ABSTRACT
Objective To establish a nomogram for overall survival rate after liver resection for primary small hepatocellular carcinoma based on SEER data and external validation of Chinese data. Methods The data of 1809 patients, registered in National Cancer Institute SEER database in 2004-2015, who underwent hepatectomy for primary small hepatocellular carcinoma were extracted as modeling group, and 158 patients with small hepatocellular carcinoma who underwent hepatectomy in Affiliated Hospital of North Sichuan Medical College from 2010 to 2017 were collected as validation group. The univariate Cox risk regression analysis, lasso regression analysis, and multivariate Cox hazard regression analysis were used to investigate the influencing factors for OS after hepatectomy in patients with small hepatocellular carcinoma. A nomogram was established based on the independent influencing factors for OS, and index of concordance (C-index), calibration curves, and receiver operating characteristic (ROC) curve were used to analyze the predictive ability of the nomogram. The Kaplan-Meier survival analysis and the log-rank test were used to investigate the difference in survival between the high- and low-risk groups. Results The multivariate Cox hazard regression analysis showed that sex (hazard ratio [ HR ]=1.22, 95% confidence interval [ CI ]: 1.05-1.41, P =0.010), Seer stage ( HR =1.51, 95% CI : 1.23-1.85, P < 0.001; HR =10.31, 95% CI : 2.53-42.04, P =0.001), tumor diameter ( HR =1.22, 95% CI : 1.06-1.39, P =0.004), vascular invasion or metastasis ( HR =1.43, 95% CI : 1.24-1.65, P < 0.001), and alpha-fetoprotein ( HR =1.33, 95% CI : 1.16-1.54, P < 0.001) were independent risk factors for OS after hepatectomy for small hepatocellular carcinoma. The modeling group had a C-index of 0.621, and its area under the ROC curve at 1, 2, and 3 years was 0.666(95% CI 0.628-0.704), 0.678(95% CI 0.647-0.708), and 0.663(95% CI : 0.635-0.690), respectively; the validation group had a C-index of 0.718, and its area under the ROC curve at 1, 2, and 3 years was 0.695(95% CI : 0.593-0.797), 0.781(95% CI : 0.706-0.856), and 0.759(95% CI 0.669-0.848), respectively. Risk stratification was performed based on the nomogram, and the Kaplan-Meier survival analysis showed that for both the modeling group and the validation group, the low-risk group had a significantly better prognosis than the high-risk group ( P < 0.01). Conclusion The model established for survival rate after liver resection for primary small hepatocellular carcinoma can predict the 1-, 2-, and 3-year OS rates and can thus be used in clinical practice in China.
ABSTRACT
Objective To establish an Early Warning System for Recurrence Scoring after Radical Resection of BCLC stage 0/A Primary Liver Cancer (PLC-EWSPRS), and to investigate its predictive value. Methods A retrospective analysis was performed for the clinical data of 232 patients with BCLC stage 0/A liver cancer who underwent radical resection in Affiliated Hospital of Chuanbei Medical College from January 2009 to January 2015, and according to the presence or absence of recurrence within 5 years after surgery based on telephone or outpatient follow-up data, the patients were divided into recurrence group with 103 patients and non-recurrence group with 129 patients. The t -test was used for comparison of normally distributed continuous data between groups, and the Mann-Whitney U test was used for comparison of non-normally distributed continuous data between groups; the chi-square test or Fisher's exact test was used for comparison of categorical data between groups. The indices with statistical significance were included in the binary logistic regression analysis to investigate the risk factors for recurrence of BCLC stage 0/A liver cancer after surgery. Two points were assigned for independent risk factors and one point was assigned for risk factors to establish the PLC-EWSPRS system. The receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) were used to evaluate the diagnostic efficiency of this system. Results Compared with the non-recurrence group, the recurrence group had significantly higher levels of aspartate aminotransferase and alanine aminotransferase (ALT) and a significantly lower level of albumin (Alb) before surgery ( Z =3.864 and 4.587, t =-5.628, all P < 0.001), as well as a significantly higher proportion of patients with positive HBsAg, capsular invasion, microvascular invasion (MVI), tumor diameter ≥5 cm, liver cirrhosis (moderate-to-severe), non-R0 resection, or death within 5 years ( χ 2 =35.539, 22.325, 13.398, 7.130, 4.312, 4.034, and 18.527, all P < 0.05). The regression analysis showed that preoperative Alb < 40 g/L (odds ratio [ OR ]=5.796, P < 0.001), preoperative ALT ≥40 U/L ( OR =3.029, P =0.002), MVI ( OR =3.981, P =0.003), positive HBsAg ( OR =7.829, P < 0.001), capsular invasion ( OR =5.357, P < 0.001), and non-R0 resection ( OR =3.048, P =0.018) were independent risk factors for recurrence of BCLC stage 0/A liver cancer within 5 years after surgery. According to the assignment criteria of the PLC-EWSPRS system, the recurrence group had the lowest score of 2 points and the highest score of 14 points, while the non-recurrence had the lowest score of 0 point and the highest score of 11 points, and the recurrence group had a significantly higher score than the non-recurrence group ( P < 0.05). The ROC curve analysis showed that the PLC-EWSPRS system had an AUC of 0.918 (95% confidence interval [ CI ]: 0.883-0.953, P < 0.001) in predicting recurrence within 5 years after surgery in patients with BCLC stage 0/A liver cancer undergoing radical resection, and subgroup analysis showed that the system had an AUC of 0.796 (95% CI : 0.695-0.896, P =0.002), 0.859 (95% CI : 0.791-0.927, P < 0.001), and 0.944 (95% CI : 0.839-1.000, P =0.044), respectively, in predicting recurrence within 5 years after surgery in patients with a low score of 0-5 points, a moderate score of 6-10 points, and a high score of 11-14 points. Conclusion The PLC-EWSPRS system has a good value in predicting the recurrence of BCLC stage 0/A liver cancer within 5 years after surgery and thus has important guiding significance for postoperative reexamination and treatment strategy for patients with BCLC stage 0/A liver cancer undergoing radical resection.