Thin-section multiplanar reformats from multidetector-row CT data: utility for assessment of regional tumor extent in non-small cell lung cancer.

Purpose: To determine the clinical utility of thin-section multiplanar reformats (MPRs) from multidetector-row CT (MDCT) data sets for assessing the extent of regional tumors in non-small cell lung cancer (NSCLC) patients.
Materials and Methods: Sixty consecutive NSCLC patients, who were considered candidates for surgical treatment, underwent contrast-enhanced MDCT examinations, surgical resection and pathological examinations. All MDCT examinations were performed with a 4-detector row computed tomography (CT). From each raw CT data set, 5mm section thickness CT images (routine CT), 1.25 mm section thickness CT images (thin-section CT) and 1.25 mm section thickness sagittal (thin-section sagittal MPR) and coronal images (thin-section coronal MPR) were reconstructed. A 4-point visual score was used to assess mediastinal, interlobar and chest wall invasions on each image set. For assessment of utility in routine clinical practice, mean reading times for each image set were compared by means of Fisher's protected least significant difference (PLSD) test. A receiver operator characteristic (ROC) analysis was performed to determine the diagnostic capability of each of the image data sets. Finally, sensitivity, specificity and accuracy of the reconstructed images were compared by McNemar test.
Results: Mean reading times for thin-section sagittal and coronal MPRs were significantly shorter than those for routine CT and thin-section CT (p<0.05). Areas under the curve (Azs) showing interlobar invasion on thin-section sagittal and coronal MPRs were significantly larger than that on routine CT (p=0.03), and the Az on thin-section sagittal MPR was also significantly larger than that on routine CT (p=0.02). Accuracy of chest wall invasion by thin-section sagittal MPR was significantly higher than that by routine CT (p=0.04).
Conclusion: Thin-section multiplanar reformats from multidetector-row CT data sets are useful for assessing the extent of regional tumors in non-small cell lung cancer patients.