1 Comparison of target coverage for two motion-encompassing methods for lung stereotactic body radiotherapy by using in-treatment 4D cone-beam CT

Abstract

Introduction Recently, in-treatment 4D cone-beam CT (4D-CBCTin-treat) has been implemented in clinic allowing the acquisition of respiratory correlated CBCT projections concurrently with the beam delivery. With this imaging modality, the actual 4D target locations during treatment can be visualized at the end of each fraction. 4D-CBCTin-treat provides reliable information to evaluate actual target coverage during treatment and to verify planning target volume (PTV) adequacy for lung stereotactic body radiotherapy (SBRT). The purpose of this study was to compare in-treatment target coverage for two motion-encompassing PTV strategies: an internal target volume (ITV) based approach and the mid-ventilation (MidV) based approach. Methods Data from 27 lung cancer patients treated with a frameless SBRT technique using a volumetric modulated arc therapy (VMAT) were selected. For each patient, a 4D-CT was used for treatment planning. For the ITV-based approach, we applied the PTV setting currently used at our institute for lung SBRT: ITV was determined by contouring the target on 10 respiration phases and a generic 4 mm margin was then added to create the PTV. For the MidV-based approach, PTV was defined around the tumor closest to its time-weighted mean position (MidV CT phase) using the Van Herk margin recipe [ [1] Van Herk M B. The probability of correct target dosage: dose-population histograms for deriving treatment margins in radiotherapy. Google Scholar ]. At each fraction, the target was localized using a 4D-CBCT on-line registration procedure and a 4D-CBCTin-treat was acquired during the VMAT delivery. The clinical target volume (CTV) was delineated in 10 respiration phases on all 4D-CBCTin-treat images reconstructed during the treatment course. We defined target coverage per treatment as the percentage of the CTV included within the PTV averaged over respiration phases and fractions. Results A total of 93 4D-CBCTin-treat were delineated. Mean target coverage per treatment was 98.6% (min: 93.4%) and 98.9% (min: 93.5%) for ITV and MidV approaches respectively. Mean PTV volume was 18.1 cc (min: 2.9 cc, max: 50.7 cc) and 18.6 cc (min: 5.4 cc, max: 49.9 cc) for ITV and MidV approaches respectively. Compared to the MidV, the ITV-based strategy using a 4 mm PTV margin resulted to a non-statistically significant difference neither for target coverage (p = 0.401) nor for PTV volume (p = 0.186). Conclusions With a fast VMAT delivery technique and a 4D image guidance protocol, a 4 mm PTV margin can be safely applied when using the ITV-based strategy for frameless lung SBRT. This ITV-based PTV setting provides equivalent irradiated volumes and tumor coverage than the MidV approach. Recently, in-treatment 4D cone-beam CT (4D-CBCTin-treat) has been implemented in clinic allowing the acquisition of respiratory correlated CBCT projections concurrently with the beam delivery. With this imaging modality, the actual 4D target locations during treatment can be visualized at the end of each fraction. 4D-CBCTin-treat provides reliable information to evaluate actual target coverage during treatment and to verify planning target volume (PTV) adequacy for lung stereotactic body radiotherapy (SBRT). The purpose of this study was to compare in-treatment target coverage for two motion-encompassing PTV strategies: an internal target volume (ITV) based approach and the mid-ventilation (MidV) based approach. Data from 27 lung cancer patients treated with a frameless SBRT technique using a volumetric modulated arc therapy (VMAT) were selected. For each patient, a 4D-CT was used for treatment planning. For the ITV-based approach, we applied the PTV setting currently used at our institute for lung SBRT: ITV was determined by contouring the target on 10 respiration phases and a generic 4 mm margin was then added to create the PTV. For the MidV-based approach, PTV was defined around the tumor closest to its time-weighted mean position (MidV CT phase) using the Van Herk margin recipe [ [1] Van Herk M B. The probability of correct target dosage: dose-population histograms for deriving treatment margins in radiotherapy. Google Scholar ]. At each fraction, the target was localized using a 4D-CBCT on-line registration procedure and a 4D-CBCTin-treat was acquired during the VMAT delivery. The clinical target volume (CTV) was delineated in 10 respiration phases on all 4D-CBCTin-treat images reconstructed during the treatment course. We defined target coverage per treatment as the percentage of the CTV included within the PTV averaged over respiration phases and fractions. A total of 93 4D-CBCTin-treat were delineated. Mean target coverage per treatment was 98.6% (min: 93.4%) and 98.9% (min: 93.5%) for ITV and MidV approaches respectively. Mean PTV volume was 18.1 cc (min: 2.9 cc, max: 50.7 cc) and 18.6 cc (min: 5.4 cc, max: 49.9 cc) for ITV and MidV approaches respectively. Compared to the MidV, the ITV-based strategy using a 4 mm PTV margin resulted to a non-statistically significant difference neither for target coverage (p = 0.401) nor for PTV volume (p = 0.186). With a fast VMAT delivery technique and a 4D image guidance protocol, a 4 mm PTV margin can be safely applied when using the ITV-based strategy for frameless lung SBRT. This ITV-based PTV setting provides equivalent irradiated volumes and tumor coverage than the MidV approach.