Background and Objective:In patients with hepatocellular carcinoma(HCC) receiving potentially curative minimally invasive therapy,autologous cytokine-induced killer(CIK) cells were used to reduce recurrence.In this study we observed the changes in serum alpha-fetoprotein(AFP) after the treatment with CIK cells to explore if AFP could serve as a marker for predicting immunotherapeutic clinical outcome.Methods:A total of 122 patients with HCC and elevated AFP(> 25 ng/mL) received a curative treatment of transcatheter arterial chemoembolization(TACE) plus radiofrequency ablation(RFA) at the Sun Yat-sen University Cancer Center.Of these patients,83 patients without residual tumor or extrahepatic metastasis and with AFP level less than 1.5 times the normal range(AFP < 37.5 ng/mL) were randomly assigned to the study group(n = 42) and the control group(n = 41).In the study group,CIK cells were transfused intravenously or via common hepatic arteries every week for at least 4 times,and the T-lymphocyte subset data before and after CIK cell infusions was examined by flow cytometry.All the two groups of patients were screened by tomography every 2 months to observe tumor recurrence.Serum AFP was collected at baseline and at different time points after treatment in parallel with radiologic response and clinical outcome.Results:Two patients in the control group were lost to follow-up after treatment.After CIK cell infusions,the downtrend of the AFP level was observed in the study group and not in the control group.There was a significant difference in the level of AFP between different time points after CIK infusions in both groups.The 1-year recurrence rate was 7.14 % for the study group and 23.1% for the control group(P = 0.044).In subgroup analysis,for patients with a slightly high level of AFP(25 ng/mL < AFP < 37.5 ng/mL) after curative TACE plus RFA treatment,the 1-year recurrence rate was 28.57% for the study group and 80% for the control group.The time to recurrence in the study group was also longer than that in the control group(mean 10.2 months vs.6.8 months).After CIK cell infusions,the percent of CD3+CD4+ T cells and CD4+ /CD8+ T cells increased from 28.1 ± 5.9% and 0.9 ± 0.3% to 32.7 ± 3.6% and 1.2 ± 0.2%(P < 0.001 and = 0.004,respectively),while the percent of CD3+CD8+ T cells decreased from 32.9 ± 8.4% to 28.8 ± 2.2%(P = 0.046).Also the percentage of patients with hepatitis B virus(HBV)-DNA content less than 1 × 103 copies/mL was 73.5% in the study group and 9.1% in the control group.Conclusions:CIK cells transfusion may reduce the level of serum AFP and anti-HBV and decrease the 1-year recurrence rate of patients with HCC after curative TACE plus RFA.Serum AFP decrease after CIK cell treatment may serve as a useful marker for predicting immunotherapy clinical outcome in patients with HCC undergone curative minimally invasive therapy. Background and Objective: In patients with hepatocellular carcinoma (HCC) receiving potentially curative minimally invasive therapy, autologous cytokine-induced killer (CIK) cells were used to reduce recurrence. In this study we observed the changes in serum alpha-fetoprotein (AFP) after the treatment with CIK cells to explore if AFP could serve as a marker for predicting immunotherapeutic clinical outcome. Methods: A total of 122 patients with HCC and elevated AFP (> 25 ng / mL) received a curative treatment of transcatheter arterial chemoembolization (TACE) plus radiofrequency ablation (RFA) at the Sun Yat-sen University Cancer Center. Of these patients, 83 patients without residual tumor or extrahepatic metastasis and with AFP level less than 1.5 times the normal range (AFP <37.5 ng / mL) were randomly assigned to the study group (n = 42) and the control group (n = 41). In the study group, CIK cells were transfused intravenously or via common hepatic arteries every week for at least 4 times, and the T-ly mphocyte subset data before and after CIK cell infusions was examined by flow cytometry. All the two groups of patients were screened by tomography every 2 months to observe tumor recurrence. Serum AFP was collected at baseline and at different time points after treatment in parallel with radiologic response and clinical outcome. Results: Two patients in the control group were lost to follow-up after treatment. After CIK cell infusions, the downtrend of the AFP level was observed in the study group and not in the control group. There was a significant difference in the level of AFP between different time points after CIK infusions in both groups. The 1-year recurrence rate was 7.14% for the study group and 23.1% for the control group (P = 0.044). In subgroup analysis, for patients with a slightly high level of AFP (25 ng / mL