Aucun titre de diapositive

1

Can Direct Aperture Optimized IMRT using only
Conventional Jaws produce satisfactory treatment
plans for Head and Neck Cancer?
Guangwei Mu, Ping Xia Department of Radiation
Oncology, University of California San Francisco,
San Francisco, CA, USA
INTRODUCTION
RESULTS

• Background
• IMRT has become the choice of treatment for
  disease sites that require critical structure
  sparing such as head and neck cancer.
• It has evolved into a standard practice in many
  radiation therapy centers. However, due to the
  cost and knowledge required with the
  implementation of IMRT, many centers worldwide
  still use cobalt-60 machines or early versions of
  LINAC without multi-leaf collimators, and do not
  have the capability of providing IMRT to the
  needed patients.
• DAO enabled JO-IMRT has been proven to be
  effective for simpler cases such as breast and
  prostate. But it was suspected that Head and Neck
  cancer might be too complex for the JO-IMRT.
• It was believed that to reach a clinically
  acceptable JO-IMRT plan for a relatively complex
  case, 20 segments per beam angle was required.
• Objective
• To investigate the feasibility of DAO enabled
  Jaws-only IMRT (JO-IMRT) treatment planning for
  complex Head and Neck cancer patients.
• To determine the proper number of segments needed
  for a clinically acceptable plan for JO-IMRT.

• With a randomly selected patient, 8 or 9 segments
  per beam enabled the successful reaching a
  clinically satisfactory JO-IMRT plan for head and
  neck patients within reasonable time frame.
• Six randomly selected HNC patients were
  re-planned with inverse-planned JO-IMRT and
  MLC-IMRT. The JO-IMRT plans were shown to be
  effective. Of the six HNC patients, all had at
  least one JOIMRT plan, with 9 segments per beam,
  to meets the acceptance criteria. Four acceptable
  MLC-IMRT plans with 9 segments per beam were
  achieved, with another one very close. The amount
  of time to reach an acceptable JO-IMRT varied
  widely but on average required around 15 times of
  iterations as for other simpler anatomic sites.
  The number of segments for quality plan did not
  have to be high. Nine segments per beam was a
  good choice to arrive at quality plan in most
  cases.
• Among the JO-IMRT and MLC-IMRT plans with
  identical number of apertures per beam angle, no
  significant difference was observed in the
  average DVHs, and the plan conformal index.
• The averaged plan quality evaluation metrics

METHODS AND MATERIALS

• Using a commercial Treatment Planning System
  Prowess Panther, which utilizes Direct Aperture
  Optimization (DAO) algorithm, a set of randomly
  selected six head and neck patients previously
  treated at our institution were planned for
  JO-IMRT and MLC based IMRT (MLC-IMRT). The
  identical clinical acceptance criteria, identical
  set of beam orientations as the clinical plans
  were used.
• Number of segments per beam from 5 to 9 was
  experimented to try to achieve acceptable plans.
  For each patient, the best efforts were recorded
  and analyzed to verify the validity of possible
  clinical application.
• The plan acceptance criteria were established
  according to the RTOG-0225 protocol.

Figure 2. DVHs of the targets and some key
sensitive structures for the JO-IMRT plans with 9
to 5 segments per beam angle. For comparison
reason, the MLC-IMRT plan with 9 segments per
beam was also shown.

Figure 3. Part of an actual JO-IMRT delivery
sequence, the intensity map and the monitor units
(MU) associated with each aperture.
Figure 4. Part of an actual MLC-IMRT delivery
sequence, the intensity map and the monitor units
(MU) associated with each aperture.
CONCLUSIONS
Table 1. Example of the key endpoint doses and
plan quality evaluation metrics for three
nasopharyngeal cancer patient planned with DAO
JO-IMRT and MLC-IMRT, with 9 segments per beam
angle.

• JO-IMRT is feasible for head and neck tumors.
  The plan quality of JO-IMRT plans was comparable
  to that of traditional IMRT or DAO optimized
  MLC-IMRT plans with regards to plan conformity
  and uniformity as well as sensitive structure
  sparing.
• JO-IMRT plans require nearly twice numbers of MUs
  than the MLC IMRT plans.
• Nine segments per beam were sufficient to obtain
  acceptable jaw-only IMRT plans.

Figure 1. Dose distribution of a specific patient
at two locations. On left is the IMRT plan
delivered with MLC and on the right with
conventional jaws only.
UCSF
.