Early clinical experience with Ethos Therapy system for delivery of urgent palliative IMRT plans
Dom Withers,
United Kingdom
PO-1742
Abstract
Early clinical experience with Ethos Therapy system for delivery of urgent palliative IMRT plans
Authors: Dom Withers1, Amy Ward2, Siobhan Graham3, Ghirmay Kidane1, Ewan Almond1
1Queen's Hospital, Radiotherapy Physics, Romford, United Kingdom; 2Queen's Hospital, Clinical Oncology, Romford, United Kingdom; 3Queen's Hospital, Radiotherapy, Romford, United Kingdom
Show Affiliations
Hide Affiliations
Purpose or Objective
Our
clinic upgraded an existing Varian Halcyon linear accelerator to Ethos Therapy.
Whilst the primary purpose of Ethos is for daily on-line adaptive radiotherapy,
we have also recognised that its efficient automated plan generation is likely
to confer other benefits. There is often a need to treat with palliative RT for
urgent symptom control. Typically there is short notice involved and limited
additional capacity for both clinicians and treatment planners. As such virtual
simulation conformal radiotherapy techniques are often used despite the desire
for the better dose distributions that could be achieved with IMRT. This work
describes our early clinical experience of using the ETHOS system to provide
IMRT plans in a timely manner along with a streamlined administrative and
checking workflow.
Material and Methods
To
date, eight treatments have been delivered to seven patients (see Table 1). Planning
directives were created with goals for PTV coverage. Priority was given to uniform
target coverage, with limited hotspots. Typically the only OAR considered was
spinal canal which was delineated automatically by Ethos. Contours were
delineated directly in ETHOS or on Eclipse and imported. Plan previews were
reviewed and the priority of goals changed on a case-by-case basis. IMRT plans with
seven equidistant fields were generated automatically by the ETHOS Intelligent
Optimisation Engine (IOE). This was reviewed and clinically approved by the
oncologist.
Results
When
there were no interruptions during and between tasks the overall time for
treatment planning was typically 40-60 minutes from import of planning images
through contouring, plan creation and plan sign off (similar to the lead time
for virtual simulation). Within this time, automated plan generation took no
more than 5 minutes. Treatment preparation tasks typically took 15-20 minutes. This
compares favorably to previous cases planned via a standard IMRT workflow which
took up to 3 hours for treatment planning and checking tasks and 1 hour for
treatment preparation tasks. Dose distributions were clinically approved and
noted to achieve a more uniform dose distribution to the intended target,
compared with the traditional parallel-opposed or single applied field
techniques that tend to produce hotspots outside the target and offer poor dose
fall-off. Treatments were all delivered in standard 20-minute appointment slots.
Conclusion
Whilst
there will still be occasions when single fields are an optimal solution, we
found that using the ETHOS workflow in IGRT mode allowed the creation of good-quality
IMRT plans in time frames comparable to those using virtual simulation
techniques. This could allow for optimal choice of planning technique for each
case rather than choosing the fastest technique for clinical need. Our aim is
to develop the workflow further for it to become our default choice for urgent
plans with a contoured target volume.