Page 13 - Read Online

Page 2 of 12 Nishioka et al. Hepatoma Res 2018;4:1 I http://dx.doi.org/10.20517/2394-5079.2017.46

Conclusion: Distinct sub-classes of HCC associated with different molecular features and survival outcomes can
be detected with statistical confidence in a Pacific Island cohort. Molecular classification signatures and other
predictive markers for HCC that are valid for all patient populations are needed to support multi-center efforts to
develop targeted therapies for HCC.

Keywords: Hepatocellular carcinoma, alpha-fetoprotein, survival, gene expression, enrichment analysis, molecular
signature, Asia-Pacific, Hawaii

INTRODUCTION
Hepatocellular carcinoma (HCC) remains a leading cause of cancer-related mortality worldwide despite
[1,2]
current extensive knowledge about its preventable risk factors . The highest incidence rates of HCC are in
areas with endemic hepatitis B virus (HBV) infection such as China and Sub-Saharan Africa, however HCC
incidence has been increasing in Oceania, Europe, and the United States (US) owing to the rising prevalence
[1,3]
of other HCC risk factors such as hepatitis C virus (HCV) infection and steatohepatitis . In the US alone,
HCC diagnoses have tripled since 1975, and with a 5-year survival rate as low as 12%, HCC is fast becoming
[3]
a leading cause of cancer-related mortality in this region . Reflecting its predominantly Asian and Pacific
Islander demographic, Hawaii has one of the highest incidence rates of liver cancer in the US, with an age-
adjusted incidence rate of 11.0 per 100,000 that is considerably higher than the overall US rate of 7.6 per
[4]
100,000 . Given that the distribution of HCC risk factors in Hawaii differs from that of both Asia and the
[5-7]
continental US , studies involving cohorts from Hawaii may provide additional insights into the disease.
HCC is renowned for its genomic and molecular diversity. Recent attempts at HCC molecular sub-
classification have produced multiple, sometimes orthogonal, classification systems that associate with
various clinical, histological, and molecular features [8-12] . The molecular diversity of HCC makes targeted
therapy challenging, since it dilutes any individual therapeutic target within the patient population, leading
to weaker overall benefit in conventional clinical trials [13-15] . Consequently, it is not surprising that among
HCC clinical trials to date, all molecularly-specific agents have failed, and only multi-targeting agents such as
[16]
sorafenib have shown efficacy . A robust molecular sub-classification system for HCC could enable clinical
trials to enrich study cohorts according to tumoral expression of targeted molecular pathways [11,15-17] . In fact,
this may be the most important next step in advancing patient-individualized treatment of HCC. It would
therefore be prudent to validate HCC sub-classification systems across many different patient populations
worldwide.

Serum alpha-fetoprotein (AFP) measurement has been used extensively for HCC screening and diagnosis,
despite being associated with a limited diagnostic sensitivity of approximately 66% [18,19] . Possibly explaining
this limited sensitivity, different molecular sub-classes of HCC have been associated with different degrees
of AFP production [10-12] . Clinically, differences in AFP production have also been associated with gross
and histopathologic tumor differences, including differences in tumor size, multinodular appearance, and
[20]
vascular invasion . AFP may also be directly involved in tumor pathogenesis through its involvement
in several mitogen and anti-apoptotic pathways, as well as potentially by exerting paracrine effects on
immune and other non-tumor cells [21,22] . Given these associations, AFP could have significant value beyond
that of a diagnostic marker. While several molecular classification systems for HCC have been associated
with differences in AFP levels across their respective sub-classes [10,11] , these associations along with the
classification systems themselves are in need of further validation across many different population cohorts.
Since Hawaii has a unique and diverse patient population, we assessed the feasibility of applying HCC
molecular classification systems derived from other patient populations to those in Hawaii and examined the
relationship of AFP and other clinical parameters to the transcriptomic features of HCC.