Diagnostic value of ${}^{18}$F-fluorodeoxyglucose positron emission tomography/computed tomography imaging in pediatric opsoclonus myoclonus ataxia syndrome presenting with neuroblastoma

Early precision diagnosis and effective treatment of opsoclonus myoclonus ataxia syndrome (OMAS) patients presenting with neuroblastoma can prevent serious neurological outcomes. To assess the diagnostic value of ${}^{18}$F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) imaging in pediatric OMAS with neuroblastoma. A retrospective evaluation of 45 patients diagnosed with OMAS who underwent ${}^{18}$F-FDG PET/CT was performed. A univariate analysis was performed to compare clinical characteristics between OMAS with and without neuroblastoma. Univariate and multivariate logistic regression analyses were applied to identify independent risk factors for OMAS with neuroblastoma and to develop the clinical model. Finally, independent risk factors and PET/CT were fitted to build the combined model for the diagnosis of OMAS with neuroblastoma and presented as a nomogram. Receiver operating characteristic curve, decision curve, and calibration curve analyses were conducted to evaluate the performance of the models. Among 45 patients, 27 were PET/CT-positive, 23/27 lesions were neuroblastoma, and four were false positives. One of the false positive patients was confirmed to be adrenal reactive hyperplasia by postoperative pathology, and the symptoms of OMAS disappeared in the remaining three cases during clinical follow-up. The average maximal standardized uptake value of PET/CT-positive lesions was 2.6. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of PET/CT were 100%, 81.8%, 85.2%, 100%, and 91.1%, respectively. Age at diagnosis, lactate dehydrogenase, and neuron-specific enolase showed statistically significant differences between OMAS with and without neuroblastoma. Lactate dehydrogenase was identified as the independent risk factor to develop the clinical model, and the clinical model demonstrated an area under the curve (AUC) of 0.82 for the diagnosis of OMAS with neuroblastoma, with an AUC as high as 0.91 when combined with PET/CT. The decision curve analysis and calibration curve demonstrated that the nomogram had good consistency and clinical usefulness. In patients with OMAS, ${}^{18}$F-FDG PET/CT has a high diagnostic accuracy in detecting tumors of the neuroblastoma, especially when combined with the independent risk factor serum lactate dehydrogenase.