Diarrhea predicts a positive response to sorafenib in patients with advanced hepatocellular carcinoma


  • These authors contributed equally.

  • Potential conflict of interest: Hans Christian Spangenberg: teaching and speaking for Bayer Healthcare. The other authors have no conflict of interest to disclose.

To the Editor:

With great interest we appreciated the recent article in HEPATOLOGY by Iavarone et al.,1 who provided indirect evidence for the efficacy of sorafenib in a multicenter field-practice study in Italy finding a median overall survival (OS) of 10.5 months compared with 10.7 months in the SHARP trial.2 In addition, the authors set out to determine the differences in OS concerning Barcelona Clinic Liver Cancer (BCLC) stage, observing a significantly higher survival rate in patients with BCLC-B compared with BCLC-C (26.0 versus 8.4 months, P < 0.001). They reproduced the data from the SHARP trial in terms of OS and adverse events. The most common adverse events in their study were fatigue, weight loss, hand-foot syndrome, and diarrhea. Interestingly, recent studies have indicated that adverse events may be of prognostic and predictive importance in patients treated with sorafenib.3 Vincenzi et al.4 reported that early hand-foot reaction may be a positive predictive factor for response to sorafenib showing a prolongation of the time to progression (TTP) in patients with confirmed hand-foot syndrome.

We retrospectively analyzed 112 patients with advanced hepatocellular carcinoma (HCC) treated with sorafenib with respect to OS, TTP, and prognostic importance of the most common adverse events. We found a median OS of 9.6 months and a median TTP of 3.9 months. The median duration of sorafenib treatment was 3 months and 75 patients (66.4%) reported adverse events. The most common ones were diarrhea in 36/112 patients (32.1%), hand-foot syndrome (15.2%), fatigue (13.4%), and loss of appetite (7.2%). Multivariate Cox regression models revealed BCLC stage as a negative independent prognostic factor (hazard ratio [HR]: 3.08; P = 0.001), while diarrhea was a positive independent prognostic factor (HR: 0.41; P = 0.001). Patients with diarrhea had a significantly longer median OS of 14.1 months as compared with 7.1 months in patients without diarrhea (P = 0.011; Fig. 1), while hand-foot syndrome was not associated with OS or TTP (P = 0.15 and P = 0.59, respectively).

Figure 1.

Overall survival stratified to diarrhea during sorafenib treatment. Patients with diarrhea during sorafenib treatment had a longer median OS (14.1 months, 95% confidence interval [CI]: 5.84; 22.62) as compared with patients without diarrhea (7.1 months, 95% CI: 5.52; 8.61).

In comparison with Iavarone et al. and the SHARP trial, we were able to reproduce their data in our retrospective study. Interestingly, we were able to show for the first time that diarrhea is associated with prolonged OS and may be an independent positive prognostic factor. These data suggest that patients with diarrhea during sorafenib therapy should receive sufficient symptomatic therapy in order to prevent early termination of sorafenib treatment.

Dominik Bettinger*, Michael Schultheiβ MD*, Eva Knüppel*, Robert Thimme MD*, Hubert E. Blum MD*, Hans Christian Spangenberg MD*, * University Hospital Freiburg, Department of Medicine II, Freiburg, Germany.