Magnetic resonance imaging in peripheral lung adenocarcinoma: correlation with histopathologic features.

2009 
Objective Magnetic resonance imaging (MRI) with various technologic advancements has generally been used to elevate the accuracy of diagnosis for several malignant tumors. This study retrospectively evaluated the efficacy of newer MRI techniques for differentiating among the different types of invasiveness in lung adenocarcinoma by comparing the MRI findings with the pathologic findings. Materials and Methods From May 2005 to April 2007, 46 patients with lung adenocarcinoma measuring 3 cm or less across the greatest dimension underwent a surgical operation including preoperative MRI study in this hospital. The MR imaging protocol included 3 pulse sequences: (1) respiratory-triggered T2-weighted short TI inversion recovery; (2) respiratory-triggered high b-value diffusion-weighted imaging (DWI); (3) gadolinium-enhanced dynamic MRI studies. Results Of all the tumors, 13 were bronchioloalveolar carcinoma (BAC), 24 were adenocarcinomas with mixed subtypes (advanced BAC), and 9 were other histologic subtypes (non-BAC). Both the moderate and strong signal intensity on DWI was significantly greater in the advanced BAC (79.2%) and the non-BAC (88.9%) than in the BAC (38.5%). In the dynamic study, a strong enhancement on the time-intensity curve was significantly greater in the advanced BAC (95.2%) and the non-BAC (87.5%) than in the BAC (25%). When the lesions demonstrated a strong enhancement in dynamic study or showed strong signal intensity on DWI, they were judged to be positive. Sensitivity, specificity, and accuracy were 97%, 76.9%, and 91.3%, respectively. Conclusions DWI could therefore be a useful diagnostic modality for differentiating the subtypes of lung adenocarcinomas, and the MRI finding may thus provide useful supplementary information before surgery comprising limited resections.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    24
    References
    36
    Citations
    NaN
    KQI
    []