Basic data of HCC patients
A total of 217 patients were included in this study, including 191 (88.02%) men and 26 (11.98%) women, with a mean age of 50.12 ± 12.07 years. Among them, 194 patients (94.47%) were complicated by HBV infection, and no patient with hepatitis C was found. 47 (21.66%) patients received regular antiviral treatment (> 6 months) and 119 patients (45.16%) had active viral replication preoperatively (HBV-DNA > 1000 IU/mL). 33 (15.2%) patients had multiple tumors, mean maximum tumor diameter was 5.93 ± 3.17 cm, 67 (37.89%) patients with AFP > 400 ng/ml, 168 patients with cirrhosis, 172 (79.26%) patients with grade 0/A BCLC and 45 (20.74%) patients with BCLC grade CB. At the end of follow-up, the median duration of follow-up is 21.32 ± 14.40 months. 127 patients (58.53%) relapsed, of whom 111 patients (51.15%) relapsed within 2 years after surgery, most patients had intrahepatic recurrence (Table 1).
The threshold value and the area under the curve of the inflammation index and the TBS
The patients’ ROC curve was plotted against preoperative MLR, NLR, PLR, SII, and TBS values. The results of the ROC curves showed that MLR = 0.39, NLR = 2.63, PLR = 134, SII = 428 and TBS = 8.06 were the optimal cutoff values (Table 2). Based on the optimal cutoff value, patients with a value less than or equal to the cutoff value were assigned to the low level group, and patients with a value above the cutoff value were assigned to the high level group. The AUC of SII, PLR, NLR, MLR, and TBS were 0.643, 0.642, 0.642, 0.618, and 0.724, respectively. The sensitivity of SII, MLR, PLR, NLR and TBS was 38.6%, 32.3%, 39.4.9%, 44.9% and 38.6%, respectively, and the specificity was 85, 6%, 88.9%, 84.4%, 81.1% and 95.6%, respectively. By comparing AUC (Fig. 2), we found that TBS was superior to other inflammatory indicators in predicting recurrence in patients with HCC postoperatively. Moreover, among the four inflammatory indicators, SII, PLR, and NLR had similar prognostic value for postoperative HCC recurrence, but they were all superior to MLR.
Comparison of TBS value with maximum tumor diameter, tumor count and TTV values in prognostic accuracy
Next, we plotted ROC curves for TBS values, maximum tumor diameter and tumor counts and TTV values of all HCC patients. According to their ROC curves, the AUC of TBS, tumor diameter, tumor count and TTV values were 0.724, 0.715, 0.554, 0.719, respectively (Fig. 3). Comparing the AUC of these indicators, the predictive value of TBS for postoperative HCC recurrence was better than that of tumor diameter, tumor count, and TTV values. This advantage may be more pronounced in a large sample of patients with multiple tumors.
Univariate and multivariate analysis of disease-free survival
Univariate Cox analysis was performed for sex, age, albumin, total bilirubin, ALBI, TBS values, TTV values, ALT, AST, PT, APTT, tumor diameter, tumor count, AFP, degree of differentiation, HBV infection, HBV DNA quantification, cirrhosis, MVI, satellite nodules, BCLC stage, SII group, NLR, MLR, PLR and TBS, We found values of TBS (PP 400ng/mL (PPPPPPP 428 (P 2.63 (P 0.39 (P 134 (P 8.06 (P
Considering TBS calculated as a function of tumor diameter and tumor count, a multivariate COX risk regression analysis was performed after removing tumor size and diameter. Only MVI (HR =P= 0.005), satellite nodule (P= 0.017), stage BCLC BC (P= 0.013), NLR > 2.63 (P= 0.013), TBS > 8.06 (P
The Prognostic Value of the Combined Index Model (NLR-TBS) for HCC
The optimal inflammatory immune index (NLR) and TBS obtained in this study were combined to stratify the risk of postoperative recurrence of HCC: low-risk group (TBS ≤ 8.06 and NLR ≤ 2.63), high-risk group medium (TBS > 8.06 and NLR ≤ 2.63 or TBS ≤ 8.06 and NLR > 2.63), high-risk group (TBS > 8.06 and NLR > 2.63); According to the recurrence of the patients, ROC curves were drawn for the TBS group, the NLR group and the NLR-TBS groups (Fig. 4). The areas under the ROC curves of the TBS, NLR and NLR–TBS groups were as follows: 0.671, 0.630, 0.762, according to the area under the ROC curve, the prognostic value of NLR–TBS for postoperative recurrence of HCC was greater than the simple index of the NLR and TBS cluster. Finally, the established NLR-TBS prediction model was compared to the traditional HCC stage system (BCLC stage, TNM stage). In plotting the ROC curves, the areas under the ROC curves of the NLR–TBS model, TNM and BCLC stages were 0.762, 0.587 and 0.620, respectively (Fig. 5). The prognostic value of NLR-TBS model for postoperative liver cancer recurrence was better than that of TNM stage and BCLC stage. Survival curves at the risk of recurrence of TBS (Fig. 6A), NLR (Fig. 6B) and TBS + NLR (Fig. 6C) were plotted according to the recurrence situation of the patients.