IROC REPORT Paige Taylor, MS, IROC Houston QA Center

Proton therapy practices are growing and evolving. There are over 20 proton therapy centers treating patients in the USA, many of which are contributing to NCI-funded clinical trials. IROC Houston monitors proton centers that wish to participate in such protocols. To this end, we perform phantom audits and on-site dosimetry audits to ensure comparable dose across proton centers. One area of interest is the margins used by proton centers to cover target volumes. These parameters, both lateral and range margins, are collected from the institutions through our electronic proton facility questionnaire.

The typical lateral margins added to the CTV were collected for several anatomical disease sites from 24 proton centers. The median and range for the lateral margins are shown in Table 1. For brain targets, there was a 3-mm difference between the smallest and largest standard lateral margins. For thoracic targets, up to a 12-mm difference in standard lateral margins was reported. Overall, depending on the treatment site, the proton lateral margins varied from 2 mm to 15 mm.

The range margins added to the CTV were reported as either a percentage of the planned beam range, or a percentage of the planned beam range plus a fixed value. For example, a 1.5% margin for a beam with a range of 28 cm would be 4.2 mm water equivalent depth. The proton therapy centers' range margins varied widely, from a variable 1% of the range + an additional 1 mm, to a variable 3.5% of the range + 3 mm. Table 2 shows the calculation of the median depth margin for a proton beam with an initial range of 12 cm (lung) and 28 cm (prostate). For these examples, the depth margins used by the centers varied by 5 mm between institutions for the lung, and 10 mm for the prostate.

We might expect small variation in margins between institutions based on their image guidance or beam delivery system, but most of the centers in this study were using the same orthogonal kV method for patient setup, thus the variation in lateral and depth margins across centers was not expected. In addition, many of the centers share similar proton delivery systems, and even these facilities did not all use the same margins.

This difference in treatment margins is particularly important to the cooperative groups when designing clinical trials and combining patient treatment data from multiple proton therapy centers. If a trial group designs a multi-institutional clinical trial to include proton therapy for the prostate, it will be valuable to know that unless penetration uncertainty margins are specified in the protocol, institutions' depth margins may vary by as much as 10 mm. This means that institutions would be treating different target volumes, and this variation might result in very different normal tissue outcomes.

Thought should be given to which margins best suit each institution's proton therapy equipment capabilities, but also the specifications of the trials. Several new proton centers have implemented variable range margins based on anatomical target location, changes of the pencil beam spot size with energy and air gap, or robustness evaluations of target coverage. Similarly, the increased use of CBCT or CT-on-rails systems may improve patient setup accuracy and reduce the necessary target margins. In the future, clinicians and physicists can work together to develop consensus values for clinical proton margins for treatment consistency across proton therapy centers.

Anatomical Site-Specific Lateral Margins [mm]
	Brain	H&N	Thorax	Abdomen	Pelvis
Median [range]	3 [2-5]	3 [2-7]	5 [3-15]	5 [4-10]	5 [2-10]

Table 1. Lateral margins for proton therapy treatment of various anatomical disease sites.

	Total Depth Margin for Lung [mm]	Total Depth Margin for Prostate [mm]
Median [range]	5 [2-7]	10 [4-14]

Table 2. Total additional range margin for proton therapy treatment of lung and prostate. Back to Article List Next Previous