Total Diffusion Volume Evaluated By Whole-Body Diffusion Weighted Magnetic Resonance Image for the Assessment of Tumor Burden in Multiple Myeloma; Relations with Measurable Residual Disease Evaluated By Next-Generation Flow Cytometry and Bone Marrow CD138 Staining Assessed By Qupath

Introduction: Whole-body diffusion-weighted magnetic resonance imaging (WBDWI) is an emerging imaging technique for evaluating tumor burden in patients with multiple myeloma (MM). Given MM’s patchy distribution, response assessment of residual disease solely through one site, such as minimal residual disease (MRD) by bone marrow aspiration, may be inadequate. Estimating the residual tumor volume of the entire body at the time of response is necessary. Total diffusion volume (tDV) is a potential parameter to quantify the whole body-tumor volume obtained by WBDWI.

Methods: We retrospectively analyzed the tDV obtained by WBDWI before and after treatment, which was the first-best response in patients with newly diagnosed MM from 2017 to 2023. WBDWI was performed on a 1.5-Tesla unit (Magnetom Vision; Siemens Healthcare) with b-value of 0 and 900 s/mm². The regions of interest, including areas with abnormally high signal intensity of diffusion-weighted imaging considered as accumulations of myeloma cells in the bone marrow (BM) or extramedullary tissue, were automatically identified. The sum of the volumes of lesions throughout the body with high signal intensity was calculated as tDV. Physiological areas with high signal intensity were removed manually. The median apparent diffusion coefficient (ADC) value and tDV, representing all pixels in the bone marrow and extramedullary disease with ADC value from 300 to 2000µm²/sec, were quantitatively calculated by using imaging software of BD score (PixSpace, Fukuoka, Japan). MRD from BM samples, simultaneously obtained with WBDWI at the first-best response was assessed using 8-color next-generation flow cytometry (DuraClone, Beckman Coulter, Brea, CA), with a sensitivity equivalent to EuroFlow standard procedure (1.0×10^-5). Digital analysis using whole-slide imaging was employed to quantify the proportions of BM-CD138 positive plasma cells (BMPC) by immunohistochemistry clot sections objectively, using a high-resolution Ultra-Fast Scanner (Philips, Best,17 The Netherlands) and QuPath software (version 0.5.1; University of Edinburgh, Scotland). We compared tDV and ADC values before and after treatment, and across response categories. We evaluated the relationship between tDV, residual CD138 positive BMPC, and MRD after treatment and their impact on progression-free survival (PFS).

Result We analyzed 103 patients whose pre- and post-treatment tDV values were available. The median tDV significantly decreased from 231.5mL before treatment to 31.7mL after treatment (p<0.001). Patients who achieved less than complete response (CR) (partial response [PR] and very good partial response [VGPR], n=24) had a median tDV of 90.56 mL, while those achieving CR or better (≥CR, n=79) had a median tDV of 27.56 mL (p=0.04). The ADC value did not significantly differ before and after treatment (median 890 and 810 µm²/sec, p=0.55) and the ADC value after treatment between patients who achieved less than CR and those who achieved CR or better was not also different (median 760 and 850 µm²/sec, p=0.32). Patients with low tDV (<40 mL) after treatment exhibited significantly longer PFS compared to those with high tDV (≥40mL) (not reached vs, 35 months p< 0.01). No correlation was found between tDV and residual CD138 positive BMPC (r=-0.08, p=0.40). The correlation between tDV and MRD was low (r=0.33, p<0.01) as 21 patients (20.3%) demonstrated high tDV even though the MRD was negative after treatment. Among MRD negative-patients, those with high tDV after treatment had shorter PFS compared to those with low tDV after treatment, and MRD-positive patients had the shortest PFS (not reached, 51, and 35 months, respectively; MRD negative with low tDV vs MRD negative with high tDV; p=0.02, MRD negative with high tDV vs MRD positive; p=0.04). In multivariate analysis, low tDV was a significant prognostic factor for PFS (HR, 0.23, 95%CI, 0.08-0.64; p<0.01).

Conclusion: Our findings indicate that high tDV correlates with shorter PFS, even in patients with MRD negative. The PFS results suggest that WBDWI identifies residual disease outside the site where MRD from the BM sample was examined, with poor prognosis for the residual tumor on MRD or WBDWI and favorable prognosis when both are negative. A combined analysis of MRD assessment from BM samples with WBDWI is a potentially useful method for optimizing the assessment of the treatment response of patients with MM.