(68)Gallium Ventilation/Perfusion Pet-Ct And Ct Pulmonary Angiography For Pulmonary Embolism Diagnosis: An Interobserver Agreement Study.

1495 Objectives: 68Ga Ventilation/Perfusion (V/Q) PET-CT is a promising imaging tool for pulmonary embolism (PE) diagnosis. However, no study has verified whether the interpretation was reproducible between different observers. Difference in interpretation is an important factor that could interfere with the right diagnosis. The aim of this study was to assess the interobserver agreement in the interpretation of V/Q PET-CT for the diagnosis of acute PE, and to compare it to the interobserver agreement of CT pulmonary angiography (CTPA) interpretation. Methods: 24 cancer patients with suspected acute PE underwent V/Q PET-CT and CTPA within 24 hours as part of a prospective pilot study evaluating V/Q PET-CT for the management of patients with suspected PE (PMID: 31044265). V/Q PET-CT and CTPA scans were reassessed independently by 4 nuclear medicine physicians and 4 radiologists, respectively. Physicians had different expertise in reading V/Q scintigraphy and CTPA. For each modality, 2 readers were from an Australian institution and 2 readers from a French institution. Interpretation was blinded to the initial interpretation and any clinical information or imaging test result. For each modality, results were reported on a binary fashion, i.e. “PE” or “no PE”. V/Q PET-CT scans were read as positive if there was at least one segmental or two subsegmental mismatched perfusion defects. CTPA scans were interpreted as positive if there was a constant intraluminal filling defect with a configuration consistent with thrombus. Interobserver agreement was assessed by calculating kappa (κ) coefficient.Results: Out of the 24 V/Q PET-CT scans, the diagnostic conclusion was concordantly negative in 22 patients and concordantly positive in 1 patient. The remaining scan was interpreted as positive by 1 reader and negative by 3 readers. Out of the 24 CTPA scans, the diagnostic conclusion was concordantly negative in 16 and concordantly positive in 1. Out of the 7 remaining scans, PE was reported by 1 reader in 4 cases, by 2 readers in 2 cases, by 3 readers in 1 case. Most of discordant results on CTPA were related to clots reported on subsegmental arteries. Mean kappa coefficient was 0.82 for V/Q PET-CT interpretation and 0.39 for CTPA interpretation. Conclusions: Interobserver agreement in the interpretation of V/Q PET-CT for PE diagnosis was almost perfect. Further studies are now needed to assess the diagnostic accuracy of the test and the impact in the management of patients with suspected PE. Agreement was lower with CTPA, mainly as a result of discrepancies in the subsegmental arteries.