Contents lists available at ScienceDirect Physica Medica journal homepage: www.elsevier.com/locate/ejmp Original paper Impact of spot size variations on dose in scanned proton beam therapy A.C. Kraan a, ⁎ , N. Depauw b , B. Clasie b , T. Madden b , H.M. Kooy b a Istituto Nazionale di Fisica Nucleare, Sezione di Pisa, Italy b Massachusetts General Hospital, Department of Physics, Boston, USA ARTICLE INFO Keywords: Spot size Variation IMPT Robustness ABSTRACT Background: In scanned proton beam therapy systematic deviations in spot size at iso-center can occur as a result of changes in the beam-line optics. There is currently no general guideline of the spot size accuracy required clinically. In this work we quantify treatment plan robustness to systematic spot size variations as a function of spot size and spot spacing, and we suggest guidelines for tolerance levels for spot size variations. Methods: Through perturbation of spot size in treatment plans for 7 patients and a phantom, we evaluated the dose impact of systematic spot size variations of 5% up to 50%. We investigated the dependence on nominal spot size by studying scenarios with small, medium and large spot sizes for various inter-spot spacings. To come to tolerance levels, we used the Γ passing rate and dose-volume-histograms. Results: Limits on spot size accuracy were extracted for 8 sites, 3 different spot sizes and 3 different inter-spot spacings. While the allowable spot size variation strongly depends on the spot size, the inter-spot spacing turned out to be only of limited influence. Conclusions: Plan robustness to spot size variations strongly depend on spot size, with small spot plans being much more robust than larger spots plans. Inter-spot spacing did not influence plan robustness. Combining our results with existing literature, we propose limits of ± 25%, ± 20% and ± 10% of the spot width σ , for spots with σ of 2.5, 5.0 and 10 mm in proton therapy spot scanning facilities, respectively. 1. Introduction In Intensity Modulated Proton Therapy (IMPT), dose is delivered to the patient by combining the dose from numerous small proton beams (spots) with a certain lateral size, energy, position, and number of protons. To ensure that the planned and delivered dose correspond, the spot characteristics must be stable. The lateral size of the spots is a parameter for which it is challenging to guarantee perfect stability over time [1–5]. Beam size changes could occur as a result of variations in the proton accelerator (e.g. beam energy, divergence, offset) and in the beam transport (magnet currents, gantry). If beam size modifications persist over many fractions, dose modifications in the patient can occur, with the risk of compromised target coverage and/or overdosage in critical structures [1–4]. Although the importance of spot size stability is known, literature is scarse and there are no general guidelines available on recommended values of this parameter for existing and future proton therapy spot scanning facilities. Chanrion et al.[1] report that dose modifications can occur for beam size changes ⩽25%, based on dose parameters for prostate and skull-base patients. Parodi et al. [3] suggest ± 50% as tolerance limit, based on target coverage for a spherical phantom. Finally Lin et al. [4] report ± 10%, based on the Γ analysis of 28 pa- tients. None of these studies systematically studied the dependence on beam width and inter-spot distance, and moreover none of these studies reported both dose parameters and the Γ analysis together. The latter is useful to understand the full impact of spot size inaccuracies, both in view of machine commissioning as well as patient safety. The goal of this work is twofold. First, we intend to quantify the clinical influence of spot size changes as a function of spot size and inter-spot distance. This will be done by performing a robustness ana- lysis for 7 patients and a phantom. Second, we combine our results with the existing literature to extract tolerance levels for spot size changes. 2. Methods and materials Our patient group (Table 1) consists of 7 patients (pelvis, chest-wall, rectum, chordoma, cardiac, retro-peritoneal, spinal sarcoma) and 1 phantom. For each case we created treatment plans with the Astroid treatment planning system [7,7]. For optimizing the target and organ- at-risk dose, multi-criteria optimization is used, based on the compu- tation of a set of Pareto optimal plans [8]. Plans were made with 3 https://doi.org/10.1016/j.ejmp.2018.12.011 Received 17 July 2018; Received in revised form 9 October 2018; Accepted 15 December 2018 ⁎ Corresponding author. E-mail address: aafke.kraan@pi.infn.it (A.C. Kraan). Physica Medica 57 (2019) 58–64 1120-1797/ © 2018 Published by Elsevier Ltd on behalf of Associazione Italiana di Fisica Medica. T