A potential to reduce pulmonary toxicity: The use of perfusion SPECT with IMRT for functional lung avoidance in radiotherapy of non-smalt cell lung cancer

Background and purpose: The study aimed to examine specific avoidance of functional lung (FL) defined by a single photon emission computerized tomography (SPECT) lung perfusion scan, using intensity modulated radiotherapy (IMRT) and three-dimensional conformal radiotherapy (3-DCRT) in patients with non-smail cell lung cancer (NSCLC). Materials and methods: Patients with NSCLC underwent planning computerized tomography (CT) and lung perfusion SPECT scan in the treatment position using fiducial markers to allow co-registration in the treatment planning system. Radiotherapy (RT) volumes were delineated on the CT scan. FIL was defined using co-registered SPECT images. Two inverse coplanar RT plans were generated for each patient: 4-field 3-DCRT and 5-field step-and-shoot IMRT. 3-DCRT plans were created using automated AutoPlan optimisation software, and IMRT plans were generated employing Pinnacle 3 treatment planning system (Philips Radiation Oncology Systems). All plans were prescribed to 64 Gy in 32 fractions using data for the 6 MV beam from an Elekta linear accelerator. The objectives for both plans were to minimize the volume of FIL irradiated to 20 Gy (fV(20)) and dose variation within the planning target volume (PTV). A spinal cord dose was constrained to 46 Gy. Volume of PTV receiving 90% of the prescribed dose (PTV90), fV20, and functional mean lung dose (fMLD) were recorded. The PTV90/fV(20) ratio was used to account for variations in both measures, where a higher value represented a better plan. Results: Thirty-four RT plans of 17 patients with stage I-IIIB NSCLC suitable for radical RT were analysed. In 6 patients with stage I-II disease there was no improvement in PTV90, fV20, PTV/fV(20) ratio and fMLD using IMRT compared to 3-DCRT. In 11 patients with stage IIIA-B disease, the PTV was equally well covered with IMRT and 3-DCRT plans, with IMRT producing better PTV90/fV20 ratio (mean ratio - 7.2 vs. 5.3, respectively, p = 0.001) and reduced fMLD figures compared to 3-DCRT (mean value - 11.5 vs. 14.3 Gy, p = 0.001). This was due to reduction in fV20 while maintaining PTV coverage. Conclusion: The use of IMRT compared to 3-DCRT improves the avoidance of FIL defined by perfusion SPECT scan in selected patients with locally advanced NSCLC. If the dose to FL is shown to be the primary determinant of lung toxicity, IMRT would allow for effective dose escalation by specific avoidance of FL. 0 2007 Elsevier Ireland Ltd. Alt rights reserved.