We agree that baseline liver function correlates with mortality, and we reported in detail no significant differences in baseline characteristics between groups; in particular, Model for End-Stage Liver Disease, Child-Pugh score, and age did not differ significantly, whereas albumin was higher in the groups with higher mortality (table 4).Moreover, Luo et al. stated that bilirubinemia >7 mg/dl is not a contraindication for transjugular intrahepatic portosystemic shunt (TIPS) insertion for ascites. This is a dangerous and misleading statement because studies so far have excluded those patients. [1] There is evidence that bilirubin below 3 mg/dl is a good biomarker to predict outcome in patients with ascites. [2] The strong association of bilirubin levels with post-TIPS outcome reinforces the recommendation of guidelines for extreme caution for TIPS insertion for ascites if bilirubin >3-5 mg/dl. [3,4] Although the quality of stents has improved over time, we included many patients and followed them over a long period of time. Indeed, we have reported the stent types used here (figure S7).Luo et al. raised an important issue of portal hepatic pressure gradient (PPG) variation and changes over time in our patients. It is very unlikely that in the real world outside of clinical trials longitudinal invasive measurements of PPG after TIPS are routinely performed. Ultrasound-based assessment of patency is the routine follow-up of choice. Therefore, we try to elaborate an easy-to-obtain surrogate marker at the time of TIPS insertion that is applicable for most cases in real-life practice. Hereby, percentage decrease in PPG is a good marker for response to treatment in ascites control in our cohort.Finally, we agree that in recently compensated cases with ascites as the first decompensating event, the amount of ascites may be prognostic. [5] However, all patients in our study received TIPS for therapy of recurrent and refractory ascites. Thus, the cited study does not apply to our patients and therefore the concern is unjustified.We thank again Luo et al. and hope that our answers could address their concerns and avoid confusion in further readers.
CONFLICT OF INTERESTNothing to report.