Uncertainties in planned dose due to the limited voxel size of the planning CT when treating lung tumors with proton therapy

On the Way to Accounting for Lung Modulation Effects in Particle Therapy of Lung Cancer Patients-A Review

Particle therapy presents a promising alternative to conventional photon therapy for treating non-small cell lung cancer (NSCLC). However, the heterogeneous structure of lung tissue leads to the degradation of the Bragg peak and thereby to the degradation of the dose distribution. This review offers a comprehensive overview of the models developed to account for these modulation effects. It summarizes studies focused on determining modulation power as a predictor of this so-called lung modulation. In addition, the review covers early investigations on dose uncertainties caused by lung modulation in CT-based lung phantoms and patient anatomies and discusses future challenges in integrating these solutions into clinical treatment planning routines.

Deep learning-based voxel sampling for particle therapy treatment planning

Objective.Scanned particle therapy often requires complex treatment plans, robust optimization, as well as treatment adaptation. Plan optimization is especially complicated for heavy ions due to the variable relative biological effectiveness. We present a novel deep-learning model to select a subset of voxels in the planning process thus reducing the planning problem size for improved computational efficiency.Approach.Using only a subset of the voxels in target and organs at risk (OARs) we produced high-quality treatment plans, but heuristic selection strategies require manual input. We designed a deep-learning model based onP-Net to obtain an optimal voxel sampling without relying on patient-specific user input. A cohort of 70 head and neck patients that received carbon ion therapy was used for model training (50), validation (10) and testing (10). For training, a total of 12 500 carbon ion plans were optimized, using a highly efficient artificial intelligence (AI) infrastructure implemented into a research treatment planning platform. A custom loss function increased sampling density in underdosed regions, while aiming to reduce the total number of voxels.Main results.On the test dataset, the number of voxels in the optimization could be reduced by 84.8% (median) at <1% median loss in plan quality. When the model was trained to reduce sampling in the target only while keeping all voxels in OARs, a median reduction up to 71.6% was achieved, with 0.5% loss in the plan quality. The optimization time was reduced by a factor of 7.5 for the total AI selection model and a factor of 3.7 for the model with only target selection.Significance.The novel deep-learning voxel sampling technique achieves a significant reduction in computational time with a negligible loss in the plan quality. The reduction in optimization time can be especially useful for future real-time adaptation strategies.

A novel, end-to-end framework for avoiding collisions between the patient's body and gantry in proton therapy

BACKGROUND

Administration of external radiation therapy via proton therapy systems carries a risk of occasional collisions between the patient's body and gantry, which is increased by the snout placed near the patient for better dose distribution. Although treatment planning software (TPS) can simulate controlled collisions, the computed tomography (CT) data used for treatment planning are insufficient given that collisions can occur outside the CT imaging region. Thus, imaging the three-dimensional (3D) surface outside the CT range and combining the data with those obtained by CT are essential for avoiding collisions.

PURPOSE

To construct a prototype for 3D surface imaging and an end-to-end framework for preventing collisions between the patient's body and the gantry.

METHODS

We obtained 3D surface data using a light sectioning method (LSM). By installing only cameras in front of the CT, we achieved LSM using the CT couch motion and preinstalled patient-positioning lasers. The camera image contained both sagittal and coronal lines, which are unnecessary for LSM and were removed by deep learning. We combined LSM 3D surface data and original CT data to create synthetic Digital Imaging and Communications in Medicine (DICOM) data. Subsequently, we compared the TPS snout auto-optimization using the original CT data with the synthetic DICOM data.

RESULTS

The mean positional error for LSM of the arms and head was 0.7 ± 0.8  and 0.8 ± 0.8 mm for axial and sagittal imaging, respectively. The TPS snout auto-optimization indicated that the original CT data would cause collisions; however, the synthetic DICOM data prevented these collisions.

CONCLUSIONS

The prototype system's acquisition accuracy for 3D surface data was approximately 1 mm, which was sufficient for the collision simulation. The use of a TPS with collision avoidance can help optimize the snout position using synthetic DICOM data. Our proposed method requires no external software for collision simulation and can be integrated into the clinical workflow to improve treatment planning efficiency.

Experimental comparison of photon versus particle computed tomography to predict tissue relative stopping powers

PURPOSE

Measurements comparing relative stopping power (RSP) accuracy of state-of-the-art systems representing single-energy and dual-energy computed tomography (SECT/DECT) with proton CT (pCT) and helium CT (HeCT) in biological tissue samples.

METHODS

We used 16 porcine and bovine samples of various tissue types and water, covering an RSP range from 0.90 ± 0.06 to 1.78 ± 0.05. Samples were packed and sealed into 3D-printed cylinders ( d = 2  cm, h = 5  cm) and inserted into an in-house designed cylindrical polymethyl methacrylate (PMMA) phantom ( d = 10  cm, h = 10  cm). We scanned the phantom in a commercial SECT and DECT (120 kV; 100  and 140 kV/Sn (tin-filtered)); and acquired pCT and HeCT ( E ∼ 200  MeV/u, 2 ∘ steps, ∼ 6.2 × 10 6 (p)/ ∼ 2.3 × 10 6 (He) particles/projection) with a particle imaging prototype. RSP maps were calculated from SECT/DECT using stoichiometric methods and from pCT/HeCT using the DROP-TVS algorithm. We estimated the average RSP of each tissue per modality in cylindrical volumes of interest and compared it to ground truth RSP taken from peak-detection measurements.

RESULTS

Throughout all samples, we observe the following root-mean-squared RSP prediction errors ± combined uncertainty from reference measurement and imaging: SECT 3.10 ± 2.88%, DECT 0.75 ± 2.80%, pCT 1.19 ± 2.81%, and HeCT 0.78 ± 2.81%. The largest mean errors ± combined uncertainty per modality are SECT 8.22 ± 2.79% in cortical bone, DECT 1.74 ± 2.00% in back fat, pCT 1.80 ± 4.27% in bone marrow, and HeCT 1.37 ± 4.25% in bone marrow. Ring artifacts were observed in both pCT and HeCT reconstructions, imposing a systematic shift to predicted RSPs.

CONCLUSION

Comparing state-of-the-art SECT/DECT technology and a pCT/HeCT prototype, DECT provided the most accurate RSP prediction, closely followed by particle imaging. The novel modalities pCT and HeCT have the potential to further improve on RSP accuracies with work focusing on the origin and correction of ring artifacts. Future work will study accuracy of proton treatment plans using RSP maps from investigated imaging modalities.

Experimental determination of modulation power of lung tissue for particle therapy

In particle therapy of lung tumors, modulating effects on the particle beam may occur due to the microscopic structure of the lung tissue. These effects are caused by the heterogeneous nature of the lung tissue and cannot be completely taken into account during treatment planning, because these micro structures are too small to be fully resolved in the planning CT. In several publications, a new material parameter called modulation power (P_mod) was introduced to characterize the effect. For various artificial lung surrogates, this parameter was measured and published by other groups and ranges up to approximately 1000μm. Studies investigating the influence of the modulation power on the dose distribution during irradiation are using this parameter in the rang of 100-800μm. More precise measurements forP_modon real lung tissue have not yet been published. In this work, the modulation power of real lung tissue was measured using porcine lungs in order to produce more reliable data ofP_modfor real lung tissue. For this purpose,ex-vivoporcine lungs were frozen in a ventilated state and measurements in a carbon ion-beam were performed. Due to the way the lungs were prepared and transferred to a solid state, the lung structures that modulate the beam could also be examined in detail using micro CT imaging. An optimization of the established methods of measuring the modulation power, which takes better account of the typical structures within lung tissue, was developed as well.

Compensating for beam modulation due to microscopic lung heterogeneities in carbon ion therapy treatment planning

PURPOSE

To predict and mitigate for the degradation in physical and biologically effective dose distributions of particle beams caused by microscopic heterogeneities in lung tissue.

MATERIALS AND METHODS

The TRiP98 treatment planning system was adapted to account for the beam-modulating effect of heterogeneous lung tissue in physical and biological inverse treatment planning. The implementation employs an analytical model that derives the degradation from the established "modulation power" parameter P mod and the total water-equivalent thickness of lung parenchyma traversed by the beam. Beam modulation was reproduced through an on-the-fly convolution of the reference Bragg curve with Gaussian kernels depending on the modulation power of lung tissue (upstream). For biological doses, the degradation was determined by modulating dose-averaged α , β , and LET distributions. Carbon SOBP measurements behind lung substitute material were performed to validate the code. The implementation was then applied to a phantom and patient case.

RESULTS

Experimental results show the passage through a 20-cm Gammex LN300 slab led to a decrease in target coverage and broadening of the SOBP distal fall-off. However, dose coverage was regained through optimization. A good agreement between calculated and measured SOBPs was also found. In addition, a patient case study revealed a 3.2% decrease in D 95 from degradation ( P mod = 450 μ m), which was reduced to a 0.4% difference after optimization. Furthermore, widening of the RBE distribution beyond the target distal edge was observed. This implies an increased degradation in the biological dose, which could be harmful to healthy tissues distal to the target.

CONCLUSIONS

This is the first implementation capable of compensating for lung dose perturbations, which is more effective than margin extensions. A larger patient study is needed to examine the observed modulation in the RBE distribution and judge the clinical relevance also in IMPT, where margins might prove insufficient to recover target coverage.

Estimating the modulating effect of lung tissue in particle therapy using a clinical CT voxel histogram analysis

To treat lung tumours with particle therapy, different additional tasks and challenges in treatment planning and application have to be addressed thoroughly. One of these tasks is the quantification and consideration of the Bragg peak degradation due to lung tissue: As lung is an heterogeneous tissue, the Bragg peak is broadened when particles traverse the microscopic alveoli. These are not fully resolved in clinical CT images and thus, the effect is not considered in the dose calculation. In this work, a correlation between the CT histograms of heterogeneous material and the impact on the Bragg peak curve is presented. Different inorganic materials were scanned with a conventional CT scanner and additionally, the Bragg peak degradation was measured in a proton beam and was then quantified. A model is proposed that allows an estimation of the modulation power by performing a histogram analysis on the CT scan. To validate the model for organic samples, a second measurement series was performed with frozen porcine lunge samples. This allows to investigate the possible limits of the proposed model in a set-up closer to clinical conditions. For lung substitutes, the agreement between model and measurement is within ±0.05 mm and for the organic lung samples, within ±0.15 mm. This work presents a novel, simple and efficient method to estimate if and how much a material or a distinct region (within the lung) is degrading the Bragg peak on the basis of a common clinical CT image. Up until now, only a direct in-beam measurement of the region or material of interest could answer this question.

A Monte Carlo Determination of Dose and Range Uncertainties for Preclinical Studies with a Proton Beam

Proton therapy (PRT) is an irradiation technique that aims at limiting normal tissue damage while maintaining the tumor response. To study its specificities, the ARRONAX cyclotron is currently developing a preclinical structure compatible with biological experiments. A prerequisite is to identify and control uncertainties on the ARRONAX beamline, which can lead to significant biases in the observed biological results and dose-response relationships, as for any facility. This paper summarizes and quantifies the impact of uncertainty on proton range, absorbed dose, and dose homogeneity in a preclinical context of cell or small animal irradiation on the Bragg curve, using Monte Carlo simulations. All possible sources of uncertainty were investigated and discussed independently. Those with a significant impact were identified, and protocols were established to reduce their consequences. Overall, the uncertainties evaluated were similar to those from clinical practice and are considered compatible with the performance of radiobiological experiments, as well as the study of dose-response relationships on this proton beam. Another conclusion of this study is that Monte Carlo simulations can be used to help build preclinical lines in other setups.

Application of lung substitute material as ripple filter for multi-ion therapy with helium-, carbon-, oxygen-, and neon-ion beams

A development project for hypo-fractionated multi-ion therapy has been initiated at the National Institute of Radiological Sciences in Japan. In the treatment, helium, carbon, oxygen, and neon ions will be used as primary beams with pencil beam scanning. A ripple filter (RiFi), consisting of a thin plastic or aluminum plate with a fine periodic ridge and groove structure, has been used to broaden the Bragg peak of heavy-ion beams in the beam direction. To sufficiently broaden the Bragg peak of helium-, carbon-, oxygen-, and neon-ion beams with suppressed lateral scattering and surface dose inhomogeneity, in this study, we tested a plate made of a lung substitute material, Gammex LN300, as the RiFi. The planar integrated dose distribution of a 183.5-MeV/u neon-ion beam was measured behind a 3-cm-thick LN300 plate in water. The Bragg peak of the pristine beam was broadened following the normal distribution with the standard deviation value of 1.29 mm, while the range of the beam was reduced by 8.8 mm by the plate. To verify the LN300 performance as the RiFi in multi-ion therapy, we measured the pencil beam data of helium-, carbon-, oxygen, and neon-ion beams penetrating the 3-cm-thick LN300 plate. The data were then modeled and used in a treatment planning system to achieve a uniform 10% survival of human undifferentiated carcinoma cells within a cuboid target by the beam for each of the different ion species. The measured survival fractions were reasonably reproduced by the planned ones for all the ion species. No surface dose inhomogeneity was observed for any ion species even when the plate was placed close to the phantom surface. The plate made of lung substitute material, Gammex LN300, is applicable as the RiFi in multi-ion therapy with helium-, carbon-, oxygen, and neon-ion beams.

Investigation of the Bragg peak degradation caused by homogeneous and heterogeneous lung tissue substitutes: proton beam experiments and comparison to current clinical dose calculation

Submillimetre structures of lung tissue are not represented in computed tomography images used for radiotherapeutic dose calculation. In order to study the effect experimentally, lung substitutes with properties similar to lung tissue were chosen, namely two types of commercial lung tissue equivalent plates (LTEPs) (CIRS, USA), two types of cork, balsawood, floral foam and konjac sponge. Laterally integrated dose profiles were measured as a function of depth for proton pencil beams (PBs) with an initial nominal energy of 97.4 and 148.2 MeV, respectively. The obtained dose profiles were investigated for their shifting and degradation of the Bragg peak (BP) caused by the materials, expressed as water equivalent thickness (WET) and full width half maximum. The set-up was simulated in the treatment planning system (TPS) RayStation using the Monte Carlo (MC) dose calculation algorithm. While the WET between experiment and dose calculation agreed within 0.5 mm, except for floral foam, the full width half maximum was underestimated in the TPS by up to 2.3 mm. Normalisation to the same mass thickness of the lung substitutes allowed to classify LTEPs and balsawood as homogeneous and cork, floral foam and konjac sponge as heterogeneous materials. The material specific BP degradation was up to 3.4 times higher for the heterogeneous samples. The modulation power as a measure for the heterogeneity was compared to the spectrum of Hounsfield units (HU) of the materials. A clear correlation was not found, but with further improvements the HU spectrum may serve as an indicator for the material heterogeneity. Further, MC simulations of binary voxel models using GATE/Geant4 were performed to investigate the influence of grain size and mass density. For mass densities similar to lung tissue the BP degradation had a maximum at 3 and 7 mm grain size.

Pro-con of proton: Dosimetric advantages of intensity-modulation over passive scatter for thoracic malignancies

•

Intensity Modulated Proton Therapy (IMPT) results in significant reduction of dose to organ at risk.

•

Improving plan robustness mitigates interplay effects.

•

Blanket use of small spots on a group of patients may severely worsen interplay in selected patients.

•

Hypofractionated regimes have fewer interplay effects in both fractional and overall simulations.

•

Randomised control trials are required before any clinical benefit of IMPT can be confirmed.

The use of passively scattered proton therapy (PSPT) or intensity modulated proton therapy (IMPT) opens the potential for dose escalation or critical structure sparing in thoracic malignancies. While the latter offers greater dose conformality, dose distributions are subjected to greater uncertainties, especially due to interplay effects. Exploration in this area is warranted to determine if there is any dosimetric advantages in using IMPT for thoracic malignancies. This review aims to both compare organs-at-risk sparing and plan robustness between PSPT and IMPT and examine the mitigation strategies for the reduction of interplay effects currently available. Early evidence suggests that IMPT is dosimetrically superior to PSPT in thoracic malignancies. Randomised control trials are required before any clinical benefit of IMPT can be confirmed.

Statistical limitations in proton imaging

Proton imaging is a promising technology for proton radiotherapy as it can be used for: (1) direct sampling of the tissue stopping power, (2) input information for multi-modality RSP reconstruction, (3) gold-standard calibration against concurrent techniques, (4) tracking motion and (5) pre-treatment positioning. However, no end-to-end characterization of the image quality (signal-to-noise ratio and spatial resolution, blurring uncertainty) against the dose has been done. This work aims to establish a model relating these characteristics and to describe their relationship with proton energy and object size. The imaging noise originates from two processes: the Coulomb scattering with the nucleus, producing a path deviation, and the energy loss straggling with electrons. The noise is found to increases with thickness crossed and, independently, decreases with decreasing energy. The scattering noise is dominant around high-gradient edge whereas the straggling noise is maximal in homogeneous regions. Image quality metrics are found to behave oppositely against energy: lower energy minimizes both the noise and the spatial resolution, with the optimal energy choice depending on the application and location in the imaged object. In conclusion, the model presented will help define an optimal usage of proton imaging to reach the promised application of this technology and establish a fair comparison with other imaging techniques.

The Promise of Proton Therapy for Central Nervous System Malignancies

Radiation therapy plays a significant role in management of benign and malignant diseases of the central nervous system. Patients may be at risk of acute and late toxicity from radiation therapy due to dose deposition in critical normal structures. In contrast to conventional photon delivery techniques, proton therapy is characterized by Bragg peak dose deposition which results in decreased exit dose beyond the target and greater sparing of normal structure which may reduce the rate of late toxicities from treatment. Dosimetric studies have demonstrated reduced dose to normal structures using proton therapy as compared to photon therapy. In addition, clinical studies are being reported demonstrating safety, feasibility, and low rates of acute toxicity. Technical challenges in proton therapy remain, including full understanding of depth of proton penetration and the biological activity in the distal Bragg peak. In addition, longer clinical follow-up is required to demonstrate reduction in late toxicities as compared to conventional photon-based radiation techniques. In this review, we summarize the current clinical literature and areas of active investigation in proton therapy for adult central nervous system malignancies.

Comparing 2 Monte Carlo Systems in Use for Proton Therapy Research

Purpose

Several Monte Carlo transport codes are available for medical physics users. To ensure confidence in the accuracy of the codes, they must be continually cross-validated. This study provides comparisons between MC² and Tool for Particle Simulation (TOPAS) simulations, that is, between medical physics applications for Monte Carlo N-Particle Transport Code (MCNPX) and Geant4.

Materials and Methods

Monte Carlo simulations were repeated with 2 wrapper codes: TOPAS (based on Geant4) and MC² (based on MCNPX). Simulations increased in geometrical complexity from a monoenergetic beam incident on a water phantom, to a monoenergetic beam incident on a water phantom with a bone or tissue slab at various depths, to a spread-out Bragg peak incident on a voxelized computed tomography (CT) geometry. The CT geometry cases consisted of head and neck tissue and lung tissue. The results of the simulations were compared with one another through dose or energy deposition profiles, r ₉₀ calculations, and γ-analyses.

Results

Both codes gave very similar results with monoenergetic beams incident on a water phantom. Systematic differences were observed between MC² and TOPAS simulations when using a lung or bone slab in a water phantom, particularly in the r ₉₀ values, where TOPAS consistently calculated r ₉₀ to be deeper by about 0.4%. When comparing the performance of the 2 codes in a CT geometry, the results were still very similar, exemplified by a 3-dimensional γ-analysis pass rate > 95% at the 2%-2-mm criterion for tissues from both head and neck and lung.

Conclusion

Differences between TOPAS and MC² were minor and were not considered clinically relevant.

Effects of the Bragg peak degradation due to lung tissue in proton therapy of lung cancer patients

Purpose

To quantify the effects of the Bragg peak degradation due to lung tissue on treatment plans of lung cancer patients with spot scanning proton therapy and to give a conservative approximation of these effects.

Methods and materials

Treatment plans of five lung cancer patients (tumors of sizes 2.7–46.4 cm³ at different depths in the lung) were optimized without consideration of the Bragg peak degradation. These treatment plans were recalculated with the Monte Carlo code TOPAS in two scenarios: in a first scenario, the treatment plans were calculated without including the Bragg peak degradation to reproduce the dose distribution predicted by the treatment-planning system (TPS). In a second scenario, the treatment plans were calculated while including the Bragg peak degradation. Subsequently, the plans were compared by means of D_mean, D_98% and D_2% in the clinical target volume (CTV) and organs at risk (OAR). Furthermore, isodose lines were investigated and a gamma index analysis was performed.

Results

The Bragg peak degradation leads to a lower dose in the CTV and higher doses in OARs distal to the CTV compared to the prediction from the TPS. The reduction of the mean dose in the CTV was − 5% at maximum and − 2% on average. The deeper a tumor was located in the lung and the smaller its volume the bigger was the effect on the CTV. The enhancement of the mean dose in OARs distal to the CTV was negligible for the cases investigated.

Conclusions

Effects of the Bragg peak degradation due to lung tissue were investigated for lung cancer treatment plans in proton therapy. This study confirms that these effects are clinically tolerable to a certain degree in the current clinical context considering the various more critical dose uncertainties due to motion and range uncertainties in proton therapy.

3D-printable lung phantom for distal falloff verification of proton Bragg peak

In proton therapy, the Bragg peak of a proton beam reportedly deteriorates when passing though heterogeneous structures such as human lungs. Previous studies have used heterogeneous random voxel phantoms, in which soft tissues and air are randomly allotted to render the phantoms the same density as human lungs, for conducting Monte Carlo (MC) simulations. However, measurements of these phantoms are complicated owing to their difficult‐to‐manufacture shape. In the present study, we used Voronoi tessellation to design a phantom that can be manufactured, and prepared a Voronoi lung phantom for which both measurement and MC calculations are possible. Our aim was to evaluate the effectiveness of this phantom as a new lung phantom for investigating proton beam Bragg peak deterioration. For this purpose, we measured and calculated the percentage depth dose and the distal falloff widths (DFW) passing through the phantom. For the 155 MeV beam, the measured and calculated DFW values with the Voronoi lung phantom were 0.40 and 0.39 cm, respectively. For the 200 MeV beam, the measured and calculated DFW values with the Voronoi lung phantom were both 0.48 cm. Our results indicate that both the measurements and MC calculations exhibited high reproducibility with plastinated lung sample from human body in previous studies. We found that better results were obtained using the Voronoi lung phantom than using other previous phantoms. The designed phantom may contribute significantly to the improvement of measurement precision. This study suggests that the Voronoi lung phantom is useful for simulating the effects of the heterogeneous structure of lungs on proton beam deterioration.

Quantification of the dependencies of the Bragg peak degradation due to lung tissue in proton therapy on a CT-based lung tumor phantom

The fine, sub-millimeter sized structure of lung tissue causes a degradation of the Bragg peak curve in particle therapy. The Bragg peak is degraded because particles of the same energy traverse lung tissue of different compositions of high and low density materials. Hence, they experience different energy losses resulting in variable ranges and a broadened Bragg peak. Since this fine structure of lung tissue is not resolved in standard treatment-planning CTs, current state-of-the-art dose calculation procedures used in the clinical routine are unable to account for this degradation. Neglecting this Bragg peak degradation in treatment planning can lead to an underdose in the target volume and an overdose distal to the target. Aim of this work is to systematically investigate the potential effects of the Bragg peak degradation on the dose distribution in dependence of different parameters like the tumor volume and its depth in lung. Proton plans were optimized on CT based phantoms without considering the Bragg peak degradation and afterwards recalculated with the Monte Carlo toolkit TOPAS: first, without consideration of the degradation and second, with the Bragg peak degradation accounted for. The direct comparison of these two dose distributions enables a quantification of the degradation effect. To carve out the dependencies of various parameters that could influence the Bragg peak degradation and thus the target dose, the simulations were performed for a variety of tumor sizes and shapes, as well as different positions within the lung. The results show that due to the Bragg peak degradation the mean dose in the target volume can be reduced by a few percent up to 14% (for extreme cases) depending on the geometry. It was shown that this effect increases with a decreasing tumor volume and increasing depth of the tumor. For the first time, a tumor specific estimation of the effect on the dose distribution due to the Bragg peak degradation in lung tissue is presented.

Motion effects in proton treatments of hepatocellular carcinoma-4D robustly optimised pencil beam scanning plans versus double scattering plans

Pencil beam scanning (PBS) proton therapy enables better dose conformality for complex anatomical geometries than passive proton scattering techniques, but is more susceptible to organ motion. This becomes an issue when treating moving tumours in the thorax or abdomen. Novel four-dimensional treatment planning approaches have been developed to increase the robustness of PBS plans against motion. However, their efficacy still needs to be examined by means of 4D dynamically accumulated dose (4DDD) analyses. This study investigates the potential use of 4D robust optimisation to maintain sufficient target coverage in the presence of organ motion, while sparing surrounding healthy tissue, for hepatocellular carcinoma (HCC). The liver is particularly suited to study motion interplay effects since the treatment region exhibits smaller density gradients and more homogeneous tissue than targets in the thorax, making it less prone to range errors. A facility-specific beam time model, developed and experimentally validated previously, was used for the clinical evaluation. 4DDD analyses of eleven target volumes did not show a significant improvement of the target coverage using 4D robust optimisation, but a reduction of the dose to close-by organs at risk. Interplay effects were averaged out for the applied fractionation scheme of 15 fractions. Contrary to PBS, passive double scattering (DS) plans yielded homogeneous 4DDD dose distributions in a single fraction. But, in some cases, they exceeded organ at risk dose limits, which were only satisfied in PBS. The average normal liver dose could be decreased by almost 6% compared to non-robustly optimised PBS plans and by 16% compared to DS plans when implementing 4D robust optimisation. Except for some very small tumours with large motion amplitudes, 4D robustly optimised PBS plans were found to be clinically acceptable even without supplementary motion mitigation techniques.

Advances in the use of motion management and image guidance in radiation therapy treatment for lung cancer

The development of advanced radiation technologies, including intensity-modulated radiation therapy (IMRT), stereotactic body radiation therapy (SBRT) and proton therapy, has resulted in increasingly conformal radiation treatments. Recent evidence for the importance of minimizing dose to normal critical structures including the heart and lungs has led to incorporation of these advanced treatment modalities into radiation therapy (RT) for non-small cell lung cancer (NSCLC). While such technologies have allowed for improved dose delivery, implementation requires improved target accuracy with treatments, placing increasing importance on evaluating tumor motion at the time of planning and verifying tumor position at the time of treatment. In this review article, we describe issues and updates related both to motion management and image guidance in the treatment of NSCLC.

Effectiveness of different rescanning techniques for scanned proton radiotherapy in lung cancer patients

Non-small cell lung cancer (NSCLC) is a tumour type thought to be well-suited for proton radiotherapy. However, the lung region poses many problems related to organ motion and can for actively scanned beams induce severe interplay effects. In this study we investigate four mitigating rescanning techniques: (1) volumetric rescanning, (2) layered rescanning, (3) breath-sampled (BS) layered rescanning, and (4) continuous breath-sampled (CBS) layered rescanning. The breath-sampled methods will spread the layer rescans over a full breathing cycle, resulting in an improved averaging effect at the expense of longer treatment times. In CBS, we aim at further improving the averaging by delivering as many rescans as possible within one breathing cycle. The interplay effect was evaluated for 4D robustly optimized treatment plans (with and without rescanning) for seven NSCLC patients in the treatment planning system RayStation. The optimization and final dose calculation used a Monte Carlo dose engine to account for the density heterogeneities in the lung region. A realistic treatment delivery time structure given from the IBA ScanAlgo simulation tool served as basis for the interplay evaluation. Both slow (2.0 s) and fast (0.1 s) energy switching times were simulated. For all seven studied patients, rescanning improves the dose conformity to the target. The general trend is that the breath-sampled techniques are superior to layered and volumetric rescanning with respect to both target coverage and variability in dose to OARs. The spacing between rescans in our breath-sampled techniques is set at planning, based on the average breathing cycle length obtained in conjunction with CT acquisition. For moderately varied breathing cycle lengths between planning and delivery (up to 15%), the breath-sampled techniques still mitigate the interplay effect well. This shows the potential for smooth implementation at the clinic without additional motion monitoring equipment.

Differences in Normal Tissue Response in the Esophagus Between Proton and Photon Radiation Therapy for Non-Small Cell Lung Cancer Using In Vivo Imaging Biomarkers

PURPOSE

To determine whether there exists any significant difference in normal tissue toxicity between intensity modulated radiation therapy (IMRT) or proton therapy for the treatment of non-small cell lung cancer.

METHODS AND MATERIALS

A total of 134 study patients (n=49 treated with proton therapy, n=85 with IMRT) treated in a randomized trial had a previously validated esophageal toxicity imaging biomarker, esophageal expansion, quantified during radiation therapy, as well as esophagitis grade (Common Terminology Criteria for Adverse Events version 3.0), on a weekly basis during treatment. Differences between the 2 modalities were statically analyzed using the imaging biomarker metric value (Kruskal-Wallis analysis of variance), as well as the incidence and severity of esophagitis grade (χ² and Fisher exact tests, respectively). The dose-response of the imaging biomarker was also compared between modalities using esophageal equivalent uniform dose, as well as delivered dose to an isotropic esophageal subvolume.

RESULTS

No statistically significant difference in the distribution of esophagitis grade, the incidence of grade ≥3 esophagitis (15 and 11 patients treated with IMRT and proton therapy, respectively), or the esophageal expansion imaging biomarker between cohorts (P>.05) was found. The distribution of imaging biomarker metric values had similar distributions between treatment arms, despite a slightly higher dose volume in the proton arm (P>.05). Imaging biomarker dose-response was similar between modalities for dose quantified as esophageal equivalent uniform dose and delivered esophageal subvolume dose. Regardless of treatment modality, there was high variability in imaging biomarker response, as well as esophagitis grade, for similar esophageal doses between patients.

CONCLUSIONS

There was no significant difference in esophageal toxicity from either proton- or photon-based radiation therapy as quantified by esophagitis grade or the esophageal expansion imaging biomarker.

Methodology for analysis and reporting patterns of failure in the Era of IMRT: head and neck cancer applications

Background

The aim of this study is to develop a methodology to standardize the analysis and reporting of the patterns of loco-regional failure after IMRT of head and neck cancer.

Material and Methods

Twenty-one patients with evidence of local and/or regional failure following IMRT for head-and-neck cancer were retrospectively reviewed under approved IRB protocol. Manually delineated recurrent gross disease (rGTV) on the diagnostic CT documenting recurrence (rCT) was co-registered with the original planning CT (pCT) using both deformable (DIR) and rigid (RIR) image registration software. Subsequently, mapped rGTVs were compared relative to original planning target volumes (TVs) and dose using a centroid-based approaches. Failures were then classified into five types based on combined spatial and dosimetric criteria; A (central high dose), B (peripheral high dose), C (central elective dose), D (peripheral elective dose), and E (extraneous dose).

Results

A total of 26 recurrences were identified. Using DIR, recurrences were assigned to more central TVs compared to RIR as detected using the spatial centroid-based method (p = 0.0002). rGTVs mapped using DIR had statistically significant higher mean doses when compared to rGTVs mapped rigidly (mean dose 70 vs. 69 Gy, p = 0.03). According to the proposed classification 22 out of 26 failures were of type A (central high dose) as assessed by DIR method compared to 18 out of 26 for the RIR because of the tendencey of RIR to assign failures more peripherally.

Conclusions

RIR tends to assigns failures more peripherally. DIR-based methods showed that the vast majority of failures originated in the high dose target volumes and received full prescribed doses suggesting biological rather than technology-related causes of failure. Validated DIR-based registration is recommended for accurate failure characterization and a novel typology-indicative taxonomy is recommended for failure reporting in the IMRT era.

Electronic supplementary material

The online version of this article (doi:10.1186/s13014-016-0678-7) contains supplementary material, which is available to authorized users.

Degradation of proton depth dose distributions attributable to microstructures in lung-equivalent material

PURPOSE

The purpose of the work reported here was to investigate the influence of sub-millimeter size heterogeneities on the degradation of the distal edges of proton beams and to validate Monte Carlo (MC) methods' ability to correctly predict such degradation.

METHODS

A custom-designed high-resolution plastic phantom approximating highly heterogeneous, lung-like structures was employed in measurements and in Monte Carlo simulations to evaluate the degradation of proton Bragg curves penetrating heterogeneous media.

RESULTS

Significant differences in distal falloff widths and in peak dose values were observed in the measured and the Monte Carlo simulated curves compared to pristine proton Bragg curves. Furthermore, differences between simulations of beams penetrating CT images of the phantom did not agree well with the corresponding experimental differences. The distal falloff widths in CT image-based geometries were underestimated by up to 0.2 cm in water (corresponding to 0.8-1.4 cm in lung tissue), and the peak dose values of pristine proton beams were overestimated by as much as ˜35% compared to measured curves or depth-dose curves simulated on the basis of true geometry. The authors demonstrate that these discrepancies were caused by the limited spatial resolution of CT images that served as a basis for dose calculations and lead to underestimation of the impact of the fine structure of tissue heterogeneities. A convolution model was successfully applied to mitigate the underestimation.

CONCLUSIONS

The results of this study justify further development of models to better represent heterogeneity effects in soft-tissue geometries, such as lung, and to correct systematic underestimation of the degradation of the distal edge of proton doses.

Site-specific range uncertainties caused by dose calculation algorithms for proton therapy

The purpose of this study was to assess the possibility of introducing site-specific range margins to replace current generic margins in proton therapy. Further, the goal was to study the potential of reducing margins with current analytical dose calculations methods. For this purpose we investigate the impact of complex patient geometries on the capability of analytical dose calculation algorithms to accurately predict the range of proton fields. Dose distributions predicted by an analytical pencil-beam algorithm were compared with those obtained using Monte Carlo (MC) simulations (TOPAS). A total of 508 passively scattered treatment fields were analyzed for seven disease sites (liver, prostate, breast, medulloblastoma-spine, medulloblastoma-whole brain, lung and head and neck). Voxel-by-voxel comparisons were performed on two-dimensional distal dose surfaces calculated by pencil-beam and MC algorithms to obtain the average range differences and root mean square deviation for each field for the distal position of the 90% dose level (R90) and the 50% dose level (R50). The average dose degradation of the distal falloff region, defined as the distance between the distal position of the 80% and 20% dose levels (R80-R20), was also analyzed. All ranges were calculated in water-equivalent distances. Considering total range uncertainties and uncertainties from dose calculation alone, we were able to deduce site-specific estimations. For liver, prostate and whole brain fields our results demonstrate that a reduction of currently used uncertainty margins is feasible even without introducing MC dose calculations. We recommend range margins of 2.8% + 1.2 mm for liver and prostate treatments and 3.1% + 1.2 mm for whole brain treatments, respectively. On the other hand, current margins seem to be insufficient for some breast, lung and head and neck patients, at least if used generically. If no case specific adjustments are applied, a generic margin of 6.3% + 1.2 mm would be needed for breast, lung and head and neck treatments. We conclude that the currently used generic range uncertainty margins in proton therapy should be redefined site specific and that complex geometries may require a field specific adjustment. Routine verifications of treatment plans using MC simulations are recommended for patients with heterogeneous geometries.

Range uncertainties in proton therapy and the role of Monte Carlo simulations

The main advantages of proton therapy are the reduced total energy deposited in the patient as compared to photon techniques and the finite range of the proton beam. The latter adds an additional degree of freedom to treatment planning. The range in tissue is associated with considerable uncertainties caused by imaging, patient setup, beam delivery and dose calculation. Reducing the uncertainties would allow a reduction of the treatment volume and thus allow a better utilization of the advantages of protons. This paper summarizes the role of Monte Carlo simulations when aiming at a reduction of range uncertainties in proton therapy. Differences in dose calculation when comparing Monte Carlo with analytical algorithms are analyzed as well as range uncertainties due to material constants and CT conversion. Range uncertainties due to biological effects and the role of Monte Carlo for in vivo range verification are discussed. Furthermore, the current range uncertainty recipes used at several proton therapy facilities are revisited. We conclude that a significant impact of Monte Carlo dose calculation can be expected in complex geometries where local range uncertainties due to multiple Coulomb scattering will reduce the accuracy of analytical algorithms. In these cases Monte Carlo techniques might reduce the range uncertainty by several mm.

Range uncertainties in proton therapy and the role of Monte Carlo simulations

The main advantages of proton therapy are the reduced total energy deposited in the patient as compared to photon techniques and the finite range of the proton beam. The latter adds an additional degree of freedom to treatment planning. The range in tissue is associated with considerable uncertainties caused by imaging, patient setup, beam delivery and dose calculation. Reducing the uncertainties would allow a reduction of the treatment volume and thus allow a better utilization of the advantages of protons. This paper summarizes the role of Monte Carlo simulations when aiming at a reduction of range uncertainties in proton therapy. Differences in dose calculation when comparing Monte Carlo with analytical algorithms are analyzed as well as range uncertainties due to material constants and CT conversion. Range uncertainties due to biological effects and the role of Monte Carlo for in vivo range verification are discussed. Furthermore, the current range uncertainty recipes used at several proton therapy facilities are revisited. We conclude that a significant impact of Monte Carlo dose calculation can be expected in complex geometries where local range uncertainties due to multiple Coulomb scattering will reduce the accuracy of analytical algorithms. In these cases Monte Carlo techniques might reduce the range uncertainty by several mm.