Respiratory Gating Facilitating 4D Imaging in MRI, CT, and FDG-PET for GTV Definition in Patients with NSCLC

PURPOSE
Respiratory motion blurs PET-images and may cause localization errors in studies acquired using breath holding such as CT and MRI. To reduce motion related influences on radiation therapy planning of lung cancer, image acquisition may be gated. The aim of our study was to compare motion of primary lung tumors as detected by respiration gated data acquired with PET, CT, MRI.

METHOD AND MATERIALS
So far, this ongoing study included four patients (median age 72.5 years) with NSCLC scheduled for radiation therapy. All patients were investigated under free breathing conditions and respiratory gated FDG-PET, CT and MRI. Image data was separated into eight comparable gates uniformly distributed over the breathing cycle. After blinded GTV-definition, volumes and centers of volumes (COV) were generated from all gates and all modalities. Comparative statistics were done using t-tests, confidence intervals and Lin's concordance analysis.

RESULTS
Median tumor volumes in PET, CT and MRI were 28 ml, 69 ml, and 46 ml, respectively. t-tests revealed significant differences for the comparison CT/MRI (p=0.04) and PET/CT (p=0.05), but not for MRI/PET (p=0.09). The medians of the maximum distances the COV traveled during the entire breathing cycle in PET, CT and MRI were 0.42 cm, 0.95 cm, 0.93 cm, respectively. t-tests revealed nearly significant differences between PET and CT (p=0.06) and MRI and PET (p=0.06), while the result for CT and MRI was in good agreement (p=0.74). Lin’s concordance analysis (OCCC=0.172) revealed best congruency between CT and MRI (rho=0.55), only fair congruency between MRI and PET (rho=0.2) and poor congruency between PET and CT (rho=0.02).

CONCLUSION
In our early experience, the differences in GTV-definition of NSCLC in 4D imaging by PET, CT and MRI are mainly related to the observer’s experience. As shown in other tumors before, PET revealed the smallest tumor volume. Due to the low movement of the tumors evaluated in the patients included so far, results are preliminary. Nevertheless, results suggest that MRI may be applicable in radiation therapy planning for GTV-definition. The inclusion of a higher number of patients with pronounced tumor movement during breathing is required.

CLINICAL RELEVANCE/APPLICATION
Gated 4D imaging in NSCLC may increase planning accuracy in radiation oncology. In selected patients, gated radiation therapy might be advantageous and may reduce toxicity in normal tissues.