Page 112 - Read Online

Chen et al. Hepatoma Res 2019;5:25 I http://dx.doi.org/10.20517/2394-5079.2019.12 Page 3 of 8

Identification of hepatocellular carcinoma
One of the main aims of this study was to determine the incidence of HCC in this cohort. Diagnosis of HCC
was ascertained based on cytology or histology reports. We also included subjects with at least two findings
from the AFP, liver US, computer tomography (CT), magnetic resonance imaging (MRI), and angiography
tests indicating they were positive for HCC . In addition, disease progression based on one imaging-
[16]
modality study was accepted as a diagnosis of HCC.

In addition to the medical records maintained in this clinic, the study subjects had linked records in the
Cancer Registration Database of Chang Gung Memorial Hospital. This database contains all information
recorded about all cancers diagnosed or treated at this hospital since 1987. We linked our cohort with this
database to retrieve all HCC cases identified between 1987 and 2007 in other departments.

The subjects’ national identification numbers were also linked with our national mortality database, which
has been maintained by the Statistics Office of Taiwan’s Department of Health since 1985. Cause of death
was classified using the ninth revision of the International Classification of Diseases, Injuries and Causes of
Death (ICD-9, World Health Organization, 1977). We matched our cohort with this database between 1985
and 2007. For patients lost to follow-up in our hospital, the national mortality database was the only available
source of information regarding diagnosis of HCC.

Classification of HCC according to follow-up schedule
The patients in this study differed with respect to the point at which HCC was detected during their follow-
up schedule, and they were therefore classified into three groups. The first group comprised patients with
HCC identified within one year of last follow-up (regular follow-up group). The second group comprised
patients who were lost to follow-up for more than one year but returned with HCC, and those for whom
HCC was identified upon initial presentation (out-of-schedule group). The third group comprised patients
who were completely lost to follow-up, but who were identified as having died from HCC based on the
national mortality database (lost group).

Statistical analysis
To compare the characteristics of the three groups, the χ test, Fisher’s exact test, Student’s t-test, linear-
2
by-linear association, Pearson’s correlation, and ANOVA were used as appropriate. Survival analysis was
done using the Kaplan-Meier actuarial curve method with the log-rank test. Univariate and multivariate
analyses were performed to identify factors associated with mortality and HCC or development of cirrhosis.
Variables found to be significant in the univariate models were tested in a multivariate setting using the Cox
proportional hazards regression model. All statistical analyses were performed using SPSS software (version
22; SPSS Inc., Chicago, IL) and P < 0.05 was considered significant.

RESULTS
During the study period, 238 HCC cases were identified (1.5% or 156.2/100,000 person-years, Table 1). Male
sex (89.1% in the HCC group vs. 58.6% in the non-HCC HBsAg carriers, P < 0.001), age at enrollment (43.32
vs. 34.96 years, P < 0.001), prevalence of liver cirrhosis (70.2% vs. 2.7%, P < 0.001), and initial ALT (123.72 vs.
63.22 U/L, P < 0.001) and maximal ALT levels (310.73 vs. 130.63 U/L, P < 0.001) were significantly associated
with HCC [Table 1 and Figure 1]. In the HBsAg carriers without HCC, 72.6% had initial ALT levels lower
than 2 × the upper limit of normal (ULN; Figure 2A) and 69.3% had a maximal ALT level lower than 5 ×
ULN [Figure 2B]. In contrast, 63.5% HCC patients had an initial ALT level greater than or equal two 2 ×
ULN and 71.5% had a maximal ALT level greater than or equal to 5 × ULN.

Of the patients with HCC, 130 (54.6%) cases were identified during regular follow-up, 55 (23.1%) were in the
out-of-schedule group, and 53 (22.3%) cases were lost to follow-up [Table 2]. The diagnosis was established