A machine-learning-based combination of criteria to detect bladder cancer lymph node metastasis on [18F]FDG PET/CT: a pathology-controlled study.

OBJECTIVES: Initial pelvic lymph node (LN) staging is pivotal for treatment planification in patients with muscle-invasive bladder cancer (MIBC), but [18F]FDG PET/CT provides insufficient and variable diagnostic performance. We aimed to develop and validate a machine-learning-based combination of criteria on [18F]FDG PET/CT to accurately identify pelvic LN involvement in bladder cancer patients. METHODS: Consecutive patients with localized MIBC who performed preoperative [18F]FDG PET/CT between 2010 and 2017 were retrospectively assigned to training (n = 129) and validation (n = 44) sets. The reference standard was the pathological status after extended pelvic LN dissection. In the training set, a random forest algorithm identified the combination of criteria that best predicted LN status. The diagnostic performances (AUC) and interrater agreement of this combination of criteria were compared to a consensus of experts. RESULTS: The overall prevalence of pelvic LN involvement was 24% (n = 41/173). In the training set, the top 3 features were derived from pelvic LNs (SUVmax of the most intense LN, and product of diameters of the largest LN) and primary bladder tumor (product of diameters). In the validation set, diagnostic performance did not differ significantly between the combination of criteria (AUC = 0.59 95%CI [0.43-0.73]) and the consensus of experts (AUC = 0.64 95%CI [0.48-0.78], p = 0.54). The interrater agreement was equally good with Κ = 0.66 for both. CONCLUSION: The developed machine-learning-based combination of criteria performs as well as a consensus of experts to detect pelvic LN involvement on [18F]FDG PET/CT in patients with MIBC. KEY POINTS: • The developed machine-learning-based combination of criteria performs as well as experts to detect pelvic LN involvement on [18F]FDG PET/CT in patients with muscle-invasive bladder cancer. • The top 3 features to predict LN involvement were the SUVmax of the most intense LN, the product of diameters of the largest LN, and the product of diameters of the primary bladder tumor.

Integrated analysis of 18F-FDG PET/CT improves preoperative lymph node staging for patients with invasive bladder cancer.

OBJECTIVES: Preoperative 18F-fluoro-2-deoxy-D-glucose (18F-FDG) positron emission tomography/computed tomography (PET/CT) is controversial to assess lymph node (LN) staging in patients with invasive bladder cancer. We proposed to use the maximum standardized uptake value (SUVmax) associated with axial-based LN size to improve the detection of regional LN metastasis. METHODS: From May 2015 to May 2017, we prospectively included patients with urothelial bladder cancer who underwent radical cystectomy with extended pelvic LN dissection. All patients underwent preoperative 18F-FDG PET/CT staging before surgery. The gold standard comparator was the pathological examination of resected LNs. The data were reported on a regional per area- and patient-based model according to SUVmax values and axial-based LN size criteria. RESULTS: In total, 1012 LNs were identified in 61 patients with clinically localized invasive bladder cancer who underwent radical cystectomy and extended pelvic LN dissection. Loco-regional involvement of 24 LN areas was confirmed in 17 patients. In per area analysis, diagnostic accuracy of PET/CT and CT alone were respectively 84% and 78% (p = 0.039). On patient-based analysis, combined PET/CT correctly classified pelvic LN status in 5/61 (+ 8%) additional patients using optimal thresholds compared to CT alone, with accuracies of 82% and 74%, respectively (p = 0.13). CONCLUSION: Combining SUVmax and axial-based LN size criteria using 18F-FDG PET/CT improved the diagnostic accuracy for preoperative LN staging in patients with invasive bladder cancer, in per area analysis. KEY POINTS: • Combining metabolical and morphological features using18F-FDG PET/CT improves the detection of malignant lymph node in patients with bladder cancer. • 18 F-FDG PET/CT may help for initial staging of patients with muscle invasive bladder cancer.

Functional Alterations Due to COVID-19 Lung Lesions-Lessons From a Multicenter V/Q Scan-Based Registry.

PURPOSE: In coronavirus disease 2019 (COVID-19) patients, clinical manifestations as well as chest CT lesions are variable. Lung scintigraphy allows to assess and compare the regional distribution of ventilation and perfusion throughout the lungs. Our main objective was to describe ventilation and perfusion injury by type of chest CT lesions of COVID-19 infection using V/Q SPECT/CT imaging. PATIENTS AND METHODS: We explored a national registry including V/Q SPECT/CT performed during a proven acute SARS-CoV-2 infection. Chest CT findings of COVID-19 disease were classified in 3 elementary lesions: ground-glass opacities, crazy-paving (CP), and consolidation. For each type of chest CT lesions, a semiquantitative evaluation of ventilation and perfusion was visually performed using a 5-point scale score (0 = normal to 4 = absent function). RESULTS: V/Q SPECT/CT was performed in 145 patients recruited in 9 nuclear medicine departments. Parenchymal lesions were visible in 126 patients (86.9%). Ground-glass opacities were visible in 33 patients (22.8%) and were responsible for minimal perfusion impairment (perfusion score [mean ± SD], 0.9 ± 0.6) and moderate ventilation impairment (ventilation score, 1.7 ± 1); CP was visible in 43 patients (29.7%) and caused moderate perfusion impairment (2.1 ± 1.1) and moderate-to-severe ventilation impairment (2.5 ± 1.1); consolidation was visible in 89 patients (61.4%) and was associated with moderate perfusion impairment (2.1 ± 1) and severe ventilation impairment (3.0 ± 0.9). CONCLUSIONS: In COVID-19 patients assessed with V/Q SPECT/CT, a large proportion demonstrated parenchymal lung lesions on CT, responsible for ventilation and perfusion injury. COVID-19-related pulmonary lesions were, in order of frequency and functional impairment, consolidations, CP, and ground-glass opacity, with typically a reverse mismatched or matched pattern.

Lung Scintigraphy for Pulmonary Embolism Diagnosis in COVID-19 Patients: A Multicenter Study.

In patients with novel coronavirus disease 2019 (COVID-19) referred for lung scintigraphy because of suspected pulmonary embolism (PE), there has been an ongoing debate within the nuclear medicine community as to whether and when the ventilation imaging should be performed. Indeed, whereas PE diagnosis typically relies on the recognition of ventilation-perfusion (V/P) mismatched defects, the ventilation procedure potentially increases the risk of contamination to health-care workers. The primary aim of this study was to assess the role of ventilation imaging when lung scintigraphy is performed because of suspected PE in COVID-19 patients. The secondary aim was to describe practices and imaging findings in this specific population. Methods: A national registry was created in collaboration with the French Society of Nuclear Medicine to collect lung scans performed on COVID-19 patients for suspected PE. The practices of departments were assessed regarding imaging protocols and aerosol precautions. A retrospective review of V/P SPECT/CT scans was then conducted. Two physicians masked to clinical information reviewed each case by sequentially viewing perfusion SPECT, perfusion SPECT/CT, and V/P SPECT/CT images. The scans were classified into 1 of the 4 following categories: patients for whom PE could reasonably be excluded on the basis of perfusion SPECT only, perfusion SPECT/CT, or V/P SPECT/CT and patients with mismatched defects suggestive of PE according to the European Association of Nuclear Medicine criteria. Results: Data from 12 French nuclear medicine departments were collected. Lung scans were performed between March 2020 and April 2021. Personal protective equipment and dedicated cleaning procedures were used in all departments. Of the 145 V/Q SPECT/CT scans included in the central review, PE could be excluded using only perfusion SPECT, perfusion SPECT/CT, or V/P SPECT/CT in 27 (19%), 55 (38%), and 45 (31%) patients, respectively. V/P SPECT/CT was positive for PE in 18 (12%) patients, including 12 (67%) with a low burden of PE (≤10%). Conclusion: In this population of COVID-19 patients assessed with lung scintigraphy, PE could confidently be excluded without the ventilation imaging in only 57% of patients. Ventilation imaging was required to confidently rule out PE in 31% of patients. Overall, the prevalence of PE was low (12%).

[Unexpected focal bowel 18-FDG uptake sites: should they be systematically investigated?]. / Hyperfixations intestinales focales de découverte fortuite sur la TEP au 18-FDG : faut-il proposer une exploration de façon systématique ?

The purpose of the present study was to retrospectively evaluate the nature and significance of unexpected focal 18-FDG uptake localized by PET/CT within the intestinal tract. Methods. The data of 4,033 PET/CT were retrospectively reviewed. One hundred and eighty PET/CT showed unexpected focal uptakes (patients with known intestinal neoplasia were excluded). Among them, 42 patients corresponding to 47 focal uptake sites were investigated by endoscopy or surgery. Results. Height endoscopy results were negative (17%). We found 25 adenomatous polyps (53.2%), 10 neoplasms (21.3%) and 4 inflammatory lesions (8.5%). 18-FDG uptake values were not statistically different between the 4 groups. Conclusion. Eighty-three percent of unexpected intestinal foci of hypermetabolism are either inflammatory, malignant or premalignant lesions. These results justify systematic investigation of these lesions.