Click here to


Are you sure ?

Yes, do it No, cancel

Quality of Head and Neck VMAT Treatment Plans Transferred Between Machines

R Zakariaee*, J DeMarco, J Lewis, Z Zumsteg, B Fraass, Cedars-Sinai Medical Center, Los Angeles, CA


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

To evaluate quality of head and neck (H&N) VMAT treatment plans originally optimized for one machine and converted to a different machine for time-efficient high-quality transfer of time-sensitive H&N treatments due to machine downtime.

Simultaneously integrated boost (SIB) H&N plans originally planned using Eclipse™ on machine M1 (120-leaf MLC) were “re-optimized” (i.e. a single optimization run using the final objective set from the original plan) while changing machine type (Plan “P-M2”, 160-leaf MLC). To decompose various interplaying effects (changing machines, restarting optimization, different optimization strategies), the initial plans were also re-optimized (“P-M1”) on the original machine (M1). Both re-optimized plans were normalized to the same (high-risk) target coverage as the original clinical plan (“P-Clin”). Differences between organ-at-risk (OAR) dose-parameters for a single fraction of P-M2 were compared against P-Clin and P-M1. OAR doses for the treatment course with one fraction of P-Clin replaced with P-M2 (“P-Total”) were also assessed.

Large variations in OAR dose-parameters in P-M2 compared to P-Clin were observed (up to +154% for parotid) with a mean 19%(±32%) across all OARs for the first 10 cases. However, the mean difference between the two re-optimized plans P-M2 and P-M1 was only -1%(±32%). The largest variation of OAR dose-parameters in P-Total relative to P-Clin was 4.5%, 3.2% and 3.0% for parotids, brachial plexus and spinal cord, respectively. Significant differences in results also depended on planner optimization strategies.

This study demonstrates that transfer of complicated H&N plans to another linac type without detailed re-optimization can result in significant degradation of plan quality (e.g. OAR dose). However, this degradation mostly stems from restarting the optimization rather than different MLC designs or other machine differences. Continuing investigation will further study influence of different optimization strategies and evaluate methods to minimize this effect while creating high quality transferred plans efficiently.


Treatment Planning, Linear Accelerator, Optimization


TH- External Beam- Photons: General (most aspects)

Contact Email