Skip to content

Brock (PanCan) Pulmonary Nodule Malignancy Risk

McWilliams 2013 (NEJM), models 1a/1b/2a/2b

Estimates the probability that a pulmonary nodule found on a baseline screening low-dose CT is malignant, using the McWilliams (PanCan/Brock University) logistic models. Implements all four published variants (parsimonious and full, each with or without spiculation).

Also searched as: lung nodule · lung cancer screening · logistic model · BTS · lung cancer · pan can

Model variant

Full models add age, family history, emphysema, nodule type and nodule count and performed best in external validation.

Patient sex
Nodule maximum diameter

Largest transverse (long-axis) diameter of the index nodule.

mm
Located in an upper lobe

Either upper lobe; reference category is middle or lower lobe.

Spiculated margin

Required for models 1b and 2b.

Patient age

Full models only (centered at 62 years).

years
Family history of lung cancer

Full models only.

Emphysema on CT

Visually detected emphysema; full models only.

Nodule type

Full models only; solid is the reference. Perifissural nodules short-circuit to ≈0% (0 of 571 malignant in the source cohorts).

Total nodule count on scan

Full models only (centered at 4 nodules).

Information
Incomplete input
  • Please provide: model version (parsimonious vs full, with/without spiculation), patient sex, nodule maximum diameter (mm), upper-lobe location (yes/no).

How this tool works

The Brock (PanCan) model was derived on 7008 nodules from the Pan-Canadian Early Detection of Lung Cancer Study and externally validated on 5021 nodules from British Columbia Cancer Agency screening cohorts, with discrimination above AUC 0.90 even for nodules 10 mm or smaller. Nodule size enters through a nonlinear fractional-polynomial transformation, (size/10)^(−0.5) − 1.58113883, and continuous predictors are centered (age at 62 years, nodule count at 4). The parsimonious models retain only female sex, transformed size and upper-lobe location (plus spiculation in model 1b); the full models add age, family history of lung cancer, CT emphysema, nodule type and nodule count. Coefficients here are transcribed directly from Table 2 of the NEJM 2013 paper. The British Thoracic Society recommends this model for assessing solid nodules ≥8 mm, using a 10% risk threshold to trigger PET-CT evaluation. Perifissural nodules are handled separately: none of the 571 perifissural nodules in the derivation and validation cohorts was malignant.

References

Implemented from the cited published sources. Educational use only; confirm against current guidelines before clinical use.

Clinical questions that use this

Vetting guides that reach for this tool.