We appreciate the letter from Deng and colleagues regarding our study. They raised several questions related to the treatment of hepatocellular carcinoma with sorafenib or hepatic arterial infusion chemotherapy (HAIC). 1 First, they were curious about combination therapies. As we described in the combination and second-line therapy section, patients in the HAIC group more frequently (68%) received combination such as transarterial chemolipiodolization (15.3%), radiotherapy (26.3%), or both therapies (23.7%) as well as radiotherapy plus percutaneous ethanol injection (2.6%). Among the patients in the sorafenib group, only 17% received combination therapy using transarterial chemolipiodolization (5.7%) and radiotherapy (11.4%). To prevent ischemic insults to the liver in the presence of portal vein tumor thrombosis, we chose transarterial chemolipiodolization for patients needing local tumor control. 2 However, overall survival or time to progression was not different between the monotherapy and combination therapy groups. During the study period, as there was no approved immune-oncologic agent, we did not have a chance to evaluate the combination of HAIC with immunotherapies such as atezolizumab plus bevacizumab. However, this is an interesting topic that has recently yielded favorable data, so further evaluation of HAIC in combination with the newer drugs is warranted. 3 Second, treatment-associated adverse events were more common in the HAIC group than in the sorafenib group. Most grade 3 or 4 adverse events were hematologic or portrelated events, which are mostly manageable. 1 Regarding the underlying liver function and performance, one-third had Child-Pugh class B and another one-third showed Eastern Cooperative Oncology Group grade 2 or 3 status in our study. These patients were associated with worse survival than those with well-preserved liver function and performance status as Deng and colleagues indicated. However, we cannot exclude these patients from the treatment in the real-life setting as their survival is too short with supportive care only. In our study, compared with sorafenib, HAIC showed trends of benefit in overall survival in patients with Child-Pugh B liver function (5.9 months vs 15.3 months for Child-Pugh 7 and 3.4 months vs 5.9 months for Child-Pugh 8 or 9) and performance status 3 (0.8 months vs 6.3 months, respectively). 1 However, there were no statistical differences, suggesting further studies with lager number of patients are needed in these subgroups of the population.Third, although we used cisplatin and 5-fluorouracil for HAIC, there are reports of combining new cytotoxic agents such as oxaliplatin. 4,5 We appreciate the suggestion of performing HAIC using FOLFOX (folinic acid, 5-fluorouracil, and oxaliplatin), which significantly improved overall survival over sorafenib. Furthermore, improved local tu-