also are discussed. Evidence-based recommendations are highlighted when deemed appropriate. Diagnosis of Hepatocellular Carcinoma The diagnosis of HCC is of considerable interest to clinicians evaluating patients with liver cirrhosis. 1,2 Based on the growing availability of potential curative therapies that may improve survival in selected patients, 3 a consensus statement from the European Association for the Study of the Liver (EASL) has been formulated to standardize diagnostic approaches in patients with HCC. 4 This statement recognizes both histological and radiological criteria as sufcient evidence for identifying HCC in patients with cirrhosis. Recommendations were categorized based on size of the primary lesion when detected by conventional imaging modalities. Hepatocellular Carcinoma Lesions >2 cm in Diameter The accurate diagnosis of HCC lesions 2 cm in diameter can be made noninvasively, based on radio- graphic criteria in patients with cirrhosis. The ability to diagnose these tumors noninvasively is based on the high prevalence of HCC in patients with cirrhosis, the neo- vascularity of HCC (a biological hallmark of neoplasia), and the ability to equate neovascularity with contrast enhancement on rapid-sequence cross-sectional imaging studies (Figure 1). Consensus diagnostic criteria have been developed and widely accepted (Table 1). Require- ments for establishing a radiologic diagnosis of HCC include the presence of a focal lesion 2 cm in diameter, identied by using 2 imaging modalities and showing arterial hypervascularization in at least 1 of the imaging modalities. Contrast-enhanced computed tomography (CT), magnetic resonance (MR) imaging (MRI), and angiography are used most commonly to identify these features in clinical practice. If the lesion is identied in only a single imaging modality, the presence of a serum -fetoprotein (AFP) level 400 ng/mL also is considered diagnostic of HCC in the setting of cirrhosis. 4 Given the superb diagnostic accuracy of these criteria, 48 percuta- Abbreviations used in this paper: AFP, -fetoprotein; CT, computed tomography; DCP, des-carboxyprothrombin; EASL, European Associa- tion for the Study of the Liver; FISH, uorescent in situ hybridization; HCC, hepatocellular carcinoma; MR, magnetic resonance; MRI, mag- netic resonance imaging. 2004 by the American Gastroenterological Association 0016-5085/04/$30.00 doi:10.1053/j.gastro.2004.09.026 GASTROENTEROLOGY 2004;127:S126S132 neous biopsy for histological conrmation is not manda- tory. Among 160 patients with 225 focal liver lesions evaluated by means of sequential radiological imaging studies in preparation for planned surgical therapy, 5 the preoperative diagnostic accuracy rate without histologi- cal conrmation was reported to be 98.2%. For patients with HCC, sensitivity, specicity, and accuracy of diag- nosis by means of radiological criteria were 100%, 98.8%, and 99.1%, respectively. Other cited advantages for using noninvasive criteria to diagnose HCC in lesions 2 cm in diameter include avoidance of a 10%20% false-negative rate from histo- logical sampling. 911 This phenomenon has been ob- served even when cytological analysis is combined with ne-needle biopsy. 10 In clinical practice, because of this high false-negative rate, a negative biopsy result in the setting of cirrhosis does not conclusively exclude the possibility of HCC. 4 Moreover, tumor seeding from an invasive biopsy has been well described, 1113 with prev- alence rates between 1.5% and 5%. Of note, the majority of lesions in which needle-track seeding was reported were 2 cm in size. 13 The reported incidence of tumor seeding from an invasive biopsy is based solely on the presence of needle-track seeding, which may underesti- mate this problem. All biopsies disrupt the tumor-vas- cular integrity and may promote more widely dissemi- nated seeding. Thus, many experienced medical centers contemplating curative therapeutic approaches use the consensus criteria and do not routinely subject patients to needle biopsies. In the case of indeterminate radiolog- ical ndings (ie, a lesion identied on only 1 imaging modality with no contrast enhancement) and lesions that are obviously advanced and require systemic therapies, an image-guided biopsy of suspicious lesions 2 cm in size is reasonable. 11,13 Limitations associated with noninvasive diagnostic criteria include the misclassication of dysplastic nod- ules 14,15 as malignant. Consequences of this misdiagnosis are unclear. Less than 20% of patients with high-grade dysplastic nodules remain cancer free during 3 years. 16 Perhaps these lesions should be treated aggressively; however, the decision is more complex when contem- plating liver transplantation compared with surgical re- section or ablative therapies. Additionally, other malig- nant hepatic lesions, such as non-Hodgkins lymphoma, recently described in association with hepatitis C induced cirrhosis, could be missed without histological conrmation and result in misguided therapy. 17 How- ever, in the United States, intrahepatic hepatitis C associated lymphoma appears to be rare. Hepatocellular Carcinoma Lesions <2 cm in Diameter Criteria for the diagnosis of HCC are less well dened for nodules 2 cm in diameter. As a result, a number of recall procedures have been outlined by the EASL consensus statement to initiate follow-up. How- ever, evidence to support the clinical effectiveness and widespread use of these recommendations is lacking. 4 Focal hepatic lesions 1 cm in diameter carry a 50% risk for harboring malignancy. 1,2,18 To exclude interval growth, serial abdominal ultrasound examina- tions every 3 months are recommended until the lesion exceeds 1 cm in size. For nodules between 1 and 2 cm in size, the use of ne-needle aspiration with biopsy is recommended for diagnosis, according to EASL consen- sus guidelines. This is based on evidence suggesting that the sensitivity and diagnostic accuracy with existing imaging techniques is reduced for lesions 2 cm. 2 How- ever, limitations of invasive biopsy include a 30%40% false-negative diagnostic rate 12 and the risk for tumor seeding, as described. Given the inadequate test perfor- mance of an invasive biopsy and serum AFP level (dis- cussed next), novel strategies are required to improve the diagnosis of HCC in lesions 2 cm. Better cytological techniques may enhance the utility of ne-needle aspirate specimens in making the diagnosis of HCC. For example, gains of chromosome 1 and 8 are common features of HCC and are diagnostic of aneuploidy, 19 a diagnostic feature of neoplasia, when present. Fluorescent in situ Table 1. Noninvasive Diagnostic Criteria for Hepatocellular Carcinoma Mass in a cirrhotic liver, 2 cm, identied by means of 2 coincident imaging techniques (ultrasonography, spiral computed tomography, magnetic resonance imaging, angiography) One imaging study showing contrast enhancement Figure 1. Computed tomographic scan of the liver in a patient with cirrhosis attributable to hepatitis Crelated disease shows a contrast- enhanced lesion typical of hepatocellular carcinoma (arrow). November Supplement 2004 HCC DIAGNOSIS S127 hybridization (FISH) techniques using probes for the centromeres of these 2 chromosomes have been developed for HCC. 20 In other hepatobiliary neoplasia, namely cholangiocarcinoma, FISH techniques are clinically available to aid in the cytological diagnosis of cancer. 21 The role of FISH in the diagnosis of small HCC needs to be explored and represents a high priority investigative opportunity. Serum Markers for Hepatocellular Carcinoma: -Fetoprotein and Des-Carboxyprothrombin Serum Levels Selection of a serum AFP cutoff level 400 ng/mL as a complementary method for the diagnosis of HCC is based on experience from patients with viral- induced cirrhosis. 2225 However, serum AFP levels 400 ng/mL also may be observed in patients with active viral replication without evidence for HCC. 25 Furthermore, serum AFP levels often are normal (20 ng/mL) with lesions between 1 and 2 cm and do not help clarify the status of most indeterminate nodules. 24 Finally, a recent systematic review conrmed the poor diagnostic ability of serum AFP level for detecting HCC at any level of pretest risk. 25 However, as a conrmatory test in patients with a mass on imaging studies, AFP determination remains clinically useful 24 despite the less than optimal correlation of serum AFP level with HCC. It should be emphasized that the reported poor performance of serum AFP determinations for HCC pertains to patient popu- lations with primarily viral liver diseases. The utility of serum AFP determinations in patients with cholestatic liver diseases, alcohol and nonalcohol steatohepatitis syn- dromes, hemochromatosis, cryptogenic cirrhosis, and 1 - antitrypsin deciency has not been established. The des-carboxyprothrombin (DCP) test has been available for a couple of decades, but interest in its performance as an HCC biomarker has resurfaced. A recent study suggests it has a high specicity for differ- entiating malignant from benign liver disease. 26 Its abil- ity to specically identify early HCC 2 cm is unknown and currently the subject of investigation. Individual Cross-Sectional Imaging Techniques for the Diagnosis of Hepatocellular Carcinoma Earlier studies attempting to document the diag- nostic accuracy of imaging techniques were limited by retrospective study design and problems related to veri- cation bias. However, more recent studies have used explant histological evaluation as a comparative gold standard. 2731 Abdominal ultrasonography is associated with a sensitivity and specicity of 20%51% and 92% 96% for detecting lesions consistent with HCC in pa- tients with cirrhosis, respectively. 2729 Further reduc- tions in test performance result when additional lesions are sought with this modality. 27 Detection rates for lesions between 2 and 3 cm and 1 and 2 cm with ultrasonography are estimated at 20% and 13%, respec- tively. 28 However, the high specicity with ultrasound implies that any recognized lesion in patients with cir- rhosis requires additional evaluation to exclude malig- nancy. 29 Computed tomographic imaging was associated with similar test performance characteristics from earlier liver explant comparative studies. 30,31 However, recent investigations documented sensitivity and specicity rates of 71%80% and 80%96% for contrast-enhanced CT compared with explant histological evaluation, re- spectively. 3234 This probably is related to an increased proportion of larger tumors among patients referred for possible liver transplantation. 33 However, Lipiodol (Guerbet Andre, Aulnay-Sous-Bois, France)-enhanced techniques are less accurate. 35 MRI-angiography appears to have greater sensitivity (84%) compared with CT for the detection of nodules between 1 and 2 cm. However, lesions 1 cm continued to be missed in 70% of cases with both imaging techniques. 36 Furthermore, in the absence of a mass lesion on T2 MR images, only 30% of hypervascular lesions 2 cm eventually are diagnosed as HCC on follow-up, with the remainder disappearing or remaining stable in size. 37 Positron emission tomography is not an effective technique for the detection of HCC lesions 2 cm. 32 Role of Cross-Sectional Imaging for Early Detection of Hepatocellular Carcinoma Despite a complete lack of evidence from random- ized controlled trials, many centers around the world have implemented early HCC detection surveillance pro- grams for patients with cirrhosis. Diagnostic test perfor- mance characteristics of imaging techniques have been extrapolated to justify the effectiveness of surveillance without evidence-based foundations. Based on its cost and availability, ultrasonography remains the modality of choice. Both CT and MRI generally are used as conrmatory tools for diagnosis, given their increased cost and limited availability. However, the use of CT for HCC surveillance had a greater sensitivity (88%) com- pared with ultrasonography (50%) when compared di- rectly in a single investigation. 34 Enthusiasm for widespread HCC surveillance is tem- pered by recognized limitations in diagnosis. Ultrasound remains unable to reliably distinguish macroregenerative and low-grade dysplastic nodules, which often are 5 mm and have a low risk for malignant transforma- S128 TALWALKAR AND GORES GASTROENTEROLOGY Vol. 127, No. 5 tion, from true HCC. 38 High-grade dysplastic nodules, which are usually 1 cm in size and often missed by ultrasonography, CT, and MRI, develop into HCC in 30% of cases. 15,38 The ability to exclude multicentric HCC, which occurs in 20%40% of cases, also severely limits the effectiveness of ultrasonography for detecting early-stage disease. 39 Staging of Hepatocellular Carcinoma There are 2 components to the staging of HCC: (1) intrahepatic staging and (2) extrahepatic staging. Intrahepatic staging requires delineation of the number and size of lesions and whether vascular invasion is or is not present. Optimally, if 1 lesion is present, staging technology would permit the distinction between de novo lesions and intrahepatic metastases. Extrahepatic staging is performed to determine whether metastasis outside the liver is or is not present. Each of these issues is discussed in further detail. For critical decision making, the number and location of all intrahepatic lesions needs to be carefully assessed. For example, current transplant criteria for HCC include 1 lesion 5 cm in diameter or 3 lesions, all 3 cm. Precise and accurate staging would determine the accu- rate size of a lesion and the number of lesions present in the liver. As best can be determined, no published stud- ies correlated the actual size of lesions noted in explanted livers with size estimated from preoperative imaging studies. There also is a tendency to simply measure the largest diameter of a lesion. Volumetric studies should be possible and likely would be more reective of tumor burden. Genetic studies are necessary to distinguish between synchronous primary lesions and intrahepatic metastases. Comparative genomic hybridization and allelotyping with determination of loss of heterozygosity can be used for this purpose. Fractional allelic loss (quantitation of loss of heterozygosity) has been reported to make this distinction. 40 Furthermore, fractional allelic loss corre- lates with biological aggressiveness and outcome; high- frequency fractional allelic loss is associated with reduced disease-free survival after orthotopic liver transplanta- tion. 41 Diagnosis of recurrent disease, metachronous vs metastatic, also has been accomplished by means of com- parative genomic hybridization. 40 In this study, approx- imately one third of lesions were metachronous, one third were metastatic, and one third were not classiable. Thus, genetic approaches hold great promise for helping diagnose and stage HCC. Extrahepatic staging of HCC currently is based solely on such imaging studies as cross-sectional imaging of the chest and abdomen and bone scans. However, despite negative imaging study results, recurrent disease is com- mon after resection and liver transplantation for large tumors. 42 Moreover, recurrence in the graft is a common presentation in this context. These data suggest micro- metastases are common and often return to the fertile garden, the liver. Detection of micrometastases is an important clinical goal and will help dictate which ther- apies are optimal for a given patient. Attempts to cor- relate polymerase chain reactionbased detection of AFP and albumin levels in peripheral-blood samples and pa- tient outcome were unsuccessful and have been largely abandoned. In particular, these techniques were sensitive and detected circulating gene products in a wide variety of clinical circumstances. 43 However, a recent study re- ported the detection of circulating tumor cells in the blood of patients with HCC. 44 The presence of circulat- ing tumor cells was associated with limited survival compared with patients who did not have circulating cells present in peripheral blood. This nding needs to be veried. It also may be that micrometastases reside in bone marrow. In patients with breast cancer, detection of epithelial markers within bone marrow (ie, cytokeratin- positive cells) is associated with greater recurrence and lower survival rates in patients with apparent early-stage disease. 45 Again, this approach needs to be assessed in the context of HCC. Until the presence and clinical signif- icance of micrometastases can be determined, it remains difcult to establish which patients should be considered for locoregional therapies (eg, surgery, transplantation, ablation, and chemoembolization). Prognostic Staging Systems for Hepatocellular Carcinoma The natural history of HCC depends on the se- verity of the underlying cirrhosis, tumor characteristics (ie, size, multicentricity, presence or absence of vascular invasion, pathological grade, and extrahepatic metasta- ses), performance status of the patient, comorbid condi- tions, and the efcacy of treatment interventions. Four prognostic classications have been developed for HCC (Table 2). 4649 None of them is perfect. The Okuda classication uses a gross assessment of tumor factors and is not very useful in the clinical evaluation of patients. The Barcelona Clinic Liver Cancer, Cancer of the Liver Italian Program, and Groupe dEtude et de Traitement du Carcinome Hpatocellulaire classications use more rened assessments. These models cannot provide a pre- cise prognosis for individual patients, but nevertheless are helpful as a guide for prognosis and treatment of November Supplement 2004 HCC DIAGNOSIS S129 patient subgroups. For example, according to the Barce- lona Clinic Liver Cancer classication, patients with uni- centric disease, no symptoms, and no evidence of vascular invasion or extrahepatic spread have a 3-year survival rate of 50%. 46 Conversely, patients with severe cancer-related symptoms reected by deteriorated performance status and marked impairment of their liver function have a 50% survival rate at 3 months, regardless of tumor characteristics. The information provided by these mod- els needs to be kept in mind when considering treatment options for each patient. Summary Considerable advances have been made in the ability to diagnose HCC. These advances relate to the recognition that patients with cirrhosis are at high risk for this disease, resulting in the implementation of screening strategies. Cross-sectional imaging modalities, including CT and MR, have improved in resolution and speed, permitting imaging during the arterial phase of contrast administration. Given that most HCC lesions are highly vascular, small tumors frequently are identi- ed by means of contrast enhancement during MRI and CT. In a cirrhotic liver, the presence of a mass lesion 2 cm in diameter with contrast enhancement is virtu- ally diagnostic of HCC, and consensus criteria have been established for the noninvasive diagnosis of HCC. Diag- nosis of a small HCC nodule 2 cm in diameter is still challenging and requires advances in imaging modalities and/or cytological approaches (Table 3). Likewise, greater effort is required to diagnose micrometastases to aid in staging and therapeutic selection (Table 3). References 1. El-Serag HB. Hepatocellular carcinoma: an epidemiologic view. J Clin Gastroenterol 2002;35(Suppl):S72S78. 2. Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lancet 2003;362:19071917. 3. Llovet JM, Bruix J. Systematic review of randomized trials for unresectable hepatocellular carcinoma: chemoembolization im- proves survival. Hepatology 2003;37:429442. 4. Bruix J, Sherman M, Llovet JM, Beaugrand M, Lencioni R, Burroughs AK, Christensen E, Pagliaro L, Colombo M, Rodes J. Clinical management of hepatocellular carcinoma. Conclusions of the Barcelona-2000 EASL conference. European Association for the Study of the Liver. J Hepatol 2001;35:421430. 5. Torzilli G, Minagawa M, Takayama T, Inoue K, Hui AM, Kubota K, Ohtomo K, Makuuchi M. Accurate preoperative evaluation of liver mass lesions without ne-needle biopsy. Hepatology 1999;30: 889893. 6. Yu JS, Kim KW, Kim EK, Lee JT, Yoo HS. Contrast enhancement of small hepatocellular carcinoma: usefulness of three succes- sive early image acquisitions during multiphase dynamic MR imaging. AJR Am J Roentgenol 1999;173:597604. 7. Lim JH, Kim CK, Lee WJ, Park CK, Koh KC, Paik SW, Joh JW. Detection of hepatocellular carcinomas and dysplastic nodules in cirrhotic livers: accuracy of helical CT in transplant patients. AJR Am J Roentgenol 2000;175:693698. 8. Kim T, Murakami T, Takahashi S, Tsuda K, Tomoda K, Narumi Y, Oi H, Sakon M, Nakamura H. Optimal phases of dynamic CT for detecting hepatocellular carcinoma: evaluation of unenhanced and triple-phase images. Abdom Imaging 1999;24:473480. 9. Bru C, Maroto A, Bruix J, Faus R, Bianchi L, Calvet X, Ayuso C, Vilana R, Gilabert R, Rodes J. Diagnostic accuracy of ne-needle aspiration biopsy in patients with hepatocellular carcinoma. Dig Dis Sci 1989;34:17651769. 10. Longchampt E, Patriarche C, Fabre M. Accuracy of cytology vs microbiopsy for the diagnosis of well-differentiated hepatocellular carcinoma and macroregenerative nodule. Denition of standard- ized criteria from a study of 100 cases. Acta Cytol 2000;44:515 523. 11. Huang GT, Sheu JC, Yang PM, Lee HS, Wang TH, Chen DS. Ultrasound-guided cutting biopsy for the diagnosis of hepatocel- Table 3. Future Directions in the Diagnosis and Staging of Hepatocellular Carcinoma Diagnosis Genetic denition of hepatocellular carcinoma Cytological techniques for aneuploidy (eg, uorescent in situ hybridization) Staging Micrometastasis detection by means of molecular techniques Detection of circulating tumor cells Molecular techniques for staging intrahepatic lesions Table 2. Parameters Used in Prognostic Classications for Hepatocellular Carcinoma Parameters Okuda CLIP GETCHC BCLC Patient factors Performance status X X Tumor characteristics Tumor size X X X Vascular invasion X X X Extrahepatic spread X -Fetoprotein X X Liver function Child-Turcotte-Pugh stage X X Portal hypertension X Ascites X Bilirubin X X X Albumin X -Fetoprotein X Abbreviations: BCLC, Barcelona Clinic Liver Cancer; CLIP, Cancer of the Liver Italian Program; GETCHC, Groupe dEtude et de Traitement du Carcinome Hpatocellulaire. S130 TALWALKAR AND GORES GASTROENTEROLOGY Vol. 127, No. 5 lular carcinomaa study based on 420 patients. J Hepatol 1996;25:334338. 12. Durand F, Regimbeau JM, Belghiti J, Sauvanet A, Vilgrain V, Terris B, Moutardier V, Farges O, Valla D. Assessment of the benets and risks of percutaneous biopsy before surgical resection of hepatocellular carcinoma. J Hepatol 2001;35: 254258. 13. Takamori R, Wong LL, Dang C, Wong L. Needle-tract implantation from hepatocellular cancer: is needle biopsy of the liver always necessary? Liver Transpl 2000;6:6772. 14. Mion F, Grozel L, Boillot O, Paliard P, Berger F. Adult cirrhotic liver explants: precancerous lesions and undetected small hepatocel- lular carcinomas. Gastroenterology 1996;111:15871592. 15. Takayama T, Makuuchi M, Hirohashi S, Sakamoto M, Okazaki N, Takayasu K, Kosuge T, Motoo Y, Yamazaki S, Hasegawa H. Malignant transformation of adenomatous hyperplasia to hepa- tocellular carcinoma. Lancet 1990;336:11501153. 16. Borzio M, Fargion S, Borzio F, Fracanzani AL, Croce AM, Stroffolini T, Oldani S, Cotichini R, Roncalli M. Impact of large regenerative, low grade and high grade dysplastic nodules in hepatocellular carci- noma development. J Hepatol 2003;39:208214. 17. Caturelli E, Bartolucci F, Biasini E, Vigliotti ML, Andriulli A, Siena DA, Attino V, Bisceglia M. Diagnosis of liver nodules observed in chronic liver disease patients during ultrasound screening for early detec- tion of hepatocellular carcinoma. AmJ Gastroenterol 2002;97:397 405. 18. Nakashima T, Kojiro M. Pathologic characteristics of hepatocel- lular carcinoma. Semin Liver Dis 1986;6:259266. 19. Thorgeirsson SS, Grisham JW. Molecular pathogenesis of human hepatocellular carcinoma. Nat Genet 2002;31:339346. 20. Wilkens L, Flemming P, Gebel M, Bleck J, Terkamp C, Wingen L, Kreipe H, Schlegelberger B. Induction of aneuploidy by increasing chromosomal instability during dedifferentiation of hepatocellular carcinoma. Proc Natl Acad Sci U S A 2004;101: 13091314. 21. Baron TH, Harewood GC, Rumalla A, Pochron NL, Stadheim LM, Gores GJ, Therneau TM, De Groen PC, Sebo TJ, Salomao DR, Kipp BR. A prospective comparison of digital image analysis and routine cytology for the identication of malignancy in biliary tract strictures. Clin Gastroenterol Hepatol 2004;2:214219. 22. Soresi M, Magliarisi C, Campagna P, Leto G, Bonssuto G, Riili A, Carroccio A, Sesti R, Tripi S, Montalto G. Usefulness of alpha- fetoprotein in the diagnosis of hepatocellular carcinoma. Antican- cer Res 2003;23:17471753. 23. Maringhini A, Cottone M, Sciarrino E, Marceno MP, La Seta F, Fusco G, Rinaldi F, Pagliaro L. Ultrasonography and alpha-feto- protein in diagnosis of hepatocellular carcinoma in cirrhosis. Dig Dis Sci 1988;33:4751. 24. Sherman M. Alphafetoprotein: an obituary. J Hepatol 2001;34: 603605. 25. Gupta S, Bent S, Kohlwes J. Test characteristics of alpha-feto- protein for detecting hepatocellular carcinoma in patients with hepatitis C. A systematic review and critical analysis. Ann Intern Med 2003;139:4650. 26. Marrero JA, Su GL, Wei W, Emick D, Conjeevaram HS, Fontana RJ, Lok AS. Des-gamma carboxyprothrombin can differentiate hepatocellular carcinoma from nonmalignant chronic liver dis- ease in American patients. Hepatology 2003;37:11141121. 27. Kim CK, Lim JH, Lee WJ. Detection of hepatocellular carcinomas and dysplastic nodules in cirrhotic liver: accuracy of ultrasonog- raphy in transplant patients. J Ultrasound Med 2001;20:99 104. 28. Bennett GL, Krinsky GA, Abitbol RJ, Kim SY, Theise ND, Teper- man LW. Sonographic detection of hepatocellular carcinoma and dysplastic nodules in cirrhosis: correlation of pretransplantation sonography and liver explant pathology in 200 patients. AJR Am J Roentgenol 2002;179:75180. 29. Dodd GD III, Miller WJ, Baron RL, Skolnick ML, Campbell WL. Detection of malignant tumors in end-stage cirrhotic livers: ef- cacy of sonography as a screening technique. AJR Am J Roent- genol 1992;159:727733. 30. Taourel PG, Pageaux GP, Coste V, Fabre JM, Pradel JA, Ramos J, Larrey D, Domergue J, Michel H, Bruel JM. Small hepatocellular carcinoma in patients undergoing liver transplantation: detection with CT after injection of iodized oil. Radiology 1995;197:377 380. 31. Gambarin-Gelwan M, Wolf DC, Shapiro R, Schwartz ME, Min AD. Sensitivity of commonly available screening tests in detecting hepatocellular carcinoma in cirrhotic patients undergoing liver transplantation. Am J Gastroenterol 2000;95:15351538. 32. Teefey SA, Hildeboldt CC, Dehdashti F, Siegel BA, Peters MG, Heiken JP, Brown JJ, McFarland EG, Middleton WD, Balfe DM, Ritter JH. Detection of primary hepatic malignancy in liver trans- plant candidates: prospective comparison of CT, MR imaging, US, and PET. Radiology 2003;226:533542. 33. Yao FY, Ferrell L, Bass NM, Watson JJ, Bacchetti P, Venook A, Ascher NL, Roberts JP. Liver transplantation for hepatocellular carcinoma: expansion of the tumor size limits does not adversely impact survival. Hepatology 2001;33:13941403. 34. Chalasani N, Horlander JC Sr, Said A, Hoen H, Kopecky KK, Stockberger SM Jr, Manam R, Kwo PY, Lumeng L. Screening for hepatocellular carcinoma in patients with advanced cirrhosis. Am J Gastroenterol 1999;94:29882993. 35. Bizollon T, Rode A, Bancel B, Gueripel V, Ducerf C, Baulieux J, Trepo C. Diagnostic value and tolerance of Lipiodol-computed tomography for the detection of small hepatocellular carcinoma: correlation with pathologic examination of explanted livers. J Hepatol 1998;28:491496. 36. Burrel M, Llovet JM, Ayuso C, Iglesias C, Sala M, Miquel R, Caralt T, Ayuso JR, Sole M, Sanchez M, Bru C, Bruix J. MRI angiography is superior to helical CT for detection of HCC prior to liver transplantation: an explant correlation. Hepatology 2003;38:10341042. 37. Shimizu A, Ito K, Koike S, Fujita T, Shimizu K, Matsunaga N. Cirrhosis or chronic hepatitis: evaluation of small ( or 2-cm) early-enhancing hepatic lesions with serial contrast-enhanced dynamic MR imaging. Radiology 2003;226:550555. 38. Libbrecht L, Bielen D, Verslype C, Vanbeckevoort D, Pirenne J, Nevens F, Desmet V, Roskams T. Focal lesions in cirrhotic ex- plant livers: pathological evaluation and accuracy of pretrans- plantation imaging examinations. Liver Transpl 2002;8: 749761. 39. Bolondi L, Soa S, Siringo S, Gaiani S, Casali A, Zironi G, Piscaglia F, Gramantieri L, Zanetti M, Sherman M. Surveillance programme of cirrhotic patients for early diagnosis and treat- ment of hepatocellular carcinoma: a cost effectiveness anal- ysis. Gut 2001;48:251259. 40. Chen YJ, Yeh SH, Chen JT, Wu CC, Hsu MT, Tsai SF, Chen PJ, Lin CH. Chromosomal changes and clonality relationship be- tween primary and recurrent hepatocellular carcinoma. Gastro- enterology 2000;119:431440. 41. Finkelstein SD, Marsh W, Demetris AJ, Swalsky PA, Sasatomi E, Bonham A, Subotin M, Dvorchik I. Microdissection-based allelo- typing discriminates de novo tumor from intrahepatic spread in hepatocellular carcinoma. Hepatology 2003;37:871879. 42. Adam R, Del Gaudio M. Evolution of liver transplantation for hepatocellular carcinoma. J Hepatol 2003;39:888895. 43. Carr BI. An assay for micrometastatic hepatocellular carcinoma. Gastroenterology 1994;106:264267. 44. Vona G, Estepa L, Beroud C, Damotte D, Capron F, Nalpas B, Mineur A, Franco D, Lacour B, Pol S, Brechot C, Paterlini-Brechot P. Impact of cytomorphological detection of circulating tumor cells in patients with liver cancer. Hepatology 2004;39:792797. November Supplement 2004 HCC DIAGNOSIS S131 45. Braun S, Pantel K, Muller P, Janni W, Hepp F, Kentenich CR, Gastroph S, Wischnik A, Dimp T, Kindermann G, Riethmuller G, Schlimok G. Cytokeratin-positive cells in the bone marrow and survival of patients with stage I, II, or III breast cancer. N Engl J Med 2000;342:525533. 46. Llovet JM, Bustamante J, Castells A, Vilana R, Ayuso Mdel C, Sala M, Bru C, Rodes J, Bruix J. Natural history of untreated non- surgical hepatocellular carcinoma: rationale for the design and eval- uation of therapeutic trials. Hepatology 1999;29:6267. 47. Levy I, Sherman M. Staging of hepatocellular carcinoma: assess- ment of the CLIP, Okuda, and Child-Pugh staging systems in a cohort of 257 patients in Toronto. Gut 2002;50:881885. 48. Cancer of the Liver Italian Program Investigators. A new prognos- tic system for hepatocellular carcinoma: a retrospective study of 435 patients: the Cancer of the Liver Italian Program (CLIP) Investigators. Hepatology 1998;28:751755. 49. Chevret S, Trinchet JC, Mathieu D, Rached AA, Beaugrand M, Chastang C. A new prognostic classication for predicting sur- vival in patients with hepatocellular carcinoma. Groupe dEtude et de Traitement du Carcinome Hepatocellulaire. J Hepatol 1999; 31:133141. Address requests for reprints to: Gregory J. Gores, MD, Professor of Medicine, Mayo Medical School, Clinic, and Foundation, 200 First Street SW, Rochester, Minnesota 55905. e-mail: gores.gregory@ mayo.edu; fax: (507) 284-0762. Supported in part by grant no. DK41876 from the National Institutes of Health (to G.J.G.) and the Mayo and Palumbo Foundations. The excellent secretarial services of Erin Bungum are gratefully acknowledged. S132 TALWALKAR AND GORES GASTROENTEROLOGY Vol. 127, No. 5