An Intensive Surveillance Program Detected a High Incidence of Hepatocellular Carcinoma Among Hepatitis B Virus Carriers With Abnormal Alpha
http://www.100md.com
《临床肿瘤学》
the Departments of Clinical Oncology, Diagnostic Radiology and Organ Imaging, Surgery, and Medicine and Therapeutics, Prince of Wales Hospital, and Centre for Clinical Trials, School of Public Health, Chinese University of Hong Kong, Hong Kong Special Administrative Region, China
Department of Oncology, University of Birmingham, Birmingham, United Kingdom
ABSTRACT
PURPOSE: To study the incidence and treatment outcomes of hepatocellular carcinoma (HCC) detected in an intensive surveillance program (ISP) of hepatitis B virus (HBV) carriers.
PATIENTS AND METHODS: We screened 1,018 HBV carriers by serum alpha-fetoprotein (AFP) measurement and abdominal ultrasonography (AUS). Patients with an abnormal AFP level or AUS result were enrolled in an ISP that included Lipiodol computed tomography followed by AFP measurement/AUS every 3 months for 2 years and then every 6 months thereafter. The rest were on routine surveillance for 2 years.
RESULTS: A total of 9,849 serum AFP measurements and 3,053 AUSs were performed. After a median follow-up of 4.12 years, we diagnosed 24 HCCs among 78 patients with abnormal screening test results at enrollment (group A); 23 HCCs among 93 patients with only abnormal surveillance test results during follow-up (group B); and nine HCCs among 847 patients with 2 years of normal surveillance (group C). Annual incidence of HCC in the ISP was 760.2 (95% CI, 538.4 to 1,073.7) per 100,000. Mean tumor sizes were 3.02, 2.91, and 4.82 cm in groups A, B, and C, respectively (P = .01). Tumor resection rate of the ISP was 36.2%, although another 29.8% of the patients were eligible for locoregional ablative therapy.
CONCLUSION: This study illustrated that a high incidence of relatively small HCCs may be detected by using intensive surveillance of high-risk HBV carriers. However, the surgical resection rate was low, and we were not able to demonstrate clinical benefit with the early detection. Future surveillance studies should consider incorporation of therapy aimed at long-term control of small-sized tumors.
INTRODUCTION
Hepatocellular carcinoma (HCC) is one of the most common malignancies in the world.1 Thousands of patients die as a result of the disease because of late presentation and lack of effective therapies for advanced disease, and the incidence of HCC in Western countries continues to rise.2,3 In Asia, the incidence of HCC is high at more than 30 cases per 100,000 population per year in some countries. Chronic hepatitis B virus (HBV) infection and cirrhosis are the main reasons for the high incidence of HCC in Asia. Beasley et al4 first established the etiologic link between HBV and HCC and subsequently confirmed that patients with positive hepatitis B surface antigen (HBsAg) are at a higher risk of developing HCC.5 The annual incidence of HCC in this group of patients ranges between 1% and 6%. Therefore, a surveillance program targeted at this population with chronic viral hepatitis may detect early resectable HCC and has the potential to reduce mortality.
Serum alpha-fetoprotein (AFP) measurement and abdominal ultrasonography (AUS) are widely accepted as screening tests for the early detection of HCC. It is common practice in the United States to routinely screen patients with cirrhosis for early HCC.6 Large-scale screening programs that target HBsAg-positive patients have defined the validity of these screening tests.7-9 Sensitivity and specificity of serum AFP measurements ranges from 64.3% to 96.9% and 93.8% to 95%, respectively. AUS has also been considered to be an effective surveillance tool with sensitivity and specificity ranging from 71.4% to 84.3% and 93.8% to 97%, respectively. Early detection of HCC by serum AFP and AUS is feasible, but no studies have demonstrated a definite improvement in survival or a reduction in the mortality rate. To date, the clinical benefits of screening programs for early detection of HCC remain controversial.
We hypothesized that an intensive surveillance program (ISP) that includes quarterly surveillance and the addition of a more sensitive imaging technique for a high-risk group of HBV carriers may detect a higher number of early HCCs for curative surgery and possibly result in an improvement in survival outcomes. In this study, Chinese HBV carriers with either an elevation of serum AFP or abnormal AUS results were enrolled into an ISP to determine incidence, tumor size, resectability, and survival outcome.
PATIENTS AND METHODS
Eligibility
Between October 1997 and November 2000, all patients attending the hepatology clinic at the Prince of Wales Hospital (the Chinese University of Hong Kong, Hong Kong Special Administrative Region, China) were evaluated for eligibility for the study. The hepatology clinic treats patients with liver disease from the general population of 1.4 million people residing in New Territories East in Hong Kong. The sources of referral include family doctors, general internists, and other subspecialists from the hospital catchment area. Reasons for referral ranged from mild hepatic dysfunction to advanced cirrhosis. Eligible patients were screened according to the following criteria: age between 40 and 70 years; seropositive for HBsAg; clinical grade A or B cirrhosis according to Child's classification; and life expectancy of more than 2 years. Patients with non–HBV-related cirrhosis or history of malignancy were excluded. The study was approved by the Clinical Research Ethics Committee of the Chinese University of Hong Kong, and written informed consent was obtained from all patients before entry into the trial.
Routine Investigation and Standard Surveillance
Evaluation at enrollment included a full medical history, physical examination, urinalysis, CBC, liver-function test, and assessment of serum electrolytes, creatinine, International Normalized Ratio (INR), HBsAg, hepatitis B e antigen (HBeAg), and antibody to HBeAg (anti-HBeAg). Initial screening serum AFP measurements and AUSs were performed after enrollment. The microparticles enzyme immunoassay method (Abbott Laboratories, Chicago, IL) was used for qualitative measurement of serum AFP. Serum AFP greater than 20 μg/L was considered to be abnormally elevated. AUS was performed by one of the two designated radiologists at the Department of Diagnostic Imaging, Prince of Wales Hospital, each of whom had more than 10 years of experience. The standard ultrasound system equipment was the HDI 3000 (Advance Technology Laboratories, Bothwell, WA). An electronic curvilinear 3.5-MHz real-time transducer was used to scan subcostally and intercostally with the patient in the supine position first and then in the left decubitus position. Positive AUS implied the presence of a focal solid nodule with hypoechoic, hyperechoic, or mixed sonographic pattern, and probable AUS implied the presence of an atypical lesion. Both positive and probable AUSs were considered to be abnormal.
Patients with either elevated serum AFP or abnormal AUS results at initial screening were admitted to the ISP. Patients with normal initial screening serum AFP and AUS were followed every 6 months for 2 years with repeated surveillance of serum AFP (Fig 1). AUS was repeated after 2 years for this group of patients, after which they had no additional routine surveillance in the program. The patients were referred back to the hepatology clinic at the Prince of Wales Hospital, at which serum AFP measurements or AUSs were performed by indication only.
ISP
Lipiodol computed tomography (LCT) was considered to be a sensitive imaging modality for diagnosis of small HCC when the study was first initiated.10 All patients enrolled onto the ISP were offered additional investigation by LCT. The hepatic angiogram involved access to the arterial system via the femoral artery using the Seldinger technique. Celiac axis, superior mesenteric artery, and selective hepatic artery angiograms were performed as contrast material was injected into the respective arteries. Angiographic images were examined for evidence of hypervascular areas in the liver that would suggest HCC. Lipiodol, an oily contrast medium that tends to be accumulated and retained selectively by tumor tissue, was selectively injected into the hepatic arteries during the procedure. A CT scan was performed 10 days after arterial Lipiodol administration to detect any Lipiodol retention. The presence of hypervascular lesion(s) with corresponding Lipiodol retention was an indication for histologic confirmation. Liver biopsy was offered to all patients with hypervascular lesions on LCT or the presence of space-occupying lesions on AUS. After exclusion of a diagnosis of HCC, AFP measurement and AUS were repeated every 3 months for 2 years and subsequently every 6 months. If a new abnormality was detected in either surveillance test, the patient was reinvestigated by LCT and/or biopsy. Liver biopsies were fixed in 10% formalin and stained with hematoxylin and eosin. A designated hepatic pathologist was responsible for evaluating all liver biopsies for this study.
Surgical and Nonsurgical Treatment for HCC
The hepatobiliary surgical team assessed all patients who had either a hypervascular lesion on LCT or a confirmed diagnosis of HCC. Tumor resectability was determined by tumor size, tumor location, proximity to major vessels, multifocal disease, and hepatic reserve. Surgery was offered to all patients who had a potentially resectable tumor. Surgical options included wedge resection, subsegmentectomy, and segmentectomy. Postoperative radioactive iodine-131 was offered to selected patients according to institutional protocol. Patients who had declined or were not eligible for surgery were evaluated by the oncology team for consideration of locoregional ablative therapy including alcohol injection, radio-frequency ablation, microwave ablation and selective internal radiation, systemic chemotherapy, or supportive care. All patients with confirmed HCC were followed up at the hepatoma clinic until death.
Statistical Analysis
Patients were classified into three groups to clearly delineate the effect of the surveillance program. Patients with elevated serum AFP and/or abnormal AUS results at initial screening represented those with pre-existing HCC or who were at high risk of HCC (group A). Patients with normal serum AFP and AUS at initial screening but subsequently had an abnormal surveillance test result were also considered to be at high risk for developing HCC (group B). Both groups A and B were enrolled to the ISP. Patients with normal serum AFP and AUS at initial screening and normal surveillance for 2 years were considered to be at low risk for developing HCC (group C). Comparisons among the three groups with respect to sex, age, bilirubin, alkaline phosphatase, albumin, INR, ascites, encephalopathy, cirrhosis, and Child's stage were made by using the 2 test. Survival was defined as the time between the date of diagnosis of HCC and the last report date. All comparisons of overall survival were done by using log-rank test; Kaplan-Meier estimates of patients with confirmed diagnosis of HCC are shown graphically to indicate the survival patterns between groups (Fig 2).
This prospective surveillance study was intended to target a population of high-risk HBV carriers for the ISP. Providing the best estimate that up to half of all HBV carriers with elevated serum AFP or abnormal AUS results would eventually develop HCC and operationally defining a 10% error rate as acceptably accurate, the minimal sample size for the ISP should be 96. As established from our previous study,11 approximately 9.1% of consecutive attendees at the hepatology clinic were found to have elevated serum AFP. We therefore calculated the sample size for the ISP to be 1,055.
RESULTS
Patients' Characteristics
Between October 1997 and November 2000, 1,018 (88.4%) of 1,152 eligible patients were enrolled onto the study. The male to female ratio was 688:330, and the mean age was 48.7 years (range, 40 to 69 years). The number of patients in groups A, B, and C was 78, 93, and 847, respectively. All patients were followed up from enrollment until death or last follow-up. The data set was frozen for analysis in April 2004. Groups A and B (171 patients) were enrolled onto the ISP. The total duration of follow-up for this group of patients within the program was 704.8 person-years. For group C, the total duration of follow-up was 3,504.7 person-years. The average duration of follow-up for the ISP group and group C was 4.12 and 4.13 years, respectively.
Patients' characteristics and liver functions at enrollment are listed in Table 1. Groups A and B had a significantly higher percentage of patients with elevated total bilirubin as per Child's classification (> 34 μmol/L; normal range, 5 to 21 μmol/L) (P < .001) and low albumin (< 35 g/L; normal range, 35 to 50 g/L) than group C (P < .001). There were also significant differences in the number of patients with an elevated INR (P < .001), but the relevance of the difference in history of encephalopathy is limited by the small number of patients. This finding corresponded to the high percentage of ultrasonographic evidence of cirrhosis in groups A and B. However, the overall number of patients with elevated total bilirubin, prolonged INR, ascites, or encephalopathy was relatively small, and there was no significant difference in the portion of patients classified as having Child's class B cirrhosis among the three groups (P = .06).
Surveillance Tests and HCC
A total of 9,849 serum AFP measurements (group A, 1,205; group B, 1,481; group C, 7,163) and 3,053 AUSs (group A, 841; group B, 979; group C, 1,233) were performed in this study. Group B had the highest number of surveillance tests per patient, with an average of 15.4 serum AFP measurements and 10.5 AUSs per patient. Comparatively, group A had an average of 15.4 serum AFP measurements and 10.5 AUSs per patient; and group C had an average of 8.5 serum AFP measurements and 1.46 AUSs per patient. An LCT scan was offered to all 171 patients enrolled onto the ISP, but 35 patients declined. In total, 110 patients had one LCT scan, 20 patients had two LCT scans, and six patients had three LCT scans. Patients with positive or suspicious findings underwent biopsy. Patients with negative findings continued with surveillance every 3 months, and LCT would be repeated on two consecutive elevations of serum AFP, notable enlargement of the lesion, or presence of new lesion(s) on AUS.
Fifty-seven (73.1%) of the 78 patients from group A had elevation of serum AFP above 20 μg/L, 15 (19.2%) had abnormal AUS results, and six (7.7%) had abnormalities in both tests at initial screening. Twenty-four HCCs were diagnosed in this group, 15 of which (62.5%) were diagnosed within 6 months from initial enrollment and were considered to be pre-existing HCC. The other nine patients had negative LCT at initial assessment, and HCC was diagnosed during the subsequent follow-up. Twelve (21.0%) of 57 patients with elevated serum AFP, seven (46.7%) of 15 patients with abnormal lesions on AUS, and five (83.3%) of six patients with both elevated serum AFP and abnormal AUS results were diagnosed with HCC.
Group B comprised 93 patients with normal initial screening tests but abnormal surveillance test results during the 2-year follow-up. Twenty-three HCCs were confirmed in this group. Sixty-nine patients (74.2%) in group B had elevated serum AFP, 23 patients (24.7%) had abnormal AUS results, and one patient (1.1%) had both abnormalities. The number of HCCs detected in groups A, B, and C was 12 (17.1%), 10 (41.7%), and one (100%), respectively. The median duration from enrollment to diagnosis of HCC was 2.47 years.
Eight hundred forty-seven patients from group C had normal serum AFP and AUS during the 2 years of standard surveillance, of whom 9 were confirmed to have HCC. The median duration from enrollment to diagnosis was 3.56 years, which was significantly longer than the patients with HCC in group B (P = .015).
The 15 patients with HCC diagnosed within 6 months of enrollment were considered to have pre-existing disease. Therefore, 32 new cases of HCC (including nine patients from group A with negative LCT at initial assessment) were diagnosed within the ISP. The incidence of HCC in this high-risk group of patients was 760.2 per 100,000 per year (95% CI, 538.4 to 1,073.7 per 100,000). By comparison, the incidence of HCC in HBsAg carriers with 2 years of negative screening (group C) was significantly lower at 213.8 per 100,000 per year (95% CI, 111.3 to 410.7 per 100,000).
Tumor Size and Management
The mean tumor size of the 47 HCCs detected in the ISP was 2.98 cm (range, 0.6 to 6.0 cm), and the mean tumor size in group A, B, and C was 3.02, 2.91, and 4.82 cm, respectively. There was a significant difference between groups A and B combined as compared to group C for HCC patients (P = .01, Wilcoxon rank-sum test). The tumor size of the patients with a detectable lesion on surveillance AUS was compared with that of patients without ultrasonographic evidence of tumor. The mean tumor size of the two respective groups was 3.50 versus 2.55 cm (P = .14) in group A, and 2.81 versus 3.02 cm (P = .71) in group B. Twenty-one patients (44.7%) from the ISP were confirmed to have two or more primary tumor sites at the time of diagnosis. Seven patients from group C (77.8%) had unifocal disease, and two patients (22.2%) had multifocal disease at the time of diagnosis.
Surgical resection was feasible for 17 patients (36.2%; 95% CI, 23% to 51%) in the ISP (groups A and B). The resection rate in group A was 33.3% (8 of 24 patients; 95% CI, 16% to 55%); the reasons for nonsurgical intervention included poor liver function (six patients), small cirrhotic liver (two patients), multifocal disease (seven patients), and patient refusal (one patient). Similarly, the resection rate in group B was 39.1% (9 of 23 patients; 95% CI, 21% to 61%), and the reasons for nonsurgical intervention were poor liver function (two patients), poor performance status (one patient), small cirrhotic liver (two patients), patient refusal (one patient), multifocal disease (seven patients), and poor hematologic function (one patient). Fourteen patients (29.8%) received locoregional ablative therapy including local ablation (seven patients) and selective internal radiation (seven patients). The remaining 16 patients (34.0%) were treated with either systemic chemotherapy (four patients) or supportive care only (12 patients). Four (44.4%) of nine patients diagnosed with HCC in group C underwent curative surgery. One patient presented with a ruptured HCC and required emergency partial resection. The remaining four patients had large tumors that were not suitable for either surgery or locoregional ablative therapy and therefore received best supportive care only.
Survival
The median survival of 47 patients with confirmed HCC from the ISP was 1.73 years (95% CI, 0.84 to 3.73 years), and the 3-year survival rate was 38.4% (95% CI, 22.7% to 54.2%). There was no significant difference between the 15 patients with HCC detected within 6 months from initial screening and 32 patients diagnosed during the ISP. The median survival of the two groups was 1.63 years (95% CI, 0.09 to 4.05 years) and 1.73 years (95% CI, 0.75 to 4.76 years), respectively (P = .69). The 3-year survival rate for patients with HCC detected within 6 months was 32.1% (95% CI, 9.4% to 52.2%) and for the ISP group was 37.1% (95% CI, 15% to 60.8%). There is no statistically significant difference in survival between patients who received surgical resection, locoregional ablative therapy, or supportive care (P = .59) (Fig 2). The median survival of patients who received surgical resection, locoregional ablative therapy, or supportive care was 1.73 years (95% CI, 0.09 to 4.13 years), 3.73 years (95% CI, 1.02 to 4.36 years), and 1.63 years (95% CI, 0.32 to 3.17 years), respectively, and the 3-year survival rates were 23.5% (95% CI, 0.0% to 58.7%), 54.5% (95% CI, 25.1% to 83.9%), and 34.4% (95% CI, 9.4% to 59.4%), respectively. HCC was the primary cause of death registered for 77.8% of the patients in group A and for 91% of the patients in group B. Other causes of death included cirrhosis and cerebral vascular accident. Five of the nine patients with confirmed HCC in group C are still alive at the time of this writing; thus, the median survival has not been reached. The four deaths were directly related to tumor progression.
DISCUSSION
The high incidence of HCC detected in patients in the ISP can be attributed to detection of subclinical tumors by additional imaging techniques and frequent surveillance of this high-risk group. The annual incidence of HCC in this subgroup of patients is 760.2 per 100,000 per year, which is significantly higher than that for the group of patients with normal surveillance tests for 2 years. Additional imaging by LCT was capable of detecting tumors that were not apparent on screening AUS.12 Ebara et al13 reported the change of sonographic patterns of small HCC over time with repeated examination. As the tumor size increased to approximately 2.5 cm, the tumor nodule changed from a hypoechoic pattern to an isoechoic pattern, which can be missed by AUS alone. Only when the nodule reached 3 cm did it show a hyperechoic or mixed pattern and became detectable by AUS. During this phase of transition, the tumor can be detected by an additional imaging technique such as a CT or LCT scan. Elevated AFP and negative AUS at initial screening and subsequent surveillance were noted in 57 patients in group A and 69 patients in group B. LCT was capable of detecting Lipiodol-retaining lesions in 24 patients (12 from each group) that were subsequently confirmed histologically to be HCC. The respective mean tumor size in the two groups was 3.02 and 2.91 cm, respectively, falling in the 2.5 to 3.0 cm range when the tumor could be isoechoic.
Intensity of surveillance of a target population is the other reason for the observed high incidence. Groups A and B comprised preselected high-risk HBV carriers with elevated serum AFP or abnormal AUS results. Colombo et al14 detected 29 HCCs after surveillance of a group of Italian patients with well-compensated cirrhosis. Of these patients, 12 (41%) had persistently elevated serum AFP and 11 (38%) had a fluctuating level. Similar to our program, these authors performed quarterly surveillance serum AFP measurements and AUS in this subgroup. Izzo et al15 also performed quarterly surveillance serum AFP measurements and AUS in a group of 325 high-risk patients with viral hepatitis or cirrhosis. HCC was found in 20.3% of the patients with histologic evidence of severe chronic active hepatitis or cirrhosis, among which 74.6% had an elevated serum AFP level. In our study, after excluding the 15 patients with HCC that was diagnosed at initial screening, 32 HCCs (nine from group A and 23 from group B) were detected in a group of 156 HBV carriers with abnormal screening test results (20.5%). Being able to identify this population with a high incidence of HCC makes it possible to filter the general population of patients with HBV and select a more specific target group for ISP.
The primary objectives of a screening-and-surveillance program are to detect early-stage tumors, increase resection rates, and improve survival. In this study, small tumors (mean size, 2.98 cm) were detected, but the resection rate remained relatively low at 36.2%. This prospective study confirmed the results of a retrospective series from Hong Kong in which the resection rate was 26.8% in 142 HCCs that were detected by screening a similar population with a mean tumor size of 3.1 cm.16 The best explanation for the unsatisfactory outcome is coexisting cirrhosis and multifocal disease. Two Italian studies targeted a population with established histologic diagnosis of cirrhosis, and both reported low resection rates and poor survival.14,15 Izzo et al15 detected 67 HCCs in 1,125 patients with chronic hepatitis and cirrhosis. Fourteen of their patients with solitary tumors were not operable because of poor hepatic reserve, and 29 patients (43.2%) had unresectable multifocal or extrahepatic disease. In contrast, McMahon et al9 studied a group of Alaskan HBV carriers without established diagnosis of cirrhosis and reported a much higher resection rate of 68.8%. The randomized study from China also reported a high resection rate, and again cirrhosis was not an inclusion criterion.7 Our ISP targeted high-risk HBV carriers. The liver function of patients in the ISP was worse than in the group with normal surveillance for 2 years, and there were also more patients with ultrasonographic evidence of cirrhosis (Table 1). Although histologic evidence was lacking, the clinical findings were suggestive of more advanced cirrhosis among patients in the ISP. As a result, a high proportion of patients with multifocal disease and poor hepatic reserves who were not fit for surgery were observed.
Early detection by surveillance is capable of identifying small tumors for locoregional therapy even if surgical resection is not feasible. Locoregional therapy is a feasible therapeutic option for this group of patients and may improve survival. Two meta-analyses have reported improvement in the 2-year survival rates for transarterial chemoembolization over best supportive care for patients with unresectable HCC.17,18 Sangiovanni et al19 updated the survival outcomes of a prospective surveillance program from Milan.14 These authors reported a significant reduction of yearly mortality rates for treated patients in the third quinquennia. Their therapeutic strategy had shifted from more radical treatment such as hepatic resection or liver transplantation to less aggressive locoregional ablative therapies. The proportion of patients undergoing radical treatment decreased from 20% to 8%, whereas the proportion of patients receiving locoregional ablative therapies increased from 8% to 35%. The shift in paradigm explained the reduction in mortality. Fourteen patients (29.8%) from our study received locoregional ablative therapies. The median survival was numerically longer than that for the surgical group, but the difference was not statistically different (P = .59). Limited by the small sample size, a meaningful comparison cannot be made. The overall survival outcome from both surgical resection and locoregional ablative therapy was disappointing for patients with relatively early detection of illness.
Orthotopic liver transplantation (OLT) is by far the only intervention that can potentially cure both HCC and the underlying liver disease.20 Patients who had successful liver transplantation for small HCC attained a 4-year survival rate of 75%. Mazzaferro et al21 proposed the "Milan criteria" for selection of the best candidates for OLT. They suggested that patients should have either presence of a solitary tumor less than 5 cm in diameter or, if multifocal, no more than three lesions that are less than 3 cm in diameter. Twenty-three of the 47 patients (48.9%) diagnosed with HCC in our program had a solitary lesion less than 5 cm in diameter. Only approximately half (11) of the patients had surgical resection, and the rest (12 patients) were not eligible because of poor hepatic reserve, small cirrhotic liver, or refusal. Most would fulfill the Milan criteria and might potentially benefit from OLT. However, access to OLT is limited by scarcity of donor organs, long, unpredictable waiting time, and funding shortage.22 The role of OLT in surveillance programs of HBV carriers remains to be defined.
In conclusion, intensive surveillance is capable of detecting a high incidence of HCC in this population of HBV carriers with abnormal screening test results. Poor hepatic reserve and presence of multifocal disease at presentation were the main reasons for the low resection rate. We were not able to demonstrate a significant clinical benefit despite detection of small-sized tumors. Those performing surveillance studies in the future should consider incorporation of therapy aimed at long-term control of small-sized tumors.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by a research grant from the Hong Kong Cancer Fund, Hong Kong Special Administrative Region, China.
Presented at the 40th Annual Meeting of the American Society of Clinical Oncology, June 5-8, 2004, New Orleans, LA.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
El-Serag HB: Hepatocellular carcinoma: An epidemiologic view. J Clin Gastroenterol 35:S72-S78, 2002 (suppl 2)
Mericanc I, Guan R, Amarapuka D, et al: Chronic hepatitis B virus infection in Asian countries. J Gastroenterol Hepatol 15:1356-1361, 2000
El-Serag HB, Davila JA, Petersen NJ, et al: The continuing increase in the incidence of hepatocellular carcinoma in the United States: An update. Ann Intern Med 18:817-823, 2003 [Erratum: Ann Intern Med 140:151, 2004; Summary for patients: Ann Intern Med 139:I28, 2003]
Beasley RP, Hwang LY, Lin CC, et al: Hepatocellular carcinoma and hepatitis B virus: A prospective study of 22,707 men in Taiwan. Lancet 2(8256):1129-1133, 1981
Liaw YF, Tai DI, Chu CM, et al: Early detection of hepatocellular carcinoma in patients with chronic type B hepatitis: A prospective study. Gastroenterology 90:263-267, 1986
Chalasani N, Said A, Ness R, et al: Screening for hepatocellular carcinoma in patients with cirrhosis in the United States: Results of a national survey. Am J Gastroenterol 94:2224-2229, 1999
Yang B, Zhang B, Xu Y, et al: Prospective study of early detection for primary liver cancer. J Cancer Res Clin Oncol 123:357-360, 1997
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 22:432-438, 1995
McMahon BJ, Bulkow L, Harpster A, et al: Screening for hepatocellular carcinoma in Alaska natives infected with chronic hepatitis B: A 16-year population-based study. Hepatology 32:842-847, 2000
Palma LD: Diagnostic imaging and interventional therapy of hepatocellular carcinoma. Br J Radiol 71:808-818, 1998
Johnson PJ, Leung N, Cheng P, et al: "Hepatoma-specific" alphafetoprotein may permit preclinical diagnosis of malignant change in patients with chronic liver disease. Br J Cancer 75:236-240, 1997
Mok TS, Yu SC, Lee C, et al: False-negative rate of abdominal sonography for detection hepatocellular carcinoma in patients with hepatitis B and elevated serum alpha-fetoprotein level. AJR Am J Roentgenol 183:453-458, 2004
Ebara M, Ohto M, Shinagawa T, et al: Natural history of minute hepatocellular carcinoma smaller than three centimeters complicating cirrhosis: A study in 22 patients. Gastroenterology 90:289-298, 1986
Colombo M, de Franchis R, Del Ninno E, et al: Hepatocellular carcinoma in Italian patients with cirrhosis. N Engl J Med 325:675-680, 1991
Izzo F, Cremona F, Ruffolo F, et al: Outcome of 67 patients with hepatocellular cancer detected during screening of 1125 patients with chronic hepatitis. Ann Surg 227:513-518, 1998
Yuen MF, Cheng CC, Lauder IJ, et al: Early detection of hepatocellular carcinoma increases the chance of treatment: Hong Kong experience. Hepatology 31:330-335, 2000
Camma C, Schepis F, Orlando A, et al: Transarterial chemoembolization for unresectable hepatocellular carcinoma: Meta-analysis of randomized controlled trials. Radiology 224:47-54, 2002
Llovet JM, Bruix J: Systematic review of randomized trials for unresectable hepatocellular carcinoma: Chemoembolization improves survival. Hepatology 37:429-442, 2003
Sangiovanni A, Del Ninno E, Fasani P, et al: Increased survival of cirrhotic patients with a hepatocellular carcinoma detected during surveillance. Gastroenterology 126:1005-1014, 2004
Cillo U, Vitale A, Bassanello M, et al: Liver transplantation for the treatment of moderately or well-differentiated hepatocellular carcinoma. Ann Surg 239:150-159, 2004
Mazzaferro V, Regalia E, Doci R, et al: Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med 334:693-699, 1996
Yao FY, Bass NM, Nikolai B, et al: Liver transplantation for hepatocellular carcinoma: Analysis of survival according to the intention-to-treat principle and dropout from the waiting list. Liver Transpl 8:873-883, 2002(Tony S.K. Mok, Winnie Yeo)
Department of Oncology, University of Birmingham, Birmingham, United Kingdom
ABSTRACT
PURPOSE: To study the incidence and treatment outcomes of hepatocellular carcinoma (HCC) detected in an intensive surveillance program (ISP) of hepatitis B virus (HBV) carriers.
PATIENTS AND METHODS: We screened 1,018 HBV carriers by serum alpha-fetoprotein (AFP) measurement and abdominal ultrasonography (AUS). Patients with an abnormal AFP level or AUS result were enrolled in an ISP that included Lipiodol computed tomography followed by AFP measurement/AUS every 3 months for 2 years and then every 6 months thereafter. The rest were on routine surveillance for 2 years.
RESULTS: A total of 9,849 serum AFP measurements and 3,053 AUSs were performed. After a median follow-up of 4.12 years, we diagnosed 24 HCCs among 78 patients with abnormal screening test results at enrollment (group A); 23 HCCs among 93 patients with only abnormal surveillance test results during follow-up (group B); and nine HCCs among 847 patients with 2 years of normal surveillance (group C). Annual incidence of HCC in the ISP was 760.2 (95% CI, 538.4 to 1,073.7) per 100,000. Mean tumor sizes were 3.02, 2.91, and 4.82 cm in groups A, B, and C, respectively (P = .01). Tumor resection rate of the ISP was 36.2%, although another 29.8% of the patients were eligible for locoregional ablative therapy.
CONCLUSION: This study illustrated that a high incidence of relatively small HCCs may be detected by using intensive surveillance of high-risk HBV carriers. However, the surgical resection rate was low, and we were not able to demonstrate clinical benefit with the early detection. Future surveillance studies should consider incorporation of therapy aimed at long-term control of small-sized tumors.
INTRODUCTION
Hepatocellular carcinoma (HCC) is one of the most common malignancies in the world.1 Thousands of patients die as a result of the disease because of late presentation and lack of effective therapies for advanced disease, and the incidence of HCC in Western countries continues to rise.2,3 In Asia, the incidence of HCC is high at more than 30 cases per 100,000 population per year in some countries. Chronic hepatitis B virus (HBV) infection and cirrhosis are the main reasons for the high incidence of HCC in Asia. Beasley et al4 first established the etiologic link between HBV and HCC and subsequently confirmed that patients with positive hepatitis B surface antigen (HBsAg) are at a higher risk of developing HCC.5 The annual incidence of HCC in this group of patients ranges between 1% and 6%. Therefore, a surveillance program targeted at this population with chronic viral hepatitis may detect early resectable HCC and has the potential to reduce mortality.
Serum alpha-fetoprotein (AFP) measurement and abdominal ultrasonography (AUS) are widely accepted as screening tests for the early detection of HCC. It is common practice in the United States to routinely screen patients with cirrhosis for early HCC.6 Large-scale screening programs that target HBsAg-positive patients have defined the validity of these screening tests.7-9 Sensitivity and specificity of serum AFP measurements ranges from 64.3% to 96.9% and 93.8% to 95%, respectively. AUS has also been considered to be an effective surveillance tool with sensitivity and specificity ranging from 71.4% to 84.3% and 93.8% to 97%, respectively. Early detection of HCC by serum AFP and AUS is feasible, but no studies have demonstrated a definite improvement in survival or a reduction in the mortality rate. To date, the clinical benefits of screening programs for early detection of HCC remain controversial.
We hypothesized that an intensive surveillance program (ISP) that includes quarterly surveillance and the addition of a more sensitive imaging technique for a high-risk group of HBV carriers may detect a higher number of early HCCs for curative surgery and possibly result in an improvement in survival outcomes. In this study, Chinese HBV carriers with either an elevation of serum AFP or abnormal AUS results were enrolled into an ISP to determine incidence, tumor size, resectability, and survival outcome.
PATIENTS AND METHODS
Eligibility
Between October 1997 and November 2000, all patients attending the hepatology clinic at the Prince of Wales Hospital (the Chinese University of Hong Kong, Hong Kong Special Administrative Region, China) were evaluated for eligibility for the study. The hepatology clinic treats patients with liver disease from the general population of 1.4 million people residing in New Territories East in Hong Kong. The sources of referral include family doctors, general internists, and other subspecialists from the hospital catchment area. Reasons for referral ranged from mild hepatic dysfunction to advanced cirrhosis. Eligible patients were screened according to the following criteria: age between 40 and 70 years; seropositive for HBsAg; clinical grade A or B cirrhosis according to Child's classification; and life expectancy of more than 2 years. Patients with non–HBV-related cirrhosis or history of malignancy were excluded. The study was approved by the Clinical Research Ethics Committee of the Chinese University of Hong Kong, and written informed consent was obtained from all patients before entry into the trial.
Routine Investigation and Standard Surveillance
Evaluation at enrollment included a full medical history, physical examination, urinalysis, CBC, liver-function test, and assessment of serum electrolytes, creatinine, International Normalized Ratio (INR), HBsAg, hepatitis B e antigen (HBeAg), and antibody to HBeAg (anti-HBeAg). Initial screening serum AFP measurements and AUSs were performed after enrollment. The microparticles enzyme immunoassay method (Abbott Laboratories, Chicago, IL) was used for qualitative measurement of serum AFP. Serum AFP greater than 20 μg/L was considered to be abnormally elevated. AUS was performed by one of the two designated radiologists at the Department of Diagnostic Imaging, Prince of Wales Hospital, each of whom had more than 10 years of experience. The standard ultrasound system equipment was the HDI 3000 (Advance Technology Laboratories, Bothwell, WA). An electronic curvilinear 3.5-MHz real-time transducer was used to scan subcostally and intercostally with the patient in the supine position first and then in the left decubitus position. Positive AUS implied the presence of a focal solid nodule with hypoechoic, hyperechoic, or mixed sonographic pattern, and probable AUS implied the presence of an atypical lesion. Both positive and probable AUSs were considered to be abnormal.
Patients with either elevated serum AFP or abnormal AUS results at initial screening were admitted to the ISP. Patients with normal initial screening serum AFP and AUS were followed every 6 months for 2 years with repeated surveillance of serum AFP (Fig 1). AUS was repeated after 2 years for this group of patients, after which they had no additional routine surveillance in the program. The patients were referred back to the hepatology clinic at the Prince of Wales Hospital, at which serum AFP measurements or AUSs were performed by indication only.
ISP
Lipiodol computed tomography (LCT) was considered to be a sensitive imaging modality for diagnosis of small HCC when the study was first initiated.10 All patients enrolled onto the ISP were offered additional investigation by LCT. The hepatic angiogram involved access to the arterial system via the femoral artery using the Seldinger technique. Celiac axis, superior mesenteric artery, and selective hepatic artery angiograms were performed as contrast material was injected into the respective arteries. Angiographic images were examined for evidence of hypervascular areas in the liver that would suggest HCC. Lipiodol, an oily contrast medium that tends to be accumulated and retained selectively by tumor tissue, was selectively injected into the hepatic arteries during the procedure. A CT scan was performed 10 days after arterial Lipiodol administration to detect any Lipiodol retention. The presence of hypervascular lesion(s) with corresponding Lipiodol retention was an indication for histologic confirmation. Liver biopsy was offered to all patients with hypervascular lesions on LCT or the presence of space-occupying lesions on AUS. After exclusion of a diagnosis of HCC, AFP measurement and AUS were repeated every 3 months for 2 years and subsequently every 6 months. If a new abnormality was detected in either surveillance test, the patient was reinvestigated by LCT and/or biopsy. Liver biopsies were fixed in 10% formalin and stained with hematoxylin and eosin. A designated hepatic pathologist was responsible for evaluating all liver biopsies for this study.
Surgical and Nonsurgical Treatment for HCC
The hepatobiliary surgical team assessed all patients who had either a hypervascular lesion on LCT or a confirmed diagnosis of HCC. Tumor resectability was determined by tumor size, tumor location, proximity to major vessels, multifocal disease, and hepatic reserve. Surgery was offered to all patients who had a potentially resectable tumor. Surgical options included wedge resection, subsegmentectomy, and segmentectomy. Postoperative radioactive iodine-131 was offered to selected patients according to institutional protocol. Patients who had declined or were not eligible for surgery were evaluated by the oncology team for consideration of locoregional ablative therapy including alcohol injection, radio-frequency ablation, microwave ablation and selective internal radiation, systemic chemotherapy, or supportive care. All patients with confirmed HCC were followed up at the hepatoma clinic until death.
Statistical Analysis
Patients were classified into three groups to clearly delineate the effect of the surveillance program. Patients with elevated serum AFP and/or abnormal AUS results at initial screening represented those with pre-existing HCC or who were at high risk of HCC (group A). Patients with normal serum AFP and AUS at initial screening but subsequently had an abnormal surveillance test result were also considered to be at high risk for developing HCC (group B). Both groups A and B were enrolled to the ISP. Patients with normal serum AFP and AUS at initial screening and normal surveillance for 2 years were considered to be at low risk for developing HCC (group C). Comparisons among the three groups with respect to sex, age, bilirubin, alkaline phosphatase, albumin, INR, ascites, encephalopathy, cirrhosis, and Child's stage were made by using the 2 test. Survival was defined as the time between the date of diagnosis of HCC and the last report date. All comparisons of overall survival were done by using log-rank test; Kaplan-Meier estimates of patients with confirmed diagnosis of HCC are shown graphically to indicate the survival patterns between groups (Fig 2).
This prospective surveillance study was intended to target a population of high-risk HBV carriers for the ISP. Providing the best estimate that up to half of all HBV carriers with elevated serum AFP or abnormal AUS results would eventually develop HCC and operationally defining a 10% error rate as acceptably accurate, the minimal sample size for the ISP should be 96. As established from our previous study,11 approximately 9.1% of consecutive attendees at the hepatology clinic were found to have elevated serum AFP. We therefore calculated the sample size for the ISP to be 1,055.
RESULTS
Patients' Characteristics
Between October 1997 and November 2000, 1,018 (88.4%) of 1,152 eligible patients were enrolled onto the study. The male to female ratio was 688:330, and the mean age was 48.7 years (range, 40 to 69 years). The number of patients in groups A, B, and C was 78, 93, and 847, respectively. All patients were followed up from enrollment until death or last follow-up. The data set was frozen for analysis in April 2004. Groups A and B (171 patients) were enrolled onto the ISP. The total duration of follow-up for this group of patients within the program was 704.8 person-years. For group C, the total duration of follow-up was 3,504.7 person-years. The average duration of follow-up for the ISP group and group C was 4.12 and 4.13 years, respectively.
Patients' characteristics and liver functions at enrollment are listed in Table 1. Groups A and B had a significantly higher percentage of patients with elevated total bilirubin as per Child's classification (> 34 μmol/L; normal range, 5 to 21 μmol/L) (P < .001) and low albumin (< 35 g/L; normal range, 35 to 50 g/L) than group C (P < .001). There were also significant differences in the number of patients with an elevated INR (P < .001), but the relevance of the difference in history of encephalopathy is limited by the small number of patients. This finding corresponded to the high percentage of ultrasonographic evidence of cirrhosis in groups A and B. However, the overall number of patients with elevated total bilirubin, prolonged INR, ascites, or encephalopathy was relatively small, and there was no significant difference in the portion of patients classified as having Child's class B cirrhosis among the three groups (P = .06).
Surveillance Tests and HCC
A total of 9,849 serum AFP measurements (group A, 1,205; group B, 1,481; group C, 7,163) and 3,053 AUSs (group A, 841; group B, 979; group C, 1,233) were performed in this study. Group B had the highest number of surveillance tests per patient, with an average of 15.4 serum AFP measurements and 10.5 AUSs per patient. Comparatively, group A had an average of 15.4 serum AFP measurements and 10.5 AUSs per patient; and group C had an average of 8.5 serum AFP measurements and 1.46 AUSs per patient. An LCT scan was offered to all 171 patients enrolled onto the ISP, but 35 patients declined. In total, 110 patients had one LCT scan, 20 patients had two LCT scans, and six patients had three LCT scans. Patients with positive or suspicious findings underwent biopsy. Patients with negative findings continued with surveillance every 3 months, and LCT would be repeated on two consecutive elevations of serum AFP, notable enlargement of the lesion, or presence of new lesion(s) on AUS.
Fifty-seven (73.1%) of the 78 patients from group A had elevation of serum AFP above 20 μg/L, 15 (19.2%) had abnormal AUS results, and six (7.7%) had abnormalities in both tests at initial screening. Twenty-four HCCs were diagnosed in this group, 15 of which (62.5%) were diagnosed within 6 months from initial enrollment and were considered to be pre-existing HCC. The other nine patients had negative LCT at initial assessment, and HCC was diagnosed during the subsequent follow-up. Twelve (21.0%) of 57 patients with elevated serum AFP, seven (46.7%) of 15 patients with abnormal lesions on AUS, and five (83.3%) of six patients with both elevated serum AFP and abnormal AUS results were diagnosed with HCC.
Group B comprised 93 patients with normal initial screening tests but abnormal surveillance test results during the 2-year follow-up. Twenty-three HCCs were confirmed in this group. Sixty-nine patients (74.2%) in group B had elevated serum AFP, 23 patients (24.7%) had abnormal AUS results, and one patient (1.1%) had both abnormalities. The number of HCCs detected in groups A, B, and C was 12 (17.1%), 10 (41.7%), and one (100%), respectively. The median duration from enrollment to diagnosis of HCC was 2.47 years.
Eight hundred forty-seven patients from group C had normal serum AFP and AUS during the 2 years of standard surveillance, of whom 9 were confirmed to have HCC. The median duration from enrollment to diagnosis was 3.56 years, which was significantly longer than the patients with HCC in group B (P = .015).
The 15 patients with HCC diagnosed within 6 months of enrollment were considered to have pre-existing disease. Therefore, 32 new cases of HCC (including nine patients from group A with negative LCT at initial assessment) were diagnosed within the ISP. The incidence of HCC in this high-risk group of patients was 760.2 per 100,000 per year (95% CI, 538.4 to 1,073.7 per 100,000). By comparison, the incidence of HCC in HBsAg carriers with 2 years of negative screening (group C) was significantly lower at 213.8 per 100,000 per year (95% CI, 111.3 to 410.7 per 100,000).
Tumor Size and Management
The mean tumor size of the 47 HCCs detected in the ISP was 2.98 cm (range, 0.6 to 6.0 cm), and the mean tumor size in group A, B, and C was 3.02, 2.91, and 4.82 cm, respectively. There was a significant difference between groups A and B combined as compared to group C for HCC patients (P = .01, Wilcoxon rank-sum test). The tumor size of the patients with a detectable lesion on surveillance AUS was compared with that of patients without ultrasonographic evidence of tumor. The mean tumor size of the two respective groups was 3.50 versus 2.55 cm (P = .14) in group A, and 2.81 versus 3.02 cm (P = .71) in group B. Twenty-one patients (44.7%) from the ISP were confirmed to have two or more primary tumor sites at the time of diagnosis. Seven patients from group C (77.8%) had unifocal disease, and two patients (22.2%) had multifocal disease at the time of diagnosis.
Surgical resection was feasible for 17 patients (36.2%; 95% CI, 23% to 51%) in the ISP (groups A and B). The resection rate in group A was 33.3% (8 of 24 patients; 95% CI, 16% to 55%); the reasons for nonsurgical intervention included poor liver function (six patients), small cirrhotic liver (two patients), multifocal disease (seven patients), and patient refusal (one patient). Similarly, the resection rate in group B was 39.1% (9 of 23 patients; 95% CI, 21% to 61%), and the reasons for nonsurgical intervention were poor liver function (two patients), poor performance status (one patient), small cirrhotic liver (two patients), patient refusal (one patient), multifocal disease (seven patients), and poor hematologic function (one patient). Fourteen patients (29.8%) received locoregional ablative therapy including local ablation (seven patients) and selective internal radiation (seven patients). The remaining 16 patients (34.0%) were treated with either systemic chemotherapy (four patients) or supportive care only (12 patients). Four (44.4%) of nine patients diagnosed with HCC in group C underwent curative surgery. One patient presented with a ruptured HCC and required emergency partial resection. The remaining four patients had large tumors that were not suitable for either surgery or locoregional ablative therapy and therefore received best supportive care only.
Survival
The median survival of 47 patients with confirmed HCC from the ISP was 1.73 years (95% CI, 0.84 to 3.73 years), and the 3-year survival rate was 38.4% (95% CI, 22.7% to 54.2%). There was no significant difference between the 15 patients with HCC detected within 6 months from initial screening and 32 patients diagnosed during the ISP. The median survival of the two groups was 1.63 years (95% CI, 0.09 to 4.05 years) and 1.73 years (95% CI, 0.75 to 4.76 years), respectively (P = .69). The 3-year survival rate for patients with HCC detected within 6 months was 32.1% (95% CI, 9.4% to 52.2%) and for the ISP group was 37.1% (95% CI, 15% to 60.8%). There is no statistically significant difference in survival between patients who received surgical resection, locoregional ablative therapy, or supportive care (P = .59) (Fig 2). The median survival of patients who received surgical resection, locoregional ablative therapy, or supportive care was 1.73 years (95% CI, 0.09 to 4.13 years), 3.73 years (95% CI, 1.02 to 4.36 years), and 1.63 years (95% CI, 0.32 to 3.17 years), respectively, and the 3-year survival rates were 23.5% (95% CI, 0.0% to 58.7%), 54.5% (95% CI, 25.1% to 83.9%), and 34.4% (95% CI, 9.4% to 59.4%), respectively. HCC was the primary cause of death registered for 77.8% of the patients in group A and for 91% of the patients in group B. Other causes of death included cirrhosis and cerebral vascular accident. Five of the nine patients with confirmed HCC in group C are still alive at the time of this writing; thus, the median survival has not been reached. The four deaths were directly related to tumor progression.
DISCUSSION
The high incidence of HCC detected in patients in the ISP can be attributed to detection of subclinical tumors by additional imaging techniques and frequent surveillance of this high-risk group. The annual incidence of HCC in this subgroup of patients is 760.2 per 100,000 per year, which is significantly higher than that for the group of patients with normal surveillance tests for 2 years. Additional imaging by LCT was capable of detecting tumors that were not apparent on screening AUS.12 Ebara et al13 reported the change of sonographic patterns of small HCC over time with repeated examination. As the tumor size increased to approximately 2.5 cm, the tumor nodule changed from a hypoechoic pattern to an isoechoic pattern, which can be missed by AUS alone. Only when the nodule reached 3 cm did it show a hyperechoic or mixed pattern and became detectable by AUS. During this phase of transition, the tumor can be detected by an additional imaging technique such as a CT or LCT scan. Elevated AFP and negative AUS at initial screening and subsequent surveillance were noted in 57 patients in group A and 69 patients in group B. LCT was capable of detecting Lipiodol-retaining lesions in 24 patients (12 from each group) that were subsequently confirmed histologically to be HCC. The respective mean tumor size in the two groups was 3.02 and 2.91 cm, respectively, falling in the 2.5 to 3.0 cm range when the tumor could be isoechoic.
Intensity of surveillance of a target population is the other reason for the observed high incidence. Groups A and B comprised preselected high-risk HBV carriers with elevated serum AFP or abnormal AUS results. Colombo et al14 detected 29 HCCs after surveillance of a group of Italian patients with well-compensated cirrhosis. Of these patients, 12 (41%) had persistently elevated serum AFP and 11 (38%) had a fluctuating level. Similar to our program, these authors performed quarterly surveillance serum AFP measurements and AUS in this subgroup. Izzo et al15 also performed quarterly surveillance serum AFP measurements and AUS in a group of 325 high-risk patients with viral hepatitis or cirrhosis. HCC was found in 20.3% of the patients with histologic evidence of severe chronic active hepatitis or cirrhosis, among which 74.6% had an elevated serum AFP level. In our study, after excluding the 15 patients with HCC that was diagnosed at initial screening, 32 HCCs (nine from group A and 23 from group B) were detected in a group of 156 HBV carriers with abnormal screening test results (20.5%). Being able to identify this population with a high incidence of HCC makes it possible to filter the general population of patients with HBV and select a more specific target group for ISP.
The primary objectives of a screening-and-surveillance program are to detect early-stage tumors, increase resection rates, and improve survival. In this study, small tumors (mean size, 2.98 cm) were detected, but the resection rate remained relatively low at 36.2%. This prospective study confirmed the results of a retrospective series from Hong Kong in which the resection rate was 26.8% in 142 HCCs that were detected by screening a similar population with a mean tumor size of 3.1 cm.16 The best explanation for the unsatisfactory outcome is coexisting cirrhosis and multifocal disease. Two Italian studies targeted a population with established histologic diagnosis of cirrhosis, and both reported low resection rates and poor survival.14,15 Izzo et al15 detected 67 HCCs in 1,125 patients with chronic hepatitis and cirrhosis. Fourteen of their patients with solitary tumors were not operable because of poor hepatic reserve, and 29 patients (43.2%) had unresectable multifocal or extrahepatic disease. In contrast, McMahon et al9 studied a group of Alaskan HBV carriers without established diagnosis of cirrhosis and reported a much higher resection rate of 68.8%. The randomized study from China also reported a high resection rate, and again cirrhosis was not an inclusion criterion.7 Our ISP targeted high-risk HBV carriers. The liver function of patients in the ISP was worse than in the group with normal surveillance for 2 years, and there were also more patients with ultrasonographic evidence of cirrhosis (Table 1). Although histologic evidence was lacking, the clinical findings were suggestive of more advanced cirrhosis among patients in the ISP. As a result, a high proportion of patients with multifocal disease and poor hepatic reserves who were not fit for surgery were observed.
Early detection by surveillance is capable of identifying small tumors for locoregional therapy even if surgical resection is not feasible. Locoregional therapy is a feasible therapeutic option for this group of patients and may improve survival. Two meta-analyses have reported improvement in the 2-year survival rates for transarterial chemoembolization over best supportive care for patients with unresectable HCC.17,18 Sangiovanni et al19 updated the survival outcomes of a prospective surveillance program from Milan.14 These authors reported a significant reduction of yearly mortality rates for treated patients in the third quinquennia. Their therapeutic strategy had shifted from more radical treatment such as hepatic resection or liver transplantation to less aggressive locoregional ablative therapies. The proportion of patients undergoing radical treatment decreased from 20% to 8%, whereas the proportion of patients receiving locoregional ablative therapies increased from 8% to 35%. The shift in paradigm explained the reduction in mortality. Fourteen patients (29.8%) from our study received locoregional ablative therapies. The median survival was numerically longer than that for the surgical group, but the difference was not statistically different (P = .59). Limited by the small sample size, a meaningful comparison cannot be made. The overall survival outcome from both surgical resection and locoregional ablative therapy was disappointing for patients with relatively early detection of illness.
Orthotopic liver transplantation (OLT) is by far the only intervention that can potentially cure both HCC and the underlying liver disease.20 Patients who had successful liver transplantation for small HCC attained a 4-year survival rate of 75%. Mazzaferro et al21 proposed the "Milan criteria" for selection of the best candidates for OLT. They suggested that patients should have either presence of a solitary tumor less than 5 cm in diameter or, if multifocal, no more than three lesions that are less than 3 cm in diameter. Twenty-three of the 47 patients (48.9%) diagnosed with HCC in our program had a solitary lesion less than 5 cm in diameter. Only approximately half (11) of the patients had surgical resection, and the rest (12 patients) were not eligible because of poor hepatic reserve, small cirrhotic liver, or refusal. Most would fulfill the Milan criteria and might potentially benefit from OLT. However, access to OLT is limited by scarcity of donor organs, long, unpredictable waiting time, and funding shortage.22 The role of OLT in surveillance programs of HBV carriers remains to be defined.
In conclusion, intensive surveillance is capable of detecting a high incidence of HCC in this population of HBV carriers with abnormal screening test results. Poor hepatic reserve and presence of multifocal disease at presentation were the main reasons for the low resection rate. We were not able to demonstrate a significant clinical benefit despite detection of small-sized tumors. Those performing surveillance studies in the future should consider incorporation of therapy aimed at long-term control of small-sized tumors.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by a research grant from the Hong Kong Cancer Fund, Hong Kong Special Administrative Region, China.
Presented at the 40th Annual Meeting of the American Society of Clinical Oncology, June 5-8, 2004, New Orleans, LA.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
El-Serag HB: Hepatocellular carcinoma: An epidemiologic view. J Clin Gastroenterol 35:S72-S78, 2002 (suppl 2)
Mericanc I, Guan R, Amarapuka D, et al: Chronic hepatitis B virus infection in Asian countries. J Gastroenterol Hepatol 15:1356-1361, 2000
El-Serag HB, Davila JA, Petersen NJ, et al: The continuing increase in the incidence of hepatocellular carcinoma in the United States: An update. Ann Intern Med 18:817-823, 2003 [Erratum: Ann Intern Med 140:151, 2004; Summary for patients: Ann Intern Med 139:I28, 2003]
Beasley RP, Hwang LY, Lin CC, et al: Hepatocellular carcinoma and hepatitis B virus: A prospective study of 22,707 men in Taiwan. Lancet 2(8256):1129-1133, 1981
Liaw YF, Tai DI, Chu CM, et al: Early detection of hepatocellular carcinoma in patients with chronic type B hepatitis: A prospective study. Gastroenterology 90:263-267, 1986
Chalasani N, Said A, Ness R, et al: Screening for hepatocellular carcinoma in patients with cirrhosis in the United States: Results of a national survey. Am J Gastroenterol 94:2224-2229, 1999
Yang B, Zhang B, Xu Y, et al: Prospective study of early detection for primary liver cancer. J Cancer Res Clin Oncol 123:357-360, 1997
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 22:432-438, 1995
McMahon BJ, Bulkow L, Harpster A, et al: Screening for hepatocellular carcinoma in Alaska natives infected with chronic hepatitis B: A 16-year population-based study. Hepatology 32:842-847, 2000
Palma LD: Diagnostic imaging and interventional therapy of hepatocellular carcinoma. Br J Radiol 71:808-818, 1998
Johnson PJ, Leung N, Cheng P, et al: "Hepatoma-specific" alphafetoprotein may permit preclinical diagnosis of malignant change in patients with chronic liver disease. Br J Cancer 75:236-240, 1997
Mok TS, Yu SC, Lee C, et al: False-negative rate of abdominal sonography for detection hepatocellular carcinoma in patients with hepatitis B and elevated serum alpha-fetoprotein level. AJR Am J Roentgenol 183:453-458, 2004
Ebara M, Ohto M, Shinagawa T, et al: Natural history of minute hepatocellular carcinoma smaller than three centimeters complicating cirrhosis: A study in 22 patients. Gastroenterology 90:289-298, 1986
Colombo M, de Franchis R, Del Ninno E, et al: Hepatocellular carcinoma in Italian patients with cirrhosis. N Engl J Med 325:675-680, 1991
Izzo F, Cremona F, Ruffolo F, et al: Outcome of 67 patients with hepatocellular cancer detected during screening of 1125 patients with chronic hepatitis. Ann Surg 227:513-518, 1998
Yuen MF, Cheng CC, Lauder IJ, et al: Early detection of hepatocellular carcinoma increases the chance of treatment: Hong Kong experience. Hepatology 31:330-335, 2000
Camma C, Schepis F, Orlando A, et al: Transarterial chemoembolization for unresectable hepatocellular carcinoma: Meta-analysis of randomized controlled trials. Radiology 224:47-54, 2002
Llovet JM, Bruix J: Systematic review of randomized trials for unresectable hepatocellular carcinoma: Chemoembolization improves survival. Hepatology 37:429-442, 2003
Sangiovanni A, Del Ninno E, Fasani P, et al: Increased survival of cirrhotic patients with a hepatocellular carcinoma detected during surveillance. Gastroenterology 126:1005-1014, 2004
Cillo U, Vitale A, Bassanello M, et al: Liver transplantation for the treatment of moderately or well-differentiated hepatocellular carcinoma. Ann Surg 239:150-159, 2004
Mazzaferro V, Regalia E, Doci R, et al: Liver transplantation for the treatment of small hepatocellular carcinomas in patients with cirrhosis. N Engl J Med 334:693-699, 1996
Yao FY, Bass NM, Nikolai B, et al: Liver transplantation for hepatocellular carcinoma: Analysis of survival according to the intention-to-treat principle and dropout from the waiting list. Liver Transpl 8:873-883, 2002(Tony S.K. Mok, Winnie Yeo)