The Indication Of Hepatectomy For Multiple Hepatocellular Carcinoma And The Preoperative Predictors Of Short-term Survival And Microvascular Invasion

Background and aims:Hepatocellular carcinoma (HCC) is the third leading cause of cancer mortalityworldwide, with the highest rate reported in East Asia. Multiple lesions are detectedin about40%of HCC patients. Hepatic resection is the mainstay of curativetreatment. The indications for surgery in patients with multiple HCC have not yetbeen established. The aims of these studies were to identify the criteria of surgicalresection for multinodular hepatocellular carcinoma and the preoperative predictorsof short-term survival and microvascular invasion.Methods:We retrospectively analyzed patients with multinodular HCC and liver function ofChild-Pugh A who underwent surgical resection or simple TACE as initial treatment.Continuous data were expressed as mean±SD and were compared by unpaired t test.Categorical data were compared by χ2test or Fisher's exact test. The survival curveswere analyzed by the Kaplan-Meier method and log-rank test. The risk factors wereassessed by Logistic regression or Cox regression analysis. Propensity scoresmatching model was used to control selection bias between different groups. A seriesof receiver operating characteristics (ROC) curves were used to identify the cutoffvalues with optimal discriminatory ability. The factors with a P-value less than0.1inunivariate analysis were included in the multivariate analysis. A P-value <0.05wasconsidered statistically significant. All statistical processing was performed by SPSS18.0(SPSS Inc., Chicago, IL, United States) or SAS9.0.Results:Part1Hepatic resection improves the survival of multiple hepatocellularcarcinoma patients within the UCSF criteria compared with transcatheterarterial chemoembolization In198patients within BCLC intermediate stage,the1-,2-,3-year overall survival rates were86%,63%,50%respectively in SRgroup and85%,52%,26%respectively in TACE group (P=0.002). Tumors withinUCSF criteria (HR=0.324, P<0.001) and treatment strategy (TACE vs. SR,HR=1.698, P=0.041) were proved to be independent predictors of survival. Afterperforming scores matching model analysis, resection brought better survival than TACE for patients within UCSF criteria (1-year:83%vs.63%,3-year:76%vs.16%,P<0.001) while resection and TACE had similar effect for patients exceeding UCSFcriteria (1-year:77%vs.85%,3-year:37%vs.22%, P=0.794).Part2Patients with Multiple Hepatocellular Carcinoma within the UCSFCriteria Have Similar Outcomes after Curative Resection to Patients within theBCLC Early Stage Criteria In162patients who underwent potentially curativeresection as initial treatment, the median overall survival was38.3months (range,3-80months), while the median disease-free survival was18.6months (range:1-79).The UCSF criteria were shown to independently predict overall and disease-freesurvival. In patients within the UCSF criteria,3-year overall and disease-freesurvival were significantly better than in those exceeding the UCSF criteria (68%vs.34%and54%vs.26%, respectively; both P <0.001). There were no significantdifferences in3-year overall and disease-free survival between patients within theUCSF criteria but exceeding the BCLC early stage, and patients within the BCLCearly stage (71%vs.66%, P=0.506and57%vs.50%, P=0.666, respectively).Tumors within the UCSF criteria were associated with a lower incidence ofhigh-grade tumor (P=0.009), microvascular invasion (P=0.005),3-month death(P=0.046), prolonged Pringle's maneuver (P=0.005), and surgical margin <0.5cm(P<0.001) than those exceeding the UCSF criteria. Tumors within the UCSF criteriabut exceeding the BCLC early stage had similar invasiveness and surgical difficultyto those within the BCLC early-stage criteria.Part3Preoperative predictors of short-term survival after hepatectomy formultinodular hepatocellular carcinoma In162patients, the1-year mortalitywas14%. Independent prognostic risk factors of1-year death included prealbumin<170mg/L (hazard ratio (HR)=5.531, P<0.001), alkaline phosphatase (ALP)>129U/L (HR=3.252, P=0.005), alpha-fetoprotein (AFP)>20μg/L (HR=7.477,P=0.011), total tumor size>8cm (HR=10.543, P<0.001), platelet count <100×109/L(HR=9.937, P<0.001), and gamma-glutamyl transpeptidase (GGT)>64U/L(HR=3.791, P<0.001). The scoring model had strong ability to predict1-yearsurvival (area under ROC was0.925, P<0.001). Patients with score≥5hadsignificantly poorer short-term outcome than those with score<5(1-year mortality:62%vs.5%, P<0.001;1-year recurrence rate:86%vs.33%, P<0.001). Patients withscore≥5had greater possibilities of microvascular invasion (P<0.001), poor tumordifferentiation (P=0.003), liver cirrhosis with small nodules (P<0.001), and intraoperative blood transfusion (P=0.010).Part4Preoperative predictors of microvascular invasion for multinodularhepatocellular carcinoma After supplementing104patients who underwentpotentially curative resection, the total sample size increased to266. The patientswith MVI had a significantly poorer overall and recurrence-free survival than thosewithout MVI (survival rates:1-year:86%vs.71%,3-year:58%vs.16%;recurrence-free survival rates:1year:69%vs.12%;3year:48%vs.12%; both P<0.001). Multivariate analysis showed that alpha-fetoprotein (AFP)>400μg/L(OR=3.732, P=0.016), gamma-glutamyl transpeptidase (GGT)>130U/L (OR=19.779,P<0.001), total tumor diameter>8cm (OR=5.545, P=0.010) and tumor number>3(OR=11.566, P=0.007) were independent predictors of MVI. After constructing thescoring analysis, the incidence of MVI in patients with high score (≥3) wassignificantly greater than those with low score (<3,64.1%vs.10.9%, P<0.001). Theoverall and recurrence-free survival were significantly poorer in patients with highscore (both P <0.001).Conclusion:1. In patients with multinodular HCC in BCLC intermediate stage andwell-preserved liver function, surgical resection achieved better survival outcomethan TACE when patients were within UCSF criteria. Multiple HCC patients withinthe UCSF criteria benefit from curative resection. Measured expansion of curativetreatment is justified.2. For multiple HCC patients with Child-Pugh A liver function, ALP, GGT,platelet, prealbumin, AFP and total tumor size can be enrolled in preoperativeestimation. Resection should be indicated cautiously for patients with score≥5.3. AFP>400μg/L, GGT>130U/L, total tumor diameter>8cm and tumornumber>3were preoperative predictors of MVI for multinodular HCC. For patientswith high risk and well-preserved liver function, anatomic resection should beconsidered.