The Kinetics Of AFP After Liver Transplantation For Hepatocellular Carcinoma And Its Clinical Significance

Hepatocellular carcinoma (HCC) is the second most common malignant tumor in terms of mortality in China. The prognosis of HCC is poor for that patients are often at late stage of HCC at diagnosis and/or suffer from liver cirrhosis simultaneously. The clinical application of liver transplantation has brought hope for many HCC patients. Now the most important factor affecting the prognosis after liver transplantation is the high rate of tumor recurrence. So it is of great importance and interest to find ways to predict the tumor recurrence after transplantation, besides making more strict indication. Alpha-fetoprotein(AFP) is the most frequently used tumor marker for HCC in diagnosis and monitoring recurrence after treatment. Tumor maker kinetics has been used to predict the outcome in many malignancies after treatment, especially after chemotherapy. Whether the AFP kinetics plays a role in predicting the recurrence of HCC after transplantation is still unknown so far?[Objective]The study is to explore the kinetics of AFP after liver transplantation and evacuate the methodology and the clinical significance of its utility.[Methods] The AFP decline process of 61 HCC patients were studied retrospectively who had underwent liver transplantation between April 2001 and September 2005 in Liver Cancer Institute of Zhongshan Hospital, Fudan University. The eligibility criteria for the patients was: 1.elevated pre-transplantation AFP, 2. a minimum of 2 elevated AFP values after transplantation, 3. alive for at least 3 months after transplantation, 4. complete follow-up document. The patients were divided into two groups according to tumor recurrence or recurrence-free during the 2-year follow-up. The biological half life (t_1/2) was analyzed in the recurrence-free group with a method of semi-logarithmic linear regression. Calculate Apparent Half Life (AHL) using the two end points of AFP within one month after transplantation to evacuate its sensitivity and specificity for recurrence. The cutoff value is determined by the receiver operator characteristic curve (ROC curve). The recurrence group was further divided into two groups according to whether or not AFP could return to reference range after transplantation. The progression-free survival was compared between the two subgroups. Apparent half life (AHL) calculated using the two ends of AFP values within one month and any two consecutive AFP values serially was evaluatedrespectively in each patient in the two subgroups. Formulae: AHL = 0.693×(t1-t2)/(lnCt1-lnCt2).[Result] AFP values of 91.3%(21/23) patients in the recurrence-free group fit the semi-logarithmic linear regression satisfactorily (α<0.1), the average coefficient was 0.145 ( 95% range : 0.08 , 0.21). The corresponding t_1/2 was 4.8d in average (95% range: 3.3, 8.6 d). We find no significance between pre-treatment AFP concentration and t_1/2 by the criteria of pre-treatment AFP>1000 or >10000 ng/ml (p>0.1). With the cutoff value of 7 , 8 , 9 d, the sensitivity and specificity of AHL for recurrence calculating using the two end points of AFP within one month after transplantation were 13.3% and 100%, 13.3% and 96.8%, 16.7% and 90.3% respectively. The progression-free survival (90% within 3 months) for patients whose AFP could not normalize after transplantation was significantly shorter than those whose AFP could (90% beyond 5 months). 80% (8/10) had their AFP rebound within two months after transplantation. There is no significant difference in the decline patterns between the recurrence and recurrence-free groups if their AFP could normalize after transplantation. Calculating using two end points of AFP within one month after transplantation, the sensitivity and specificity of AHL for predicting that AFP could not return to normal (residual tumor) were 30.0% and 98.4% ,30.0% and 96.1% , 40.0% and 92.2% respectively With the cutoff value of 7 , 8 , 9 d. Calculating using the two consecutive AFP values serially, the sensitivity, specificity and the lead time of AHL for predicting residual tumor were 100% , 95% and 1～2w, 100% ,95% and 1～3w, 100% , 90% and 1～3w, respectively With the cutoff value of 7 , 8 , 9 d.[Conclusion] AFP declines in a first-order rate of clearance after liver transplantation with an average half life of 4.8 d. The patients whose AFP can not return to reference range after liver transplantation could be diagnosed as having residual tumor. The significance of monitoring AFP decline is to find those patients whose AFP could not normalize after transplantation, but it has no predictive value for those recurrence patients whose AFP can normalize. The methodology of AHL can predict residual tumor. AHL can predict residual tumor at one month after transplantation with sensitivity 40% and specificity 92.2% , when cutoff value is set at 7d. Calculating AHL using the two consecutive AFP values serially, With the cutoff value of 8d, can predict (100%) residual tumor with a lead time of 1～3 weeks before a rebound of AFP.