Technical Quality Control Guidelines for Helical TomoTherapy
Part of the Technical Quality Control Guidelines for Canadian Radiation Treatment Programs Suite
Canadian Partnership for Quality Radiotherapy
Technical Quality Control Guidelines for Helical TomoTherapy
A guidance document on behalf of:
Canadian Association of Radiation Oncology
Canadian Organization of Medical Physicists
Canadian Association of Medical Radiation Technologists
Canadian Partnership Against Cancer
November 6, 2015
HTT.2015.11.06
Introduction
This document contains detailed performance objectives and safety criteria for Helical TomoTherapy. Please refer to the overarching document Technical Quality Control Guidelines for Canadian Radiation Treatment Centres for a programmatic overview of technical quality control, and a description of how the performance objectives and criteria listed in this document should be interpreted.
Expert Reviewer(s)
Dany Simard
Centre hospitalier de l'Université de Montréal (CHUM)
Montreal, QC
Noel Blais
CISSS Montérégie-Centre
Centre intégré de cancérologie de la Montérégie (CICM)
Greenfield Park, QC
Emilie Soisson
McGill University
Montreal, QC
Brad Warkentin
Cross Cancer Institute
Edmonton, AB
Slav Yartsev
London Health Sciences Centre
London, ON
System Description
Helical TomoTherapy units are hybrid machines that combine a medical linear accelerator (linac) with components of a helical third generation CT scanner to deliver arc-based, intensity-modulated radiotherapy and on-board image guidance. The technology was devised and developed by Rock Mackie and collaborators at the University of Wisconsin in the 1980s and 1990s (Mackie et al., 1993; Mackie, 2006). The linac and CT detector are mounted opposite each other on a ring gantry that continuously rotates while the couch translates at a constant speed through the gantry, enabling the characteristic helical imaging and treatment deliveries. This differs from conventional medical linacs that use a C-arm gantry design. The ring gantry design ensures minimal gantry sag and, provided proper alignment; the centre of rotation for radiation and mechanical components should be within 1 mm (Balog, Mackie, et al., 2003). TomoTherapy systems have been used clinically in Canada since 2004, and are currently marketed by Accuray Incorporated (Sunnyvale, CA, USA).
TomoTherapy machines use a conventional S-band medical linac with a magnetron RF (radiofrequency) source. Only a single energy X-ray beam (nominally 6 MV) with a source-to-axis distance of 85 cm is available for treatment (no electron beam capabilities). This beam is unflattened (i.e. no flattening filter) and has a nominal dose rate of about 850 cGy/min. The collimation systems used to generate TomoTherapy intensity modulated fan-beams differ significantly from those of conventional C-arm linacs. A 64-leaf binary (i.e. either open or shut) MLC composed of two opposing banks (32 leaves each) is used to collimate and modulate the beam in the lateral direction (patient left-right) over a distance of 40 cm (defined at isocenter). The leaves are divergent with the source in the lateral direction, and typically have opening and closing times of 20 - 30 ms. A single set of front and back y-jaws defines the fan width in the superior-inferior direction. Until recently, all TomoTherapy systems were commissioned with three discrete fan widths for treatment: 1.0, 2.5, and 5.0 cm (width at isocenter). The choice of y-jaw setting largely dictates the tradeoff between the dose fall-off in the superior-inferior direction and treatment delivery time. Treatment planning and delivery incorporating dynamic movement of the y-jaws (“TomoEdge” functionality) is now available on some TomoTherapy systems, which mitigates this tradeoff.
A TomoTherapy treatment plan is defined by a treatment sinogram, as well as the period of gantry rotation (between 12 and 60 seconds) and the couch travel per rotation. The sinogram specifies the fraction of time each of the 64 leaves is open for each beam projection in a treatment, with each rotation being divided into 51 projections. Treatment plans assume a constant linac output; recent TomoTherapy units employ a dose control servo (DCS) system to ensure this constancy. Like conventional linacs, TomoTherapy systems use ionization chambers to monitor beam output and to interlock the beam if the output deviates from set limits (typically ±5% over 12s and ±50% over 3 s). Calculated beam-on time, rather than MUs (monitor units), is used to terminate TomoTherapy delivery. For absolute dose calibration, dose rate is adjusted; this affects all treatment deliveries, since a common treatment beam is defined for all three possible jaw settings. While the majority of TomoTherapy deliveries are rotational, TomoTherapy systems can also deliver plans comprised of beams at fixed gantry angles (if equipped with the optional “TomoDirect” functionality).
For megavoltage computed-tomography (MVCT) imaging, the TomoTherapy system uses the same linac, operated at a nominal energy of approximately 3.5 MV and a decreased pulse repetition frequency. The resulting reduced output (~ 2 to 4 % of the treatment beam) keeps the typical imaging dose below 2 cGy. In its current imaging configuration, the y-jaw setting for imaging defines a fan width of 4 mm (defined at the source-axis-distance of 85 cm); user-selectable nominal slice thicknesses of 2, 4, or 6 mm are provided by varying the scan pitch. The CT detector system is a Xenon-filled ionization chamber array located at the beam exit side, 145 cm from the source. The detectors are not focused on the source, and it is primarily the interaction of the MV beam with the tungsten septa that produces the secondary electrons detected by the array. In addition to patient set-up verification, pre-treatment MVCT images may also be used post-treatment for evaluation of the dosimetric effect of changes in patient anatomy and for adaptive radiotherapy planning. Data from the detector array and the monitor chambers can be acquired continuously during treatment, which serves as a useful tool for commissioning and quality assurance. [Balog, Olivera, et al., 2003].
While MVCT imaging is generally relied on for accurate patient positioning, TomoTherapy machines also provide two sets of lasers (red and green) to assist in patient set-up and to define the machine geometry. Since the ring gantry system doesn’t have a field light to visualize treatment fields, a virtual isocenter is located outside the bore at a distance 70 cm from the treatment isocenter in the longitudinal direction. The fixed green lasers project lines to this virtual isocenter and are mainly used during quality assurance measurements. Movable red lasers are positioned during treatment planning (e.g. to coincide with patient BB’s), and are used for the initial patient setup and straightening prior to MVCT imaging. The plan-independent “home” position of the red lasers coincides with the green lasers.
The testing schedule assumes that a patient specific QA is performed on a routine basis. In addition, an assumption is made that patients are setup at each fraction using the MVCT imaging system with only occasional use of MVCT images for dose calculation and plan evaluation. If these practices are not followed, additional QA would be required to ensure treatment delivery accuracy (MLC specific tests, movable lasers QA frequency, patient setup accuracy without MVCT, MVCT HU accuracy for planning, ...). In addition, testing should be performed following significant machine repairs. The specific tests performed would be repair dependent and up to the discretion of the supervising physicist and field service engineer. Further information on test design can be found in the publication by Fenwick et al. (Fenwick, 2004) and the report of AAPM task group 148 (Langen, 2010).
Test Tables
The following section includes test descriptions and testing frequency for recommended daily, monthly, and annual quality control tests.
Table 1: Daily Quality Control Tests
Designator / Test / PerformanceTolerance / Action
Daily
DL1 / Treatment beam output—Rotational or static / 2% / 3%
DL2 / Image/laser coordinate coincidence (lateral,longitudinal,vertical) / 1.5 mm for non-SRS/SBRT / 2 mm for non-SRS/SBRT
1 mm for SRS/SBRT
DL3 / Movable Red laser home position / 1.0 mm for non-SRS/SBRT / 1.5 mm for non-SRS/SBRT
1.0 mm for SRS/SBRT
DL4 / Image registration/alignment (in x,y, and z directions)
(i.e. Position/Reposition Test) / 1 mm
DL5 / MVCT quality / No artefact
DL6 / Backup tape status / No error, changed
DL7 / Safety System Systems TQC / Complete
Table2: Monthly Quality Control Tests
Designator / Test / PerformanceTolerance / Action
Monthly
ML1 / Static treatment beam output / 2% / 3%
ML2 / Rotational treatment beam output / 1.5% / 2%
ML3 / Monitor chamber constancy / 1.5% / 2%
ML4 / Treatment beam output variation with gantry angle / 1.5%
0.8% with DCS / 2%
1% with DCS
ML5 / Beam quality / 0.8% / 1% (PDD10 or TMR2010 )
ML6 / Transverse profile / 1% / 1.5%
ML7 / Longitudinal profiles (each slice width) / 0.8% / 1%
ML8 / Interrupted treatment procedure / 2%/0.8 mm / 3% / 1 mm
ML9 / Red laser movement accuracy / 0.8 mm / 1 mm
ML10 / Treatment couch movement accuracy / 0.8 mm / 1 mm
ML11 / Treatment couch Level / 0.4° / 0.5°
ML12 / MVCT image quality / Consistency
ML13 / MVCT image value density table / 10 HU for water / 20 HU for water
ML14 / MLC specific tests / see description below.
Table3: Annual Quality Control Tests
AL1 / Gantry angle consistency / 0.8° / 1°AL2 / Couch speed uniformity / 1.5% / 2%
AL3 / Couch translation per gantry rotation / 0.8 mm per 5 cm / 1 mm per 5 cm
AL4 / MVCT Dose / Consistency (<4 cGy)
AL5 / Radiation source alignment against y-jaw / 0.3 mm (no Dynamic Jaws)
0.2 mm (Dynamic Jaws) / 0.3 mm (no Dynamic Jaws)
0.2 mm (Dynamic Jaws)
AL6 / x-alignment of source / 0.25 mm / 0.34 mm
AL7 / y-jaw divergence/beam centering / 0.4 mm at iso / 0.5 mm at iso
AL8 / y-jaw/gantry rotation plane alignment / 0.4° / 0.5°
AL9 / Treatment beam field centering / 0.4 mm at iso / 0.5 mm at iso
AL10 / MLC alignment / 1 mm at iso
0.4° / 1.5 mm at iso
0.5°
AL11 / Beam quality / 0.8% / 1% (PDD10 or TMR2010)
AL12 / Transverse profile / 1%
AL13 / Longitudinal profiles / 1% DTA, 1% Dose
AL14 / Imaging/treatment/ laser coordinate coincidence / 1 mm non-SRS
0.8 mm SBRT-SRS / 2 mm non-SRS
1 mm SBRT-SRS
AL15 / TG-51 treatment beam output calibration verification / 1% / 2%
AL16 / Axial green laser / 0.8 mm / 0.2° / 1 mm / 0.3°
AL17 / Treatment couch Sag / 4 mm / 5 mm
AL18 / Treatment Planning Systems TQC / Complete
AL19 / Independent quality control review / Complete
Daily Tests
DL1Measurement of the daily output consistency should be monitored with static or rotational delivery. Among the daily measurements for a system with TomoDirect, at least one static and one rotational delivery should be done during a week. It is a good practice to alternate the jaw size used for the daily output test to cycle through all of the configurations used clinically on a weekly basis. For output measurement on non-TomoDirect units, helical procedures are preferred to account for rotational variation of the output and field width.
DL2A test should demonstrate the coincidence of the image center with the green laser located at the virtual isocenter. The green laser is a surrogate of the treatment isocenter, and is independent of the TomoTherapy software. The coincidence (taking into account the 70 cm displacement) of the green laser and treatment isocenter should be tested annually. Note that the vertical coincidence of the green laser and image centre is affected by the couch sag, and thus the test should be conducted at a consistent couch location.
DL3The red lasers, when at their home position, should overlay the static green lasers.
DL4Based on a daily MVCT phantom image registration, the automatic couch and red laser positioning adjustment should be tested. This is typically done by positioning a phantom with a known setup error, subsequently performing image registration and shifting of the phantom, and evaluating the values of the image shifts and the final red laser positioning.
DL5 The manufacturer recommends a daily MVCT acquisition with no couch in the bore (airscan) to maintain MVCT image quality. Following the recommended daily airscan procedure, ensure with a phantom that MVCT imaging is free of abnormal artefacts.
DL6 The backup tape should be changed daily. The usual setup is to have 5 backup tapes, a different one for each weekday. If an error related to the backup is reported, it should be investigated and rectified, but treatment may proceed, at the discretion of a qualified medical physicist.
DL7Ensure safety systems are in compliance with CPQR guideline SST.2015.03.01
Monthly Tests
ML1The static treatment beam output is checked for the field width against the baseline reference. This test typically uses an ionization chamber calibrated against the local secondary standard.
ML2Using an ionization chamber calibrated against the local secondary standard, the output for rotational deliveries is checked for all commissioned field widths against the treatment planning system. One option for this calibration is to use a set of standard plans provided by the manufacturer (referred to as “TomoPhant” plans). This is the vendor recommended absolute dose calibration procedure for the treatment beam.
ML3The two monitor chambers’ MU displays should agree with each other and with the expected value within 2%. The vendor recommends calibrating the MU displays using a static beam and the 5 cm jaws opening. MU calibration doesn’t affect treatment beam delivery, as time (not MU) is used to terminate TomoTherapy deliveries.
ML4The variation in output as the gantry rotates is evaluated. One potential method is to place an ionization chamber with buildup cap at isocenter, suspended off the end of the couch (to avoid gantry angle-dependent couch attenuation); an alternative is to assess the variation in monitor chamber readings as a function of gantry angle during a rotational delivery. .
ML5Measurement of the PDD10 or TMR2010 (or similar) is checked against the baseline reference.
ML6The transverse profile (lateral) consistency is checked against the baseline reference for the largest commissioned field width. The average absolute difference with respect to the baseline for multiple off axis ratio measurements, within 80% of the field size, should be within the specified value. The goal is to ensure that profiles are delivered in a manner consistent with that modelled in the treatment planning system.
ML7The longitudinal profile consistency is checked against the baseline reference for all commissioned field widths. The profile’s FWHM (full width at half maximum) should not vary more than the specified value.
ML8When a treatment is interrupted, a “completion procedure” has to be generated at the operator station to complete the treatment delivery. The correct generation and delivery of a “completion procedure” should be checked against an uninterrupted procedure for all commissioned field widths (1 field width per month). The delivered dose agreement should be within 3% of each other and the overall length of the dose distribution in the longitudinal direction should be within 1 mm.
ML9Translations of the movable red lasers should be verified against a ruler or fiducials on a phantom.
ML10The physical displacement of the couch should agree with the digital readout within the specified value over a travel of 20 cm. Over the same travel, the longitudinal couch movement should also be perpendicular to the treatment plane within the same specified value.
ML11The stationary couch should be level within the specified value in both the pitch and roll directions.
ML12 Image noise, uniformity and contrast should be checked for consistency with baseline. The spatial resolution of MVCT images should be at least 1.6 mm for a high contrast object. Distances measured on the images for the x-, y- and z- directions should be within 2 mm for non-SRS/SBRT and within 1 mm for SRS/SBRT treatments.
ML13Using density plugs in a phantom, the MVCT image value density table (IVDT) table should be verified monthly. A Houndsfield units (HU) drift of about 20 HU corresponds to a density error of about 2% and a calculation error of about 1% in a typical clinical case (varies with patient dimensions and tumor location). Testing more frequently may be warranted if MVCTs are used for planning (adaptive or otherwise). To help correct for IVDT drift, a built-in TomoTherapy procedure is available at the operator station for HU rescaling based on measurement of water and air HUs. If IVDT drifts cannot be adequately corrected, updating the TPS MVCT IVDT table may be necessary.
ML14MLC performance is a critical component of dose delivery accuracy on TomoTherapy systems. Typically, most centres perform patient-specific delivery quality assurance (DQA) for each patient plan to ensure this accuracy, thus indirectly verifying the capability of the MLC. This is still recommended practice. However, if a centre decides to deviate from patient-specific DQA, additional MLC-specific QA tests should be introduced. These tests are not prescribed in this TQC guideline.
Annual Tests
AL1 Verification of the ability of the system to reproduce the same gantry angles after a certain number of rotations.
AL2 Verification of the couch translation speed uniformity.
AL3 Verification of the synchronization of the couch translation and the gantry rotation.
AL4 MVCT imaging dose should be verified for its consistency over time for all imaging modes. A properly calibrated ion chamber and appropriate sized phantom should be used for these measurements. A dose of <4 cGy measured in the center of the vendor supplied cylindrical Virtual Water phantom is expected. Significant deviations over time should be reported.
AL5 The source y-direction (longitudinal) position should be within the specified value. The longitudinal alignment of the radiation source is checked against the y-jaw center. Action and tolerance levels refer to source shifts in the y-direction at the source position (no magnification).
AL6 The source x-direction (lateral) position should be within the specified value. The lateral alignment of the radiation source is checked against the MLC center. Action and tolerance levels refer to source shifts in the x-direction at the source position (no magnification).
AL7 Verification that the beam diverges symmetrically around the mechanical rotation plane. The alignment of the y-jaw with the beam plane must be checked. This test ensures that the central transverse axis of the treatment beam intersects the mechanical rotational axis perpendicularly.
AL8 Verification that the y-jaw is parallel to the plane of rotation. AL7 and AL8 are necessary to test that the mechanical and radiation rotational axis are coincident within the specified value.
AL9 All field widths should share the same longitudinal center. This test should include the imaging beam, since patient setup with MVCT imaging implicitly assumes the coincidence of the MV imaging and MV treatment beams.
AL10 Verification of the lateral alignment of the MLC (offset and twist) relative to the center of rotation.
AL11 Using a water tank system, the beam quality, defined as a ratio of doses at different depths, is checked against the beam model data. PDD10 or TMR2010 should be within the specified value. The goal is to ensure that treatment beam energy is consistent with that modelled in the treatment planning system.