|
Abstract: Objective To construct a nomogram risk model that can be used to predict the
prognosis of hepatocellular carcinoma (HCC) patients with portal vein tumor thrombus
(PVTT). Methods Total of 370 HCC patients with PVTT in the Third Hospital of Hebei
Medical University from January 2010 to December 2019 were retrospectively involved and
divided into training set (224 cases) and test set (92 cases) according to the randomization
principle of 7∶3. The clinical data including age, gender, family history of liver cancer,
history of smoking and drinking were collected. Laboratory indicators included white blood
cell (WBC), neutrophil-to-lymphocyte ratio (NLR), hemoglobin (HGB), platelet count (PLT),
alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (TBil),
albumin (ALB), lactate dehydrogenase (LDH), gamma-glutamyltransferase (GGT), alkaline
phosphatase (ALP), creatinine (Cr), prothrombin (PTA), international normalized ratio
(INR), alpha-fetoprotein (AFP) and C-reactive protein (CRP). Univariate and multivariate
Cox regression were used to analyze the factors affecting 1-year mortality of HCC patients
with PVTT, R software was used to build and verify the nomogram mode. Receiver operator
characteristic (ROC) curve was used to evaluate the predict value of the nomogram model,
BCLC stage, MELD score, Child-Pugh score, and ALBI grade. Results Ascites (HR = 1.46,
95%CI: 1.07～1.99), upper gastrointestinal bleeding (HR = 2.54, 95%CI: 1.62～3.99), PLT ＞
100 × 109
/L (HR = 1.53, 95%CI: 1.11～2.11), ALT ＞ 50 U/L (HR = 1.41, 95%CI: 1.00～2.08),
TBil ＞ 18.8 μmol/L (HR = 1.61, 95%CI: 1.13～2.29), AFP ＞ 400 μg/L (HR = 1.49, 95%CI:
1.07～2.07) and CRP ＞ 5 mg/L (HR = 2.85, 95%CI: 1.72～4.72) were independent risk
factors for the prognosis of HCC patients with PVTT (P ＜ 0.05). The area under ROC curve
of nomogram model in model group and verification group based on the above seven factors
were 0.787 (95%CI: 0.713～0.860) and 0.840 (95%CI: 0.740～0.940), respectively, the
difference was not statistically significant (z = -0.842, P = 0.4). The area under ROC curve of
nomogram model in model group and verification group were significantly higher than those
of MELD scores (z = 4.012, P ＜ 0.01; z = 2.569, P ＜ 0.01), ALBI grade (z = 5.333, P ＜ 0.01;
z = 3.562, P ＜ 0.01), Child-Pugh score (z = 4.596, P ＜ 0.01; z = 3.056, P ＜ 0.01) and BCLC
stage (z = 5.206, P ＜ 0.01; z = 4.392, P ＜ 0.01). Conclusions The column line graph constructed
by 7 key factors, including ascites, upper gastrointestinal bleeding, PLT, ALT, TBil, AFP and CRP
was valuable in predicting the risk of prognosis in HCC patients with PVTT.
|