Oncology · Medical Literature
[Correspondence] Perspective on proton beam therapy for oropharyngeal cancer
Abstract
The TORPEdO trial1 represents a landmark contribution to oropharyngeal cancer management, demonstrating similar late quality-of-life outcomes between intensity-modulated proton therapy (IMPT) and intensity-modulated radiotherapy (IMRT). However, as head and neck surgeons involved in the salvage of recurrent disease, we wish to draw attention to a signal deserving closer scrutiny.
Clinical relevance
From the perspective of head and neck salvage surgeons, does the TORPEdO trial's demonstration of similar 12-month quality-of-life outcomes between intensity-modulated proton therapy (IMPT) and intensity-modulated radiotherapy (IMRT) for locally advanced oropharyngeal squamous cell carcinoma (OPSCC) adequately capture all clinically relevant signals — specifically those pertaining to disease recurrence patterns and surgical salvageability? __SNAPSHOT__ {"studyDesign":"Published correspondence (expert perspective/commentary) responding to the TORPEdO phase 3, multicenter, open-label, randomized controlled trial (Thomson et al., Lancet 2026)","setting":"Not applicable (correspondence); the referenced TORPEdO trial was conducted across 20 UK National Health Service hospitals","sampleSize":"Not applicable (correspondence); the referenced TORPEdO trial enrolled n=205 (IMPT: n=136, IMRT: n=69)","population":"Patients with newly diagnosed locally advanced oropharyngeal squamous cell carcinoma suitable for definitive concurrent chemoradiotherapy including bilateral neck treatment; key exclusion criteria included pre-treatment feeding tube dependence, N3 disease, upfront neck dissection, induction chemotherapy, contraindication to cisplatin, and prior head and neck radiotherapy","intervention":"Intensity-modulated proton therapy (IMPT): 70 Gy in 33 fractions over 6.5 weeks with two cycles of high-dose cisplatin (100 mg/m² every 3 weeks)","comparator":"Intensity-modulated radiotherapy (IMRT): same dose, fractionation, and concurrent cisplatin regimen","primaryOutcome":"Co-primary endpoints at 12 months: (1) clinician-reported gastrostomy-tube dependence or CTCAE Grade 3 weight loss (≥20% from baseline); (2) patient-reported University of Washington Quality of Life (UW-QoL) physical composite score (saliva, taste, chewing, swallowing, speech, appearance). The correspondence draws attention to signals within the trial's secondary and ancillary data — specifically recurrence patterns and their implications for salvage surgery — not captured by the co-primary endpoints.","followUp":"Median follow-up of approximately 27–28.3 months (as reported in the primary trial); follow-up duration for salvage-relevant outcomes referenced in the correspondence is not separately reported in the abstract","fundingSource":"The correspondence itself: not reported. The referenced TORPEdO trial was funded by Cancer Research UK with additional support from The Taylor Family Foundation (non-profit/government)"}
Key findings
This is a published correspondence, not an original study; it generates no independent primary data. The authors comment on the TORPEdO trial (Thomson et al., Lancet 2026), which found no statistically significant difference between IMPT and IMRT on either co-primary endpoint at 12 months: (1) UW-QoL physical composite score: 78.3 (IMPT) vs. 77.1 (IMRT), difference 1.3 points, 97.5% CI −3.7 to 6.2, p=0.563; (2) gastrostomy-tube dependence: 1.7% vs. 1.7%, respectively. Of note, grade 3 weight loss was numerically higher in the IMPT arm (18.2% vs. 5.7%; OR 2.8, 97.5% CI 0.8–10.4, p=0.080), a finding that did not meet the pre-specified 2.5% significance threshold but represents a clinically non-trivial absolute difference of 12.5 percentage points (calculated). At a median follow-up of ~27 months, 2-year freedom from loco-regional recurrence was 94.8% (IMPT) vs. 96.8% (IMRT) and 2-year overall survival was 94.4% vs. 94.9%, respectively — no meaningful between-group difference. The correspondence authors, writing from the perspective of head and neck salvage surgeons, flag a signal within the trial's data that they argue deserves closer scrutiny, though the specific nature of that signal (e.g., patterns of recurrence, salvageability, wound healing, or tissue characteristics relevant to surgical salvage) is not detailed in the supplied abstract and cannot be further characterized without access to the full correspondence text.
Practice implications
This correspondence represents an expert surgical perspective on the TORPEdO trial and does not establish new evidence independently. The TORPEdO trial itself established that, in a rigorously quality-assured UK NHS setting, IMPT does not improve 12-month patient-reported quality of life or reduce long-term gastrostomy dependence compared with high-quality IMRT for locally advanced OPSCC requiring bilateral chemoradiation. The trial's authors concluded that IMRT remains the standard of care in settings where IMPT is not routinely available. What the trial did NOT establish: (1) superiority or inferiority in recurrence patterns specific to surgical salvage candidacy; (2) long-term (>2 year) outcomes, given the relatively short median follow-up; (3) outcomes in N3 disease, patients receiving induction chemotherapy, or those requiring upfront neck dissection; (4) relative outcomes in HPV-negative OPSCC, which were underrepresented given enrollment demographics (67.8% had <10 pack-years). The correspondence authors' concern — drawn from their surgical salvage experience — highlights that aggregate 12-month QoL and short-term locoregional control rates may not fully capture outcomes relevant to patients who recur: specifically, whether the tissue biology, dosimetric footprint, or immune microenvironment modifications differ between IMPT and IMRT in ways that affect salvageability or surgical morbidity after recurrence. This is a clinically important, underexplored question that the TORPEdO trial was not designed to answer. Only the abstract of this correspondence was available for review; full clinical arguments, any supporting data, and specific proposed signals could not be independently assessed. __LIMITATIONS__ 1. This is a correspondence letter, the lowest tier of evidence; it introduces no primary data, statistical analyses, or pre-specified hypotheses. 2. Only the abstract text was available for analysis; the full text of the letter — including the specific 'signal deserving closer scrutiny' identified by the authors — was not accessible, preventing complete critical appraisal. 3. The authors' perspective is drawn from their role as salvage surgeons, introducing a potential selection bias: clinicians who manage the most complex and morbid recurrences may perceive the radiotherapy modality question differently than those treating the broader population. 4. The referenced TORPEdO trial itself has important limitations relevant to the correspondence's concerns: (a) relatively short median follow-up (~27 months) limits assessment of late recurrences, second primaries, and very late toxicities; (b) the 2:1 randomization design, chosen for operational efficiency, produced an underpowered control arm (n=69), reducing precision for subgroup analyses; (c) the trial was open-label, precluding blinding; (d) the trial was restricted to patients requiring bilateral neck treatment, limiting generalizability to unilateral disease where dosimetric advantages of IMPT may be more pronounced; (e) the trial was conducted exclusively in UK NHS centers with stringent quality assurance, potentially overestimating IMRT performance in lower-resource settings; (f) salvage surgery rates, outcomes, and tissue characteristics post-recurrence were not reported as primary or secondary endpoints. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ De Luca, Bussu, and Camaioni raise a clinically important surgical perspective: the TORPEdO trial's demonstration of equivalent 12-month quality-of-life outcomes between IMPT and IMRT for locally advanced OPSCC does not preclude the possibility of meaningful differences in recurrence patterns or surgical salvageability that aggregate patient-reported outcome measures would not capture. Clinicians should interpret TORPEdO's equivalence finding within its defined endpoints and follow-up window, recognizing that the question of whether radiation modality affects the biology or technical feasibility of salvage after recurrence remains unanswered by current phase 3 data. Further research specifically addressing recurrence patterns, surgical margins, wound healing, and functional outcomes after salvage surgery stratified by primary radiation modality is warranted before modality selection decisions can be optimized for the subset of patients who will ultimately require salvage.