Advertisement

Proton beam radiotherapy vs. radiofrequency ablation for recurrent hepatocellular carcinoma: A randomized phase III trial

  • Author Footnotes
    † T.H.K. and Y.H. K. contributed equally to this study.
    Tae Hyun Kim
    Footnotes
    † T.H.K. and Y.H. K. contributed equally to this study.
    Affiliations
    Center for Liver and Pancreatobiliary Cancer, National Cancer Center, Goyang, Republic of Korea

    Center for Proton Therapy, National Cancer Center, Goyang, Republic of Korea
    Search for articles by this author
  • Author Footnotes
    † T.H.K. and Y.H. K. contributed equally to this study.
    Young Hwan Koh
    Footnotes
    † T.H.K. and Y.H. K. contributed equally to this study.
    Affiliations
    Center for Liver and Pancreatobiliary Cancer, National Cancer Center, Goyang, Republic of Korea

    Department of Radiology, National Cancer Center, Goyang, Republic of Korea
    Search for articles by this author
  • Bo Hyun Kim
    Affiliations
    Center for Liver and Pancreatobiliary Cancer, National Cancer Center, Goyang, Republic of Korea
    Search for articles by this author
  • Min Ju Kim
    Affiliations
    Department of Radiology, National Cancer Center, Goyang, Republic of Korea
    Search for articles by this author
  • Ju Hee Lee
    Affiliations
    Center for Liver and Pancreatobiliary Cancer, National Cancer Center, Goyang, Republic of Korea

    Department of Radiology, National Cancer Center, Goyang, Republic of Korea
    Search for articles by this author
  • Boram Park
    Affiliations
    Biostatistics Collaboration Team, Research Core Center, National Cancer Center, Goyang, Republic of Korea
    Search for articles by this author
  • Joong-Won Park
    Correspondence
    Corresponding author. Address: Center for Liver and Pancreatobiliary Cancer, Research Institute and Hospital, National Cancer Center, 323 Ilsan-ro, Ilsandong-gu, Goyang-si, Gyeonggi-do, 10408, South Korea. Tel.: +82-31-920-1605; fax: +82-31-920-0149.
    Affiliations
    Center for Liver and Pancreatobiliary Cancer, National Cancer Center, Goyang, Republic of Korea
    Search for articles by this author
  • Author Footnotes
    † T.H.K. and Y.H. K. contributed equally to this study.
Open AccessPublished:October 05, 2020DOI:https://doi.org/10.1016/j.jhep.2020.09.026

      Highlights

      • This is the 1st phase III RCT to evaluate proton beam radiotherapy vs. radiofrequency ablation in recurrent HCC.
      • Proton beam radiotherapy was non-inferior to radiofrequency ablation in terms of local progression-free survival and safety.
      • Proton beam radiotherapy is a promising treatment option for small HCC.

      Background & Aims

      Proton beam radiotherapy (PBT) has recently been applied to treat hepatocellular carcinoma (HCC); however, there is no randomized controlled trial-based evidence on its safety and efficacy. We compared the outcomes of PBT and radiofrequency ablation (RFA) in patients with recurrent/residual HCC (rHCC) in a phase III non-inferiority trial.

      Methods

      Patients with rHCC (size <3 cm, number ≤2) were randomly assigned to receive PBT or RFA according to Child-Pugh score and tumor stage. After randomization, if the assigned treatment was technically infeasible, crossover was allowed. The primary endpoint was 2-year local progression-free survival (LPFS), with a non-inferiority margin of 15% in the per-protocol (PP) population; a complementary analysis was performed in the intention-to-treat (ITT) population (NCT01963429).

      Results

      The ITT population comprised 144 patients receiving either PBT (n = 72) or RFA (n = 72). Six patients switched from the PBT arm to the RFA arm and 19 patients switched from the RFA arm to the PBT arm. In the PP population, the 2-year LPFS rate with PBT (n = 80) vs. RFA (n = 56) was 94.8% vs. 83.9%, a difference of 10.9 percentage points (90% CI 1.8–20.0; p <0.001); in the ITT population, the 2-year LPFS rate with PBT vs. RFA was 92.8% vs. 83.2%, a difference of 9.6 percentage points (90% CI 0.7–18.4; p <0.001), meeting the criteria for non-inferiority. The 3- and 4-year LPFS rates for PBT were also non-inferior to those for RFA. The most common adverse events were radiation pneumonitis (32.5%) and decreased leukocyte counts (23.8%) for PBT and increased alanine aminotransferase levels (96.4%) and abdominal pain (30.4%) for RFA. No Grade 4 adverse events or mortality were noted.

      Conclusions

      PBT showed LPFS values that were non-inferior to those for RFA; in addition, PBT was tolerable and safe.

      Clinical trial number

      Lay summary

      Radiofrequency ablation is the standard of care for patients with small hepatocellular carcinoma in whom surgery is not feasible. This study is the first phase III randomized controlled trial to evaluate the clinical outcomes of proton beam radiotherapy vs. radiofrequency ablation in patients with recurrent small HCC. Our findings show that this new technique is not inferior and can be applied safely in patients with small recurrent hepatocellular carcinoma.

      Graphical abstract

      Keywords

      Introduction

      Primary liver cancer, including hepatocellular carcinoma (HCC), is the fifth most commonly observed cancer worldwide. HCC-related prognoses remain poor owing to the presence of underlying chronic liver disease, late diagnosis, and frequent recurrence or progression after treatment.
      • Akinyemiju T.
      • Abera S.
      • Ahmed M.
      • Alam N.
      • Alemayohu M.A.
      • Allen C.
      • et al.
      Global Burden of Disease Liver Cancer Collaboration
      The burden of primary liver cancer and underlying etiologies from 1990 to 2015 at the global, regional, and national level: results from the Global Burden of Disease study 2015.
      ,
      • Park J.W.
      • Chen M.
      • Colombo M.
      • Roberts L.R.
      • Schwartz M.
      • Chen P.J.
      • et al.
      Global patterns of hepatocellular carcinoma management from diagnosis to death: the BRIDGE Study.
      The initial treatment modality for HCC is generally based on tumor stage, underlying liver function, and performance.
      European Association for the Study of the Liver
      EASL clinical practice guidelines: management of hepatocellular carcinoma.
      Korean Liver Cancer Association, National Cancer Center
      2018 Korean Liver Cancer Association-National Cancer Center Korea practice guidelines for the management of hepatocellular carcinoma.
      • Marrero J.A.
      • Kulik L.M.
      • Sirlin C.B.
      • Zhu A.X.
      • Finn R.S.
      • Abecassis M.M.
      • et al.
      Diagnosis, staging, and management of hepatocellular carcinoma: 2018 practice guidance by the American Association for the Study of Liver Diseases.
      Surgical resection is considered the best treatment for HCC, but for small HCC cases, in which resection is not feasible, radiofrequency ablation (RFA) is recommended; RFA is also the primary recommendation for tumors smaller than 2 cm.
      European Association for the Study of the Liver
      EASL clinical practice guidelines: management of hepatocellular carcinoma.
      ,
      Korean Liver Cancer Association, National Cancer Center
      2018 Korean Liver Cancer Association-National Cancer Center Korea practice guidelines for the management of hepatocellular carcinoma.
      The recurrence rate of HCC after resection or RFA is significant. In the case of local recurrence, despite the lack of sufficient evidence, the treatment method is generally selected according to the first-line treatment principle.
      European Association for the Study of the Liver
      EASL clinical practice guidelines: management of hepatocellular carcinoma.
      ,
      Korean Liver Cancer Association, National Cancer Center
      2018 Korean Liver Cancer Association-National Cancer Center Korea practice guidelines for the management of hepatocellular carcinoma.
      Proton beam radiotherapy (PBT) is a type of radiation therapy that has been proven to be effective as a primary treatment for ocular tumor and prostatic cancer.
      • Levin W.P.
      • Kooy H.
      • Loeffler J.S.
      • DeLaney T.F.
      Proton beam therapy.
      ,
      • Zietman A.L.
      • Bae K.
      • Slater J.D.
      • Shipley W.U.
      • Efstathiou J.A.
      • Coen J.J.
      • et al.
      Randomized trial comparing conventional-dose with high-dose conformal radiation therapy in early-stage adenocarcinoma of the prostate: long-term results from proton radiation oncology group/american college of radiology 95-09.
      PBT has recently been applied to HCC treatment, and its safety and local control effects have been reported in various studies.
      • Bush D.A.
      • Kayali Z.
      • Grove R.
      • Slater J.D.
      The safety and efficacy of high-dose proton beam radiotherapy for hepatocellular carcinoma: a phase 2 prospective trial.
      • Kim T.H.
      • Park J.W.
      • Kim Y.J.
      • Kim B.H.
      • Woo S.M.
      • Moon S.H.
      • et al.
      Phase I dose-escalation study of proton beam therapy for inoperable hepatocellular carcinoma.
      • Hong T.S.
      • Wo J.Y.
      • Yeap B.Y.
      • Ben-Josef E.
      • McDonnell E.I.
      • Blaszkowsky L.S.
      • et al.
      Multi-institutional phase II study of high-dose hypofractionated proton beam therapy in patients with Localized, unresectable hepatocellular carcinoma and intrahepatic cholangiocarcinoma.
      A previous phase II study in HCC patients demonstrated the safety and promising outcomes associated with PBT.
      • Bush D.A.
      • Kayali Z.
      • Grove R.
      • Slater J.D.
      The safety and efficacy of high-dose proton beam radiotherapy for hepatocellular carcinoma: a phase 2 prospective trial.
      ,
      • Hong T.S.
      • Wo J.Y.
      • Yeap B.Y.
      • Ben-Josef E.
      • McDonnell E.I.
      • Blaszkowsky L.S.
      • et al.
      Multi-institutional phase II study of high-dose hypofractionated proton beam therapy in patients with Localized, unresectable hepatocellular carcinoma and intrahepatic cholangiocarcinoma.
      • Kim T.H.
      • Park J.W.
      • Kim B.H.
      • Oh E.S.
      • Youn S.H.
      • Moon S.H.
      • et al.
      Phase II study of hypofractionated proton beam therapy for hepatocellular carcinoma.
      • Kawashima M.
      • Furuse J.
      • Nishio T.
      • Konishi M.
      • Ishii H.
      • Kinoshita T.
      • et al.
      Phase II study of radiotherapy employing proton beam for hepatocellular carcinoma.
      However, as there is no strong randomized controlled trial (RCT)-based evidence on PBT, it is not generally recommended for HCC treatment. Therefore, we conducted an RCT to evaluate and compare the local efficacy and clinical outcomes of PBT and RFA in patients with recurrent or residual HCC (rHCC).

      Patients and methods

      Study population

      This phase III investigator-initiated, randomized, single-center, open-label clinical trial was performed at the National Cancer Center (NCC), Goyang, Republic of Korea. The eligibility criteria for this study were as follows: HCC diagnosis was confirmed either histologically or clinically according to the Korean Liver Cancer Study Group and NCC Korea guidelines
      Korean Liver Cancer Study Group, National Cancer Center
      ,
      Korean Liver Cancer Study Group, National Cancer Center
      2014 KLCSG-NCC Korea practice guideline for the management of hepatocellular carcinoma.
      ; presence of recurrent or residual HCC lesions without vascular invasion after other treatment; the largest diameter and number of target lesion(s) were <3 cm and ≤2, respectively; no history of prior radiotherapy to targeted lesion(s); no evidence of extrahepatic metastasis; Child-Pugh score ≤7 without uncontrolled ascites; Eastern Cooperative Oncology Group performance status ≤2; age ≥18 years; adequate bone marrow (white blood cell count ≥2,000/μl, platelet count ≥50,000/μl, and hemoglobin level ≥7.5 g/dl) and liver function (total bilirubin level ≤3.0 mg/dl, and aspartate aminotransferase and alanine aminotransferase level <5.0× the upper limit of normal). All patients were unresectable or unwilling to undergo resection. All patients provided written informed consent before enrollment (supplementary information; study protocol). This study was approved by the NCC institutional review board and complied with the Declaration of Helsinki and Good Clinical Practice guideline.

      Study design and treatment

      Eligible patients were randomly assigned (1:1) to the PBT arm or RFA arm with stratification according to the Child-Pugh classification (A vs. B7) and tumor stage (American Joint Committee on Cancer [AJCC] 7th edition stage I–II vs. III). After randomization to each treatment arm, if the assigned method was not technically feasible, the patients were allowed to be treated with the other method. Patients and all investigators were unmasked to the treatment assignment (supplementary information; study protocol).
      RFA was performed percutaneously using a monopolar water-cooling electrode system (Covidien-Medtronic, Minneapolis, MN, USA; and STARmed, Goyang, Korea) under ultrasound or CT guidance. The number of electrodes for single or multiple electrodes and overlapping RFA procedures were determined by the tumor size. RFA was performed under local anesthesia with intravenous analgesia or monitored anesthesia care. During RFA procedures, the output power was initially set at 50W and gradually increased from 90W to 200W, where it was maintained until the impedance reached a maximum value. Cold saline was infused into the electrode lumen using the pump to keep the tip temperature below 20°C. RFA procedures were continued until the entire tumor and border area sizes greater than 5 mm were included in the detected target lesion(s) on ultrasound or CT. Immediately after RFA, contrast-enhanced dynamic liver CT was conducted, and the results were carefully reviewed by the radiologist for the assessment of the technical success and procedure-related complications. If the level of ablation was considered insufficient, additional RFA was repeated during the same hospital stay.
      PBT was performed using 230 MeV passively scattered proton beams (Proteus 235; Ion Beam Applications, S.A., Louvain-la-Neuve, Belgium). A contrast-enhanced four-dimensional CT scan was obtained in each patient. The internal target volume (ITV) was calculated as the sum of the gross tumor volumes in each CT image during the gated (exhalation) phases (30% of the total respiratory cycle). The planning target volumes (PTVs) were defined as the ITVs plus 5–7 mm margins in all directions. The prescribed dose to the PTV was 66 Gray equivalent (GyE) in 10 fractions, 5 fractions per week.
      • Fukumitsu N.
      • Sugahara S.
      • Nakayama H.
      • Fukuda K.
      • Mizumoto M.
      • Abei M.
      • et al.
      A prospective study of hypofractionated proton beam therapy for patients with hepatocellular carcinoma.
      ,
      • Mizumoto M.
      • Okumura T.
      • Hashimoto T.
      • Fukuda K.
      • Oshiro Y.
      • Fukumitsu N.
      • et al.
      Proton beam therapy for hepatocellular carcinoma: a comparison of three treatment protocols.
      All patients was asked to fast for at least 4 hours before PBT at each treatment, and radiation was delivered under gated phases with a respiratory-gated technique.
      Planned clinical, laboratory, and tumor assessment via contrast-enhanced multiphasic CT or MRI was performed within 2 weeks before each treatment, at the first month after the completion of RFA or PBT, every 3 months for the following 2 years, and every 6 months thereafter. Clinical and laboratory tests were performed before discharge after RFA treatment and during PBT treatment every week.

      Outcomes and assessments

      The primary endpoint was 2-year local progression-free survival (LPFS), defined as the time from the commencement date of each intervention to the date of local progression, and it was censored at the date of the last follow-up when the patients had no evidence of local progression. The secondary endpoints were progression-free survival (PFS), defined as time from the commencement date of each treatment to the date of local, intrahepatic or distant progression or death from any cause; overall survival (OS), defined as time from the commencement date of each treatment to date of death from any cause; and safety, defined as the presentation of adverse events (AEs) related to the treatments that were assessed according to the Common Terminology Criteria for Adverse Events version 3.0. Tumor assessment including size and response, and the evaluation of disease progression were conducted by 2 radiologists (Y-HK and JHL), and reviewed by an independent radiologist (MJK) according to RECIST version 1.1.
      • Eisenhauer E.A.
      • Therasse P.
      • Bogaerts J.
      • Schwartz L.H.
      • Sargent D.
      • Ford R.
      • et al.
      New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1).
      Local progression was defined as the presence of regrowth or new tumor growth within 1 cm from the margin of the ablative zone in RFA or 1 cm from the margin of the PTV in PBT, respectively. Intrahepatic progression was defined as the regrowth of a previously treated non-target tumor or new tumor growth within the liver, except for local progression.

      Statistical analysis

      Efficacy was assessed in all the randomized patients (intention-to-treat [ITT] population) and patients treated per the protocol (per-protocol [PP] population), and safety was assessed in the PP population. Since this RCT had a non-inferiority crossover design, the analysis of the PP population was given priority, and the ITT analysis was performed complementarily.
      • Parpia S.
      • Julian J.A.
      • Thabane L.
      • Gu C.
      • Whelan T.J.
      • Levine M.N.
      Treatment crossovers in time-to-event non-inferiority randomised trials of radiotherapy in patients with breast cancer.
      Previous studies reported 2-year LPFS rates of 62.5–96.8% for RFA
      • Choi D.
      • Lim H.K.
      • Rhim H.
      • Kim Y.S.
      • Yoo B.C.
      • Paik S.W.
      • et al.
      Percutaneous radiofrequency ablation for recurrent hepatocellular carcinoma after hepatectomy: long-term results and prognostic factors.
      • Kim Y.S.
      • Lim H.K.
      • Rhim H.
      • Lee M.W.
      • Choi D.
      • Lee W.J.
      • et al.
      Ten-year outcomes of percutaneous radiofrequency ablation as first-line therapy of early hepatocellular carcinoma: analysis of prognostic factors.
      • Koh Y.H.
      • Choi J.I.
      • Kim H.B.
      • Kim M.J.
      Computed tomographic-guided radiofrequency ablation of recurrent or residual hepatocellular carcinomas around retained iodized oil after transarterial chemoembolization.
      • Lencioni R.A.
      • Allgaier H.P.
      • Cioni D.
      • Olschewski M.
      • Deibert P.
      • Crocetti L.
      • et al.
      Small hepatocellular carcinoma in cirrhosis: randomized comparison of radio-frequency thermal ablation versus percutaneous ethanol injection.
      • Lin S.M.
      • Lin C.J.
      • Lin C.C.
      • Hsu C.W.
      • Chen Y.C.
      Randomised controlled trial comparing percutaneous radiofrequency thermal ablation, percutaneous ethanol injection, and percutaneous acetic acid injection to treat hepatocellular carcinoma of 3 cm or less.
      • Nakazawa T.
      • Kokubu S.
      • Shibuya A.
      • Ono K.
      • Watanabe M.
      • Hidaka H.
      • et al.
      Radiofrequency ablation of hepatocellular carcinoma: correlation between local tumor progression after ablation and ablative margin.
      • Shiina S.
      • Tateishi R.
      • Arano T.
      • Uchino K.
      • Enooku K.
      • Nakagawa H.
      • et al.
      Radiofrequency ablation for hepatocellular carcinoma: 10-year outcome and prognostic factors.
      and 54.6–96.0% for PBT
      • Bush D.A.
      • Kayali Z.
      • Grove R.
      • Slater J.D.
      The safety and efficacy of high-dose proton beam radiotherapy for hepatocellular carcinoma: a phase 2 prospective trial.
      ,
      • Hong T.S.
      • Wo J.Y.
      • Yeap B.Y.
      • Ben-Josef E.
      • McDonnell E.I.
      • Blaszkowsky L.S.
      • et al.
      Multi-institutional phase II study of high-dose hypofractionated proton beam therapy in patients with Localized, unresectable hepatocellular carcinoma and intrahepatic cholangiocarcinoma.
      • Kim T.H.
      • Park J.W.
      • Kim B.H.
      • Oh E.S.
      • Youn S.H.
      • Moon S.H.
      • et al.
      Phase II study of hypofractionated proton beam therapy for hepatocellular carcinoma.
      • Kawashima M.
      • Furuse J.
      • Nishio T.
      • Konishi M.
      • Ishii H.
      • Kinoshita T.
      • et al.
      Phase II study of radiotherapy employing proton beam for hepatocellular carcinoma.
      ,
      • Fukumitsu N.
      • Sugahara S.
      • Nakayama H.
      • Fukuda K.
      • Mizumoto M.
      • Abei M.
      • et al.
      A prospective study of hypofractionated proton beam therapy for patients with hepatocellular carcinoma.
      in treatment-naïve or rHCC (Tables S1 and S2). Because there was no LPFS data for RFA vs. placebo, this study was designed with reference to a previous RCT of RFA vs. percutaneous ethanol injection (PEI) in patients with HCCs <3 cm
      • Lin S.M.
      • Lin C.J.
      • Lin C.C.
      • Hsu C.W.
      • Chen Y.C.
      Randomised controlled trial comparing percutaneous radiofrequency thermal ablation, percutaneous ethanol injection, and percutaneous acetic acid injection to treat hepatocellular carcinoma of 3 cm or less.
      ; the difference of 2-year LRFS rate between RFA vs. PEI was 20%. We assumed that PBT could be more effective than PEI; the non-inferiority margin was calculated as 15%. A total of 144 patients (72 patients per arm) were required for the achievement of a power of 80%, a type I error level of 5%, and a follow-up loss rate of 5%. The primary endpoint, 2-year LPFS, was tested for non-inferiority with a margin of 15% – preferentially in the PP population and complementarily in ITT population – and its critical value was evaluated using the Com-Nougue approach
      • Com-Nougue C.
      • Rodary C.
      • Patte C.
      How to establish equivalence when data are censored: a randomized trial of treatments for B non-Hodgkin lymphoma.
      calculating the Z-statistic with standard error estimated by Greenwood's formula.
      Differences in the incidence of AEs between the arms were evaluated using the χ2 test and Fisher's exact test. Survival outcomes and hazard ratios (HRs) were estimated using the Kaplan-Meier method and a Cox proportional hazard model, respectively. A p value <0.05 was considered statistically significant, and all statistical tests were performed using SAS software (version 9.4; SAS Institute Inc., Cary, NC, USA) and R software (version 3.6.2; R Project for Statistical Computing).

      Results

      Patients

      Of the 154 patients assessed for eligibility between December 2013 and December 2017, 144 were randomly assigned to the PBT arm (n = 72) and RFA arm (n = 72), and they comprised the ITT population (Fig. 1). The last patient completed the trial on January 2020. In the PBT arm, due to the proximity of the gastrointestinal organ to the targeted tumor(s) (n = 9) or withdrawal of informed consent (n = 2), finally 6 (8.3%) patients crossed over to the RFA arm and 5 patients received another treatment. In the RFA arm, due to the presence of a non-echogenic tumor, inadequate electrode path or the proximity of major vessels and bile ducts to the targeted tumor(s) (n = 22), 19 (26.4%) patients crossed over to the PBT arm and 3 received another treatment (Fig. 1). Thus, the PP population comprised 80 patients in the PBT arm and 56 patients in the RFA arm. The rate of crossover to the other arm due to technical infeasibility was significantly higher in the RFA arm (26.4%) than the PBT arm (8.3%) (p = 0.004). The median follow-up duration was 51.6 months (90% CI 45.6–59.5) (IQR 39.3–67.8) in the PBT arm and 50.7 months (90% CI 45.8–57.7) (IQR 41.4–67.4) in the RFA arm. Between both arms, the baseline characteristics of the ITT population were well-balanced (Tables 1 and S3).
      Figure thumbnail gr1
      Fig. 1CONSORT diagram.
      PBT, proton beam radiotherapy; RFA, radiofrequency ablation; TACE, transarterial chemoembolization.

      Outcomes

      In the PP population, for all patients (n = 136), the 2-year LPFS rate associated with PBT vs. RFA was 94.8% vs. 83.9% (difference of 10.9%; 90% CI 1.8–20.0), and for all target lesions (n = 143), the corresponding values were 95.1% vs. 84.4%, respectively. In the ITT population, for all patients, the 2-year LPFS rate associated with PBT vs. RFA was 92.8% vs. 83.2% (difference of 9.6%; 90% CI 0.7–18.4), and for all target lesions, the corresponding values were 93.3% vs. 83.2%, respectively; both analyses met the criteria for non-inferiority (Table 2, Fig. 2). The HRs of LPFS between the PBT and RFA arms for all target lesions and all patients were not significantly different (p >0.05 each) (Fig. 3A–D). The lower boundary of the CI of the differences (PBT minus RFA) in the 2-year LPFS did not include the non-inferiority margin of −15% (Fig. 2); thus, the non-inferiority of PBT in terms of LPFS was declared. In addition, in the PP population, for all patients, the 3- and 4-year LPFS rates were 88.3% (79.9–93.3%) and 85.8% (76.3–91.7%) in the PBT arm and 77.6% (66.4–85.5%) and 77.6% (66.4–85.5%) in the RFA arm, respectively, and the differences in the 3- and 4-year LPFS rates between the 2 arms were 10.7% (−0.8 to 22.1%) and 8.2% (−3.9 to 21.2%), respectively (Table 2, Fig. 2). The median PFS values in the PBT arm were 13.4 months (90% CI 10.32–16.76) and 13.4 months (90% CI 7.69–16.76), and those in the RFA arm were 13.6 months (90% CI 9.03–21.72) and 13.7 months (90% CI 9.86–18.89) in the PP and ITT populations, respectively. In the PP and ITT populations, the 2-year PFS rates were 28.7% and 31.9% in the PBT arm, and 37.5% and 31.9% in the RFA arm, respectively (Table 2); the HRs of the 2-year PFS rate between the PBT and RFA arms were not different (HR 1.10; 95% CI 0.76–1.59; HR 0.99; 95% CI 0.70–1.33, respectively) (Fig. 3E–F). The 3- and 4-year PFS rates were 20.9% and 15.6% in the PBT arm, and 23.1% and 13.8% in RFA arm, respectively, in the PP population (Table 2).
      Table 2Local progression-free survival, progression-free survival and overall survival in the proton beam radiotherapy and radiofrequency ablation arms in the per-protocol and intention-to-treat populations.
      NPBTRFAHazard ratiop value
      Cox proportional hazards model.
      Per-protocol population
       LPFS (for all target lesions), % (90% CI)143
      2 year95.1 (89.1–97.8)84.4 (74.6–90.7)0.51 (0.26–1.03)0.114
      3 year88.9 (80.9–93.7)78.3 (67.3–86.0)
      4 year86.5 (77.4–92.1)78.3 (67.3–86.0)
       LPFS (for all patients), % (90% CI)136
      2 year94.8 (88.5–97.7)83.9 (73.8–90.4)0.52 (0.26–1.05)0.123
      3 year88.3 (79.9–93.3)77.6 (66.4–85.5)
      4 year85.8 (76.3–91.7)77.6 (66.4–85.5)
       PFS (for all patients), % (95% CI)136
      2 year28.7 (19.3–38.9)37.5 (25.0–49.9)1.10 (0.76–1.59)0.623
      3 year20.9 (12.7–30.5)23.1 (13.1–34.8)
      4 year15.6 (8.3–25.0)13.8 (6.0–24.9)
       OS (for all patients), % (95% CI)136
      2 year88.8 (79.5–94.0)92.9 (82.1–97.3)1.19 (0.62–2.27)0.600
      3 year79.0 (67.9–86.6)87.2 (74.9–93.7)
      4 year74.0 (62.1–82.7)78.0 (63.6–87.3)
      Intention-to-treat population
       LPFS (for all target lesions), % (90% CI)155
      2 year93.3 (86.5–96.7)83.2 (74.7–89.0)0.68 (0.37–1.26)0.306
      3 year86.8 (78.2–92.2)78.4 (69.1–85.2)
      4 year84.3 (74.7–90.5)78.4 (69.1–85.2)
       LPFS (for all patients),% (90% CI)144
      2 year92.8 (85.6–96.5)83.2 (74.4–89.2)0.73 (0.39–1.38)0.419
      3 year85.8 (76.5–91.6)78.3 (68.6–85.3)
      4 year83.0 (72.6–89.7)78.3 (68.6–85.3)
       PFS (for all patients), % (95% CI)144
      2 year31.9 (21.6–42.8)31.9 (21.6–42.8)0.99 (0.70–1.41)0.958
      3 year26.3 (16.8–36.8)17.9 (10.1–27.6)
      4 year18.7 (10.2–29.1)12.6 (5.9–21.8)
       OS (for all patients), % (95% CI)144
      2 year91.7 (82.4–96.2)90.3 (80.7–95.2)1.07 (0.58–1.98)0.821
      3 year80.8 (69.2–88.4)86.0 (69.2–88.4)
      4 year75.4 (62.8–84.2)77.0 (64.5–85.6)
      LPFS, local progression-free survival; OS, overall survival; PBT, proton beam radiotherapy; PFS, progression-free survival; RFA, radiofrequency ablation.
      Cox proportional hazards model.
      Figure thumbnail gr2
      Fig. 2Forest plots of local progression-free survival in all patients.
      p value was calculated for the hypothesis of non-inferiority with a non-inferiority margin of −15.0%. A 1-sided p value <0.05 was considered statistically significant (Cox-proportional hazard model). ITT, intention-to-treat; PBT, proton beam radiotherapy; PP, per-protocol; RFA, radiofrequency ablation.
      Figure thumbnail gr3
      Fig. 3Survival curves in the PP population and ITT population.
      (A) LPFS curves for all target lesions in the PP population. (B) LPFS curves in the ITT population. (C) LPFS curves for all patients in the PP population. (D) LPFS curves for all patients in the ITT population. (E) PFS curves in the PP population. (F) PFS curves in the ITT population. (G) OS curves in the PP population. (H) OS curves in the ITT population. Kaplan-Meier method and a Cox proportional hazard model was used. HR, hazard ratio; ITT, intention-to-treat; LPFS, local progression-free survival; OS, overall survival; PFS, progression-free survival; PBT, proton beam radiotherapy; PP, per-protocol; RFA, radiofrequency ablation.
      The median OS value was not reached in both arms and both populations. In the PP and ITT populations, the 2-year OS rates were 88.8% and 91.7% in the PBT arm, and 92.9% and 90.3% in the RFA arm, respectively (Table 2); the HRs of the OS rate between the PBT and RFA arms were not different (HR 1.19; 95% CI 0.62–2.27; HR 1.07; 95% CI 0.58–1.98, respectively) (Fig. 3G–H). The 3- and 4-year OS rates were 79.0% and 74.0%, respectively, in the PBT arm and 87.2% and 78.0%, respectively, in the RFA arm in the PP population (Table 2).
      In the PP population, the best tumor responses in the PBT arm and RFA arm for all target lesions (n = 155) were complete response (CR) in 71 patients (83.5%) and 56 patients (96.6%); partial response in 6 patients (7.1%) and 1 patient (1.7%); stable disease in 8 patients (9.4%) and 1 patient (1.7%); and progressive disease in no patient, respectively (p = 0.063) (Fig. S1). Median time to the best tumor response was 4.4 months (range 1–13) in the PBT arm and 1 month (range 0.8–1.2) in the RFA arm (p <0.001). In the PP population, of the 80 patients in the PBT group, 65 (81.3%) had disease progression and 24 (30%) died from disease progression (n = 21), an unknown cause (n = 1), chronic renal failure (n = 1), and pneumonia (n = 1) unrelated to the treatment. Cumulative local, intrahepatic and extrahepatic progression were observed in 10 (12.5%), 65 (81.3%), and 10 patients (12.5%), respectively (Fig. S2). Of the 56 patients in the RFA group, 42 (82.1%) had disease progression and 15 (26.8%) died from disease progression unrelated to the treatment. Cumulative local, intrahepatic and extrahepatic progression occurred in 13 (23.2%), 41 (73.2%), and 11 patients (19.6%), respectively (Fig. S2). After disease progression development, subsequent treatment was performed; there was no significant difference between the PBT and RFA arms (p >0.05) (Table S4).

      Safety

      In the PP population, the incidence rates of leukopenia, thrombocytopenia, hyperbilirubinemia and hypoalbuminemia were similar in both arms (p >0.05 each) (Table 3). The incidences of elevated alanine aminotransferase levels were higher in the RFA arm than the PBT arm (grade 1, 25% vs. 12.5%; grade 2, 57.1% vs. 3.8%; and grade 3, 14.3% vs. 0%; p <0.001), and those of increased Child-Pugh score were higher in the RFA arm than the PBT arm (1-point decrease, 0% vs. 2.5%; no change, 80.4% vs. 90%; 1-point increase, 19.6% vs. 7.5%; p = 0.049). The incidence of dermatitis and radiation pneumonitis were higher in the PBT group than the RFA group (grade 1, 17.5% vs. 0%, p <0.001; and grade 1, 32.5% vs. 0%, p <0.001, respectively). The incidence of radiation pneumonitis is relatively high, but all were asymptomatic, radiographic changes (grade 1). Although grade 3 AEs occurred more frequently in the RFA group than the PBT group (16.1% vs. 0%, p <0.001), these toxicities were transient and all patients recovered. The RFA arm showed no major biliary and pulmonary AEs, except for 1 (1.8%) patient who had bleeding in the inferior phrenic artery after RFA that was controlled by embolization. The median length of hospital stay in the RFA arm was 3.1 days (range 2–8), and the median out-patient clinic duration in the PBT arm was 12.7 days (p <0.001). In both arms, treatment-related late hepatic failure and death without evidence of disease progression and/or subsequent treatment were not observed.
      Table 3Adverse events after proton beam radiotherapy and radiofrequency ablation.
      CTCAE gradePBT (n = 80), n (%)RFA (n = 56), n (%)p value
      Grade 1Grade 2Grade 3Grade 4Grade 1Grade 2Grade 3Grade 4
      WBC increase0 (0.0)0 (0.0)0 (0.0)0 (0.0)1 (1.8)0 (0.0)0 (0.0)0 (0.0)0.412
      Fisher's exact test.
      WBC decrease18 (22.5)1 (1.3)0 (0.0)0 (0.0)9 (16.1)0 (0.0)0 (0.0)0 (0.0)0.569
      Fisher's exact test.
      PLT decrease15 (18.8)0 (0.0)0 (0.0)0 (0.0)16 (28.6)0 (0.0)0 (0.0)0 (0.0)0.179
      Chi-square test.
      ALT/AST increase10 (12.5)3 (3.8)0 (0.0)0 (0.0)14 (25.0)32 (57.1)8 (14.3)0 (0.0)<0.001
      Fisher's exact test.
      Albumin decrease5 (6.3)0 (0.0)0 (0.0)0 (0.0)3 (5.4)0 (0.0)0 (0.0)0 (0.0)1.000
      Fisher's exact test.
      Bilirubin increase8 (10.0)0 (0.0)0 (0.0)0 (0.0)8 (14.3)0 (0.0)0 (0.0)0 (0.0)0.445
      Chi-square test.
      Fever0 (0.0)0 (0.0)0 (0.0)0 (0.0)5 (8.9)1 (1.8)0 (0.0)0 (0.0)0.004
      Fisher's exact test.
      Pain0 (0.0)1 (1.3)0 (0.0)0 (0.0)10 (17.9)7 (12.5)0 (0.0)0 (0.0)<0.001
      Fisher's exact test.
      Nausea0 (0.0)1 (1.3)0 (0.0)0 (0.0)5 (8.9)1 (1.8)0 (0.0)0 (0.0)0.011
      Fisher's exact test.
      Bleeding0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)1 (1.8)0 (0.0)0.412
      Fisher's exact test.
      Dermatitis14 (17.5)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0.001
      Chi-square test.
      Radiation pneumonitis26 (32.5)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)<0.001
      Chi-square test.
      Ascites0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)-
      Upper gastrointestinal ulcer0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)0 (0.0)-
      No. of patients with ≥ Grade 3 AEs0 (0.0)9 (16.1)<0.001
      Fisher's exact test.
      Change of Child-Pugh score−10+1+2−10+1+20.049
      Fisher's exact test.
      2 (2.5)72 (90.0)6 (7.5)0 (0)0 (0)45 (80.4)11 (19.6)0 (0)
      AE, adverse event; ALT, alanine aminotransferase; AST, aspartate aminotransferase; PBT, proton beam radiotherapy; PLT, platelet; RFA, radiofrequency ablation; WBC, white blood cell.
      Chi-square test.
      Fisher's exact test.

      Discussion

      This study demonstrated that PBT yielded LPFS values that were comparable to those obtained on RFA in patients with rHCC. Although the rate of crossover to other treatments after randomization was relatively high in both arms, PBT showed consistently non-inferior values, in terms of LPFS, to RFA in both the ITT and PP analyses. Despite crossover, patient characteristics between both arms in ITT and PP population were not significantly different (Table 1). In confirmatory trials with a superiority design, the ITT set is usually used in the primary analysis. However, in non-inferiority, crossover clinical trials, such as the present one, the analysis of the PP population was given priority and ITT analysis was performed complementarily.
      • Parpia S.
      • Julian J.A.
      • Thabane L.
      • Gu C.
      • Whelan T.J.
      • Levine M.N.
      Treatment crossovers in time-to-event non-inferiority randomised trials of radiotherapy in patients with breast cancer.
      Until the planning of this study, we lacked sufficient evidence for the application of PBT as a first-line treatment, so we performed this RCT with PBT as a second-line treatment in patients with rHCC. Because a non-inferiority trial is a research method that is selected when the treatment to be studied is considered to be equivalent and to be complementary to the standard treatment (for cases of technical infeasibility). We assumed that PBT had an equivalent effect on local control and could be applied as complementary to RFA treatment. Thus, we selected a non-inferiority crossover trial for this study. Additionally, selecting patients eligible for both therapies could have excluded the majority of real-world patients who may benefit from either treatment; this study compared the outcomes of PBT vs. RFA for tumors with the same tumor conditions.
      Table 1Patients' characteristics between the proton beam radiotherapy and radiofrequency ablation arms.
      CharacteristicsIntention-to-treat populationp valuePer-protocol populationp value
      PBT (n = 72)RFA (n = 72)PBT (n = 80)RFA (n = 56)
      Sex
       Male61 (84.7)59 (81.9)0.655
      Pearson's Chi-square test.
      69 (86.3)45 (80.4)0.358
      Pearson's Chi-square test.
       Female11 (15.3)13 (18.1)11 (13.8)11 (19.6)
      Age, years
      Median (range)60 (46-78)61.5 (40-82)0.351
      Wilcoxon rank-sum test.
      61 (40-82)61 (44-77)0.974
      Wilcoxon rank-sum test.
       <6033 (45.8)28 (38.9)0.399
      Pearson's Chi-square test.
      34 (42.5)23 (41.1)0.868
      Pearson's Chi-square test.
       ≥6039 (54.2)44 (61.1)46 (57.5)33 (58.9)
      ECOG PS
       065 (90.3)67 (93.1)0.547
      Pearson's Chi-square test.
      73 (91.3)51 (91.1)1.000
      Fisher's exact test.
       17 (9.7)5 (6.9)7 (8.8)5 (8.9)
      Etiology
       HBV61 (84.7)60 (83.3)0.820
      Pearson's Chi-square test.
      67 (83.8)47 (83.9)0.978
       Others11 (15.3)12 (16.7)13 (16.3)9 (16.1)
      Child-Pugh classification
       A70 (97.2)70 (97.2)1.000
      Fisher's exact test.
      77 (96.3)55 (98.2)0.643
      Fisher's exact test.
       B72 (2.8)2 (2.8)3 (3.8)1 (1.8)
      AFP, ng/ml
      Continuous variables presented as median (range).
      4.9 (0.7-411.8)5.1 (1.4-963.6)0.984
      Wilcoxon rank-sum test.
      5.1 (0.7-411.8)5.1 (1.5-963.6)0.804
      Wilcoxon rank-sum test.
       <1049 (68.1)49 (68.1)1.000
      Pearson's Chi-square test.
      57 (71.3)36 (64.3)0.390
      Pearson's Chi-square test.
       ≥1023 (31.9)23 (31.9)23 (28.8)20 (35.7)
      Tumor size, cm
      Continuous variables presented as median (range).
      1.2 (1.0-2.5)1.2 (1.0-2.9)0.439
      Wilcoxon rank-sum test.
      1.2 (1.0-2.9)1.2 (1.0-2.2)0.956
      Wilcoxon rank-sum test.
       <263 (88.7)65 (90.3)0.763
      Pearson's Chi-square test.
      68 (86.1)52 (92.9)0.217
      Pearson's Chi-square test.
       ≥28 (11.3)7 (9.7)11 (13.9)4 (7.1)
      No. of treated lesion(s)
       167 (93.1)66 (91.7)0.754
      Pearson's Chi-square test.
      75 (93.8)54 (96.4)0.700
      Fisher's exact test.
       25 (6.9)6 (8.3)5 (6.3)2 (3.6)
      AJCC stage
       I20 (27.8)20 (27.8)1.000
      Fisher's exact test.
      21 (26.3)18 (32.1)0.476
      Fisher's exact test.
       II50 (69.4)50 (69.4)58 (72.5)36 (64.3)
       III2 (2.8)2 (2.8)1 (1.3)2 (3.6)
      mUICC stage
       I15 (20.8)16 (22.2)0.693
      Fisher's exact test.
      16 (20.0)14 (25.0)0.868
      Fisher's exact test.
       II24 (33.3)30 (41.7)30 (37.5)21 (37.5)
       III32 (44.4)25 (34.7)33 (41.3)20 (35.7)
       IVA1 (1.4)1 (1.4)1 (1.3)1 (1.8)
      BCLC stage
       04 (5.6)5 (6.9)0.908
      Fisher's exact test.
      5 (6.3)4 (7.1)0.988
      Fisher's exact test.
       A39 (54.2)38 (52.8)43 (53.8)31 (55.4)
       B24 (33.3)26 (36.1)28 (35.0)18 (32.1)
       C5 (6.9)3 (4.2)4 (5.0)3 (5.4)
      Diagnosis
       Pathologic22 (30.6)28 (38.9)0.294
      Pearson's Chi-square test.
      23 (28.8)22 (39.3)0.199
      Pearson's Chi-square test.
       Imaging-based50 (69.4)44 (61.1)57 (71.3)34 (60.7)
      Pre-Tx to target site(s)
       No41 (56.9)33 (45.8)0.182
      Pearson's Chi-square test.
      42 (52.5)30 (53.6)0.902
      Pearson's Chi-square test.
       Yes31 (43.1)39 (54.2)38 (47.5)26 (46.4)
      TACE29 (40.3)35 (48.6)37 (46.3)21 (37.5)
      RFA1 (1.4)3 (4.2)0 (0.0)4 (7.1)
      TACE + RFA1 (1.4)1 (1.4)1 (1.3)1 (1.8)
      Pre-Tx to non-target site(s)
       No3 (4.2)4 (5.6)1.000
      Fisher's exact test.
      4 (5.0)3 (5.4)1.000
      Fisher's exact test.
       Yes69 (95.8)68 (94.4)76 (95.0)53 (94.6)
      AFP, α-fetoprotein; AJCC stage, American Joint Committee on Cancer stage; BCLC stage, Barcelona Clinic Liver Cancer stage; ECOG PS, Eastern Cooperative Oncology Group performance status; mUICC stage, modified International Union Against Cancer stage; PBT, proton beam radiotherapy; RFA, radiofrequency ablation; TACE, transarterial chemoembolization; Tx, treatment.
      # Continuous variables presented as median (range).
      Pearson's Chi-square test.
      Fisher's exact test.
      Wilcoxon rank-sum test.
      RFA is frequently not feasible for HCC lesion(s) due to their invisibility under ultrasound guidance and limited accessibility of the tumor location.
      • Rhim H.
      • Lee M.H.
      • Kim Y.S.
      • Choi D.
      • Lee W.J.
      • Lim H.K.
      Planning sonography to assess the feasibility of percutaneous radiofrequency ablation of hepatocellular carcinomas.
      In the present study, RFA was not feasible in 30.6% (22 of 72) of the patients in the RFA arm. Of these patients, PBT was feasible in 86.4% (19 of 22) of the patients and not feasible in 13.6% (3 of 22) patients due to proximity to the gastrointestinal organs (Fig. 1). In the PBT arm, PBT was not feasible in 15.3% (11 of 72) of the patients due to the technical infeasibility of PBT (12.5% [9/72]) and withdrawal of informed consent (2.7% [2/72]). Of the patients showing infeasibility to PBT, RFA was feasible in 66.7% (6 of 9) of the patients and infeasible in 33.3% (3 of 9). The rate of crossover to the other treatment due to technical infeasibility was significantly higher in RFA than PBT (19/72 [26.4%] vs. 6/72 [8.3%], p <0.05); owing to the single-institution nature of this study and the fact that RFA was conducted by an expert with more than 10 years of experience, PBT showed better feasibility than RFA. RFA is generally limited by echogenicity, sub-phrenic location and proximity of vascular and biliary structures and PBT can be limited by the proximity of tumor(s) to radiosensitive gastrointestinal structures. Conversely, RFA is relatively not limited by proximity to gastrointestinal structures and PBT is relatively not limited by echogenicity, sub-phrenic location, and proximity of vascular and biliary structures. These findings suggested that RFA and PBT could be used in a complementary approach by cross-covering the technical infeasibility of each arm. However, 3 patients from each arm received another treatment because neither PBT nor RFA were feasible after randomization (Fig. 1); this is a limitation of study design.
      RFA, as a first-line treatment for small HCC, has shown excellent outcomes, with 2- or 3-year LPFS rates of 70.2–96.8%.
      • Kim Y.S.
      • Lim H.K.
      • Rhim H.
      • Lee M.W.
      • Choi D.
      • Lee W.J.
      • et al.
      Ten-year outcomes of percutaneous radiofrequency ablation as first-line therapy of early hepatocellular carcinoma: analysis of prognostic factors.
      ,
      • Lencioni R.A.
      • Allgaier H.P.
      • Cioni D.
      • Olschewski M.
      • Deibert P.
      • Crocetti L.
      • et al.
      Small hepatocellular carcinoma in cirrhosis: randomized comparison of radio-frequency thermal ablation versus percutaneous ethanol injection.
      • Lin S.M.
      • Lin C.J.
      • Lin C.C.
      • Hsu C.W.
      • Chen Y.C.
      Randomised controlled trial comparing percutaneous radiofrequency thermal ablation, percutaneous ethanol injection, and percutaneous acetic acid injection to treat hepatocellular carcinoma of 3 cm or less.
      • Nakazawa T.
      • Kokubu S.
      • Shibuya A.
      • Ono K.
      • Watanabe M.
      • Hidaka H.
      • et al.
      Radiofrequency ablation of hepatocellular carcinoma: correlation between local tumor progression after ablation and ablative margin.
      • Shiina S.
      • Tateishi R.
      • Arano T.
      • Uchino K.
      • Enooku K.
      • Nakagawa H.
      • et al.
      Radiofrequency ablation for hepatocellular carcinoma: 10-year outcome and prognostic factors.
      ,
      • Kono M.
      • Inoue T.
      • Kudo M.
      • Chishina H.
      • Arizumi T.
      • Takita M.
      • et al.
      Radiofrequency ablation for hepatocellular carcinoma measuring 2 cm or smaller: results and risk factors for local recurrence.
      • Lee L.H.
      • Hwang J.I.
      • Cheng Y.C.
      • Wu C.Y.
      • Lee S.W.
      • Yang S.S.
      • et al.
      Comparable outcomes of ultrasound versus computed tomography in the guidance of radiofrequency ablation for hepatocellular carcinoma.
      • Lee D.H.
      • Lee J.M.
      • Lee J.Y.
      • Kim S.H.
      • Han J.K.
      • Choi B.I.
      Radiofrequency ablation for intrahepatic recurrent hepatocellular carcinoma: long-term results and prognostic factors in 168 patients with cirrhosis.
      As a second-line treatment for rHCC, it has shown a 2-year LPFS rate of 62.5–68.5%.
      • Koh Y.H.
      • Choi J.I.
      • Kim H.B.
      • Kim M.J.
      Computed tomographic-guided radiofrequency ablation of recurrent or residual hepatocellular carcinomas around retained iodized oil after transarterial chemoembolization.
      ,
      • Kim N.
      • Kim H.J.
      • Won J.Y.
      • Kim D.Y.
      • Han K.H.
      • Jung I.
      • et al.
      Retrospective analysis of stereotactic body radiation therapy efficacy over radiofrequency ablation for hepatocellular carcinoma.
      Although this study enrolled rHCC patients and 46.4% (26 of 56) of the patients in the RFA arm had a history of other local treatments for target lesions (Table 1), RFA showed comparable outcomes with 2- and 3-year LPFS rates of 83.9% and 77.6%, respectively, similar to previous reports.
      • Kim Y.S.
      • Lim H.K.
      • Rhim H.
      • Lee M.W.
      • Choi D.
      • Lee W.J.
      • et al.
      Ten-year outcomes of percutaneous radiofrequency ablation as first-line therapy of early hepatocellular carcinoma: analysis of prognostic factors.
      ,
      • Koh Y.H.
      • Choi J.I.
      • Kim H.B.
      • Kim M.J.
      Computed tomographic-guided radiofrequency ablation of recurrent or residual hepatocellular carcinomas around retained iodized oil after transarterial chemoembolization.
      ,
      • Shiina S.
      • Tateishi R.
      • Arano T.
      • Uchino K.
      • Enooku K.
      • Nakagawa H.
      • et al.
      Radiofrequency ablation for hepatocellular carcinoma: 10-year outcome and prognostic factors.
      ,
      • Kim N.
      • Kim H.J.
      • Won J.Y.
      • Kim D.Y.
      • Han K.H.
      • Jung I.
      • et al.
      Retrospective analysis of stereotactic body radiation therapy efficacy over radiofrequency ablation for hepatocellular carcinoma.
      Other studies with PBT showed outcomes with 2- or 3-year LPFS rates of 71.4–96% in patients with HCC.
      • Bush D.A.
      • Kayali Z.
      • Grove R.
      • Slater J.D.
      The safety and efficacy of high-dose proton beam radiotherapy for hepatocellular carcinoma: a phase 2 prospective trial.
      • Kim T.H.
      • Park J.W.
      • Kim Y.J.
      • Kim B.H.
      • Woo S.M.
      • Moon S.H.
      • et al.
      Phase I dose-escalation study of proton beam therapy for inoperable hepatocellular carcinoma.
      • Hong T.S.
      • Wo J.Y.
      • Yeap B.Y.
      • Ben-Josef E.
      • McDonnell E.I.
      • Blaszkowsky L.S.
      • et al.
      Multi-institutional phase II study of high-dose hypofractionated proton beam therapy in patients with Localized, unresectable hepatocellular carcinoma and intrahepatic cholangiocarcinoma.
      ,
      • Kawashima M.
      • Furuse J.
      • Nishio T.
      • Konishi M.
      • Ishii H.
      • Kinoshita T.
      • et al.
      Phase II study of radiotherapy employing proton beam for hepatocellular carcinoma.
      ,
      • Fukumitsu N.
      • Sugahara S.
      • Nakayama H.
      • Fukuda K.
      • Mizumoto M.
      • Abei M.
      • et al.
      A prospective study of hypofractionated proton beam therapy for patients with hepatocellular carcinoma.
      In this study, 47.5% (38 of 80) of the patients in the PBT arm had a history of prior local treatment for target lesions, and despite the inclusion of a broader PTV than that used in other studies, PBT showed comparable outcomes, with 2- and 3-year LPFS rates of 94.8% and 88.3%, respectively, similar to previous reports.
      On comparing the LPFS, PFS, and OS values, we observed different patterns in the PP and ITT populations (Fig. 2). Although there was no statistical difference (Table 1), the subtle discrepancies in the patients' characteristics between the 2 populations may contribute to these outcome differences; after crossover, the proportion of patients with tumor sizes > 2 cm decreased (PBT 13.9% vs. RFA 7.1%) and that of patients with AJCC stage I disease increased (PBT 26.3% vs. RFA 32.1%) in the RFA arm (Table 1). These results are considered to be limitations associated with the crossover trial design. However, PBT and RFA yielded OS rates that were similar to those observed for RFA in randomized trials comparing the efficacy of RFA and surgical resection in early HCC.
      • Feng K.
      • Yan J.
      • Li X.
      • Xia F.
      • Ma K.
      • Wang S.
      • et al.
      A randomized controlled trial of radiofrequency ablation and surgical resection in the treatment of small hepatocellular carcinoma.
      • Chen M.S.
      • Li J.Q.
      • Zheng Y.
      • Guo R.P.
      • Liang H.H.
      • Zhang Y.Q.
      • et al.
      A prospective randomized trial comparing percutaneous local ablative therapy and partial hepatectomy for small hepatocellular carcinoma.
      • Xia Y.
      • Li J.
      • Liu G.
      • Wang K.
      • Qian G.
      • Lu Z.
      • et al.
      Long-term effects of repeat hepatectomy vs. percutaneous radiofrequency ablation among patients with recurrent hepatocellular carcinoma: a randomized clinical trial.
      • Huang J.
      • Yan L.
      • Cheng Z.
      • Wu H.
      • Du L.
      • Wang J.
      • et al.
      A randomized trial comparing radiofrequency ablation and surgical resection for HCC conforming to the Milan criteria.
      Although the non-inferiority of PBT compared to RFA was not shown in terms of the 2-year PFS as well as 3- and 4-year OS in the PP analysis (Fig. 2), the HRs in the PFS and OS between RFA and PBT were consistently not significant in both the ITT and PP analyses (p >0.05 each) (Fig. 3E–H). In the best tumor response evaluation, there was a lag in local control with PBT, unlike RFA, so the median time to the best tumor response was 4.4 months (range 1–13). The CR of the PBT arm was only 83.5%, because the criterion was RECIST. If it were based on modified RECIST, it would be possible to obtain a higher CR rate (Fig. S1).
      This study has several limitations. First, the primary outcome measure was 2-year LPFS, rather than PFS or OS. As frequent intrahepatic recurrence is a biological characteristic of HCC, which results in subsequent treatments that may affect survival outcomes,
      • Moon H.
      • Choi J.E.
      • Lee I.J.
      • Kim T.H.
      • Kim S.H.
      • Ko Y.H.
      • et al.
      All-treatment array of hepatocellular carcinoma from initial diagnosis to death: observation of cumulative treatments.
      the most important goal of locoregional therapy such as RFA and PBT is local control, including LPFS. Considering the presence of a moderate correlation between PFS or time to progression and OS in advanced HCC,
      • Llovet J.M.
      • Montal R.
      • Villanueva A.
      Randomized trials and endpoints in advanced HCC: role of PFS as a surrogate of survival.
      LPFS was used as the primary outcome measure in this RCT. Second, this study had a single-center design and predominantly included patients with chronic hepatitis B; in order for our results to be generalizable, further studies must be conducted across other institutions including patients with various etiologies. Only a percutaneous RFA with ultrasound or CT guidance (the most popular method in real-world practice) was performed in this study. However, laparoscopic RFA and/or multi-bipolar RFA with real-time ultrasound/CT or ultrasound/MRI fusion image navigation can improve the feasibility and outcomes of RFA.
      • Seror O.
      • N'Kontchou G.
      • Nault J.C.
      • Rabahi Y.
      • Nahon P.
      • Ganne-Carrie N.
      • et al.
      Hepatocellular carcinoma within Milan criteria: no-touch multibipolar radiofrequency ablation for treatment-long-term results.
      • Ahn S.J.
      • Lee J.M.
      • Lee D.H.
      • Lee S.M.
      • Yoon J.H.
      • Kim Y.J.
      • et al.
      Real-time US-CT/MR fusion imaging for percutaneous radiofrequency ablation of hepatocellular carcinoma.
      • Song K.D.
      • Lee M.W.
      • Rhin H.
      • Kang T.W.
      • Cha D.I.
      • Sinn D.H.
      • et al.
      Percutaneous US/MRI fusion-guided radiofrequency ablation for recurrent subcentimeter hepatocellular carcinoma: technical feasibility and therapeutic outcomes.
      During study design, due to a lack of LPFS data on RFA vs. placebo, we assumed a non-inferiority margin of 15% based on the RCT data of RFA vs. PEI.
      • Lin S.M.
      • Lin C.J.
      • Lin C.C.
      • Hsu C.W.
      • Chen Y.C.
      Randomised controlled trial comparing percutaneous radiofrequency thermal ablation, percutaneous ethanol injection, and percutaneous acetic acid injection to treat hepatocellular carcinoma of 3 cm or less.
      This margin seems to be relatively wide, but the lower boundary of CI of actual differences (PBT minus RFA) of 2-, 3-, and 4-year LPFS in the PP population was within - 5%. Lastly, PBT entails high costs and exclusive equipment, so our results cannot be applied to all patients with HCC.
      In conclusion, this prospective randomized study demonstrated that PBT is associated with LPFS rates that are comparable to those observed for RFA in patients with rHCC with ≤2 tumor(s) of <3 cm. PBT was also tolerable and safe, consistent with the known profile. The associated good feasibility and comparable clinical outcomes suggest that PBT may be a promising treatment option for small HCC.

      Abbreviations

      AE, adverse events; AJCC, American Joint Committee on Cancer; CR, complete response; HCC, hepatocellular carcinoma; HR, hazard ratio; ITT, intention-to-treat; ITV, internal target volume; LPFS, local progression-free survival; NCC, National Cancer Center; OS, overall survival; PEI, percutaneous ethanol injection; PBT, proton beam radiotherapy; PFS, progression-free survival; PP, per-protocol; PTV, planning target volumes; RCT, randomized controlled trial; RFA, radiofrequency ablation; rHCC, recurrent/residual HCC.

      Financial support

      This study was supported by the National Cancer Center Grant, Korea (NCC 1810271 , 1810031 and 1710030 ). The funding source had no role in the study design, data curation, or the analysis and interpretation of data.

      Authors' contributions

      Conception and study design: J-WP, THK, YHK; Funding acquisition: J-WP; Supervision: J-WP; Investigation and data curation: J-WP, THK, YHK, BHK, JHL, and MJK, Data analysis and interpretation: BP, THK, JWP, YHK, BHK; Writing – original draft: THK; Writing-review and editing: J-WP, THK, YHK, BHK, JHL, MJK, and BP. All authors reviewed and approved the final draft.

      Data availability statement

      Informed consent was not obtained from patients to share raw information in a public repository.

      Conflict of interest

      J-WP has served in a consulting or advisory role for Roche, Genetech, BMS, Bayer, Eisai, Ipsen, AstraZeneca; received honoraria from Bayer, Eisai; and participated in research sponsored by Ono-BMS, AstraZeneca, Blueprint, Roche, Eisai, Exelicis, Kowa, Merk. BHK has participated in research sponsored by Ono-BMS and received honoraria from Abbvie. The other authors declare that they have no conflict of interest related to this manuscript.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Acknowledgments

      Byung-Ho Nam, Ph.D., President, HERINGS, Korea, assisted with the study design, sample size calculation and analysis plan. Ju Hee Lee's current affiliation is the Department of Radiology, Asan Medical Center, Korea, and Min Ju Kim's current affiliation is the Department of Radiology, Ehwa's Women's University Hospital, Korea. We are grateful to all APROH trial patients, their family and collaborators of APROH trials.

      References

        • Akinyemiju T.
        • Abera S.
        • Ahmed M.
        • Alam N.
        • Alemayohu M.A.
        • Allen C.
        • et al.
        • Global Burden of Disease Liver Cancer Collaboration
        The burden of primary liver cancer and underlying etiologies from 1990 to 2015 at the global, regional, and national level: results from the Global Burden of Disease study 2015.
        JAMA Oncol. 2017; 3: 1683-1691
        • Park J.W.
        • Chen M.
        • Colombo M.
        • Roberts L.R.
        • Schwartz M.
        • Chen P.J.
        • et al.
        Global patterns of hepatocellular carcinoma management from diagnosis to death: the BRIDGE Study.
        Liver Int. 2015; 35: 2155-2166
        • European Association for the Study of the Liver
        EASL clinical practice guidelines: management of hepatocellular carcinoma.
        J Hepatol. 2018; 69: 182-236
        • Korean Liver Cancer Association, National Cancer Center
        2018 Korean Liver Cancer Association-National Cancer Center Korea practice guidelines for the management of hepatocellular carcinoma.
        Gut and liver. 2019; 13: 227-299
        • Marrero J.A.
        • Kulik L.M.
        • Sirlin C.B.
        • Zhu A.X.
        • Finn R.S.
        • Abecassis M.M.
        • et al.
        Diagnosis, staging, and management of hepatocellular carcinoma: 2018 practice guidance by the American Association for the Study of Liver Diseases.
        Hepatology. 2018; 68: 723-750
        • Levin W.P.
        • Kooy H.
        • Loeffler J.S.
        • DeLaney T.F.
        Proton beam therapy.
        Br J Cancer. 2005; 93: 849-854
        • Zietman A.L.
        • Bae K.
        • Slater J.D.
        • Shipley W.U.
        • Efstathiou J.A.
        • Coen J.J.
        • et al.
        Randomized trial comparing conventional-dose with high-dose conformal radiation therapy in early-stage adenocarcinoma of the prostate: long-term results from proton radiation oncology group/american college of radiology 95-09.
        J Clin Oncol. 2010; 28: 1106-1111
        • Bush D.A.
        • Kayali Z.
        • Grove R.
        • Slater J.D.
        The safety and efficacy of high-dose proton beam radiotherapy for hepatocellular carcinoma: a phase 2 prospective trial.
        Cancer. 2011; 117: 3053-3059
        • Kim T.H.
        • Park J.W.
        • Kim Y.J.
        • Kim B.H.
        • Woo S.M.
        • Moon S.H.
        • et al.
        Phase I dose-escalation study of proton beam therapy for inoperable hepatocellular carcinoma.
        Cancer Res Teat. 2015; 47: 34-45
        • Hong T.S.
        • Wo J.Y.
        • Yeap B.Y.
        • Ben-Josef E.
        • McDonnell E.I.
        • Blaszkowsky L.S.
        • et al.
        Multi-institutional phase II study of high-dose hypofractionated proton beam therapy in patients with Localized, unresectable hepatocellular carcinoma and intrahepatic cholangiocarcinoma.
        J Clin Oncol. 2016; 34: 460-468
        • Kim T.H.
        • Park J.W.
        • Kim B.H.
        • Oh E.S.
        • Youn S.H.
        • Moon S.H.
        • et al.
        Phase II study of hypofractionated proton beam therapy for hepatocellular carcinoma.
        Front Oncol. 2020; 10: 542
        • Kawashima M.
        • Furuse J.
        • Nishio T.
        • Konishi M.
        • Ishii H.
        • Kinoshita T.
        • et al.
        Phase II study of radiotherapy employing proton beam for hepatocellular carcinoma.
        J Clin Oncol. 2005; 23: 1839-1846
        • Korean Liver Cancer Study Group, National Cancer Center
        Korean J Hepatol. 2009; 15: 391-423
        • Korean Liver Cancer Study Group, National Cancer Center
        2014 KLCSG-NCC Korea practice guideline for the management of hepatocellular carcinoma.
        Gut Liver. 2015; 9: 267-317
        • Fukumitsu N.
        • Sugahara S.
        • Nakayama H.
        • Fukuda K.
        • Mizumoto M.
        • Abei M.
        • et al.
        A prospective study of hypofractionated proton beam therapy for patients with hepatocellular carcinoma.
        Int J Radiat Oncol Biol Phys. 2009; 74: 831-836
        • Mizumoto M.
        • Okumura T.
        • Hashimoto T.
        • Fukuda K.
        • Oshiro Y.
        • Fukumitsu N.
        • et al.
        Proton beam therapy for hepatocellular carcinoma: a comparison of three treatment protocols.
        Int J Radiat Oncol Biol Phys. 2011; 81: 1039-1045
        • Eisenhauer E.A.
        • Therasse P.
        • Bogaerts J.
        • Schwartz L.H.
        • Sargent D.
        • Ford R.
        • et al.
        New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1).
        Eur J Cancer. 2009; 45: 228-247
        • Parpia S.
        • Julian J.A.
        • Thabane L.
        • Gu C.
        • Whelan T.J.
        • Levine M.N.
        Treatment crossovers in time-to-event non-inferiority randomised trials of radiotherapy in patients with breast cancer.
        BMJ Open. 2014; 4: e006531
        • Choi D.
        • Lim H.K.
        • Rhim H.
        • Kim Y.S.
        • Yoo B.C.
        • Paik S.W.
        • et al.
        Percutaneous radiofrequency ablation for recurrent hepatocellular carcinoma after hepatectomy: long-term results and prognostic factors.
        Ann Surg Oncol. 2007; 14: 2319-2329
        • Kim Y.S.
        • Lim H.K.
        • Rhim H.
        • Lee M.W.
        • Choi D.
        • Lee W.J.
        • et al.
        Ten-year outcomes of percutaneous radiofrequency ablation as first-line therapy of early hepatocellular carcinoma: analysis of prognostic factors.
        J Hepatol. 2013; 58: 89-97
        • Koh Y.H.
        • Choi J.I.
        • Kim H.B.
        • Kim M.J.
        Computed tomographic-guided radiofrequency ablation of recurrent or residual hepatocellular carcinomas around retained iodized oil after transarterial chemoembolization.
        Korean J Radiol. 2013; 14: 733-742
        • Lencioni R.A.
        • Allgaier H.P.
        • Cioni D.
        • Olschewski M.
        • Deibert P.
        • Crocetti L.
        • et al.
        Small hepatocellular carcinoma in cirrhosis: randomized comparison of radio-frequency thermal ablation versus percutaneous ethanol injection.
        Radiology. 2003; 228: 235-240
        • Lin S.M.
        • Lin C.J.
        • Lin C.C.
        • Hsu C.W.
        • Chen Y.C.
        Randomised controlled trial comparing percutaneous radiofrequency thermal ablation, percutaneous ethanol injection, and percutaneous acetic acid injection to treat hepatocellular carcinoma of 3 cm or less.
        Gut. 2005; 54: 1151-1156
        • Nakazawa T.
        • Kokubu S.
        • Shibuya A.
        • Ono K.
        • Watanabe M.
        • Hidaka H.
        • et al.
        Radiofrequency ablation of hepatocellular carcinoma: correlation between local tumor progression after ablation and ablative margin.
        AJR Am J Roentgenol. 2007; 188: 480-488
        • Shiina S.
        • Tateishi R.
        • Arano T.
        • Uchino K.
        • Enooku K.
        • Nakagawa H.
        • et al.
        Radiofrequency ablation for hepatocellular carcinoma: 10-year outcome and prognostic factors.
        Am J Gastroenterol. 2012; 107 (quiz 578): 569-577
        • Com-Nougue C.
        • Rodary C.
        • Patte C.
        How to establish equivalence when data are censored: a randomized trial of treatments for B non-Hodgkin lymphoma.
        Stat Med. 1993; 12: 1353-1364
        • Rhim H.
        • Lee M.H.
        • Kim Y.S.
        • Choi D.
        • Lee W.J.
        • Lim H.K.
        Planning sonography to assess the feasibility of percutaneous radiofrequency ablation of hepatocellular carcinomas.
        AJR Am J Roentgenol. 2008; 190: 1324-1330
        • Kono M.
        • Inoue T.
        • Kudo M.
        • Chishina H.
        • Arizumi T.
        • Takita M.
        • et al.
        Radiofrequency ablation for hepatocellular carcinoma measuring 2 cm or smaller: results and risk factors for local recurrence.
        Dig Dis. 2014; 32: 670-677
        • Lee L.H.
        • Hwang J.I.
        • Cheng Y.C.
        • Wu C.Y.
        • Lee S.W.
        • Yang S.S.
        • et al.
        Comparable outcomes of ultrasound versus computed tomography in the guidance of radiofrequency ablation for hepatocellular carcinoma.
        PLoS one. 2017; 12: e0169655
        • Lee D.H.
        • Lee J.M.
        • Lee J.Y.
        • Kim S.H.
        • Han J.K.
        • Choi B.I.
        Radiofrequency ablation for intrahepatic recurrent hepatocellular carcinoma: long-term results and prognostic factors in 168 patients with cirrhosis.
        Cardiovasc Intervent Radiol. 2014; 37: 705-715
        • Kim N.
        • Kim H.J.
        • Won J.Y.
        • Kim D.Y.
        • Han K.H.
        • Jung I.
        • et al.
        Retrospective analysis of stereotactic body radiation therapy efficacy over radiofrequency ablation for hepatocellular carcinoma.
        Radiother Oncol. 2019; 131: 81-87
        • Feng K.
        • Yan J.
        • Li X.
        • Xia F.
        • Ma K.
        • Wang S.
        • et al.
        A randomized controlled trial of radiofrequency ablation and surgical resection in the treatment of small hepatocellular carcinoma.
        J Hepatol. 2012; 57: 794-802
        • Chen M.S.
        • Li J.Q.
        • Zheng Y.
        • Guo R.P.
        • Liang H.H.
        • Zhang Y.Q.
        • et al.
        A prospective randomized trial comparing percutaneous local ablative therapy and partial hepatectomy for small hepatocellular carcinoma.
        Ann Surg. 2006; 243: 321-328
        • Xia Y.
        • Li J.
        • Liu G.
        • Wang K.
        • Qian G.
        • Lu Z.
        • et al.
        Long-term effects of repeat hepatectomy vs. percutaneous radiofrequency ablation among patients with recurrent hepatocellular carcinoma: a randomized clinical trial.
        JAMA Oncol. 2020; 6: 255-263
        • Huang J.
        • Yan L.
        • Cheng Z.
        • Wu H.
        • Du L.
        • Wang J.
        • et al.
        A randomized trial comparing radiofrequency ablation and surgical resection for HCC conforming to the Milan criteria.
        Ann Surg. 2010; 252: 903-912
        • Moon H.
        • Choi J.E.
        • Lee I.J.
        • Kim T.H.
        • Kim S.H.
        • Ko Y.H.
        • et al.
        All-treatment array of hepatocellular carcinoma from initial diagnosis to death: observation of cumulative treatments.
        J Cancer Res Clin Oncol. 2017; 143: 2327-2339
        • Llovet J.M.
        • Montal R.
        • Villanueva A.
        Randomized trials and endpoints in advanced HCC: role of PFS as a surrogate of survival.
        J Hepatol. 2019; 70: 1262-1277
        • Seror O.
        • N'Kontchou G.
        • Nault J.C.
        • Rabahi Y.
        • Nahon P.
        • Ganne-Carrie N.
        • et al.
        Hepatocellular carcinoma within Milan criteria: no-touch multibipolar radiofrequency ablation for treatment-long-term results.
        Radiology. 2016; 180: 611-621
        • Ahn S.J.
        • Lee J.M.
        • Lee D.H.
        • Lee S.M.
        • Yoon J.H.
        • Kim Y.J.
        • et al.
        Real-time US-CT/MR fusion imaging for percutaneous radiofrequency ablation of hepatocellular carcinoma.
        J Hepatol. 2017; 66: 347-354
        • Song K.D.
        • Lee M.W.
        • Rhin H.
        • Kang T.W.
        • Cha D.I.
        • Sinn D.H.
        • et al.
        Percutaneous US/MRI fusion-guided radiofrequency ablation for recurrent subcentimeter hepatocellular carcinoma: technical feasibility and therapeutic outcomes.
        Radiology. 2018; 288: 878-886