Click here to


Are you sure ?

Yes, do it No, cancel

Adapting to the Motion of Multiple Independent Targets Using Multi Leaf Collimator Tracking for Locally Advanced Prostate Cancer

E Hewson1*, Y Ge2 , R O'Brien1 , S Roderick3 , L Bell3 , A Dipuglia3 , J Booth3,4 , P Keall1 , D Nguyen1 , (1) ACRF Image X Institute, University of Sydney, Sydney, NSW, (2) Prince of Wales Hospital, Randwick, NSW, (3) Royal North Shore Hospital, St Leonards, NSW, (4) School of Physics, University of Sydney, Sydney, NSW


(Thursday, 7/18/2019) 10:00 AM - 12:00 PM

Room: 303

Purpose: With the introduction of prostate-specific membrane antigen (PSMA) PET into standard practice the number of patients with nodal boost treatments has increased. However, independent motion of the prostate and nodes can degrade the dose delivered. This study addressed this problem for the first time by developing and evaluating an MLC tracking algorithm that adapts to multiple targets.

Methods: The multi-target tracking algorithm divides the MLC field into two segments assigned to each target, calculates the ideal MLC positions by optimising each target’s aperture separately, and recombines the segments into one deliverable MLC field. Six clinical prostate cancer treatment plans with lymph node metastases were simulated and compared to no tracking with the patient setup to the prostate. Two prostate motion traces were simulated with mean 3D displacements of 1.7mm and 5.4mm, and interfraction shifts 5mm inferior and 10mm posterior. Tracking error was evaluated as the area incorrectly inside (overexposed) and outside (underexposed) the aperture. Dose reconstruction was used to assess the dose that would have been delivered to the targets and organs at risk (OARs).

Results: Across all patients the error was lower using multi-target tracking compared to no tracking, with reductions in over and underexposed area of 87±5%. Multi-target tracking maintained the prostate PTV V95% and node PTV V75% to within -0.6±1.9% and -0.3±2.1%. Without tracking, the prostate and nodes were underdosed by -9.3±9.8% and -30±14%. Using multi-target tracking the bladder V65%, rectum V65% and small bowel V57% differed from the plan by 14±4.1%, -1.4±2.4% and -0.3±0.3%. Without tracking, the bladder, rectum and small bowel differed by 5.5±4.8%, -7.3±3.2% and -9.7±7.4%.

Conclusion: This study demonstrated the dosimetric improvement of multi-target MLC tracking. The algorithm optimises for prostate and node coverage resulting in target doses close to the plan but may increase the dose to the OARs.

Funding Support, Disclosures, and Conflict of Interest: Nguyen is supported by NHMRC and Cancer Institute NSW early career fellowships. O'Brien is supported by a Cancer Institute NSW Career Development fellowship. Keall is supported by an NHMRC Senior Principal Research Fellowship. Keall is an inventor on one licensed patent and one unlicensed patent related to MLC tracking.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email