Liver cirrhosis can cause portal hypertension with refractory ascites and variceal bleeding as well as hepatocellular carcinoma (HCC). Therefore, there is a rising patient population previously treated with transjugular intrahepatic portosystemic stent (TIPS) for portal hypertension suffering from HCC. So far a negative influence of TIPS on HCC concerning treatment options has been suspected, since due to reduced portal liver perfusion only transarterial chemotherapy (TAC) instead of additional embolization (TACE) is usually performed. Therefore, the effect of embolization, which has a higher antitumoral potency than intra-arterial chemotherapy itself, is missing. To evaluate treatment modalities in patients with TIPS and HCC we analyzed firstline treatment and overall survival (OS).
Between April 1995 and May 2011 we recruited 60 newly diagnosed HCC patients of the Liver Center of the University Hospital Freiburg who previously (>6 months) had been treated with bare metal stent TIPS implantation. For controls we analyzed 60 consecutive HCC patients without previous TIPS implantation. These patients were matched 1:1 for age (±5 years), sex, etiology of liver disease, and Child-Pugh score at the time of HCC diagnosis. In all, 51/60 patients (85.0%) in the TIPS group and 42/60 patients (70.0%) in the non-TIPS group died within the observation time. Tumor stages were assessed using the established Barcelona Clinic Liver Cancer (BCLC) classification. In both groups the majority of patients presented with BCLC stage A (48.3% and 44.3%) and BCLC stage B (30.0% and 35.7%) without statistically significant differences (P = 0.966).
TIPS patients had a median OS of 17.0 months (95% confidence interval [CI]: 10.21; 23.79) compared to 24.0 months (95% CI: 9.39; 38.61) of non-TIPS patients (P = 0.040, Fig. 1A). A multivariate Cox regression model identified multifocal hepatic tumor manifestation (hazard ratio [HR] 2.13, P = 0.012), TIPS (HR 1.74; P = 0.040), Child-Pugh B (HR 1.98; P = 0.008), and C (HR: 3.30; P = 0.004), alpha-fetoprotein (AFP) >20 ng/mL (HR: 1.94; P = 0.008), and metastasis (HR 5.20; P = 0.001) as significant independent negative predictors of OS.
Moreover, we analyzed firstline treatment in TIPS and non-TIPS patients. A majority of patients with TIPS were treated by best supportive care (BSC) and did not receive any HCC-specific treatment compared to patients in the non-TIPS group (26.8% versus 6.2%). Interestingly, 28 (46.6%) TIPS patients were treated with TAC compared to 49 (81.6%) non-TIPS patients who had been treated with TACE (P = 0.002, Fig. 1B). None of the TIPS patients developed severe hepatotoxicity as a possible reason for impaired OS. No statistical differences concerning surgical approaches, percutaneous therapies (radiofrequency ablation [RFA]) and sorafenib application, were found.
In conclusion, our findings indicate that TIPS patients have limited therapeutic possibilities concerning HCC-specific therapies, resulting in impaired OS. Especially, transarterial chemotherapies are less often administered in patients with TIPS. Therapy strategies in TIPS patients with HCC should be reassessed, since treatment options are expanded: (super)selective TACE yttrium-90 radioembolization or percutaneous ethanol injection in combination with TACE might be alternative approaches. Therefore, prospective studies are needed to determine the effectiveness and the safety of therapeutic approaches using embolization in patients with TIPS and HCC and to establish treatment guidelines for HCC in these patients.
Eva Knüppel, M.D.1*
Dominik Bettinger, M.D.1*
Wulf Euringer, M.D.2
Robert Thimme, M.D.1
Martin Roessle, M.D.3
Hans Christian Spangenberg, M.D.1
Michael Schultheiβ, M.D.1
1Department of Medicine IIUniversity Hospital FreiburgFreiburg, Germany
2Department of Diagnostic and Interventional RadiologyUniversity Hospital FreiburgFreiburg, Germany
3PraxisZentrum für Gastroenterologie und EndokrinologieFreiburg, Germany
*These authors contributed equally.