RT Journal Article SR Electronic T1 Correlation between whole-bone marrow 18F-FDG uptake and prognostic factors in newly diagnosed multiple myeloma patients JF Journal of Nuclear Medicine JO J Nucl Med FD Society of Nuclear Medicine SP 99 OP 99 VO 61 IS supplement 1 A1 Mahdi Zirakchian Zadeh A1 Brian Ostergaard A1 William Raynor A1 Esha Kothekar A1 Oswaldo Acosta-Montenegro A1 Chaitanya Rojulpote A1 Siavash Mehdizadeh Seraj A1 Fatemeh Kaghazchi A1 Leila Arani A1 Thomas Werner A1 Niels Abildgaard A1 Mona-Elisabeth Revheim A1 Poul Flemming Hoilund-Carlsen A1 Abass Alavi YR 2020 UL http://jnm.snmjournals.org/content/61/supplement_1/99.abstract AB 99Introduction: Investigators in the previous studies evaluated the correlation of focal lesion 18F-FDG uptake with prognostic factors in MM patients. However, according to the International Myeloma Group, poor reproducibility of conventional methods of PET quantification coupled with lack of standard imaging criteria represent a serious limitation to implementing the conventional approaches for MM patients. Our group recently proposed a novel method of 18F-FDG PET quantification for MM patients based on CT threshold to assess 18F-FDG uptake in the whole-bone marrow (WBM) of the patients, instead of focusing on focal osteolytic lesions. In this study, we aimed to assess the correlation of WBM 18F-FDG uptake (expressed as global mean standardized uptake value or GSUVmean) in the MM patients with prognostic factors. Methods: The pre-treatment 18F-FDG PET/CT scans of 42 MM patients were collected. A threshold algorithm was applied based on Hounsfield units on CT images to segment the bony structures and to quantify the 18F-FDG uptake in the entire skeleton of the subjects examined (OsiriX software, Pixmeo; Bernex, Switzerland). Target-to-background ratio (TBR) was performed for normalization. Image analysis was replicated by two independent observers. Results: In both univariate and multivariate regression analyses, beta 2-microglobulin (B2M), hemoglobin (Hgb) and calcium were significantly correlated with TBR-GSUVmean (CE (CI) for multivariate analysis: B2M: 0.02 (0.01-0.03), p< 0.001, Hgb: - 0.31(-0.048- -0.13), p= 0.01, calcium: 0.002 (0.001-0.009), p= 0.03). The correlation between TBR-GSUVmean and albumin was not significant in univariate analysis, but the association became statistically significant in multivariate analysis (CE for multivariate analysis: -0.04 (-0.01-0.07), p= 0.04). A high level of agreement was observed between two operators (ICC: 0.970, 95% CI: 0.953-0.980; p< 0.001). Conclusions: We used a semi-automated method of PET/CT quantification and assessed the correlation of WBM 18F-FDG uptake with prognostic factors. The importance of this approach becomes more evident considering that diffuse bone marrow involvement is present in some MM patients, and the global approach would appear more practical for the assessment of 18F-FDG PET scans of the patients. In addition, high reproducibility that we observed for our proposed methodology indicates that this approach has the potential to overcome the poor reproducibility of conventional methods of PET quantification.