Comparison of Tumor Volumes as Determined by Pathologic Examination and FLT-PET/CT Images of Non–Small Cell Lung Cancer: A Pilot Study

Abstract

To determine the cut-off standardized uptake value (SUV) on 3-deoxy-3-18F-fluorothymidine (FLT) positron emission tomography/computed tomography (FLT-PET/CT) images that generates the best volumetric match to pathologic gross tumor volume (GTVpath) for non–small cell lung cancer (NSCLC). Twelve patients with NSCLC who, without any antineoplastic therapy before, underwent FLT-PET/CT scans one or two days followed by lobectomy were enrolled. The surgical specimen was dissected into 5-7 μm sections at approximately 4-mm intervals and stained with hematoxylin and eosin. The tumor-containing area was outlined slice by slice and the GTVpath determined by summing over all the slices, taking into account the interslice thickness and fixation-induced volume reduction. The gross tumor volume from the PET images, GTVPET, was determined as a function of cut-off SUV. Then, the GTV PET(SUV) value was determined at different cut-off values of the absolute SUV, from 0.9 to 1.8 at 0.1 intervals. Similarly, GTVPET(%SUV) was determined at various percentage thresholds of the maximal SUV, from 15% to 60% at 5% intervals. The optimal threshold or optimal absolute SUV was defined as the value at which the GTVPET was the same as the GTVpath. The patients included 5 men and 7 women, with a median age of 62 years (range, 43–71). The pathologic classifications were 8 adenocarcinoma and 4 squamous cell carcinoma. The locations of the tumors were 3 superior lobe of left lung, 2 inferior lobe of left lung, 5 superior lobe of right lung, and 2 inferior lobe of right lung, respectively. The fixation process induced a volumetric reduction to 70.84% ± 5.71% (range, 60.4%-79.7%) of the original tumor volumes. The maximal SUV was 3.50 ± 0.97 (range, 1.96-5.05). The GTVpath was 11.07± 10.30 cm3 (range, 2.39-33.10cm3). The mean optimal threshold and absolute SUV that produced the best agreement between GTVpath and GTVPET were 56% ± 11% and 1.84 ± 0.33, respectively. The optimal absolute SUV had a positive correlation with maximal SUV (P=.037), whereas the optimal threshold had no significant correlation with it (P=.52). On the other hand, the optimal threshold was correlated with volume retraction (P<.05), but the optimal absolute SUV had no significant correlation with volume retraction (P>.05). This study evaluated the use of GTVpath as a criterion for determining the FLT-PET/CT images’ optimal cut-off SUV for NSCLC. Absolute SUV 1.8 and threshold 56% might be used as the optimal cut-off SUV for NSCLC target volume delineation.