F. Tixier

Characterization of PET/CT images using texture analysis: the past, the present… any future?

After seminal papers over the period 2009xa0–xa02011, the use of texture analysis of PET/CT images for quantification of intratumour uptake heterogeneity has received increasing attention in the last 4xa0years. Results are difficult to compare due to the heterogeneity of studies and lack of standardization. There are also numerous challenges to address. In this review we provide critical insights into the recent development of texture analysis for quantifying the heterogeneity in PET/CT images, identify issues and challenges, and offer recommendations for the use of texture analysis in clinical research. Numerous potentially confounding issues have been identified, related to the complex workflow for the calculation of textural features, and the dependency of features on various factors such as acquisition, image reconstruction, preprocessing, functional volume segmentation, and methods of establishing and quantifying correspondences with genomic and clinical metrics of interest. A lack of understanding of what the features may represent in terms of the underlying pathophysiological processes and the variability of technical implementation practices makes comparing results in the literature challenging, if not impossible. Since progress as a field requires pooling results, there is an urgent need for standardization and recommendations/guidelines to enable the field to move forward. We provide a list of correct formulae for usual features and recommendations regarding implementation. Studies on larger cohorts with robust statistical analysis and machine learning approaches are promising directions to evaluate the potential of this approach.

Development of a nomogram combining clinical staging with (18)F-FDG PET/CT image features in non-small-cell lung cancer stage I-III.

Our goal was to develop a nomogram by exploiting intratumour heterogeneity on CT and PET images from routine (18)F-FDG PET/CT acquisitions to identify patients with the poorest prognosis. This retrospective study included 116 patients with NSCLC stage I, II or III and with staging (18)F-FDG PET/CT imaging. Primary tumour volumes were delineated using the FLAB algorithm and 3D Slicer™ on PET and CT images, respectively. PET and CT heterogeneities were quantified using texture analysis. The reproducibility of the CT features was assessed on a separate test-retest dataset. The stratification power of the PET/CT features was evaluated using the Kaplan-Meier method and the log-rank test. The best standard metric (functional volume) was combined with the least redundant and most prognostic PET/CT heterogeneity features to build the nomogram. PET entropy and CT zone percentage had the highest complementary values with clinical stage and functional volume. The nomogram improved stratification amongst patients with stage II and III disease, allowing identification of patients with the poorest prognosis (clinical stage III, large tumour volume, high PET heterogeneity and low CT heterogeneity). Intratumour heterogeneity quantified using textural features on both CT and PET images from routine staging (18)F-FDG PET/CT acquisitions can be used to create a nomogram with higher stratification power than staging alone.

Confrontation de la TEP/TDM au 18FDG initiale aux statuts p16 (INK4a) et HPV des cancers des VADS localement avancés

Objectifs La surexpression de la proteine p16 et le statut HPV sont reconnus comme facteurs de bon pronostic independants dans les cancers des voies aerodigestives superieures (VADS). Il a ete suggere que les parametres de la TEP/TDM au 18FDG initiale seraient egalement des facteurs pronostiques independants. Notre objectif etait d’etudier le lien entre les donnees de la virologie et les donnees de la TEP/TDM pretherapeutique dans les cancers des VADS localement avances, traites par radio-chimiotherapie. Methodes Quarante patients presentant un cancer des VADS avec des volumes tumoraux superieurs a 3 cm3 ont ete inclus prospectivement. Ils avaient tous beneficie d’une TEP/TDM initiale dont ont ete extraits le volume metabolique, l’intensite de fixation (SUV), l’activite globale et des parametres d’heterogeneite a l’echelle locale, regionale et globale ainsi que des parametres de forme. Ces parametres ont ete confrontes a l’analyse virologique des biopsies pretraitementxa0: recherche de l’expression de p16 en IHC, recherche de l’ADN HPV 16 et determination du statut HPV (p16xa0+xa0ADN HPV 16). Resultats Les patients p16+ presentaient des tumeurs plus hypermetaboliques avec des SUVmax (pxa0=xa00,028) et SUVmean (pxa0=xa00,02) plus eleves que les tumeurs p16−. Elles etaient par ailleurs plus heterogenes localement en TEP avec par exemple une correlation plus faible (pxa0=xa00,004). Enfin, les formes des lesions initiales etaient moins complexes chez les patients p16+ et HPV+, avec un D2Bmax moins eleve (pxa0=xa00,03). Conclusion Les cancers des VADS localement avances ont des caracteristiques differentes en TEP lorsque les tumeurs sont p16+. Le lien entre ces caracteristiques biologiques et le devenir des patients reste maintenant a etablir.

MO-DE-207B-11: Reliability of PET/CT Radiomics Features in Functional and Morphological Components of NSCLC Lesions: A Repeatability Analysis in a Prospective Multicenter Cohort

PURPOSEnThe goal of this study was to evaluate the repeatability of radiomics features (intensity, shape and heterogeneity) in both PET and low-dose CT components of test-retest FDG-PET/CT images in a prospective multicenter cohort of 74 NSCLC patients from ACRIN 6678 and a similar Merck trial.nnnMETHODSnSeventy-four patients with stage III-IV NCSLC were prospectively included. The primary tumor and up to 3 additional lesions per patient were analyzed. The Fuzzy Locally Adaptive Bayesian algorithm was used to automatically delineate metabolically active volume (MAV) in PET. The 3D SlicerTM software was exploited to delineate anatomical volumes (AV) in CT. Ten intensity first-order features, as well as 26 textural features and four 3D shape descriptors were calculated from tumour volumes in both modalities. The repeatability of each metric was assessed by Bland-Altman analysis.nnnRESULTSnOne hundred and five lesions (primary tumors and nodal or distant metastases) were delineated and characterized. The MAV and AV determination had a repeatability of -1.4±11.0% and -1.2±18.7% respectively. Several shape and heterogeneity features were found to be highly or moderately repeatable (e.g., sphericity, co-occurrence entropy or intensity size-zone matrix zone percentage), whereas others were confirmed as unreliable with much higher variability (more than twice that of the corresponding volume determination).nnnCONCLUSIONnOur results in this large multicenter cohort with more than 100 measurements confirm the PET findings in previous studies (with <30 lesions). In addition, our study is the first to explore the repeatability of radiomics features in the low-dose CT component of PET/CT acquisitions (previous studies considered dosimetry CT, CE-CT or CBCT). Several features were identified as reliable in both PET and CT components and could be used to build prognostic models. This work has received a French government support granted to the CominLabs excellence laboratory and managed by the National Research Agency in the Investing for the Future program under reference ANR-10-LABX-07-01, and support from the city of Brest.

Impact thérapeutique de la TEP/TDM à la 18fluorocholine dans la prise en charge de la récidive du cancer de la prostate. Étude rétrospective sur 89 patients

Objectifs L’objectif principal etait d’evaluer l’impact therapeutique de la TEP/TDM a la FCH dans la prise en charge des patients presentant une recidive biologique de cancer de la prostate. Les objectifs secondaires etaient d’evaluer les performances diagnostiques de l’examen et les facteurs clinico-biologiques predictifs du site de recidive. Materiels et methodes Cette etude retrospective sur 89xa0patients ayant beneficie d’une TEP/TDM a la FCH pour une suspicion de recidive biologique d’un adenocarcinome prostatique a ete realisee au CHU de Poitiers (octobre 2011–juin 2014). Les examens ont ete realises soit sur une camera Philips Gemini, soit une camera Siemens Biograph mCT. Les resultats de la TEP ont ete confrontes aux donnees biologiques (PSA et PSAdt) et histopathologiques initiales. Les decisions therapeutiques obtenues sans ou avec connaissance des resultats TEP ont ete determinees au moyen d’un algorithme therapeutique etabli a partir des dernieres recommandations en vigueur. Resultats Les PSA et PSAdt moyens etaient, respectivement, de 8,58xa0ng/mL et 12xa0moisxa0; 95,5xa0% des TEP etaient positives, dont 43,5xa0% recidives locales, 15,3xa0% recidives ganglionnaires regionales, 17,6xa0% recidives locoregionales et 23,5xa0% recidives a distance. Le PSA ( p xa0=xa00,004) et le PSAdt ( p xa0=xa00,0002) etaient correles au site de recidive. Seul le PSAdt etait un facteur predictif independant d’une recidive locale ( p xa0=xa00,0001) ou a distance ( p xa0=xa00,0006). Les sensibilites et specificites de la TEP pour la detection des lesions prostatiques, ganglionnaires et osseuses etaient, respectivement, de 100xa0% et 73,7xa0%, 95,6xa0% et 100xa0%, 69,2xa0% et 100xa0%. Un changement de l’attitude therapeutique a ete induit par la TEP chez 51,7xa0% des patients explores. Conclusions Cette etude confirme l’interet pratique de la TEP/TDM a la Fluorocholine dans la localisation du site de recidive du cancer de la prostate, avec un changement de la strategie therapeutique interessant pres d’un patient sur deux.