Abstract
241894
Introduction: Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome marked by elevated somatostatin receptor expression in causative tumors. SSTR imaging aids in locating and diagnosing these elusive tumors. Compared to 68Ga-labeled tracers, 18F-labeled probes offer advantages including increased production yield, lower positron energy, and longer half-life, potentially enhancing image quality. This study aimed to compare the diagnostic efficacy of Al18F-NOTA-LM3 with 68Ga-DOTA-TATE PET/CT for detecting the culprit lesion of TIO.
Methods: 10 patients with clinically suspected TIO were prospectively recruited and underwent whole-body 68Ga-DOTA-TATE and Al18F-NOTA-LM3 PET/CT scans PET/CT within 3 days. The diagnosis of the culprit lesion was confirmed by the combination of PET/CT results and clinical information. All images were independently interpreted by two experienced nuclear medicine specialists. Abnormal findings on PET/CT images were compared with pathological examinations and clinical manifestations.
Results: 68Ga-DOTA-TATE PET/CT and Al18F-NOTA-LM3 PET/CT both identified a solitary tumor lesion exhibiting elevated expression of somatostatin receptors in seven patients. The SUVmax of 7 culprit tumors was higher on Al18F-NOTA-LM3 PET/CT compared with that on 68Ga-DOTA-TATE PET/CT (34.4 ± 5.9 vs. 27.8 ± 3.7; p = 0.157). Four of these patients underwent surgical excision of the culprit tumor, followed by subsequent alleviation of systemic bone pain and fatigue symptoms as well as an elevation in serum phosphorus level. The remaining three patients opted for conservative management, with one case presenting tumor invaded the occipital bone slope. Furthermore, three patients demonstrated negative findings on both PET/CT examinations.
Conclusions: The preliminary study indicated that Al18F-NOTA-LM3 PET/CT is effective in detecting TIO pathogenic tumors, with comparable diagnostic efficiency to 68Ga-DOTA-TATE PET/CT. Further research necessitates additional clinical studies with larger sample sizes.