Phase 2 Trial of Consolidation Pembrolizumab After Proton Reirradiation for Thoracic Recurrences of Non-Small Cell Lung Cancer

PURPOSE

Reirradiation (reRT) with proton beam therapy (PBT) may offer a chance of cure while minimizing toxicity for patients with isolated intrathoracic recurrences of non-small cell lung cancer (NSCLC). However, distant failure remains common, necessitating strategies to integrate more effective systemic therapy.

METHODS AND MATERIALS

This was a phase 2, single-arm trial (NCT03087760) of consolidation pembrolizumab after PBT reRT for locoregional recurrences of NSCLC. Four to 12 weeks after completion of 60 to 70 Gy PBT reRT, patients without progressive disease received pembrolizumab for up to 12 months. Primary endpoint was progression-free survival (PFS), measured from the start of reRT. Secondary endpoints were overall survival (OS) and National Cancer Institute Common Terminology Criteria for Adverse Events, version 5.0 toxicity.

RESULTS

Between 2017 and 2021, 22 patients received PBT reRT. Median interval from prior radiation end to reRT start was 20 months. Most recurrences (91%) were centrally located. Most patients received concurrent chemotherapy (95%) and pencil beam scanning PBT (77%), and 36% had received prior durvalumab. Fifteen patients (68%) initiated consolidation pembrolizumab on trial and received a median of 3 cycles (range, 2-17). Pembrolizumab was discontinued most commonly due to toxicity (n = 5; 2 were pembrolizumab-related), disease progression (n = 4), and completion of 1 year (n = 3). Median follow-up was 38.7 months. Median PFS and OS were 8.8 months (95% CI, 4.2-23.7) and 22.8 months (95% CI, 6.9-not reached), respectively. There was only one isolated in-field failure after reRT. Grade ≥3 toxicities occurred in 10 patients (45%); 2 were pembrolizumab-related. There were 2 grade 5 toxicities, an aorto-esophageal fistula at 6.9 months and hemoptysis at 46.8 months, both probably from reRT. The trial closed early due to widespread adoption of immunotherapy off-protocol.

CONCLUSIONS

In the first-ever prospective trial combining PBT reRT with consolidation immunotherapy, PFS was acceptable and OS favorable. Late grade 5 toxicity occurred in 2 of 22 patients. This approach may be considered in selected patients with isolated thoracic recurrences of NSCLC.

Pneumonitis Rates Before and After Adoption of Immunotherapy Consolidation in Patients With Locally Advanced Non-Small Cell Lung Cancer Treated With Concurrent Chemoradiation

PURPOSE

We hypothesized that after adoption of immune checkpoint inhibitor (ICI) consolidation for patients with locally advanced non-small cell lung cancer (LA-NSCLC) receiving concurrent chemoradiation therapy (cCRT), rates of symptomatic pneumonitis would increase, thereby supporting efforts to reduce lung radiation dose.

METHODS AND MATERIALS

This single institution, multisite retrospective study included 783 patients with LA-NSCLC treated with definitive cCRT either before introduction of ICI consolidation (pre-ICI era cohort [January 2011-September 2017]; N = 448) or afterward (ICI era cohort [October 2017-December 2021]; N = 335). Primary endpoint was grade ≥2 pneumonitis (G2P) and secondary endpoint was grade ≥3 pneumonitis (G3P), per Common Terminology Criteria for Adverse Events v5.0. Pneumonitis was compared between pre-ICI era and ICI era cohorts using the cumulative incidence function and Gray's test. Inverse probability of treatment weighting (IPTW)-adjusted Fine-Gray models were generated. Logistic models were developed to predict the 1-year probability of G2P as a function of lung dosimetry.

RESULTS

G2P was higher in the ICI era than in the pre-ICI era (1-year cumulative incidence 31.4% vs 20.1%; P < .001; IPTW-adjusted multivariable subdistribution hazard ratio, 2.03; 95% confidence interval, 1.53-2.70; P < .001). There was no significant interaction between ICI era treatment and either lung volume receiving ≥20 Gy (V20) or mean lung dose in Fine-Gray regression for G2P; however, the predicted probability of G2P was higher in the ICI era at clinically relevant values of lung V20 (≥24%) and mean lung dose (≥14 Gy). Cut-point analysis revealed a lung V20 threshold of 28% in the ICI era (1-year G2P rate 46.0% above vs 19.8% below; P < .001). Among patients receiving ICI consolidation, lung V5 was not associated with G2P. G3P was not higher in the ICI era (1-year cumulative incidence 7.5% vs 6.0%; P = .39; IPTW-adjusted multivariable subdistribution hazard ratio, 1.12; 95% confidence interval, 0.63-2.01; P = .70).

CONCLUSIONS

In patients with LA-NSCLC treated with cCRT, the adoption of ICI consolidation was associated with an increase in G2P but not G3P. With ICI consolidation, stricter lung dose constraints may be warranted.

FLASH radiotherapy for the treatment of symptomatic bone metastases in the thorax (FAST-02): protocol for a prospective study of a novel radiotherapy approach

BACKGROUND

FLASH therapy is a treatment technique in which radiation is delivered at ultra-high dose rates (≥ 40 Gy/s). The first-in-human FAST-01 clinical trial demonstrated the clinical feasibility of proton FLASH in the treatment of extremity bone metastases. The objectives of this investigation are to assess the toxicities of treatment and pain relief in study participants with painful thoracic bone metastases treated with FLASH radiotherapy, as well as workflow metrics in a clinical setting.

METHODS

This single-arm clinical trial is being conducted under an FDA investigational device exemption (IDE) approved for 10 patients with 1-3 painful bone metastases in the thorax, excluding bone metastases in the spine. Treatment will be 8 Gy in a single fraction administered at ≥ 40 Gy/s on a FLASH-enabled proton therapy system delivering a single transmission proton beam. Primary study endpoints are efficacy (pain relief) and safety. Patient questionnaires evaluating pain flare at the treatment site will be completed for 10 consecutive days post-RT. Pain response and adverse events (AEs) will be evaluated on the day of treatment and on day 7, day 15, months 1, 2, 3, 6, 9, and 12, and every 6 months thereafter. The outcomes for clinical workflow feasibility are the occurrence of any device issues as well as time on the treatment table.

DISCUSSION

This prospective clinical trial will provide clinical data for evaluating the efficacy and safety of proton FLASH for palliation of bony metastases in the thorax. Positive findings will support the further exploration of FLASH radiation for other clinical indications including patient populations treated with curative intent.

REGISTRATION

ClinicalTrials.gov NCT05524064.

Deformable lung 4DCT image registration via landmark-driven cycle network

BACKGROUND

An automated, accurate, and efficient lung four-dimensional computed tomography (4DCT) image registration method is clinically important to quantify respiratory motion for optimal motion management.

PURPOSE

The purpose of this work is to develop a weakly supervised deep learning method for 4DCT lung deformable image registration (DIR).

METHODS

The landmark-driven cycle network is proposed as a deep learning platform that performs DIR of individual phase datasets in a simulation 4DCT. This proposed network comprises a generator and a discriminator. The generator accepts moving and target CTs as input and outputs the deformation vector fields (DVFs) to match the two CTs. It is optimized during both forward and backward paths to enhance the bi-directionality of DVF generation. Further, the landmarks are used to weakly supervise the generator network. Landmark-driven loss is used to guide the generator's training. The discriminator then judges the realism of the deformed CT to provide extra DVF regularization.

RESULTS

We performed four-fold cross-validation on 10 4DCT datasets from the public DIR-Lab dataset and a hold-out test on our clinic dataset, which included 50 4DCT datasets. The DIR-Lab dataset was used to evaluate the performance of the proposed method against other methods in the literature by calculating the DIR-Lab Target Registration Error (TRE). The proposed method outperformed other deep learning-based methods on the DIR-Lab datasets in terms of TRE. Bi-directional and landmark-driven loss were shown to be effective for obtaining high registration accuracy. The mean and standard deviation of TRE for the DIR-Lab datasets was 1.20 ± 0.72 mm and the mean absolute error (MAE) and structural similarity index (SSIM) for our datasets were 32.1 ± 11.6 HU and 0.979 ± 0.011, respectively.

CONCLUSION

The landmark-driven cycle network has been validated and tested for automatic deformable image registration of patients' lung 4DCTs with results comparable to or better than competing methods.

NRG Oncology and PTCOG Patterns of Practice Survey and Consensus Recommendations on Pencil-Beam Scanning Proton Stereotactic Body Radiation Therapy and Hypofractionated Radiation Therapy for Thoracic Malignancies

Stereotactic body radiation therapy (SBRT) and hypofractionation using pencil-beam scanning (PBS) proton therapy (PBSPT) is an attractive option for thoracic malignancies. Combining the advantages of target coverage conformity and critical organ sparing from both PBSPT and SBRT, this new delivery technique has great potential to improve the therapeutic ratio, particularly for tumors near critical organs. Safe and effective implementation of PBSPT SBRT/hypofractionation to treat thoracic malignancies is more challenging than the conventionally fractionated PBSPT because of concerns of amplified uncertainties at the larger dose per fraction. The NRG Oncology and Particle Therapy Cooperative Group Thoracic Subcommittee surveyed proton centers in the United States to identify practice patterns of thoracic PBSPT SBRT/hypofractionation. From these patterns, we present recommendations for future technical development of proton SBRT/hypofractionation for thoracic treatment. Among other points, the recommendations highlight the need for volumetric image guidance and multiple computed tomography-based robust optimization and robustness tools to minimize further the effect of uncertainties associated with respiratory motion. Advances in direct motion analysis techniques are urgently needed to supplement current motion management techniques.

Proton Pencil-Beam Scanning Stereotactic Body Radiation Therapy and Hypofractionated Radiation Therapy for Thoracic Malignancies: Patterns of Practice Survey and Recommendations for Future Development from NRG Oncology and PTCOG

Stereotactic body radiation therapy (SBRT) and hypofractionation using pencil-beam scanning (PBS) proton therapy (PBSPT) is an attractive option for thoracic malignancies. Combining the advantages of target coverage conformity and critical organ sparing from both PBSPT and SBRT, this new delivery technique has great potential to improve the therapeutic ratio, particularly for tumors near critical organs. Safe and effective implementation of PBSPT SBRT/hypofractionation to treat thoracic malignancies is more challenging than the conventionally-fractionated PBSPT due to concerns of amplified uncertainties at the larger dose per fraction. NRG Oncology and Particle Therapy Cooperative Group (PTCOG) Thoracic Subcommittee surveyed US proton centers to identify practice patterns of thoracic PBSPT SBRT/hypofractionation. From these patterns, we present recommendations for future technical development of proton SBRT/hypofractionation for thoracic treatment. Amongst other points, the recommendations highlight the need for volumetric image guidance and multiple CT-based robust optimization and robustness tools to minimize further the impact of uncertainties associated with respiratory motion. Advances in direct motion analysis techniques are urgently needed to supplement current motion management techniques.

Acute hospitalizations after proton therapy versus intensity-modulated radiotherapy for locally advanced non-small cell lung cancer in the durvalumab era

INTRODUCTION

It was hypothesized that use of proton beam therapy (PBT) in patients with locally advanced non-small cell lung cancer treated with concurrent chemoradiation and consolidative immune checkpoint inhibition is associated with fewer unplanned hospitalizations compared with intensity-modulated radiotherapy (IMRT).

METHODS

Patients with locally advanced non-small cell lung cancer treated between October 2017 and December 2021 with concurrent chemoradiation with either IMRT or PBT ± consolidative immune checkpoint inhibition were retrospectively identified. Logistic regression was used to assess the association of radiation therapy technique with 90-day hospitalization and grade 3 (G3+) lymphopenia. Competing risk regression was used to compare G3+ pneumonitis, G3+ esophagitis, and G3+ cardiac events. Kaplan-Meier method was used for progression-free survival and overall survival. Inverse probability treatment weighting was applied to adjust for differences in PBT and IMRT groups.

RESULTS

Of 316 patients, 117 (37%) received PBT and 199 (63%) received IMRT. The PBT group was older (p < .001) and had higher Charlson Comorbidity Index scores (p = .02). The PBT group received a lower mean heart dose (p < .0001), left anterior descending artery V15 Gy (p = .001), mean lung dose (p = .008), and effective dose to immune circulating cells (p < .001). On inverse probability treatment weighting analysis, PBT was associated with fewer unplanned hospitalizations (adjusted odds ratio, 0.55; 95% CI, 0.38-0.81; p = .002) and less G3+ lymphopenia (adjusted odds ratio, 0.55; 95% CI, 0.37-0.81; p = .003). There was no difference in other G3+ toxicities, progression-free survival, or overall survival.

CONCLUSIONS

PBT is associated with fewer unplanned hospitalizations, lower effective dose to immune circulating cells and less G3+ lymphopenia compared with IMRT. Minimizing dose to lymphocytes may be warranted, but prospective data are needed.

Cardiac radiation dose is associated with inferior survival but not cardiac events in patients with locally advanced non-small cell lung cancer in the era of immune checkpoint inhibitor consolidation

PURPOSE

We assessed the association of cardiac radiation dose with cardiac events and survival post-chemoradiation therapy (CRT) in patients with locally advanced non-small cell lung cancer (LA-NSCLC) after adoption of modern radiation therapy (RT) techniques, stricter cardiac dose constraints, and immune checkpoint inhibitor (ICI) consolidation.

METHODS AND MATERIALS

This single-institution, multi-site retrospective study included 335 patients with LA-NSCLC treated with definitive, concurrent CRT between October 2017 and December 2021. All patients were evaluated for ICI consolidation. Planning dose constraints included heart mean dose < 20 Gy (<10 Gy if feasible) and heart volume receiving ≥ 50 Gy (V50Gy) < 25 %. Twenty-one dosimetric parameters for three different cardiac structures (heart, left anterior descending coronary artery [LAD], and left ventricle) were extracted. Primary endpoint was any major adverse cardiac event (MACE) post-CRT, defined as acute coronary syndrome, heart failure, coronary revascularization, or cardiac-related death. Secondary endpoints were: grade ≥ 3 cardiac events (per CTCAE v5.0), overall survival (OS), lung cancer-specific mortality (LCSM), and other-cause mortality (OCM).

RESULTS

Median age was 68 years, 139 (41 %) had baseline coronary heart disease, and 225 (67 %) received ICI consolidation. Proton therapy was used in 117 (35 %) and intensity-modulated RT in 199 (59 %). Median LAD V15Gy was 1.4 % (IQR 0-22) and median heart mean dose was 8.7 Gy (IQR 4.6-14.4). Median follow-up was 3.3 years. Two-year cumulative incidence of MACE was 9.5 % for all patients and 14.3 % for those with baseline coronary heart disease. Two-year cumulative incidence of grade ≥ 3 cardiac events was 20.4 %. No cardiac dosimetric parameter was associated with an increased risk of MACE or grade ≥ 3 cardiac events. On multivariable analysis, cardiac dose (LAD V15Gy and heart mean dose) was associated with worse OS, driven by an association with LCSM but not OCM.

CONCLUSIONS

With modern RT techniques, stricter cardiac dose constraints, and ICI consolidation, cardiac dose was associated with LCSM but not OCM or cardiac events in patients with LA-NSCLC.

Plunging Into the PACIFIC: Outcomes of Patients With Unresectable KRAS-Mutated Non-Small Cell Lung Cancer Following Definitive Chemoradiation and Durvalumab Consolidation

BACKGROUND

Immune checkpoint inhibitor (ICI) consolidation following concurrent chemoradiotherapy (CRT) substantially improved progression free survival (PFS) and overall survival (OS) in the PACIFIC trial becoming the standard of care in locally-advanced, unresectable NSCLC. KRAS mutation may influence response to ICI.

METHODS

In this single-institution, retrospective analysis, we compared treatment outcomes for patients with unresectable KRAS mutated (KRAS-mt) and wild-type (KRAS-wt) NSCLC treated with CRT between October 2017 and December 2021. Kaplan-Meier analysis was conducted comparing median progression free survival and median overall survival from completion of radiotherapy in all KRAS-mt patients and KRAS-G12C-mutated patients. Outcomes were also compared with and without ICI consolidation.

RESULTS

Of 156 patients, 42 (26.9%) were KRAS-mt and 114 (73.1%) were KRAS-wt. Baseline characteristics differed only in histology; KRAS-mt NSCLC more likely to be adenocarcinoma. KRAS-mt patients had worse PFS (median 6.3 vs. 10.7 months, P = .041) but similar OS (median 23.1 vs. 27.3 months, P = .237). KRAS-mt patients were more likely to not receive ICI due to rapid disease progression post-CRT (23.8% vs. 4.4%, P = .007). Among patients who received ICI (n = 114), KRAS-mt was not associated with inferior PFS (8.1 vs. 11.9 months, P = .355) or OS (30.5 vs. 31.7 months, P = .692). KRAS-G12C patients (n = 22) had similar PFS and OS to other KRAS-mt.

CONCLUSION

In one of the largest post-CRT KRAS-mt cohort published, KRAS-mt was associated with inferior PFS, largely due to rapid progression prior to ICI consolidation, but did not affect OS. Among those who received ICI consolidation, outcomes were comparable regardless of KRAS-mt status.

A Novel Inverse Algorithm To Solve the Integrated Optimization of Dose, Dose Rate, and Linear Energy Transfer of Proton FLASH Therapy With Sparse Filters

PURPOSE

The recently proposed Integrated Physical Optimization Intensity Modulated Proton Therapy (IPO-IMPT) framework allows simultaneous optimization of dose, dose rate, and linear energy transfer (LET) for ultra-high dose rate (FLASH) treatment planning. Finding solutions to IPO-IMPT is difficult because of computational intensiveness. Nevertheless, an inverse solution that simultaneously specifies the geometry of a sparse filter and weights of a proton intensity map is desirable for both clinical and preclinical applications. Such solutions can reduce effective biologic dose to organs at risk in patients with cancer as well as reduce the number of animal irradiations needed to derive extra biologic dose models in preclinical studies.

METHODS AND MATERIALS

Unlike the initial forward heuristic, this inverse IPO-IMPT solution includes simultaneous optimization of sparse range compensation, sparse range modulation, and spot intensity. The daunting computational tasks vital to this endeavor were resolved iteratively with a distributed computing framework to enable Simultaneous Intensity and Energy Modulation and Compensation (SIEMAC). SIEMAC was demonstrated on a human patient with central lung cancer and a minipig.

RESULTS

SIEMAC simultaneously improves maps of spot intensities and patient-field-specific sparse range compensators and range modulators. For the patient with lung cancer, at our maximum nozzle current of 300 nA, dose rate coverage above 100 Gy/s increased from 57% to 96% in the lung and from 93% to 100% in the heart, and LET coverage above 4 keV/µm dropped from 68% to 9% in the lung and from 26% to <1% in the heart. For a simple minipig plan, the full-width half-maximum of the dose, dose rate, and LET distributions decreased by 30%, 1.6%, and 57%, respectively, again with similar target dose coverage, thus reducing uncertainty in these quantities for preclinical studies.

CONCLUSIONS

The inverse solution to IPO-IMPT demonstrated the capability to simultaneously modulate subspot proton energy and intensity distributions for clinical and preclinical studies.

Safety and efficacy of durvalumab after concurrent chemoradiation in Black patients with locally advanced non-small cell lung cancer

BACKGROUND

The PACIFIC trial established consolidative durvalumab after concurrent chemoradiation as standard-of-care in patients with stage III or unresectable non-small cell lung cancer (NSCLC). Black patients, however, comprised just 2% (n = 14) of randomized patients in this trial, warranting real-world evaluation of the PACIFIC regimen in these patients.

METHODS

This single-institution, multi-site study included 105 patients with unresectable stage II/III NSCLC treated with concurrent chemoradiation followed by durvalumab between 2017 and 2021. Overall survival (OS), progression-free survival (PFS), and grade ≥3 pneumonitis-free survival (PNFS) were compared between Black and non-Black patients using Kaplan-Meier and Cox regression analyses.

RESULTS

A total of 105 patients with a median follow-up of 22.8 months (interquartile range, 11.3-37.3 months) were identified for analysis, including 57 Black (54.3%) and 48 (45.7%) non-Black patients. The mean radiation prescription dose was higher among Black patients (61.5 ± 2.9 Gy vs. 60.5 ± 1.9 Gy; p = .031), but other treatment characteristics were balanced between groups. The median OS (not-reached vs. 39.7 months; p = .379) and PFS (31.6 months vs. 19.3 months; p = .332) were not statistically different between groups. Eight (14.0%) Black patients discontinued durvalumab due to toxicity compared to 13 (27.1%) non-Black patients (p = .096). The grade ≥3 pneumonitis rate was similar between Black and non-Black patients (12.3% vs. 12.5%; p = .973), and there was no significant difference in time to grade ≥3 PNFS (p = .904). Three (5.3%) Black patients and one (2.1%) non-Black patient developed grade 5 pneumonitis.

CONCLUSIONS

The efficacy and tolerability of consolidative durvalumab after chemoradiation appears to be comparable between Black and non-Black patients.

Interpretable Machine Learning for Choosing Radiation Dose-volume Constraints on Cardio-pulmonary Substructures Associated with Overall Survival in NRG Oncology RTOG 0617

PURPOSE

Our objective was to use interpretable machine learning for choosing dose-volume constraints on cardiopulmonary substructures (CPSs) associated with overall survival (OS) in radiation therapy for locally advanced non-small cell lung cancer.

METHODS AND MATERIALS

A total of 428 patients with non-small cell lung cancer were randomly divided into training/validation/test subsets (n = 230/149/49) in Radiation Therapy Oncology Group 0617. Manual or automated contouring was performed to segment CPSs, including heart, atria, ventricles, aorta, left/right ventricle/atrium (LV+RV+LA+RA), inferior/superior vena cava, pulmonary artery, and pericardium. Peri (pericardium-heart), rest (heart-[LV+RV+LA+RA]), clinical target volume (CTV), and lungs-CTV contours were also obtained. Dose-volume histogram features were extracted, including minimum/mean dose to the hottest x% volume (Dx%[Gy]/MOHx%[Gy]), minimum/mean/maximum dose, percent volume receiving at least xGy (VxGy[%]), and overlapping volume of each CPS with planning target volume (PTV_Voverlap[%]). Clinical parameters were collected from the National Clinical Trials Network/Community oncology research program data archive. Feature selection was performed using a series of multiblock sparse partial least squares regression, stability selection supervised principal component analysis, and Boruta. Explainable boosting machine (EBM) was trained using a conditional survival distribution-based approach for imputing censored data, treating survival analysis as a regression problem. Harrell's C-index was used to evaluate OS discrimination performance of EBM, Cox proportional hazards (CPH), random survival forest, extreme gradient boosting survival embeddings, and CPH deep neural network (DeepSurv) models in the test set. Dose-volume constraints were selected using the binary change point detection algorithm in Shapley additive explanations-based partial dependence functions.

RESULTS

Selected features included LA_V60Gy(%), pericardium_D30%(Gy), lungs-CTV_PTV_Voverlap(%), RA_V55Gy(%), and received_cons_chemo. All models ranked LA_V60Gy(%) as the most important feature. EBM achieved the best performance for predicting OS, followed by extreme gradient boosting survival embeddings, random survival forest, DeepSurv, and CPH (C-index = 0.653, 0.646, 0.642, 0.638, and 0.632). EBM global explanations suggested that LA_V60Gy(%) < 25.6, lungs-CTV_PTV_Voverlap(%) < 1.1, pericardium_D30%(Gy) < 18.9, RA_V55Gy(%) < 19.5, and received_cons_chemo = 'Yes' for improved OS.

CONCLUSIONS

EBM can be used to discriminate OS while also guiding dose-volume constraint selection for optimal management of cardiac toxicity in lung cancer radiation therapy.

Multimodal imaging-based material mass density estimation for proton therapy using supervised deep learning

OBJECTIVE

Mapping CT number to material property dominates the proton range uncertainty. This work aims to develop a physics-constrained deep learning-based multimodal imaging (PDMI) framework to integrate physics, deep learning, MRI, and advanced dual-energy CT (DECT) to derive accurate patient mass density maps.

METHODS

Seven tissue substitute MRI phantoms were used for validation including adipose, brain, muscle, liver, skin, spongiosa, 45% hydroxyapatite (HA) bone. MRI images were acquired using T1 weighted Dixon and T2 weighted short tau inversion recovery sequences. Training inputs are from MRI and twin-beam dual-energy images acquired at 120 kVp with gold/tin filters. The feasibility investigation included an empirical model and four residual networks (ResNet) derived from different training inputs and strategies by PDMI framework. PRN-MR-DE and RN-MR-DE denote ResNet (RN) trained with and without a physics constraint (P) using MRI (MR) and DECT (DE) images. PRN-DE stands for RN trained with a physics constraint using only DE images. A retrospective study using institutional patient data was also conducted to investigate the feasibility of the proposed framework.

RESULTS

For the tissue surrogate study, PRN-MR-DE, PRN-DE, and RN-MR-DE result in mean mass density errors: -0.72%/2.62%/-3.58% for adipose; -0.03%/-0.61%/-0.18% for muscle; -0.58%/-1.36%/-4.86% for 45% HA bone. The retrospective patient study indicated that PRN-MR-DE predicted the densities of soft tissue and bone within expected intervals based on the literature survey, while PRN-DE generated large density deviations.

CONCLUSION

The proposed PDMI framework can generate accurate mass density maps using MRI and DECT images. The supervised learning can further enhance model efficacy, making PRN-MR-DE outperform RN-MR-DE. The patient investigation also shows that the framework can potentially improve proton range uncertainty with accurate patient mass density maps.

ADVANCES IN KNOWLEDGE

PDMI framework is proposed for the first time to inform deep learning models by physics insights and leverage the information from MRI to derive accurate mass density maps.

Single energy CT-based mass density and relative stopping power estimation for proton therapy using deep learning method

Background

The number of patients undergoing proton therapy has increased in recent years. Current treatment planning systems (TPS) calculate dose maps using three-dimensional (3D) maps of relative stopping power (RSP) and mass density. The patient-specific maps of RSP and mass density were obtained by translating the CT number (HU) acquired using single-energy computed tomography (SECT) with appropriate conversions and coefficients. The proton dose calculation uncertainty of this approach is 2.5%-3.5% plus 1 mm margin. SECT is the major clinical modality for proton therapy treatment planning. It would be intriguing to enhance proton dose calculation accuracy using a deep learning (DL) approach centered on SECT.

Objectives

The purpose of this work is to develop a deep learning method to generate mass density and relative stopping power (RSP) maps based on clinical single-energy CT (SECT) data for proton dose calculation in proton therapy treatment.

Methods

Artificial neural networks (ANN), fully convolutional neural networks (FCNN), and residual neural networks (ResNet) were used to learn the correlation between voxel-specific mass density, RSP, and SECT CT number (HU). A stoichiometric calibration method based on SECT data and an empirical model based on dual-energy CT (DECT) images were chosen as reference models to evaluate the performance of deep learning neural networks. SECT images of a CIRS 062M electron density phantom were used as the training dataset for deep learning models. CIRS anthropomorphic M701 and M702 phantoms were used to test the performance of deep learning models.

Results

For M701, the mean absolute percentage errors (MAPE) of the mass density map by FCNN are 0.39%, 0.92%, 0.68%, 0.92%, and 1.57% on the brain, spinal cord, soft tissue, bone, and lung, respectively, whereas with the SECT stoichiometric method, they are 0.99%, 2.34%, 1.87%, 2.90%, and 12.96%. For RSP maps, the MAPE of FCNN on M701 are 0.85%, 2.32%, 0.75%, 1.22%, and 1.25%, whereas with the SECT reference model, they are 0.95%, 2.61%, 2.08%, 7.74%, and 8.62%.

Conclusion

The results show that deep learning neural networks have the potential to generate accurate voxel-specific material property information, which can be used to improve the accuracy of proton dose calculation.

Advances in knowledge

Deep learning-based frameworks are proposed to estimate material mass density and RSP from SECT with improved accuracy compared with conventional methods.

The Inherited KRAS-variant as a Biomarker of Cetuximab Response in NSCLC

PURPOSE

RTOG 0617 was a phase III randomized trial for patients with unresectable stage IIIA/IIIB non-small cell lung cancer comparing standard-dose (60 Gy) versus high-dose (74 Gy) radiotherapy and chemotherapy, plus or minus cetuximab. Although the study was negative, based on prior evidence that patients with the KRAS-variant, an inherited germline mutation, benefit from cetuximab, we evaluated KRAS-variant patients in RTOG 0617.

EXPERIMENTAL DESIGN

From RTOG 0617, 328 of 496 (66%) of patients were included in this analysis. For time-to-event outcomes, stratified log-rank tests and multivariable Cox regression models were used. For binary outcomes, Cochran-Mantel-Haenzel tests and multivariable logistic regression models were used. All statistical tests were two sided, and a P value <0.05 was considered significant.

RESULTS

A total of 17.1% (56/328) of patients had the KRAS-variant, and overall survival rates were similar between KRAS-variant and non-variant patients. However, there was a time-dependent effect of cetuximab seen only in KRAS-variant patients-while the hazard of death was higher in cetuximab-treated patients within year 1 [HR = 3.37, 95% confidence interval (CI): 1.13-10.10, P = 0.030], death was lower from year 1 to 4 (HR = 0.33, 95% CI: 0.11-0.97, P = 0.043). In contrast, in non-variant patients, the addition of cetuximab significantly increased local failure (HR = 1.59, 95% CI: 1.11-2.28, P = 0.012).

CONCLUSIONS/DISCUSSION

Although an overall survival advantage was not achieved in KRAS-variant patients, there is potential impact of cetuximab for this genetic subset of patients. In contrast, cetuximab seems to harm non-variant patients. These findings further support the importance of genetic patient selection in trials studying the addition of systemic agents to radiotherapy.

SIGNIFICANCE

The KRAS-variant is the first functional, inherited miRNA-disrupting variant identified in cancer. Our findings support that cetuximab has a potentially beneficial impact on KRAS-variant patients treated with radiation. The work confirms prior evidence that KRAS-variant patients are a subgroup who are especially sensitive to radiation. These findings further support the potential of this class of variants to enable true treatment personalization, considering the equally important endpoints of response and toxicity.

Comparing Changes in Medicare Reimbursement for Radiation Oncology and Medical Oncology (2010-2020)

PURPOSE/OBJECTIVE(S)

A recent study found that radiation oncology (RO) has seen significant declines in Medicare reimbursement (MCR) from 2010-2019. While it is presumed that other cancer subspecialties have seen decreasing MCR, to our knowledge, there are no studies directly comparing changes in MCR between RO and other oncology subspecialties. In this study, we analyze changes in MCR from 2010-2020 for both RO and medical oncology. We hypothesized that the declines in MCR will be similar between the two fields.

MATERIALS/METHODS

The publicly available Physician/Supplier Procedure Summary (PSPS) database was used for all years from 2010-2020. All reimbursement for providers with primary provider codes 92 (RO) and 83 and 90 (heme/onc and medical oncology, respectively) were analyzed. For the 150 most highly-reimbursed HCPCS codes for each specialty in 2010, the total allowed charge for each code was corrected for inflation and then divided by the number of submitted claims to calculate average MCR per code for each year. For each code and each specialty, the 2020 billing frequency was multiplied by the calculated average reimbursement per claim in a given year to calculate what the reimbursement would have been in that year using 2020 dollars and utilization rates (projected reimbursement). The projected reimbursement was summed for all HCPCS codes in each year for each specialty to calculate an aggregate MCR for that specialty for that year. This aggregate MCR was then compared with the actual 2020 reimbursement for that basket of codes to calculate the change in MCR over time.

RESULTS

Both medical and radiation oncology saw decreases in projected vs. actual MCR from 2010-2020 for this basket of services (Table). Adjusting for inflation and utilization, RO MCR declined by $0.7 billion (B) (-29.0%) from 2010 to 2020 and by $0.2B (-10.5%) from 2015 to 2020 while medical oncology MCR declined by $0.8B (-14.7%) from 2010-2020 and by $0.4B (-6.6%) from 2015-2020. The average decrease per year in projected vs. actual reimbursement for RO was 2.9% (2010 to 2015) and 1.05% (2015 to 2020) and for medical oncology was 1.5% (2010-2015) and 0.7% (2015-2020), respectively.

CONCLUSION

Adjusting for inflation, Medicare reimbursement for a large array of services has declined for both medical oncology and RO from 2010 - 2020. Contrary to our hypothesis, RO reported a 97% greater relative decline in reimbursement compared with medical oncology from 2010 - 2020. Significant decreases in reimbursement to both fields and their potential implications on patient care and access to care should be considered by policymakers while shifting towards an episode-based Alternative Payment Model and when considering further cuts to Medicare reimbursement.

Progression-Free Survival as a Surrogate Endpoint of Overall Survival in Patients with Locally Advanced Non-Small Cell Lung Cancer Treated with Chemoradiotherapy: Trial-Level Meta-Analysis and Individual-Level Analysis of NRG/RTOG 0617 and PROCLAIM

PURPOSE/OBJECTIVE(S)

Overall Survival (OS) is the gold standard endpoint in randomized clinical trials (RCTs) of Locally Advanced Non-Small Cell Lung Cancer (LA-NSCLC). Intermediate endpoints that can be observed at earlier time points and predict OS would improve trial efficiency and expedite the adoption of proven interventions.

MATERIALS/METHODS

Atrial-level meta-analysis was conducted using a weighted regression analysis to quantify the correlation between PFS and OS hazard ratios (HRs). Large (n≥ 100) contemporary RCTs in LA-NSCLC that used platinum-based chemoradiation were included. An individual-level surrogacy analysis based on Prentice criteria was performed to evaluate if PFS could reliably predict OS using NRG/RTOG 0617 (NCT00533949), a phase III RCT of dose escalated CRT. The individual-level correlation between PFS and OS was validated using PROCLAIM (NCT00686959) control arm.

RESULTS

Nineteen RCTs comprising a total of 5525 patients (pts) were included in the trial-level meta-analysis. A moderately high correlation was observed between PFS HR and OS HR (R2 = 0.68, 95% CI = 0.42-0.94). Individual-level analysis of NRG/RTOG 0617 showed that, as reported, RT dose was associated with OS (HR = 1.28, 95% CI = 1.04-1.58, p = 0.02) and PFS (HR = 1.21, 95% CI = 0.99-1.46, p = 0.06). Progressive disease (PD) was highly associated with OS, where pts having PD within 6mo or 12mo had a significantly higher mortality risk than those not having PD within 6mo or 12 mo, respectively, in landmark analysis (PD within 6mo: HR = 2.56, 95% CI = 1.82-3.59, p<0.0001; PD within 12mo: HR = 3.18, 95% CI = 2.45-4.12, p<0.0001). Accounting for PD moderately reduced RT dose effect on OS (HR = 1.21, 95% CI = 0.98-1.49), suggesting RT dose effect on OS may be mediated partially through PD. The association between OS and PD occurrence within 6mo or 12mo was similar in PROCLAIM control arm (PD within 6mo: HR = 2.06, 95% CI = 1.48-2.86, p<0.0001; PD within 12mo: HR = 2.02, 95% CI = 1.38-2.95, p<0.0001).

CONCLUSION

A moderately high trial-level surrogacy between PFS and OS was identified in trial-level meta-analysis. PD occurrence also reliably predicted OS at the individual patient level in both NRG/RTOG 0617 and PROCLAIM. These results support the use of PFS as a valid endpoint in clinical trials of LA-NSCLC.

Treatment of Thymoma and Thymic Carcinoma with Proton Beam Therapy: Outcomes from the Proton Collaborative Group Prospective Registry

PURPOSE/OBJECTIVE(S)

Given the generally long natural history of thymic malignancies, proton beam therapy (PBT) is advocated to minimize the risk of long-term toxicities to mediastinal organs. Adverse events (AE) and long-term clinical outcomes for this population have not been well-characterized.

MATERIALS/METHODS

The Proton Collaborative Group registry (NCT01255748), a multi-institutional prospective database of academic and community proton centers in the US, was queried for patients with thymomas and thymic carcinomas treated with PBT. Patients with recurrent/metastatic disease, non-thymic histology, received either prior or palliative radiotherapy (dose < 40 Gy RBE) were excluded. Overall survival (OS) and local control (LC) were estimated using Kaplan-Meier methods.

RESULTS

A total of 97 patients were identified in the PCG registry. After applying relevant exclusion criteria, 70 patients from 12 proton centers treated from 2011-2021 were included for analysis. Median follow-up length was 16 months. Median age was 58.5 years (IQR 46-63), and 60% were female. 81.4% had a diagnosis of thymoma, and 18.6% thymic carcinoma. 59 patients underwent surgical resection. 11 were treated with definitive PBT, of which 5 received concurrent chemotherapy. Median dose was 54 Gy RBE (range 41.4 - 70 Gy RBE), median number of fractions was 30 (range 21 - 38). 73.4% received pencil beam scanning and 23% uniform scanning PBT. Treatment was overall well-tolerated: a single patient developed grade 4 pneumonitis. Grade 3 AEs were seen in 3 patients - dyspnea, anorexia, and heart failure. Highest grade toxicity experienced was grade 2 for 47.1% and grade 1 for 42.9% of patients. 3-year overall survival (OS) was 82.6% for the entire cohort. 3-year OS was 94% for resected/adjuvant cohort and 35.6% in the non-surgical/definitive cohort. 3-year local control (LC) was 91.7% for the entire cohort. By surgery/margin status, 3-year LC was 96.8% in patients with close or negative margins (a single failure in a patient with close margins), whereas 3-year LC was 55.1% for patients with positive margins/unresectable disease.

CONCLUSION

Thymic malignancies treated with PBT appear to have favorable outcomes, especially in the adjuvant setting, in this cohort representing the largest series of such patients.

Automatic segmentation of neurovascular bundle on mri using deep learning based topological modulated network

PURPOSE

Radiation damage on neurovascular bundles (NVBs) may be the cause of sexual dysfunction after radiotherapy for prostate cancer. However, it is challenging to delineate NVBs as organ-at-risks from planning CTs during radiotherapy. Recently, the integration of MR into radiotherapy made NVBs contour delineating possible. In this study, we aim to develop an MRI-based deep learning method for automatic NVB segmentation.

METHODS

The proposed method, named topological modulated network, consists of three subnetworks, that is, a focal modulation, a hierarchical block and a topological fully convolutional network (FCN). The focal modulation is used to derive the location and bounds of left and right NVBs', namely the candidate volume-of-interests (VOIs). The hierarchical block aims to highlight the NVB boundaries information on derived feature map. The topological FCN then segments the NVBs inside the VOIs by considering the topological consistency nature of the vascular delineating. Based on the location information of candidate VOIs, the segmentations of NVBs can then be brought back to the input MRI's coordinate system.

RESULTS

A five-fold cross-validation study was performed on 60 patient cases to evaluate the performance of the proposed method. The segmented results were compared with manual contours. The Dice similarity coefficient (DSC) and 95th percentile Hausdorff distance (HD95 ) are (left NVB) 0.81 ± 0.10, 1.49 ± 0.88 mm, and (right NVB) 0.80 ± 0.15, 1.54 ± 1.22 mm, respectively.

CONCLUSION

We proposed a novel deep learning-based segmentation method for NVBs on pelvic MR images. The good segmentation agreement of our method with the manually drawn ground truth contours supports the feasibility of the proposed method, which can be potentially used to spare NVBs during proton and photon radiotherapy and thereby improve the quality of life for prostate cancer patients.

Deep learning-based fast volumetric imaging using kV and MV projection images for lung cancer radiotherapy: A feasibility study

PURPOSE

The long acquisition time of CBCT discourages repeat verification imaging, therefore increasing treatment uncertainty. In this study, we present a fast volumetric imaging method for lung cancer radiation therapy using an orthogonal 2D kV/MV image pair.

METHODS

The proposed model is a combination of 2D and 3D networks. The proposed model consists of five major parts: (1) kV and MV feature extractors are used to extract deep features from the perpendicular kV and MV projections. (2) The feature-matching step is used to re-align the feature maps to their projection angle in a Cartesian coordinate system. By using a residual module, the feature map can focus more on the difference between the estimated and ground truth images. (3) In addition, the feature map is downsized to include more global semantic information for the 3D estimation, which is useful to reduce inhomogeneity. By using convolution-based reweighting, the model is able to further increase the uniformity of image. (4) To reduce the blurry noise of generated 3D volume, the Laplacian latent space loss calculated via the feature map that is extracted via specifically-learned Gaussian kernel is used to supervise the network. (5) Finally, the 3D volume is derived from the trained model. We conducted a proof-of-concept study using 50 patients with lung cancer. An orthogonal kV/MV pair was generated by ray tracing through CT of each phase in a 4D CT scan. Orthogonal kV/MV pairs from nine respiratory phases were used to train this patient-specific model while the kV/MV pair of the remaining phase was held for model testing.

RESULTS

The results are based on simulation data and phantom results from a real Linac system. The mean absolute error (MAE) values achieved by our method were 57.5 HU and 77.4 HU within body and tumor region-of-interest (ROI), respectively. The mean achieved peak-signal-to-noise ratios (PSNR) were 27.6 dB and 19.2 dB within the body and tumor ROI, respectively. The achieved mean normalized cross correlation (NCC) values were 0.97 and 0.94 within the body and tumor ROI, respectively. A phantom study demonstrated that the proposed method can accurately re-position the phantom after shift. It is also shown that the proposed method using both kV and MV is superior to current method using kV or MV only in image quality.

CONCLUSION

These results demonstrate the feasibility and accuracy of our proposed fast volumetric imaging method from an orthogonal kV/MV pair, which provides a potential solution for daily treatment setup and verification of patients receiving radiation therapy for lung cancer.

Benchmarking the Radiation Oncology Alternative Payment Model: Changes in Medicare Reimbursement for 16 Common Radiation Therapy Treatment Courses

Radiation oncology (RO) has seen declines in Medicare reimbursement (MCR) in the past decade under the current fee-for-service model. Although studies have explored decline in reimbursement at a per-code level, to our knowledge there are no recent studies analyzing changes in MCR over time for common RO treatment courses. By analyzing changes in MCR for common treatment courses, our study had 3 objectives: (1) to provide practitioners and policymakers with estimates of recent reimbursement changes for common treatment courses; (2) to provide an estimate of how reimbursement will change in the future under the current fee-for-service model if current trends continue; and (3) to provide a baseline for treatment episodes in the event that the episode-based Radiation Oncology Alternative Payment Model is eventually implemented. Specifically, we quantified inflation- and utilization-adjusted changes in reimbursement for 16 common radiation therapy (RT) treatment courses from 2010 to 2020. Centers for Medicare & Medicaid Services Physician/Supplier Procedure Summary databases were used to obtain reimbursement for all RO procedures in 2010, 2015, and 2020 for free-standing facilities. Inflation-adjusted average reimbursement (AR) per billing instance was calculated for each Healthcare Common Procedure Coding System code using 2020 dollars. For each year, the billing frequency of each code was multiplied by the AR per code. Results were summed per RT course per year, and AR for RT courses were compared. Sixteen common RO courses for head and neck, breast, prostate, lung, and palliative RT were analyzed. AR decreased for all 16 courses from 2010 to 2020. From 2015 to 2020, the only course that increased in AR was palliative 2-dimensional 10-fraction 30 Gy, which increased by 0.4%. Courses using intensity modulated RT saw the largest AR decline from 2010 to 2020, ranging from 38% to 39%. We report significant declines in reimbursement from 2010 to 2020 for common RO courses, with the largest declines for intensity modulated RT. Policymakers should consider the significant cuts to reimbursement that have already occurred when considering future reimbursement adjustment under the current fee-for-service model or when considering mandatory adoption of a new payment system with further cuts and the negative effect of such cuts on quality and access to care.

One-step Iterative Estimation of Effective Atomic Number and Electron Density for Dual Energy CT

Dual-energy computed tomography (DECT) is a promising technology that has shown a number of clinical advantages over conventional X-ray CT, such as improved material identification, artifact suppression, etc. For proton therapy treatment planning, besides material-selective images, maps of effective atomic number (Z) and relative electron density to that of water ($\rho_e$) can also be achieved and further employed to improve stopping power ratio accuracy and reduce range uncertainty. In this work, we propose a one-step iterative estimation method, which employs multi-domain gradient $L_0$-norm minimization, for Z and $\rho_e$ maps reconstruction. The algorithm was implemented on GPU to accelerate the predictive procedure and to support potential real-time adaptive treatment planning. The performance of the proposed method is demonstrated via both phantom and patient studies.

High-Dose Once-Daily Thoracic Radiotherapy in Limited-Stage Small-Cell Lung Cancer: CALGB 30610 (Alliance)/RTOG 0538

PURPOSE

Although level 1 evidence supports 45-Gy twice-daily radiotherapy as standard for limited-stage small-cell lung cancer, most patients receive higher-dose once-daily regimens in clinical practice. Whether increasing radiotherapy dose improves outcomes remains to be prospectively demonstrated.

METHODS

This phase III trial, CALGB 30610/RTOG 0538 (ClinicalTrials.gov identifier: NCT00632853), was conducted in two stages. In the first stage, patients with limited-stage disease were randomly assigned to receive 45-Gy twice-daily, 70-Gy once-daily, or 61.2-Gy concomitant-boost radiotherapy, starting with either the first or second (of four total) chemotherapy cycles. In the second stage, allocation to the 61.2-Gy arm was discontinued following planned interim toxicity analysis, and the study continued with two remaining arms. The primary end point was overall survival (OS) in the intention-to-treat population.

RESULTS

Trial accrual opened on March 15, 2008, and closed on December 1, 2019. All patients randomly assigned to 45-Gy twice-daily (n = 313) or 70-Gy once-daily radiotherapy (n = 325) are included in this analysis. After a median follow-up of 4.7 years, OS was not improved on the once-daily arm (hazard ratio for death, 0.94; 95% CI, 0.76 to 1.17; P = .594). Median survival is 28.5 months for twice-daily treatment, and 30.1 months for once-daily treatment, with 5-year OS of 29% and 32%, respectively. Treatment was tolerable, and the frequency of severe adverse events, including esophageal and pulmonary toxicity, was similar on both arms.

CONCLUSION

Although 45-Gy twice-daily radiotherapy remains the standard of care, this study provides the most robust information available to help guide the choice of thoracic radiotherapy regimen for patients with limited-stage small-cell lung cancer.

Inter-fraction deformable image registration using unsupervised deep learning for CBCT-guided abdominal radiotherapy

CBCTs in image-guided radiotherapy provide crucial anatomy information for patient setup and plan evaluation. Longitudinal CBCT image registration could quantify the inter-fractional anatomic changes, e.g. tumor shrinkage, daily OAR variation throughout the course of treatment. The purpose of this study is to propose an unsupervised deep learning based CBCT-CBCT deformable image registration which enables quantitative anatomic variation analysis. The proposed deformable registration workflow consists of training and inference stages that share the same feed-forward path through a spatial transformation-based network (STN). The STN consists of a global generative adversarial network (GlobalGAN) and a local GAN (LocalGAN) to predict the coarse- and fine-scale motions, respectively. The network was trained by minimizing the image similarity loss and the deformable vector field (DVF) regularization loss without the supervision of ground truth DVFs. During the inference stage, patches of local DVF were predicted by the trained LocalGAN and fused to form a whole-image DVF. The local whole-image DVF was subsequently combined with the GlobalGAN generated DVF to obtain final DVF. The proposed method was evaluated using 100 fractional CBCTs from 20 abdominal cancer patients in the experiments and 105 fractional CBCTs from a cohort of 21 different abdominal cancer patients in a holdout test. Qualitatively, the registration results show good alignment between the deformed CBCT images and the target CBCT image. Quantitatively, the average target registration error (TRE) calculated on the fiducial markers and manually identified landmarks was 1.91±1.18 mm. The average mean absolute error (MAE), normalized cross correlation (NCC) between the deformed CBCT and target CBCT were 33.42±7.48 HU, 0.94±0.04, respectively. In summary, an unsupervised deep learning-based CBCT-CBCT registration method is proposed and its feasibility and performance in fractionated image-guided radiotherapy is investigated. This promising registration method could provide fast and accurate longitudinal CBCT alignment to facilitate inter-fractional anatomic changes analysis and prediction.

Definitive intensity modulated proton re-irradiation for lung cancer in the immunotherapy era

Introduction

As immunotherapy has improved distant metastasis-free survival (DMFS) in Non-Small Cell Lung Cancer (NSCLC), isolated locoregional recurrences have increased. However, management of locoregional recurrences can be challenging. We report our institutional experience with definitive intent re-irradiation using Intensity Modulated Proton Therapy (IMPT).

Method

Retrospective cohort study of recurrent or second primary NSCLC or LS-SCLC treated with IMPT. Kaplan-Meier method and log-rank test were used for time-to-event analyses.

Results

22 patients were treated from 2019 to 2021. After first course of radiation (median 60 Gy, range 45-70 Gy), 45% received adjuvant immunotherapy. IMPT re-irradiation began a median of 28.2 months (8.8-172.9 months) after initial radiotherapy. The median IMPT dose was 60 GyE (44-60 GyE). 36% received concurrent chemotherapy with IMPT and 18% received immunotherapy after IMPT. The median patient's IMPT lung mean dose was 5.3 GyE (0.9-13.9 GyE) and 5 patients had cumulative esophagus max dose >100 GyE with 1-year overall survival (OS) 68%, 1-year local control 80%, 1-year progression free survival 45%, and 1-year DMFS 60%. Higher IMPT (HR 1.4; 95% CI 1.1-1.7, p=0.01) and initial radiotherapy mean lung doses (HR 1.3; 95% CI 1.0-1.6, p=0.04) were associated with worse OS. Two patients developed Grade 3 pneumonitis or dermatitis, one patient developed Grade 2 pneumonitis, and seven patients developed Grade 1 toxicity. There were no Grade 4 or 5 toxicities.

Discussion

Definitive IMPT re-irradiation for lung cancer can prolong disease control with limited toxicity, particularly in the immunotherapy era.

Sins of omission: A meta-research study evaluating the omission of operability in published retrospective comparisons of surgery with stereotactic body radiotherapy in patients with early-stage non-small cell lung cancer

INTRODUCTION

Patients receiving stereotactic body radiotherapy (SBRT) for early-stage non-small cell lung cancer (NSCLC) are typically inoperable, in concordance with guidelines that advocate surgical resection as preferred treatment for operable patients. This differential treatment allocation complicates retrospective comparisons of surgery with SBRT by introducing the potential for confounding by operability.

METHODS

PubMed was queried for manuscripts reporting primary data from retrospective comparisons of overall survival (OS) between patients undergoing surgery versus SBRT for early-stage NSCLC. Each manuscript was categorized for two outcomes: (1) whether treatment allocation was based on a determination of patient operability, and (2) whether a direct OS comparison between operable SBRT patients and surgically treated patients was included. Associations with variables of interest were measured with statistical significance prespecified at p < 0.10.

RESULTS

From 3,072 manuscripts identified in our query, sixty-one analyses met screening criteria. Twenty-one (34 %) reported operability status influencing treatment allocation. These were more likely to be published in journals with a surgical focus (52 vs 20 %) and impact factor < 5 (81 vs 58 %), and to contain cohorts from institutional datasets (81 vs 55 %), and to have a radiation oncologist as first (43 vs 25 %) or senior (43 vs 28 %) author. Seven (11 %) manuscripts featured a direct OS comparison between SBRT and surgery.

CONCLUSION

Nearly-two-thirds of peer-reviewed retrospective studies that have compared OS between surgery and SBRT for early-stage NSCLC lack information on patient operability status, and nearly 90% lack a direct comparison between operable SBRT patients and those receiving surgery.

Automated lung tumor delineation on positron emission tomography/computed tomography via a hybrid regional network

BACKGROUND

Multimodality positron emission tomography/computed tomography (PET/CT) imaging combines the anatomical information of CT with the functional information of PET. In the diagnosis and treatment of many cancers, such as non-small cell lung cancer (NSCLC), PET/CT imaging allows more accurate delineation of tumor or involved lymph nodes for radiation planning.

PURPOSE

In this paper, we propose a hybrid regional network method of automatically segmenting lung tumors from PET/CT images.

METHODS

The hybrid regional network architecture synthesizes the functional and anatomical information from the two image modalities, whereas the mask regional convolutional neural network (R-CNN) and scoring fine-tune the regional location and quality of the output segmentation. This model consists of five major subnetworks, that is, a dual feature representation network (DFRN), a regional proposal network (RPN), a specific tumor-wise R-CNN, a mask-Net, and a score head. Given a PET/CT image as inputs, the DFRN extracts feature maps from the PET and CT images. Then, the RPN and R-CNN work together to localize lung tumors and reduce the image size and feature map size by removing irrelevant regions. The mask-Net is used to segment tumor within a volume-of-interest (VOI) with a score head evaluating the segmentation performed by the mask-Net. Finally, the segmented tumor within the VOI was mapped back to the volumetric coordinate system based on the location information derived via the RPN and R-CNN. We trained, validated, and tested the proposed neural network using 100 PET/CT images of patients with NSCLC. A fivefold cross-validation study was performed. The segmentation was evaluated with two indicators: (1) multiple metrics, including the Dice similarity coefficient, Jacard, 95th percentile Hausdorff distance, mean surface distance (MSD), residual mean square distance, and center-of-mass distance; (2) Bland-Altman analysis and volumetric Pearson correlation analysis.

RESULTS

In fivefold cross-validation, this method achieved Dice and MSD of 0.84 ± 0.15 and 1.38 ± 2.2 mm, respectively. A new PET/CT can be segmented in 1 s by this model. External validation on The Cancer Imaging Archive dataset (63 PET/CT images) indicates that the proposed model has superior performance compared to other methods.

CONCLUSION

The proposed method shows great promise to automatically delineate NSCLC tumors on PET/CT images, thereby allowing for a more streamlined clinical workflow that is faster and reduces physician effort.

Proton FLASH Radiotherapy for the Treatment of Symptomatic Bone Metastases: The FAST-01 Nonrandomized Trial

Importance

To our knowledge, there have been no clinical trials of ultra-high-dose-rate radiotherapy delivered at more than 40 Gy/sec, known as FLASH therapy, nor first-in-human use of proton FLASH.

Objectives

To assess the clinical workflow feasibility and treatment-related toxic effects of FLASH and pain relief at the treatment sites.

Design, Setting, and Participants

In the FAST-01 nonrandomized trial, participants treated at Cincinnati Children's/UC Health Proton Therapy Center underwent palliative FLASH radiotherapy to extremity bone metastases. Patients 18 years and older with 1 to 3 painful extremity bone metastases and life expectancies of 2 months or more were eligible. Patients were excluded if they had foot, hand, and wrist metastases; metastases locally treated in the 2 weeks prior; metal implants in the treatment field; known enhanced tissue radiosensitivity; and implanted devices at risk of malfunction with radiotherapy. One of 11 patients who consented was excluded based on eligibility. The end points were evaluated at 3 months posttreatment, and patients were followed up through death or loss to follow-up for toxic effects and pain assessments. Of the 10 included patients, 2 died after the 2-month follow-up but before the 3-month follow-up; 8 participants completed the 3-month evaluation. Data were collected from November 3, 2020, to January 28, 2022, and analyzed from January 28, 2022, to September 1, 2022.

Interventions

Bone metastases were treated on a FLASH-enabled (≥40 Gy/sec) proton radiotherapy system using a single-transmission proton beam. This is consistent with standard of care using the same prescription (8 Gy in a single fraction) but on a conventional-dose-rate (approximately 0.03 Gy/sec) photon radiotherapy system.

Main Outcome and Measures

Main outcomes included patient time on the treatment couch, device-related treatment delays, adverse events related to FLASH, patient-reported pain scores, and analgesic use.

Results

A total of 10 patients (age range, 27-81 years [median age, 63 years]; 5 [50%] male) underwent FLASH radiotherapy at 12 metastatic sites. There were no FLASH-related technical issues or delays. The average (range) time on the treatment couch was 18.9 (11-33) minutes per patient and 15.8 (11-22) minutes per treatment site. Median (range) follow-up was 4.8 (2.3-13.0) months. Adverse events were mild and consistent with conventional radiotherapy. Transient pain flares occurred in 4 of the 12 treated sites (33%). In 8 of the 12 sites (67%) patients reported pain relief, and in 6 of the 12 sites (50%) patients reported a complete response (no pain).

Conclusions and Relevance

In this nonrandomized trial, clinical workflow metrics, treatment efficacy, and safety data demonstrated that ultra-high-dose-rate proton FLASH radiotherapy was clinically feasible. The treatment efficacy and the profile of adverse events were comparable with those of standard-of-care radiotherapy. These findings support the further exploration of FLASH radiotherapy in patients with cancer.

Trial Registration

ClinicalTrials.gov Identifier: NCT04592887.

Enhanced cardiac substructure sparing through knowledge-based treatment planning for non-small cell lung cancer radiotherapy

Radiotherapy (RT) doses to cardiac substructures from the definitive treatment of locally advanced non-small cell lung cancers (NSCLC) have been linked to post-RT cardiac toxicities. With modern treatment delivery techniques, it is possible to focus radiation doses to the planning target volume while reducing cardiac substructure doses. However, it is often challenging to design such treatment plans due to complex tradeoffs involving numerous cardiac substructures. Here, we built a cardiac-substructure-based knowledge-based planning (CS-KBP) model and retrospectively evaluated its performance against a cardiac-based KBP (C-KBP) model and manually optimized patient treatment plans. CS-KBP/C-KBP models were built with 27 previously-treated plans that preferentially spare the heart. While the C-KBP training plans were created with whole heart structures, the CS-KBP model training plans each have 15 cardiac substructures (coronary arteries, valves, great vessels, and chambers of the heart). CS-KBP training plans reflect cardiac-substructure sparing preferences. We evaluated both models on 28 additional patients. Three sets of treatment plans were compared: (1) manually optimized, (2) C-KBP model-generated, and (3) CS-KBP model-generated. Plans were normalized to receive the prescribed dose to at least 95% of the PTV. A two-tailed paired-sample t-test was performed for clinically relevant dose-volume metrics to evaluate the performance of the CS-KBP model against the C-KBP model and clinical plans, respectively. Overall results show significantly improved cardiac substructure sparing by CS-KBP in comparison to C-KBP and the clinical plans. For instance, the average left anterior descending artery volume receiving 15 Gy (V15 Gy) was significantly lower (p &lt; 0.01) for CS-KBP (0.69 ± 1.57 cc) compared to the clinical plans (1.23 ± 1.76 cc) and C-KBP plans (1.05 ± 1.68 cc). In conclusion, the CS-KBP model significantly improved cardiac-substructure sparing without exceeding the tolerances of other OARs or compromising PTV coverage.

Deformable CT image registration via a dual feasible neural network

PURPOSE

A quality assurance (QA) CT scans are usually acquired during cancer radiotherapy to assess for any anatomical changes, which may cause an unacceptable dose deviation and therefore warrant a replan. Accurate and rapid deformable image registration (DIR) is needed to support contour propagation from the planning CT (pCT) to the QA CT to facilitate dose volume histogram (DVH) review. Further, the generated deformation maps are used to track the anatomical variations throughout the treatment course and calculate the corresponding accumulated dose from one or more treatment plans.

METHODS

In this study, we aim to develop a deep learning (DL)-based method for automatic deformable registration to align the pCT and the QA CT. Our proposed method, named dual-feasible framework, was implemented by a mutual network that functions as both a forward module and a backward module. The mutual network was trained to predict two deformation vector fields (DVFs) simultaneously, which were then used to register the pCT and QA CT in both directions. A novel dual feasible loss was proposed to train the mutual network. The dual-feasible framework was able to provide additional DVF regularization during network training, which preserves the topology and reduces folding problems. We conducted experiments on 65 head-and-neck cancer patients (228 CTs in total), each with 1 pCT and 2-6 QA CTs. For evaluations, we calculated the mean absolute error (MAE), peak-signal-to-noise ratio (PSNR), structural similarity index (SSIM), target registration error (TRE) between the deformed and target images and the Jacobian determinant of the predicted DVFs.

RESULTS

Within the body contour, the mean MAE, PSNR, SSIM, and TRE are 122.7 HU, 21.8 dB, 0.62 and 4.1 mm before registration and are 40.6 HU, 30.8 dB, 0.94, and 2.0 mm after registration using the proposed method. These results demonstrate the feasibility and efficacy of our proposed method for pCT and QA CT DIR.

CONCLUSION

In summary, we proposed a DL-based method for automatic DIR to match the pCT to the QA CT. Such DIR method would not only benefit current workflow of evaluating DVHs on QA CTs but may also facilitate studies of treatment response assessment and radiomics that depend heavily on the accurate localization of tissues across longitudinal images.

Validation of a deep learning-based material estimation model for Monte Carlo dose calculation in proton therapy

Objective. Computed tomography (CT) to material property conversion dominates proton range uncertainty, impacting the quality of proton treatment planning. Physics-based and machine learning-based methods have been investigated to leverage dual-energy CT (DECT) to predict proton ranges. Recent development includes physics-informed deep learning (DL) for material property inference. This paper aims to develop a framework to validate Monte Carlo dose calculation (MCDC) using CT-based material characterization models.Approach.The proposed framework includes two experiments to validatein vivodose and water equivalent thickness (WET) distributions using anthropomorphic and porcine phantoms. Phantoms were irradiated using anteroposterior proton beams, and the exit doses and residual ranges were measured by MatriXX PT and a multi-layer strip ionization chamber. Two pre-trained conventional and physics-informed residual networks (RN/PRN) were used for mass density inference from DECT. Additional two heuristic material conversion models using single-energy CT (SECT) and DECT were implemented for comparisons. The gamma index was used for dose comparisons with criteria of 3%/3 mm (10% dose threshold).Main results. The phantom study showed that MCDC with PRN achieved mean gamma passing rates of 95.9% and 97.8% for the anthropomorphic and porcine phantoms. The rates were 86.0% and 79.7% for MCDC with the empirical DECT model. WET analyses indicated that the mean WET variations between measurement and simulation were -1.66 mm, -2.48 mm, and -0.06 mm for MCDC using a Hounsfield look-up table with SECT and empirical and PRN models with DECT. Validation experiments indicated that MCDC with PRN achieved consistent dose and WET distributions with measurement.Significance. The proposed framework can be used to identify the optimal CT-based material characterization model for MCDC to improve proton range uncertainty. The framework can systematically verify the accuracy of proton treatment planning, and it can potentially be implemented in the treatment room to be instrumental in online adaptive treatment planning.

Male pelvic multi-organ segmentation using token-based transformer Vnet

Objective. This work aims to develop an automated segmentation method for the prostate and its surrounding organs-at-risk in pelvic computed tomography to facilitate prostate radiation treatment planning.Approach. In this work, we propose a novel deep learning algorithm combining a U-shaped convolutional neural network (CNN) and vision transformer (VIT) for multi-organ (i.e. bladder, prostate, rectum, left and right femoral heads) segmentation in male pelvic CT images. The U-shaped model consists of three components: a CNN-based encoder for local feature extraction, a token-based VIT for capturing global dependencies from the CNN features, and a CNN-based decoder for predicting the segmentation outcome from the VIT's output. The novelty of our network is a token-based multi-head self-attention mechanism used in the transformer, which encourages long-range dependencies and forwards informative high-resolution feature maps from the encoder to the decoder. In addition, a knowledge distillation strategy is deployed to further enhance the learning capability of the proposed network.Main results. We evaluated the network using: (1) a dataset collected from 94 patients with prostate cancer; (2) and a public dataset CT-ORG. A quantitative evaluation of the proposed network's performance was performed on each organ based on (1) volume similarity between the segmented contours and ground truth using Dice score, segmentation sensitivity, and precision, (2) surface similarity evaluated by Hausdorff distance (HD), mean surface distance (MSD) and residual mean square distance (RMS), (3) and percentage volume difference (PVD). The performance was then compared against other state-of-art methods. Average volume similarity measures obtained by the network overall organs were Dice score = 0.91, sensitivity = 0.90, precision = 0.92, average surface similarities were HD = 3.78 mm, MSD = 1.24 mm, RMS = 2.03 mm; average percentage volume difference was PVD = 9.9% on the first dataset. The network also obtained Dice score = 0.93, sensitivity = 0.93, precision = 0.93, average surface similarities were HD = 5.82 mm, MSD = 1.16 mm, RMS = 1.24 mm; average percentage volume difference was PVD = 6.6% on the CT-ORG dataset.Significance. In summary, we propose a token-based transformer network with knowledge distillation for multi-organ segmentation using CT images. This method provides accurate and reliable segmentation results for each organ using CT imaging, facilitating the prostate radiation clinical workflow.

A component method to delineate surgical spine implants for proton Monte Carlo dose calculation

PURPOSE

Metallic implants have been correlated to local control failure for spinal sarcoma and chordoma patients due to the uncertainty of implant delineation from computed tomography (CT). Such uncertainty can compromise the proton Monte Carlo dose calculation (MCDC) accuracy. A component method is proposed to determine the dimension and volume of the implants from CT images.

METHODS

The proposed component method leverages the knowledge of surgical implants from medical supply vendors to predefine accurate contours for each implant component, including tulips, screw bodies, lockers, and rods. A retrospective patient study was conducted to demonstrate the feasibility of the method. The reference implant materials and samples were collected from patient medical records and vendors, Medtronic and NuVasive. Additional CT images with extensive features, such as extended Hounsfield units and various reconstruction diameters, were used to quantify the uncertainty of implant contours.

RESULTS

For in vivo patient implant estimation, the reference and the component method differences were 0.35, 0.17, and 0.04 cm³ for tulips, screw bodies, and rods, respectively. The discrepancies by a conventional threshold method were 5.46, 0.76, and 0.05 cm³ , respectively. The mischaracterization of implant materials and dimensions can underdose the clinical target volume coverage by 20 cm³ for a patient with eight lumbar implants. The tulip dominates the dosimetry uncertainty as it can be made from titanium or cobalt-chromium alloys by different vendors.

CONCLUSIONS

A component method was developed and demonstrated using phantom and patient studies with implants. The proposed method provides more accurate implant characterization for proton MCDC and can potentially enhance the treatment quality for proton therapy. The current proof-of-concept study is limited to the implant characterization for lumbar spine. Future investigations could be extended to cervical spine and dental implants for head-and-neck patients where tight margins are required to spare organs at risk.

A potential revolution in cancer treatment: A topical review of FLASH radiotherapy

FLASH radiotherapy (RT) is a novel technique in which the ultrahigh dose rate (UHDR) (≥40 Gy/s) is delivered to the entire treatment volume. Recent outcomes of in vivo studies show that the UHDR RT has the potential to spare normal tissue without sacrificing tumor control. There is a growing interest in the application of FLASH RT, and the ultrahigh dose irradiation delivery has been achieved by a few experimental and modified linear accelerators. The underlying mechanism of FLASH effect is yet to be fully understood, but the oxygen depletion in normal tissue providing extra protection during FLASH irradiation is a hypothesis that attracts most attention currently. Monte Carlo simulation is playing an important role in FLASH, enabling the understanding of its dosimetry calculations and hardware design. More advanced Monte Carlo simulation tools are under development to fulfill the challenge of reproducing the radiolysis and radiobiology processes in FLASH irradiation. FLASH RT may become one of standard treatment modalities for tumor treatment in the future. This paper presents the history and status of FLASH RT studies with a focus on FLASH irradiation delivery modalities, underlying mechanism of FLASH effect, in vivo and vitro experiments, and simulation studies. Existing challenges and prospects of this novel technique are discussed in this manuscript.

Impact of Radiation Dose to the Immune Cells in Unresectable or Stage III Non-Small Cell Lung Cancer in the Durvalumab Era

BACKGROUND /PURPOSE

Higher estimated radiation doses to immune cells (EDIC) have correlated with worse overall survival (OS) in patients with locally-advanced non-small cell lung cancer (NSCLC) prior to the PACIFIC trial, which established consolidative durvalumab as standard-of-care. Here, we examine the prognostic impact of EDIC in the durvalumab era.

MATERIALS/METHODS

This single-institution, multi-center study included patients with unresectable stage II/III NSCLC treated with chemoradiation followed by durvalumab. Associations between EDIC [analyzed continuously and categorically (≤6 Gy vs. >6 Gy)] and OS, progression-free survival (PFS), and locoregional control (LRC) were evaluated by Kaplan-Meier and Cox proportional methods.

RESULTS

100 patients were included with median follow-up of 23.7 months. The EDIC >6 Gy group had a significantly greater percentage of stage IIIB/IIIC disease (76.0% vs. 32.6%; p<0.001) and larger tumor volumes (170cc vs. 42cc; p<0.001). There were no differences in early durvalumab discontinuation from toxicity (24.1% vs. 15.2%; p=0.27). Median OS was shorter among the EDIC >6 Gy group (29.6 months vs. not reached; p<0.001). On multivariate analysis, EDIC >6 Gy correlated with worse OS (HR: 4.15, 95%CI: 1.52-11.33; p=0.006), PFS (HR: 3.79; 95%CI: 1.80-8.0; p<0.001), and LRC (HR: 2.66, 95%CI: 1.15-6.18; p=0.023). Analyzed as a continuous variable, higher EDIC was associated with worse OS (HR: 1.34; 95%CI: 1.16-1.57; p<0.001), PFS (HR: 1.52; 95%CI: 1.29-1.79; p<0.001), and LRC (HR: 1.34, 95%CI: 1.13-1.60; p=0.007).

CONCLUSIONS

In the immunotherapy era, EDIC is an independent predictor of OS and disease control in locally advanced NSCLC, warranting investigation into techniques to reduce dose to the immune compartment.

Synchronized high-speed scintillation imaging of proton beams, generated by a gantry-mounted synchrocyclotron, on a pulse-by-pulse basis

BACKGROUND

With the emergence of more complex and novel proton delivery techniques, there is a need for quality assurance (QA) tools with high spatiotemporal resolution to conveniently measure the spatial and temporal properties of the beam. In this context, scintillation-based dosimeters, if synchronized with the radiation beam and corrected for ionization quenching, are appealing.

PURPOSE

To develop a synchronized high-speed scintillation imaging system for characterization and verification of the proton therapy beams on a pulse-by-pulse basis.

MATERIALS AND METHODS

A 30 cm × 30 cm × 5 cm block of BC-408 plastic scintillator placed in a light-tight housing was irradiated by proton beams generated by a Mevion S250^TM proton therapy synchrocyclotron. A high-speed camera system, placed perpendicular to the beam direction and facing the scintillator, was synchronized to the accelerator's pulses to capture images. Opening and closing of the camera's shutter was controlled by setting a proper time delay and exposure time, respectively. The scintillation signal was recorded as a set of two-dimensional (2D) images. Empirical correction factors were applied to the images to correct for the non-uniformity of the pixel sensitivity and quenching of the scintillator. Proton range and modulation were obtained from the corrected images.

RESULTS

The camera system was able to capture all data on a pulse-by-pulse basis at a rate of ∼504 frames per second. The applied empirical correction method for ionization quenching was effective and the corrected composite image provided a 2D map of dose distribution. The measured range (depth of distal 90%) through scintillation imaging agreed within 1.2 mm with that obtained from ionization chamber measurement.

CONCLUSION

A high-speed camera system capable of capturing scintillation signals from individual proton pulses was developed. The scintillation imaging system is promising for rapid proton beam characterization and verification. This article is protected by copyright. All rights reserved.

Safety results of NRG-LU004: Phase I trial of accelerated or conventionally fractionated radiotherapy combined with durvalumab in PD-L1–high locally advanced non-small cell lung cancer

8513

Background: In advanced non-small cell lung cancer (NSCLC), high Programmed-Death-1 Ligand (PD-L1) (>50%) expression demonstrate superior response and survival with immune checkpoint inhibitors compared to chemotherapy. We hypothesize that it is safe and feasible to substitute durvalumab instead of chemotherapy concurrently with radiotherapy (RT) in patients with Locally Advanced-NSCLC (LA-NSCLC) and high PD-L1. Methods: NRG-LU004 (NCT03801902) is a Phase I study for patients with stage II-III unresectable or inoperable, LA-NSCLC with PD-L1> 50% (Dako 22C3 or Ventana SP263) expression. There were safety and expansion phases with a primary endpoint of safety. Patients started with 1500 mg durvalumab Q4 weeks and thoracic RT within 2 weeks from 1st infusion. Durvalumab continued once a month up to 1 year. In the safety cohort, 6 patients in cohort 1 were treated with accelerated fractionated RT (ACRT) to 60 Gy in 15 fractions, followed by a required safety hold for 90 days. During cohort 1 safety hold, cohort 2 patients were treated with conventional RT 60 Gy in 30 fractions (CONV) followed by a 60-day safety hold. A cohort advanced to the expansion phase to enroll 6 more patients if safety criteria (0-1 patients with a dose limiting toxicity [DLT]) were met. If both cohorts were deemed safe, patients would be randomized 1:1 to ACRT or CONV with safety defined as < 4 of 12 evaluable patients per arm experiencing a DLT. Feasibility was defined as at least 80% of patients in each arm receiving at least 80% of the planned dose of durvalumab during the first 8 weeks. Results: 24 evaluable patients enrolled between January 2019 and June 2021. No DLTs were reported in cohort 1, and 1 (unrelated bronchopulmonary hemorrhage leading to discontinuation of durvalumab) in cohort 2. Both safety cohorts advanced to the expansion phase. All but one patient (CONV) received RT per protocol/with an acceptable variation. At the time of analysis, 24% had received all 13 cycles of durvalumab. For the ACRT cohort, there were 4 grade 3, 1 grade 4 (lymphopenia), and 1 grade 5 AE (lung infection, assessed as unrelated to therapy). For CONV, there were 8 grade 3, 0 grade 4, and 1 grade 5 AE (respiratory failure, unrelated to therapy). For feasibility, 10 of 12 (85%) patients in the ACRT cohort received the second dose of durvalumab (2 not received due to shingles and unrelated death), while 9 of 12 (75%) of the CONV cohort received the second dose (reasons for not receiving: viral hepatitis, bronchopulmonary hemorrhage, and respiratory failure, all assessed as unrelated to therapy). Conclusions: Chemotherapy-free thoracic RT approaches (ACRT or CONV RT) are safe, when given with concurrent durvalumab in patients with PD-L1 high LA-NSCLC. A trial to compare immunoradiotherapy and consolidation durvalumab to standard chemoradiation and consolidation durvalumab is planned. Clinical trial information: NCT03801902.

Dual-energy CT based mass density and relative stopping power estimation for proton therapy using physics-informed deep learning

Proton therapy requires accurate dose calculation for treatment planning to ensure the conformal doses are precisely delivered to the targets. The conversion of CT numbers to material properties is a significant source of uncertainty for dose calculation. The aim of this study is to develop a physics-informed deep learning (PIDL) framework to derive accurate mass density and relative stopping power maps from dual-energy computed tomography (DECT) images. The PIDL framework allows deep learning (DL) models to be trained with a physics loss function, which includes a physics model to constrain DL models. Five DL models were implemented including a fully connected neural network (FCNN), dual-FCNN (DFCNN), and three variants of residual networks (ResNet): ResNet-v1 (RN-v1), ResNet-v2 (RN-v2), and dual-ResNet-v2 (DRN-v2). An artificial neural network (ANN) and the five DL models trained with and without physics loss were explored to evaluate the PIDL framework. Two empirical DECT models were implemented to compare with the PIDL method. DL training data were from CIRS electron density phantom 062M (Computerized Imaging Reference Systems, Inc., Norfolk, VA). The performance of DL models was tested by CIRS adult male, adult female, and 5-year-old child anthropomorphic phantoms. For density map inference, the physics-informed RN-v2 was 3.3%, 2.9% and 1.9% more accurate than ANN for the adult male, adult female, and child phantoms. The physics-informed DRN-v2 was 0.7%, 0.6%, and 0.8% more accurate than DRN-v2 without physics training for the three phantoms, respectfully. The results indicated that physics-informed training could reduce uncertainty when ANN/DL models without physics training were insufficient to capture data structures or derived significant errors. DL models could also achieve better image noise control compared to the empirical DECT parametric mapping methods. The proposed PIDL framework can potentially improve proton range uncertainty by offering accurate material properties conversion from DECT.

Mutual enhancing learning-based automatic segmentation of CT cardiac substructure

Objective. Current segmentation practice for thoracic cancer RT considers the whole heart as a single organ despite increased risks of cardiac toxicities from irradiation of specific cardiac substructures. Segmenting up to 15 different cardiac substructures can be a very time-intensive process, especially due to their different volume sizes and anatomical variations amongst different patients. In this work, a new deep learning (DL)-based mutual enhancing strategy is introduced for accurate and automatic segmentation, especially of smaller substructures such as coronary arteries. Approach. Our proposed method consists of three subnetworks: retina U-net, classification module, and segmentation module. Retina U-net is used as a backbone network architecture that aims to learn deep features from the whole heart. Whole heart feature maps from retina U-net are then transferred to four different sets of classification modules to generate classification localization maps of coronary arteries, great vessels, chambers of the heart, and valves of the heart. Each classification module is in sync with its corresponding subsequent segmentation module in a bootstrapping manner, allowing them to share their encoding paths to generate a mutual enhancing strategy. We evaluated our method on three different datasets: institutional CT datasets (55 subjects) 2) publicly available Multi-Modality Whole Heart Segmentation (MM-WHS) challenge datasets (120 subjects), and Automated Cardiac Diagnosis Challenge (ACDC) datasets (100 subjects). For institutional datasets, we performed five-fold cross-validation on training data (45 subjects) and performed inference on separate hold-out data (10 subjects). For each subject, 15 cardiac substructures were manually contoured by a resident physician and evaluated by an attending radiation oncologist. For the MM-WHS dataset, we trained the network on 100 datasets and performed an inference on a separate hold-out dataset with 20 subjects, each with 7 cardiac substructures. For ACDC datasets, we performed five-fold cross-validation on 100 datasets, each with 3 cardiac substructures. We compared the proposed method against four different network architectures: 3D U-net, mask R-CNN, mask scoring R-CNN, and proposed network without classification module. Segmentation accuracies were statistically compared through dice similarity coefficient, Jaccard, 95% Hausdorff distance, mean surface distance, root mean square distance, center of mass distance, and volume difference. Main results. The proposed method generated cardiac substructure segmentations with significantly higher accuracy (P < 0.05) for small substructures, especially for coronary arteries such as left anterior descending artery (CA-LADA) and right coronary artery (CA-RCA) in comparison to four competing methods. For large substructures (i.e. chambers of the heart), our method yielded comparable results to mask scoring R-CNN method, resulting in significantly (P < 0.05) improved segmentation accuracy in comparison to 3D U-net and mask R-CNN. Significance. A new DL-based mutual enhancing strategy was introduced for automatic segmentation of cardiac substructures. Overall results of this work demonstrate the ability of the proposed method to improve segmentation accuracies of smaller substructures such as coronary arteries without largely compromising the segmentation accuracies of larger substructures. Fast and accurate segmentations of up to 15 substructures can possibly be used as a tool to rapidly generate substructure segmentations followed by physicians’ reviews to improve clinical workflow.

Magnetic resonance imaging contrast enhancement synthesis using cascade networks with local supervision

PURPOSE

Gadolinium-based contrast agents (GBCAs) are widely administrated in MR imaging for diagnostic studies and treatment planning. Although GBCAs are generally thought to be safe, various health and environmental concerns have been raised recently about their use in MR imaging. The purpose of this work is to derive synthetic contrast enhance MR images from unenhanced counterpart images, thereby eliminating the need for GBCAs, using a cascade deep learning workflow that incorporates contour information into the network.

METHODS AND MATERIALS

The proposed workflow consists of two sequential networks: (1) a retina U-Net, which is first trained to derive semantic features from the non-contrast MR images in representing the tumor regions; and (2) a synthesis module, which is trained after the retina U-Net to take the concatenation of the semantic feature maps and non-contrast MR image as input and to generate the synthetic contrast enhanced MR images. After network training, only the non-contrast enhanced MR images are required for the input in the proposed workflow. The MR images of 369 patients from the multimodal brain tumor segmentation challenge 2020 (BraTS2020) dataset were used in this study to evaluate the proposed workflow for synthesizing contrast enhanced MR images (200 patients for five-fold cross-validation and 169 patients for hold-out test). Quantitative evaluations were conducted by calculating the normalized mean absolute error (NMAE), structural similarity index measurement (SSIM), and Pearson correlation coefficient (PCC). The original contrast enhanced MR images were considered as the ground truth in this analysis.

RESULTS

The proposed cascade deep learning workflow synthesized contrast enhanced MR images that are not visually differentiable from the ground truth with and without supervision of the tumor contours during the network training. Difference images and profiles of the synthetic contrast enhanced MR images revealed that intensity differences could be observed in the tumor region if the contour information was not incorporated in network training. Among the hold-out test patients, mean values and standard deviations of the NMAE, SSIM, and PCC were 0.063±0.022, 0.991±0.007 and 0.995±0.006, respectively, for the whole brain; and were 0.050±0.025, 0.993±0.008 and 0.999±0.003, respectively, for the tumor contour regions. Quantitative evaluations with five-fold cross-validation and hold-out test showed that the calculated metrics can be significantly enhanced (p-values ≤ 0.002) with the tumor contour supervision in network training.

CONCLUSION

The proposed workflow was able to generate synthetic contrast enhanced MR images that closely resemble the ground truth images from non-contrast enhanced MR images when the network training included tumor contours. These results suggest that it may be possible to minimize the use of GBCAs in cranial MR imaging studies.

Decreases in Radiation Oncology Medicare Reimbursement over time: Analysis by Billing Code

PURPOSE

Radiation oncology (RO) has seen declines in Medicare reimbursement (MCR). However, there are no recent studies analyzing the contributions of specific billing codes to overall RO reimbursement. We compared total MCR for specific Healthcare Common Procedure Coding System (HCPCS) codes in 2019 with MCR for those codes in 2010 and 2015, corrected for inflation, to see how the same basket of RO services in 2019 would have been reimbursed in 2010 and 2015 (adjusted MCR).

METHODS AND MATERIALS

The Centers for Medicare & Medicaid Services Physician/Supplier Procedure Summary database was used to obtain MCR data for RO HCPCS codes in 2010, 2015, and 2019. For each code, the total allowed charge was divided by the number of submitted claims to calculate the average MCR per claim in 2010, 2015, and 2019. The 2019 billing frequency for each code was then multiplied by the inflation-adjusted average MCR for those codes in 2010 and 2015 to determine what the MCR would have been in 2010 and 2015 using 2019 dollars and utilization rates. Results were compared with actual 2019 MCR to calculate the projected difference.

RESULTS

Total inflation-adjusted RO MCR was $2281 million (M), $1991 M, and $1848 M in 2010, 2015, and 2019 respectively. This represents a cut of $433 M (19%) and $143 M (7%) from 2010 and 2015, respectively, to 2019. After utilization adjustment, total reimbursement was $2534 M, $2034 M, and $1848 M for 2010, 2015, and 2019, respectively, representing a cut of $686 M (27%) and $186 M (9%) from 2010 and 2015, respectively, to 2019. Intensity modulated radiation therapy (IMRT) treatment delivery and planning accounted for $917 M (36%), $670 M (33%), and $573 M (31%) of the adjusted MCR in 2010, 2015, and 2019, respectively.

CONCLUSIONS

Medicare reimbursement decreased substantially from 2010 to 2019. A decline in IMRT treatment reimbursement was the primary driver of MCR decline. When considering further cuts, policymakers should consider these trends and their consequences for health care quality and access.

Simultaneous dose and dose rate optimization (SDDRO) of the FLASH effect for pencil-beam-scanning proton therapy

PURPOSE

Compared to CONV-RT (with conventional dose rate), FLASH-RT (with ultra-high dose rate) can provide biological dose sparing for organs-at-risk (OARs) via the so-called FLASH effect, in addition to physical dose sparing. However, the FLASH effect only occurs, when both dose and dose rate meet certain minimum thresholds. This work will develop a simultaneous dose and dose rate optimization (SDDRO) method accounting for both FLASH dose and dose rate constraints during treatment planning for pencil-beam-scanning proton therapy.

METHODS

SDDRO optimizes the FLASH effect (specific to FLASH-RT) as well as the dose distribution (similar to CONV-RT). The nonlinear dose rate constraint is linearized, and the reformulated optimization problem is efficiently solved via iterative convex relaxation powered by alternating direction method of multipliers. To resolve and quantify the generic tradeoff of FLASH-RT between FLASH and dose optimization, we propose the use of FLASH effective dose based on dose modifying factor (DMF) owing to the FLASH effect.

RESULTS

FLASH-RT via transmission beams (TB) (IMPT-TB or SDDRO) and CONV-RT via Bragg peaks (BP) (IMPT-BP) were evaluated for clinical prostate, lung, head-and-neck (HN), and brain cases. Despite the use of TB, which is generally suboptimal to BP for normal tissue sparing, FLASH-RT via SDDRO considerably reduced FLASH effective dose for high-dose OAR adjacent to the target. For example, in the lung SBRT case, the max esophageal dose constraint 27 Gy was only met by SDDRO (24.8 Gy), compared to IMPT-BP (35.3 Gy) or IMPT-TB (36.6 Gy); in the brain SRS case, the brain constraint V12Gy≤15cc was also only met by SDDRO (13.7cc), compared to IMPT-BP (43.9cc) or IMPT-TB (18.4cc). In addition, SDDRO substantially improved the FLASH coverage from IMPT-TB, e.g., an increase from 37.2% to 67.1% for lung, from 39.1% to 58.3% for prostate, from 65.4% to 82.1% for HN, from 50.8% to 73.3% for the brain.

CONCLUSIONS

Both FLASH dose and dose rate constraints are incorporated into SDDRO for FLASH-RT that jointly optimizes the FLASH effect and physical dose distribution. FLASH effective dose via FLASH DMF is introduced to reconcile the tradeoff between physical dose sparing and FLASH sparing, and quantify the net effective gain from CONV-RT to FLASH-RT.

Surgical Outcomes for Early Stage Non-small Cell Lung Cancer at Facilities With Stereotactic Body Radiation Therapy Programs

BACKGROUND

Patients undergoing surgery for early stage non-small cell lung cancer (NSCLC) may be at high risk for postoperative mortality. Access to stereotactic body radiation therapy (SBRT) may facilitate more appropriate patient selection for surgery.

RESEARCH QUESTION

Is postoperative mortality associated with early stage NSCLC lower at facilities with higher use of SBRT?

STUDY DESIGN AND METHODS

Patients with early stage NSCLC reported to the National Cancer Database between 2004 and 2015 were included. Use of SBRT was defined by each facility's SBRT experience (in years) and SBRT to surgery volume ratios. Multivariate logistic regression was used to test for the associations between SBRT use and postoperative mortality.

RESULTS

The study cohort consisted of 202,542 patients who underwent surgical resection of cT1-T2N0M0 NSCLC tumors. The 90-day postoperative mortality rate declined during the study period from 4.6% to 2.6% (P < .001), the proportion of facilities that used SBRT increased from 4.6% to 77.5% (P < .001), and the proportion of patients treated with SBRT increased from 0.7% to 15.4% (P < .001). On multivariate analysis, lower 90-day postoperative mortality rates were observed at facilities with > 6 years of SBRT experience (OR, 0.84; 95% CI, 0.76-0.94; P = .003) and SBRT to surgery volume ratios of more than 17% (OR, 0.85; 95% CI, 0.79-0.92; P < .001). Ninety-day mortality also was associated with surgical volume, region, year, age, sex, and race, among other covariates. Interaction testing between these covariates showed negative results.

INTERPRETATION

Patients who underwent resection for early stage NSCLC at facilities with higher SBRT use showed lower rates of postoperative mortality. These findings suggest that the availability and use of SBRT may improve the selection of patients for surgery who are predicted to be at high risk of postoperative mortality.

Association of Operability with Post-Treatment Mortality in Early-Stage Non-Small Cell Lung Cancer

BACKGROUND

Operability is both a crucial determinant in treatment selection and a potential confounder in analyses comparing surgery with non-surgical approaches such as stereotactic body radiotherapy (SBRT). We aimed to assess the association between operability status and intervention with post-treatment mortality in early-stage non-small cell lung cancer (NSCLC).

PATIENTS AND METHODS

We defined four groups of patients with cT1-T2N0M0 NSCLC diagnosed 2010 to 2014 from the National Cancer Database: SBRT patients deemed operable vs. inoperable and surgery patients receiving open vs. minimally-invasive approaches. Mortality rates at 30, 60, and 90 days post-treatment were calculated and compared.

RESULTS

We abstracted 80,108 patients, 0.8% undergoing SBRT and operable, 13.2% undergoing SBRT and inoperable, 52.4% undergoing open surgery, and 33.7% undergoing minimally-invasive surgery. Mortality rates were highest among open surgery patients and lowest among operable SBRT patients (2.0% vs. 0.2% at 30 days and 3.7% vs. 0.7% at 90 days), with intermediate results in the other two groups. These findings persisted on multivariate Cox regression: compared to patients undergoing minimally-invasive surgery, mortality risk was highest among open surgery patients (30 days HR 1.32, 95%CI 1.16-1.51; 90 days HR 1.36, 95%CI 1.24-1.50; both P < .001) and lowest among operable SBRT patients (30 days HR 0.09, 95%CI 0.01-0.64; 90 days HR 0.15, 95%CI 0.05-0.46; both P ≤ .016). These associations were maintained in a propensity score-matched subset.

CONCLUSION

Operable patients undergoing SBRT experience minimal post-treatment mortality compared to their inoperable counterparts. These findings illustrate the potential for confounding by operability to bias results in cohort studies that compare surgical vs. non-surgical approaches in early-stage NSCLC.

Synthesizing high-resolution magnetic resonance imaging using parallel cycle-consistent generative adversarial networks for fast magnetic resonance imaging

PURPOSE

The common practice in acquiring the magnetic resonance (MR) images is to obtain two-dimensional (2D) slices at coarse locations while keeping the high in-plane resolution in order to ensure enough body coverage while shortening the MR scan time. The aim of this study is to propose a novel method to generate HR MR images from low-resolution MR images along the longitudinal direction. In order to address the difficulty of collecting paired low- and high-resolution MR images in clinical settings and to gain the advantage of parallel cycle consistent generative adversarial networks (CycleGANs) in synthesizing realistic medical images, we developed a parallel CycleGANs based method using a self-supervised strategy.

METHODS AND MATERIALS

The proposed workflow consists of two parallely trained CycleGANs to independently predict the HR MR images in the two planes along the directions that are orthogonal to the longitudinal MR scan direction. Then, the final synthetic HR MR images are generated by fusing the two predicted images. MR images, including T1-weighted (T1), contrast enhanced T1-weighted (T1CE), T2-weighted (T2), and T2 Fluid Attenuated Inversion Recovery (FLAIR), of the multimodal brain tumor segmentation challenge 2020 (BraTS2020) dataset were processed to evaluate the proposed workflow along the cranial-caudal (CC), lateral, and anterior-posterior directions. Institutional collected MR images were also processed for evaluation of the proposed method. The performance of the proposed method was investigated via both qualitative and quantitative evaluations. Metrics of normalized mean absolute error (NMAE), peak signal-to-noise ratio (PSNR), edge keeping index (EKI), structural similarity index measurement (SSIM), information fidelity criterion (IFC), and visual information fidelity in pixel domain (VIFP) were calculated.

RESULTS

It is shown that the proposed method can generate HR MR images visually indistinguishable from the ground truth in the investigations on the BraTS2020 dataset. In addition, the intensity profiles, difference images and SSIM maps can also confirm the feasibility of the proposed method for synthesizing HR MR images. Quantitative evaluations on the BraTS2020 dataset shows that the calculated metrics of synthetic HR MR images can all be enhanced for the T1, T1CE, T2, and FLAIR images. The enhancements in the numerical metrics over the low-resolution and bi-cubic interpolated MR images, as well as those genearted with a comparative deep learning method, are statistically significant. Qualitative evaluation of the synthetic HR MR images of the clinical collected dataset could also confirm the feasibility of the proposed method.

CONCLUSIONS

The proposed method is feasible to synthesize HR MR images using self-supervised parallel CycleGANs, which can be expected to shorten MR acquisition time in clinical practices.

Deep learning-based motion tracking using ultrasound images

PURPOSE

Ultrasound (US) imaging is an established imaging modality capable of offering video-rate volumetric images without ionizing radiation. It has the potential for intra-fraction motion tracking in radiation therapy. In this study, a deep learning-based method has been developed to tackle the challenges in motion tracking using US imaging.

METHODS

We present a Markov-like network, which is implemented via generative adversarial networks, to extract features from sequential US frames (one tracked frame followed by untracked frames) and thereby estimate a set of deformation vector fields (DVFs) through the registration of the tracked frame and the untracked frames. The positions of the landmarks in the untracked frames are finally determined by shifting landmarks in the tracked frame according to the estimated DVFs. The performance of the proposed method was evaluated on the testing dataset by calculating the tracking error (TE) between the predicted and ground truth landmarks on each frame.

RESULTS

The proposed method was evaluated using the MICCAI CLUST 2015 dataset which was collected using seven US scanners with eight types of transducers and the Cardiac Acquisitions for Multi-structure Ultrasound Segmentation (CAMUS) dataset which was acquired using GE Vivid E95 ultrasound scanners. The CLUST dataset contains 63 2D and 22 3D US image sequences respectively from 42 and 18 subjects, and the CAMUS dataset includes 2D US images from 450 patients. On CLUST dataset, our proposed method achieved a mean tracking error of 0.70 ± 0.38 mm for the 2D sequences and 1.71 ± 0.84 mm for the 3D sequences for those public available annotations. And on CAMUS dataset, a mean tracking error of 0.54 ± 1.24 mm for the landmarks in the left atrium was achieved.

CONCLUSIONS

A novel motion tracking algorithm using US images based on modern deep learning techniques has been demonstrated in this study. The proposed method can offer millimeter-level tumor motion prediction in real time, which has the potential to be adopted into routine tumor motion management in radiation therapy.

NRG Oncology Survey of Monte Carlo Dose Calculation Use in US Proton Therapy Centers

Purpose/Objective(s)

Monte Carlo (MC) dose calculation has appeared in primary commercial treatment-planning systems and various in-house platforms. Dual-energy computed tomography (DECT) and metal artifact reduction (MAR) techniques complement MC capabilities. However, no publications have yet reported how proton therapy centers implement these new technologies, and a national survey is required to determine the feasibility of including MC and companion techniques in cooperative group clinical trials.

Materials/Methods

A 9-question survey was designed to query key clinical parameters: scope of MC utilization, validation methods for heterogeneities, clinical site-specific imaging guidance, proton range uncertainties, and how implants are handled. A national survey was distributed to all 29 operational US proton therapy centers on 13 May 2019.

Results

We received responses from 25 centers (86% participation). Commercial MC was most commonly used for primary plan optimization (16 centers) or primary dose evaluation (18 centers), while in-house MC was used more frequently for secondary dose evaluation (7 centers). Based on the survey, MC was used infrequently for gastrointestinal, genitourinary, gynecology and extremity compared with other more heterogeneous disease sites (P < .007). Although many centers had published DECT research, only 3/25 centers had implemented DECT clinically, either in the treatment-planning system or to override implant materials. Most centers (64%) treated patients with metal implants on a case-by-case basis, with a variety of methods reported. Twenty-four centers (96%) used MAR images and overrode the surrounding tissue artifacts; however, there was no consensus on how to determine metal dimension, materials density, or stopping powers.

Conclusion

The use of MC for primary dose calculation and optimization was prevalent and, therefore, likely feasible for clinical trials. There was consensus to use MAR and override tissues surrounding metals but no consensus about how to use DECT and MAR for human tissues and implants. Development and standardization of these advanced technologies are strongly encouraged for vendors and clinical physicists.