Evaluación de la exploración PET/TC con 18F-TFB en pacientes con cáncer diferenciado de tiroides

Patients receive fluorine F18 tetrafluoroborate IV and undergo PET/CT scan on study.

Ancillary studies

Undergo a PET/CT scan

• Connecticut
• GATO
• Análisis de gato
• Tomografía Axial Computarizada
• Tomografía axial computarizada
• Tomografía computarizada
• tomografía

Given IV

• 18F-tetrafluoroborato
• 18F-TFB

Undergo a PET/CT scan

• Imágenes médicas, tomografía por emisión de positrones
• MASCOTA
• Escaneo de mascotas
• Tomografía por emisión de positrones
• Tomografía de emisión de positrones
• Imágenes espectroscópicas de resonancia magnética de protones

Descriptive summaries of detected lesions (presence/absence) will include counts and proportions at the patient and lesion level. Overall concordance of these measures across modalities will be estimated using Cohen's kappa with 95% confidence intervals generated by clustered bootstrapping. Per-lesion analysis of detection rates by modality will be performed using cluster-adjusted McNemar's test, accounting for the paired nature of the data (i.e., two scans) and potential within-sample correlation for patients with multiple lesions.

Conspicuities and diagnostic confidence scores for a given lesion will be averaged across readers. Descriptive summaries of these scores will include medians and interquartile ranges. Concordance of these measures across modality will be evaluated using Lin's concordance correlation coefficient with 95% confidence intervals generated by clustered bootstrapping. Differences in conspicuities and diagnostic confidence scores will be tested using cluster-adjusted Wilcoxon signed-rank tests.

Given true positive and negative lesions determined by histopathology and/or imaging, clinical and laboratory follow-up, we will estimate sensitivity, specificity, net present value, and positive predictive value on a per-lesion and patient basis for 18F TFB PET/CT and 123I single-photon emission computerized tomography/CT along with the corresponding two-sided 95% confidence intervals. The confidence intervals will be constructed using clustered bootstrapping.

The impact of PET on clinical management in DTC patients will be evaluated using descriptive statistics. Evidence of discordance among the modalities will be assessed using a McNemar- Bowker test of symmetry.

Inter-reader reproducibility will be assessed for positivity at the patient level and region level (nominal) as well as standardized uptake values (SUVs) for positive lesions (quantitative). Reproducibility will be reported using the Fleiss' Kappa test for multiple readers for positivity and Pearson intra-class correlation for SUVs.