Click here to

Session: Therapy Radiotherapy Planning and Delivery [Return to Session]

Clinical Implementation of Knowledge-Based Planning (KBP) Reduces Inter-Planner Variations for Prostate Cancer Treatments

H Xie1, J Roper1,2, J Wolf2, O Kayode2, X Yang1,2, T Liu1,2, A B Jani1,2, J D Bradley1,2, J Zhang1,2*, (1) Department of Radiation Oncology, (2) Winship Cancer Institute, Emory University, Atlanta, GA 30322


WE-C1000-IePD-F7-5 (Wednesday, 7/13/2022) 10:00 AM - 10:30 AM [Eastern Time (GMT-4)]

Exhibit Hall | Forum 7

Purpose: The plan quality of volumetric modulated arc therapy (VMAT) varies by the treatment planner due to individual preferences and prior experiences. RapidPlan (Varian, Palo Alto, CA) is a commercially-available knowledge-based planning (KBP) module recently implemented in our institution to reduce variations in plan quality of prostate cancer treatments by predicting the best achievable dose volume histograms (DVHs). The purpose of this work is to investigate whether clinical implementation of RapidPlan has improved plan quality and reduced plan quality variations among planners.

Methods: Prostate VMAT plans were identified for the two prescription (Rx) schemes commonly used at our institution: 312 plans (52 KBP) at 45 Gy in 25 fractions and 397 plans (153 KBP) at 70 Gy in 28 fractions. Four planners were identified with ≥10 conventionally optimized plans and ≥10 KBP plans for each Rx. Plan quality metrics were assessed including the volumetric ratio of the 50% isodose line to the PTV (R50) and doses to the organs at risk.

Results: RapidPlan implementation reduced the R50: 6.26±4.35 vs 7.87±6.25, p<0.001 for 70 Gy Rx; 3.69±0.45 vs 3.98±0.43, p<0.001 for 45 Gy Rx) and improved OAR sparing (V40Gy,Bladder (%): 27.32±13.02 vs 32.39±16.52, p=0.002 for 70 Gy Rx; 19.67±12.39 vs 26.59±18.06, p=0.010 for 45 Gy Rx; V45Gy,Rectum (%): 17.32±9.25 vs 25.74±10.40, p<0.001 for 70 Gy Rx; 4.87±3.87 vs 8.33±4.93, p<0.001 for 45 Gy Rx). The main tradeoff is a 15-18% increase in monitor units. With RapidPlan, two out of four planners significantly improved their plans in terms of rectum Dmean (planner1: w/o KBP 28.7±8.0 Gy, w/ KBP 22.1±6.0 Gy; planner2: w/o KBP 30.4±5.6 Gy, w/ KBP 25.43±5.1 Gy) and R50. The other two planners maintained similar plan quality after RapidPlan implementation.

Conclusion: The results demonstrate that successful clinical implementation of KBP models can improve plan quality and reduce inter-planner variations.

Funding Support, Disclosures, and Conflict of Interest: Supported by Winship Cancer Institute # IRG-21-137-07 -IRG from the American Cancer Society


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email