Adaptive Planning of Pancreatic Stereotactic Body Radiation Therapy Patients Based on Target Excursion and Deformation

Abstract

The primary objective of this study is to derive and validate planning target volume (PTV) margins for patients undergoing SBRT for pancreatic cancer using electromagnetic transponders to evaluate target motion and deformation and to facilitate adaptive planning. Pancreatic SBRT patients were implanted with 3 electromagnetic transponders for treatment setup, breathing motion tracking, and delivery gating. For each patient, a simulation CT scan and a 10-phase 4DCT scans were acquired. Soft Tissue Beacon Transponders were used to track tumor position and gate treatment delivery based on previously defined preset limits. All initial plans utilized a uniform 3 mm PTV margin expansion from the clinical target volume (CTV), and treatment-gating limit was set to 3 mm. The prescription was 650 cGy per fraction to a total of 3250 cGy. The patient’s target motion and intertransponders distances were measured daily. Treatment motion preset limits and target deformation were used to calculate new margins. Retrospectively, patients were re-planned using calculated margins. The mean CTV, PTV, normal tissue toxicity were evaluated. Friedman’s chi-squire test was used to determine if average inter-fraction target deformation differed significantly for each patient and between patients. Statistical analyses were performed with two-tailed p-values reported and a p-value of <0.05 considered significant. Average intrafraction target motions were 4.95 ±2.26 mm, 5.59 ±1.93, and 5.40±1.98 mm in the left/right, superior/inferior, and anterior/posterior directions respectively. Average target deformation was 2.39 ±1.46 mm for all patients. Calculated margins based on treatment motion limit and target deformation was 5.39 ± 1.46 mm in all directions. With the implementation of these revised plans based on these data, 100% of the CTV, and 95% of PTV were encompassed by the prescription dose. Duodenum D5cc dose was 1509 ± 443 cGy and D10cc dose was 1146 ± 203 cGy, small bowl V20 was significantly less than 30 cc. Large Bowl V28.5 was significantly less than 20 cc. For a single patient, inter-fraction tumor deformation does not change significantly (p=0.7425). Tumor deformation does not change significantly from patient to patient (p=0.8794). Implanted transponder beacons during pancreatic SBRT can provide useful information on tumor deformation and can be used to calculate new PTV margins to adapt the plan and keep normal tissue toxicity within tolerance. Because inter-fraction tumor deformation does not change significantly, revised margins after first fraction provides a reasonable estimate for the remaining fractions.