KEY POINTS
- The review organizes toxicity-reduction strategies into three dimensions: spatial precision, including response-guided target reduction and adaptive radiotherapy; temporal optimization, including treatment sequencing and systemic-therapy de-escalation; and energy selection, principally intensity-modulated proton therapy.
- The strongest spatial evidence comes from a randomized phase 3 trial of 449 patients after induction chemotherapy. Three-year locoregional recurrence-free survival was 94.3% with reduced volumes versus 90.5% with conventional volumes, meeting the prespecified non-inferiority criterion.
- Response-guided target reduction lowered grade 3 or higher oral mucositis from 56.9% to 34.0% and also reduced severe xerostomia and hearing loss. This approach has subsequently been incorporated into international nasopharyngeal-carcinoma contouring recommendations.
- A randomized phase 3 trial of 468 patients found three-year failure-free survival of 83.7% with induction chemotherapy followed by radiotherapy and adjuvant chemotherapy versus 79.5% with induction followed by concurrent chemoradiotherapy. Grade 3 or higher acute toxicity during radiotherapy fell from 58.6% to 34.8%.
- Immunotherapy-based chemotherapy de-escalation is presented as a promising temporal strategy, but the supporting studies are heterogeneous and largely phase 2 or conducted in recurrent or metastatic disease. They do not yet establish replacement of concurrent cisplatin in routine curative treatment.
- Intensity-modulated proton therapy can reduce mean doses to structures such as the parotids, cochlea, swallowing muscles, brainstem and temporal lobes by approximately 30–50% in planning studies. However, there is no phase 3 comparative trial in newly diagnosed nasopharyngeal carcinoma demonstrating improved survival or patient-reported outcomes.
- The proposed integration of Epstein-Barr-virus DNA, radiomics, multi-omics, patient-reported outcomes and artificial intelligence remains conceptual. The review used no formal systematic-search or risk-of-bias methodology, most evidence came from endemic-region populations, and one proton-therapy comparison in its summary table relied on analogous prostate-cancer data rather than nasopharyngeal-carcinoma evidence.
CLINICAL TAKEAWAY
Response-guided target reduction after induction chemotherapy has the most convincing evidence for reducing toxicity without compromising disease control. Sequential chemoradiotherapy may be reasonable when concurrent treatment is poorly tolerated, while proton therapy and immunotherapy-based de-escalation should remain selectively applied or investigational; this narrative review is a useful framework, not a clinical guideline.