AUTHORS
Ioannis Ntanasis-Stathopoulos, MD, Vassilis Koutoulidis2, Charalampos Filippatos3, Maria Gavriatopoulou, MD4, Panagiotis Malandrakis, MD, MSc, PhD5, Evangelos Eleutherakis-Papaiakovou6, Nikolaos Kanellias, MD, PhD7, Magdalini Migkou8, Ilias Katsadouros, MD9, Panagiotis Boutsikos, MD10, Eirini Solia, MD11, Foteini Theodorakakou, MD12, Efstathios Kastritis, MD13, Meletios-Athanasios (Thanos) C. Dimopoulos, MD14, Lia-Angela Moulopoulos15, Evangelos Terpos, MD16
AFFILIATIONS
1 National and Kapodistrian University of Athens, Athens, Greece
2 National and Kapodistrian University of Athens, Athens, Greece
3 National and Kapodistrian University of Athens, Athens, Greece
4 National and Kapodistrian University of Athens, Athens, Greece
5 National and Kapodistrian University of Athens, Athens, Greece
6 National and Kapodistrian University of Athens, Athens, Greece
7 National and Kapodistrian University of Athens, Athens, Greece
8 National and Kapodistrian University of Athens, Athens, Greece
9 National and Kapodistrian University of Athens, Athens, Greece
10 National and Kapodistrian University of Athens, Athens, Greece
11 National and Kapodistrian University of Athens, Athens, Greece
12 National and Kapodistrian University of Athens, Athens, Greece
13 National and Kapodistrian University of Athens, Athens, Greece
14 National and Kapodistrian University of Athens, Athens, Greece
15 National and Kapodistrian University of Athens, Athens, Greece
16 National and Kapodistrian University of Athens, Athens, Greece
ABSTRACT
Introduction: While modern therapies have revolutionized the treatment landscape of newly diagnosed multiple myeloma (NDMM), predicting early relapse remains a critical clinical challenge. Traditional risk-stratification models often overlook the spatial heterogeneity of the bone marrow microenvironment. While whole-body magnetic resonance imaging (WBMRI) is the gold standard for detecting marrow infiltration, the prognostic value of its quantitative metrics remains underutilized.
Methods: In this prospective single-center study, 91 consecutive patients with newly diagnosed multiple myeloma (NDMM) treated with modern therapies underwent baseline whole-body MRI (WBMRI). Bone marrow fat fraction (FF) and apparent diffusion coefficient (ADC) were measured in the thoracolumbar spine; the presence and number of focal lesions (FL) and paramedullary disease (PMD) sites were also recorded.
Results: The median age was 67.5 years and there was a male predominance (61.5%), with nearly half of the cohort (46.2%) classified as having high-risk disease per IMS/IMWG criteria. Baseline whole-body MRI evaluation revealed a predominantly focal infiltration pattern (50.5%), while normal, diffuse, and variegated patterns were observed in 25.3%, 18.7%, and 5.5% of patients, respectively. First-line treatment was anti-CD38-based in 52.7%, anti-BCMA-based in 27.5%, PI+IMiD-based in 16.5% and IMiD-based in 3.3%. The 75.8% of the cohort achieved ≥VGPR, which included sCR in 16.5% and CR in 6.6%. The ORR was 97.8%. The median follow-up was 33.3 months.
The median baseline marrow FF was 43.0% (IQR, 33.2–58.8). Maximally-ranked statistics revealed a cutoff of 20% as the optimal threshold for PFS curve separation in our cohort. A total of 11 (12.4%) patients had a median FF of less than 20% at baseline, while 78 (87.6%) above 20%. Those with FF<20% demonstrated a near 6-fold increase for disease progression or death (HR=5.87, 95%CI: 2.10-16.4, p=0.007), compared to those with ≥20%. The 2-year PFS rate between the two groups was 45.5% and 89.2%, respectively (p<0.001). This observed increased risk remained significant after adjusting for prespecified baseline covariates (aHR=7.36, 95%CI: 1.93-28.00, p=0.003). Moreover, when further stratifying patients in three subgroups (<20%, 20-40% and >40% baseline FF), those with <20% demonstrated significantly worse PFS compared to both other groups, with 2-year PFS rates of 45.5%, 79.9% and 94.1%, respectively (p=0.004). Multivariate analysis demonstrated a 27% reduced risk for a PFS event among those with 20-40% FF (aHR=0.73, 95%CI: 0.06-0.98, p=0.047) and a vast 92% reduced risk for a PFS event (aHR=0.08, 95%CI: 0.02-0.37, p=0.001) for those with >40% FF, compared to the <20% FF group.
We then investigated whether more optimal baseline FF thresholds exist for stratification, depending on the baseline WBMRI pattern. The baseline FF distribution varied significantly between normal, diffuse, focal and variegated patterns (p<0.001). In patients with diffuse pattern, an optimal cutoff of 11.8% baseline FF was identified, resulting in a significant 13-fold increased risk for a PFS event (HR=12.90, p=0.006), which was attenuated to a non-significant trend after multivariable adjustment (aHR=12.48, p=0.118). For patients with a focal pattern, the optimal FF cutoff was 40%, resulting in 2-year PFS rates of 69.3% and 93.8% between those with <40% and ≥40% baseline FF, respectively, but results were marginally non-significant (p=0.059).
The median baseline marrow ADC was 0.4 (IQR, 0.4-0.6) and did not seem to be associated with PFS, neither as a continuous variable (p=0.593) nor as a binary cutoff (prespecified, ADC>0.6, p=0.172 and ADC>0.7, p=0.134). Interestingly, only 1 out of the 9 patients with an ADC>0.7 also had FF<20% at baseline, highlighting a non-significant association between the two metrics (p=0.333). PMD was present in 26 patients (28.6%). No survival disadvantage was induced by the presence of PMDs (PFS, p=0.411) at diagnosis.
Conclusion: Bone marrow fat depletion (FF <20%) on WBMRI at MM diagnosis emerges as a powerful, independent imaging biomarker for early disease progression or death in NDMM in the contemporary treatment era. Moreover, PMD and FL-oriented sub-stratification at baseline also warrant consideration and further prospective research.
DISCLOSURES
Ioannis Ntanasis-Stathopoulos, MD: No relevant conflicts of interest to declare.