KEY POINTS
- The single-centre real-world cohort included 752 patients with nonmetastatic nasopharyngeal carcinoma, the majority with stage III–IVA disease. Median follow-up was 68 months.
- Elective nodal regions were treated to approximately 45 Gy EQD2, testing whether microscopic nodal disease could be controlled with a lower prophylactic dose while reducing normal-tissue exposure.
- Five-year outcomes were 97.3% regional control, 96.1% local control, 90.6% distant control and 89.7% overall survival.
- Regional recurrence occurred in only 23/752 patients (3%). Pure recurrence within the prophylactic 45-Gy volume occurred in only 5 patients, or 0.7% of the entire cohort.
- Several failures inside the elective volume arose from small equivocal nodes visible on baseline MRI that subsequently regressed after induction chemotherapy and were therefore not included in definitive-dose nodal volumes. This highlights the importance of careful pretreatment nodal assessment when de-escalating dose.
- In 103 patients who did not receive induction or adjuvant chemotherapy, five-year regional control remained 97.5%, although this subgroup contained predominantly earlier-stage disease with lower nodal burden.
- In 20 paired replans, reducing elective dose from 50 to 45 Gy EQD2 significantly lowered exposure across relevant OARs and produced statistically significant modeled NTCP reductions for parotid dysfunction, dysphagia and hypothyroidism.
- The study had no clinical 50-Gy comparator arm, longitudinal toxicity was not systematically captured and the NTCP advantage was modeled rather than observed. It was also conducted at a single centre in an NPC-endemic region.
CLINICAL TAKEAWAY
Uniform elective nodal irradiation at 45 Gy EQD2 produced excellent long-term regional control with very few failures in the elective volume and clearly improved modeled normal-tissue sparing. These data meaningfully strengthen the case for the 45-Gy lower limit in contemporary NPC guidance, but careful baseline nodal imaging remains essential and prospective comparative data are still needed.