This analysis demonstrates that significant interfractional target variation exists in the treatment of GMZL; this variation could lead to difficulties in target localization and/or reproducibility of treatment. MRIgRT is a promising method to evaluate and allow for adaptation to these variations.
Anupama Chundury, MD, Benjamin W. Fischer-Valuck, MD, Olga Green, PhD, Harold Li, PhD, Sasa Mutic, PhD, Maria Thomas, MD, PhD, Jiayi Huang, MD; Washington University School of Medicine
BACKGROUND: Definitive radiation therapy (RT) is the current therapeutic mainstay for Helicobacter pylori–negative or refractory localized gastric marginal zone lymphoma (GMZL). The purpose of this study was to evaluate interfractional target variation in patients undergoing definitive RT for GMZL utilizing a novel onboard magnetic resonance imaging–guided radiation therapy (MRIgRT) system.
METHODS: Five patients with GMZL were treated with definitive RT using MRIgRT. The prescription dose was 30 Gy delivered in 20 fractions of 1.5-Gy daily. Preceding initiation of treatment, patients underwent a treatment planning CT simulation and an MR simulation to delineate the clinical target volume (CTV), defined as the entire stomach. Prior to delivery of each RT fraction, a daily onboard pretreatment volumetric MRI (dMRI) was acquired for target localization. CTVs were retrospectively recontoured on the acquired MRIs, and volumes were analyzed using the treatment planning system.
RESULTS: One hundred dMRIs were evaluated. Mean daily CTV (dCTV) was 267.4 cc (standard deviation [SD], 144.3 cc). Minimum dCTV was 134.9 cc, while maximum dCTV was 864.8 cc. When compared with the planning CTV (as delineated on MR simulation), mean decrease in dCTV over the course of a full treatment was 256.1 cc (SD, 188.3 cc). The mean decrease in dCTV volume percentage was 46.5% (SD, 15%). Ninety-one of the 100 dCTVs had > 20% volume change when compared with the planning CTV; for 3 patients (60%), every dCTV percent change was greater than 20%. Finally, two patients (40%) required resimulation and replanning due to clinically relevant tumor regression or noncompliance to pretreatment eating instructions.
CONCLUSIONS: This analysis demonstrates that significant interfractional target variation exists in the treatment of GMZL; this variation could lead to difficulties in target localization and/or reproducibility of treatment. MRIgRT is a promising method to evaluate and allow for adaptation to these variations.
Proceedings of the 98th Annual Meeting of the American Radium Society -americanradiumsociety.org