Page 13 - Read Online

Page 8 of 13 Tong et al. Hepatoma Res 2019;5:36 I http://dx.doi.org/10.20517/2394-5079.2019.005

Yes 129 17.7 4.10-38.4 0.5418
No 35 17.9 8.10-36.9

TGR: tumor growth rate; HBV: hepatitis B virus; HCV: hepatitis C virus; HBeAg: hepatitis B e-antigen; AFP: alpha-fetoprotein

Figure 1. Regression tree analysis: predictors of hepatocellular carcinoma tumor growth rates. Each node is based on available data for
each predictive variable presented. TGR is reported as a median. TGR: tumor growth rate; AFP: alpha-fetoprotein

per month and the median TGR in the faster group was 38.9% per month. As illustrated in Figure 3, the
recurrence-free survival in patients who received OLT, surgical resection, or RFA was significantly longer in
patients with slow TGRs for each treatment modality (P = 0.029). Patients who received OLT who had slow
TGRs had the longest recurrence-free survival. Those HCC patients who received surgical resection or RFA
had similar survival rates in both the slow and fast TGR groups. The poorest recurrence-free survivals were
observed in the TACE treated or supportive care patients with fast TGRs.

DISCUSSION
Previous reports have utilized MRI or CT to find potential biomarkers to predict clinical outcomes in
patients with HCC. Using MRI, one report showed that patients with fat-containing HCC had less tumor
progression, less distant metastases, and a longer time to tumor progression when compared to patients
[16]
with non-fat containing HCC . Another report showed that patients with complete tumor encapsulation
on MRI had lower AFP levels, an absence of vascular invasion, more patients in Child-Pugh class A, and
[17]
significantly longer survivals . Further, the authors also noted that the rates of downstaging and eventual
liver transplantation were significantly higher. However, recognition of these imaging features depends on
the expertise of the interpreting radiologist and may be challenging to implement as a practical clinical
tool. Nevertheless, efforts to standardize imaging reporting (i.e., Organ Procurement and Transplantation
Network/United Network for Organ Sharing (OPTN) and Liver Imaging Reporting and Data System
(LI-RADS) criterion) may allow incorporation of additional important imaging biomarkers for tumor
prognosis [18,19] . As tumor size is already a basic measure reported with all detected tumors, the calculation of
TGR is feasible when serial imaging is available and, thus, may be considered as another potential imaging
biomarker.