Contrast Beware! Single-Institution Experience Contra-Indicating the Use of Contrast-Enhanced Simulation CT Images for Online Adaptive Treatment Planning

Abstract

<h3>Purpose/Objective(s)</h3> Online adaptive radiotherapy is in its nascent years of clinical implementation. Recently, a dedicated CBCT-based adaptive radiotherapy delivery system was cleared by the FDA for clinical use. While academic institutions have seen rapid adoption of these systems, established clinical guidelines may not be reproducible in the adaptive environment. Here we share our clinical experience on the use of intravenous (IV) contrast during CT simulation for a patient to be treated with CBCT-based adaptive therapy. <h3>Materials/Methods</h3> Recently, our institution simulated a candidate for lung SBRT with an upper left lobe mass. The intent for this treatment was to deliver 52.5 Gy to the GTV using the treatment delivery system. Per institutional protocol, all thoracic SBRT cases are simulated acquiring a free-breathing CT scan which is then followed by the acquisition of a 4DCT scan. As part of this protocol, IV-iodine contrast is applied during 4DCT acquisition for improved ITV definition. CT scans are acquired on a scanner using 120 kVp, 1400 mAs, and an IMR reconstruction filter. CT slice thickness is set to 2 mm and pixel spacing is 1.11 mm. A maximum intensity projection (MIP) and average (AVG) scans are derived from the 4DCT scan for target delineation and planning, respectively. <h3>Results</h3> As physics began planning, it was noticed that a large bolus of IV contrast remained clustered at the junction of the left subclavian and left internal jugular veins, which was located very near (< 4 cm) to the intended target volume. Furthermore, the bolus of contrast was in a position that occluded radiation entry to the target from angles anterior through the patient's left side. By design, the treatment delivery system only provides limited options for density overrides of volumes (water, bone, or titanium). Even with careful determination and assignment of materials to contoured "artifact" volumes in a planning CT scan, these density override assignments are carried over in dose calculation of online adaptive sessions. In addition, since the treatment delivery system uses deformable image registration to map the planning CT with the daily CBCT to improve the accuracy of online adaption dose calculation, errors in registration caused by the lack of presence of iodine contrast during CBCT acquisition could result in increased dose uncertainties on the order of approximately 5-7%. Based on these estimates, this case was re-planned prior to treatment delivery using the free-breathing CT scan. <h3>Conclusion</h3> Due to the treatment delivery system online adaptive methodology, errors that propagate from a bolus of contrast in proximity to the target need to be prospectively estimated using an emulator. At a minimum, planning for Ethos should be done on non-contrast scans to better represent the daily CBCT that is used for deformation.