Impact of high-dose chemotherapy on the ability to deliver subsequent local-regional radiotherapy for breast cancer: analysis of Cancer and Leukemia Group B Protocol 9082

Radiation-induced Hounsfield unit change correlates with dynamic CT perfusion better than 4DCT-based ventilation measures in a novel-swine model

To analyze radiation induced changes in Hounsfield units and determine their correlation with changes in perfusion and ventilation. Additionally, to compare the post-RT changes in human subjects to those measured in a swine model used to quantify perfusion changes and validate their use as a preclinical model. A cohort of 5 Wisconsin Miniature Swine (WMS) were studied. Additionally, 19 human subjects were recruited as part of an IRB approved clinical trial studying functional avoidance radiation therapy for lung cancer and were treated with SBRT. Imaging (a contrast enhanced dynamic perfusion CT in the swine and 4DCT in the humans) was performed prior to and post-RT. Jacobian elasticity maps were calculated on all 4DCT images. Contours were created from the isodose lines to discretize analysis into 10 Gy dose bins. B-spline deformable image registration allowed for voxel-by-voxel comparative analysis in these contours between timepoints. The WMS underwent a research course of 60 Gy in 5 fractions delivered locally to a target in the lung using an MRI-LINAC system. In the WMS subjects, the dose-bin contours were copied onto the contralateral lung, which received < 5 Gy for comparison. Changes in HU and changes in Jacobian were analyzed in these contours. Statistically significant (p < 0.05) changes in the mean HU value post-RT compared to pre-RT were observed in both the human and WMS groups at all timepoints analyzed. The HU increased linearly with dose for both groups. Strong linear correlation was observed between the changes seen in the swine and humans (Pearson coefficient > 0.97, p < 0.05) at all timepoints. Changes seen in the swine closely modeled the changes seen in the humans at 12 months post RT (slope = 0.95). Jacobian analysis showed between 30 and 60% of voxels were damaged post-RT. Perfusion analysis in the swine showed a statistically significant (p < 0.05) reduction in contrast inside the vasculature 3 months post-RT compared to pre-RT. The increases in contrast outside the vasculature was strongly correlated (Pearson Correlation 0.88) with the reduction in HU inside the vasculature but were not correlated with the changes in Jacobians. Radiation induces changes in pulmonary anatomy at 3 months post-RT, with a strong linear correlation with dose. The change in HU seen in the non-vessel lung parenchyma suggests this metric is a potential biomarker for change in perfusion. Finally, this work suggests that the WMS swine model is a promising pre-clinical model for analyzing radiation-induced changes in humans and poses several benefits over conventional swine models.

Technical advances in x-ray microbeam radiation therapy

In the last 25 years microbeam radiation therapy (MRT) has emerged as a promising alternative to conventional radiation therapy at large, third generation synchrotrons. In MRT, a multi-slit collimator modulates a kilovoltage x-ray beam on a micrometer scale, creating peak dose areas with unconventionally high doses of several hundred Grays separated by low dose valley regions, where the dose remains well below the tissue tolerance level. Pre-clinical evidence demonstrates that such beam geometries lead to substantially reduced damage to normal tissue at equal tumour control rates and hence drastically increase the therapeutic window. Although the mechanisms behind MRT are still to be elucidated, previous studies indicate that immune response, tumour microenvironment, and the microvasculature may play a crucial role. Beyond tumour therapy, MRT has also been suggested as a microsurgical tool in neurological disorders and as a primer for drug delivery. The physical properties of MRT demand innovative medical physics and engineering solutions for safe treatment delivery. This article reviews technical developments in MRT and discusses existing solutions for dosimetric validation, reliable treatment planning and safety. Instrumentation at synchrotron facilities, including beam production, collimators and patient positioning systems, is also discussed. Specific solutions reviewed in this article include: dosimetry techniques that can cope with high spatial resolution, low photon energies and extremely high dose rates of up to 15 000 Gy s-1, dose calculation algorithms-apart from pure Monte Carlo Simulations-to overcome the challenge of small voxel sizes and a wide dynamic dose-range, and the use of dose-enhancing nanoparticles to combat the limited penetrability of a kilovoltage energy spectrum. Finally, concepts for alternative compact microbeam sources are presented, such as inverse Compton scattering set-ups and carbon nanotube x-ray tubes, that may facilitate the transfer of MRT into a hospital-based clinical environment. Intensive research in recent years has resulted in practical solutions to most of the technical challenges in MRT. Treatment planning, dosimetry and patient safety systems at synchrotrons have matured to a point that first veterinary and clinical studies in MRT are within reach. Should these studies confirm the promising results of pre-clinical studies, the authors are confident that MRT will become an effective new radiotherapy option for certain patients.

CBCT correction using a cycle-consistent generative adversarial network and unpaired training to enable photon and proton dose calculation

In presence of inter-fractional anatomical changes, clinical benefits are anticipated from image-guided adaptive radiotherapy. Nowadays, cone-beam CT (CBCT) imaging is mostly utilized during pre-treatment imaging for position verification. Due to various artifacts, image quality is typically not sufficient for photon or proton dose calculation, thus demanding accurate CBCT correction, as potentially provided by deep learning techniques. This work aimed at investigating the feasibility of utilizing a cycle-consistent generative adversarial network (cycleGAN) for prostate CBCT correction using unpaired training. Thirty-three patients were included. The network was trained to translate uncorrected, original CBCT images (CBCT_org) into planning CT equivalent images (CBCT_cycleGAN). HU accuracy was determined by comparison to a previously validated CBCT correction technique (CBCT_cor). Dosimetric accuracy was inferred for volumetric-modulated arc photon therapy (VMAT) and opposing single-field uniform dose (OSFUD) proton plans, optimized on CBCT_cor and recalculated on CBCT_cycleGAN. Single-sided SFUD proton plans were utilized to assess proton range accuracy. The mean HU error of CBCT_cycleGAN with respect to CBCT_cor decreased from 24 HU for CBCT_org to  -6 HU. Dose calculation accuracy was high for VMAT, with average pass-rates of 100%/89% for a 2%/1% dose difference criterion. For proton OSFUD plans, the average pass-rate for a 2% dose difference criterion was 80%. Using a (2%, 2 mm) gamma criterion, the pass-rate was 96%. 93% of all analyzed SFUD profiles had a range agreement better than 3 mm. CBCT correction time was reduced from 6-10 min for CBCT_cor to 10 s for CBCT_cycleGAN. Our study demonstrated the feasibility of utilizing a cycleGAN for CBCT correction, achieving high dose calculation accuracy for VMAT. For proton therapy, further improvements may be required. Due to unpaired training, the approach does not rely on anatomically consistent training data or potentially inaccurate deformable image registration. The substantial speed-up for CBCT correction renders the method particularly interesting for adaptive radiotherapy.

MRI-based treatment planning for proton radiotherapy: dosimetric validation of a deep learning-based liver synthetic CT generation method

Magnetic resonance imaging (MRI) has been widely used in combination with computed tomography (CT) radiation therapy because MRI improves the accuracy and reliability of target delineation due to its superior soft tissue contrast over CT. The MRI-only treatment process is currently an active field of research since it could eliminate systematic MR-CT co-registration errors, reduce medical cost, avoid diagnostic radiation exposure, and simplify clinical workflow. The purpose of this work is to validate the application of a deep learning-based method for abdominal synthetic CT (sCT) generation by image evaluation and dosimetric assessment in a commercial proton pencil beam treatment planning system (TPS). This study proposes to integrate dense block into a 3D cycle-consistent generative adversarial networks (cycle GAN) framework in an effort to effectively learn the nonlinear mapping between MRI and CT pairs. A cohort of 21 patients with co-registered CT and MR pairs were used to test the deep learning-based sCT image quality by leave-one-out cross validation. The CT image quality, dosimetric accuracy and the distal range fidelity were rigorously checked, using side-by-side comparison against the corresponding original CT images. The average mean absolute error (MAE) was 72.87±18.16 HU. The relative differences of the statistics of the PTV dose volume histogram (DVH) metrics between sCT and CT were generally less than 1%. Mean 3D gamma analysis passing rate of 1mm/1%, 2mm/2%, 3mm/3% criteria with 10% dose threshold were 90.76±5.94%, 96.98±2.93% and 99.37±0.99%, respectively. The median, mean and standard deviation of absolute maximum range differences were 0.170 cm, 0.186 cm and 0.155 cm. The image similarity, dosimetric and distal range agreement between sCT and original CT suggests the feasibility of further development of an MRI-only workflow for liver proton radiotherapy.

Impact of backscatter material thickness on the depth dose of orthovoltage irradiators for radiobiology research

The orthovoltage x-ray energy frequently used in radiation research is prone to dosimetry errors due to insufficient backscatter conditions. In many radiobiology studies, especially for cell irradiations, precise dose calculation algorithms such as Convolution-Superposition or Monte Carlo are impractical and as such, less accurate hand calculation methods are used for dose estimation. These dose estimation methods typically assume full backscatter conditions. The purpose of this study is to demonstrate the magnitude of the dose error that results from insufficient backscatter, and to provide lookup tables to account this issue. The beam spectra of several widely used commercial systems (XRAD-225, XRAD-320, SARRP) were used in Monte Carlo (MC) simulations on a series of phantom setups to investigate the impact of varying backscatter conditions on dosimetry. The depth dose curves for different field sizes, water phantom thicknesses and beam qualities were generated. In addition, depth dependent backscatter factors for different field sizes and different beam qualities were calculated. It is demonstrated that as much as a 50% dose difference exists for different backscatter conditions at the beam qualities studied. The choice of cell dish size as well as other changes in the experiment setup can have more than 10% impact on the dose. The impact of backscatter is reduced with a decrease in field size. Further, the thickness needed to provide full backscatter can be approximated as being equal to the field size. It is imperative to ensure full backscatter conditions during system and dosimeter calibration, or to use the look-up table provided in this study.

Comparing Unet training with three different datasets to correct CBCT images for prostate radiotherapy dose calculations

Image intensity correction is crucial to enable cone beam computed tomography (CBCT) based radiotherapy dose calculations. This study evaluated three different deep learning based correction methods using a U-shaped convolutional neural network architecture (Unet) in terms of their photon and proton dose calculation accuracy. CT and CBCT imaging data of 42 prostate cancer patients were included. For target ground truth data generation, a CBCT correction method based on CT to CBCT deformable image registration (DIR) was used. The method yields a deformed CT called (i) virtual CT (vCT) which is used to generate (ii) corrected CBCT projections allowing the reconstruction of (iii) a final corrected CBCT image. The single Unet architecture was trained using these three different datasets: (Unet1) raw and corrected CBCT projections, (Unet2) raw CBCT and vCT image slices and (Unet3) raw and reference corrected CBCT image slices. Volumetric arc therapy (VMAT) and proton pencil beam scanning (PBS) single field uniform dose (SFUD) plans were optimized on the reference corrected image and recalculated on the obtained Unet-corrected CBCT images. The mean error (ME) and mean absolute error (MAE) for Unet1/2/3 were [Formula: see text] Hounsfield units (HU) and [Formula: see text] HU. The 1% dose difference pass rates were better than 98.4% for VMAT for 8 test patients not seen during training, with little difference between Unets. Gamma evaluation results were even better. For protons a gamma evaluation was employed to account for small range shifts, and [Formula: see text] mm pass rates for Unet1/2/3 were better than [Formula: see text] and 91%. A 3 mm range difference threshold was established. Only for Unet3 the 5th and 95th percentiles of the range difference distributions over all fields, test patients and dose profiles were within this threshold. A single Unet architecture was successfully trained using both CBCT projections and CBCT image slices. Since the results of the other Unets were poorer than Unet3, we conclude that training using corrected CBCT image slices as target data is optimal for PBS SFUD proton dose calculations, while for VMAT all Unets provided sufficient accuracy.

Hypo-fractionated boost in locally advanced non-small cell lung cancer: temporal distribution of boost fractions

To propose new schemas for radiation boosting of primary tumors, in locally advanced non-small cell lung cancers (NSCLC), in conjunction with standard chemoradiotherapy. To investigate the effect of temporal distributions of the boost fractions on tumor control. NSCLC cases, previously treated with 60 Gy in 30 fractions, were retrospectively planned by adding a radiation boost (25 Gy in 5 fractions) to the primary tumor. Several integrated and sequential boosting schedules were considered. Biological doses were calculated for targets and organs at risk (OAR). Tumor control probabilities (TCP) were calculated using an empirical model and a stochastic model that accounts more systematically for tumor growth kinetics and cell kill. For heterogeneous patient populations, the TCPs for different boost schedules ranged from 82% to 84% and from 73% to 74% for integrated and sequential boosting respectively. For individual tumors with specific growth parameters, the TCP varied by up to 19% between the different schedules. The TCP for sequential boosting was expected to be up to 67% lower than front integrated boosting. The gap in TCP between schedules was higher for tumors with higher clonogenic cell numbers, lower radio-sensitivity, shorter doubling times and lower cell loss. The proposed boosting schemas are dosimetrically feasible and biologically effective. We suggest that the boosts are most effective when given during the first week of treatment and least effective when given sequentially after the end of treatment. The effect of boost scheduling and the effectiveness of front boosting are expected to be most significant for tumors with high clonogenic cell numbers, fast growing rates, low cell loss and low radio-sensitivity. Ultimately, animal studies and clinical trials, guided by biology modeling as presented in the present work, will be needed to verify the effectiveness of fine tuning temporal distributions of radiotherapy fractions.

Exploration and application of phenomenological RBE models for proton therapy

The relative biological effectiveness (RBE) of protons varies with multiple physical and biological factors. Phenomenological RBE models have been developed to include such factors in the estimation of a variable RBE, in contrast to the clinically applied constant RBE of 1.1. In this study, eleven published phenomenological RBE models and two plan-based models were explored and applied to simulated patient cases. All models were analysed with respect to the distribution and range of linear energy transfer (LET) and reference radiation fractionation sensitivity ((α/β) _x ) of their respective experimental databases. Proton therapy plans for a spread-out Bragg peak in water and three patient cases (prostate adenocarcinoma, pituitary adenoma and thoracic sarcoma) were optimised using an RBE of 1.1 in the Eclipse^™ treatment planning system prior to recalculation and modelling in the FLUKA Monte Carlo code. Model estimated dose-volume parameters for the planning target volumes (PTVs) and organs at risk (OAR) were compared. The experimental in vitro databases for the various models differed greatly in the range of (α/β) _x values and dose-averaged LET (LET_d). There were significant variations between the model estimations, which arose from fundamental differences in the database definitions and model assumptions. The greatest variations appeared in organs with low (α/β) _x and high LET_d, e.g. biological doses given to late responding OARs located distal to the target in the treatment field. In general, the variation in maximum dose (D_2%) was larger than the variation in mean dose and other dose metrics, with D_2% of the left optic nerve ((α/β) _x   =  2.1 Gy) in the pituitary adenoma case showing the greatest discrepancies between models: 28-52 Gy(RBE), while D_2% for RBE_1.1 was 30 Gy(RBE). For all patient cases, the estimated mean RBE to the PTV was in the range 1.09-1.29 ((α/β) _x   =  1.5/3.1/10.6 Gy). There were considerable variations between the estimations of RBE and RBE-weighted doses from the different models. These variations were a consequence of fundamental differences in experimental databases, model assumptions and regression techniques. The results from the implementation of RBE models in dose planning studies should be evaluated in light of these deviations.

Contour scanning for penumbra improvement in pencil beam scanned proton therapy

Proton therapy, especially in the form of pencil beam scanning (PBS), allows for the delivery of highly conformal dose distributions for complex tumor geometries. However, due to scattering of protons inside the patient, lateral dose gradients cannot be arbitrarily steep, which is of importance in cases with organs at risk (OARs) in close proximity to, or overlapping with, planning target volumes (PTVs). In the PBS approach, physical pencil beams are planned using a regular grid orthogonal to the beam direction. In this work, we propose an alternative to this commonly used approach where pencil beams are placed on an irregular grid along concentric paths based on the target contour. Contour driven pencil beam placement is expected to improve dose confirmation by allowing the optimizer to best enhance the penumbra of irregularly shaped targets using edge enhancement. Its effectiveness has been shown to improve dose confirmation to the target volume and reduce doses to OARs in head-and-neck planning studies. Furthermore, the deliverability of such plans, as well as the dosimetric improvements over conventional grid-based plans, have been confirmed in first phantom based verifications.

High-dose chemotherapy and autologous bone marrow or stem cell transplantation versus conventional chemotherapy for women with early poor prognosis breast cancer

BACKGROUND

Overall survival rates are disappointing for women with early poor prognosis breast cancer. Autologous transplantation of bone marrow or peripheral stem cells (in which the woman is both donor and recipient) has been considered a promising technique because it permits use of much higher doses of chemotherapy.

OBJECTIVES

To compare the effectiveness and safety of high-dose chemotherapy and autograft (either autologous bone marrow or stem cell transplantation) with conventional chemotherapy for women with early poor prognosis breast cancer.

SEARCH METHODS

We searched the Cochrane Breast Cancer Group Specialised Register, MEDLINE (1966 to October 2015), EMBASE (1980 to October 2015), the World Health Organization's International Clinical Trials Registry Search Platform, and ClinicalTrials.gov on the 21 October 2015.

SELECTION CRITERIA

Randomised controlled trials (RCTs) comparing high-dose chemotherapy and autograft (bone marrow transplant or stem cell rescue) versus chemotherapy without autograft for women with early poor prognosis breast cancer.

DATA COLLECTION AND ANALYSIS

Two review authors selected RCTs, independently extracted data and assessed risks of bias. We combined data using a Mantel-Haenszel fixed-effect model to calculate pooled risk ratios (RRs) and 95% confidence intervals (CIs). We assessed the quality of the evidence using GRADE methods. Outcomes were survival rates, toxicity and quality of life.

MAIN RESULTS

We included 14 RCTs of 5600 women randomised to receive high-dose chemotherapy and autograft (bone marrow transplant or stem cell rescue) versus chemotherapy without autograft for women with early poor prognosis breast cancer. The studies were at low risk of bias in most areas.There is high-quality evidence that high-dose chemotherapy does not increase the likelihood of overall survival at any stage of follow-up (at three years: RR 1.02, 95% CI 0.95 to 1.10, 3 RCTs, 795 women, I² = 56%; at five years: RR 1.00, 95% CI 0.96 to 1.04, 9 RCTs, 3948 women, I² = 0%; at six years: RR 0.94, 95% CI 0.81 to 1.08, 1 RCT, 511 women; at eight years: RR1.17, 95% CI 0.95 to 1.43, 1 RCT, 344 women; at 12 years: RR 1.18, 95% CI 0.99 to 1.42, 1 RCT, 382 women).There is high-quality evidence that high-dose chemotherapy improves the likelihood of event-free survival at three years (RR 1.19, 95% CI 1.06 to 1.34, 3 RCTs, 795 women, I² = 56%) but this effect was no longer apparent at longer duration of follow-up (at five years: RR 1.04, 95% CI 0.99 to 1.09, 9 RCTs, 3948 women, I² = 14%; at six years RR 1.04, 95% CI 0.87 to 1.24, 1 RCT, 511 women; at eight years: RR 1.27, 95% CI 0.99 to 1.64, 1 RCT, 344 women; at 12 years: RR 1.18, 95% CI 0.95 to 1.45, 1 RCT, 382 women).Treatment-related deaths were much more frequent in the high-dose arm (RR 7.97, 95% CI 3.99 to 15.92, 14 RCTs, 5600 women, I² = 12%, high-quality evidence) and non-fatal morbidity was also more common and more severe in the high-dose group. There was little or no difference between the groups in the incidence of second cancers at four to nine years' median follow-up (RR 1.25, 95% CI 0.90 to 1.73, 7 RCTs, 3423 women, I² = 0%, high-quality evidence). Women in the high-dose group reported significantly worse quality-of-life scores immediately after treatment, but there were few statistically significant differences between the groups by one year.The primary studies were at low risk of bias in most areas, and the evidence was assessed using GRADE methods and rated as high quality for all comparisons.

AUTHORS' CONCLUSIONS

There is high-quality evidence of increased treatment-related mortality and little or no increase in survival by using high-dose chemotherapy with autograft for women with early poor prognosis breast cancer.

Head and neck intensity modulated radiotherapy parotid glands: time of re-planning

PURPOSE

To investigate the correct time point for re-planning by evaluating dosimetric changes in the parotid glands (PGs) during intensity-modulated radiotherapy (IMRT) in head and neck cancer patients.

MATERIALS AND METHODS

Patients with head and neck cancer treated with IMRT were enrolled. During treatment all patients underwent cone-beam computed tomography (CBCT) scans to verify the set-up. CBCT scans at treatment days 10, 15, 20 and 25 were used to transfer the original plan (CBCTplan I, II, III, IV, respectively) using rigid registration between the two. The PGs were retrospectively contoured and evaluated with the dose-volume histogram. The mean dose, the dose to 50 % of volume, and the percentage of volume receiving 30 and 50 Gy were evaluated for each PG. The Wilcoxon sign ranked test was used to evaluate the effects of dosimetric variations and values <0.05 were taken to be significant.

RESULTS

From February to June 2011, ten patients were enrolled and five IMRT plans were evaluated for each patient. All the dosimetric parameters increased throughout the treatment course. However, this increase was statistically significant at treatment days 10 and 15 (CBCTplan I, II; p = 0.02, p = 0.03, respectively).

CONCLUSION

CBCT is a feasible method to assess the dosimetric changes in the PGs. Our data showed that checking the PG volume and dose could be indicated during the third week of treatment.

What we have learned: the impact of quality from a clinical trials perspective

In this review article, we address the radiation oncology process improvements in clinical trials and review how these changes improve the quality for the next generation of trials. In recent years, we have progressed from a time of limited data acquisition to the present in which we have real-time influence of clinical trials quality. This enables immediate availability of the important elements, including staging, eligibility, response, and outcome for all trial investigators. Modern informatics platforms are well designed for future adaptive clinical trials. We review what will be needed in the informatics architecture of current and future clinical trials.