Isabelle Gardin

Comparative assessment of methods for estimating tumor volume and standardized uptake value in (18)F-FDG PET.

In 18F-FDG PET, tumors are often characterized by their metabolically active volume and standardized uptake value (SUV). However, many approaches have been proposed to estimate tumor volume and SUV from 18F-FDG PET images, none of them being widely agreed upon. We assessed the accuracy and robustness of 5 methods for tumor volume estimates and of 10 methods for SUV estimates in a large variety of configurations. Methods: PET acquisitions of an anthropomorphic phantom containing 17 spheres (volumes between 0.43 and 97 mL, sphere-to-surrounding-activity concentration ratios between 2 and 68) were used. Forty-one nonspheric tumors (volumes between 0.6 and 92 mL, SUV of 2, 4, and 8) were also simulated and inserted in a real patient 18F-FDG PET scan. Four threshold-based methods (including one, Tbgd, accounting for background activity) and a model-based method (Fit) described in the literature were used for tumor volume measurements. The mean SUV in the resulting volumes were calculated, without and with partial-volume effect (PVE) correction, as well as the maximum SUV (SUVmax). The parameters involved in the tumor segmentation and SUV estimation methods were optimized using 3 approaches, corresponding to getting the best of each method or testing each method in more realistic situations in which the parameters cannot be perfectly optimized. Results: In the phantom and simulated data, the Tbgd and Fit methods yielded the most accurate volume estimates, with mean errors of 2% ± 11% and −8% ± 21% in the most realistic situations. Considering the simulated data, all SUV not corrected for PVE had a mean bias between −31% and −46%, much larger than the bias observed with SUVmax (−11% ± 23%) or with the PVE-corrected SUV based on Tbgd and Fit (−2% ± 10% and 3% ± 24%). Conclusion: The method used to estimate tumor volume and SUV greatly affects the reliability of the estimates. The Tbgd and Fit methods yielded low errors in volume estimates in a broad range of situations. The PVE-corrected SUV based on Tbgd and Fit were more accurate and reproducible than SUVmax.

Simultaneous positron emission tomography (PET) assessment of metabolism with 18F-fluoro-2-deoxy-d-glucose (FDG), proliferation with 18F-fluoro-thymidine (FLT), and hypoxia with 18fluoro-misonidazole (F-miso) before and during radiotherapy in patients with non-small-cell lung cancer (NSCLC): A pilot study

OBJECTIVES To investigate the changes in tumour proliferation (using FLT), metabolism (using FDG), and hypoxia (using F-miso) during curative (chemo-) radiotherapy (RT) in patients with non-small-cell lung cancer (NSCLC). PATIENTS AND METHODS Thirty PET scans were performed in five patients (4 males, 1 female) that had histological proof of NSCLC and were candidates for curative-intent RT. Three PET-CT (Biograph S16, Siemens) scans were performed before (t(0)) and during (around dose 46 Gy, t(46)) RT with minimal intervals of 48 h between each PET-CT scan. The tracers used were (18)fluoro-2deoxyglucose (FDG) for metabolism, (18)fluorothymidine (FLT) for proliferation, and (18)F-misonidasole (F-miso) for hypoxia. The 3 image sets obtained at each time point were co-registered (rigid: n=9, elastic: n=1, Leonardo, TrueD, Siemens) using FDG PET-CT as reference. VOIs were delineated (40% SUV(max) values were used as a threshold) for tumours and lymph nodes on FDG PET-CT, and they were automatically pasted on FLT and F-miso PET-CT images. ANOVA and correlation analyses were used for comparison of SUV(max) values. RESULTS Four tumours and twelve nodes were identified on initial FDG PET-CT images. FLT SUV(max) values were significantly lower (p<0.0006) at t(46) in both tumours and nodes. The decrease in FDG SUV(max) values had a trend towards significance (p=0.048). F-Miso SUV(max) values were significantly higher in tumours than in nodes (p=0.02) and did not change during radiotherapy (p=0.39). A significant correlation was observed between FLT and FDG uptake (r=0.56, p<10(-4)) when all data were pooled together, and they remained similar when the before and during RT data were analysed separately. FDG and F-miso uptakes were significantly correlated (r=0.59, p=0.0004) when all data were analysed together. The best fit was obtained after adjusting for lesion type (tumour vs. node). This correlation was observed for the SUV(max) measured during RT (r=0.70, p=0.008) but not for the pre-RT data (r=0.19, p=0.35). The weak correlation between FLT and F-miso uptakes only became significant (r=0.66, p=0.002) when the analysis was restricted to the data acquired during RT. CONCLUSION Three different PET acquisitions can be performed quasi-simultaneously (4-7 days) before and during radiotherapy in patients with NSCLC. Our results at 46 Gy suggest that a fast decrease in the proliferation of both tumours and nodes exists during radiotherapy with differences in metabolism (borderline significant decrease) and hypoxia (stable).

Development of a generic thresholding algorithm for the delineation of 18FDG-PET-positive tissue: application to the comparison of three thresholding models

An iterative generic algorithm has been developed to compare three thresholding models used to delineate gross tumour volume on (18)F-FDG PET images. 3D volume was extracted and characteristic parameters were measured. Three fitting models using different parameters were studied: model 1 (volume, contrast), model 2 (contrast) and model 3 (SUV). The calibration was performed using a cylindrical phantom filled with hot spheres. To validate the models, two other phantoms were used. The calibration procedure showed a better fitting model for model 1 (R(2) from 0.94 to 1.00) than for model 3 (0.95) and model 2 (0.69). The validation study shows that model 3 yielded large volume measurement errors. Models 1 and 2 gave close results with no significant differences. Model 2 was preferred because it presents less error dispersion and needs fewer characteristic parameters, making it easier to implement. Our results show the importance of developing a generic algorithm to compare the performances of fitting models objectively and to validate results on other phantoms than the ones used during the calibration process to avoid methodological biases.

Validation of the Monte Carlo simulator GATE for indium-111 imaging

Monte Carlo simulations are useful for optimizing and assessing single photon emission computed tomography (SPECT) protocols, especially when aiming at measuring quantitative parameters from SPECT images. Before Monte Carlo simulated data can be trusted, the simulation model must be validated. The purpose of this work was to validate the use of GATE, a new Monte Carlo simulation platform based on GEANT4, for modelling indium-111 SPECT data, the quantification of which is of foremost importance for dosimetric studies. To that end, acquisitions of (111)In line sources in air and in water and of a cylindrical phantom were performed, together with the corresponding simulations. The simulation model included Monte Carlo modelling of the camera collimator and of a back-compartment accounting for photomultiplier tubes and associated electronics. Energy spectra, spatial resolution, sensitivity values, images and count profiles obtained for experimental and simulated data were compared. An excellent agreement was found between experimental and simulated energy spectra. For source-to-collimator distances varying from 0 to 20 cm, simulated and experimental spatial resolution differed by less than 2% in air, while the simulated sensitivity values were within 4% of the experimental values. The simulation of the cylindrical phantom closely reproduced the experimental data. These results suggest that GATE enables accurate simulation of (111)In SPECT acquisitions.

Evaluation of a Rigid Registration Method of Lung Perfusion SPECT and Thoracic CT

OBJECTIVE The objective of our study was to evaluate a rigid registration method in lung perfusion SPECT using thoracic CT as a standard. MATERIALS AND METHODS The reproducibility of markers selection and the robustness of the method were assessed on a torso phantom. The accuracy of registration regarding the number and location of markers and the breathing state during CT was evaluated on eight patients using 10 external markers placed around the thorax before SPECT and CT acquisitions. The accuracy of registration was assessed using the mean errors (ME) between 10 markers after registration. RESULTS Registration using external markers on a phantom was accurate (ME, < 3 mm) when rotation was less than 40 degrees (p = 0.02). The accuracy of registration improved markedly from four to six markers for phantom (5.5-3.6 mm) and patients (11.2-9.5 mm) and then remained constant up to 10 markers. The ME was less when using markers that well encompassed the thorax for phantom and patients (p = 0.02 and p = 0.05, respectively). The use of four anatomic markers was not accurate (ME, 20 mm). CONCLUSION The registration method is reproducible and accurate, and six external markers were required to get an ME of less than 10 mm in patients.

Clinical Feasibility of Fast 3-Dimensional Dosimetry of the Liver for Treatment Planning of Hepatocellular Carcinoma with 90Y-Microspheres

Several treatment strategies are used for selective internal radiation therapy with 90Y-microspheres. The diversity of approaches does not favor the standardization of the prescribed activity calculation. To this aim, a fast 3-dimensional (3D) dosimetry method was developed for 90Y-microsphere treatment planning and was clinically evaluated retrospectively. Methods: Our 3D approach is based on voxel S values (VSVs) and has been implemented in the software tool VoxelDose. VSVs were previously calculated at a fine voxel size. The time-integrated activity (TIA) map is derived from pretherapeutic 99mTc-macroaggregated-albumin SPECT/CT. The fine VSV map is resampled at the voxel size of the TIA map. Then, the TIA map is convolved with the resampled VSV map to construct the 3D dose map. Data for 10 patients with 12 tumor sites treated by 90Y-microspheres for hepatocellular carcinoma were collected retrospectively. 3D dose maps were computed for each patient, and tumoral liver and nontumoral liver (TL and NTL, respectively) were delineated, allowing the computation of descriptive statistics (i.e., mean absorbed dose, minimum absorbed dose, and maximum absorbed dose) and dose–volume histograms. Mean absorbed doses in TL and NTL from VoxelDose were compared with those calculated with the standard partition model. Results: The estimated processing time for a complete 3D dosimetry calculation is on the order of 15 min, including 10 s for the dose calculation (i.e., VSV resampling and convolution). An additional 45 min was needed for the semiautomatic and manual segmentation of TL and NTL. The mean absorbed dose (±SD) was 422 ± 263 Gy for TL and 50.1 ± 36.0 Gy for NTL. The comparison between VoxelDose and partition model shows a mean relative difference of 1.5% for TL and 4.4% for NTL. Results show a wide spread of voxel-dose values around mean absorbed dose. The minimum absorbed dose within TL ranges from 32 to 267 Gy (n = 12). The fraction of NTL volume irradiated with at least 80 Gy ranges from 4% to 70% (n = 10), and the absorbed dose from which 25% of NTL was the least irradiated ranges from 14 to 178 Gy. Conclusion: This article demonstrates the feasibility of a fast 3D dosimetry method for 90Y-microspheres and highlights the potential value of a 3D treatment planning strategy.

Fine-Resolution Voxel S Values for Constructing Absorbed Dose Distributions at Variable Voxel Size

This article presents a revised voxel S values (VSVs) approach for dosimetry in targeted radiotherapy, allowing dose calculation for any voxel size and shape of a given SPECT or PET dataset. This approach represents an update to the methodology presented in MIRD pamphlet no. 17. Methods: VSVs were generated in soft tissue with a fine spatial sampling using the Monte Carlo (MC) code MCNPX for particle emissions of 9 radionuclides: 18F, 90Y, 99mTc, 111In, 123I, 131I, 177Lu, 186Re, and 201Tl. A specific resampling algorithm was developed to compute VSVs for desired voxel dimensions. The dose calculation was performed by convolution via a fast Hartley transform. The fine VSVs were calculated for cubic voxels of 0.5 mm for electrons and 1.0 mm for photons. Validation studies were done for 90Y and 131I VSV sets by comparing the revised VSV approach to direct MC simulations. The first comparison included 20 spheres with different voxel sizes (3.8–7.7 mm) and radii (4–64 voxels) and the second comparison a hepatic tumor with cubic voxels of 3.8 mm. MC simulations were done with MCNPX for both. The third comparison was performed on 2 clinical patients with the 3D-RD (3-Dimensional Radiobiologic Dosimetry) software using the EGSnrc (Electron Gamma Shower National Research Council Canada)-based MC implementation, assuming a homogeneous tissue-density distribution. Results: For the sphere model study, the mean relative difference in the average absorbed dose was 0.20% ± 0.41% for 90Y and −0.36% ± 0.51% for 131I (n = 20). For the hepatic tumor, the difference in the average absorbed dose to tumor was 0.33% for 90Y and −0.61% for 131I and the difference in average absorbed dose to the liver was 0.25% for 90Y and −1.35% for 131I. The comparison with the 3D-RD software showed an average voxel-to-voxel dose ratio between 0.991 and 0.996. The calculation time was below 10 s with the VSV approach and 50 and 15 h with 3D-RD for the 2 clinical patients. Conclusion: This new VSV approach enables the calculation of absorbed dose based on a SPECT or PET cumulated activity map, with good agreement with direct MC methods, in a faster and more clinically compatible manner.

Areas of High 18F-FDG Uptake on Preradiotherapy PET/CT Identify Preferential Sites of Local Relapse After Chemoradiotherapy for Non–Small Cell Lung Cancer

The high rates of failure in the radiotherapy target volume suggest that patients with stage II or III non–small cell lung cancer (NSCLC) should receive an increased total dose of radiotherapy. Areas of high 18F-FDG uptake on preradiotherapy 18F-FDG PET/CT have been reported to identify intratumor subvolumes at high risk of relapse after radiotherapy. We wanted to confirm these observations on a cohort of patients included in 3 sequential prospective studies. Our aim was to assess an appropriate threshold (percentage of maximum standardized uptake value [SUVmax]) to delineate subvolumes on staging 18F-FDG PET/CT scans assuming that a smaller target volume would facilitate isotoxic radiotherapy dose escalation. Methods: Thirty-nine patients with inoperable stage II or III NSCLC, treated with chemoradiation or with radiotherapy alone, were extracted from 3 prospective studies (ClinicalTrials.gov identifiers NCT01261585, NCT01261598, and RECF0645). All patients underwent 18F-FDG PET/CT at initial staging, before radiotherapy, during radiotherapy, and during systematic follow-up in a single institution. All 18F-FDG PET/CT acquisitions were coregistered on the initial scan. Various subvolumes in the initial acquisition (30%, 40%, 50%, 60%, 70%, 80%, and 90% SUVmax thresholds) and in the 3 subsequent acquisitions (40% and 90% SUVmax thresholds) were pasted on the initial scan and compared. Results: Seventeen patients had a local relapse. The SUVmax measured during radiotherapy was significantly higher in locally relapsed tumors than in locally controlled tumors (mean, 6.8 vs. 4.6; P = 0.02). The subvolumes delineated on initial PET/CT scans with 70%–90% SUVmax thresholds were in good agreement with the recurrent volume at a 40% SUVmax threshold (common volume/baseline volume, 0.60–0.80). The subvolumes delineated on initial PET/CT scans with 30%–60% SUVmax thresholds were in good to excellent agreement with the core volume of the relapse (90% SUVmax threshold) (common volume/recurrent volume and overlap fraction indices, 0.60–0.93). The agreement was moderate (>0.51) when a 70% SUVmax threshold was used to delineate on initial PET/CT scans. Conclusion: High 18F-FDG uptake areas on pretreatment PET/CT scans identify tumor subvolumes at greater risk of relapse in patients with NSCLC treated by concomitant chemoradiation. We propose a 70% SUVmax threshold to delineate areas of high 18F-FDG uptake on initial PET/CT scans as the target volumes for potential radiotherapy dose escalation.

Fusion of multi-tracer PET images for Dose Painting

PET imaging with FluoroDesoxyGlucose (FDG) tracer is clinically used for the definition of Biological Target Volumes (BTVs) for radiotherapy. Recently, new tracers, such as FLuoroThymidine (FLT) or FluoroMisonidazol (FMiso), have been proposed. They provide complementary information for the definition of BTVs. Our work is to fuse multi-tracer PET images to obtain a good BTV definition and to help the radiation oncologist in dose painting. Due to the noise and the partial volume effect leading, respectively, to the presence of uncertainty and imprecision in PET images, the segmentation and the fusion of PET images is difficult. In this paper, a framework based on Belief Function Theory (BFT) is proposed for the segmentation of BTV from multi-tracer PET images. The first step is based on an extension of the Evidential C-Means (ECM) algorithm, taking advantage of neighboring voxels for dealing with uncertainty and imprecision in each mono-tracer PET image. Then, imprecision and uncertainty are, respectively, reduced using prior knowledge related to defects in the acquisition system and neighborhood information. Finally, a multi-tracer PET image fusion is performed. The results are represented by a set of parametric maps that provide important information for dose painting. The performances are evaluated on PET phantoms and patient data with lung cancer. Quantitative results show good performance of our method compared with other methods.

Comparison Between 2D and 3D Dosimetry Protocols in 90Y-Ibritumomab Tiuxetan Radioimmunotherapy of Patients with Non-Hodgkin's Lymphoma

UNLABELLED We compared the radiation-absorbed dose obtained from a two dimensional (2D) protocol, based on planar whole-body (WB) scans and fixed reference organ masses with dose estimates, using a 3D single-photon emission computed tomography (SPECT) imaging protocol and patient-specific organ masses. METHODS Six (6) patients with follicular non-Hodgkins lymphoma underwent a computed tomography (CT) scan, 5 2D planar WB, and 5 SPECT scans between days 0 and 6 after the injection of (111)In-ibritumomab tiuxetan. The activity values in the liver, spleen, and kidneys were calculated from the 2D WB scans, and also from the 3D SPECT images reconstructed, using quantitative image processing. Absorbed doses after the administration of (90)Y-ibritumomab tiuxetan were calculated from the (111)In WB activity values combined with reference organ masses and also from the SPECT activity values and organ masses as estimated from the patient CT scan. To assess the quantitative accuracy of the WB and SPECT scans, an abdominal phantom was also studied. RESULTS The differences between organ masses estimated from the patient CT and from the reference MIRD models were between -10% and +98%. Using the phantom, errors in organ and tumor activity estimates were between -86% and 10% for the WB protocol and between -43% and -6% for the SPECT protocol. Patient liver, spleen, and kidney activity values determined from SPECT were systematically less than those from the WB scans. Radiation-absorbed doses calculated with the 3D protocol were systematically lower than those calculated from the WB protocol (29%+/-26%, 73%+/-26%, and 33%+/-53% differences in the liver, spleen, and kidney, respectively), except in the kidneys of 2 patients and in the liver of 1 patient. CONCLUSIONS Accounting for patient-specific organ mass and using SPECT activity quantification have both a great impact on estimated absorbed doses.