Session: Radiobiology and Special Topics in Medical Physics - The Cellar

June 24, 2022 from 11:00am EST to 12:00pm EST

Scientific Session 2 – Radiobiology and Special Topics in Medical Physics
Friday, June 24, 2022, 11:00-12:00

Scientific Session 2:Radiobiology and Special Topics in Medical Physics – Presentation 1

Relating ionizing radiation damage quantified by a plasmid DNA-based detector to cells

Tracey Ng, Sunil Advani, Mauro Tambasco
SDSU, UCSD

Purpose: To provide a method for relating ionizing radiation damage quantified by a plasmid DNA-based detector intended for radiation beam assessment to that of cellular DNA damage.

Methods: Circular polycarbonate phantoms constructed with an induction sealed cavity in the center were seeded with 24 ng/μL pBR322 plasmid DNA. It was irradiated using a 137 Cs-irradiator source (Model Mark I-68 irradiator, J.L. Shepherd & Associates) to doses ranging from 2 to 30 Gy. The rate of single and double strand breaks per Gy, SSB/Gy and DSB/Gy, respectively were quantified. Agarose gel electrophoresis was used to quantify the relative yields of supercoiled (no damage), open circular (SSB), and linear (DSB) plasmid DNA conformations. The HCT-116 cell line was irradiated with the same irradiator to doses ranging from 1 to 6 Gy, and DSB/Gy quantified at the peak DSB/Gy (1.5 hours) using the γH2AX assay. The rates of DSB/Gy/Gbp of plasmid versus HCT-116 cellular DNA were compared to find the factor relating them.

Results: The factor relating the DSB/Gy/Gbp of plasmid DNA to cellular DNA was (1.2 ±0.6)×10^(-2). This greater damage in the plasmid DNA is expected as the free radical scavenging capacity of the plasmid DNA was ~400 times less than the cells.

Conclusion: We have demonstrated an approach for relating DSB damage of a plasmid DNA-based detector to that of cellular DNA damage.

 

Scientific Session 2:Radiobiology and Special Topics in Medical Physics – Presentation 2

Enhancing nanoparticle accumulation in two dimensional, three dimensional, and xenograft mouse cancer models in the presence of docetaxel

Kyle Bromma, Devika Chithrani, Wayne Beckham
University of Victoria, BC Cancer

Purpose: Docetaxel (DTX) given in conjunction with radiation therapy improves survival in high-risk prostate cancer. Addition of gold nanoparticles (GNPs) is expected to further improve therapeutic benefits remarkably due to GNPs being a strong radiosensitizer. We lay the foundation for the combination of GNP/DTX/RT in three different prostate cancer cell models: monolayer, spheroid, and an in vivo xenograft mouse model.

Methods: Two cell lines, PC-3 and LNCaP, were used in monolayer and in spheroids. Male NRC mice had PC-3 cells implanted for the animal xenograft model. All models were treated with GNPs of 12 nm in size and low doses of DTX. Gold content of cells was measured using inductively coupled plasma mass spectrometry. Cell cycle synchronization, proliferation effects, and cell damage measurements due to DTX was also measured.

Results: Use of DTX was found to synchronize the cells in the G2/M phase effectively after 24 hours. The G2/M phase is the most radiosensitive phase of the cell cycle, allowing for a synergistic response of DTX with GNPs. GNP uptake with docetaxel did not have a significant increase in monolayer in both cell lines, due to the low dose. However, in spheroids and in vivo, there was a 130% increase in GNP uptake in both cell models.

Conclusion: The combination of DTX and GNPs lead to a further increase in GNP uptake and synchronized cells in the most radiosensitive phase, the G2/M phase. Together, this should lead to a synergistic outcome when combined with radiotherapy to effectively treat prostate cancer.

 

Scientific Session 2:Radiobiology and Special Topics in Medical Physics – Presentation 3

Early prediction of skin toxicities during radiotherapy using optical and infrared imaging

Abby Yashayaeva, James Robar, Hannah Dahn, Michelle Svatos
Dalhousie University, University of Wisconsin

Purpose: Approximately 90% of breast cancer radiation therapy patients experience skin toxicities that are difficult to classify and predict ahead of time. A prediction of severe toxicity at the early stages of the treatment may alert clinicians to intervene. The objective of this study was to evaluate the correlation between the skin toxicity occurrences and radiomic features extracted from optical and infrared (thermal) images of skin, and develop a model for predicting patients’ skin toxicity grade.

Methods: Optical and infrared images of breast cancer patients' chest wall were acquired daily during the course of radiation therapy (RT), as well as weekly on the three weeks following treatment. Skin toxicity assessments were conducted weekly until the final visit. The trends of skin colour and temperature radiomic features of the treated area with respect to the untreated area were analyzed, reduced, and used in a machine learning model to predict the patients’ skin toxicity grade.

Results: Analyzing each patient individually and the entire group, 166 colour and temperature features were reduced to seven independent features with significant p-values. The cross-validation accuracy of the machine learning model remained above 80% when reducing the input data until only a biologically effective dose of 30 Gy was included (approximately the first half of treatment fractions).

Conclusion: The quantitative analysis of radiomic features was shown to be promising for predicting skin toxicities. This study will continue toward the development of a practical and effective tool that can be used daily in the RT treatment room.

 

Scientific Session 2:Radiobiology and Special Topics in Medical Physics – Presentation 4

Congruence of skin reactions with measured and predicted dose: Toward a dose area treatment planning metric to predict moist desquamation in breast radiotherapy

Aria Malhotra, Emilie Carpentier, Cheryl Duzenli
BC Cancer

Purpose: High skin dose can cause potentially severe radiotherapy induced skin reactions. The goal of this study is to determine a treatment planning metric to prospectively predict the likelihood and location of moist desquamation (MD) for breast radiotherapy patients.

Methods: The congruence of treatment planning system (TPS) and in-vivo EBT3 film surface dose measurements was assessed for 19 breast radiotherapy patients.  A skin dose evaluation structure (DEV) using an extended margin on the body contour was constructed in the Eclipse TPS. The region of contact between the film and the skin on the inferior breast surface was extracted and a dose area analysis was conducted to compare the measured and predicted dose distributions in this region. Cumulative EQD2 dose-area data using α/β= 11 Gy for MD was compared between patients who developed MD and those that did not, for both TPS and in-vivo measurements.

Results: The maximum dose to >=1 cm2 was over predicted by TPS compared with measurement by approximately 7%. Cumulative EQD2 dose-area data showed a separation between the curves for the +/- MD populations starting at A75% = 38 +/- 2 cm2 and A85% = 17 +/- 1 cm2, for in vivo measurements and TPS calculations respectively.

Conclusions: This study suggests that there is potential to use a treatment planning dose-area metric to predict MD in breast radiotherapy. The analysis was performed in the inferior breast region where skin reactions are most prevalent.

 

Scientific Session 2:Radiobiology and Special Topics in Medical Physics – Presentation 5

Four-dimensional dose calculations and Monte Carlo quality assurance for dynamic tumour tracking on a gimballed linear accelerator

Emilie Carpentier, Maryam Rostamzadeh, Alanah Bergman, Shiqin Su, Tony Popescu, Marie-Laure Camborde, Tania Karan, Ronan McDermott, Emma Dunne, Tony Mestrovic
University of British Columbia, BC Cancer, University of Michigan

Purpose: The Vero4DRT linac is capable of dynamic tumour tracking (DTT) by panning/tilting the radiation beam. Currently, no treatment planning systems (TPSs) model panning/tilting. Here, a 4D dose calculation protocol is presented that models panning/tilting within the TPS to create a 4D dose distribution. Panning/tilting is incorporated into Monte Carlo (MC) for quality assurance (QA) of these 4D dose distributions.

Methods: Step-and-shoot intensity modulated radiation therapy (sIMRT) plans were optimized for 10 liver patients. Within the TPS, the plan was transferred to all 10 phases of their 4DCTs while uniquely modelling panning/tilting, and the dose was re-calculated. These dose distributions were accumulated to create two 4D dose distributions using 1) all ten phases, and 2) only the inhale and exhale phases. These 4D dose distributions were then re-calculated with MC. Differences between MC and the TPS’ dose algorithm were analyzed.

Results: 26 OARs were investigated among the 10 patients. The 2-phase 4D dose distributions showed six of the 26 OARs would exceed their dose constraints. The 10-phase 4D dose distributions showed five of the same OARs would exceed their dose constraints. MC verified that the same OARs exceeded their dose limits. Dose differences between MC and the TPS were prominent in the beam penumbra region.

Conclusion: Calculating 4D dose distributions while properly modeling the actual beam geometry during DTT and verifying such dose distributions with an independent MC calculation allows accurate confirmation of the safety of a DTT treatment prior to delivery.

 

Scientific Session 2:Radiobiology and Special Topics in Medical Physics – Presentation 6

Implementation and validation of beam current transformer for Mobetron ultra-high dose rate electron beam monitoring using multi-detector approach

Gabriel Famulari, Karim Zerouali, James Renaud, Bryan Muir, Jean-Francois Aubry, Francois DeBlois, Jean-Francois Carrier
Centre Hospitalier de l'Université de Montréal (CHUM), National Research Council Canada, Université de Montréal

Purpose: To evaluate the performance of a custom beam current transformer (BCT) as a beam monitoring tool for the Mobetron electron radiation therapy system at ultra-high dose rates (UHDR) using a multi-detector comparison.

Methods: The BCT was placed at the exit window of the primary collimator using a custom 3D-printed holder. An Exradin A17 ion chamber was placed under a 12-cm stack of Solid Water slabs. A plastic scintillation detector (PSD) or EBT-XD film were placed at depth of maximum dose. The beam pulse characteristics were measured by BCT for various beam parameters. Reproducibility was evaluated with BCT and A17 using 10 measurements of N=5 pulses at variable pulse width (PW) and pulse repetition frequency (PRF) between 1.2-4 µs and 15-60 Hz, respectively. Dependence of detector response on PW was evaluated using all detectors.

Results: The BCT counted the correct number of pulses consistent with the delivery settings. The response (peak-to-peak) measured by BCT (A17) was within 1% (2%) at 9 MeV and 4% (4%) at 6 MeV. Maximum deviations in the BCT-A17 ratio did not exceed 2% at 9 MeV and 1% at 6 MeV. While the relative response of BCT, PSD and EBT-XD with PW agreed within 3%, the A17 response deviated from the other detectors by up to 13%.

Conclusion: The BCT shows promise as a pulse-per-pulse beam monitoring tool for UHDR electron beams. The ion chamber measurements in the bremsstrahlung tail region suffer from beam energy fluctuations which cannot be decoupled from fluence variations.

 

Scientific Session 2:Radiobiology and Special Topics in Medical Physics – Presentation 7

Validation of TOPAS/GEANT4 electron fluence calculations in megavoltage photon beams coupled to magnetic fields

Yunuen Cervantes, Francisco Berumen, Luc Beaulieu
Université Laval

Purpose: Monte Carlo (MC) calculations of electron fluence spectra in magnetic fields have been validated using special conditions to fulfil Fano’s theorem. Under these conditions, the electron spectrum is the same with and without magnetic fields. This work aims to benchmark electron fluence MC calculations in megavoltage photon beams coupled to magnetic fields in TOPAS (TOol for PArticle Simulation). 

Methods: MC simulations were performed in TOPAS and in EGSnrc in a geometry fulfilling the special conditions. Electron fluence was scored in a water disk (r = 5 mm, h = 0.25 mm) placed at 5 cm depth in a water phantom (20x20x15 cm3) and irradiated with a 1.25 MeV monoenergetic photon beam with a field size of 10x10 cm2, at 0 T and 1.5 T. Three TOPAS physics modules were explored g4em-standard_opt3, g4em-standard_opt4 and g4em-penelope. The production cut for all particles was set to 65 nm. The fluence was scored using the PreStep mode with a 1 keV bin-width.

Results: Above 65 keV, there is a good agreement (>95%) between the three TOPAS simulations and the reference (EGSnrc), at 0 T and 1.5 T. Below this energy, there is a disagreement as high as 15% for the constructors g4em-standard_opt4 and g4em-penelope, and 25% for g4em-standard_opt3. However, only the TOPAS physics constructor g4em-standard_opt3 presents the expected behavior in magnetic fields.

Conclusion: TOPAS simulations of electrons fluence in the absence and presence of magnetic fields are validated for energies above 65 keV. The low-energy region present discrepancies that need further investigations.