Triple Strategies Of Sorafenib Combined With Chemoembolization And Radiofrequency Ablation For Large, Unresectable Hepatocellular Carcinomas:a Pilot Study

BACKGROUND and PURPOSE:Liver cancer is the fifth most common cancer in the world and the second leading cause of cancer death. Hepatocellular carcinomas (HCCs) represent the major histological subtype, accounting for70-85%of the total liver cancer burden. Due to the elevated prevalence of chronic hepatitis B virus (HBV), HCCs have a high incidence in china. Hepatic resection has been considered a likely curative treatment for HCCs. However, only20%of patients with HCCs are candidates because of poor hepatic reserve, tumor multicentricity or large tumor size at the first diagnosis. According to the Barcelona Clinic Liver Cancer staging system (BCLC), transcatheter arterial chemoembolization (TACE) is the first option for BCLC-B hepatocellular carcinoma, and administration of Sorafenib is the first option for BCLC-C hepatocellular carcinoma. Because of the high heterogeneity of the HCCs, it is impossible to either gain a good result of complete chemoembolization or well preserve the liver with only single TACE treatment, especially for the large hepatocellular carcinomas(the maximum diameter of the tumor lesion>5cm). Some studies have reported that Radiofrequency ablation (RFA) combined with TACE can produce a1-year complete response (CR) rate of 90-100%for patients with tumors larger than5cm, which was equivalent to the efficacy of treatment of3cm lesions, and there were no increasing risks. But even TACE combined with RFA is still a local treatment, which can’t prevent metastasis and recurrence efficiently, and may fails improving long-term effects eventually, especially for the HCCs with microvascular invasion or portal vein tumor thrombus. Studies have shown that Sorafenib is a dual tyrosinkinase inhibitor with activity against the vascular endothelial growth factor receptor (VEGFR) and Raf kinase, which can suppress the increased serum levels of angiogenic factors induced by the local treatment from hypoxic cells at the periphery of the treated lesion. The preliminary result of the phase II trial (START, a phase II trial of Study in Asia of the Combination of TACE with Sorafenib in patients with HCCs) had shown that combination of TACE with Sorafenib was safe and efficient in improving survival rate. But, until today, there is still no research reported about the triple combination of Sorafenib, TACE and RFA, and there are no integrated and standardized triple strategies using them. Given the superposition effect to the liver function by combination of Sorafenib, TACE and RFA, the safety and applicability of these triple strategies are yet to be validated. Nowadays, there is a debate on when to use Sorafenib as combined with local interventional treatment, because some experts tend to think that the antiangiogenesis effect of Sorafenib takes certain times and it’s better to use Sorafenib before the local treatment (RFA), but other experts consider the early use Sorafenib may reduce the effect of TACE. In our study, we aimed to make a preliminary research of the two different triple strategies of sorafenib combined with chemoembolization and radiofrequency ablation (starting administration of sorafenib after TACE combined RFA, or starting administration of sorafenib before TACE combined RFA) for large, unresectable hepatocellular carcinomas, to observe its applicability, safety and preliminary efficacy, providing clinical evidences to support the future multicenter large-scale randomized clinical trials.METHODS:This research was a single center prospective randomized controlled clinical trial, started in November2008, recruiting patients of advanced stage HCCs(BCLC stage B or C; Child-pugh class A or B; the maximum diameter of the HCC>5cm; the number of tumor lesions≤3) according to the inclusion and exclusion criteria, who agreed to receive the combination treatment of sorafenib with TACE and RFA. All the patients recruited were randomly divided into two groups receiving different triple strategies:Group A, administration of Sorafenib after the TACE combined RFA; Group B, administration of Sorafenib1-2weeks before the TACE combined RFA. To avoid possible adverse events, we treated patients with intermittent method using Sorafenib between local therapy sessions and suspended temporarily during TACE or RFA. The termination, suspension and reduction of Sorafenib were conducted following the specific schedule made before this research. Regular follow-up started in all the patients after the first treatment. In hospital, laboratory tests including a liver function test, a complete blood count were closely monitored, in case of occurring severe complications. After discharge, a routine blood tests continued every4weeks for observing any possible adverse events. The first radiographic examination should be conducted4weeks after the first TACE combined RFA, then every8-12weeks thereafter, to evaluate the local efficiency of this combined treatment. Follow-up TACE or RFA or TACE combined RFA should be done if necessary when the appearance of new lesions or nodular, irregular enhancement around or within the ablation zone showed in CT or MRI images. All the patients were followed up until death or2013-2-21. The primary objective of the study was to assess the safety profile of these two different strategies. All the adverse events (AEs) were monitored according to the National Cancer Institute common terminology criteria for adverse events (CTCAE) version3.0, and were correlated to the post-procedure complication or a drug-related adverse effect. The AEs of3/4stages leading to the termination, suspension and reduction of Sorafenib were recorded, so was the time when this AEs occurred. The incidents of AEs were compared between the two groups using Fisher’s exact probability method. The second objective of the study was to assess the efficiency profile of these two different strategies, including technical effectiveness rate (Technical effectiveness was defined as complete devascularization of the primary lesions during the arterial phase), objective reaction rate (ORR)(ORR is defined as the proportion of patients with confirmed CR or confirmed PR according to mRECIST, relative to all randomized patients), overall survival (OS) and the time to tumor progression (TTP). The comparison between the efficiency rates of these two groups were conducted using Fisher’s exact probability method. Kaplan-Meier method was used in the analysis of survival curves, with the comparison of OS and TTP between the two groups carried out by Log-rank test. Significant prognostic factors were assessed by using univariate analysis with all the patients. All statistical analyses were performed using SPSS. All reported P values are2-sided, with P<.05considered statistically significant.RESULTS:Until the end of follow up (date:2013-02-21), The A group and B group all got15eligible patients recruited respectively, total30patients, of whom23patients died (A group11cases, B group12cases) and7survived, during a follow-up of9-49months. The average diameter of HCC lesions was8.96±3.071cm (5.1-16.5cm)(Group A8.81±3.160cm, Group B9.10±3.082cm, P=0.803). There were11cases of BCLC C stage HCC in group A, and10cases of BCLC C stage HCC in group B (73.3%vs66.7%, P=0.999). Patients in group B started the first TACE in10.73±2.120days (8-16days) after starting administration of Sorafenib. The period between first TACE and starting administration of Sorafenib in group A was8.67±2.610days (6-16days), and patients in group B resumed administration of Sorafenib in10.00±1.964days (8-14days), with no significant difference bwteen these two goups (P=0.125). The results of safty analysis:There was no therapy related death occurred in this study; No severe post-procedure complication occurred. One case of biloma was observed1month after the first RFA. This was successfully treated by antibiotic therapy and percutaneous drainage. Three minor complications (hydrothorax (n=1) and subcapsular hematoma (n=2)) developed after the first RFA, and all of these were self-limited without treatments. Hand-foot skin reaction (53.3%VS66.7%, P=0.710), anepithymia (53.3%VS46.7%, P=0.999), fatigue (53.3%VS40.0%, P=0.715), and diarrhea (33.3%VS46.7%, P=0.710) were the most common AEs after the use of Sorafenib in patients of goup A and B, the incidences of which were not statistically different between these two groups. Grade3/4toxicities observed in group A and group B were HFSR (Grade3,13.2%VS6.7%, P=0.999), diarrhea (Grade3,13.3%VS6.7%, P=0.999), and gastrointestinal hemorrhage (Grade4,0.0%VS6.7%, P=0.999), there was no statistical difference between two groups. Only1patient in group B with Grade4AE was withdrawn from administration of Sorafenib. The dose of Sorafenib was suspended temporarily or reduced in these patients of Grade3AEs (n=3). At the late stage of the treatment, Sorafenib was withdrawn from24of30patients because of the deterioration of ECOG performance status and liver functions due to the cirrhosis or tumor progression. The rates of withdrawn due to the certain reason between the two groups were similar (due to decompensated liver cirrhosis, Goup A40%(6/15) VS Goup B20%(3/15), P=0.427; due to tumour progression, Goup A40%(6/15) VS Goup B60%(9/15), P=0.466), so was the maintenance time of Sorafenib(Group A23.13±11.600months (6-45months), Group B24.27±10.613month (4-41months), P=0.782). To sum up, the safety profiles were similar between these two groups no matter when Sorafenib was used (before TACE combined with RFA). The overall safety profile of grade3/4AEs in this study (HFSR10%, diarrhea10%, gastrointestinal hemorrhage3.3%) was similar to the result of SHARP study (Sorafenib Hepatocellular Carcinoma Assessment Randomized Protocol).The results of efficacy analysis:Technical effectiveness was achieved in12patients (40.0%) at the first CT check, with7patients (40.0%) in Group A and5patients in Group B (46.7%). Using the Fisher’s exact probability method, the technical effectiveness rates between these two groups were no statistically different (P=0.710). The ORR of6months after treatment started was40.0%in group A (CR2/15, PR4/15), and33.3%in group B (CR1/15, PR4/15), there was no significantly different (P=0.999). The mOS of patients in group A was31months (95%CI:24.295～37.708months), the overall survival rates of12months,24months and36months were92.9%,64.3%, and30.1%, respectively. The mOS of patients in group B was32months (95%CI:23.705～40.295months), the overall survival rates of12months,24months and36months were86.7%,72.7%and27.3%, respectively. There was no significantly difference of the OS between patients of two groups (x2=0.050, P=0.822). The mTTP of patients in group A was18months (95%CI:16.201-19.799months), the tumor progression free rates of6months,12months and24months were93.3%,65.0%and28.9%, respectively. The mTTP of patients in group B was19months (95%CI:8.647～29.353months), the tumor progression free rates of6months,12months and24months were93.3%,66.7%and37.5%, respectively. There was no significantly difference of the TTP between patients of two groups (x2=0.000, P=0.988). Univariate analysis identified serum AFP level (P=0.040), maximum tumor diameter (P=0.000), portal vein involvement or BCLC stage (P=0.011), and number of lesions (P=0.006) as predictive factors for overall survival of this kind of HCCs in our study. Child-pugh class A or B before treatment didn’t significantly affect the overall survival time (P=0.060), as shown in univariate analysis. Using Cox regression analysis for overall survival (Method:Forward Stepwise, Enter:P≤0.05, Removal:P>0.10), the multivariate analysis only identified maximum tumor diameter (P=0.001, BB=9.954(95%CI:2.492-39.763)) and portal vein involvement or BCLC stage (P=0.014, BB=4.887(95%CI:1.373～17.399)) as independent prognostic factors for the combined treatment. HCCs which were more than10cm or in BCLC stage C (portal vein involvement) indicated poor prognosis.CONCLUSIONS:Two different combined strategies using Sorafenib after TACE combined RFA or using Sorafenib before TACE combined RFA are both feasible and safe in treating unresectable hepatocellular carcinomas large than5cm. These two different strategies treating recruited patients in our study had similar safety profile. The tolerance of Sorafenib showed no difference between patients using these two strategies. Administration of Sorafenib before local treatment didn’t improve the short-term and the long-term efficiency compared to using Sorafenib after TACE combined RFA. If useing Sorafenib before local treatment could cast a waste of medical resources needs further discussion. This study showed that Sorafenib combined with transcatheter arterial chemoembolization and radiofrequency ablation could be both safely implied in Child-pugh A and Child -pugh B cirrhosis patients, and patients with HCC lesions more than10cm might not be the most eligible population for these triple strategies. Due to the limited cases in our study, the results still need to be proved by the following large-scale randomized clinical trials.