Close association between high serum alanine aminotransferase levels and multicentric hepatocarcinogenesis in patients with hepatitis C virus-associated cirrhosis

Authors


Abstract

BACKGROUND

Multicentric development of hepatocellular carcinoma (HCC) is a characteristic feature of hepatitis C virus (HCV)-associated cirrhosis (HCV-LC). In this study, the objective was to determine whether the persistent elevation of the serum alanine aminotransferase (ALT) level, which represents the inflammatory necrosis of hepatocytes, is correlated with the multicentric development of hepatocellular carcinoma (HCC) in patients with early-stage HCV-LC.

METHODS

Ninety-three consecutive patients with biopsy proven HCV-LC (Child Stage A) who had been followed for > 5 years for the development of HCC were studied. They were subdivided into three groups according to their serum ALT level: Group A included 33 patients with annual average serum ALT levels that were persistently high (≥ 80 IU; high ALT group), Group B included 41 patients with annual average serum ALT levels that were persistently low (< 80 IU; low ALT group), and Group C included 19 unclassified patients. The patients had been studied prospectively with frequent ultrasonography and magnetic resonance imaging or computed tomography (CT) scans for > 5years. When the development of HCC was suspected, angiography, infusion of lipiodol into the hepatic artery, and lipiodol-CT scans were performed in all patients to determine the number of HCC nodules.

RESULTS

In Group A, 27 patients (81.8%) developed HCC. Seventeen of 27 patients (63.0%) had multiple nodules. In contrast, in Group B, only 12 patients (29.3%) developed HCC, and only 1 of these 12 patients (8.3%) had multiple nodules. There was a significant difference between Groups A and B in the incidence of developing HCC (P < 0.001) and developing multiple nodules (P = 0.006). In addition, among the male patients, the incidence of developing multiple HCC nodules in Group A (12 of 19 patients; 63.2%) was significantly higher (P < 0.05) compared with the incidence in Group B (0 of 6 patients; 0%). The same tendency was observed among the female patients.

CONCLUSIONS

These results showed a close correlation between multicentric hepatocarcinogenesis and sustained necroinflammation of the liver in patients with HCV-LC. Cancer 2002;94:1787–95. © 2002 American Cancer Society.

DOI 10.1002/cncr.10391

Hepatocellular carcinoma (HCC) can develop multicentrically in hepatitis C virus (HCV)-associated cirrhosis (HCV-LC).1 Kubo et al.2 studied the risk factors for multicentric carcinogenesis of HCC in patients who underwent liver resection for HCC and found, by multivariate analysis, that the odds ratio for multicentric occurrence in patients who were infected with HCV was 4.30 compared with patients without either HCV or hepatitis B virus infection, and they concluded that HCV infection is a strong risk factor for multicentric carcinogenesis. However, the reason why multicentric hepatocarcinogenesis occurs in patients with HCV-LC remains unknown.

Recently, Ishikawa et al.3 used rat liver to study the correlation between the number of enzyme-altered (adenosine triphosphate-deficient) islands that were considered the immediate progeny of initiated cells and increased DNA synthesis. Those authors found that the number of enzyme-altered islands was inversely proportional to the time between the carcinogen treatment and subsequent partial hepatectomy, suggesting the importance of increased DNA synthesis in multicentric hepatocarcinogenesis. Thus, it has been suggested that increased DNA synthesis in hepatocytes (hepatocellular proliferation) is one of the major causes of multicentric hepatocarcinogenesis. In addition, we demonstrated clinically that the incidence of developing a second hepatoma after hepatectomy increased significantly when DNA synthesis of hepatocytes in the residual liver was increased in patients with HCV-LC and HCC who had neither hepatic vein invasion nor portal vein invasion of HCC4; nonsynchronous, multicentric hepatocarcinogenesis was promoted in patients with HCV-LC who had increased DNA synthesis of hepatocytes in the residual liver. Moreover, it was suggested that the increased DNA synthesis of hepatocytes in patients with HCV-LC is a result of sustained inflammatory necrosis of hepatocytes,4 and we demonstrated that the majority of patients with HCV-LC who had high DNA synthesis showed severe inflammatory reactions in liver biopsy specimens compared with the majority of patients who had low DNA synthesis activity, who showed slight inflammatory reactions.5 We also demonstrated previously that there is a close association between sustained high serum alanine aminotransferase (ALT) levels and the higher rate of incidence of HCC in patients with HCV-LC.6

Based on these findings, a strong correlation between the development of multicentric hepatocarcinogenesis and sustained liver cell necrosis is suggested. To demonstrate this hypothesis, annual biopsies of cirrhotic liver would provide accurate information on the status of sustained inflammatory necrosis; however, these biopsies are not allowed on ethical grounds. It is feasible that, if serum ALT levels remain high for many years, then necrosis of hepatocytes may be severe. In fact, Zelman and Wang7 demonstrated that elevations in serum aminotransferase levels are markers of liver cell necrosis. Moreover, Schmidt et al.8 demonstrated a close correlation between elevated transaminases and histologic necroinflammation in biopsied specimens from patients with chronic liver disease, including cirrhosis. Subsequently, Baier et al.9 studied the correlation between histologic necroinflammation in biopsied specimens and elevated levels of transaminases in patients with chronic liver diseases and found that, in patients with cirrhosis, the histologic finding of necroinflammation coincided with a significant elevation in transaminases in 53 of 64 patients (83%), although the tendency was not so prominent in patients with chronic hepatitis. Those authors also mentioned that, in patients with chronic liver disease who show subsided inflammation in liver biopsy specimens, almost normal transaminase levels were recognized in the majority of patients.

Therefore, in place of liver biopsies, we monitored the time course of serum ALT levels and assumed that, if the annual average serum ALT level was persistently high (≥ 80 IU), then the level of liver cell necrosis would be severe, and, conversely, if the annual average serum ALT level was persistently low (< 80 IU), then the level of liver cell necrosis would be slight. A cut-off level of 80 IU was adopted, because the annual average ALT level in patients with HCV-LC who had high DNA synthesis activity of hepatocytes, as estimated by bromodeoxyuridine (BrdU) uptake10 in vitro (BrdU labeling index ≥ 1.5%), was > 80 IU in all patients.11 Our previous study5 showed that these patients had a high risk of developing HCC. In addition, we examined the number of HCC nodules precisely when HCC nodules were found for the first time after patients were diagnosed with Child Stage A12 HCV-LC by using angiography and computed tomography (CT) scans after lipiodol infusion (lipiodol-CT)13 in patients with sustained high and low ALT levels.

MATERIALS AND METHODS

Study Population

One hundred three consecutive, posthepatic, cirrhotic patients with Child Stage A12 HCV-LC who were either HCV antibody positive (C-100 antibody or the second-generation tests) or HCV-RNA positive (polymerase chain reaction analysis) were studied. The study was confined to patients with Child Stage A HCV-LC, because long-term follow-up was required to observe the occurrence of HCC. Habitual alcohol drinkers were omitted from the study. We excluded patients who took > 40 g of ethanol daily and those who drank > 3 days per a week. Patients who had been exposed previously to hepatitis B virus and patients who were positive for hepatitis B surface antigen, hepatitis B core antibody, and hepatitis B surface antibody also were excluded from this study. Patients who had undergone prior hepatic resection for HCC also were excluded. All patients were hospitalized at Kanagawa Cancer Center Hospital, Tokai University Hospital, or Japanese Red Cross Center Hospital between July 17, 1985 and January 31, 1995. The diagnosis of cirrhosis was made by liver biopsy in all patients. This study was performed according to the Declaration of Helsinki; all patients provided informed consent, and the study protocol was approved by the Human Research Review Committee of Kanagawa Cancer Center Hospital, Tokai University, and Japan Red Cross Center Hospital.

Follow-Up Study

Upon admission, a liver biopsy was performed, and ultrasonography (US) and either CT scans or magnetic resonance imaging (MRI) were carried out to exclude any patients who already had been diagnosed with HCC. Also upon admission, plasma retention of indocyanine green at 15 minutes (normal, < 10%) was examined in most patients, and HCV-RNA levels (using reverse transcriptase-polymerase chain reaction analysis) also were evaluated in eligible patients. After the patient was discharged, for the purpose of detecting small HCC nodules, all patients were examined by US every 3 months and CT scan or MRI every 6 months.

The criteria for a suspicious HCC lesion on US, CT scan, or MRI were as follows: On US, an HCC nodule usually appears as a small, solitary, strongly hypoechoic mass lesion measuring 5–10 mm in greatest dimension, although it sometimes appears as a bright loop lesion. On CT scan, an HCC nodule appears as a low-density mass lesion of small size on an unenhanced image, which usually is enhanced on early phase after a bolus injection of contrast medium. On unenhanced MRI, an HCC nodule appears as a small mass lesion that is hypointense on a T1-weighted image and moderately hyperintense on a T2-weighted image.

Serum α-fetoprotein (AFP) levels (radioimmunoassay < 15.0 ng/mL) were measured; and biochemical tests, including serum albumin (3.8–5.3 g/dL), serum ALT (5–40 IU), serum aspartate aminotransferase (10–30 IU), the thymol turbidity test (0.0–5.0 Kilo IU [KIU]), and the zinc turbidity test (4.0–12.0 KIU), and peripheral blood counts were performed every month. The patients were followed carefully during the long-term observation period, and a careful endoscopic examination of esophageal varices was performed every 6 months in all patients. If esophageal varices with signs of redness were found, then either sclerotherapy or ligation was carried out to avoid esophageal bleeding.

If the US, CT, MRI, or AFP studies suggested the development of HCC, then further imaging examinations, including helical dynamic CT scans, lipiodol-CT,13 angiography, and biopsy of the tumor, were carried out to confirm the diagnosis of HCC. In particular, angiography and lipiodol infusion into the hepatic artery were performed in all patients with suspected development of HCC, and the numbers of HCC nodules were confirmed mainly by angiography and lipiodol-CT (the final decision was based on lipiodol-CT). For the lipiodol-CT technique, a conventional hepatic angiogram was performed first; then, 3–5 mL of lipiodol ultrafluid were infused selectively into the common hepatic artery or separately into both the right and left hepatic arteries in patients with aberrant vessels. CT scanning of the upper abdomen was performed 10–14 days after the injection of the lipiodol with contiguous slices through the liver at 10-mm intervals.10 We identified the deposit of lipiodol in dense, homogenous uptake or dense, patchy uptake as true HCC nodules. The greatest dimensions of the HCC tumors at the first examination, as determined with the various imaging modalities, were estimated in all patients who developed HCC, with the exception of one patient who developed the diffuse type of HCC. All patients were followed for > 5 years. Four male patients and two female patients who died of liver failure within 5 years from the beginning of the study were excluded, and four patients were excluded who developed HCC within the first year after the initial biopsy, because it is likely that they had minute HCC nodules when the initial biopsy had been performed. A total of 93 patients were examined for > 5 years.

Laboratory Analyses

The annual average of serum ALT levels, starting from the time of liver biopsy, were calculated each year for all patients. In calculating the annual average ALT levels, 12 measurements were usually included.

For serum ALT, if the annual average ALT level was < 80 IU, then an open circle was marked; if the annual average ALT level was ≥ 80 IU, then a solid circle was marked. We divided the 93 patients into 3 groups: patients with predominantly solid circles (n = 33 patients; the high ALT group), patients with predominantly open circles (n = 41 patients; the low ALT group), and patients with mixed solid and open circles (n = 19 patients; the unclassified ALT group). The characteristics of patients who developed HCC in sustained high, sustained low, and unclassified ALT groups are shown in Table 1.

Table 1. Characteristics of Patients in Sustained High, Sustained Low, and Unclassified Alanine Aminotransferase Groups at the Beginning of the Study
CharacteristicALT groupP value
Sustained high (A)Sustained low (B)Unclassified (C)
  • SD: standard deviation; M: male; F: female; ICG R15: indocyanine green retention rate at 15 minutes; AST: aspartate aminotransferase; ALT: alanine aminotoransferase; ZTT: zinc turbidity test; TTT: thymol turbidity test; HCV: hepatitis C virus; KIU: Kilo international unit.

  • a

    Student t test.

  • b

    Pearson chi-square test.

No. of patients334119
Age in yrs (mean ± SD)60.2 ± 5.560.6 ± 8.762.5 ± 6.0NSa
Gender (M/F)21/1216/2510/9NSb
Child classification (A/B)33/041/019/0NSb
Liver function (mean ± SD)
 Albumin (g/dL)3.8 ± 0.43.8 ± 0.43.7 ± 0.4NSa
 Total bilirubin (mg/dL)0.8 ± 0.21.1 ± 0.71.0 ± 0.5NSa
 Direct bilirubin (mg/dL)0.2 ± 0.10.3 ± 0.30.3 ± 0.2NSa
 ICG R15 (%)22.8 ± 10.123.0 ± 12.724.0 ± 13.1NSa
 AST (IU)123.2 ± 67.188.9 ± 51.0112.5 ± 58.80.021 (A:B)a
 ALT (IU)158.3 ± 80.287.5 ± 60.5129.4 ± 76.8< 0.0001 (A:B)a
 ZTT (Kunkel unit)18.4 ± 2.918.1 ± 5.118.6 ± 7.6NSa
 TTT (Kunkel unit)8.6 ± 3.79.4 ± 5.111.6 ± 6.7NSa
 Prothrombin time (seconds)13.1 ± 1.213.2 ± 1.213.2 ± 0.9NSa
 Platelet counts/mm3 (× 104)10.4 ± 3.411.2 ± 3.510.8 ± 3.8NSa
α fetoproteinn (ng/mL)72.5 ± 89.726.3 ± 32.337.4 ± 30.90.009 (A:B)a
HCV-RNA level (KIU/mL)317.6 ± 297.0301.6 ± 250.9263.5 ± 343.6NSa
Greatest tumor dimension (mm)20 ± 718 ± 618 ± 8NSa

Statistical Analysis

The Student t test (Welch method), the chi-square test (Fisher exact probability), and the Kaplan–Meier method (log-rank test) were used for statistical analyses.

RESULTS

Of the 33 patients with HCV-LC who had sustained high ALT levels, 27 patients (81.8%) developed HCC during an average observation period of 8.7 ± 3.2 years (mean ± standard deviation [SD]). Of the 27 patients who developed HCC, it was found that 17 patients (63.0%) had multiple HCC nodules at the time they first developed HCC (Table 2).

Table 2. Multiplicity of Nodules of Hepatocellular Carcinoma in Patients with Hepatitis C Virus-Associated Cirrhosis with Sustained High Alanine Aminotransferase Levels
PatientGenderAge (yrs)ALT groupChild's classificationYrs after biopsyaHCC tumor
At beginningAt HCC developmentAt 5 yrs after12345678910111213MultiplicityLargest size (cm)Location (segment)
  • HCC: hepatocellular carcinoma; ALT: alanine aminotransferase; M: male; F: female; H: high ALT group; m: multiple nodules; s: single nodule.

  • a

    Solid circles represent years in which the annual average serum ALT level was ≥ 80 IU, and open circles represent years in which the annual average ALT level was < 80 IU. Stars represent the year in which HCC nodules were detected. A dagger represents a patient who died of liver failure.

1M61HAAAm2.0S5,S8
2M63HAABm1.5S6,S7,S8
3M65HAAm2.0S6,S8
4M74HAAm3.0S5,S8
5M57HABmDiffuseDiffuse
6M54HAAAm1.8S6,S7,S8
7M59HABAm1.5S6,S8
8M57HAABm2.0S3,S8
9M69HAAs3.0S8
10M56HAAAs1.0S8
11M63HABCm0.9S4,S8
12M66HAAAs1.7S4
13M60HAAAs3.3S4
14M51HAAAs2.4S4
15M58HAAAs0.5S6
16M60HAAAm2.0S6,S8
17M66HAAs1.5S8
18M52HABAm2.0S2,S6
19M58HAAAm2.2S1,S5,S8
20F59HAAAm2.2S4,S6,S7,S8
21F73HAAAs1.5S5
22F62HAAAs1.5S5
23F58HAAm2.0S4,S7
24F66HABBm1.3S2,S5
25F62HACBm3.0S3,S5
26F53HABBs2.0S8
27F63HAAAm2.9S2,S8
28M56HAA
29M50HAA
30M60HAA
31F57HAA
32F60HAA
33F65HAA

In contrast, of the 41 patients with HCV-LC who had sustained low ALT levels, 12 patients (29.3%) developed HCC during an average observation period of 9.4 ± 2.7 years. Of the 12 patients who developed HCC, only 1 patient (8.3%) showed multiple nodules, and the remaining 11 patients (91.7%) had only one HCC nodule (Table 3). Among patients in the sustained high ALT group, HCC developed as much as 2.8-fold more frequently compared with patients in the sustained low ALT group. There was a significant difference between patients in the sustained high and sustained low ALT groups in the incidence of developing HCC (P < 0.001). Moreover, there was a significant difference between patients in the sustained high and sustained low ALT groups in the incidence of developing multiple HCC nodules (P = 0.006) (Fig. 1). In the male patients, the incidence of developing multiple HCC nodules in the high ALT group (12 of 19 patients; 63.2%) was significantly greater (P < 0.05) among patients in the low ALT group (0 of 6 patients; 0%). The same tendency also was observed in the female patients, although the difference was not significant. Although the proportion of males was greater than the proportion of females in the high ALT group, and the reverse was true in the low ALT group, the rate of developing multiple HCC nodules in the high ALT group was very high for patients of both genders: Twelve of 19 male patients (63.2%) and 5 of 8 female patients (62.5%) developed multiple HCC nodules. Furthermore, the interval between the diagnosis of cirrhosis (Child Stage A) and the development of HCC among the patients who developed HCC was 4.8 ± 0.6 years (mean ± standard error) in the high ALT group and 7.6 ± 1.0 years in the low ALT group (P < 0.05).

Table 3. Multiplicity of Nodules of Hepatocellular Carcinoma in Patients with Hepatitis C Virus-Associated Cirrhosis with Sustained Low Alanine Aminotransferase Levels
PatientGenderAge (yrs)ALT groupChild classificationYrs after biopsyaHCC tumor
At beginningAt HCC developmentAt 5 yrs after12345678910111213MultiplicityLargest size (cm)Location (segment)
  • HCC: hepatocellular carcinoma; ALT: alanine aminotransferase; M: male; F: female; L: low ALT group; s: single nodule; m: multiple nodules.

  • a

    Solid circles represent years in which the annual average serum ALT level was ≥ 80 IU, and open circles represent years in which the annual average ALT level was < 80 IU. Solid stars represent the year in which HCC nodules were detected. Open stars represent patients who developed adenomatous hyperplasia. Daggers represent patients who died of liver failure.

1M61LAAs3.0S7
2M72LAAAs1.3S5
3M62LABAs1.2S8
4M58LAAAs1.9S3
5M53LAAAs2.8S8
6M61LAAAs1.3S8
7F75LAAAs1.5S6
8F69LAAAs1.2S4
9F65LAAAs1.8S3
10F65LAAAm1.2S2,S3,S4
11F70LAAAs1.9S6
12F61LABAs2.0S8
13M61LAA
14M65LAB○ ○☆
15M59LAA
16M68LAA
17M68LAA
18M57LAA
19M43LAA
20M56LAC
21M66LAA
22M56LAA
23F65LAA
24F60LAA
25F66LAA
26F62LAB
27F60LAA
28F54LAA
29F59LAC
30F64LAA
31F63LAA
32F53LAA
33F71LAA
34F59LAA
35F52LAA
36F55LAA
37F66LAA
38F66LAA
39F43LAB
40F55LAA
41F41LAA
Figure 1.

This chart indicates the incidence of developing multiple hepatocellular carcinoma nodules among three groups of patients with different serum alanine aminotransferase (ALT) levels.

Of the 19 patients with unclassified ALT levels, 8 patients (42.1%) developed HCC during an average observation period of 8.9 ± 2.3 years. It was found that four of the eight patients who developed HCC (50.0%) had multiple HCC nodules at the time they first developed HCC (Table 4). The slightly shorter follow-up for the sustained high ALT group and the unclassified group, compared with the sustained low ALT group, was due to the short survival of patients who developed HCC in both groups. Comparing the baseline and follow-up Child scores between the sustained high ALT group and the sustained low ALT group revealed no significant difference (Table 5).

Table 4. Multiplicity of Nodules of Hepatocellular Carcinoma in Patients with Hepatitis C Virus-Associated Cirrhosis with Unclassified Alanine Aminotransferase Levels
PatientGenderAge (yrs)ALT groupChild classificationYrs after biopsyaHCC tumor
At beginningAt HCC developmentAt 5 yrs after12345678910111213MultiplicityLargest size (cm)Location (segment)
  • HCC: hepatocellular carcinoma; ALT: alanine aminotransferase; M: male; F: female; U: unclassified ALT group; m: multiple nodules; s: single nodules.

  • a

    Solid circles represent years in which the annual average serum ALT level was ≥ 80 IU, and open circles represent years in which the annual average ALT level was < 80 IU. Stars represent the year in which HCC nodules were detected. A dagger represents a patient who died of liver failure.

1M64UAAAm2.0S4
2M67UAAAm1.1S7
3M69UAAAs1.0S8
4M62UAAAs3.0S5
5F60UABm1.0S8
6F54UAAAs2.5S6–S7
7F64UAAAs1.0S8
8F79UABAm2.5S2–S3,S7
9M63UAA
10M56UAA
11M56UAA
12M65UAA
13M62UAA
14M63UAA
15F55UAA
16F56UAC
17F66UAA
18F62UAA
19F63UAA
Table 5. Changes in Child Score in the Course of Follow-up of Patients in Sustained High, Sustained Low, and Unclassified Alanine Aminotransferase Groups
ALT groupProportion of patients' number in Child's score A/B/C
At beginningAt HCC occurenceAt 5 yrs after LC diagnosis
  1. ALT: alanine aminotransferase; HCC: hepatocellular carcinoma; LC: cirrhosis.

Sustained high33/0/020/6/121/5/1
Sustained low41/0/010/2/035/3/2
Unclassified19/0/06/2/017/0/1

There was no significant difference between the sustained high ALT group and the sustained low ALT group in the HCV-RNA level (317.6 ± 297.0 vs. 301.6 ± 250.9 KIU/mL, respectively). The mean greatest dimension of the largest HCC nodules at the time patients first developed HCC in the sustained high ALT group (20 ± 7 mm; mean ± SD) was nearly equal to that measured in the sustained low ALT group (18 ± 6 mm) (Table 1).

DISCUSSION

We previously demonstrated that the proliferation of hepatocytes was important in the development of HCC in patients with anti-HCV-LC and that patients with HCV-LC who had high DNA synthesis activity (BrdU labeling index ≥ 1.5%) developed HCC significantly more frequently than patients who had low DNA synthesis activity (BrdU labeling index < 1.5%).5 Recently, we demonstrated that patients with HCV-LC who had persistently high serum ALT levels (annual average ALT ≥ 80 IU) developed HCC significantly more frequently than patients who had persistently low serum ALT levels (annual average ALT < 80 IU).6 Concerning this phenomenon, it would be expected that patients with elevated ALT levels may have a faster progression to end stage liver disease and that more advanced cirrhosis is more likely to correlate with carcinogenesis. In this study, we also determined the Child stage of cirrhosis at the time when HCC was developing and found that it remained in Child Stage A when it was developing in the majority of patients in the high ALT group (Table 5). However, there is a possibility that viral activity itself actually may become the important link to increased incidence of carcinogenesis,14, 15 but there was no need for considering this possibility in the current study, because no significant difference was found between the sustained high ALT group and the sustained low ALT group with regard to the HCV-RNA level.

Based on these findings, a persistent necroinflammatory process and subsequent proliferation of hepatocytes (increased DNA synthesis) appeared to be very important in the development of HCC in patients with HCV-LC. Because the necroinflammatory process may include the entire liver in patients with HCV-LC, it is feasible that an increased necroinflammatory process and subsequent increased proliferation of hepatocytes may be an important cause of multicentric hepatocarcinogenesis in patients with HCV-LC.

The current findings clearly demonstrate that multicentric carcinogenesis occurs significantly more frequently in patients with HCV-LC who have persistently high ALT levels compared with patients who have persistently low ALT levels in whom HCC nodules appear singularly in most cases. These findings indicate the need for more intensive surveillance by US, CT scanning, and MRI for the multicentric development of HCC in patients with HCV-LC who have persistently high serum ALT levels.

Kubo et al.16 investigated the correlation between the multicentric occurrence of HCC and the histology of nonmalignant hepatic tissue in 252 patients who were infected with HCV and underwent surgery for HCC. Those authors found that high grading score and high staging core were independent risk factors for multicentric carcinogenesis. Their findings indicate that active hepatitis and extensive fibrosis are responsible for the development of multicentric HCC. The current findings are in accordance with these previous observations.

Of course, there is a possibility that some HCC nodules that were found in the current study may have been metastatic lesions from a primary HCC nodule. In this respect, however, Okuda et al.17 studied 31 patients histopathologically who had recurrent tumors measuring < 20 mm in greatest dimension that developed after the patients underwent hepatectomy. Those authors found that 16 patients (51.6%) had newly developed HCC, whereas, in the other 14 patients (48.4%), it was believed that the recurrence was metastatic. Kajiwara18 conducted a histologic study of biopsy specimens from patients with small HCC tumors measuring < 20 mm in greatest dimension and found that approximately 50% of patients had differentiated HCC tumors, and the other 50% of patients had moderately to poorly differentiated HCC nodules. In addition, Kenmochi et al.19 studied the correlation between the histologic tumor grade and tumor size in patients with HCC and found that 9 of 11 tumors measuring < 30 mm in greatest dimension were Edmondson–Steiner Grade 1 carcinoma, which has less ability to metastasize. Kim et al.20 reported similar findings in patients with small HCC tumors measuring < 15 mm in greatest dimension. Because the mean tumor measurements in our patients were ≤ 20 mm in greatest dimension for all three groups, at least half of the HCC nodules found in the current study suggested differentiated HCC, which has less ability to metastasize, even in the sustained high ALT group. Thus, at least half of the patients with multiple nodules had nodules of suggested multicentric origin rather than metastatic origin, and the difference remains in the incidence of multicentric carcinogenesis between patients in the sustained high and sustained low ALT groups.

Finally, the adequacy of using lipiodol-CT for detecting small HCC nodules must be discussed. The procedure of CT scanning after the intra-arterial injection of lipiodol ultrafluid into the hepatic artery (lipiodol-CT) has been reported by Yumoto et al.,13 Ohishi et al.,21 Choi et al.,22 and Ngan23 as a useful method of detecting small HCC nodules.

Ngan23 demonstrated that CT scanning after intra-arterial injection of iodized poppy seed oil (lipiodol-CT) had an overall sensitivity of 97.1% and an accuracy of 88.3% in the diagnosis of HCC and that lesions with dense, homogenous uptake or dense, patchy uptake were true HCC nodules in very high percentages; the author concluded that lipiodol-CT was useful for the early detection of HCC. Choi et al.22 also demonstrated that sensitivity for detecting small HCC nodules was 96% with lipiodol-CT and that it was superior to sonography (P < 0.01), CT scanning (P < 0.01), and angiography (P < 0.05) for detecting small HCC nodules. Therefore, we used lipiodol-CT to finally detect small HCC nodules.

Taking all of these observations into consideration, the current findings demonstrated a close correlation between multicentric hepatocarcinogenesis and sustained necroinflammation of the liver in patients with HCV-LC after long-term observation of about 9 years. Furthermore, the current study confirmed the importance of sustained inflammation for the development of HCC and for multicentric carcinogenesis in HCV-LC. Patients with HCV-LC who have sustained elevated ALT levels should be screened more aggressively for HCC if current modalities of therapy are under consideration.

Ancillary