The multicenter NRG-GI003 trial, presented at the ASTRO meeting, failed to show a survival benefit for proton therapy over conventional photon radiotherapy in locally advanced hepatocellular carcinoma. Patients receiving photons lived more than twice as long, with median overall survival of 54.9 months versus 26.6 months for protons. The trial was halted at an interim analysis after meeting futility criteria, and progression-free survival and local progression rates did not differ between the groups.
Proton therapy did show a dosimetric advantage, significantly reducing radiation exposure to uninvolved liver tissue and yielding a numerically lower rate of severe treatment-related adverse events (11% versus 24%). However, a subgroup analysis found that photons dramatically outperformed protons for tumors 5 cm or larger, a result that surprised investigators. Theodore Hong of Dana-Farber Cancer Institute noted that the photon arm performed "noticeably better than expected," likely due to improvements in systemic therapy.
Discussants raised questions about the trial's design. Erqi Pollom of Stanford Medicine pointed out that only 58% of proton patients received the highest planned dose, and the isotoxic design may have limited dose escalation. Marco Schwarz of the University of Washington argued that the trade-offs between target dose, liver dose, and tumor volume meant protons spared the liver rather than boosting tumor dose. Both agreed the role of proton therapy in liver radiation remains an open and worthwhile question.