Phase III trial result — conference presentation (ASTRO, 2026-09-28), not yet published as a paper Japan — proton therapy covered by public insurance for HCC ≥4 cm unsuitable for surgery (since 2022-04)
The first randomized phase III comparison of proton therapy for liver cancer versus standard photon (X-ray) radiation, the US trial NRG-GI003, has reported initial results. In 115 patients with advanced hepatocellular carcinoma, proton therapy did not improve overall survival, progression-free survival or local progression compared with photons, and the results were released early after crossing a prespecified futility boundary. Protons did clearly lower the radiation dose to the healthy liver. The data were presented at the American Society for Radiation Oncology (ASTRO) meeting in September 2026 and have not yet been published as a peer-reviewed paper.
What was presented: protons vs photons, the first phase III comparison in liver cancer
Proton beams deposit most of their energy at a set depth and then stop (the Bragg peak). This can reduce the dose to normal tissue behind the tumour, and many expected a particular advantage in the liver, an organ sensitive to radiation. Until now, support for this came mainly from single-centre retrospective analyses.
NRG-GI003 (NCT03186898), funded by the US National Cancer Institute and run by NRG Oncology, was designed to test the question by randomization. The hypothesis was that sparing the normal liver would reduce liver decompensation and therefore improve survival. The principal investigator is Dr. Theodore S. Hong of Dana-Farber Cancer Institute and Beth Israel Deaconess Medical Center. Results were presented in the plenary session of the ASTRO Annual Meeting in Boston on 28 September 2026.
Who took part: advanced HCC not suitable for surgery
- Unresectable or locally recurrent hepatocellular carcinoma (the most common primary liver cancer)
- 1–3 tumours, liver function Child-Pugh A or B7 (relatively preserved), good performance status
- Tumour vascular thrombus allowed when there was a single lesion
- Prior systemic or liver-directed treatment allowed
Enrolment ran from June 2017 to January 2026; 115 patients were analysed (57 protons, 58 photons). The median age was 72 and 93% were Child-Pugh A. Treatment was given in 5 fractions (30–50 Gy) or 15 fractions (37.5–67.5 Gy), with the dose individualized according to mean normal-liver dose. The original target of 167 patients was reduced because of slow accrual.
Results in numbers
| Measure | Proton therapy | Photon (X-ray) therapy |
|---|---|---|
| 2-year overall survival | 56.2% | 69.3% |
| Overall survival hazard ratio (HR) | 1.30 (90% CI 0.83–2.04) — no benefit from protons | |
| 2-year progression-free survival | 31.6% | 30.4% |
| 2-year local progression (treated tumour regrew) | 19.8% | 22.3% |
| Treatment-related grade ≥3 adverse events | 11% | 24% |
| Mean normal-liver dose (5 fractions) | 8.5 Gy | 12.8 Gy |
Source: ASTRO 2026 plenary abstract (Hong TS et al.), median follow-up 20 months. The difference in adverse events was not statistically significant (p=0.09). Dose figures as reported from the presentation by CancerNetwork.
In short, proton therapy was not better than photons for survival or recurrence. Median survival was reported as 26.6 months with protons and 54.9 months with photons, but because this is an interim analysis it is also hard to conclude that protons are worse. The investigators noted that the photon arm did far better than the roughly 14-month median assumed when the trial was designed, which may reflect progress in photon stereotactic radiotherapy (SBRT) and overall care.
Protons did reduce the normal-liver dose (12.7 Gy vs 20.3 Gy with 15 fractions), and severe side effects were numerically about half (mainly laboratory changes such as low lymphocyte counts). There were no treatment-related deaths in either arm. A post hoc analysis suggested a trend favouring protons for tumours under 5 cm and photons for tumours of 5 cm or more, but Dr. Hong stressed that this is hypothesis-generating rather than definitive. Quality-of-life and fatigue results have not yet been reported.
Status in Japan and abroad: an approved treatment, with evidence still accumulating
- Japan: under the April 2022 fee-schedule revision, proton therapy for hepatocellular carcinoma with a longest diameter of 4 cm or more that cannot be cured by surgery became covered by public health insurance (Fukui Prefectural Hospital Proton Therapy Center). Tumours under 4 cm are not covered, so the route and cost must be checked with each institution.
- United States: proton therapy is already in clinical use; this result means it has not shown longer survival than photons in liver cancer, which may affect insurance coverage decisions.
- Limitations: interim analysis, fewer patients than planned, and no peer-reviewed paper yet. It does not mean that patients who have had or are planning proton therapy must change treatment immediately.
The differences between insurance, Advanced Medical Care and self-pay proton therapy in Japan are explained in our guide to proton therapy in Japan and proton therapy information. For translation of Japanese medical documents or medical interpreting, please contact FTH.
What to check and questions for your doctor
- Given my tumour size, number and any vessel invasion, which radiation therapy suits me? What are the pros and cons of protons vs photons (including SBRT)?
- With my liver function (Child-Pugh score), how important is sparing the normal liver?
- What other treatments could be used instead of, or with, radiation (surgery, radiofrequency ablation, embolization, drug therapy)?
- If treated in Japan, do I meet the insurance conditions (longest diameter 4 cm or more, etc.)? If not, what will it cost?
- Which tests will check my liver function and recurrence after treatment, and how often?
Frequently asked questions
Is proton therapy better than standard radiation for liver cancer?
In the initial NRG-GI003 results, overall survival, progression-free survival and local progression were not better than with photon radiation. Because the dose to the healthy liver was lower, possible advantages in individual cases, such as weaker liver function, should be discussed with your doctor.
Does this mean proton therapy is dangerous?
No. Severe side effects were numerically lower with protons (11% vs 24%, not statistically significant), and there were no treatment-related deaths in either arm. The gap in median survival comes from an interim analysis and should be interpreted with caution.
Is proton therapy for liver cancer covered by insurance in Japan?
Since April 2022, hepatocellular carcinoma with a longest diameter of 4 cm or more that cannot be cured by surgery is covered by public insurance. For international patients, costs depend on whether they are enrolled in Japanese public insurance, so check directly with the treating institution.
When will the results be published as a paper?
As of the source check date (2026-10-15), only the conference abstract and presentation are available; no full paper has been published. Quality-of-life results are expected later.
Related: Proton therapy in Japan · Colon cancer immunotherapy FDA approval · Preparing for interpreting before a hospital visit in Japan
Sources
- ASTRO 2026 Annual Meeting, Plenary — Hong TS, Winter K, Koay EJ, et al. Initial Results of NRG Oncology NRG-GI003: A Phase III Randomized Trial of Protons versus Photons for Hepatocellular Carcinoma (2026-09-28)
amportal.astro.org
Checked on: 2026-10-15 - ClinicalTrials.gov — Radiation Therapy With Protons or Photons in Treating Patients With Liver Cancer (NCT03186898)
clinicaltrials.gov/study/NCT03186898
Checked on: 2026-10-15 - 福井県立病院 陽子線がん治療センター — ふくい陽子線治療ニュース 第11号（2022年5月、保険適用の拡大）
pref.fukui.lg.jp
Checked on: 2026-10-15 - CancerNetwork — Proton Therapy Does Not Improve Survival in Hepatocellular Carcinoma (2026-09-29, report of the presentation)
cancernetwork.com
Checked on: 2026-10-15
Written by the FTH editorial team. Sources checked 2026-10-15 · Medical review: not applicable
This article is general information based on public sources and does not replace diagnosis or treatment. Please discuss treatment decisions with your medical team.
