Most lung nodules are benign. The ones that are not need to reach a specialist assessment before growth changes the prognosis. Within a single stage letter, five-year survival varies by 16 percentage points — from 92% at T1a (≤1 cm) to 76% at T1c (2–3 cm). Finding it at 8 mm and operating at 8 mm is not the same as finding it at 8 mm and reaching surgery at 25 mm. The referral threshold matters as much as the scan finding.
Last reviewed: May 2026 · Mr Lawrence Okiror FRCS(CTh) FRCSEd(CTh) · GMC 6150382
A brief referral note to Jo Mitchelson with the CT report and recent spirometry is all that is needed. Mr Okiror reviews the imaging personally at the first appointment.
10-second summary
Solid or part-solid ≥6 mm. Any growth. Persistent GGN. Brock score doesn’t match concern. If any of these apply — or the patient cannot tolerate surveillance uncertainty — refer. Jo Mitchelson: 020 7952 2882 · pa@lungsurgeon.co.uk. Seen within 2–3 working days.
The current standard
The Brock model is the most widely used clinical risk calculator for pulmonary nodule malignancy probability in UK practice. It combines patient characteristics — age, sex, smoking history, family history — with nodule size, type (solid or sub-solid), location, and spiculation to produce a single probability score. It is embedded in BTS guidelines and is the basis of most UK MDT triage decisions.
It has two structural limitations. First, it relies on patient demographics and coarse morphological categories rather than the rich image-derived features that the CT itself contains. Second, it systematically underscores ground-glass and sub-solid nodules — precisely the nodule types most likely to represent adenocarcinoma in situ, minimally invasive adenocarcinoma, or lepidic-predominant invasive adenocarcinoma. Many GGNs that Brock rates below the 10% intervention threshold turn out to be cancers.
What is emerging
AI-augmented risk stratification reads the CT image directly — extracting quantitative features from the nodule itself using radiomics: texture, surface morphology, internal heterogeneity, vessel contact patterns, and density gradients that human readers describe qualitatively but cannot codify. The resulting probability score does not depend on patient demographics. It analyses what the nodule looks like, not who the patient is.
This matters most for the nodules Brock handles worst. A 7 mm ground-glass nodule in a 52-year-old non-smoker might generate a Brock score of 3–4% and be placed on surveillance. Radiomics-based analysis of the same nodule — assessing internal architecture, boundary characteristics, and comparison with thousands of histologically confirmed training cases — may identify features associated with a substantially higher malignancy probability, triggering earlier specialist review and tissue diagnosis.
AI risk stratification tools are entering clinical use at major UK thoracic centres in 2026, starting with retrospective validation against institutional nodule CT databases before prospective clinical deployment. At GSTT, a pilot programme is underway. In time, these tools are expected to complement and in some scenarios replace Brock as the primary risk stratification method for indeterminate pulmonary nodules — particularly for sub-solid and ground-glass lesions where the current model underperforms.
What this means for the referring clinician today: if a nodule’s Brock score says “low risk” but the clinical picture, the imaging morphology, or the patient’s own concern says otherwise — trust the concern, not the score. Specialist assessment with access to AI-augmented tools, HRCT review, and MDT discussion is the right next step. The referral is the intervention.
The NHS National Optimal Lung Cancer Pathway standard is 72 hours from chest X-ray to CT scan. The median achieved nationally is 53 days. The private pathway at London Bridge Hospital collapses this to days.
The combined biopsy-to-resection pathway — ION navigational bronchoscopy followed by robotic anatomic resection, potentially under a single anaesthetic — is an active pathway at London Bridge Hospital. For appropriate nodules, this means tissue diagnosis and curative surgery in a single visit rather than two separate procedures weeks apart. Combined ION-to-surgery pathway →
For the full evidence base
The full Stages I–IIIA pathway, trial data, and outcomes
AI in the Lung Cancer PathwayFive AI decision points from screening shadow to operative field
Lung Nodules & Cancer (Patients)Patient-facing guide to what a shadow or nodule means
Shadow on a ScanFor patients told a shadow was found on their chest scan
Private appointments at London Bridge Hospital within 2–3 working days. Clinic letters to the referring GP electronically within 2 working days. 153 personal anatomic resections in 2024–25 at 99.59% operative survival rate.
Jo Mitchelson, PA · 020 7952 2882 · pa@lungsurgeon.co.uk
St Thomas’ #1 UK · Guy’s #2 UK · London Bridge Hospital #10 UK · Newsweek World’s Best Hospitals 2026