The correct answer is edge biopsy of the ulcer. A persistent non-healing lower alveolar ulcer in a 52-year-old man who has smoked for 30 years and chews tobacco is oral squamous cell carcinoma until proved otherwise; tobacco nitrosamines, areca nut and alcohol act synergistically, and the gingivobuccal complex is the commonest oral subsite in India. Treatment depends entirely on histology, so tissue diagnosis must precede staging and therapy. Biopsy is taken from the edge, including a rim of normal mucosa, because the centre contains only necrotic slough and yields false negatives.

An X-ray of the mandible (or better, CT/MRI) assesses bone erosion and is part of staging after malignancy is confirmed; it cannot make the diagnosis. Local gum paint with astringent and analgesic is purely symptomatic and dangerously delays diagnosis. Systemic antibiotics would only suit an infective or dental-abscess ulcer and are inappropriate empirical treatment for a suspicious lesion.

Key point: Any oral ulcer failing to heal within 3 weeks in a tobacco user needs an edge biopsy; imaging is for staging and never substitutes for histopathology.