Prospective Study of Early Detection of Hepatocellular Carcinoma in Patients with Cirrhosis HIROKOOKA, NARITOKURIOKA, KOSYUN KIM, TORUKA”0, TETSUOKUROKI,YASUHIRO MIZOGUCHIAND mNZ0 KOBAYASHI Third Department of Internal Medicine, Osaka City University Medical School, Osaka, 545, Japan

We prospectively monitored 140 cirrhotic patients for the development of hepatocellular carcinoma for 6 yr, using periodical screening by high-resolution convex-array ultrasonography and a-fetoprotein. Twenty-eight patients were positive for HBs antigen, 26 patients had received blood transfusions and were negative for H B s antigen and 26 patients had a history of heavy drinking. We detected hepatocellular carcinoma in 40 patients during this period. The overall cumulative incidence of hepatocellular carcinoma in the 6 yr was 39%;the cumulative incidence was 59%in patients with HBsAg, 53%in patients who had had blood transfusions and were negative for HBsAg and 22%in patients who had a history of heavy drinking and who were without HBsAg. Detection of the carcinoma in 85%of these 40 patients was based on results of ultrasonography. Twenty-six of the patients (65%) had a small hepatocellular carcinoma of 2 cm or less. a-Fetoproteinlevels were lower than 100 ng/d in 56% of these 40 patients. Patients with cirrhosis are at high risk of developing hepatocellular carcinoma, especially patients with HBsAg or with a history of blood transfusion who are negative for HBsAg. Periodic monitoring by use of ultrasonography in particular is recommended for early detection of hepatocellular carcinoma. (HEPA~LOGY 1990;12:680-687.)

With the advent of high-resolution convex-array real-time US, the entire liver can be scrutinized quickly and a hepatic tumor as small as 0.5 cm across can be seen. In showing the uppermost portion of the right lobe of the liver, convex-array scanners are superior to linear-array scanners. We did this 6-yr prospective study of patients with cirrhosis using AFP and convex-array US scanners for the early detection of HCC and examined the incidence of HCC according to the different backgrounds. PATIENTS AND METHODS

Patients. During the 72 mo from January 1983 to December 1988, we studied the 140 consecutive outpatients with cirrhosis (histologically confirmed in 60 patients and clinically and biochemically evident in the remainder) from whom consent was obtained. At the time of their entry into the study, HCC was not suspected in the 95 men and 45 women with a mean age of 54.9 yr. The clinical background of the patients was as follows: twenty-eight patients (20%)were positive for HBsAg; 26 patients (19%)had had blood transfusions and were negative for HBsAg, and 26 patients (19%) had a history of heavy drinking, having consumed more than 75 gm of alcohol per day for more than 10 yr; they were negative for HBsAg. Three of the 28 patients who had HBsAg were heavy drinkers. As a rule, all patients were prospectively monitored every 2 mo HCC is one of the most prevalent malignancies in the by AFP and every 3 mo by US. The mean length of follow-up world, particularly in Asia and sub-Saharan Africa (1,2). was 41.1 mo (median = 36.2 mo), with a range of 2 mo to The incidence of HCC is low in some areas, but it has a 72 mo. Methods. US was done with linear-array real-time scanners poor prognosis everywhere because of the frequently associated cirrhosis (3-6). With new diagnostic proce- (3.5 MHz, model SSD 256, Aloka Co., Ltd., Tokyo, and models dures, ultrasonography (US) in particular, small HCCs SAL 30A, 50A, Toshiba Systems Co., Ltd., Tokyo) and convex-arrayreal-time scanners (3.5 MHz, models EUB 40 and less than 2 cm in diameter can now be detected. EUB 340, Hitachi Medical Corp., Tokyo; models SSD 280,650, Several prospective studies of the incidence of HCC in Aloka Co. Ltd.; and model SSA 90A, Toshiba Systems Co.). patients with cirrhosis have been performed (7-12). Selective hepatic arteriography and computed tomography Most of the studies used only the a-fetoprotein WFP) (CT) were performed if HCC was suspected. A US-guided level to screen for HCC, or else both AF’P assay and biopsy was done for confirmation of the diagnosis of HCC when linear-array real-time US were used. The incidence of indicated. AFP levels were measured by RIA with a commercial kit HCC in patients positive for HBsAgis high (74,131.The incidence of HCC in patients with a history of blood (Dainabot Radioisotope Laboratory Ltd., Tokyo). Statistical transfusion who are negative for HBsAg is not known. analysis was done by the Kaplan-Meier method and by life table analysis. Received September 13, 1989; accepted May 2, 1990. Address reprint requests to: Dr.Hiroko Oh,Third Department of Internal Medicine, Osaka City University Medical School, 1-5-7 Asahi-machi, Abeno-ku, Osaka, 545, Japan. 31/1/23247

RESULTS

HCC was found in 40 patients during this 6-yrstudy: seven patients were diagnosed in the first year, nine in the second year, eight in the third year, eight in the

680

Vol. 12, No. 4, 1990

EARLY DETECTION OF HEPATOMA IN PATIENTS WITH CIRRHOSIS

68 1

50

0

1

2

3

4

5

6 Year

FIG.1 . Cumulative incidence of HCC in 140 patients with cirrhosis

fourth year, five in the fifth year and three in the sixth year. The incidence of HCC per year was 5.3%,8.1%, 8.896,11%, 8.5% and 6%, respectively. HCC was confirmed pathologically in 15 patients and clinically by an increase in AFP to diagnostic levels in two patients during the follow-up, by results of angiography in 18 patients and by increase in tumor size (detected by US) in five patients. Another 29 patients died because of liver failure or bleeding from the gastrointestinal tract. We lost contact with 26 patients. Table 1 summarizes the clinical features of the 40 patients who were found to have HCC. The size of the tumor detected was 2 cm in diameter or less in 25 of the patients (62%).Five patients had HCC of the diffuse type, which is difficult to detect by US. US was the main diagnostic clue in 34 patients (85%).The AFP titer was less than 100 ng/ml in 23 (58%)of the 40 patients. An abnormally high AFP titer (above 400 ng/ml) was found in only 10 patients. In the 25 patients with carcinoma measuring 2 cm or smaller, 17 (68%)had a low titer of less than 100 ng/ml. In some patients (cases 10, 15 and 16), examinations were not done on schedule and the carcinoma was detected after it had grown fairly large. Table 2 shows the cumulative incidence of HCC with use of a life table. The cumulative incidence for the 6 yr was 39%. Figure 1 shows the cumulative incidence according to the background of the patients. The incidence of HCC in the patients who were positive for HBsAg (59%) was higher than in patients who were negative for HBsAg (36%). In particular, all three patients who had HBsAg and a history of heavy drinking were found to have HCC within 4 yr of the start of this study. The cumulative incidence of HCC in the patients who had had blood transfusions and were negative for HBsAg was 53%. Figure 2A shows the overall survival rate of the patients with HCC. The l-yr survival rate was 72’26, at 2 yr, it was 53%and at 3 yr it was 41%.In terms of Child’s criteria, the survival rate of patients in class C was very low: 56%at 1 yr, 11%at 2 yr and 11%at 3 yr (Fig. 2B). C h e Report. A small hyperechoic lesion had first been detected in Patient 27 in 1982 (Fig. 3A). The US

diagnosis at that time was hemangioma. The size of the mass increased year by year (Fig. 3B and 3C).In 1984, angiography was done, but a tumor stain was not seen. Liver biopsy was done in 1985, and the histological diagnosis was of focal fatty change of the liver (Fig. 4A). In 1986, the size of the mass further increased, and the echo pattern of the mass had changed (Fig. 3D),so angiography was repeated. This time, a clear tumor stain was detected and surgery was done. The tumor was well-differentiated HCC (Fig. 4B). The doubling time of this tumor was calculated to be 313 days. Reexamined retrospectively, the histological diagnosis of the specimen obtained by the first liver biopsy was of welldifferentiated HCC. DISCUSSION

Cirrhosis is a high-risk factor for HCC; 60%to 86%of HCC coexists with cirrhosis (13-20). In the several studies of patients with cirrhosis, only AFP was used to screen for HCC in most subjects (8,9),so the size of HCC at detection was relatively large. Sheu et al. (11) used both US and AFP to monitor 236 patients with cirrhosis for 1.4 yr and found seven patients to have HCC. The size of the tumor when detected was 1.7 to 4.7 cm. Cottone et al. (12) monitored 147 patients with cirrhosis by screening with AFP and US every 6 mo and detected five patients with HCC during the 2-yr study. The diameter of the tumor was 5 cm or less in 80%of these patients. Kobayashi et al. (21) studied 95 patients with cirrhosis by infusion hepatic arteriography and CT yearly, performing US every 3 mo and AFP assays every 2 mo. They found eight patients (8%)to have HCC. The follow-up period for the patients ranged from 1to 8 yr, with a mean of 3.5 yr in the eight patients with HCC. The diameter of the tumor was 2 cm or less in 25%of the patients. In this study, the incidence of HCC was higher and the size of the tumors was smaller than in previous reports. The racial difference in the subjects may be one factor, but probably the main reason for the high incidence of HCC found in this study was the h g h sensitivity of our screening method, which uses AFP and a high-resolution convex-array US scanner. If we had

OKA ET AL.

682

HEPATOLOGY

TABLE1. Clinical features of the 40 patients found to have HCC during the 6-yrstudy Test

Patient numbem

ABe

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

63 55 55 60 55 61 62 55 59 56 61 58 68 56 62 45 69 80 64 52 59 74 61 52 75 64 62 44 58 54 59 61 56 53 58 59 53 52 61 57

Tumor Size (ern)

AEP Sex

BSAg

Alcohol"

F

M M M M F

M F M M F

M F

M M F

F M M M M F

M M M F F

M M M M M M F

F M M M M M

BP

(Wml)'

CT = computed tomography; Angio = angiography; TAE = transcatheter arterial embolization; PEIT "History of heavy drinking. 'History of blood transfusion. 'Level of AFP on detection of HCC. dAtumor stain waa detected in the second angiography. 'Not done. This patient declined all examinations and treatment. 'Diagnosed by increase in tumor size and/or presence of tumor thrombus in portal vein. #Single injection of anticancer drugs.

used only the criterion of a high titer of AFP of more than 400 ng/ml in screening, 75% of the cases of HCC would have been detected later or not at all. If AFP of more than 100 ng/ml had been used, 65% of the HCCs 2 cm or less would have been missed. Takashima et al. (22)reported in 1982that the sensitivity of US is inferior to angiography and CT; however, when we began to use a convex-array scanner in 1985,we could detect HCC smaller than 1.5 cm in diameter more easily than with

4.3 2.5 1.5 1.8 1.7 2.3 Diffuse 2.0 3.9

578.0 21.2 115.0 49.1 52.4 1,004.9 15.7 494.0 45.4 6,857.9 13.2 34.4 311.7 90.0 2,460.0 1,510.0 432.0 81.6 70.5 14.6 6.6 152.5 1,677.8 64.6 13.3 7.5 < 5.0 137.0 104.8 370.8 20.1 883.4 29.0 < 5.0 17.1 178.7 23.9 24.0 1,430.0 < 5.0

NF 1.2 2.0 1.7 2.0 4.4 6.0 1.5 1.0 3.4 2.0 Diffuse 1.8 Diffuse Diffuse 1.8 1.0 1.0 Diffuse 1.5 1.0 1.5 2.1 0.5 1.3 1.o 1.2 1.3 1.0 4.7 1.4 =

causing

suspicion of HCC

AFP

us CT us us AFP us us us AFP us us us us AFP us us us us us us us us us us us us us us us us AFP us us us us us us AFP us

percutaneous ethanol injection.

earlier methods. The sensitivity of CT and angiography was lower than that of US for HCC less than 2 cm in diameter in this study. In case 39, the tumor was large when detected (47 mm). However, the doubling time of the tumor, which is parallel to that for AFP (23-251, was calculated to be 15.6 days. This very rapidly growing tumor was probably 3 mm across or less when a US examination had been done 4 mo earlier. The patient reported no symptoms at

Vol. 12, No. 4, 1990

683

EARLY DETECTION OF HEPATOMA IN PATIENTS WITH CIRRHOSIS

Detection by imaging Diegnostic method AngiO

Histology AngiO AngiO AngiO

CT

hgi0

+ +

+

+ -d

AngiO

+ + + +

AFP

ND

Histology AngiO AngiO

Histology hgi0 Angio AngiO Angio AngiO

Histology Histology Histology Histology US + CT' AngiO

AFP UY Histology

+ +

+ + +

+ + -

+ + ND

+

ND ND

+

AngiO

-d

Histology

-

AngiO AngiO AngiO

+

Histology AngiO

Histology Histology Histology UY UY Histology Histology

us

+ +

+ + -

+ -

ND -

+ ND

the time of detection, so if periodical screening had not been done, the detection of the tumor would have been further delayed and the operation could not have been succ4?ssful. There is also a type of HCC with very slow growth (261, such as in case 27.In this case, we took 4 yr to confirm the diagnosis of HCC. Results in this case suggestedthat not only periodical screening every 3 mo by US and AFP but also long-term monitoring is essential for patients in this hqh-risk group. Recent changes in the morphological criteria of well-differentiated HCC (27,28) will shorten the time before a definite diagnosis can be made for patients who develop this disease, but it is very diacult to distinguish HCC from adenomatous regenerative nodules. Close clinical observation will still be useful for diagnosis.

Treatment

TAE Resection TAE

TAE TAE Resection TAE TAE Chemotherapy ND Resection ND TAE TAE TAE TAE Resection PEIT PEIT PEIT ND

TAE ND Chemotherapy

PEIT TAE Resection TAE TAE Bolusg Resection Bolus PEIT PEIT PEIT ND

ND PEIT Resection ND

Outcome

Time (mo)

Died Died Died Died Died survived Died Died Died Died Died Died Died Died Died Died survived Died Survived Died Died Died Died Survived Survived survived survived Died Died Died survived Died Died survived survived Died Died Survived survived Survived

39 62

21 25 27 64

15 16 22 11

40 10 32

10 5 10 54

36 44 14 7 38 1

37 34 51 81 2 22

7 33

3 41 24 25

19 4 5 13 36

A high incidence of HCC of 32% to 67% in patients with cirrhosis who are positive for HBsAg has been reported by many authors (7, 13-15, 17, 18, 291, and Hino, Kitagawa and Sugano (30)have reported that of 23 HCCs from carriers of HBsAg, 83% had hepatitis B viral DNA integrated into the tumor DNA. However, Zaman et al. (10)concluded by multifactor analysis in their prospective study of cirrhosis that HBsAg is not a risk factor in HCC. Okuda et al. (31)showed that a group of patients who were positive for anti-HBsAg, anti-HBcAg, or both, with low titers was more similar in terms of the occurrence of HCC to a group negative for all seromarkers for hepatitis B virus (HBV) than to a group positive for HBaAg, so we used only HBsAg as an HBV marker in this study.

OKA ET AL.

684

HEPATOLOGY

Survival rate (%)

m

A O'

loo0

1500

d

2Qlm

25M) (Days)

Survival rate

(%I

"B

500

1000

1500

i

2500

2@30

(Days)

FIG.2.(A)Overall survival rate of the patients with HCC. (B) Survival rate of patients with HCC in terms of Child's criteria.

TABLE 2. Cumulative incidence of HCC with use of a life table Incidence of HCC

lYr

lx = number without HCC dx = number with HCC wx = number of deaths or drop-outs lx' = corrected number without HCC" qx = incidence of HCC px = non-HCC rate Px = cumulative nonHCC rate

1-Px Cumulative incidence of HCC (%) 1 dx %' = lx - -WX; qx = -. px = 1 2 1'X'

-

2Yr

3yr

4Yr

SYr

6Yr

140 7 16

117 9 11

97 8 12

77 8 8

61 5 4

52 3 4

132

111.5

91

73

59

50

0.0530 0.9470 0.9470

0.0807 0.9193 0.8706

0.8790 0.9120 0.7940

0.1095 0.8904 0.7070

0.0847 0.9153 0.6471

0.0600 0.9400 0.6083

0.0530 5.3

0.1294 12.9

0.2059 20.6

0.2930 29.3

0.3529 35.3

0.3917 39.2

qx; Px = (Pxof last year) x px.

Vol. 12, No. 4,1990

EARLY DETECTION OF HEPATOMA IN PATIENTS WITH CIRRHOSIS

685

FIG.3. Ultrasonogram, case 27. (A) Small hyperechoic lesion had been found in case 27 in March 1982 and was believed to be a hemangioma at the time of entry to this study. (BI Size of the mass had increased by March 1983. (C)Small hypoechoic area had appeared in the maw in December 1984. A tumor stain was not seen by angiography at that time. (D) The size of the mass had further increased and the echo pattern had changed to a mixed pattern in July 1986.

Here, the cumulative incidence of HCC in the patients with HBsAg was high, and all three heavy drinkers positive for HBsAg were found to have HCC within 4 yr of the start of the study. On the other hand, the cumulative incidence of HCC in heavy drinkers without HBsAg was low. This is evidence of a synergistic relationship between HBV and alcohol intake (32,33).

Prospective studies of patients with cirrhosis and a history of blood transfusion have not been reported yet. This is the first prospective study dealing with the incidence of HCC in patients with cirrhosis who had a history of blood transfusion. The cumulative incidence of HCC in patients negative for HBsAg and with a history of blood transfusion was about as high as that for patients positive for HBsAg. This result strongly sug-

686

OKA ET AL.

HEPATOLOGY

..

A

’’

FIG.4. Micrographs, case 27. (A) Histological diagnosiswas of focal fatty change at the time of the first biopsy, but well-differentiated HCC with some fatty change is seen retrospectively in this specimen. The border between HCC and the noncancerous area (*) is clear but is without fibrous tissue (H & E, original magnification x 85). (B)In the resected specimen, well-differentiated HCC with some fatty change is growing with cornmession of the noncancerous area (*). A thin fibrous seDtum exists in the border between the two areas (H& E, original magnification x 85).

gested a relationship between the non-A, non-B virus

and the occurrence of HCC. In the near future, with the increasing use of HB vaccine, the number of carriers of HBV will decrease, and the proportion of patients with cirrhosis positive for HBsAg will also decrease. Consequently, the relative importance of non-A, non-B viral infection and later cirrhosis and HCC will increase. The survival rate of patients with HCC in this study was higher than in other reports (34-36).Ebara et al. (37)reported that even when the HCC was smaller than 3 cm at detection, the 3-yrsurvival of untreated patients was 13%. Cottone et al. (38) compared the natural history of 25 patients with HCC and cirrhosis of Child’s class A who were not surgicallytreated with the outcome of 12similar patients who underwent resection. The 2-yr survival rate was 50% in the untreated group and 39% in the treated group. This result suggested that early detection and consequent early treatment had no effect

on the outcome. In our study, the 3-yrsurvival rate of the patients was 41%’ and only 17 of 40 patients (43%) were in Child’s class A, so the long survival of our patients with HCC was owing not only to early detection but also to early treatment. Only selected patients with cancer and cirrhosis in Child’s class A or B undergo surgery: patients with decompensated cirrhosis can be treated by percutaneous ethanol injection therapy (39,401.In our series, the time of detection was used for calculations of the survival rate; otherwise, comparison of untreated and treated patients would not be possible. Our study population was not large enough for a thoroughgoing statistical evaluation of the effects of various kinds of medical treatment. Our findings showed that it was possible to detect HCC before it grew to 2 cm in diameter with use of a combination of US and AFP assay every 3 mo. The incidence of HCC in patients with cirrhosis negative for

Vol. 12,No. 4, 1990

EARLY DETECTION OF HEPATOMA IN PATIENTS WITH CIRRHOSIS

687

21. Kobayashi K, Sugimoto T, Makino H, Kumagai M, Unoura M, Tanaka M, Kato Y, et al. Screening methods for early detection of hepatocellular carcinoma. HEPATOLOGY 1985;5:1100-1105. 22. Takashima T, Mataui 0, Suzuki M, Ida M. Diagnosis and screening of small hepatocellular carcinoma. Radiology 1982;145: 635-638. REFERENCES 23. Shen JC, Sung LL, Chen DS, Yang PM, Lai MY, Lee CS, Hsu HC, 1. Liaw YF, Chu CM, Lin DY, Sheen IS, Yang CY, Huang MJ. et al. Growth rate of asymptomatic hepatocellular carcinoma and its clinical implications. Gastroenterology 1985;89:259-266. Age-specific prevalence and sigmfmnce of hepatitis B e antigen and antibody in chronic hepatitis B virus infection in Taiwan: a 24. Yoshino M. Growth kinetics of hepatocellular carcinoma. Jpn J Clin Oncol 1983;13:45-52. comparison among asymptomatic carriers, chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma. J Med Virol 1984;13:385- 25. Johnson PJ, Williams R. Serum alfa-fetoprotein estimation and 391. doubling time in hepatocellular carcinoma: influences of therapy 2. Steiner PE. Cancer of the liver and cirrhosis in trans-Saharan and possible value in early detection. J Natl Cancer Inst 1980;64: Africa and the United States of America. Cancer 1960;13:10851329-1332. 1166. 26. Yoshida T, Okazaki N, Yoshino M, Kitaoka H, Hirohashi S, 3. Luna G, Florence L, Johansen K. Hepatocellular carcinoma: a Shimozato Y. Minute hepatocellular carcinoma without appreciable change in size for seven years: a case report. Cancer 5-year institutional experience. Am J Surg 1985;149591-594. 4. Sandler DP, Sandler RS, Homey LF. Primary liver cancer 1982;49:1491-1495. 27. Kondo F,Wada K, Nagato Y, NakaJima T, Kondo Y, Hirooka N, mortality in the United States. J Chron Dis 1983;36:227-236. 5. B u d L, Sbolli G, Cavanna L, Givardi G, Stasi MD, Buscarini Ebara M, et al. Biopsy dmgnosis of well-differentiated hepatocelE, Fornali F. Clinical and diagnostic features of 67 cases of lular carcinoma based on new morphological criteria. HEPATOLOGY hepatocellular carcinoma. Oncology 1987;44:93-97. 19899751-755. 6. Gandolfi L, Solmi L, Bertoni F, Muratori R, Stasi G. Small 28. Kenmochi K, Sugihara S, Kojiro M. Relationship of histologic hepatocellular carcinoma: an Italian experience. Hepatogastroengrade of hepatocellular carcinoma (HCC) to tumor size, and terology 1987;34:100-102. demonstration of tumor cells of multiple different grades in single 7. Liaw YF, Tai DI, Chu CM, Lin DY, Sheen IS, Chen TJ, Pao CC. small HCC. Liver 1987;7:18-26. Early detection of hepatocellular carcinoma in patients with 29. Kobayashi K, Kumagai M, Kameda S, Sugimoto T, Suzuki K, chronic type B hepatitis: a prospective study. Gastroenterology Nishimura K, Kato Y, et d. Hepatoma development during 1986;90:263-267. long-term follow-up period of liver cirrhosis. A d a Hepatogastro8. Obata H, Hayashi N, Motoike Y, Hisamatsu T, Okuda H, enterol 1978;25:344-349. Kobayashi S, Nishioka K. A prospective study on the development 30. Hino 0,Kitagawa T, Sugano H. Relationship between serum and of hepatocellular carcinoma from liver cirrhosis with persistent histochemical markers for hepatitis B virus and rate of viral hepatitis B virus infection. Int J Cancer 1980;25:741-747. integration in hepatocellular carcinoma in Japan. Int J Cancer 9. Mueca A, Cordova C, Barnaba V, Bonavita MS, Levrero M, Zaccari 1985;35:5-10. C, Balsano F. W A n t i - H B e system and development of 31. Okuda H, Obata H, Motoike Y, Hisamitsu T. Clinicopathological features of hepatocellular carcinoma: comparison of hepatitis B primary hepatocellular carcinoma in patients with HBsAg-positive seropositive and seronegative patients. Hepatogastroenterology liver cirrhosis. Hepatogastroenterology 1983;30:3-5. 10. Zaman SN, Melia WM,Johnson RD, Portman BC, Johnson PJ, 1984;31:64-68. Williams R. Risk factors in development of hepatocellular car- 32. Ohnishi K, Iida S, Iwama S, Goto N, Nomura F, Takashi M, cinoma in cirrhosis: prospective study of 613 patients. Lancet Mishima A, et al. The effect of chronic habitual alcohol intake on 1985;1:1357-1359. the development of liver cirrhosis and hepatocellular carcinoma: 11. Sheu JC, Sung JL,Chen DS, Lai MY, Wang TH, Yu JY,Yang PM, relation to hepatitis B surface antigen carriage. Cancer 1982;49: 672-677. et al. Early detection of hepatocellular carcinoma by real-time ultrasonography: a prospective study. Cancer 1985;56:660-666. 33. Br6chot C, Nalpas B, Courouce AM, Duhamel G, Callard PC, 12. Cottone M, Turri M, Caltagirone M, Maringhini A, Sciarrino E, Carnot F, Tiollais P, et al. Evidence that hepatitis B virus has a role in liver-cell carcinoma in alcoholic liver disease.N Engl J Med Virdone R, Fusco G, et al. Early detection of hepatocellular 1982;306:1384-1387. carcinoma associated with cirrhosis by ultrasound and alfafetoprotein: a prospective study. Hepatogastroenterology 1988;35: 34. Okuda K,Obata H, N h j i m a Y, Ohtsuki T, Okazaki N, Ohnishi 101-103. K. Prognosis of primary hepatocellular carcinoma. HEPATOLOGY 1984;4:38-68. 13. Beasley Rp, Hwang LY, Lin CC, Chien CS. Hepatocellular carcinomaand hepatitis B virus: a prospechve study of 22,707men 35. Chlebowski RT, Tong M, Weissman J , Block JB,b a d KP, in Taiwan. Lancet 1981;2:1129-1133. Weiner JM, Bateman JR,et al. Hepatocellular carcinoma: diag14. Okuda K and Liver Cancer Study Group of Japan. Primary liver nostic and prognostic features in North American patients. Cancer 1984;53:2701-2706. cancer in Japan. Cancer 1980;45:2663-2669. 15. Liver Cancer Study Group of Japan. Primary liver cancer in 36. Okuda K, Ohtuki T, Obata H, Tomimatau M, Okazaki N, Japan. Cancer 1984;54:1747-1755. Hasejpwa H, Nakajima Y, et al. Natural history of hepatocellular 16. Nakashima T, Okuda K, Kojiro M, Jimi A, Yamaguchi R, carcinoma and prognosis in relation to treatment: study of 850 Sakamoto K, Ikari T. Pathology of hepatocellular carcinoma in patients. Cancer 1985;56:918-928. Japan: 232 consecutive cases autopsied in 10 years. Cancer 37. Ebara M, Ohto M, Shinagawa T, Sugiura N, Kimura K, Matsutani 1983;51:863-877. S, Morita M, et al. Natural history of minute hepatocellular 17. Shinagawa T, Ohto M, Kimura K, Tsunetomi S,Morita M, Saisho carcinoma smaller than three centimeters complicating cirrhosis: H, Tsuchiya Y, et al. Diagnosis and clinical features of small a study in 22 patients. Gaetmenterology 1986;90:289-298. hepatocellular carcinoma with emphasis on the utility of real-time 38. Cottone M, Virdone R, Fusco G, Orlando A, Turri M, Cdtagirone ultrasonography: a study in 51 patients. Gastroenterology 1984; M, Maringhini A, et al. Asymptomatic hepatocellular carcinoma in 86~495-502. Child’s A cirrhosis: a comparison of natural history and surgical treatment. Gastroenterology 1989961566-1571, 18. Kew MC, Geddes EW. Hepatocellular carcinoma in rural southern 39. Livraghi T, Festi D, Monti F, Salmi A, Vettori C. US-guided African blacks. Medicine 1982;61:98-108. 19. Nagame N, Yukaya H, Hamada T, Hirose S, Kanashima R, percutaneous alcohol injection of small hepatic and abdominal tumors. Radiology 1986;161:309-312. Inohchi K. The natural history of hepatocellular carcinoma: a 40. Ohto M, Ebara M, Watanabe Y,Sugiura N, Shinagawa T, Okuda study of 100 untreated cases.Cancer 1984;54:1461-1465. 20. Kubo Y, Okuda K, Musha H, Nakashima T. Detection of K. Percutaneous ethanol injection (PEI) therapy for smaU hepahepatocellular carcinoma during a clinical follow-up of chronic tocellular carcinoma: evaluation of its utility on the basis of liver diaeaaes: observations in 31 patients. Gastroenterology tumor-images and survival after therapy. Jpn J Med Imaging 1978;74:578-582. 1988;7:25-33.

HBsAg and with a history of blood transfusion was as high as in the patients positive for HBsAg, so they should be given careful long-term monitoring.