B. Van Asselen

First patients treated with a 1.5 T MRI-Linac : Clinical proof of concept of a high-precision, high-field MRI guided radiotherapy treatment

The integration of 1.5 T MRI functionality with a radiotherapy linear accelerator (linac) has been pursued since 1999 by the UMC Utrecht in close collaboration with Elekta and Philips. The idea behind this integrated device is to offer unrivalled, online and real-time, soft-tissue visualization of the tumour and the surroundings for more precise radiation delivery. The proof of concept of this device was given in 2009 by demonstrating simultaneous irradiation and MR imaging on phantoms, since then the device has been further developed and commercialized by Elekta. The aim of this work is to demonstrate the clinical feasibility of online, high-precision, high-field MRI guidance of radiotherapy using the first clinical prototype MRI-Linac. Four patients with lumbar spine bone metastases were treated with a 3 or 5 beam step-and-shoot IMRT plan. The IMRT plan was created while the patient was on the treatment table and based on the online 1.5 T MR images; pre-treatment CT was deformably registered to the online MRI to obtain Hounsfield values. Bone metastases were chosen as the first site as these tumors can be clearly visualized on MRI and the surrounding spine bone can be detected on the integrated portal imager. This way the portal images served as an independent verification of the MRI based guidance to quantify the geometric precision of radiation delivery. Dosimetric accuracy was assessed post-treatment from phantom measurements with an ionization chamber and film. Absolute doses were found to be highly accurate, with deviations ranging from 0.0% to 1.7% in the isocenter. The geometrical, MRI based targeting as confirmed using portal images was better than 0.5 mm, ranging from 0.2 mm to 0.4 mm. In conclusion, high precision, high-field, 1.5 T MRI guided radiotherapy is clinically feasible.

Consequences of air around an ionization chamber : Are existing solid phantoms suitable for reference dosimetry on an MR-linac?

PURPOSE A protocol for reference dosimetry for the MR-linac is under development. The 1.5 T magnetic field changes the mean path length of electrons in an air-filled ionization chamber but has little effect on the electron trajectories in a surrounding phantom. It is therefore necessary to correct the response of an ionization chamber for the influence of the magnetic field. Solid phantoms are used for dosimetry measurements on the MR-linac, but air is present between the chamber wall and phantom insert. This study aimed to determine if this air influences the ion chamber measurements on the MR-linac. The absolute response of the chamber and reproducibility of dosimetry measurements were assessed on an MR-linac in solid and water phantoms. The sensitivity of the chamber response to the distribution of air around the chamber was also investigated. METHODS Measurements were performed on an MR-linac and replicated on a conventional linac for five chambers. The response of three waterproof chambers was measured with air and with water between the chamber and the insert to measure the influence of the air volume on absolute chamber response. The distribution of air around the chamber was varied indirectly by rotating each chamber about the longitudinal chamber axis in a solid phantom and a water phantom (waterproof chambers only) and measuring the angular dependence of the chamber response, and varied directly by displacing the chamber in the phantom insert using a paper shim positioned at different orientations between the chamber casing and the insert. RESULTS The responses of the three waterproof chambers measured on the MR-linac were 0.7%-1.2% higher with water than air in the chamber insert. The responses of the chambers on the conventional linac changed by less than 0.3% when air in the insert was replaced with water. The angular dependence of the chambers ranged from 0.6% to 1.9% in the solid phantom on the MR-linac but was less than 0.5% in water on the MR-linac and less than 0.3% in the solid phantom on the conventional linac. Inserting a shim around the chamber induced changes of the chamber response in a magnetic field of up to 2.2%, but the change in chamber response on the conventional linac was less than 0.3%. CONCLUSIONS The interaction between the magnetic field and secondary electrons in the air around the chamber reduces the charge collected from 0.7% to 1.2%. The large angular dependence of ion chambers measured in the plastic phantom in a magnetic field appears to arise from a change of air distribution as the chamber is moved within the insert, rather than an intrinsic isotropy of the chamber sensitivity to radiation. It is recommended that reference dosimetry measurements on the MR-linac can be performed only in water, rather than in existing plastic phantoms.

Influence of the linac design on intensity-modulated radiotherapy of head-and-neck plans

In this study, we quantify the impact of linac/MLC design parameters on IMRT treatment plans. The investigated parameters were leaf width in the MLC, leaf transmission, related to the thickness of the leaves, and penumbra related primarily to the source size. Seven head-and-neck patients with stage T1-T3N0-N2cM0 oropharyngeal cancer were studied. For each patient nine plans were made with a different set of linac/MLC parameters. The plans were optimized in Pinnacle(3) v7.6c and PLATO RTS v2.6.4, ITP v1.1.8. A hypothetical ideal linac/MLC was introduced to investigate the influence of one parameter at a time without interaction of other parameters. When any of the three parameters was increased from the ideal set-up values (leaf width 2.5 mm, transmission 0%, penumbra 3 mm), the mean dose to the parotid glands increased, given the same tumour coverage. The largest increase was found for increasing leaf transmission. The investigation showed that by changing more than one parameter of the ideal linac/MLC set-up, the increase in the mean dose was smaller than the sum of dose increments for each parameter separately. As a reference to clinical practice, we also optimized the plans of the seven patients with the clinically used Elekta SLi 15, equipped with a standard MLC with a leaf width of 10 mm. As compared to the ideal linac, this resulted in an increase of the average dose to the parotid glands of 5.8 Gy.

Evolution of motion uncertainty in rectal cancer: implications for adaptive radiotherapy.

Reduction of motion uncertainty by applying adaptive radiotherapy strategies depends largely on the temporal behavior of this motion. To fully optimize adaptive strategies, insight into target motion is needed. The purpose of this study was to analyze stability and evolution in time of motion uncertainty of both the gross tumor volume (GTV) and clinical target volume (CTV) for patients with rectal cancer. We scanned 16 patients daily during one week, on a 1.5 T MRI scanner in treatment position, prior to each radiotherapy fraction. Single slice sagittal cine MRIs were made at the beginning, middle, and end of each scan session, for one minute at 2 Hz temporal resolution. GTV and CTV motion were determined by registering a delineated reference frame to time-points later in time. The 95th percentile of observed motion (dist95%) was taken as a measure of motion. The stability of motion in time was evaluated within each cine-MRI separately. The evolution of motion was investigated between the reference frame and the cine-MRIs of a single scan session and between the reference frame and the cine-MRIs of several days later in the course of treatment. This observed motion was then converted into a PTV-margin estimate. Within a one minute cine-MRI scan, motion was found to be stable and small. Independent of the time-point within the scan session, the average dist95% remains below 3.6 mm and 2.3 mm for CTV and GTV, respectively 90% of the time. We found similar motion over time intervals from 18 min to 4 days. When reducing the time interval from 18 min to 1 min, a large reduction in motion uncertainty is observed. A reduction in motion uncertainty, and thus the PTV-margin estimate, of 71% and 75% for CTV and tumor was observed, respectively. Time intervals of 15 and 30 s yield no further reduction in motion uncertainty compared to a 1 min time interval.

Development and clinical introduction of automated radiotherapy treatment planning for prostate cancer

To develop an automated radiotherapy treatment planning and optimization workflow to efficiently create patient specifically optimized clinical grade treatment plans for prostate cancer and to implement it in clinical practice. A two-phased planning and optimization workflow was developed to automatically generate 77Gy 5-field simultaneously integrated boost intensity modulated radiation therapy (SIB-IMRT) plans for prostate cancer treatment. A retrospective planning study (n  =  100) was performed in which automatically and manually generated treatment plans were compared. A clinical pilot (n  =  21) was performed to investigate the usability of our method. Operator time for the planning process was reduced to  <5 min. The retrospective planning study showed that 98 plans met all clinical constraints. Significant improvements were made in the volume receiving 72Gy (V72Gy) for the bladder and rectum and the mean dose of the bladder and the body. A reduced plan variance was observed. During the clinical pilot 20 automatically generated plans met all constraints and 17 plans were selected for treatment. The automated radiotherapy treatment planning and optimization workflow is capable of efficiently generating patient specifically optimized and improved clinical grade plans. It has now been adopted as the current standard workflow in our clinic to generate treatment plans for prostate cancer.

Performance of a cylindrical diode array for use in a 1.5 T MR-linac.

At the UMC Utrecht, a linear accelerator with integrated magnetic resonance imaging (MRI) has been developed, the MR-linac. Patient-specific quality assurance (QA) of treatment plans for MRI-based image guided radiotherapy requires QA equipment compatible with this 1.5 T magnetic field. The purpose of this study was to examine the performance characteristics of the ArcCHECK-MR in a transverse 1.5 T magnetic field. To this end, the short-term reproducibility, dose linearity, dose rate dependence, field size dependence, dose per pulse dependence and inter-diode dose response variation of the ArcCHECK-MR diode array were evaluated on a conventional linac and on the MR-linac. The ArcCHECK-MR diode array performed well for all tests on both linacs, no significant differences in performance characteristics were observed. Differences in the maximum dose deviations between both linacs were less than 1.5%. Therefore, we conclude that the ArcCHECK-MR can be used in a transverse 1.5 T magnetic field.

Minimizing the magnetic field effect in MR-linac specific QA-tests: the use of electron dense materials

To address the quality assurance (QA) of a MR-linac which is an MRI combined with a linear accelerator (linac), the traditional linac QA-tests need to be redesigned, since the presence of the static magnetic field in the MR-linac alters the electron trajectory. The latter causes the asymmetry in the dose kernel which is introduced by the magnetic field and hinders accurate geometrical QA-tests for the MR-linac. We introduced the use of electron dense materials (e.g. copper) to reduce the size of the dose kernel and thereby the magnetic field effect on the dose deposition. Two examples of QA-tests are presented in which the geometrical accuracy of the MR-linac was addressed; beam profile and star-shot measurements. The introduced setup was compared with a reference setup and both were tested on a conventional and the MR-linac. The results showed that the symmetry of the recorded beam profile was restored in presence of the copper material and that the isocenter size of the MR-linac can be determined accurately with the introduced star-shot setup. The use of electron dense materials is not limited to the presented QA-tests but has a broad application for beam-specific QA-tests in presence of a magnetic field.

TU‐C‐T‐6E‐06: First Clinical Experience with Pre‐Treatment and in Vivo IMRT Verification Using EPID Dosimetry

Purpose: The aim of this study was to verify the first IMRT prostate plans made in our clinic with a newly commissioned treatment planning system (TPS, Pinnacle 7.4f) using an amorphous siliconelectronic portal imaging device (a‐Si EPID, Elekta iViewGT), both pre‐treatment and in vivo. Method and Materials: For pre‐treatment verification, the plans of 8 patients were re‐calculated on a polystyrene slab phantom. An in‐house developed back‐projection algorithm was used to estimate the dose distribution at the phantom/patient isocentric mid‐plane (perpendicular to the beam‐axis) with the EPID. Each plan was also validated at the isocentre with ionization chamber measurements. Separate fields were measured with film and EPID, with gantry angle =0°. The in vivo mid‐plane dose was estimated with the EPID for the first 3 fractions and weekly thereafter. γ‐evaluation was used to assess 2D dose distributions with criteria of 3% dose difference (of maximum dose) and 3mm distance‐to‐agreement. The evaluated area included all points within the 20% isodose line of each EPID field. Anatomy changes for in vivo measurements were assessed using cone‐beam CT acquired prior to each verified fraction. Results: For pre‐treatment verification, the dose distributions of EPID vs. plan and EPID vs. film agreed within 3% or 3mm for 99.2% and 100% of points, respectively. The average ratio of the measured and planned isocentre dose was 0.987 ±0.003(SD) for ionization chamber and 0.997 ±0.009(SD) for EPID. For the in vivoyfields, 96.7% of dose points were in agreement. Examples of discrepancies were due to variation in gas pockets during treatment and problems calculating the dose distribution in a small area of overlapping segments (1cm2). Conclusion: These results show that an a‐Si EPID can be used to accurately verify IMRT prostate treatments in the mid‐plane of the phantom or patient, both pre‐treatment and in vivo.

WE-G-BRD-08: Motion Analysis for Rectal Cancer: Implications for Adaptive Radiotherapy On the MR-Linac

Purpose: Purpose of this study is to find the optimal trade-off between adaptation interval and margin reduction and to define the implications of motion for rectal cancer boost radiotherapy on a MR-linac. Methods: Daily MRI scans were acquired of 16 patients, diagnosed with rectal cancer, prior to each radiotherapy fraction in one week (N=76). Each scan session consisted of T2-weighted and three 2D sagittal cine-MRI, at begin (t=0 min), middle (t=9:30 min) and end (t=18:00 min) of scan session, for 1 minute at 2 Hz temporal resolution. Tumor and clinical target volume (CTV) were delineated on each T2-weighted scan and transferred to each cine-MRI. The start frame of the begin scan was used as reference and registered to frames at time-points 15, 30 and 60 seconds, 9:30 and 18:00 minutes and 1, 2, 3 and 4 days later. Per time-point, motion of delineated voxels was evaluated using the deformation vector fields of the registrations and the 95th percentile distance (dist95%) was calculated as measure of motion. Per time-point, the distance that includes 90% of all cases was taken as estimate of required planning target volume (PTV)-margin. Results: Highest motion reduction is observed going from 9:30 minutes to 60 seconds. We observe a reduction in margin estimates from 10.6 to 2.7 mm and 16.1 to 4.6 mm for tumor and CTV, respectively, when adapting every 60 seconds compared to not adapting treatment. A 75% and 71% reduction, respectively. Further reduction in adaptation time-interval yields only marginal motion reduction. For adaptation intervals longer than 18:00 minutes only small motion reductions are observed. Conclusion: The optimal adaptation interval for adaptive rectal cancer (boost) treatments on a MR-linac is 60 seconds. This results in substantial smaller PTV-margin estimates. Adaptation intervals of 18:00 minutes and higher, show little improvement in motion reduction.

Beam characterisation of the 1.5 T MRI-linac

As a prerequisite for clinical treatments it was necessary to characterize the Elekta 1.5 T MRI-linac 7 MV FFF radiation beam. Following acceptance testing, beam characterization data were acquired with Semiflex 3D (PTW 31021), microDiamond (PTW 60019), and Farmer-type (PTW 30013 and IBA FC65-G) detectors in an Elekta 3D scanning water phantom and a PTW 1D water phantom. EBT3 Gafchromic film and ion chamber measurements in a buildup cap were also used. Special consideration was given to scan offsets, detector effective points of measurement and avoiding air gaps. Machine performance has been verified and the system satisfied the relevant beam requirements of IEC60976. Beam data were acquired for field sizes between 1  ×  1 and 57  ×  22 cm2. New techniques were developed to measure percentage depth dose (PDD) curves including the electron return effect at beam exit, which exhibits an electron-type practical range of 1.2 ± 0.1 cm. The Lorentz force acting on the secondary charged particles creates an asymmetry in the crossline profiles with an average shift of  +0.24 cm. For a 10  ×  10 cm2 beam, scatter from the cryostat contributes 1% of the dose at isocentre. This affects the relative output factors, scatter factors and beam profiles, both in-field and out-of-field. The average 20%-80% penumbral width measured for small fields with a microDiamond detector at 10 cm depth is 0.50 cm. MRI-linac penumbral widths are very similar to that of the Elekta Agility linac MLC, as is the near-surface dose PDD(0.2 cm)  =  57%. The entrance surface dose is  ∼36% of Dmax. Cryostat transmission is quantified for inclusion within the treatment planning system. As a result, the 1.5 T MRI-linac 7 MV FFF beam has been characterised for the first time and is suitable for clinical use. This was a key step towards the first clinical treatments with the MRI-linac, which were delivered at University Medical Center Utrecht in May 2017 (Raaymakers et al 2017 Phys. Med. Biol. 62 L41-50).