FDG-PET-based radiotherapy planning in lung cancer: optimum breathing protocol and patient positioning--an...

Fluoro-2-deoxy-d-glucose (FDG)-positron emission tomography (PET) and PET/computed tomography (CT) are increasingly used for radiotherapy (RT) planning in patients with non-small-cell lung carcinoma. The planning process often is based on separately acquired FDG-PET/CT and planning CT scans. We compared intraindividual differences between PET acquired in diagnostic (D-PET) and RT treatment position (RT-PET) coregistered with planning CTs acquired using different breathing protocols.

Sixteen patients with non-small-cell lung carcinoma underwent two PET acquisitions (D-PET and RT-PET) and three planning CT acquisitions (expiration [EXP], inspiration [INS], and mid-breath hold [MID]) on the same day. All scans were rigidly coregistered, resulting in six fused data sets: D-INS, D-EXP, D-MID, RT-INS, RT-EXP, and RT-MID. Fusion accuracy was assessed by three readers at eight anatomic landmarks, lung apices, aortic arch, heart, spine, sternum, carina, diaphragm, and tumor, by using an alignment score ranging from 1 (no alignment) to 5 (exact alignment).

The RT-PET showed better alignment with any CT than D-PET (p < 0.001). With regard to breathing, RT-MID showed the best mean alignment score (3.7 +/- 1.0), followed by RT-EXP (3.5 +/- 0.9) and RT-INS (3.0 +/- 0.8), with all differences significant (p < 0.001). Comparing alignment scores with regard to anatomic landmarks, the largest deviations were found at the diaphragm, heart, and apices. Overall, there was fair agreement (kappa = 0.48; p < 0.001) among the three readers.

Significantly better fusion of PET and planning CT can be reached with PET acquired in the RT position. The best intraindividual fusion results are obtained with the planning CT performed during mid-breath hold. Our data justify the acquisition of a separate planning PET in RT treatment position if only a diagnostic PET scan is available.