Whole-body MRI (WB-MRI) has gained widespread recognition in initial evaluation of patients with multiple myeloma (MM) [1]. Recently, quantitative WB-MRI imaging in monitoring response of MM have yielded promising results [2, 3]. This study aims to evaluate MRI biomarkers in focal lesions (FLs) of newly diagnosed MM patients by WB-MRI prior to (Pre) and its changes following 8 weeks (8W) of chemotherapy.Twenty-one patients (13 male, median age 52 (range 36-69)) biopsy proven MM were enrolled prospectively. WB-MRI was performed on a 3.0T scanner. T2 weighted imaging and axial DWI-MRI (4 b-values 0,100, 300, 1000) were supplemented by pre- and post-contrast coronal 2-point mDixon imaging, covering head to toe (figure 1). For each patient, up to 20 FLs were localised for analysis by 2 radiologists in consensus, who prospectively scored (0 - non-diagnostic quality images; 1=unlikely, 2=indeterminate, 3=likely and 4=highly likely disease) involvement of 10 pre-defined anatomical regions. Maximum of 4 largest FLs per anatomical location > 5mm and scored 3/4 were selected for analysis. A matched region of interest (ROI) was contoured for each FL on pre- and post-contrast water only and pre-contrast fat only mDixon images as well as b1000 DWI images. Estimated tumour volume (eTV) was calculated by three-axis measurements on post-contrast mDixon and DWI b1000. FF was derived from ROI mean signal intensity (SI) by: SIfat/SIPre-contrast-Water + SIfat [4]. Apparent diffusion coefficient (ADC) calculated by mono-exponential curve fitting of mean SIs of all b-values from DWI [5]. Finally, enhancement ratio (ER) derived by:(SI post-contrast-Water - SI pre-contrast-Water) × 100/SI pre-contrast-Water. Disease response was assessed by international myeloma working group (IMWG) criteria after termination of induction therapy and patients were assigned to responder and non-responder groups [6]. Difference in MRI biomarkers between Pre and 8W studies for each group and for individual patients were assessed by Wilcoxon test. Biomarkers’ receiver operating characteristic (ROC) and area under the curve (AUC) analysis was conducted for prediction of non-responding patients at the end of induction. There were 15 responders and 6 non-responders after induction chemotherapy. The number of lesions evaluated in responders and non-responders was 186 [median 14, range1- 18] and 68 [median 12, range 7-15] respectively. There was a significant reduction in eTV in both groups. No significant change was observed for ER for either group. ADC increased significantly in responders but not in non-responders. Fat fraction increased significantly following treatment in the responder group but not the non-responder group (Figures 2-4). One patient with a single focal lesion in responder group was excluded from the per-patient analysis. There was a significant decrease of eTV following treatment in 13/14 and 6/6 of patients in responders and non-responders, respectively. There was a significant increase in the median FF for 13/14 responders (p<0.001 to 0.03). There was no significant change in FF following treatment in all non-responders (p=0.30 to 0.60). ADC increased significantly following treatment in 7/14 responders (p=0.001 to 0.04). 6/14 responders demonstrated no significant change of ADC (p=0.12 to 1.0); and in one patient there was a significant decrease in ADC. In the non-responder group, ADC did not significantly change following treatment in 4/6 patients (p=0.50 to 0.97), and increased in 2/6 (p=0.02 and 0.03). In 6/14 responders, there was no significant change of ER following treatment (p=0.20 to 0.93). The ER significantly decreased in 2/14 (p=0.001 and <0.001) and significantly increased in 6/14 (p =0.003 to 0.03) patients. There were no significant change of ER in 3/6 (p=0.14 to 0.31) and a significant increase of ER in 3/6 (p=0.01 to 0.05) patients in the non-responder group following treatment. Percentage change of FF following treatment was the best predictor of non-responding patients with an ROC-AUC of 1.00 (95% CI 1.00-1.00) (figure 5).We observed a significant decrease in estimated tumour volume following treatment in almost all the patients (19/20). Lesion fat fraction appears a good indicator of poor response with no significant change in non-responders (6/6 patients) compared with significant increase in responders (13/14) following treatment. Although we found a significant increase of ADC in our responder group, in keeping with previous results [7], we also observed more variable results in per-patient analysis than previously reported [3]. For clinical application, we envisage that a whole body MRI study could be performed within 10-minutes to provide FF quantification of focal lesions as a response marker of treatment. Although these results remain to be confirmed in a larger cohort of patients, the early changes in FF may help to identify patients likely to respond to a particular therapy.