Hepatitis Central

Screening for Hepatocellular Carcinoma

In contrast to certain other cancers,^1-4 screening for liver cancer has become, at least among hepatologists, an accepted part of the management of patients with end-stage liver disease. Yet there have been no randomized, controlled trials of screening or surveillance for hepatocellular carcinoma (HCC), nor are there adequate analyses of cost, efficacy, and potential benefit. It is the intent of this article to critically examine the rationale for screening patients with liver disease for liver cancer.

By definition, screening is the one-time application of a test that allows the detection of a disease at a stage when intervention may significantly improve the natural course and outcome.⁵ In contrast, surveillance is the repeated application of such tests over time. The objective of both is to reduce disease-specific mortality.

Prorok⁵ has enunciated the criteria by which any screening/surveillance program can be judged: 1) The disease must be common and have a substantial mortality and morbidity. 2) The target population must be easily identifiable, and there must be an expectation that the physicians caring for the population will accept that screening is necessary and that the population will answer the call for screening. 3) The screening test must have a low morbidity and a high sensitivity and specificity. 4) There must be standardized recall procedures. 5) The screening test must be acceptable to the target population. 6) Finally, and most importantly, there must be an acceptable and effective therapy.

In this article, we will examine HCC in light of these criteria, and we will discuss some of the challenges to instituting effective screening/surveillance programs for HCC.

THE DISEASE MUST BE COMMON AND MUST HAVE A SUBSTANTIAL MORBIDITY AND/OR MORTALITY

HCC is the fourth most common cancer in the world.⁶ Age-standardized incidence rates vary from 3 per 100,000 in North American men to 80 per 100,000 in China. ^6,7 HCC causes substantial morbidity and mortality. The reported survival rates for untreated symptomatic HCC vary from 0% at 4 months to 1% at 2 years.^8-10 Even small tumors found on screening or surveillance continue to have a significant mortality of at least 50% at 5 years when treated with resection or ethanol injection. However, even with the surveillance that is currently taking place, many small HCCs are incurable at diagnosis.

THE TARGET POPULATION MUST BE EASILY IDENTIFIABLE

Chronic hepatitis B and C are recognized as the major factors increasing the risk of HCC,^11-18 the risk being greater in the presence of coinfection with hepatitis B virus (HBV) and hepatitis C virus.¹⁶ The incidence of HCC in individuals with chronic hepatitis B is as high as 0.46% per year.^19-22 Cirrhosis is also a risk factor for HCC, irrespective of etiology. The annual risk of developing HCC in cirrhosis is between 1% and 6%. ^14-16,23-29 Although several studies have shown that the risk of HCC is higher in patients with cirrhosis caused by viral infection compared with nonviral causes, ^14,16,28,29 a high rate of HCC has also been reported in patients with cirrhosis caused by genetic hemochromatosis.³⁰ In contrast, low incidence rates are seen in biliary cirrhosis.³¹

Twenty to 56% of patients presenting with HCC have previously undiagnosed cirrhosis. ^32,33 These patients would not have been recruited into a surveillance program if the presence of cirrhosis was used to define the target population. To overcome this problem, surveillance has been extended to include patients with noncirrhotic chronic viral hepatitis as well as those with overt cirrhosis. The overall reported annual detection rate of HCC in surveillance studies, which included individuals with chronic hepatitis in addition to cirrhosis, is 0.8% to 4.1% ^34-40 (table 1).

Increasing age and male gender are also risk factors for HCC in the presence of cirrhosis. ^{12,21,29,41-44} No study has directly addressed the question of whether surveillance for HCC should or could be restricted to individuals over a certain age limit. The risk of HCC in HBV carriers is negligible before the age of 30.¹¹

THE SCREENING TEST MUST HAVE A LOW MORBIDITY AND HIGH SENSITIVITY AND SPECIFICITY

The most commonly used screening tests are serum -fetoprotein (AFP) and ultrasonography. In evaluating such tests, it is important to note that the performance characteristics of a test that is used for diagnosis may differ when used for screening/surveillance. Thus, reports of the performance characteristics of AFP and ultrasonography, when applied to screening/surveillance, must be interpreted cautiously.

AFP. This was the marker used in the first reported surveillance studies of HCC. ^29,44,45 The sensitivity of AFP for detecting HCC varies widely in both HBV-positive and predominantly HBV-negative populations, ^{22,25,27,28,44,46,47} possibly because of the confusion between diagnosis and screening. If the level of AFP triggering investigation for HCC is increased, e.g., from 20 mg/L to 100 mg/L,²⁵ the sensitivity of the AFP test falls from 39% to 13%, while the specificity increases. AFP, however, is not specific for HCC. Titers also rise with flares of active hepatitis.⁴⁶ Of 44 HBV carriers with elevated AFP levels detected during surveillance for HCC, only 6 were found to have HCC on further investigation, and, in 18 (41%), the elevated AFP was associated with an exacerbation of underlying liver disease or changes in HBV replication status.⁴⁴

In individuals with viral hepatitis who do not have HCC, AFP levels may be transiently, persistently, or intermittently elevated. Increases are most likely caused by viral hepatitis when they parallel elevations in transaminases, and may represent active hepatitis or a seroconversion illness. The diagnostic dilemma comes in differentiating HCC from viral infection when AFP levels do not correlate with alanine transaminase or occur in the presence of a normal alanine transaminase. Although AFP levels increase with time in the presence of HCC, the range overlaps with that seen in the absence of tumor. There are no guidelines as to when a rise in AFP level in the presence of a normal ultrasound should trigger further investigations to exclude HCC.

In a direct test of AFP for detecting HCC in HBV-negative and -positive patients, the specificity was only 50% in HBV-positive patients compared with 78% in HBV-negative patients.⁴⁷ Performance characteristics of AFP as a screening test were reported in three studies. ^22,25,27 In these studies, the screening methodology was well described, and it was clear that AFP was used for surveillance. These studies report a sensitivity of 39% to 64%, a specificity of 76% to 91%, and a positive predictive value of 9% to 32%.

    Ultrasound. The poor sensitivity and specificity of AFP alone as a surveillance tool led to the use of ultrasound scanning in addition to, or in place of, AFP. The performance characteristics of ultrasound as a screening test for HCC have been defined in both healthy hepatitis B surface antigen carriers (noncirrhotic)²² and in patients with cirrhosis.²⁷ The sensitivity was 71% and 78%, respectively, and the specificity was 93%. The positive predictive value was 14% and 73%, respectively. It remains to be determined whether these performance characteristics make serum AFP and ultrasonography efficient, economical tests for HCC surveillance.

Surveillance Interval. Reported surveillance intervals vary from 3 to 12 months. However, reasons for choosing these intervals are often not reported. A 6-month surveillance interval may be a rational choice, based on data from China, in which tumor doubling time in asymptomatic HCC less than 5 cm was studied. ^48,49 Sheu et al.⁴⁹ found a median tumor doubling time of 117 days. AFP levels corresponded with tumor doubling time in 17 of 31 tumors studied. The most rapidly dividing tumor took 5 months to increase in size from 1 to 3 cm. Therefore, 6-month surveillance is a reasonable interval to detect tumors growing from undetectable to detectable size.

THERE MUST BE STANDARDIZED RECALL POLICIES

There is no literature to guide decision-making on the most appropriate way to deal with abnormal screening test results. Thus, physicians undertaking surveillance programs must rely on their clinical judgment. The recall policies should be sensitive to the possibility of false-positive tests so that additional investigations, particularly invasive tests, for patients with common nonmalignant lesion such as hemangioma will be minimized.

Confirmatory Tests

Radiology. A variety of radiological investigations have been used to confirm ultrasound findings in patients with cirrhosis and chronic hepatitis with an isolated raised AFP. These include computerized tomography (CT), spiral CT, magnetic resonance imaging (MRI), lipiodol-CT, and hepatic angiography. For lesions greater than 3 cm, the overall sensitivity and specificity of contrast-enhanced CT are 68% and 81%, respectively.⁵⁰ Smaller tumors are better detected by MRI, with 81% sensitivity for tumors less than 2 cm.⁵¹ Spiral CT scanning is even more sensitive, with 87% of tumors less than 1 cm being detected compared with 64% by MRI.⁵² Sensitivity can be improved still further by using techniques such as lipiodol-CT scanning. Sensitivities of 93% to 97% have been reported, ^53,54 although the level falls to 86% for tumors less than 1 cm. All of these radiological techniques are also subject to unquantified false-positive rates. The relative sensitivity of these tests in detecting HCC in noncirrhotic or cirrhotic liver is also unknown.

Ultrasound is relatively poor at identifying multifocal tumor in cirrhotic patients. In one study, explant histology was compared with pretransplant radiological imaging in 30 patients. Although the sensitivity for ultrasound was 80% overall, it was only 16% for multicentricity, compared with 86% and 58%, respectively, for CT.⁵⁵ Combined ultrasound, CT, and hepatic angiography detected only 62% of nodules. Similar studies have shown that CT and MRI have reduced sensitivity for multicentricity, with accuracy rates of 50% to 83% for CT and 57% to 81% for MRI.^56-58 Again, there are no studies directly assessing the usefulness of radiological imaging in detecting multifocal HCC in the absence of cirrhosis. The ability to detect multicentricity is important if hepatic resection is a therapeutic option, but may be less important if liver transplantation (complete hepatectomy) is the therapy of choice.

Tissue Diagnosis. The use of biopsy to confirm HCC is controversial. It can be difficult to distinguish large cirrhotic nodules from well-differentiated HCC or low-grade dysplastic nodules from HCC in either needle or wedge biopsies.⁵⁹ Liver biopsy also carries a small risk of tumor spread along the needle track.⁶⁰ Fine-needle aspirates provide cells without some of the architectural abnormalities that are important in making a diagnosis. Therefore, fine-needle aspirates is not recommended for distinguishing cirrhotic nodules from small neoplastic lesions that are likely to have subtle abnormalities.

Unfortunately, none of the reported surveillance studies have described the additional investigations needed to confirm the presence of HCC nor the amount of secondary imaging needed to exclude HCC in the case of a false-positive screening test. Such data are important in analyzing cost benefit and also in determining the acceptability of surveillance to the target population.

THE SCREENING TESTS MUST BE ACCEPtable TO THE TARGET POPULATION

The acceptability of serum AFP and ultrasound scans as screening tools can be inferred indirectly from the number of subjects lost to follow-up during a surveillance program. Generally, compliance was better in programs involving regular clinic attendees with liver disease than in those with screening asymptomatic HBV carriers exhibiting little or no liver disease. In one population-based surveillance study of 1,069 HBV carriers, 6.7% failed to attend any follow-up, and a further 17% withdrew from the study after one or more follow-up screening visits.²² Between 3% and 18% of clinic-based cirrhotic patients were noncompliant. ^{14,16,26,27,28,34,39,45,61} These results compare with a 25% lack of compliance with fecal occult blood testing, in a colorectal cancer screening program of asymptomatic individuals.⁶² Access is also an important factor in Third-World countries where screening with ultrasound is not feasible, and even AFP screening may be too expensive.

THERE MUST BE EFFECTIVE THERAPY

For screening/surveillance to improve mortality, effective therapy must be available. Therapeutic options include hepatic resection, liver transplantation, and percutaneous alcohol injection. These have recently been reviewed by Bruix.⁶³

Although studies on the efficacy of these various modalities have reported better survival rates for smaller tumors (less than 3 cm), all have been uncontrolled and nonrandomized. This makes evaluation of the efficacy of different treatment modalities for HCC extremely difficult. Many of the reported studies suffer from lead-time bias, defined as an apparent improvement in survival due only to early detection. Studies comparing treatment of small versus large tumors all have this bias. Even if treatment in both arms were equally effective, or equally ineffective, such studies would show improved survival in the small tumor group simply because of earlier diagnosis. It follows that, in a nonrandomized, nonmatched study, it is impossible to determine whether improved survival in the early diagnosis or small tumor group is caused by treatment effect or by lead-time bias. Treatment and early diagnosis programs also preferentially identify patients with more slowly progressive disease, who may survive longer for that reason rather than because of the intervention applied. Moreover, potentially curative forms of therapy including resection, liver transplantation, and various forms of local ablation have been studied in patients referred for tumor management, not those diagnosed through surveillance. Because the 5-year survival rate of a cohort with untreated HCC smaller than 3 cm is unknown, it is impossible to accurately evaluate the effectiveness of any of these treatment strategies. This again emphasizes the importance of randomized, controlled trials to show that screening followed by currently available therapies actually does extend life expectancy.

Having discussed the six criteria needed to institute a screening program for HCC, we will now turn our attention to the importance of showing that screening/surveillance reduces mortality and to the results of published screening/surveillance programs for HCC. Finally, some of the challenges, including cost, to introducing a successful surveillance program for HCC will be discussed.

SURVEILLANCE SHOULD REDUCE MORTALITY FROM THE DISEASE

Only controlled trials of surveillance versus no surveillance can demonstrate whether surveillance for HCC reduces mortality from the disease. There are no randomized, controlled studies of surveillance, and only a few studies have examined the outcomes of HCC detected by surveillance and outcomes of tumors not detected by surveillance over the same period. Several studies have attempted to compare the clinical features of asymptomatic HCC detected by screening with symptomatic tumor, showing that tumors detected through surveillance are smaller and more amenable to potentially curative therapy. ^33,38,64-66 In a study from Japan, 81% of 391 HCC detected by surveillance were considered suitable for curative resection compared with only 46% of 1,251 symptomatic HCC.⁶⁶ The overall 5-year survival rate was 51% for asymptomatic tumors compared with 21% for symptomatic HCC. In all these studies, the duration of follow-up postresection was limited (3 to 5 years), leaving doubt as to whether cure was achieved. Furthermore, all these studies are subject to lead-time bias, as discussed earlier.

It is clear from looking at the outcome data of individual surveillance studies that the tumor size at diagnosis was not the only factor directing therapy. Unfortunately, fewer than half of the published studies report outcome data (table 2). Of tumors detected by surveillance, 50% to 75% were unifocal and 3 cm or less in size, and thus potentially curable. ^{23,28,35,36,40,42} This did not translate into a comparable resection rate. In the majority of studies, the overall surgical resection rate varied from 29 to 54%. ^{22,25-27,35-37,40,42} Failure to undertake hepatic resection was due mainly to age, patient wishes, cirrhosis, and impaired synthetic function or a poor general medical condition. ^34,42 The preference for hepatic resection rather than liver transplantation in all the studies was not discussed. The recently reported high 3-year survival rates for liver transplantation for small tumors⁶⁷ may mean increased use of this form of therapy in the future, but, in the absence of randomized, controlled trials, this cannot be accurately assessed. Interestingly, there was no difference in the resectability of asymptomatic tumors detected at the start of a surveillance program (prevalent tumors) compared with those detected during continued surveillance (incident tumors).²⁸

EFFICACY OF PUBLISHED SCREENING AND SURVEILLANCE PROGRAMS

A number of screening and surveillance programs have been reported (Tables 1 and 3. The results range from very optimistic to downright pessimistic. McMahon et al. screened 1,400 HBV carriers in Alaska over a 5-year period, using AFP as the only screening test.⁴⁶ They detected 15 tumors, of which 10 were resectable . After 5 years of follow-up, 4 tumors had recurred (2 patients had been followed to 2 years only, with no recurrence). In contrast, Sherman et al. prospectively screened 1,069 HBV carriers for periods of 6 months to 5 years.²² Over this period, 14 tumors were detected. Seven were resectable , and 6 patients actually underwent surgery. There were two postoperative recurrences and one postoperative death. Only 3 patients survived more than 2 years from diagnosis.

Studies from Asia report large numbers of subjects screened, but outcome data are frequently lacking. It is clear that screening and surveillance result in finding small tumors in asymptomatic individuals, the majority of whom are treated for cure, either by resection or ethanol injection. However, results in different parts of the world vary widely, the best coming from Asia. Thus, local demographics must contribute to the outcome.

CHALLENGES IN THE EVALUATION OF SURVEILLANCE PROGRAMS

Ideally, the way to determine the efficacy of a surveillance program for HCC would be to conduct a randomized, controlled trial of surveillance versus usual care, with disease-specific and all-cause mortality as the end points. The sample size required depends on the incidence of HCC. Assuming an incidence of 0.4% to 0.5%, the sample size may be up to 12,000 subjects. Unfortunately, this may not be feasible in North America. A suggested alternative is to compare surveillance with AFP alone versus AFP plus ultrasound, or ultrasound alone. However, such a study would not establish the efficacy of surveillance.

It can be argued that screening without evidence of efficacy is unethical, because surveillance involves not only the inconvenience of regular blood tests, ultrasounds, and extensive secondary radiological imaging, but also results in the diagnosis, albeit early, of tumors that are still untreatable . However, if only small HCC are amenable to therapy, then the approach may be to use the best surveillance tools (currently AFP and ultrasound) to find small HCCs and to study the optimal treatment of these lesions through randomized, controlled therapy trials. Given the low resectability rate and survival after surveillance in most Western centers, this is the only way that continued surveillance can be justified. In other words, if treatment trials are not available in a given area for patients with small HCCs, surveillance is inappropriate, at least in North America.

CHALLENGES TO THE INSTITUTION OF EFFECTIVE SCREENING SURVEILLANCE PROGRAM

Given effective surveillance, a program still faces challenges to its eventual success. These challenges relate to cost and the acceptability of surveillance to the physician.

    Cost. Although the cost-effectiveness of screening for colon carcinoma has been presented in several articles, there are only three analyses of the cost of screening for HCC. Kang et al.,⁶⁸ in a study of HBV carriers, concluded that using AFP and ultrasound screening yearly would detect 90% of early tumors at a cost of $11,800 each. However, their assumptions about the size of tumors detectable based on presumed tumor growth rates are not supported by surveillance studies. In a large mass screening study of 8,090 Japanese patients, 70% of whom were at high risk with either a history of liver disease, or of HBV or hepatitis C virus infection or a family history of HCC, 91 tumors were detected (1.1%) at a cost of $25,000 per tumor.⁶⁹ The overall survival of the patients found to have HCC was only 19% at 5 years. Therefore, the cost per death prevented will be higher. In neither of these cost-benefit analyses were postscreening diagnostic costs included, nor was the cost per year of life gained reported. In a third study of selected individuals with Child grade A cirrhosis, the cost per life per year gained was between $26,000 and $55,000.⁷⁰ This contrasts with $25,000 per life year saved by screening for colorectal cancer, which is considered to be cost-effective.⁷¹

Screening/surveillance for HCC has become accepted practice by hepatologists worldwide, particularly in patients with cirrhosis. This is despite the lack of evidence of benefit in reducing mortality. Tumors can certainly be detected earlier through screening/surveillance, but outcomes for such tumors will always be better than for larger tumors simply because of lead-time bias. Nonetheless, it is unlikely that a randomized, controlled study of screening/surveillance for HCC will now be performed. Future controlled trials must therefore concentrate on assessing the effectiveness of different therapies (i.e., hepatic resection, liver transplantation, and percutaneous alcohol injection) for small tumors detected through screening. Until such studies are performed, we cannot know whether we are benefiting our patients by subjecting them to regular AFP/ultrasound surveillance or just falsely reassuring ourselves.

Abbreviations: HCC, hepatocellular carcinoma; HBV, hepatitis B virus; AFP, -fetoprotein; CT, computerized tomography; MRI, magnetic resonance imaging.

Received September 19, 1997; accepted November 4, 1997.

Address reprint requests to: Dr. M. Sherman, Room EN9-223, The Toronto Hospital (General Division), 200 Elizabeth St., Toronto, Ontario M5G 2C4, Canada. Fax: (416) 591-2107.

References

     1.  Miller AB, Bains CJ, To T, Wall C. Canadian National Breast Screening Study

     2.  Breast cancer detection and deaths among women aged 50-59 years. Can Med Assoc J 1992;147: 1477-1488.

3. Krahn MD, Mahoney JE, Eckman MH, Trachtenberg J, Pauker SG, Detsky AS. Screening for prostate cancer. A decision analytic review. JAMA 1994; 272: 773-780.
4. Levin B, Bond JH. American Gastroenterological Association. Colorectal Cancer Screening: recommendations of the US preventive services task force. Gastroenterology 1996; 111: 1381-1384.
5. Prorok PC. Epidemiologic approach for cancer screening. Problems in design and analysis of trials. Am J Pediatr Haematol Oncol 1992; 14: 117-128
6. Parkin DM, Muir CS, Whelan SL, Gao Y-T, Ferlay J, Powell J. Cancer incidence in five continents. Vol 6. Lyon: IARC Scientific Publications , 1992.
7. World Health Organization. 1994 World Health Statistics Annual. Geneva: WHO , 1995.
8. Okuda K, Ohtsuki T, Obata H, Tomimatsu M, Okazaki N, Hasegawa H, Nakajima Y, et al. Natural history of hepatocellular carcinoma and prognosis in relation to treatment. Study of 850 patients. Cancer 1985; 56: 918-928.
9. Calvert X, Bruix J, Gines P, Bru C, Sole M, Vilana R, Rodes J. Prognostic factors of hepatocellular carcinoma in the West: a multivariate analysis in 206 patients. HEPATOLOGY 1990; 12: 753-760.
10. Stuart KE, Anand AJ, Jenkins RL. Hepatocellular carcinoma in the United States. Prognostic features, treatment outcome and survival. Cancer 1986; 77: 2217-2222.
11. Beasley RP, Hwang LY, Lin CC, Chien CS. Hepatocellular carcinoma and hepatitis B virus: a prospective study of 22,700 men in Taiwan. Lancet 1981; 2: 1129-1133.
12. Beasley RP. Hepatitis B virus. Cancer 1988; 61: 1942-1956.
13. Sakuma K, Saitoh N, Kasai M, Jitsukawa H, Yshino I, Yamaguchi M, Nobutomo K, et al. Relative risks of death due to liver disease among Japanese male adults having various statuses for hepatitis B s and e antigen/antibody in serum: a prospective study. HEPATOLOGY 1988; 8: 1642-1646.
14. Kato Y, Nakata K, Omagari K, Furukawa R, Kusumoto Y, Mori I, Tajima H, et al. Risk of hepatocellular carcinoma in patients with cirrhosis in Japan. Cancer 1994; 74: 2234-2238.
15. Ikeda K, Saitoh S, Koida I, Arase Y, Tsubota A, Chayama K, Kumada H, et al. A multivariate analysis of risk factors for hepatocellular carcinoma: a prospective observation of 795 patients with viral and alcoholic cirrhosis. HEPATOLOGY 1993; 18: 47-53.
16. Benvegnu L, Fattovich G, Noventa F, Tremolada F, Chemello L, Cecchetto A, Alberti A. Concurrent hepatitis B and C virus infection and risk of hepatocellular carcinoma in cirrhosis. A prospective study. Cancer 1994; 74: 2442-2448.
17. Kew MC, Houghton M, Choo QL, Kuo G. Hepatitis C virus antibodies in South African blacks with hepatocellular carcinoma. Lancet 1990; 335: 873-874.
18. Colombo M, Kuo G, Choo Q, Donato MF, Del Ninno E, Tommasini MA, Dioguardi N, et al. Prevalence of antibodies to hepatitis C virus in Italian patients with hepatocellular carcinoma. Lancet 1989; 2: 1006-1008.
19. Bruix J, Barrera JM, Calvert X, Ercilla G, Costa J, Sanchez-Tapias JM, Ventura M, et al. Prevalence of antibodies to hepatitis C virus in Spanish patients with hepatocellular carcinoma and hepatitis cirrhosis. Lancet 1989; 2: 1004-1006.
20. Villeneuve J-P, Desrochers M, Infante-Rivard C, Willems B, Raymond G, Bourcier M, Cote J, et al. A long-term follow-up study of asymptomatic hepatitis B surface antigen-positive carriers in Montreal. Gastroenterology 1994; 108: 1000-1005.
21. McMahon BJ, Alberts SR, Wainwright RB, Bulkow L, Lanier AP. Hepatitis B-related sequelae. Prospective study of 1400 hepatitis B surface antigen-positive Alaska native carriers. Arch Intern Med 1990; 150: 1051-1054.
22. Sherman M, Peltekian KM, Lee C. Screening for hepatocellular carcinoma in chronic carriers of hepatitis B virus: incidence and prevalence of hepatocellular carcinoma in a North American urban population. HEPATOLOGY 1995; 22: 432-438.
23. Cottone M, Turri M, Caltagirone M, Parisi P, Orlando A, Fiorentino G, Virdone R, et al. Screening for hepatocellular carcinoma in patients with Childs A cirrhosis: an 8 year prospective study by ultrasound and alphafetoprotein. J Hepatol 1994; 21: 1029-1034.
24. Zoli M, Magolotti D, Bianchi G, Gueli C, Marchenisi G, Pisa E. Efficacy of a surveillance program for early detection of hepatocellular carcinoma. Cancer 1996; 78: 977-985.
25. Oka H, Tamori A, Kuroki T, Kobayashi K, Yamamoto S. Prospective study of alpha-fetoprotein in cirrhotic patients monitored for development of hepatocellular carcinoma. HEPATOLOGY 1994; 19: 61-66.
26. Borzio M, Bruno S, Roncalli M, Mels GC, Ramella G, Borzio F, Leandro G, et al. Liver cell dysplasia is a major risk factor for hepatocellular carcinoma in cirrhosis. A prospective study. Gastroenterology 1995; 108: 812-817.
27. Pateron D, Ganne N, Trinchet JC, Aurousseau MH, Mal F, Meicler C, Coderc E, et al. Prospective study of screening for hepatocellular carcinoma in Caucasian patients with cirrhosis. J Hepatol 1994; 20: 65-71.
28. Colombo M, De Franchis R, Del Ninno E, Sangiovanni A, De Fazio C, Tommasini M, Donato MF, et al. Hepatocellular carcinoma in Italian patients with cirrhosis. Lancet 1991; 325: 675-680.
29. Zaman SN, Melia WM, Johnson RD, Portmann BC, Johnson PJ, Williams R. Risk factors in development of hepatocellular carcinoma in cirrhosis: prospective study of 613 patients. Lancet 1985; 1: 1357-1360.
30. Farigon S, Fracanzani AL, Piperno A, Braga M, DÕAlba R, Ronchi G, Fiorelli G. Prognostic factors for hepatocellular carcinoma in genetic haemochromatosis. HEPATOLOGY 1994; 20: 1426-1431.
31. Farinati F, Floreani A, DeMaria N, Fagiuoli S, Naccarato R, Chiaramonte M. Hepatocellular carcinoma in primary biliary cirrhosis. J Hepatol 1994; 21: 315-318.
32. Zaman SN, Johnson PJ, Williams R. Silent cirrhosis in patients with hepatocellular carcinoma. Implications for screening in high-incidence and low-incidence areas. Cancer 1990; 65: 1607-1610.
33. The Liver Cancer Study Group of Japan. Primary liver cancer in Japan. Clinicopathologic features and results of surgical treatment. Ann Surg 1990; 211: 277-287.
34. Chiba T, Matsuzaki Y, Abei M, Shoda J, Tanaka N, Osuga T, Aikawa T. The role of previous hepatitis B virus infection and heavy smoking in hepatitis C virus-related hepatocellular carcinoma. Am J Gastroenterol 1996; 91: 1195-1211.
35. Tanaka S, Kitamura T, Nakanishi K, Okuda S, Yamazaki H, Hiyama T, Fujimoto I. Effectiveness of periodic checkup by ultrasonography for the early diagnosis of hepatocellular carcinoma. Cancer 1990; 66: 2210-2214.
36. Sheu J-C, Sung J-L, Chen D-S, Lai M-Y, Wang T-H, Yu J-Y, Yang P-M, et al. Early detection of hepatocellular carcinoma by real-time ultrasonography. A prospective study. Cancer 1985; 56: 660-666.
37. Curley SA, Izzo F, Gallipoli A, de Bellis M, Cremona F, Parisi V. Identification and screening of 416 patients with chronic hepatitis at high risk to develop hepatocellular carcinoma. Ann Surg 1995; 222: 375-383.
38. Tsukuma H, Hiyama T, Tanaka S, Nakao M, Yabuuchi T, Kitamura T, Nakanishi K, et al. Risk factors for hepatocellular carcinoma among patients with chronic liver disease. N Engl J Med 1993; 328: 1797-1801.
39. Takano S, Yokosuka O, Imazeki F, Tagawa M, Omata M. Incidence of hepatocellular carcinoma in chronic hepatitis B and C: a prospective study of 251 patients. HEPATOLOGY 1995; 21: 650-655.
40. Solmi L, Primerano AMM, Gandolfi L. Ultrasound follow-up of patients at risk for hepatocellular carcinoma: results of a prospective study in 360 cases. Am J Gastroenterol 1996; 91: 1189-1193.
41. Villa E, Baldini GM, Pasquinelli C, Melegari M, Cariani E, Di Chirico G, Manenti F. Risk factors for hepatocellular carcinoma in Italy: male sex, hepatitis B virus, non-A non-B infection and alcohol. Cancer 1988; 62: 611-615.
42. Liaw Y-F, Tai D-I, Chu C-M, Lin D-Y, Sheen I-S, Chen T-J, Pao CC. Early detection of hepatocellular carcinoma in patients with chronic type B hepatitis. Gastroenterology 1986; 90: 263-266.
43. Fattovich G, Brollo L, Giustina G, Noventa F, Pontisso P, Alberti A, Realdi G, et al. Natural history and prognostic factors for chronic hepatitis type B. Gut 1991; 32: 294-298.
44. Lok ASF, Lai C-L. Alpha-fetoprotein monitoring in Chinese patients with chronic hepatitis B virus infection: role in the early detection of hepatocellular carcinoma. HEPATOLOGY 1989; 9: 110-115.
45. McMahon BJ, Wainwright RW, Lanier AP. The Alaska Native HCC screening program: a population-based screening program for hepatocellular carcinoma. In: Tabor E, Di Bisceglie AM, Purcell RH, eds. Etiology, Pathology and Treatment of Hepatocellular Carcinoma in North America. Houston: Gulf, 1991:231-241.
46. Bisceglie AM, Hoofnagle JH. Elevations of serum alpha-fetoprotein levels in patients with chronic hepatitis B. Cancer 1989; 64: 2117-2120.
47. Lee H-S, Chung YH, Kim CY. Specificities of serum alpha-fetoprotein in HBsAg⁺ and HBsAgpatients in the diagnosis of hepatocellular carcinoma. HEPATOLOGY 1991; 14: 68-72.
48. Chen DS, Sung JL, Sheu JC, Lai MY, How SW, Hsu HC, Lee CS, et al. Serum alpha fetoprotein in the early stage of human hepatocellular carcinoma. Gastroenterology 1984; 86: 1404-1409.
49. Sheu JC, Sung JL, Chen DS, Yang PM, Lau MY, Lee CS, Hsu HC, et al. Growth rates of asymptomatic hepatocellular carcinoma and its clinical implications. Gastroenterology 1985; 89: 259-266.
50. Miller WJ, Baron RL, Dodd GD, Federle MP. Malignancies in patients with cirrhosis: CT sensitivity and specificity in 200 consecutive transplant patients. Radiology 1994; 193: 645-650.
51. Hirai K, Aoki Y, Majima Y, Abe H, Nakashima O, Kojiro M, Tanikawa K. Magnetic resonance imaging of small hepatocellular carcinoma. Am J Gastroenterol 1991; 86: 205-209.
52. Murakami T, Kim T, Oi H, Nakamura H, Igarashi H, Matsushita M, Okamura J, et al. Detectability of hypervascular hepatocellular carcinoma by arterial phase imaging of MRI and spiral CT. Acta Radiol 1995; 36: 372-376.
53. Ngan H. Lipiodal computerized tomography: how sensitive and specific is the technique in the diagnosis of hepatocellular carcinoma? Br J Radiol 1990; 63: 771-775.
54. Takayasu K, Moriyama N, Muramatsu Y, Makuuchi M, Hasegawa H, Okazaki N, Hirohashi S. The diagnosis of small hepatocellular carcinomas: efficacy of various imaging procedures in 100 patients. AJR 1990; 155: 49-54.
55. Rizzi PM, Kane PA, Ryder SD, Ramage JK, Gane E, Tan KC, Portmann B, et al. Accuracy of radiology in detection of hepatocellular carcinoma before liver transplantation. Gastroenterology 1994; 107: 1425-1429.
56. De Santis M, Romagnoli R, Cristani A, Cioni G, Casolo A, Vici FF, Ventura E. MRI of small hepatocellular carcinoma: comparison with US, CT, DSA and lipiodal-CT. J Comput Assist Tomogr 1992; 16: 189-195.
57. Taourel PG, Pageaux GP, Coste V, Fabre J-M, Pradel JA, Ramos J, Larrey D, et al. Small hepatocellular carcinoma in patients undergoing liver transplantation: detection with CT after injection of iodized oil. Radiology 1995; 197: 377-380.
58. Winter TC, Takayasu K, Muramatsu Y, Furukawa H, Wakao F, Koga H, Sakamoto M, et al. Early advanced hepatocellular carcinoma: evaluation of CT and MR appearance with pathological correlation. Radiology 1994; 192: 379-387.
59. International working party. Terminology of nodular hepatocellular lesions. HEPATOLOGY 1995; 22: 983-993.
60. John TG, Garden OJ. Needle track seeding of primary and secondary liver carcinoma after percutaneous liver biopsy. HPB Surgery 1993; 6: 199-203.
61. Oka H, Kurioka N, Kim K, Kanno T, Kuroki T, Mizoguchi Y, Kobayashi K. Prospective study of early detection of hepatocellular carcinoma in patients with cirrhosis. HEPATOLOGY 1990; 12: 680-687.
62. Mandel JS, Bond JH, Church TR, Snover DC, Bradley M, Schuman LM, Ederer F. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993; 328: 1365-1371.
63. Bruix J. Treatment of hepatocellular carcinoma. HEPATOLOGY 1997; 25: 259-262.
64. Lee C-S, Sung J-L, Hwang L-Y, Sheu J-C, Chen D-S, Lin T-Y, Beasley RP. Surgical treatment of 109 patients with symptomatic and asymptomatic hepatocellular carcinoma. Surgery 1986; 99: 481-489.
65. Lai CL, Lau JY, Wu PC, Hui WM, Lai EC, Fan ST, Ngan H, et al. Subclinical hepatocellular carcinoma in Hong Kong Chinese. Oncology 1992; 49: 347-353.
66. Tang ZY, Yu YQ, Zhou XD, Yang BH, Ma ZC, Lin ZY. Subclinical hepatocellular carcinoma: an analysis of 391 patients. J Surg Oncol 1993; 3: 55-58.
67. Mazzaferro V, Regalia E, Doci R, Androla S, Pulvirenti A, Bozzetti F, Montalto F, et al. Liver transplantation for the treatment of small hepatocellular carcinoma in patients with cirrhosis. N Engl J Med 1996; 334: 693-699.
68. Kang JY, Lee TP, Yap I, Lun KC. Analysis of cost-effectiveness of different strategies for hepatocellular carcinoma in screening for hepatitis B virus carriers. J Gastroenterol Hepatol 1992; 7: 463-468.
69. Mima S, Sekiya C, Kanagawa H, Kohyama H, Gotoh K, Mizuo H, Iriji M, et al. Mass screening for hepatocellular carcinoma: experience in Hokkaido, Japan. J Gastroenterol Hepatol 1994; 9: 361-365.
70. Sarasin FP, Giostra E, Hadengue A. Cost-effectiveness of screening for detection of small hepatocellular carcinoma in Western patients with Child-Pugh Class A cirrhosis. Am J Med 1996; 171: 422-434.
71. Lieberman DA. Cost-effectiveness model for colon cancer screening. Gastroenterology 1995; 109: 1781-1790.

Copyright © 1998 by the American Association for the Study of Liver Diseases.