We retrospectively studied patients with pulmonary nodules that met the following criteria: (1) lesion diameter of 鈮? cm, (2) GGO proportion of 鈮?0%, and (3) observation without treatment in the prior 6 months. Between 1999 and 2013, 120 pulmonary lesions in 67 patients fulfilled inclusion criteria. We evaluated changes in lesion size on serial computed tomography. Two endpoints, 鈥渢ime to 2-mm growth鈥?and 鈥渋ncidence of 2-mm growth鈥? were analyzed using Cox proportional hazards and logistic regression models, respectively.
At the median observation period of 4.2 years, 34 lesions exhibited growth by 鈮? mm, whereas 86 remained unchanged. Smoking history and initial lesion diameter were statistically significant variables in both time-to-event and regression analyses. Hazard ratio (HR) for smoking history was 3.67 (P < 0.01). Compared with those 鈮? cm, HRs for 1.1-2 cm and 2.1-3 cm lesions were 2.23 (P = 0.08) and 5.08 (P = 0.04), respectively. Odds ratio (OR) for smoking history was 6.51 (P < 0.01); OR for lesion diameter of 1.1-3 cm (versus 鈮? cm) was 4.06 (P = 0.02).
Smoking history and initial lesion diameter are robustly associated with GGO growth. These results suggest that large GGOs, especially in smokers, warrant close follow-up to accurately monitor lesion growth.