How is the probability of malignancy assessed in a solitary pulmonary nodule?

Updated: Aug 22, 2019
  • Author: Pujan H Patel, MD; Chief Editor: Guy W Soo Hoo, MD, MPH  more...
  • Print

Because a malignancy may be curable when present as a solitary pulmonary nodule, great care should be taken in evaluating such lesions. A comprehensive assessment generally includes history, physical examination, evaluation of previous chest radiographs, incorporation of other imaging studies (eg, CT scanning, positron-emission tomography [PET] scanning), and invasive diagnostic procedures.

Determining the pretest probability of malignancy is essential in guiding the management of solitary pulmonary nodules, and, thus, estimating the probability of benignity using a validated quantitative model might be an effective strategy. A number of quantitative risk models have been developed, but three commonly used models are the Mayo Clinic Model, [8] the Veterans Administration (VA) Cooperative Model, [9] and the Brock University Risk Model. [10] Apart from the Brock Model in the British Thoracic Society (BTS) guidelines, [11] no one risk model has been specifically ratified by any societies or guidelines. Each has their own nuances, which should be taken into account when deciding on one's preferred risk model.

The Mayo risk model, although well validated, was developed from chest radiography data of incidentally found lung nodules sized 4-30 mm. It performed well in both the derivation and validation cohorts, with the area under the curve (AUC) of the receiver operator curve (ROC) being 0.83 and 0.8, respectively. Comparison with a group of experienced physicians' estimates of malignancy showed that it predicted malignancy no better than physician's predictions. [12] The VA model was derived from a predominantly male population with a significant smoking history. Incidentally found nodules measuring 7-30 mm were included. The overall accuracy was also good, with an area under the ROC of 0.79 in the derivation cohort and 0.73 in the validation cohort. [13] The Brock University risk model was derived from analysis of the Pan-Canadian Early Detection of Lung Cancer Study (PanCan) cohort, which encompassed those who qualified for low-dose lung cancer screening CT (ie, aged 50-75 years with a current or former history of smoking). It was then validated in an independent cohort from the British Columbia Cancer Agency study (BCCA), which included current or prior smokers aged 50-74 years with a 30 pack-year smoking history. It performed very well, with an area under the ROC greater than 0.9 in both the complete risk model and the parsimonious model. Online access to this risk calculator can be found at

Because the evidence is not definitive for many of the management guidelines, clinicians should discuss with patients the risks and benefits of alternative management options and should elicit patient preferences. The probability of malignancy only provides an estimate based on previously published studies and may not be generalized to an individual patient. Therefore, patient preferences and clinician experience are important in planning further management strategies.

Did this answer your question?
Additional feedback? (Optional)
Thank you for your feedback!