Task Group 158 Measurement and Calculation of Doses Outside the Treatment Volume from External-beam Radiation Therapy Treatment Eric E Klein, Ph.D

Task Group Members Stephen F. Kry, co-chair MD Anderson Cancer Center, Houston, TX Bryan Bednarz, co-chair, University of Wisconsin, Madison, WI Rebecca M. Howell, MD Anderson Cancer Center, Houston, TX Larry Dauer, Memorial Sloan-Kettering Cancer Center, New York NY David Followill, MD Anderson Cancer Center, Houston, TX Eric Klein, Brown University, Providence, RI Harald Paganetti, Massachusetts General Hospital and Harvard Medical School; Boston, MA Brian Wang, University of Louisville, Louisville, KY Cheng-Shie Wuu, Columbia University, New York, NY X. George Xu, Rensselaer Polytechnic Institute, Troy NY

Status of Report Measurement and Calculation of Doses Outside the Treatment Volume from External-beam Radiation Therapy Treatment Med Phys. 2017 Oct;44(10):e391-e429

Scope of Report This report aims to address the following charges as they pertain to non-target radiation: Highlight major concerns Provide a rough estimate of doses associated with different treatment approaches in clinical practice Discuss the uses of dosimeters and phantoms for measuring photon, electron, and neutron exposures Discuss the use of calculation techniques (including Monte Carlo) for dosimetric evaluations Highlight techniques that may be considered for reducing non-target doses Discuss dose reporting Make recommendations for clinical and research practice

Introduction to TG-158 Non-target dose? Out-of-field dose? Dose outside the PTV (no benefit to patient) Out-of-field dose? Dose outside any primary field border TG-158 addresses non-target dose, but primarily in a low-dose context In-field non-target dose Out of field non-target dose PTV: Target dose Out of field non-target dose

Why do we care? Low doses of radiation can be bad Second cancers Cardiac toxicity Fetal damage Implantable electronic devices Cataracts Skin dose (unique considerations not addressed in this report – see TG-176)

Where does the dose come from? Patient scatter Collimator Scatter Head leakage Neutrons Different properties than in-field radiation

So how much dose is there? Simple square fields Use TG-36 1995 What about newer techniques?

Chapter 3 of report IMRT; similar for Tomo, VMAT, FFF, SBRT, electron, proton, brachytherapy, imaging.

Caution These doses provide a range that is likely to be encountered There can be a lot of variability between individual treatments These are just rough guidelines You need to determine the dose for your own case Measurements and/or calculations

Techniques to Minimize Non-target Dose Use details from the report to discuss options and strategies to reduce non-target dose: Reducing the Target Volume (CTV and PTV) Treatment Modalities (protons, FFF, electrons) Treatment Energy (6X vs 18X) Photon Wedges (hard wedges vs dynamic/universal) MLC and Collimator Orientation Coplanar vs Non-coplanar Beams Jaw Tracking Patient Shielding Accelerator Shielding Imaging Dose Management

Peripheral Dose Contributions Characteristics Room scatter <=3% Close to field collimator contributes 20-40% Machine leakage is greatest contributor at distances >30 cm from CAX Characteristics Increases with increasing field size Minimal increase with decreasing depth No significant change with change in SSD

Reducing the target volume Reduce CTV Hodgkin’s Lymphoma – treat only the involved nodes rather than the whole mantle Very high impact on dose reduction Not usually a physics decision Reduce PTV (reduce margin size) Requires improved imaging/immobilization Reduces high and intermediate dose volumes Often needs additional imaging dose

FFF Out of field dose reduced for both IMRT and SBRT procedures using FFF Less head leakage, less collimator scatter Most benefit far from the treatment field, more modulated

Collimator orientation For tertiary MLC, align leaves along patient length (vs across patient) to achieve maximum shielding Dose reduced by 6-50% Easily implemented

Recommendations Needs are different between clinical care (e.g., measuring dose to a pacemaker) and a research study to evaluate relative risks between treatment approaches. Clinical care recommendations Research focused recommendations A more stringent set, particularly for neutrons

Clinical care rec. 1 Dose assessment should be based on needed precision to yield proper clinical management of the case. While an overestimate of the dose may be adequate, the magnitude of the error should nevertheless be known so as to ensure proper clinical management.

Clinical care rec. 2 Treatment planning systems show large errors as close as 3 cm from the field edge and doses calculated by the planning system should not be relied on beyond 3 cm from the field edge (or below the 5% isodose line). The mean organ dose can generally be trusted for cases where the majority of the organ is within the 5% isodose line.

Clinical care rec. 3 Photon measurements can be conducted with a wide range of dosimeters. Physicists must be aware of the nature of the radiation field and the response of the detector so as to at least know the magnitude of the sometimes large corrections that may be necessary.

Clinical Care rec 4. Because neutron measurements are prone to very large errors, TG-158 recommends that for clinical scenarios, data from the literature should be used to estimate patient doses rather than performing patient specific measurements. Surface doses from TG-158 can be used as conservative estimates of patient dose.

Clinical care rec. 5 and 6 For structures of interest within the IGRT imaging field, the imaging dose should be included in the total out-of-field dose. When indicated or possible, the non-target dose to the patient should be minimized following the approaches outlined in report.

Research recommendations Additional recommendations are presented on Neutron dosimetry Monte Carlo methods Dose reporting

Research Recommendations for Neutron Dosimetry Optimal neutron dosimetry consists of spectra measurements in air using, for example, Bonner spheres. Other measurement systems can provide reasonable and acceptable results. Monte Carlo is well suited for in-patient/phantom neutron dose calculations, after the model has been well validated against in-air neutron measurements.

Research Recommendations for Neutron Dosimetry Neutron measurements in-patient/phantom,, require that spatial variations in the neutron energy spectrum be accounted for in the response and calibration of the detector. Low-energy neutron detectors (e.g., standard Bonner spheres, track etch detectors, bubble detectors) should not be used in proton beams. Appropriate detectors include extended-range Bonner sphere systems, SWENDI detectors, and TEPCs.

Neutron Survey Instruments Neutron Survey Measurements Rem-Meter Detector assembly is a polyethylene cylinder, containing a BF3 proportional counter. At high Dose rates: Pulse pile-up and dead time occurs Only for use outside the field (≥ 1m) Bubble detectors for Neutron head leakage Note: too much pile-up if use Rem-meter for head leakage. Note: in many states, this is not necessary measurement because state will accept manufacturer documentation of neutron head leakage. Un-irradiated detector Bubbles form after neutron exposure

Monte Carlo Simulations For photon models, full modeling of the head is necessary beyond 15 cm from the field edge For situations with neutrons, head shielding and major structural components must be included Validation of the model must rely on out of field measurements (and/or neutrons), not in-field characteristics Phantom should match patient if possible, height is the most important parameter

Dose Reporting Relative merit of point dose, integral dose, effective dose, etc. for dose assessment How to manage neutron RBE (Q, wR, ambient dose equivalent, etc.) Effective dose should not be used when doses >2Gy are involved. Fun technical reading.

Use Diode Dosimetry Silicon semiconductor diodes Real-time measurement High sensitivity Good spatial resolution Small size Simple instrumentation No bias voltage Ruggedness Independence from temperature/air pressure changes

Or OSLD 7mm diameter x 0.2 mm thick disk Dose is nearly proportional to OSL signal Linear until ~ 3 Gy Some energy dependence Z ≈ 11.4 -Image courtesy of Landauer Some research as to could this be done during Jursinic (2007) -Figure from Jursinic (2007) -Figure from Jursinic (2007)

Do’s and Don’ts Do design techniques to reduce exposure Use tables from TG-158 Confer with Clinicians Do NOT use TP data !!