Indeterminate pulmonary nodules: risk for having or for developing lung cancer?

Abstract

This perspective discusses the report by Pinsky and colleagues, which addresses whether noncalcified pulmonary nodules identified on CT screening carry short- and long-term risk for lung cancer. We are facing challenges related to distinguishing a large majority of benign nodules from malignant ones and among those a majority of aggressive from indolent cancers. Key questions in determining individual probabilities of disease, given their history, findings on CT, and upcoming biomarkers of risk, remain most challenging. Reducing the false positives associated with current low-dose computed tomography practices and identification of individuals who need therapy and at what time during tumor surveillance could reduce costs and morbidities associated with unnecessary interventions.

Figure 1

Relative risks are plotted at the midpoints of the three time periods. There are three curves (solid black, red, and blue lines) for the overall RR for any NCN versus no NCN at the person (red), lung (blue), and lobe (black) levels. There are two additional curves (black dotted lines), for the RR of either soft tissue or ground glass (GG) nodules compared with no NCN, at the lobe level (asterisk, soft tissue; triangle, GG).

Figure 2

IPNs and probability of cancer. Schematic representation of current prediction models compared with how improved prediction models could benefit clinical management. Probabilities of IPNs to represent lung cancer are based on their clinical presentation with recommended follow-up. Low probability is considered 0% to 5%, intermediate 6% to 60%, and high 61% to 100% (based on Wahidi et al; ref. 16) Improved predictive models could dramatically reclassify nodules into different probability groups to address the likelihood of cancer at initial discovery.