Abstract
BACKGROUND: This study sought to determine the rate and patterns of malignancy in patients with extrapulmonary cancers and non-calcified pulmonary nodules, and to develop a statistical model to guide clinicians regarding choice of patients for diagnostic biopsy. METHOD: The medical records of 151 patients evaluated at the Memorial Sloan-Kettering Cancer Center between January 1999 and December 2001 for non-calcified pulmonary nodules were reviewed. Nodules were considered malignant based on the results of a diagnostic biopsy, and were considered benign if their appearance remained stable 2 years after the initial study, if they resolved, or if a biopsy showed a non-malignant condition. RESULTS: Sixty four of 151 patients (42%) were diagnosed with malignant nodules; 32 had newly diagnosed lung cancers, 28 had metastatic spread of their primary cancers, and four had lesions that were either new cancers or of undetermined aetiology. On univariate analysis the likelihood of malignancy increased with nodule size, tobacco exposure, and the finding of a solitary nodule. On multivariable analysis only nodule size and tobacco exposure were predictive of malignancy. The model had good predictive accuracy (area under the curve 0.751) but had insufficient discrimination for use as a clinical tool to determine which patients should undergo diagnostic biopsy. CONCLUSION: Nearly half the non-calcified pulmonary nodules identified in this series were malignant. Lung cancer was more common than metastatic disease. These findings support the need for close interval follow up and a low threshold for diagnostic biopsy in patients with extrapulmonary cancers and non-calcified pulmonary nodules. In smokers, such lesions should raise concern for lung cancer.