Screening for Hepatocellular Carcinoma in Elderly Residents in a Hepatitis B- and C-endemic Area

From Journal of Gastroenterology and Hepatology

Community-based Screening for Hepatocellular Carcinoma in Elderly Residents in a Hepatitis B- and C-endemic Area

Yen-Chieh Huang; Chih-Fang Huang; Kuo-Chin Chang; Shu-Fen Hung; Jing-Houng Wang; Chao-Hung Hung; Chien-Hung Chen; Po-Lin Tseng; Kwong-Ming Kee; Yi-Hao Yen; Pei-Shan Tsai; Chin-Chen Tsai; Sheng-Nan Lu

Authors and Disclosures

Posted: 03/07/2011; J Gastroenterol Hepatol. 2011;26(1):129-134. © 2011 Blackwell Publishing

Abstract
Background and Aim:
The aim of the present study was to elucidate a reasonable model and the efficacy of hepatocellular carcinoma (HCC) screening on an elderly population.

Methods:
Two-stage HCC screening was conducted in a hepatitis C virus (HCV)-endemic area. First, participants underwent blood tests for hepatitis B surface antigen (HBsAg), anti-HCV antibody, serum α-fetoprotein (AFP), aspartate aminotransferase, alanine aminotransferase, and platelet count. Patients who were abnormal for any of the six markers were enrolled for second-stage ultrasonography. Suspected cases were referred for confirmation. HCC cases were followed for 4 years. All patients were linked to national mortality and cancer register databases to identify newly-developed HCC, 30 months after screening.

Results:
A total of 461 males and 541 females were screened for HCC, with 15.1% testing positive for HBsAg and 44.3% positive for anti-HCV. Among them, 619 (61.8%) met the criteria of ultrasonographic screening; 527 (85.1%) responded, and 16 confirmed HCC (male/female = 8/8, 68.8 ± 8 years) cases were detected. All tumor diameters were less than 5 cm, and six were less than 2 cm. AFP and thrombocytopenia were two independent predictive factors of HCC. The overall survival rates of detected cases were 93.8% and 56.3% was 1 and 4 years, respectively. The only good prognostic predictor was "underwent curative treatment". Another seven non-HCC residents developed HCC after screening, and five of these were with either thrombocytopenia or AFP elevation.Conclusion: Under economical consideration, AFP and platelet count should be feasible screening markers of risk identification. Early detection and prompt treatment results in good prognosis in an aged population.

Methods
Results
Discussion
References

Discussion Only;

In this community-based HCC screening on an aged population in an HCV-endemic township, we found that platelet count and AFP could be feasible markers for high-risk patient identification and early diagnosis; cases undergoing curative treatment resulted in good prognosis.
A combination of AFP and ultrasonography as two-stage screening is currently the most acceptable screening tool for high-risk HCC patients.[5,6] Another study showed that HCC screening, using both AFP and ultrasonography, appears to be of borderline cost-effectiveness or not cost-effective at all, and HCC screening with only AFP is not recommended, except when ultrasonography is either not available or of poor quality.[7] In our study, AFP was also an independent predictive factor for HCC. But the American Association for the Study of Liver Diseases (AASLD) has recommended if ultrasonography is suboptimal or not available, consideration can be given to using AFP surveillance.[25] One study from Italy also showed that ultrasonography seems to be more efficient as a screening tool.[26] However, although being an acceptable screening tool for high-risk HCC patients, ultrasonography cannot be used as first-stage screening of general elderly residents because of the following reasons: cost-benefit considerations, the result is influenced by the operator's experience, and there is not enough manpower to perform ultrasonography as a first-stage screening tool. As a result, the two-stage HCC screening design has tried to limit ultrasonography screening to members of only high-risk groups.

Platelet count is, however, a simple, cheap, and readily-available test in all laboratories. Thrombocytopenia is a valid surrogate of cirrhosis and a valid marker for the identification of individuals at high risk for HCC, especially in areas with a high prevalence of HCV.[9] In our study, we validated this again.

The incidence of HCV-related HCC has also increased in recent years.[15] In our study, the five-marker screening method had a higher sensitivity (100%), but lower PPV (3.3%) than the two-marker screening method, but the cost of the five-marker screening method is higher than the two-marker screening method. The two-marker screening could only perform 37.6% (215/572) of ultrasonography cases to detect 87.5% (14/16) HCC cases, compared with the traditional five-marker screening method. Compared with the traditional five-marker screening method,[8] two-marker screening with AFP and platelet count in our study could have a cost benefit and be more effective in detecting HCC. Therefore, under economical consideration, two-marker, two-stage screening with AFP and platelet count as first-stage screening markers of risk identification should be feasible for screening the general population, especially in HCV-endemic areas.

However, two large studies showed that the mean ages of their hepatitis B virus (HBV)- and HCV-related HCC patients were 53–55 years and 65 years,[15,27] so the mean age of HBV-related HCC patients was approximately 10 years younger than that of the HCV-related HCC patients.[15] Another study from the USA showed that the median age of HCC patients at diagnosis is 74 years, and most patients (91%) were > 65 years.[28] The healthy life expectancy (HALE) at birth (years) of females was 79.56 and 80.81 years, and for males, 73.83 and 74.57 years in Taiwan in 2000 and 2006, respectively.[29] Prolonged HALE at birth can be predicted; the incidence rate of HCC would be significantly increased with advancing age due to prolonged HALE at birth (years). Significant increases in the risk of HCC incidence are associated with increasing age.[8] Therefore, HCC screening of the elderly population should be feasible due to the increasing incidence of HCC in aged people.

However, regardless of whether the HCC prognosis after diagnosis is the same in an aged population compared with non-elderly patients, or the prognosis is poor in an aged population due to old age, we even found HCC at an early stage. An Italian study concluded that elderly patients (> 70 years) with HCC have a worse prognosis compared with non-elderly patients, but such a difference seems to be a consequence of under-treatment.[18] One study from Japan also showed that an advanced stage of HCC, not advanced age, influenced the survival rate in elderly patients.[17] In Taiwan, under meticulous preoperative assessments and postoperative care, liver resection for HCC is justified in selected octogenarians, with short- and long-term results comparable to those of younger patients.[19] Another study from Japan showed that if HCC less than 3 cm in diameter is found and liver function is well preserved, local ablation therapy or surgical treatment would ensure greater than 5 years' survival (over 60%).[5] One study showed that the majority of patients in their study presented with advanced diseases, thus reducing the chance of curative treatment.[16]

In our study, we found that early detection and prompt treatment resulted in good prognosis in an aged population. Five confirmed cases received curative treatment, that is, surgical resection and percutaneous ethanol injection, and all of our patients were still alive at 4 years. The other 11 patients received TAE or other treatments, including conservative treatment. Their 4-year survival rate was 36.4%. The overall 4-year survival rate of the HCC cases obtained from our community screening was 56.3%. Patients who had early HCC and accepted curative treatment had an excellent prognosis of survival compared with those who had advanced HCC, even in this aged population.[8,10] Therefore, the screening of HCC in an aged population would have benefits and should be encouraged.

In the follow-up study, four HCC cases were noted within 1 year after screening. As mentioned in the Results, three were non-responders second-stage ultrasonographic screening. The other case, which was negative for HCC in the ultrasonographic screening, was noted to have HCC at 7 months after screening. This follow-up study confirmed the under-estimation of sensitivities and PPV of the markers in first-stage screening.

We have just discussed HCC screening in our study, but not surveillance. We could not provide any information or suggestion of surveillance intervals in our manuscript. Based on AASLD HCC practice guidelines, the interval of surveillance should be 6–12 months in high-risk patients based on tumor doubling times.[25] In the Japan Society of Hepatology algorism, the interval of surveillance was 3–4 months in ultra-high-risk groups, including type B or C liver cirrhosis, and 6 months in high-risk groups with chronic hepatitis B, chronic hepatitis C, or cirrhosis present.[30] However, there is no large study of the interval of screening in the general population, and this is needed.

The most common limitations of community-based screening should be the response rate and representativeness of the population. National health insurance, including adult health examinations, has been provided since 1993 in Taiwan. In the study township, a total of 40 000 residents were cared for by 15 clinics, including the Health Care Center, at the time of the study. Even so, our response rates were approximately 40% of the target population, which is in agreement with previously-published studies in Taiwan.[21] In second-stage ultrasonography, our response rate was 85%. This is also in agreement to those of previous studies.[8,11] The bias of the respondent population is difficult to evaluate. However, the status was not very different in our comparison studies.[8,11,21]

Because details of the seven HCC cases identified retrospectively from data linkage were not available, only 16 confirmed cases detected in the screening were included in the survival analysis; therefore, it is quite a small sample size. Although the follow-up period was as long as 4 years, the power for detecting difference is somewhat limited, and some significant variables might be missed. However, the important finding of "early detection and prompt treatment results in good prognosis" should be significant enough to be explored under limited case numbers and power.

In conclusion, under economical considerations, AFP and platelet count should be two feasible screening markers for risk identification in elderly populations, especially in an HCV-endemic area. "Early diagnosis and prompt treatment results in good prognosis" is the goal of the secondary prevention of preventive medicine. If patients' diseases can be detected at stages early enough to undergo curative treatment, such as surgical resection and medical ablation, the prognosis should be good, even in elderly patients. HCC screening for elderly populations should, therefore, be encouraged.

http://www.medscape.com/viewarticle/737627