Landmark Trial and Expanding Horizons: Key Developments in Particle Therapy, Spring 2026

The first quarter of 2026 has brought particle therapy its most rigorously tested clinical evidence to date — a phase 3 randomised trial comparing protons with photons in a common adult malignancy — alongside large-scale registries, mature long-term outcomes, and technology that is quietly expanding the boundaries of what is possible in the treatment room. Taken together, these developments mark a transition for the field: from establishing feasibility and dosimetric advantage toward defining where proton therapy changes patient outcomes, and where it does not.

The TORPEdO Trial: A Milestone for Proton Evidence

No paper in this collection carries more weight than TORPEdO, the phase 3 randomised controlled trial comparing intensity-modulated proton therapy (IMPT) with intensity-modulated radiotherapy (IMRT) in 205 patients with locally advanced oropharyngeal squamous cell carcinoma. Conducted across 20 UK National Health Service hospitals, the trial randomly assigned patients in a 2:1 ratio to IMPT or IMRT, delivering 70 Gy in 33 fractions with concurrent cisplatin. The co-primary endpoints — assessed at 12 months — were gastrostomy-tube dependence or severe weight loss, and the University of Washington quality of life physical composite score covering saliva, taste, chewing, swallowing, speech, and appearance.

The results bear careful interpretation. The combined endpoint of gastrostomy dependence or severe weight loss occurred in 18% of the IMPT group and 6% of the IMRT group, a difference that favoured IMRT but did not reach statistical significance. Mean physical quality of life scores were nearly identical: 78.3 for IMPT versus 77.1 for IMRT. At a median follow-up of 28 months, locoregional control and overall survival were excellent and comparable between arms — 94% and 95% respectively in the IMPT group at two years. Serious treatment-related adverse events were numerically lower with IMPT (one versus four), but the numbers are small.

TORPEdO is both a landmark and a reality check. It demonstrates that IMPT can be delivered safely in a multicentre setting with disease control matching photon therapy, but it did not show the functional benefit that the field might have expected based on dosimetric comparisons. The low rates of gastrostomy dependence in both arms (2%) suggest that modern IMRT with careful organ-at-risk sparing already achieves good functional preservation in these patients. The trial does not close the door on protons for oropharyngeal cancer — the comparable efficacy is itself an important finding — but it shifts the burden of proof: if protons are to become the standard in this indication, future studies will need to identify subgroups that derive meaningful functional advantage, or demonstrate benefit in domains not captured by these endpoints.

Expanding Indications: Liver, Pancreas, and the Function-Sparing Frontier

If TORPEdO represents the field's most rigorous test in a common cancer, a complementary story is unfolding in indications where proton therapy's dosimetric advantages are harder to replicate with photons. A Japanese nationwide prospective study analysed proton beam therapy for liver metastases from pancreatic cancer — a notoriously difficult setting where treatment options are limited. The 24 patients enrolled had a median tumour size of 11.7 mL and had exhausted surgical and chemotherapeutic options. Treatment was well tolerated, with only two cases of radiation-induced dermatitis. The one-year local recurrence rate was 16.7%, corresponding to a local control rate of approximately 83%, while overall survival at one year was 37.5% with a median of 9.3 months. The study identified that levels of CA 19-9 and progression-free survival duration after PBT were significant predictors of overall survival, suggesting that patient selection will be critical as this indication is explored further.

The numbers from Samsung Medical Center in Seoul are far larger. A retrospective review of 1,823 patients with hepatocellular carcinoma who were ineligible for Barcelona Clinic Liver Cancer guideline-recommended standard treatments — a group that often has few local options — found that proton therapy achieved two-year freedom from local progression exceeding 90% across BCLC stages 0, A, B, and C. Stage C patients still managed 87.6% local control at two years and a three-year overall survival of 37.2%. With a median follow-up of 19.5 months and treatment volumes that increased annually, this study provides some of the strongest evidence to date that proton therapy can address an unmet need in a large, underserved patient population.

The function-sparing argument finds its clearest expression in craniopharyngioma management. A case series from Italy described eight patients treated with combined surgery and proton therapy, all of whom had hypothalamic or optic pathway involvement that made gross total resection unacceptably risky. Seven received 54 GyRBE in 30 fractions. At a median follow-up of 58.5 months, six patients achieved partial response, one complete response, and one remained stable. Visual acuity worsened in only one patient; endocrinological status improved in one and was stable in the rest. The authors argue — persuasively — for multimodal function-sparing management over aggressive surgery. The principle extends to paediatric neuroblastoma, where a systematic review and meta-analysis comparing proton therapy with photon radiotherapy found significantly better survival with protons: one-year overall survival of 97.2% versus 89.1%, and progression-free survival of 96.1% versus 78.6%. Meta-regression identified treatment modality as a significant factor for both two-year overall survival and progression-free survival, although the authors caution that variability in patient background and treatment protocols may have influenced the results.

Long-Term Data and Practical Reliability

Ten-year outcomes from a prospective registry of 582 prostate cancer patients treated with moderately hypofractionated proton therapy (70 GyRBE in 28 fractions for low-risk, 72.5 GyRBE in 29 fractions for intermediate-risk) provide the kind of maturity the field needs. Freedom from disease progression was 91.9% overall — 95.9% for low-risk and 88.4% for intermediate-risk patients — with local recurrence in only 3.6% of patients. Grade 3 gastrointestinal and genitourinary toxicities were low (1.4% and 2.7% respectively), and no grade 4 or higher events occurred. The proportion of patients developing second malignancies was 2.9% for cancers diagnosed more than five years after proton therapy, and only 1.7% for in-field or pelvic cancers. With a median follow-up exceeding ten years, these data establish a benchmark for both efficacy and safety in a common indication.

At the other end of the technology spectrum, a ten-year patterns-of-use analysis of the world's first commercial compact pencil beam scanning proton therapy system, installed in a community hospital in Shreveport, Louisiana, tracked 1,382 treatment courses. The system treated an average of 138 courses per year, with prostate and brain as the dominant indications. Notably, fractionation decreased steadily over the decade, and nearly 90% of patients lived within 100 miles of the facility. The system survived the COVID-19 pandemic and a three-month machine downtime in year eight, returning to its highest annual volume in year ten. This is a quiet but important piece of evidence: compact proton therapy can work in a community setting, and it can work reliably.

Technology and the Next Horizon

Preclinical capabilities continue to advance. Researchers at HollandPTC have developed a fully 3D-printed contoured double passive scattering system that shapes a 250 MeV proton beam into uniform fields suitable for ultra-high-dose-rate FLASH experiments. The system achieves dose rates above 40 Gy/s with a lateral field uniformity exceeding 95% — up to 100 Gy/s within the spread-out Bragg peak. This opens the door for systematic radiobiological investigation of the FLASH effect with protons.

A dosimetric comparison of IMPT and knowledge-based VMAT for left-sided breast cancer in 14 patients confirmed that protons reduce mean heart dose (0.6 Gy versus 1.7 Gy) but noted that with deep-inspiration breath hold and knowledge-based planning, the photon plan mean heart doses were already below 2 Gy in all cases. The clinical significance of further reduction at these levels remains uncertain, a point the authors acknowledge directly.

Radiobiology and the Future

A workshop report from the "Hadrontherapy for Life" symposium in Caen, France, articulated a vision for a new era of radiobiology in particle therapy — one that integrates molecular biology, omics technologies, and innovative experimental models such as organoids. The report identified tumour hypoxia as a challenge only partially mitigated by high-LET radiation, and argued that carbon ions — and high-LET particles more generally — should be understood as "new drugs" capable of promoting immunogenicity and systemic effects beyond local control. The C400 accelerator under development in Caen is designed to embed radiobiology from the outset, creating a European hub for coordinated investigation.

Meanwhile, narrative reviews of boron neutron capture therapy for high-grade CNS tumours and of the evolution from physical to dual targeting in particle radiotherapy point toward a future where biochemical selectivity — through tumour-specific boron-10 agents or other molecular strategies — combines with subcellular physical precision. These are early-stage visions, but they share a common thread: the field is looking beyond the single-modality proton beam toward combinations and integrations that address the fundamental limitations of radiation therapy.

What this quarter's evidence collectively suggests is that particle therapy is entering a period of productive tension. The landmark randomised trial did not deliver the expected functional advantage, but it did demonstrate that protons can match photons in efficacy for a common adult cancer — a finding that would have seemed aspirational a decade ago. The large registries in liver and pancreatic metastases show real-world value in populations with few alternatives. The long-term prostate data provide a safety benchmark that any new modality must match. And the technology continues to evolve, enabling new experimental models and expanding access to community settings.

The field now faces a clear challenge: to identify where the dosimetric advantages of protons translate into meaningful clinical benefit that justifies their cost and complexity, and to design the trials — whether randomised or registry-based — that can demonstrate that benefit with the rigour TORPEdO has set as the new standard.