Background and objective:Advances in high-resolution computed tomography (CT) scanning have increased the detection of small ground-glass opacity (GGO) nodules and also allowed such images to be investigated in detail. However, it is diffcult to differentiate atypical adenomatous hyperplasia (AAH) from adenocarcinoma in situ (AIS) with CT, even at follow-up, because they share many similar CT manifestations. While AAH is thought to be a precursor or even an early-stage lesion of lung adenocarcinoma, and the stepwise progression from AAH to AIS is thought to be reasonable. hTerefore, the hypothesis that the attenuation of GGO is increased gradually from AAH to AIS is proposed. hTe aim of this study was to distinguish AAH from AIS with CT attenuation in patients with pure GGO nodules.
Methods:Between January 2010 and December 2012, the CT ifndings in terms of the greatest diameter and mean CT attenuation (HU) were reviewed and correlated with pathology in 56 patients with AAH (n=21) and non-mucinous AIS (n=38) by two independent observers. All the 59 lesions were pure GGO nodules with size of 2 cm or smaller. To determine variability of measuring CT attenuation, we calculated the 95%conifdence interval (CI) for the limits of agreement by using Bland-Altman analysis. Student t test was used to compare AAH and AIS in terms of diameter and CT attenuation. And receiver operating characteristic (ROC) curve was used to determine the optimal cut-off value of mean CT attenuation for differentiating AAH from AIS and obtain the diagnostic value. Two-tailed P value of less than 0.05 was considered to be signiifcant.
Results:For the manually measured CT attenuation, the 95%CI for the limits of agreement was-40 HU, 50 HU for inter-observer variability. Although there was significant difference in nodule diameter between AAH and AIS (P=0.046), the overlap was considerable. hTe mean CT attenuation was (-718±53) HU (95%CI:-822,-604) for AAH, which was signiifcantly smaller than (-600±35) HU (95%CI:-669,-531) for AIS (P=0.013). hTe area under curve (AUC) from ROC was 0.903 for differentiating AAH from AIS, and the cut-off value of-632 HU was optimal for differentiation between AAH and AIS, with sensitivity of 0.79, speciifcity of 0.95, and accuracy of 0.85.
Conclusion:hTe mean CT attenuation can help the radiological differentiation between AAH and AIS.