KEY POINTS
- This prospective phase I study enrolled 10 patients with Tis–T2N0 squamous cell carcinoma of the glottic larynx; six had T1 disease and four T2, while 5/10 had recurrent disease after prior definitive endoscopic surgery. Median follow-up on study was 10.9 months.
- Rather than treating the whole larynx, the protocol targeted gross disease with a direct PTV expansion of 3 mm radially and 5 mm craniocaudally and no additional CTV. Patients with PTV <10 cc who were not active smokers received 42.5 Gy in 5 fractions, while larger-volume tumors or active smokers received 58.08 Gy in 16 fractions.
- Treatment used gated 0.35-T MR-guided RT with continuous cine-MRI tracking. All 10 patients completed treatment without delay, meeting the feasibility endpoint; median treatment time was only 11.5 minutes per fraction, despite continuous intrafraction monitoring.
- Laryngeal motion was clinically relevant. 20/94 fractions (21%) required at least one shift and re-imaging; 11/94 required a 2D couch shift and 14/94 required a >2-mm 3D shift. In 19/94 fractions (20%), beam interruption occupied more than 20% of treatment time.
- During beam-interruption events, median target displacement was 3.2 mm, with the GTV centroid exceeding 3 mm in 53%, 5 mm in 21%, and 10 mm in 3% of cine-MRI frames. Skin motion correlated only modestly with internal laryngeal motion (r=0.30; P<.001), cautioning against reliance on external surrogates alone.
- Treatment-related toxicity was mild: no grade 3–5 toxicity occurred. Nine grade 1–2 events were recorded, including dysphagia, fatigue, cough and hoarseness; patient-reported swallowing, voice, symptom burden and general health remained broadly stable through follow-up.
- The oncologic signal requires caution. 2/10 patients developed biopsy-proven in-field recurrence at 5.5 and 5.7 months, both after the 42.5-Gy five-fraction regimen; one also developed nodal disease. Both were successfully salvaged, but the small heterogeneous cohort, short follow-up and absence of a control arm make efficacy unresolved.
CLINICAL TAKEAWAY
Real-time MRI demonstrated why the larynx is challenging for very small treatment volumes: displacement beyond 3–5 mm was common enough to trigger repeated beam interruptions and repositioning. The approach produced remarkably little toxicity, but two in-field failures among ten patients mean five-fraction reduced-volume MRgRT should remain investigational until the ongoing phase II study clarifies local control.
SOURCE
International Journal of Radiation Oncology, Biology, Physics