2322 - Comparison of Automated Treatment Planning Based on External Beam Radiation Therapy with High Dose Rate Brachytherapy for Focal Intraprostatic Boost Radiation Therapy for Localized Prostate Cancer

C. E. Kehayias¹, A. R. Mahal², J. S. Bredfeldt², M. T. King², and C. V. Guthier¹; ¹Department of Radiation Oncology, Dana-Farber Cancer Institute/Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, ²Department of Radiation Oncology, Brigham and Women’s Hospital/Dana-Farber Cancer Institute, Boston, MA

Purpose/Objective(s): The FLAME trial reports a dose escalation technique in the form of external beam radiation therapy with focal boost (FB-EBRT) to dominant intraprostatic lesions (DILs). Automated treatment planning algorithms could produce dosimetrically competitive treatment plans (TPs) efficiently while simultaneously meeting standard clinical organ-at-risk (OAR) constraints. Here, we compare the quality of TPs based on (1) FB-EBRT, generated using an automated treatment planning system, and (2) combination EBRT and high dose rate brachytherapy boost (EBRT+HDRBT) clinically delivered to a cohort of patients receiving radiation treatment for localized prostate cancer. Materials/

Methods: TPs were generated using an auto-planning environment based on a technology company for 50 prostate cancer patients who were treated with dose escalation to the DIL using EBRT+HDRBT. Three treatment planning templates were used to (1) reproduce standard of care planning with a homogenous dose to 77Gy (Template 1), (2) perform simultaneous integrated boost (SIB) to the dominant lesion (GTV) according to the FLAME trial with a V95Gy=95% goal (Template 2), and (3) repeat Template 2 but with a V95Gy=85% goal for lower OAR toxicity (Template 3). All FB-EBRT and EBRT+HDRBT TPs were converted to EQD2 and evaluated based on bladder and rectal sparing (D1cc) as well as GTV dose coverage (D98%).
Results: The auto-planner successfully met all OAR sparing and target coverage goals consistent with clinical constraints outlined in FLAME. For FB-EBRT, D1cc<77Gy and D1cc<80Gy was achieved for all rectum and bladder structures, respectively, with all PTVs covered with V73Gy=100% for all three templates. All GTVs were boosted to V95Gy=100% for both boost templates (Templates 2 and 3). Mean EQD2-converted D1cc for OARs was lower in EBRT+HDRBT with mean rectal D1cc of 68.6Gy±6.1Gy and bladder D1cc of 71.5±5.8Gy compared to rectal D1cc of 75.9Gy±1.0Gy and bladder D1cc of 80.8±0.9Gy for FB-EBRT Template 2 and rectal D1cc of 75.9Gy±1.0Gy and bladder D1cc of 80.8±1.0Gy for Template 3. Furthermore, mean DIL equivalent dose in 2Gy fractions was greater in EBRT+HDRBT with D98% of 120.1±16.7Gy compared to D98% of 109.6±5.4Gy for FB-EBRT Template 2 and D98% of 103.7±5.3Gy for Template 3. Overall, Template 2 TPs exhibited superior GTV boost compared to Template 3 TPs with nearly identical OAR sparing.
Conclusion: The auto-planner consistently produced acceptable EBRT focal boost TPs that met dosimetric constraints according to FLAME for all cases. Although EBRT+HDRBT TPs were still associated with lower OAR D1cc and greater GTV D98%, success of aggressive dose escalation over more moderate escalation suggests there is room to optimize auto-planning parameters to further increase GTV boost and reduce OAR toxicity.