On the sensitivity of TG-119 and IROC credentialing to TPS commissioning errors

Our plain-English summary of On the sensitivity of TG-119 and IROC credentialing to TPS commissioning errors, by D. McVicker, F.F. Yin, J.D. Adamson (Journal of Applied Clinical Medical Physics, 2016).

The study

• Authors: D. McVicker, F.F. Yin, J.D. Adamson
• Journal: Journal of Applied Clinical Medical Physics
• Published: 2016;17(1):34-48
• Cited by: 14 articles, as counted by NIH iCite
• Read it: PubMed · Publisher · Free full text

Abstract

We investigate the sensitivity of IMRT commissioning using the TG-119 C-shape phantom and credentialing with the IROC head and neck phantom to treatment planning system commissioning errors. We introduced errors into the various aspects of the commissioning process for a 6X photon energy modeled using the analytical anisotropic algorithm within a commercial treatment planning system. Errors were implemented into the various components of the dose calculation algorithm including primary photons, secondary photons, electron contamination, and MLC parameters. For each error we evaluated the probability that it could be committed unknowingly during the dose algorithm commissioning stage, and the probability of it being identified during the verification stage. The clinical impact of each commissioning error was evaluated using representative IMRT plans including low and intermediate risk prostate, head and neck, mesothelioma, and scalp; the sensitivity of the TG-119 and IROC phantoms was evaluated by comparing dosimetric changes to the dose planes where film measurements occur and change in point doses where dosimeter measurements occur. No commissioning errors were found to have both a low probability of detection and high clinical severity. When errors do occur, the IROC credentialing and TG 119 commissioning criteria are generally effective at detecting them; however, for the IROC phantom, OAR point-dose measurements are the most sensitive despite being currently excluded from IROC analysis. Point-dose measurements with an absolute dose constraint were the most effective at detecting errors, while film analysis using a gamma comparison and the IROC film distance to agreement criteria were less effective at detecting the specific commissioning errors implemented here.

In plain English

This study is about radiation therapy planning, not asbestos. Mesothelioma appears only as one of five example treatment plans. The authors deliberately introduced errors, one at a time, into the setup (commissioning) of a commercial radiotherapy planning system (Eclipse, using the AAA dose calculation) for a 6X photon beam. They recalculated five representative plans (low and intermediate risk prostate, head and neck, mesothelioma, and scalp) and plans on the TG-119 C-shape and IROC head and neck test phantoms. The aim was to see how much doses changed and whether standard checks would catch each error. Entering TMR curves instead of PDD curves was the most severe error. It changed prostate plan dose measures by a mean of over 13%. In the mesothelioma plan, smoothed beam profiles changed the untreated left lung's V20Gy by over 6% of its original value. The IROC film analysis passed every error tested. The authors concluded that no error had both a low chance of detection and high clinical severity. They also concluded that organ-at-risk point-dose measurements, currently excluded from IROC analysis, were the most sensitive check.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Planning-system and phantom dose calculation study in which commissioning errors were deliberately introduced; no patients were exposed or followed
• Who was studied: Five representative clinical IMRT treatment plans (low risk prostate, intermediate risk prostate, right-sided head and neck, right-sided mesothelioma, scalp), plus plans on the TG-119 C-shape phantom and the IROC head and neck phantom. The beam model was for a Varian 2100 series linear accelerator (21EX) in Eclipse version 8.6 at Duke University Medical Center. The mesothelioma plan was included to test dose calculation in the presence of tissue heterogeneities.
• Exposure: No asbestos or fiber exposure was studied. The variables tested were commissioning errors in the 6X photon beam model: incorrect SAD, incorrect flattening filter material (tungsten, lead or iron instead of copper), TMR entered as PDD, smoothed dose profiles (simulating an inappropriately large Farmer chamber), a 70 mm change in an electron contamination parameter, dosimetric leaf gap errors of 0.1 to 1 mm from a nominal 1.8 mm, and MLC transmission factor raised from 0.017 to 0.022 and 0.026.
• Main results:
  • Entering TMR instead of PDD curves was the most severe error. It caused a mean difference in dosimetric indices of over 13% in the prostate plans and differences of over 5% in all plans.
  • An incorrect flattening filter material caused a mean change within 0.3% in all clinical plans. The maximum change was 2.3%, in the parotids of the head and neck plan.
  • Smoothed dose profiles had minimal clinical effect. The exception was the V20Gy of the untreated left lung in the mesothelioma plan, which changed by more than 6% of its original value. This is a relative change, not six percentage points of lung volume.
  • For dosimetric leaf gap errors, the parotid glands were the most sensitive organs: a 2% effect with a 0.2 mm change and a 5% effect with a 0.5 mm change. Prostate plans stayed within 2% for 0.5 mm changes.
  • In the head and neck plan, 0.5% and 0.9% changes in MLC transmission factor changed the mean dose to the right parotid by 4% and 7%. In the scalp plan, the same changes altered the maximum dose to the right optic nerve by 5% and 8%. The prostate and lung plans changed by less than 2%.
  • On the TG-119 C-shape film planes, the TMR-for-PDD error gave a pass rate of 53.37% plus or minus 6.10% under 3%/3 mm absolute gamma analysis, but 98.13% under relative gamma analysis. A 0.5 mm leaf gap error went completely undetected under relative gamma analysis, with a 100% pass rate.
  • In the IROC phantom, the point-dose dosimeters (TLDs) showed a mean difference of 10% for the TMR error. The organ-at-risk TLDs were more sensitive to flattening filter and transmission factor errors. All errors passed the IROC sagittal film 4 mm criterion.
• The authors' conclusion: The authors concluded that the AAA commissioning process in Eclipse was surprisingly robust to user error, and that no error had both low detection probability and high clinical severity. They found the IROC and TG-119 criteria generally effective at detecting errors, organ-at-risk TLDs the most sensitive, and IROC film analysis ineffective for the errors tested. They also concluded that absolute rather than relative dose comparison is necessary.
• Limitations the authors state:
  • The results are specific to the AAA algorithm as implemented in the Eclipse planning system.
  • The error magnitudes chosen were arbitrary for many of the errors.
  • Some of these errors would likely be caught by other quality assurance tests that were not studied.

Topics

Medical terms in this study, explained in our Lexicon: Mesothelioma.

Related studies

• Systematic evaluation of three different commercial software solutions for automatic segmentation for adaptive therapy in head-and-neck, prostate and pleural cancer — M. La Macchia, F. Fellin, et al. (2012)
• Comparison of exome-based HLA class I genotyping tools: identification of platform-specific genotyping errors — K. Kiyotani, T.H. Mai, et al. (2017)
• Research on the correction method for radiotherapy verification plans based on displaced electronic portal imaging device — J. Guo, L. Zhou, et al. (2024)
• Dose estimation for internal organs during boron neutron capture therapy for body-trunk tumors — Y. Sakurai, H. Tanaka, et al. (2014)
• A radiotherapy community data-driven approach to determine which complexity metrics best predict the impact of atypical TPS beam modeling on clinical dose calculation accuracy — F.M.D. Brooks, M.C. Glenn, et al. (2024)
• PET-guided dose escalation tomotherapy in malignant pleural mesothelioma — A. Fodor, C. Fiorino, et al. (2011)

Browse more research in Journal of Applied Clinical Medical Physics, or return to medical research.