Photon and Proton Dose Painting Based on Oxygen Distribution - Feasibility Study and Tumour Control Probability Assessment.

Solid tumours may present hypoxic sub-regions of increased radioresistance. Hypoxia quantification requires of clinically implementable, non-invasive and reproducible techniques as positron emission tomography (PET). PET-based dose painting strategies aiming at targeting those sub-regions may be limited by the resolution gap between the PET imaging resolution and the smaller scale at which hypoxia occurs. The ultimate benefit of the usage of dose painting may be reached if the planned dose distribution can be performed and delivered consistently. This study aimed at assessing the feasibility of two PET-based dose painting strategies using two beam qualities (photon or proton beams) in terms of tumour control probability (TCP), accounting for underlying oxygen distribution at sub-millimetre scale.A tumour oxygenation model at submillimetre scale was created consisting of three regions with different oxygen partial pressure distributions, being hypoxia decreasing from core to periphery. A published relationship between uptake and oxygen partial pressure was used and a PET image of the tumour was simulated. The fundamental effects that limit the PET camera resolution were considered by processing the uptake distribution with a Gaussian 3D filter and re-binning to a PET image voxel size of 2 mm. Prescription doses to overcome tumour hypoxia were calculated based on the processed images, and planned using robust optimisation.Normal tissue complication probabilities and TCPs after the delivery of the planned doses were calculated for the nominal plan and the lowest bounds of the dose volume histograms resulting from the robust scenarios planned, taking into account the underlying oxygenation at submillimetre scale. Results were presented for the two beam qualities and the two dose painting strategies: by contours (DPBC) and by using a voxel grouping-based approach (DPBOX).In the studied case, DPBOX outperforms DPBC with respect to TCP regardless the beam quality, although both dose painting strategy plans demonstrated robust target coverage.

Assessment of the Probability of Tumour Control for Prescribed Doses Based on Imaging of Oxygen Partial Pressure.

In radiotherapy, hypoxia is a known negative factor, occurring especially in solid malignant tumours. Nitroimidazole-based positron emission tomography (PET) tracers, due to their selective binding to hypoxic cells, could be used as surrogates to image and quantify the underlying oxygen distributions in tissues. The spatial resolution of a clinical PET image, however, is much larger than the cellular spatial scale where hypoxia occurs. A question therefore arises regarding the possibility of quantifying different hypoxia levels based on PET images, and the aim of the present study is the prescription of corresponding therapeutic doses and its exploration.A tumour oxygenation model was created consisting of two concentric spheres with different oxygen partial pressure (pO2) distributions. In order to mimic a PET image of the simulated tumour, given the relation between uptake and pO2, fundamental effects that limit spatial resolution in a PET imaging system were considered: the uptake distribution was processed with a Gaussian 3D filter, and a re-binning to reach a typical PET image voxel size was performed. Prescription doses to overcome tumour hypoxia and predicted tumour control probability (TCP) were calculated based on the processed images for several fractionation schemes. Knowing the underlying oxygenation at microscopic scale, the actual TCP expected after the delivery of the calculated prescription doses was evaluated. Results are presented for three different dose painting strategies: by numbers, by contours and by using a voxel grouping-based approach.The differences between predicted TCP and evaluated TCP indicate that careful consideration must be taken on the dose prescription strategy and the selection of the number of fractions, depending on the severity of hypoxia.

Mathematical Description of Changes in Tumour Oxygenation from Repeated Functional Imaging.

Functional imaging of tumour hypoxia has been suggested as a tool for refining target definition and treatment optimization in radiotherapy. The approach, however, has been slow to be adopted clinically as most of the studies on the topic do not take into account the in-treatment changes of hypoxia. The present study aimed to introduce a function that quantifies the changes of oxygen distributions in repeated PET images taken during treatment. The proposed approach for determining the reoxygenation function was tested for feasibility on patients with head and neck cancer, repeatedly imaged with FMISO PET during radiotherapy. Reoxygenation functions were derived by solving the convolution between functions describing the oxygen distributions of successive images. The method was found to be mathematically feasible. The results indicate that the reoxygenation functions describing the change in oxygenation have distinct shapes prompting the hypothesis that oxygenation changes reflected by them might have predictive power for treatment outcome. Future studies on a larger patient population to search for predictive correlations based on the reoxygenation function are planned.

Overall survival prediction for high-grade glioma patients using mathematical modeling of tumor cell infiltration.

PURPOSE: This study aimed at applying a mathematical framework for the prediction of high-grade gliomas (HGGs) cell invasion into normal tissues for guiding the clinical target delineation, and at investigating the possibility of using tumor infiltration maps for patient overall survival (OS) prediction. MATERIAL & METHODS: A model describing tumor infiltration into normal tissue was applied to 93 HGG cases. Tumor infiltration maps and corresponding isocontours with different cell densities were produced. ROC curves were used to seek correlations between the patient OS and the volume encompassed by a particular isocontour. Area-Under-the-Curve (AUC) values were used to determine the isocontour having the highest predictive ability. The optimal cut-off volume, having the highest sensitivity and specificity, for each isocontour was used to divide the patients in two groups for a Kaplan-Meier survival analysis. RESULTS: The highest AUC value was obtained for the isocontour of cell densities 1000 cells/mm3 and 2000 cells/mm3, equal to 0.77 (p < 0.05). Correlation with the GTV yielded an AUC of 0.73 (p < 0.05). The Kaplan-Meier survival analysis using the 1000 cells/mm3 isocontour and the ROC optimal cut-off volume for patient group selection rendered a hazard ratio (HR) of 2.7 (p < 0.05), while the GTV rendered a HR = 1.6 (p < 0.05). CONCLUSION: The simulated tumor cell invasion is a stronger predictor of overall survival than the segmented GTV, indicating the importance of using mathematical models for cell invasion to assist in the definition of the target for HGG patients.

CTV Delineation for High-Grade Gliomas: Is There Agreement With Tumor Cell Invasion Models?

Purpose: High-grade glioma (HGG) is a common form of malignant primary brain cancer with poor prognosis. The diffusive nature of HGGs implies that tumor cell invasion of normal tissue extends several centimeters away from the visible gross tumor volume (GTV). The standard methodology for clinical volume target (CTV) delineation is to apply a 2- to 3-cm margin around the GTV. However, tumor recurrence is extremely frequent. The purpose of this paper was to introduce a framework and computational model for the prediction of normal tissue HGG cell invasion and to investigate the agreement of the conventional CTV delineation with respect to the predicted tumor invasion. Methods and Materials: A model for HGG cell diffusion and proliferation was implemented and used to assess the tumor invasion patterns for 112 cases of HGGs. Normal brain structures and tissues as well as the GTVs visible on diagnostic images were delineated using automated methods. The volumes encompassed by different tumor cell concentration isolines calculated using the model for invasion were compared with the conventionally delineated CTVs, and the differences were analyzed. The 3-dimensional-Hausdorff distance between the CTV and the volumes encompassed by various isolines was also calculated. Results: In 50% of cases, the CTV failed to encompass regions containing tumor cell concentrations of 614 cells/mm³ or greater. In 84% of cases, the lowest cell concentration completely encompassed by the CTV was ≥1 cell/mm³. In the remaining 16%, the CTV overextended into normal tissue. The Hausdorff distance was on average comparable to the CTV margin. Conclusions: The standard methodology for CTV delineation appears to be inconsistent with HGG invasion patterns in terms of size and shape. Tumor invasion modeling could therefore be useful in assisting in the CTV delineation for HGGs.

Evolution of the hypoxic compartment on sequential oxygen partial pressure maps during radiochemotherapy in advanced head and neck cancer.

BACKGROUND AND PURPOSE: Longitudinal Positron Emission Tomography (PET) with hypoxia-specific radiotracers allows monitoring the time evolution of regions of increased radioresistance and may become fundamental in determining the radiochemotherapy outcome in Head-and-Neck Squamous Cell Carcinoma (HNSCC). The aim of this study was to investigate the evolution of the hypoxic target volume on oxygen partial pressure maps (pO2-HTV) derived from 18FMISO-PET images acquired before and during radiochemotherapy and to uncover correlations between extent and severity of hypoxia and treatment outcome. MATERIAL AND METHODS: 18FMISO-PET/CT images were acquired at three time points (before treatment start, in weeks two and five) for twenty-eight HNSCC patients treated with radiochemotherapy. The images were converted into pO2 maps and corresponding pO2-HTVs (pO2-HTV1, pO2-HTV2, pO2-HTV3) were contoured at 10 mmHg. Different parameters describing the pO2-HTV time evolution were considered, such as the percent and absolute difference between the pO2-HTVs (%HTVi,j and HTVi-HTVj with i,j = 1, 2, 3, respectively) and the slope of the linear regression curve fitting the pO2-HTVs in time. Correlations were sought between the pO2-HTV evolution parameters and loco-regional recurrence (LRR) using the Receiver Operating Characteristic method. RESULTS: The Area Under the Curve values for %HTV1,2, HTV1-HTV2, HTV1-HTV3 and the slope of the pO2-HTV linear regression curve were 0.75 (p = 0.04), 0.73 (p = 0.02), 0.73 (p = 0.02) and 0.75 (p = 0.007), respectively. Other parameter combinations were not statistically significant. CONCLUSIONS: The pO2-HTV evolution during radiochemotherapy showed predictive value for LRR. The changes in the tumour hypoxia during the first two treatment weeks may be used for adaptive personalized treatment approaches.

Early survival prediction in non-small cell lung cancer from PET/CT images using an intra-tumor partitioning method.

PURPOSE: To explore prognostic and predictive values of a novel quantitative feature set describing intra-tumor heterogeneity in patients with lung cancer treated with concurrent and sequential chemoradiotherapy. METHODS: Longitudinal PET-CT images of 30 patients with non-small cell lung cancer were analysed. To describe tumor cell heterogeneity, the tumors were partitioned into one to ten concentric regions depending on their sizes, and, for each region, the change in average intensity between the two scans was calculated for PET and CT images separately to form the proposed feature set. To validate the prognostic value of the proposed method, radiomics analysis was performed and a combination of the proposed novel feature set and the classic radiomic features was evaluated. A feature selection algorithm was utilized to identify the optimal features, and a linear support vector machine was trained for the task of overall survival prediction in terms of area under the receiver operating characteristic curve (AUROC). RESULTS: The proposed novel feature set was found to be prognostic and even outperformed the radiomics approach with a significant difference (AUROCSALoP = 0.90 vs. AUROCradiomic = 0.71) when feature selection was not employed, whereas with feature selection, a combination of the novel feature set and radiomics led to the highest prognostic values. CONCLUSION: A novel feature set designed for capturing intra-tumor heterogeneity was introduced. Judging by their prognostic power, the proposed features have a promising potential for early survival prediction.

Impact of Tumour Cell Infiltration on Treatment Outcome in Gamma Knife Radiosurgery: A Modelling Study.

BACKGROUND: High-grade gliomas with a widespread infiltration beyond the lesion detectable on diagnostic images are increasingly treated with Gamma Knife™ Radiosurgery (GKRS). The aim of this study was to assess the cell infiltration impact on the GKRS outcome for invasive gliomas. MATERIALS AND METHODS: Tumor cell distribution was predicted using a novel algorithm whose computations are iterated until they reach an agreement with histopathology results. Treatment plans with different combinations of dose prescription (20 Gy at 50%-20% isodose) and targets [Gross Tumour Volume (GTV), zone 1 with 100%-60% of the GTV cell density and zone 2 with 60%-0% of the GTV cell density] were evaluated using standard conformity indexes (CI) and radiobiological parameters. RESULTS: Considerable differences in terms of tumor control probability were found between plans having similar CI but different targets. CONCLUSION: To account for tumor cell infiltration outside the target is of key importance in GKRS and a radiobiological evaluation should accompany well-established CI.

Early tumor response prediction for lung cancer patients using novel longitudinal pattern features from sequential PET/CT image scans.

PURPOSE: A new set of quantitative features that capture intensity changes in PET/CT images over time and space is proposed for assessing the tumor response early during chemoradiotherapy. The hypothesis whether the new features, combined with machine learning, improve outcome prediction is tested. METHODS: The proposed method is based on dividing the tumor volume into successive zones depending on the distance to the tumor border. Mean intensity changes are computed within each zone, for CT and PET scans separately, and used as image features for tumor response assessment. Doing so, tumors are described by accounting for temporal and spatial changes at the same time. Using linear support vector machines, the new features were tested on 30 non-small cell lung cancer patients who underwent sequential or concurrent chemoradiotherapy. Prediction of 2-years overall survival was based on two PET-CT scans, acquired before the start and during the first 3 weeks of treatment. The predictive power of the newly proposed longitudinal pattern features was compared to that of previously proposed radiomics features and radiobiological parameters. RESULTS: The highest areas under the receiver operating characteristic curves were 0.98 and 0.93 for patients treated with sequential and concurrent chemoradiotherapy, respectively. Results showed an overall comparable performance with respect to radiomics features and radiobiological parameters. CONCLUSIONS: A novel set of quantitative image features, based on underlying tumor physiology, was computed from PET/CT scans and successfully employed to distinguish between early responders and non-responders to chemoradiotherapy.

Evaluating tumor response of non-small cell lung cancer patients with ¹8F-fludeoxyglucose positron emission tomography: potential for treatment individualization.

OBJECTIVE: To assess early tumor responsiveness and the corresponding effective radiosensitivity for individual patients with non-small cell lung cancer (NSCLC) based on 2 successive (18)F-fludeoxyglucose positron emission tomography (FDG-PET) scans. METHODS AND MATERIALS: Twenty-six NSCLC patients treated in Maastricht were included in the study. Fifteen patients underwent sequential chemoradiation therapy, and 11 patients received concomitant chemoradiation therapy. All patients were imaged with FDG before the start and during the second week of radiation therapy. The sequential images were analyzed in relation to the dose delivered until the second image. An operational quantity, effective radiosensitivity, αeff, was determined at the voxel level. Correlations were sought between the average αeff or the fraction of negative αeff values and the overall survival at 2 years. Separate analyses were performed for the primary gross target volume (GTV), the lymph node GTV, and the clinical target volumes (CTVs). RESULTS: Patients receiving sequential treatment could be divided into responders and nonresponders, using a threshold for the average αeff of 0.003 Gy(-1) in the primary GTV, with a sensitivity of 75% and a specificity of 100% (P<.0001). Choosing the fraction of negative αeff as a criterion, the threshold 0.3 also had a sensitivity of 75% and a specificity of 100% (P<.0001). Good prognostic potential was maintained for patients receiving concurrent chemotherapy. For lymph node GTV, the correlation had low statistical significance. A cross-validation analysis confirmed the potential of the method. CONCLUSIONS: Evaluation of the early response in NSCLC patients showed that it is feasible to determine a threshold value for effective radiosensitivity corresponding to good response. It also showed that a threshold value for the fraction of negative αeff could also be correlated with poor response. The proposed method, therefore, has potential to identify candidates for more aggressive strategies to increase the rate of local control and also avoid exposing to unnecessary aggressive therapies the majority of patients responding to standard treatment.

Evaluation of nuclear reaction cross-sections and fragment yields in carbon beams using the SHIELD-HIT Monte Carlo code. Comparison with experiments.

In light ion therapy, the knowledge of the spectra of both primary and secondary particles in the target volume is needed in order to accurately describe the treatment. The transport of ions in matter is complex and comprises both atomic and nuclear processes involving primary and secondary ions produced in the cascade of events. One of the critical issues in the simulation of ion transport is the modeling of inelastic nuclear reaction processes, in which projectile nuclei interact with target nuclei and give rise to nuclear fragments. In the Monte Carlo code SHIELD-HIT, inelastic nuclear reactions are described by the Many Stage Dynamical Model (MSDM), which includes models for the different stages of the interaction process. In this work, the capability of SHIELD-HIT to simulate the nuclear fragmentation of carbon ions in tissue-like materials was studied. The value of the parameter κ, which determines the so-called freeze-out volume in the Fermi break-up stage of the nuclear interaction process, was adjusted in order to achieve better agreement with experimental data. In this paper, results are shown both with the default value κ = 1 and the modified value κ = 10 which resulted in the best overall agreement. Comparisons with published experimental data were made in terms of total and partial charge-changing cross-sections generated by the MSDM, as well as integral and differential fragment yields simulated by SHIELD-HIT in intermediate and thick water targets irradiated with a beam of 400 MeV u(-1) (12)C ions. Better agreement with the experimental data was in general obtained with the modified parameter value (κ = 10), both on the level of partial charge-changing cross-sections and fragment yields.