Prediction of Histologic Neoadjuvant Chemotherapy Response in Osteosarcoma Using Pretherapeutic MRI Radiomics.

Histologic response to chemotherapy for osteosarcoma is one of the most important prognostic factors for survival, but assessment occurs after surgery. Although tumor imaging is used for surgical planning and follow-up, it lacks predictive value. Therefore, a radiomics model was developed to predict the response to neoadjuvant chemotherapy based on pretreatment T1-weighted contrast-enhanced MRI. A total of 176 patients (median age, 20 years [range, 5-71 years]; 107 male patients) with osteosarcoma treated with neoadjuvant chemotherapy and surgery between January 2007 and December 2018 in three different centers in France (Centre Léon Bérard in Lyon, Centre Hospitalier Universitaire de Nantes in Nantes, and Hôpital Cochin in Paris) were retrospectively analyzed. Various models were trained from different configurations of the data sets. Two different methods of feature selection were tested with and without ComBat harmonization (ReliefF and t test) to select the most relevant features, and two different classifiers were used to build the models (an artificial neural network and a support vector machine). Sixteen radiomics models were built using the different combinations of feature selection and classifier applied on the various data sets. The most predictive model had an area under the receiver operating characteristic curve of 0.95, a sensitivity of 91%, and a specificity 92% in the training set; respective values in the validation set were 0.97, 91%, and 92%. In conclusion, MRI-based radiomics may be useful to stratify patients receiving neoadjuvant chemotherapy for osteosarcomas. Keywords: MRI, Skeletal-Axial, Oncology, Radiomics, Osteosarcoma, Pediatrics Supplemental material is available for this article. © RSNA, 2022.

Prediction of lipomatous soft tissue malignancy on MRI: comparison between machine learning applied to radiomics and deep learning.

OBJECTIVES: Malignancy of lipomatous soft-tissue tumours diagnosis is suspected on magnetic resonance imaging (MRI) and requires a biopsy. The aim of this study is to compare the performances of MRI radiomic machine learning (ML) analysis with deep learning (DL) to predict malignancy in patients with lipomas oratypical lipomatous tumours. METHODS: Cohort include 145 patients affected by lipomatous soft tissue tumours with histology and fat-suppressed gadolinium contrast-enhanced T1-weighted MRI pulse sequence. Images were collected between 2010 and 2019 over 78 centres with non-uniform protocols (three different magnetic field strengths (1.0, 1.5 and 3.0 T) on 16 MR systems commercialised by four vendors (General Electric, Siemens, Philips, Toshiba)). Two approaches have been compared: (i) ML from radiomic features with and without batch correction; and (ii) DL from images. Performances were assessed using 10 cross-validation folds from a test set and next in external validation data. RESULTS: The best DL model was obtained using ResNet50 (resulting into an area under the curve (AUC) of 0.87 ± 0.11 (95% CI 0.65-1). For ML/radiomics, performances reached AUCs equal to 0.83 ± 0.12 (95% CI 0.59-1) and 0.99 ± 0.02 (95% CI 0.95-1) on test cohort using gradient boosting without and with batch effect correction, respectively. On the external cohort, the AUC of the gradient boosting model was equal to 0.80 and for an optimised decision threshold sensitivity and specificity were equal to 100% and 32% respectively. CONCLUSIONS: In this context of limited observations, batch-effect corrected ML/radiomics approaches outperformed DL-based models.

CT diagnosis of closed loop bowel obstruction mechanism is not sufficient to indicate emergent surgery.

PURPOSE: To identify computed tomography (CT) findings associated with successful conservative treatment of closed loop small bowel obstruction (CL-SBO) due to adhesions or internal herniation. MATERIAL AND METHODS: The local institutional review board approved this study while waiving informed consent. Clinical and CT data were collected retrospectively for 96 consecutive patients with a CT diagnosis of CL-SBO due to adhesions or internal herniation established by experienced radiologists who had no role in patient management. Mechanical obstruction with at least two transition zones on the bowel at a single site defined CL-SBO. Two radiologists blinded to patient data independently performed a retrospective review of the CT scans. The patient groups with successful versus failed initial conservative therapy were compared. Univariate and multivariate analyses were performed to look for CT findings associated with successful conservative therapy. Interobserver agreement was assessed for each CT finding. RESULTS: Of the 96 patients, 34 (35%) underwent immediate surgery and 62 (65%) received first-line conservative treatment, which succeeded in 19 (31%) and failed in 43 (69%). The distance between the transition zones was the only independent predictor of successful conservative therapy (odds ratio, 4.6 when ≥ 8 mm; 95% confidence interval [95% CI], 1.2-18.3). A distance ≥ 8 mm had 84% (95% CI, 60-97) sensitivity and 46% (95% CI, 31-62) specificity for successful conservative treatment. The correlation coefficient for the distance between transition zones between readers 1 and 2 was fair (r = 0.46). CONCLUSION: CL-SBO can be resolved without surgery. When there is no CT sign of ischemia, the distance between the transition zones should be assessed. KEY POINTS: • Twenty percent (19/96) of all cases of closed loop small bowel obstruction (CL-SBO) representing 31% of the patients given first-line conservative therapy, were resolved without surgery. • The distance between the transition zones may help to choose between conservative and surgical management in patients with a CL-SBO but no CT evidence of ischemia. • A distance < 8 mm between the transition zones suggests a need for emergent surgery.

In vitro distinction between proinflammatory and antiinflammatory macrophages with gadolinium-liposomes and ultrasmall superparamagnetic iron oxide particles at 3.0T.

BACKGROUND: Inflammation involves a heterogeneous macrophage population, for which there is no readily available MR assessment method. PURPOSE: To assess the feasibility of distinguishing proinflammatory M1 and antiinflammatory M2 macrophages at MRI enhanced with gadolinium liposomes or ultrasmall superparamagnetic iron oxide particles. STUDY TYPE: In vitro. SPECIMEN: We employed cultured RAW macrophages. M0 macrophages were polarized with lipopolysaccharide (LPS) or interleukin-4 (IL-4), resulting in M1 or M2 macrophages. The macrophages were incubated with gadolinium (±rhodamine) liposomes or iron oxide particles and cell pellets were prepared for MRI. FIELD STRENGTH/SEQUENCE: Transverse relaxation rates and quantitative susceptibility were obtained at 3.0T with multiecho turbo spin echo and spoiled gradient echo sequences. ASSESSMENT: MRI results were compared with confocal microscopy, flow cytometry, and expression of endocytosis, M1 and M2 genes. STATISTICAL TESTS: Mann-Whitney and Kruskal-Wallis tests were performed. RESULTS: Higher transverse relaxation rates and susceptibility were observed in M1 than in M2 and M0 macrophages (P < 0.01 both with liposomes and USPIO) and significantly different susceptibility in M2 and M0 macrophages (P < 0.01 both with liposomes and USPIO). These MRI results were confirmed at confocal microscopy and flow cytometry. LPS macrophages displayed M1 gene expression, whereas IL-4 macrophages showed M2 polarization and lower endocytosis gene expression rates. DATA CONCLUSION: These in vitro results show that it is feasible to distinguish between proinflammatory M1 and antiinflammatory M2 macrophages according to their level of contrast agent uptake at MRI. LEVEL OF EVIDENCE: 1 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;49:1166-1173.

Clinical Relevance of the Feces Sign in Small-Bowel Obstruction Due to Adhesions Depends on Its Location.

OBJECTIVE: The objective of our study was to evaluate if the feces sign can be used to predict successful nonoperative treatment or progression to ischemia in patients with small-bowel obstruction (SBO) due to adhesions. MATERIALS AND METHODS: For this single-center retrospective observational study involving a blinded independent review by two radiologists of 237 consecutive CT examinations of 216 patients with SBO due to adhesions (age: mean, 70.9 years; median, 74 years; interquartile range, 62-84 years), the location of the transition zone (TZ), number of TZs, and presence and location of the feces sign relative to the TZ were recorded. The reference standard for diagnosing ischemia was surgical and pathologic findings (n = 108 CT examinations) or, when treatment was nonoperative (n = 129 CT examinations), clinical outcome. Factors associated with successful nonoperative treatment and ischemia were identified by univariate and multivariate analyses. RESULTS: A feces sign was seen in 88 of 237 CT examinations (37.1%). The feces sign was at the TZ, which we refer to as the "TZ feces" sign, in 82 of 88 (93.2%) patients; between two TZs, which we refer to as the "trapped feces" sign, in 14 (15.9%) patients; and in both locations in eight (9.1%) patients. By univariate analysis, an isolated TZ feces sign was associated positively with successful nonoperative treatment (odds ratio [OR], 3.37; 95% CI, 1.71-6.66; p < 0.001) and negatively with ischemia (OR, 0.33; 95% CI, 0.13-0.85; p = 0.02). Combined TZ and trapped feces signs were associated with ischemia (OR, 24.16; 95% CI, 2.86-203.89; p = 0.003). By multivariate analysis, regardless of the location of the feces sign, the feces sign was not significantly associated with successful nonoperative treatment or progression to ischemia. CONCLUSION: The feces sign is common and helps to identify the TZ. Among the CT signs of SBO, the feces sign does not independently help to predict successful nonoperative treatment or progression to ischemia.

Variability and errors when applying the BIRADS mammography classification.

To standardize mammographic reporting, the American College of Radiology mammography developed the Breast Imaging Reporting and Data System (BIRADS) lexicon. However, wide variability is observed in practice in the application of the BIRADS terminology and this leads to classification errors. This review analyses the reasons for variations in BIRADS mammography, describes the types of errors made by readers with illustrated examples, and details BIRADS category 3 which is the most difficult category to use in practice.

[Non oncologic applications of molecular targeted therapies]. / Applications thérapeutiques non oncologiques des thérapies moléculaires ciblées.

Significant improvements in the knowledge of cancer biology have permitted the development of new molecular targeted therapies. Meanwhile, a better understanding of the physiology of various non-cancerous diseases has allowed developing these agents in other areas. This review intends to illustrate these perspectives through examples corresponding to different strategies of molecular-targeted therapies : use of a monoclonal antibody binding a receptor (rituximab and rheumatoid arthritis) or a ligand (bevacizumab and age-related macular degeneration), tyrosine kinase inhibitor (imatinib and systemic sclerosis) or inhibitor of cytoplasmic signal transduction pathways (immunosuppressive and antiproliferative effects of mammalian target of rapamycin [mTOR] inhibitors). Clinical results can draw today what could become molecular medicine of tomorrow.