Evaluation of soft-tissue match methods for utilization in CBCT guided adaptive radiotherapy of lung cancer patients – clinical benefits, limitations and margin determination

Sammanfattning: Purpose: The aim of this study is to a) study the benefits and limitations with soft-tissue match for lung cancer patients, b) evaluate five different soft-tissue match methods, c) compare soft-tissue with bony match, d) find the stable surrogates to match on for obscured malignant lymph nodes, e) calculate CTV to PTV and OAR margins and volume of PTV for the corresponding margins, and d) study the anatomical changes associated with radiotherapy for lung cancer. Material and methods: 23 lung cancer patients (16 NSCLC, 7 SCLC) treated with radiotherapy, with 135 weekly CBCT set-up images were retrospectively matched to the planning CTs by five different match methods using the registration software Offline Review, version 10.0 (Varian Medical Systems). Four match methods utilized the volume of interest (VOI) of the CT defined GTV, including the internal motion (GTV-T/IM), plus a 2, 5, 10 or 20 mm symmetrical margin, respectively. The fifth match method used a square VOI enclosing the GTV-T with a 10 mm symmetrical margin. An intensity range of [-150;150] HU was used for automatic soft-tissue matches. Bony match was retrospectively performed and compared to soft-tissue match. Residual GTV-T/IM set-up deviations in all directions were studied for each match and PTV-T margins were calculated. Additionally, stable surrogates close to GTV-N was used for the residual GTV-N set-up deviation measurements and PTVN margin calculations. Total PTV, based on the margins calculated, was measured, by adding CTV to PTV margins to the delineated CTV for bony and GTV-T/IM + 10 mm soft-tissue matches. Additional 5 patients were included and anatomical changes were observed. Results: All soft-tissue match methods gave similar residual GTV-T/IM set-up deviations, ranging between [-3;3] mm, resulting in [5.2;5.8] mm PTV-T margins compared to bony match with deviations between [-8:10] mm and PTV-T margins [7.4;8.6] mm. Match methods utilizing larger VOIs were more stable compared to match methods using smaller VOIs. Auto match on small targets (< 3 cm3) was problematic, and not possible for match method 5. For 77% of the patients with lymph nodes, the main bronchi area was a suitable stable surrogate. For the remaining lateral GVT-Ns the aortic arch and the main pulmonary artery were suitable as surrogates. Soft-tissue and bony residual GTV-N set-up deviations ranged between [-8;10] and [-20;9] mm respectively resulting in PTV-N margins between [6;9.8] and [7.1;8.1] mm respectively. Mean total PTV spare with soft-tissue match was 54 cm3. Anatomical changes, atelectasis (21%) and pneumonitis (4%) occurred randomly during the course of treatment. For the 8 patients with large anatomical changes, which required adaptive strategy, atelectasis and tumor change were the dominant reasons for adaptation. Conclusion: Using soft-tissue match reduces the required PTV-T margins. For semi-automatic softtissue match on the primary tumor, match within GTV-T/IM with a 10 or 20 mm margin extension used as match VOIs were most appropriate. For small tumors (< 3 cm3), match manually on GTVT/ IM itself is advisable. The main bronchi area is a suitable surrogate primarily for centrally positioned mediastinal GTV-N. Atelectasis was the dominant anatomical change observed. Daily CBCT prevent missing significant anatomical changes and shorten the time between the observation and the adaptive strategy implementation.