Marc De Bast

Gradient-based delineation of the primary GTV on FDG-PET in non-small cell lung cancer: A comparison with threshold-based approaches, CT and surgical specimens.

PURPOSE The aim of this study was to validate a gradient-based segmentation method for GTV delineation on FDG-PET in NSCLC through surgical specimen, in comparison with threshold-based approaches and CT. MATERIALS AND METHODS Ten patients with stage I-II NSCLC were prospectively enrolled. Before lobectomy, all patients underwent contrast enhanced CT and gated FDG-PET. Next, the surgical specimen was removed, inflated with gelatin, frozen and sliced. The digitized slices were used to reconstruct the 3D macroscopic specimen. GTVs were manually delineated on the macroscopic specimen and on CT images. GTVs were automatically segmented on PET images using a gradient-based method, a source to background ratio method and fixed threshold values at 40% and 50% of SUV(max). All images were finally registered. Analyses of raw volumes and logarithmic differences between GTVs and GTV(macro) were performed on all patients and on a subgroup excluding the poorly defined tumors. A matching analysis between the different GTVs was also conducted using Dices similarity index. RESULTS Considering all patients, both lung and mediastinal windowed CT overestimated the macroscopy, while FDG-PET provided closer values. Among various PET segmentation methods, the gradient-based technique best estimated the true tumor volume. When analysis was restricted to well defined tumors without lung fibrosis or atelectasis, the mediastinal windowed CT accurately assessed the macroscopic specimen. Finally, the matching analysis did not reveal significant difference between the different imaging modalities. CONCLUSIONS FDG-PET improved the GTV definition in NSCLC including when the primary tumor was surrounded by modifications of the lung parenchyma. In this context, the gradient-based method outperformed the threshold-based ones in terms of accuracy and robustness. In other cases, the conventional mediastinal windowed CT remained appropriate.

The limitation of PET imaging for biological adaptive-IMRT assessed in animal models

PURPOSE Biological image-guided radiotherapy aims at specifically irradiating biologically relevant sub-volumes within the tumor, as determined for instance by PET imaging. This approach requires that PET imaging be sensitive and specific enough to image various biological pathways of interest, e.g. tumor metabolism, proliferation and hypoxia. In this framework, a validation of PET imaging used for adaptive radiotherapy was undertaken in animal models by comparing small-animal PET images (2.7mm resolution) with autoradiography (AR) (100mum resolution) in various tumors under various physiological situations. METHODS A specific template for tumor-bearing mouse imaging has been designed (Christian, R&O, 2008). It allows for the registration between MRI images (Biospec, Bruker), FDG-PET images (Mosaic, Philips) and AR (FLA-5100, Fujifilm). After registration, the tumors on the PET and AR images were segmented using a threshold-based method. The thresholds were selected to obtain absolute equal volumes in the PET and AR images. Matching indexes were then calculated between the various volumes. The entire imaging process was performed for FSAII tumors (n=5), SCCVII (n=5) and irradiated (35Gy) FSAII tumors (n=5). RESULTS In regions with high FDG activity delineated using high value thresholds, low matching values of 39%+/-11% (mean+/-SD) were observed between the volumes delineated on the PET images and those delineated on AR. The matching values progressively increased when considering larger volumes obtained with lower thresholds. These findings were independent of tumor type, tumor metabolism or tumor size. The relationship between the matching values and the percentage of overall tumor volume was fitted through a power regression (r=0.93). As shown by simulations, the matching improved with higher PET resolution. The results can be extrapolated to human tumors imaged with a whole-body PET system. CONCLUSION Discrepancies were found between the PET images and the underlying microscopic reality represented by AR images. These differences, attributed to the finite resolution of PET, were important when considering small and highly active regions of the tumors. Dose painting based on PET images should therefore be carefully considered and should take these limitations into account.

The effect of 2'-2' difluorodeoxycytidine (dFdC, gemcitabine) on radiation-induced cell lethality in two human head and neck squamous carcinoma cell lines differing in intrinsic radiosensitivity.

PURPOSE The present study investigated in vitro radio-enhancement by gemcitabine (dFdC) in two head and neck squamous cell carcinomas with different intrinsic cellular radiosensitivity. MATERIALS AND METHODS Radiosensitive (SCC61, SF2=0.16) and radioresistant (SQD9, SF2=0.49) human head and neck squamous cell carcinomas were used. Confluent cells were incubated with dFdC and irradiated in drug-free medium with a single dose of 250 kV X-rays (0-12Gy). Cell survival curves were corrected for the toxicity of the drug alone. RESULTS In both cell lines, radio-enhancement was observed with 5 microM dFdC incubated for 3 h prior to irradiation. Dose modification factors (DMF) at a surviving fraction level of 0.5 reached 1.3 (95% CI 1.1-1.6) and 1.5 (95% CI 1.4-1.5) for SQD9 and SCC61 cells, respectively. Radio-enhancement was associated with a modest increase in the alpha term of the linear-quadratic model. In SQD9 cells, radio-enhancement increased with dFdC incubation time. At 24h, DMF reached a value of 1.5 (95% CI 0.9-3.2). In SCC61 cells at 24h, DMF reached a value of 1.1 (95% CI 0.9-1.2). In both cell lines, radio-enhancement increased with dFdC concentration up to 5-10 microM from which values it levelled off up to 100 microM. CONCLUSIONS The data indicated that dFdC induced a modest radio-enhancement in both cell lines. For a short incubation time, dFdC did not radio-enhance preferentially the more radio-resistant cells, whereas the opposite was observed for a longer time. In both cell lines, radio-enhancement was saturated above a dFdC concentration of 5-10 microM.

Is 18F-FDG a surrogate tracer to measure tumor hypoxia? Comparison with the hypoxic tracer 14C-EF3 in animal tumor models

INTRODUCTION Fluorodeoxyglucose (FDG) has been reported as a surrogate tracer to measure tumor hypoxia with positron emission tomography (PET). The hypothesis is that there is an increased uptake of FDG under hypoxic conditions secondary to enhanced glycolysis, compensating the hypoxia-induced loss of cellular energy production. Several studies have already addressed this issue, some with conflicting results. This study aimed to compare the tracers (14)C-EF3 and (18)F-FDG to detect hypoxia in mouse tumor models. MATERIALS AND METHODS C3H, tumor-bearing mice (FSAII and SCCVII tumors) were injected iv with (14)C-EF3, and 1h later with (18)F-FDG. Using a specifically designed immobilization device with fiducial markers, PET (Mosaic®, Philips) images were acquired 1h after the FDG injection. After imaging, the device containing mouse was frozen, transversally sliced and imaged with autoradiography (AR) (FLA-5100, Fujifilm) to obtain high resolution images of the (18)F-FDG distribution within the tumor area. After a 48-h delay allowing for (18)F decay a second AR was performed to image (14)C-EF3 distribution. AR images were aligned to reconstruct the full 3D tumor volume, and were compared with the PET images. Image segmentation with threshold-based methods was applied on both AR and PET images to derive various tracer activity volumes. The matching index DSI (dice similarity index) was then computed. The comparison was performed under normoxic (ambient air, FSAII: n=4, SCCVII, n=5) and under hypoxic conditions (10% O(2) breathing, SCCVII: n=4). RESULTS On AR, under both ambient air and hypoxic conditions, there was a decreasing similarity between (14)C-EF3 and FDG with higher activity sub-volumes. Under normoxic conditions, when comparing the 10% of tumor voxels with the highest (18)F-FDG or (14)C-EF3 activity, a DSI of 0.24 and 0.20 was found for FSAII and SCCVII, respectively. Under hypoxic conditions, a DSI of 0.36 was observed for SCCVII tumors. When comparing the (14)C-EF3 distribution in AR with the corresponding (18)F-FDG-PET images, the DSI reached values of 0.26, 0.22 and 0.21 for FSAII and SCCVII under normoxia and SCCVII under hypoxia, respectively. CONCLUSION This study showed that FDG is not a good surrogate tracer for tumor hypoxia under either ambient or hypoxic conditions. Only specific hypoxia tracers should be used to measure tumor hypoxia.

In vivo colocalization of 2-nitroimidazole EF5 fluorescence intensity and electron paramagnetic resonance oximetry in mouse tumors.

BACKGROUND AND PURPOSE The primary objective of this study was to establish in vivo the relationship between 2-2-nitro-1H-imidazol-1yl-N-(2,2,3,3,3-pentafluoropropyl)-acetamide (EF5) adduct formation and intratumoral oxygen concentrations measured by electron paramagnetic resonance (EPR) in a tumor model mimicking a clinical situation. The secondary objective was an attempt to calibrate in situ the immunofluorescence (IF) signal with EPR oximetry. MATERIALS AND METHODS IM syngeneic fibrosarcoma (NFSA) bearing C3H mice were used. Three days after injection of a paramagnetic charcoal into the tumor, the mice were anesthetized, injected with the hypoxic marker EF5, and monitored every 20 min for 3 h with a low-frequency EPR spectrometer. Animals were allowed to breath either under 21 or 100% O(2). Tumors were then harvested, frozen, cut into sections including the charcoal and processed for EF5 adducts detection using monoclonal antibodies. Slices were viewed with a fluorescence microscope and 190x140 micrometer areas surrounding the charcoal were digitized and analyzed with the NIH-Image and Adobe Photoshop software. The fluorescence intensity (FI) was measured in the whole pictures and in strips of 10 micrometer around the charcoal. RESULTS EF5 binding increased with decreasing pO(2), most substantially at pO(2) below 5 mm Hg. Baseline (ambient air) pO(2) reached 3.2+/-2.1 mm Hg in NFSA tumors. It increased to 9.8+/-3.2 mm Hg under 100% O(2). A statistically significant correlation was observed on an individual tumor basis between the FI in the first 10 micrometer strip around the charcoal and the pO(2) determined by EPR oximetry (Wilcoxon signed rank test: P<0.001). CONCLUSIONS The present study confirms the intrinsic relationship between EF5 adduct binding and intratumoral pO(2) in an in vivo environment under biologically-relevant pO(2) values of less than 10 mm Hg.

Immobilization device for in vivo and in vitro multimodality image registration of rodent tumors

Biological image-guided radiotherapy requires that PET accurately identifies biologically relevant sub-volumes within a tumor. In this framework, an immobilization device was developed to study multi-imaging (CT, micro-MRI, micro-PET, and autoradiography) registration of mouse tumors. The registration accuracy assessed by calculating the average minimal distance between two skew lines was in the order of 0.2-0.3 mm.