Radiation treatment planning

Medical physics

Radiation therapy

Radiology

Leiomyosarcoma

Liposarcoma

Medicine

Nuclear medicine

Sarcoma

Pathology

International Journal of Radiation Oncology*Biology*Physics

Retroperitoneal Sarcomas (RPS) are often large and abut multiple abdominal structures, making high dose radiation therapy (RT) challenging. The purpose of this study was to track changes in gross tumor volume (GTV) of RPS from diagnosis through treatment to assess the need for improved adaptive planning protocols in these tumors. We retrospectively identified 34 patients with primary RPS diagnosed between 2012-2017 treated with preop RT with simultaneous integrated boost at our institution with available scans for analysis. GTVs were contoured on all available diagnostic CT (DxCT), CT simulations (CTsim), CBCT scans (taken 2 weeks after start of RT), and pre-op CT to track volumetric changes. Association between GTV, growth rate, changes, and clinical parameters were assessed by ANOVA and clinical outcomes by Kaplan-Meier analysis. Median age was 61.9 years (31-75 y) and median follow up was 24.3 months (3.3-123.6 mo). Median RT dose was 62.9 Gy (50.4-63). Histologies included dedifferentiated liposarcoma (44.1%), leiomyosarcoma (32.3%), well-differentiated liposarcoma (17.6%) and others. Median time was 39 days from Dx scan to CTsim and 77.5 days from CTsim to preop CT. Table 1 shows volume changes at different time points and rate of changes. KV portal images were inadequate for detecting tumor volume or abdominal contour changes. CBCT was limited to 9 cases treated on machines with CBCT capabilities. CT re-simulation (CTresim) performed on 3 patients was prompted by worsening symptoms. Modeled dose distribution of initial RT plan on CTresim of a representative case with 647 mL (7.5%) volume change showed 16 Gy (25%) potential increase in hot spots and 10% under-coverage of target volumes, while dose to OARs such as V45 stomach, increased by 30%. On ANOVA, higher tumor grade was associated with higher growth rate between DxCT and CTsim (p=0.036). On univariate analysis, larger GTV size at diagnosis and at CTsim was associated with poor local control (HR=1.0003, p=0.049, HR=1.0003, p=0.039) and increased GTV growth rate between DxCT and CT sim was associated with worse metastasis-free survival (HR=2.83, p=0.006).Abstract MO_15_2601; Table 1Volume measurement and changes N*MeanMedianMinMaxGTV on DxCT (mL)311563.181202.8418.069650.77GTV on CTsim (mL)342167.991763.6318.9910299.19GTV on Preop CT (mL)272398.331635.9623.6912400.26%Δ DxCT to CTSim3150.8629.92-1.98397.57%Δ CTSim to Preop CT276.52-2.23-27.59110.24%Δ CTsim to CBCT915.088.20-8.2081.5%Δ/day DxCT to CTSim311.140.75-0.067.10%Δ/day CTsim to Preop CT280.10-0.03-0.351.87*Number with available scans for volume analysis. Open table in a new tab *Number with available scans for volume analysis. RPS tumors can change significantly over the course of treatment and radiation plans need to adapt to treat changing GTV to avoid under-coverage or excessive hot spots and OAR dose. High tumor grade is a risk factor for high GTV growth rate and can be considered an indication for planned re-simulation CT during RT to adapt RT plan based on volume changes.

Volumetric Changes in Retroperitoneal Sarcoma and the Implications for Adaptive Radiation Therapy Planning