Predicting Patient-specific Dosimetric Benefits of Proton Therapy for Skull-base Tumors Using a Geometric Knowledge-based Method

Implementation of Machine Learning Models to Ensure Radiotherapy Quality for Multicenter Clinical Trials: Report from a Phase III Lung Cancer Study

The outcome of the patient and the success of clinical trials involving RT is dependent on the quality assurance of the RT plans. Knowledge-based Planning (KBP) models using data from a library of high-quality plans have been utilized in radiotherapy to guide treatment. In this study, we report on the use of these machine learning tools to guide the quality assurance of multicenter clinical trial plans. The data from 130 patients submitted to RTOG1308 were included in this study. Fifty patient cases were used to train separate photon and proton models on a commercially available platform based on principal component analysis. Models evaluated 80 patient cases. Statistical comparisons were made between the KBP plans and the original plans submitted for quality evaluation. Both photon and proton KBP plans demonstrate a statistically significant improvement of quality in terms of organ-at-risk (OAR) sparing. Proton KBP plans, a relatively emerging technique, show more improvements compared with photon plans. The KBP proton model is a useful tool for creating proton plans that adhere to protocol requirements. The KBP tool was also shown to be a useful tool for evaluating the quality of RT plans in the multicenter clinical trial setting.

A Decision Support Tool to Optimize Selection of Head and Neck Cancer Patients for Proton Therapy

Simple Summary

A decision support tool was developed to select head and neck cancer patients for proton therapy. The tool uses delineation data to predict expected toxicity risk reduction with proton therapy and can be used before a treatment plan is created. The positive predictive value of the tool is >90%. This tool significantly reduces delays in commencing treatment and avoid redundant photon vs. proton treatment plan comparison.

Abstract

Selection of head and neck cancer (HNC) patients for proton therapy (PT) using plan comparison (VMAT vs. IMPT) for each patient is labor-intensive. Our aim was to develop a decision support tool to identify patients with high probability to qualify for PT, at a very early stage (immediately after delineation) to avoid delay in treatment initiation. A total of 151 HNC patients were included, of which 106 (70%) patients qualified for PT. Linear regression models for individual OARs were created to predict the D_mean to the OARs for VMAT and IMPT plans. The predictors were OAR volume percentages overlapping with target volumes. Then, actual and predicted plan comparison decisions were compared. Actual and predicted OAR D_mean (VMAT R² = 0.953, IMPT R² = 0.975) and NTCP values (VMAT R² = 0.986, IMPT R² = 0.992) were highly correlated. The sensitivity, specificity, PPV and NPV of the decision support tool were 64%, 87%, 92% and 51%, respectively. The expected toxicity reduction with IMPT can be predicted using only the delineation data. The probability of qualifying for PT is >90% when the tool indicates a positive outcome for PT. This tool will contribute significantly to a more effective selection of HNC patients for PT at a much earlier stage, reducing treatment delay.

Proton therapy with a fixed beamline for skull-base chordomas and chondrosarcomas: outcomes and toxicity

Aim

This study presents an analysis (efficacy and toxicity) of outcomes in patients with skull-base chordomas or chondrosarcomas treated with a fixed horizontal pencil proton beam.

Background

Chordomas (CAs) and chondrosarcomas (CSAs) are rare tumours that are usually located near the base of the skull and very close to the brain's most critical structures. Proton therapy (PT) is often considered the best radiation treatment for these diseases, but it is still a limited resource. Active scanning PT delivered via a fixed pencil beamline might be a promising option.

Methods

This is a single-centre experience describing the results of proton therapy for 31 patients with CA (n = 23) or CSA (n = 8) located near the base of the skull. Proton therapy was utilized by a fixed pencil beamline with a chair to position the patient between May 2016 and November 2020. Ten patients underwent resection (32.2%), 15 patients (48.4%) underwent R2 resection, and 6 patients had unresectable tumours (19.4%). In 4 cases, the tumours had been previously irradiated. The median PT dose was 70 GyRBE (relative biological efficacy, 1.1) [range, 60 to 74] with 2.0 GyRBE per fraction. The mean GTV volume was 25.6 cm³ [range, 4.2–115.6]. Patient demographics, pathology, treatment parameters, and toxicity were collected and analysed. Radiation-induced reactions were assessed according to the Common Terminology Criteria for Adverse Events (CTCAE) v 4.0.

Results

The median follow-up time was 21 months [range, 4 to 52]. The median overall survival (OS) was 40 months. The 1- and 2-year OS was 100%, and the 3-year OS was 66.3%. Four patients died due to non-cancer-related reasons, 1 patient died due to tumour progression, and 1 patient died due to treatment-related injuries. The 1-year local control (LC) rate was 100%, the 2-year LC rate was 93.7%, and the 3-year LC rate was 85.3%. Two patients with CSA exhibited progression in the neck lymph nodes and lungs. All patients tolerated PT well without any treatment interruptions. We observed 2 cases of ≥ grade 3 toxicity, with 1 case of grade 3 myelitis and 1 case of grade 5 brainstem injury.

Conclusion

Treatment with a fixed proton beam shows promising disease control and an acceptable toxicity rate, even the difficult-to-treat subpopulation of patients with skull-base chordomas or chondrosarcomas requiring dose escalation.

A scoping review of patient selection methods for proton therapy

The aim was to explore various national and international clinical decision‐making tools and dose comparison methods used for selecting cancer patients for proton versus X‐ray radiation therapy. To address this aim, a literature search using defined scoping review methods was performed in Medline and Embase databases as well as grey literature. Articles published between 1 January 2015 and 4 August 2020 and those that clearly stated methods of proton versus X‐ray therapy patient selection and those published in English were eligible for inclusion. In total, 321 studies were identified of which 49 articles met the study’s inclusion criteria representing 13 countries. Six different clinical decision‐making tools and 14 dose comparison methods were identified, demonstrating variability within countries and internationally. Proton therapy was indicated for all paediatric patients except those with lymphoma and re‐irradiation where individualised model‐based selection was required. The most commonly reported patient selection tools included the Normal Tissue Complication Probability model, followed by cost‐effectiveness modelling and dosimetry comparison. Model‐based selection methods were most commonly applied for head and neck clinical indications in adult cohorts (48% of studies). While no ‘Gold Standard’ currently exists for proton therapy patient selection with variations evidenced globally, some of the patient selection methods identified in this review can be used to inform future practice in Australia. As literature was not identified from all countries where proton therapy centres are available, further research is needed to evaluate patient selection methods in these jurisdictions for a comprehensive overview.

Perspectives on the model-based approach to proton therapy trials: A retrospective study of a lung cancer randomized trial

PURPOSE

The goal of this study was to assess whether a model-based approach applied retrospectively to a completed randomized controlled trial (RCT) would have significantly altered the selection of patients of the original trial, using the same selection criteria and endpoint for testing the potential clinical benefit of protons compared to photons.

METHODS AND MATERIALS

A model-based approach, based on three widely used normal tissue complication probability (NTCP) models for radiation pneumonitis (RP), was applied retrospectively to a completed non-small cell lung cancer RCT (NCT00915005). It was assumed that patients were selected by the model-based approach if their expected ΔNTCP value was above a threshold of 5%. The endpoint chosen matched that of the original trial, the first occurrence of severe (grade ≥3) RP.

RESULTS

Our analysis demonstrates that NTCP differences between proton and photon therapy treatments may be too small to support a model-based trial approach for lung cancer using RP as the normal tissue endpoint. The analyzed lung trial showed that less than 19% (32/165) of patients enrolled in the completed trial would have been enrolled in a model-based trial, prescribing photon therapy to all other patients. The number of patients enrolled was also found to be dependent on the type of NTCP model used for evaluating RP, with the three models enrolling 3%, 13% or 19% of patients. This result does show limitations in NTCP models which would affect the success of a model-based trial approach. No conclusion regarding the development of RP in patients randomized by the model-based approach could statistically be made.

CONCLUSIONS

Uncertainties in the outcome models to predict NTCP are the inherent drawback of a model-based approach to clinical trials. The impact of these uncertainties on enrollment in model-based trials depends on the predicted difference between the two treatment arms and on the set threshold for patient stratification. Our analysis demonstrates that NTCP differences between proton and photon therapy treatments may be too small to support a model-based trial approach for specific treatment sites, such as lung cancer, depending on the chosen normal tissue endpoint.

Combined proton-photon treatments - A new approach to proton therapy without a gantry

PURPOSE

Although the number of proton therapy centres is growing worldwide, proton therapy is still a limited resource. The primary reasons are gantry size and cost. Therefore, we investigate the potential of a new design for proton therapy, which may facilitate proton treatments in conventional bunkers and allow the widespread use of protons.

MATERIALS AND METHODS

The treatment room consists of a standard Linac for IMRT, a motorized couch for treatments in lying position, and a horizontal proton beamline equipped with pencil beam scanning. As proton beams are limited to a coronal plane, treatment plans may be suboptimal for many tumour sites. However, high-quality plans may be realized by combining protons and photons. Treatment planning is performed by simultaneously optimizing IMRT and IMPT plans based on their cumulative physical dose. We demonstrate this concept for three head&neck cancer cases.

RESULTS

Optimal combinations use photons to improve dose conformity while protons reduce the integral dose to normal tissues. In fact, combined treatments improve on single-modality IMRT and fixed beamline IMPT plans for quality-of-life-limiting OARs and retain most of the integral dose reduction in the healthy tissues of the pure IMPT plans. The lower doses that can be obtained with multi-modality treatments reduce the risk for side effects compared to single-modality IMRT plans.

CONCLUSION

Combined proton-photon treatments may play a role in developing a new solution for proton therapy without a gantry. Optimal combinations improve on IMRT plans and reduce the risk of side effects while making protons available to more patients.

Radiobiological and dosimetric impact of RayStation pencil beam and Monte Carlo algorithms on intensity-modulated proton therapy breast cancer plans

PURPOSE

RayStation treatment planning system employs pencil beam (PB) and Monte Carlo (MC) algorithms for proton dose calculations. The purpose of this study is to evaluate the radiobiological and dosimetric impact of RayStation PB and MC algorithms on the intensity-modulated proton therapy (IMPT) breast plans.

METHODS

The current study included ten breast cancer patients, and each patient was treated with 1-2 proton beams to the whole breast/chestwall (CW) and regional lymph nodes in 28 fractions for a total dose of 50.4 Gy relative biological effectiveness (RBE). A total clinical target volume (CTV_Total) was generated by combining individual CTVs: AxI, AxII, AxIII, CW, IMN, and SCVN. All beams in the study were treated with a range shifter (7.5 cm water equivalent thickness). For each patient, three sets of plans were generated: (a) PB optimization followed by PB dose calculation (PB-PB), (b) PB optimization followed by MC dose calculation (PB-MC), and (c) MC optimization followed by MC dose calculation (MC-MC). For a given patient, each plan was robustly optimized on the CTVs with same parameters and objectives. Treatment plans were evaluated using dosimetric and radiobiological indices (equivalent uniform dose (EUD), tumor control probability (TCP), and normal tissue complication probability (NTCP)).

RESULTS

The results are averaged over ten breast cancer patients. In comparison to PB-PB plans, PB-MC plans showed a reduction in CTV target dose by 5.3% for D_99% and 4.1% for D_95% , as well as a reduction in TCP by 1.5-2.1%. Similarly, PB overestimated the EUD of target volumes by 1.8─3.2 Gy(RBE). In contrast, MC-MC plans achieved similar dosimetric and radiobiological (EUD and TCP) results as the ones in PB-PB plans. A selection of one dose calculation algorithm over another did not produce any noticeable differences in the NTCP of the heart, lung, and skin.

CONCLUSION

If MC is more accurate than PB as reported in the literature, dosimetric and radiobiological results from the current study suggest that PB overestimates the target dose, EUD, and TCP for IMPT breast cancer treatment. The overestimation of dosimetric and radiobiological results of the target volume by PB needs to be further interpreted in terms of clinical outcome.

Dosimetry and characterization of a 25-MeV proton beam line for preclinical radiobiology research

PURPOSE

With the increase in proton therapy centers, there is a growing need to make progress in preclinical proton radiation biology to give accessible data to medical physicists and practicing radiation oncologists.

METHODS

A cyclotron usually producing radioisotopes with a proton beam at an energy of about 25 MeV after acceleration, was used for radiobiology studies. Depleted silicon surface barrier detectors were used for the beam energy measurement. A complementary metal oxide semiconductor (CMOS) sensor and a plastic scintillator detector were used for fluence measurement, and compared to Geant4 and an in-house analytical dose modeling developed for this purpose. Also, from the energy measurement of each attenuated beam, the dose-averaged linear energy transfer (LET_d ) was calculated with Geant4.

RESULTS

The measured proton beam energy was 24.85 ± 0.14 MeV with an energy straggling of 127 ± 22 keV before scattering and extraction in air. The measured flatness was within ± 2.1% over 9 mm in diameter. A wide range of LET_d is achievable: constant between the entrance and the exit of the cancer cell sample ranging from 2.2 to 8 keV/μm, beyond 20 keV/μm, and an average of 2-5 keV/μm in a scattering spread-out Bragg peak calculated for an example of a 6-mm-thick xenograft tumor.

CONCLUSION

The dosimetry and the characterization of a 25-MeV proton beam line for preclinical radiobiology research was performed by measurements and modeling, demonstrating the feasibility of delivering a proton beam for preclinical in vivo and in vitro studies with LET_d of clinical interest.

Evaluation of an Automated Proton Planning Solution

Purpose Intensity-modulated proton therapy (IMPT) treatments are increasing, however, treatment planning remains complex and prone to variability. RapidPlanTMPT (Varian Medical Systems, Palo Alto, California, USA) is a pre-clinical, proton-specific, automated knowledge-based planning solution which could reduce variability and increase efficiency. It uses a library of previous IMPT treatment plans to generate a model which can predict organ-at-risk (OAR) dose for new patients, and guide IMPT optimization. This study details and evaluates RapidPlanTMPT. Methods IMPT treatment plans for 50 head-and-neck cancer patients populated the model-library. The model was then used to create knowledge-based plans (KBPs) for 10 evaluation-patients. Model quality and accuracy were evaluated using model-provided OAR regression plots and examining the difference between predicted and achieved KBP mean dose. KBP quality was assessed through comparison with respective manual IMPT plans on the basis of boost/elective planning target volume (PTVB/PTVE) homogeneity and OAR sparing. The time to create KBPs was recorded. Results Model quality was good, with an average R2 of 0.85 between dosimetric and geometric features. The model showed high predictive accuracy with differences of <3 Gy between predicted and achieved OAR mean doses for 88/109 OARs. On average, KBPs were comparable to manual IMPT plans with differences of <0.6% in homogeneity. Only 2 of 109 OARs in KBPs had a mean dose >3 Gy more than the manual plan. On average, dose-volume histogram (DVH) predictions required 0.7 minutes while KBP optimization and dose calculation required 4.1 minutes (a 'continue optimization' phase, if required, took an additional 2.8 minutes, on average). Conclusions RapidPlanTMPT demonstrated efficiency and consistency and IMPT KBPs were comparable to manual plans. Because worse OAR sparing in a KBP was not always associated with geometric-outlier warnings, manual plan checks remain important. Such an automated planning solution could also assist in clinical trial quality assurance and overcome the learning curve associated with IMPT.

Investigation of Optimal Physical Parameters for Precise Proton Irradiation of Orthotopic Tumors in Small Animals

PURPOSE

The lack of evidence of biomarkers identifying patients who would benefit from proton therapy has driven the emergence of preclinical proton irradiation platforms using advanced small-animal models to mimic clinical therapeutic conditions. This study aimed to determine the optimal physical parameters of the proton beam with a high radiation targeting accuracy, considering small-animal tumors can reach millimetric dimensions at a maximum depth of about 2 cm.

METHODS AND MATERIALS

Several treatment plans, simulated using Geant4, were generated with different proton beam features to assess the optimal physical parameters for small-volume irradiations. The quality of each treatment plan was estimated by dose-volume histograms and gamma index maps.

RESULTS

Because of its low-energy straggling, low-energy proton (<50 MeV) single-field irradiation can generate homogeneous spread-out Bragg peaks to deliver a uniform dose in millimeter-sized tumors, while sparing healthy tissues located within or near the target volume. However, multifield irradiation can limit the dose delivered in critical structures surrounding the target for attenuated high-energy beams (E > 160 MeV).

CONCLUSION

Low-energy proton beam platforms are suitable for precision irradiation for translational radiobiology studies.

Clinical Trial Strategies to Compare Protons With Photons

The favorable beam properties of protons can be translated into clinical benefits by target dose escalation to improve local control without enhancing unacceptable radiation toxicity or to spare normal tissues to prevent radiation-induced side effects without jeopardizing local tumor control. For the clinical validation of the added value of protons to improve local control, randomized controlled trials are required. For the clinical validation of the added value of protons to prevent side effects, both model-based validation or randomized controlled trials can be used. Model-based patient selection for proton therapy is crucial, independent of the validation approach. Combining these approaches in rapid learning health care systems is expected to yield the most efficient and scientifically sound way to continuously improve patient selection and the therapeutic window, eventually leading to more cancer survivors with better quality of life.

Contour-based lung dose prediction for breast proton therapy

PURPOSE

This study evaluates the feasibility of lung dose prediction based on target contour and patient anatomy for breast patients treated with proton therapy.

METHODS

Fifty-two randomly selected patients were included in the cohort, who were treated to 50.4-66.4 Gy(RBE) to the left (36), right (15), or bilateral (1) breast with uniform scanning (32) or pencil beam scanning (20). Anterior-oblique beams were used for each patient. The prescription doses were all scaled to 50.4 Gy(RBE) for the current analysis. Isotropic expansions of the planning target volume of various margins m were retrospectively generated and compared with isodose volumes in the ipsilateral lung. The fractional volume V of each expansion contour within the ipsilateral lung was compared with dose-volume data of clinical plans to establish the relationship between the margin m and dose D for the ipsilateral lung such that VD  = V(m). This relationship enables prediction of dose-volume VD from V(m), which could be derived from contours before any plan is generated, providing a goal of plan quality. Lung V20 Gy( RBE ) and V5 Gy( RBE ) were considered for this pilot study, while the results could be generalized to other dose levels and/or other organs.

RESULTS

The actual V20 Gy( RBE ) ranged from 6% to 23%. No statistically significant difference in V20 Gy( RBE ) was found between breast irradiation and chest wall irradiation (P = 0.8) or between left-side and right-side treatment (P = 0.9). It was found that V(1.1 cm) predicted V20 Gy( RBE ) to within 5% root-mean-square deviation (RMSD) and V(2.2 cm) predicted V5 Gy( RBE ) to within 6% RMSD.

CONCLUSION

A contour-based model was established to predict dose to ipsilateral lung in breast treatment. Clinically relevant accuracy was demonstrated. This model facilitates dose prediction before treatment planning. It could serve as a guide toward realistic clinical goals in the planning stage.

Automation in intensity modulated radiotherapy treatment planning-a review of recent innovations

Radiotherapy treatment planning of complex radiotherapy techniques, such as intensity modulated radiotherapy and volumetric modulated arc therapy, is a resource-intensive process requiring a high level of treatment planner intervention to ensure high plan quality. This can lead to variability in the quality of treatment plans and the efficiency in which plans are produced, depending on the skills and experience of the operator and available planning time. Within the last few years, there has been significant progress in the research and development of intensity modulated radiotherapy treatment planning approaches with automation support, with most commercial manufacturers now offering some form of solution. There is a rapidly growing number of research articles published in the scientific literature on the topic. This paper critically reviews the body of publications up to April 2018. The review describes the different types of automation algorithms, including the advantages and current limitations. Also included is a discussion on the potential issues with routine clinical implementation of such software, and highlights areas for future research.

Artificial intelligence in radiation oncology: A specialty-wide disruptive transformation?

Artificial intelligence (AI) is emerging as a technology with the power to transform established industries, and with applications from automated manufacturing to advertising and facial recognition to fully autonomous transportation. Advances in each of these domains have led some to call AI the "fourth" industrial revolution [1]. In healthcare, AI is emerging as both a productive and disruptive force across many disciplines. This is perhaps most evident in Diagnostic Radiology and Pathology, specialties largely built around the processing and complex interpretation of medical images, where the role of AI is increasingly seen as both a boon and a threat. In Radiation Oncology as well, AI seems poised to reshape the specialty in significant ways, though the impact of AI has been relatively limited at present, and may rightly seem more distant to many, given the predominantly interpersonal and complex interventional nature of the specialty. In this overview, we will explore the current state and anticipated future impact of AI on Radiation Oncology, in detail, focusing on key topics from multiple stakeholder perspectives, as well as the role our specialty may play in helping to shape the future of AI within the larger spectrum of medicine.

Machine Learning in Radiation Oncology: Opportunities, Requirements, and Needs

Machine learning (ML) has the potential to revolutionize the field of radiation oncology, but there is much work to be done. In this article, we approach the radiotherapy process from a workflow perspective, identifying specific areas where a data-centric approach using ML could improve the quality and efficiency of patient care. We highlight areas where ML has already been used, and identify areas where we should invest additional resources. We believe that this article can serve as a guide for both clinicians and researchers to start discussing issues that must be addressed in a timely manner.