Locoregionally advanced nasopharyngeal carcinoma (stage III to IVA)

Nasopharyngeal carcinoma in its endemic form is a non-keratinising carcinoma driven by Epstein-Barr virus, concentrated in southern China, South East Asia, North Africa and the Arctic, and it presents late because the nasopharynx is silent: a neck mass, unilateral hearing loss or blood-stained nasal discharge are the usual first signs, and by then most tumours have reached the skull base or the neck nodes. Plasma EBV DNA is a tumour marker that screens populations (a Hong Kong trial detected early tumours in asymptomatic men), stages the disease and, when still detectable after radiotherapy, identifies patients at high risk of relapse. The tumour is exquisitely sensitive to radiotherapy and chemotherapy, so surgery has no role in primary treatment.

The Intergroup 0099 trial (1998) established concurrent cisplatin chemoradiotherapy over radiotherapy alone, and intensity-modulated radiotherapy, which spares the parotids, brainstem and optic pathways around this awkward target, became standard in the 2000s. Induction chemotherapy was then shown to add benefit: the Sun Yat-sen phase 3 trial of gemcitabine and cisplatin before chemoradiotherapy (New England Journal of Medicine 2019) improved three-year recurrence-free survival from 76.5 to 85.3 percent and overall survival, and docetaxel-cisplatin-fluorouracil (TPF) induction had shown a similar effect, so induction gemcitabine-cisplatin followed by cisplatin chemoradiotherapy is the CSCO-ASCO and NCCN standard for stage III to IVA disease. Metronomic capecitabine for a year after chemoradiotherapy improved failure-free survival in a further Chinese phase 3 (Lancet 2021), and adjuvant therapy directed by post-treatment plasma EBV DNA is being tested in NRG-HN001. PD-1 antibodies have entered the curative setting: the CONTINUUM trial (Lancet 2024) added sintilimab to induction chemotherapy and chemoradiotherapy and improved three-year event-free survival from 76 to 86 percent, and trials of toripalimab and camrelizumab in the same setting are reporting. De-escalation is the other direction: trials omit concurrent cisplatin in low-risk stage II to III patients after induction, reduce radiotherapy dose in good responders and use proton therapy to cut late toxicity, because survivors live for decades with xerostomia, hearing loss, cranial neuropathy and the risk of carotid disease.