Background and Aims

Substantial normal tissue (NT) sparing by the FLASH effect was found to date predominantly for large single fraction doses (d≳5Gy) and experimental data suggest that the NT sparing increases for increasing d [1]. However, for many clinical scenarios hypofractionated treatment schedules are known to increase toxicities of late-reacting tissues compared to normofractionated schedules (d_ref=2Gy) at conventional dose rates (CONV). We have developed a formalism to compute the change in radiobiological damage of NT achieved by shifting from a hypofractionated CONV treatment to a normofractionated CONV treatment and to deduce the break-even FLASH modifying factor (FMF= (d_CONV/d_UHDR)|_isoeffect) that would be needed to be achieved by a hypofractionated ultra-high dose rate (UHDR) treatment to compensate this change. This ‘break-even FMF’ is hereafter referred to as W_BE and is confronted with amplitudes of FMF derived from experimental data.

Methods

We used biologically effective dose (BED) based on the LQ and the LQ-linear (LQ-L) model to derive W_BE for treatment schedules that are equieffective for the tumor. We varied tumor (T) and NT parameters including (α/β)_T and (α/β)_NT of the LQ and LQ-L model to assess the robustness of W_BE and compared it to experimental mammalian FMF from [1].

Results

The LQ-L model predicts higher W_BE that are often in favor of hypofractionated UHDR RT with dose-per-fraction d≳15Gy. Instead, when using the LQ model with (α/β)_T=10Gy and (α/β)_NT=3Gy, most currently available experimental evidence suggests no net NT sparing benefit for hypofractionated UHDR RT for d<25Gy compared to normofractionated CONV RT (i.e. FMF>W_BE), see Figure 1. Variations of α/β-ratios by ±30% result for d<20Gy in changes of W_BE below 0.08 for the LQ model.

Conclusions

The formalism allows assessing, which clinical indications that are routinely treated with normofractionated CONV RT might benefit from hypofractionated FLASH RT.

References: [1] Böhlen et al. (2022), accepted IJROBP.

Background and Aims

Background: Patient-specific quality assurance (psQA) is a critical step in ensuring the accuracy and safety of radiotherapy. To the best of our knowledge, there is no existing integrated system for FLASH psQA.

Aim: To design, validate and implement an integrated FLASH psQA framework. The framework includes the treatment planning system (TPS), machine log, and a measurement comparison tool. Automatic analysis of the delivery parameters against the corresponding planned values ensures accuracy of the dose rate, machine performance, and dosimeter response accuracy.

Methods

Methods: First, a transmission FLASH treatment plan is created in a FLASH-specific research version of Eclipse (FLEX TPS). Second, the dose and the dose rate from a uniform phantom is sent to FLASH QA system. Third, the plan is loaded and delivered on a FLASH-enabled Varian ProBeam proton system with a 2D high spatiotemporal resolution detector array placed at the desired depth in water. Third, the FLASH QA system retrieves data from TPS, log files, and measurements for analysis and report. The system reconstructs the dose rate volume based on log data and using an independent calculation model. Various FLASH-relevant parameters can be defined and compared between planned and delivered dose rate. The time stamp and the position from a high spatiotemporal resolution 2D strip ionization chamber array (SICA) was used to compare with the log file to assess the beam performance at ultra-high dose rates (UHDRs).

Results

Results: The FLASH QA system was designed, and data was integrated for end-to-end test. The logfile-based time stamp agreed with the measurement-based time stamp within a maximum 0.4 msec deviation on the spot dwell time and 1mm position accuracy under FLASH dose rates with nozzle currents up to 215nA.

Conclusions

Conclusions: An integrated FLASH patient QA framework was established and demonstrated the critical metrics to ensure the patient received the intended UHDR treatment.

Background and Aims

This study proposes a novel strategy for multi-beam FLASH radiotherapy that combines FLASH delivered by proton beams with photon IMRT beams to allow uniform coverage of the PTV.

Methods

The FLASH effect has been observed for a delivered minimum dose of 6 Gy and a minimum dose rate 40 Gy/s. To achieve these conditions we used matRad, an open-source code implemented using Matlab by the DKFZ. We divided the PTV into different regions, which we called patch regions (Fig.1). For a patch region close to an OAR, a proton FLASH beam is used to deliver radiation dose. For the remaining PTV volume, photon IMRT beams are used to deposit the necessary dose. Furthermore, we also studied the robustness of our new multi-beam FLASH strategy to shifts of the CTV and to set up errors in the IMRT.

Results

We were able to irradiate the OAR’s with a minimum dose of of 6, 8 and 10 Gy and dose rate above 40 Gy/s, allowing for FLASH protection of the OAR. For the PTV we were able to obtain V_{prescribed dose (PD)}>95%; D₉₈> 0.95*(PD); D₂<1.07*(PD); CI (conformity index)>0.8 and HI (homogeneity index)<0.1. As a consequence, we delivered an uniform dose distribution to the whole PTV (Fig.2), using the IMRT dose distribution patched with the FLASH proton dose distribution.For the robustness analysis, we found that the dose delivered to the CTV was not affected by shifts in the CTV and by set up errors.

Conclusions

We concluded from the obtained results that the novel strategy is suited for multi-beam FLASH radiotherapy. This patching approach allows us to achieve FLASH protection of the OARs without affecting the PTV dose coverage. In addition, we showed that our approach was robust to shifts of the CTV and to set up errors in the IMRT.

Background and Aims

To exploit the high conformality of IMPT plans for FLASH, new approaches are required, because the multi-energy-layer deliveries reduce the dose rate substantially. The use of 3D range modulators (3DRMs), which combines the dose-advantages of IMPT with the dose-rate-advantages of transmission proton plans, is a promising modality, referred to as FLASH IMPT.

Our aim was to create a complete workflow for FLASH IMPT treatment planning and dose rate modelling using modified clinical tools and for generating corresponding 3DRM designs, which was validated experimentally. Here, we focus on treatment planning.

Methods

A FLASH-specific research version of Eclipse (FLEX TPS) was modified to optimize dose distributions for 3DRMs. A FLASH IMPT plan was created to cover a complex target shape (field size: 84x84mm²) and a corresponding 3DRM design was saved as a ready-to-print STL file. The dose distribution for each beamlet was combined with the delivery timing model to estimate the 3D PBS dose rate¹ for a target dose of 15Gy. The spot MU list and the 3DRM design were used to simulate a Monte Carlo (MC) dose distribution. In addition, the RM was 3D-printed and the resulting measured 3D dose distribution was compared with the simulation. The full workflow is depicted in Figure 1a).

Results

The target geometry and beam arrangement are shown in Figure 1a) (top). Figure 1b) shows a good agreement between the target dose distributions modelled by Eclipse and by MC. The modelled dose rate distribution is shown in Figure 2a), illustrating an average dose rate of 45Gy/s (50% iso-dose minus target). The differences between the measured and the MC dose for the example slice, shown in Figure 2b), stay within +-3%.

Conclusions

We successfully created a FLASH IMPT treatment plan and modelled the dose rate for a complex target shape, showing good agreement with the corresponding measurement.

¹https://doi.org/10.1002/mp.14456

Background and Aims

In electron radiotherapy, heterogenous dose distribution is a common problem resulting from complex surfaces and tissue inhomogeneities. For the high-dose single-fractions commonly used in FLASH radiotherapy (FLASH-RT), this must be carefully considered. In this study we evaluate Bolus Electron Conformal Therapy (BECT) for FLASH-RT. The specific aim was to evaluate if optimized-thickness bolus and intensity modulation can reduce dose heterogeneities in high-dose single-fraction treatments to be used in a veterinary clinical trial in canine cancer patients.

Methods

In the treatment planning system electronRT (.decimal®, LLC, Sanford, Florida, USA), CT-based treatment planning can be utilized to design and subsequently fabricate individualized 3D bolus in machinable wax. Furthermore, the beams can be intensity modulated using tungsten pins in the electron block cut-out. To evaluate if optimized-thickness bolus and intensity modulation could reduce cold- and hotspots and improve the target dose homogeneity, treatment plans employing these two methods were created in a canine cancer patient case with a simulated superficial nasal tumor. The prescribed dose was 30 Gy to 95% of the CTV, and a circular field with a diameter of 6.5 cm was used.

Results

By adding optimized-thickness bolus to the canine patient treatment plan, the maximum relative dose was decreased from 143% to 116%, and the CTV homogeneity index (HI) was reduced from 0.30 to 0.15 (Figure 1). By adding intensity modulation, the HI could be further reduced to 0.10.

Conclusions

We have evaluated the potential of optimized-thickness bolus and intensity modulation for FLASH-RT. We found that unwanted hot spots in heterogenous tissue could effectively be reduced in a canine patient treatment plan. Next, we will verify that these plans can be accurately delivered. This technique will help us avoid overdosing and possibly related toxicity in future clinical studies with electron FLASH-RT.