Potential of hybrid computational phantoms for retrospective heart dosimetry after breast radiation therapy: a feasibility study

Cardiotoxicity in breast cancer treatment: Causes and mitigation

Survivorship issues and treatment related toxicities have considerably increased in breast cancer patients following improved therapeutic options. Cardiotoxicity has been a major treatment related side effects in these patients. Despite this being a well-known entity, the real magnitude of the problem remains an enigma. The amount of research in mitigation of cardiotoxicity or its management in breast cancer survivors is limited and there is an urgent need for finding solutions for the problem. In this article, we are reviewing the agents that cause cardiotoxicity and suggesting a proposal for follow up of breast cancer survivors in an attempt to reduce the magnitude of impact on their quality of life.

Mesh-type reference Korean phantoms (MRKPs) for adult male and female for use in radiation protection dosimetry

In the present study, to overcome the dosimetric limitations of the previous voxel-type reference Korean computational phantoms due to their limited voxel resolutions (i.e. on the order of millimeters) and the nature of voxel geometry, a pair of new reference Korean phantoms, called mesh-type reference Korean phantoms (MRKPs), were developed for the adult male and female in a high-quality/fidelity mesh format. The developed phantoms include all target and source regions required for effective dose calculation, even micrometer-scale target and source regions of the respiratory and alimentary tract organs, skin, urinary bladder, and eye lens. The developed phantoms, which are in either the polygon-mesh (PM) format or the tetrahedral-mesh (TM) format as necessary, can be directly used in several general-purpose Monte Carlo codes (e.g. Geant4, MCNP6, and PHITS) without voxelization. In order to understand the dosimetric impact of the new phantoms, the dose coefficients (=fluence-to-effective dose conversion coefficients) were calculated for photons and electrons with energies ranging from 10 keV to 10 GeV for the anterior-posterior (AP) irradiation geometry and compared with those of the previous voxel-type reference Korean phantoms. The results demonstrate that the effective dose coefficients of the MRKPs were generally similar to those of the previous voxel-type reference phantoms for photons; however, for electrons, significant differences were observed at energies lower than 1 MeV that were mainly due to the explicit definition of the 50 µm-thick radiosensitive target layer in the skin of the new mesh phantoms.

Is mean heart dose a relevant surrogate parameter of left ventricle and coronary arteries exposure during breast cancer radiotherapy: a dosimetric evaluation based on individually-determined radiation dose (BACCARAT study)

Background

Intra-individual heterogeneity of cardiac exposure is an issue in breast cancer (BC) radiotherapy that was poorly considered in previous cardiotoxicity studies mainly based on mean heart dose (MHD). This dosimetric study analyzes the distribution of individually-determined radiation doses to the heart and its substructures including coronary arteries and evaluate whether MHD is a relevant surrogate parameter of dose for these substructures.

Methods

Data were collected from the BACCARAT prospective study that included left or right unilateral BC patients treated with 3D-Conformal Radiotherapy (RT) between 2015 and 2017 and followed-up for 2 years with repeated cardiac imaging examinations. A coronary computed tomography angiography (CCTA) was performed before RT for all patients. Registration of the planning CT and CCTA images allowed delineation of the coronary arteries on the planning CT images. Using the 3D dose matrix generated during treatment planning and the added coronary contours, dose distributions were generated for whole heart and the following substructures: left ventricle (LV), left main coronary artery (LMCA), left anterior descending artery (LAD), left circumflex artery (LCX) and right coronary artery (RCA). A descriptive analysis of the physical doses in Gray (Gy) was performed, Dmean was the volume-weighted mean dose.

Results

Dose distributions were generated for 89 left-sided BC patients (MHD = 2.9 ± 1.5 Gy, Dmean_LAD = 15.7 ± 3.1 Gy) and 15 right-sided BC patients (MHD = 0.5 ± 0.1 Gy; Dmean_RCA = 1.2 ± 0.4 Gy). For left-sided BC patients, the ratio Dmean_LAD/MHD was around 5. Pearson correlation coefficients between MHD and Dmean for delineated substructures were all statistically significant. However, for all substructures, the coefficient of determination R² indicated that the proportion of the variance in Dmean of the substructure predictable from MHD was moderate to low (in particular R² = 0.45 for LAD). Among left-sided BC patients with MHD < 3Gy, 56% of patients could nevertheless receive LAD doses above 40Gy (V40 > 0).

Conclusion

Our study illustrates that MHD is not enough to predict with confidence individual patient dose to the LV and coronary arteries, in particular the LAD. For precise radiotherapy-induced cardiotoxicity studies it would be necessary to consider the distribution of doses within these cardiac substructures rather than just the MHD.

Trial registration

ClinicalTrials.gov: NCT02605512, Registered 6 November 2015 - Retrospectively registered.

Retrospective estimation of heart and lung doses in pediatric patients treated with spinal irradiation

BACKGROUND AND PURPOSE

The purpose of this study was to investigate whether treatment information from medical records can be used to estimate radiation doses to heart and lungs retrospectively in pediatric patients receiving spinal irradiation with conventional posterior fields.

MATERIAL AND METHODS

An algorithm for retrospective dosimetry in children treated with spinal irradiation was developed in a cohort of 21 pediatric patients with available CT-scans and treatment plans. We developed a multivariable linear regression model with explanatory variables identifiable in case note review for retrospective estimation of minimum, maximum, mean and V10%-V80% doses to the heart and lungs. Doses were estimated for both linear accelerator (Linac) and 60Co radiation therapy modalities.

RESULTS

Age and spinal field width were identified as statistically significant predictors of heart and lung doses in multivariable analyses (p < 0.01 in all models). Models showed excellent predictive performance with R2 = 0.70 for mean heart dose and 0.79 for mean lung dose, for Linac plans. In leave-one-out cross-validation analysis the average difference between predicted and actual mean heart dose was 6.7% and 7.6% of the prescription dose for Linac and 60Co plans, respectively, and 5.2% and 4.9% for mean lung dose. Due to the small sample size and large inter-patient variation in heart and lung dose, prospective studies validating these findings are highly warranted.

CONCLUSIONS

The models presented here provide retrospective estimates of heart and lung doses for historical cohorts of pediatric patients, thus facilitating studies of long-term adverse effects of radiation.

A review of uncertainties in radiotherapy dose reconstruction and their impacts on dose-response relationships

Proper understanding of the risk of radiation-induced late effects for patients receiving external photon beam radiotherapy requires the determination of reliable dose-response relationships. Although significant efforts have been devoted to improving dose estimates for the study of late effects, the most often questioned explanatory variable is still the dose. In this work, based on a literature review, we provide an in-depth description of the radiotherapy dose reconstruction process for the study of late effects. In particular, we focus on the identification of the main sources of dose uncertainty involved in this process and summarise their impacts on the dose-response relationship for radiotherapy late effects. We provide a number of recommendations for making progress in estimating the uncertainties in current studies of radiotherapy late effects and reducing these uncertainties in future studies.

Early detection and prediction of cardiotoxicity after radiation therapy for breast cancer: the BACCARAT prospective cohort study

Background

Radiotherapy (RT) for breast cancer presents a benefit in terms of reducing local recurrence and deaths resulting from breast cancer but it can lead to secondary effects due to the presence of neighboring cardiac normal tissues within the irradiation field. Breast RT has been shown to be associated with long-term increased risk of heart failure, coronary artery disease, myocardial infarction and finally cardiovascular death more than 10 years after RT. However, there is still a lack of knowledge for early cardiotoxicity induced by breast RT that can appear long before the onset of clinically significant cardiac events. Based on a 2-year follow-up prospective cohort of patients treated with breast RT, the BACCARAT (BreAst Cancer and CArdiotoxicity Induced by RAdioTherapy) study aims to enhance knowledge on detection and prediction of early subclinical cardiac dysfunction and lesions induced by breast RT and on biological mechanisms potentially involved, based on functional and anatomical cardiac imaging combined with simultaneous assessment of multiple circulating biomarkers and accurate heart dosimetry.

Methods/Design

BACCARAT study consists in a monocentric prospective cohort study that will finally include 120 women treated with adjuvant 3D CRT for breast cancer, and followed for 2 years after RT. Women aged 50 to 70 years, treated for breast cancer and for whom adjuvant 3D CRT is indicated, without chemotherapy are eligible for the study. Baseline (before RT) and follow-up data include measurements of functional myocardial dysfunction including strain and strain rate based on 2D-speckle tracking echocardiography, anatomical coronary lesions including description of plaques in segments of coronary arteries based on Coronary computed tomography angiography, and a wide panel of circulating biomarkers. The absorbed dose is evaluated for the whole heart and its substructures, in particular the coronary arteries. Analysis on occurrence and evolution of subclinical cardiac lesions and biomarkers will be performed and completed with dose-response relationship. Multivariate model of normal tissue complication probability (NTCP) will also be proposed.

Discussion

Tools and results developed in the BACCARAT study should allow improving prediction and prevention of potential lesions to cardiac normal tissues surrounding tumors and ultimately enhance patients’ care and quality of life.

Trial registration

ClinicalTrials.gov: NCT02605512

Coronary stenosis risk analysis following Hodgkin lymphoma radiotherapy: A study based on patient specific artery segments dose calculation

BACKGROUND AND PURPOSE

The dose effect-effect relationship for cardiac diseases following radiotherapy suffers from uncertainties. Three dimensional coronary artery (CA) dose calculation after mediastinal Hodgkin lymphoma radiotherapy was performed, using the patient's coronary CT angiography (CCTA), and the relationship between the coronary arteries' radiation doses and the risk of stenosis was estimated.

MATERIALS AND METHODS

Radiotherapy simulation CT scans and CCTAs of patients treated for a mediastinal Hodgkin lymphoma were used to merge thoracic and detailed cardiovascular anatomies. Radiation treatment parameters were used to estimate CA radiation doses. Twenty-one patients without coronary stenosis (controls) were matched with twelve patients with stenosis (cases). CA segments were considered as sub-volumes of interest. Radiation doses to stenotic segments were compared with those received by normal segments (from cases and controls) using a logistic regression.

RESULTS

In eleven cases out of twelve, the highest of the coronary dose distribution was on a damaged segment. Logistic regression with CA segments yielded an odds ratio associated with the risk of coronary stenosis of 1.049 per additional gray with the CA segment median dose (95% confidence interval, 1.004-1.095; p-value <0.05).

CONCLUSION

The CA segment dose significantly increased the risk of stenosis on the segment. Such personalized CA dose calculations on larger cohorts are expected to improve the understanding of the cardiovascular radiation dose-effect relationship.

Reconstruction of organ dose for external radiotherapy patients in retrospective epidemiologic studies

Organ dose estimation for retrospective epidemiological studies of late effects in radiotherapy patients involves two challenges: radiological images to represent patient anatomy are not usually available for patient cohorts who were treated years ago, and efficient dose reconstruction methods for large-scale patient cohorts are not well established. In the current study, we developed methods to reconstruct organ doses for radiotherapy patients by using a series of computational human phantoms coupled with a commercial treatment planning system (TPS) and a radiotherapy-dedicated Monte Carlo transport code, and performed illustrative dose calculations. First, we developed methods to convert the anatomy and organ contours of the pediatric and adult hybrid computational phantom series to Digital Imaging and Communications in Medicine (DICOM)-image and DICOM-structure files, respectively. The resulting DICOM files were imported to a commercial TPS for simulating radiotherapy and dose calculation for in-field organs. The conversion process was validated by comparing electron densities relative to water and organ volumes between the hybrid phantoms and the DICOM files imported in TPS, which showed agreements within 0.1 and 2%, respectively. Second, we developed a procedure to transfer DICOM-RT files generated from the TPS directly to a Monte Carlo transport code, x-ray Voxel Monte Carlo (XVMC) for more accurate dose calculations. Third, to illustrate the performance of the established methods, we simulated a whole brain treatment for the 10 year-old male phantom and a prostate treatment for the adult male phantom. Radiation doses to selected organs were calculated using the TPS and XVMC, and compared to each other. Organ average doses from the two methods matched within 7%, whereas maximum and minimum point doses differed up to 45%. The dosimetry methods and procedures established in this study will be useful for the reconstruction of organ dose to support retrospective epidemiological studies of late effects in radiotherapy patients.

An exponential growth of computational phantom research in radiation protection, imaging, and radiotherapy: a review of the fifty-year history

Radiation dose calculation using models of the human anatomy has been a subject of great interest to radiation protection, medical imaging, and radiotherapy. However, early pioneers of this field did not foresee the exponential growth of research activity as observed today. This review article walks the reader through the history of the research and development in this field of study which started some 50 years ago. This review identifies a clear progression of computational phantom complexity which can be denoted by three distinct generations. The first generation of stylized phantoms, representing a grouping of less than dozen models, was initially developed in the 1960s at Oak Ridge National Laboratory to calculate internal doses from nuclear medicine procedures. Despite their anatomical simplicity, these computational phantoms were the best tools available at the time for internal/external dosimetry, image evaluation, and treatment dose evaluations. A second generation of a large number of voxelized phantoms arose rapidly in the late 1980s as a result of the increased availability of tomographic medical imaging and computers. Surprisingly, the last decade saw the emergence of the third generation of phantoms which are based on advanced geometries called boundary representation (BREP) in the form of Non-Uniform Rational B-Splines (NURBS) or polygonal meshes. This new class of phantoms now consists of over 287 models including those used for non-ionizing radiation applications. This review article aims to provide the reader with a general understanding of how the field of computational phantoms came about and the technical challenges it faced at different times. This goal is achieved by defining basic geometry modeling techniques and by analyzing selected phantoms in terms of geometrical features and dosimetric problems to be solved. The rich historical information is summarized in four tables that are aided by highlights in the text on how some of the most well-known phantoms were developed and used in practice. Some of the information covered in this review has not been previously reported, for example, the CAM and CAF phantoms developed in 1970s for space radiation applications. The author also clarifies confusion about 'population-average' prospective dosimetry needed for radiological protection under the current ICRP radiation protection system and 'individualized' retrospective dosimetry often performed for medical physics studies. To illustrate the impact of computational phantoms, a section of this article is devoted to examples from the author's own research group. Finally the author explains an unexpected finding during the course of preparing for this article that the phantoms from the past 50 years followed a pattern of exponential growth. The review ends on a brief discussion of future research needs (a supplementary file '3DPhantoms.pdf' to figure 15 is available for download that will allow a reader to interactively visualize the phantoms in 3D).

Comparison of organs' shapes with geometric and Zernike 3D moments

The morphological similarity of organs is studied with feature vectors based on geometric and Zernike 3D moments. It is particularly investigated if outliers and average models can be identified. For this purpose, the relative proximity to the mean feature vector is defined, principal coordinate and clustering analyses are also performed. To study the consistency and usefulness of this approach, 17 livers and 76 hearts voxel models from several sources are considered. In the liver case, models with similar morphological feature are identified. For the limited amount of studied cases, the liver of the ICRP male voxel model is identified as a better surrogate than the female one. For hearts, the clustering analysis shows that three heart shapes represent about 80% of the morphological variations. The relative proximity and clustering analysis rather consistently identify outliers and average models. For the two cases, identification of outliers and surrogate of average models is rather robust. However, deeper classification of morphological feature is subject to caution and can only be performed after cross analysis of at least two kinds of feature vectors. Finally, the Zernike moments contain all the information needed to re-construct the studied objects and thus appear as a promising tool to derive statistical organ shapes.