Bone disease in multiple myeloma (MM) patients is usually assessed by skeletal X-ray (XR) and magnetic resonance imaging (MRI) of the spine (MRIS). Diffusion-weighted MRI (DW-MRI) is an innovative whole-body MRI that detects malignant lesions studying water diffusion in tissues. This prospective study compared DW-MRI with XR and MRIS for the assessment of lytic bone lesions in MM patients. Patients had symptomatic MM at diagnosis (stage I-III Durie and Salmon [D&S]) requiring the first treatment, or MM at relapse before the start of the salvage treatment. An exploratory substudy enrolled asymptomatic patients with D&S stage I MM. The primary objective of the study was to assess whether DW-MRI could detect a higher number of lytic bone lesions than XR and MRIS. The secondary objectives were: to assess whether there was a correlation between the number of lesions and response to therapy; to assess whether the number of lesions could be correlated with progression; and, to assess whether the DW-MRI could detect more lesions than standard whole-body MRI (WB-MRI). The explorative objective of the substudy was to evaluate whether DW-MRI could detect more lytic bone lesions than standard imaging in asymptomatic MM patients. Along with clinical objectives, the study was aimed at correlating the data coming from the experimental imaging technique with the biologic characteristics of disease. Patients performed XR, MRIS, WB-MRI, and DW-MRI at diagnosis or at relapse, after the treatment and 6 months thereafter (symptomatic MM) or every 6 months for 1 year (asymptomatic MM). MRIS and DW-MRI were done in a single 45-minutes session by a standard 1.5 Tesla MRI scanner. DW-MRI consisted of multiple stacked axial EPI sequences at 4 b-values, evaluated by PET-like MIP and MPR reconstructions at the highest b-value (1000). Along with the radiology exams, patients performed at the same timepoints serological and histological evaluations of disease, including cytogenetics with FISH before starting treatment. The study enrolled 50 patients between 2008 and 2010, 36 of them in the main study, 14 in the substudy. Patients in the main study evaluable for the results were 35. Their median age was 65 years (range, 33-81), D&S stage was I for 48.6% of them, II for 5.7%, and III for 45.7%; ISS staging was I for 71.4% of patients. The 57.4% of patients had IgG, 25.7% had IgA MM, 8.6% had non-secerning and micromolecolar MM. 42.9% of patients were at diagnosis, 57.1% at relapse. Median bone marrow infiltration was 30%. FISH on selected CD138+ plasma cells was normal in 28.6% of patients; 42.9% of them had a del(13), 22.9% of them had t(11;14); 8.6% had t(4;14) and 5.7% del(17). Between the 1st and the 2nd exam, patients received a treatment including lenalidomide (35.3%), thalidomide (23.5%), or bortezomib (47.0%). 97% of patients received steroids (dexametasone or metilprednisone). 26.5% of patients received an autologous transplant, and 5.9% an allogeneic transplant. Response was as following: sCR 10.3%, VGPR 24.1%, PR 37.9%, SD 10.3%, PD 17.2%. At the 3rd radiology exam, the response status was: 8% sCR, 36% VGPR, 28% PR, 20% SD and 8% PD. At the first radiology exam, XR showed a median of 1 bones with focal lesions (range, 0-10). MRIS showed a median of 0 lesions (range, 0.0-4.0). The association of XR and MRIS (XR+MRIS) showed a median of 2 lesions (range, 0-10), WB-MRI a median of 2 lesions (range, 0-18, p=0.03), and DW-MRI a median of 5.8 lesions (range, 0-18, p=5 lesions/patient detected by DW-MRI. OS was similar in these 2 groups (p=0.48), whereas patients with >=5 lesions had a worse PFS (p=0.018) and RFS (p=0.009), and a higher RI (p=0.002). After the treatment, patients is CR had a median of 0 lesions, those in VGPR 2 lesions, those in PR 3 lesions, and those in SD in PD 2 and 3 lesions, respectively. The average of lesions progressively rose from 0.7 (patients in CR) to 4.8 lesions (patients in PD). Standard methods showed a median of 0-2 lesions less than DW-MRI consistently with response. This association between response and DW-MRI was observed also in patients at the 3rd exam (0 lesions for patients in CR, 5 in patients in PD), with an increasing difference in detecting lesions according to disease status compared to the standard methods. The DW-MRI showed less lesions in the 2nd timepoint compared to the 1st timepoint (p=0.04), consistently with the fact that most patients responded to therapy, and remained stable between the 2nd and 3rd exam (p=0.21). The XR and XR+MRI remained the same during all the longitudinal 3 evaluations (p=1.0, p=0.12, and p=0.27, p=0.10, respectively). There was a trend for patients with IgA MM to have more lesions in DW-MRI (87.5% vs 50% of IgG MM, p=0.15). Patients with positive Bence-Jones had more lesions than those with negative BJ (60% vs 21.4%, p=0.06), and similarly those with positive urinary immunofixation (63.3% vs 30%, p=0.15). Freelite ratio was weakly correlated with the bone with lesions in DW-MRI (p=0.11, corr=0.29), but correlation increased considering the microscopic lesions detected by DW-MRI (corr=0.33, p=0.06). Patients at diagnosis had a 45.4% of lesions >=5 compared to 75.0% of those relapsed (p=0.21), and patients with D&S stage III had frequently >=5 lesions compared to lower stages (52.6% vs 37.5%). Patients with FISH at higher cytogetic risk (Mayo stratification) had more lesions compared to those at standard risk (73.6 vs 45.4%, p=0.24). There was a high correlation between the number of lesions showed by DW-MRI and LDH value before treatment (corr=0.44, p=0.008). Patients with high level of ICTP (>4.5) had >=5 lesions in DW-MRI. The substudy including asymptomatic patients showed a median of 0 lesions (range, 0-2) at XR, 0 at MRIS (range, 0-3), 1 lesion at XR+MRIS (range, 0-3), 0.5 lesions at WB-MRI (range, 0-6, p=0.81) and 2.5 lesions at DW-MRI (range, 0-8, p=0.04). After 6 and 12 months of follow-up the lesions were in median 0 for XR, MRIS and XR+MRIS, 0 for WB-MRI (p=0.67 and p=0.62), and 1.0 for DW-MRI at both timepoints (p=0.17 and p=0.22). In conclusion, DW-MRI was superior than standard radiological methods in detecting lytic bone lesions for symptomatic MM patients at diagnosis requiring treatment or at relapse. The number of lesions were correlated with PFS, RFS and RI. The lesions detected by the new technique correlated with response as categorized by IMWG criteria. The number of lesions were also correlated with LDH, and less strongly with Freelite ratio, Bence-Jones proteinuria or urinary immunofixation, high risk FISH, IgA isotype, relapsed disease. DW-MRI showed to be superior also in detecting lytic bone lesions in asymptomatic patients. The exams were feasible and well tolerated. DW-MRI is a new radiological method that is very promising in an highly specialized setting for the evaluation of bone lesions in MM.