Sequential Imaging with Diffusion-Weighted Whole-Body MRI (DW-MRI) and PET-CT Identifies Patients at High Risk of Relapse in Multiple Myeloma

Introduction: Although the vast majority of patients with newly diagnosed multiple myeloma (NDMM) achieve deep responses during first-line therapy, most of them face relapse over time. Focal lesions, which are seen in up to 80% of NDMM, play a crucial role in MM evolution and can act as reservoirs for relapse disease. As a consequence, the combination of imaging- and minimal residual disease (MRD) negativity has been defined as a new response category. However, the dynamics and location of focal lesions during the treatment course remains poorly characterized and the answer to what imaging modality should be used to visualize focal residual disease remains elusive. In this study, we employed two techniques, PET-CT and DW-MRI, at multiple pre-defined time points to longitudinally track resistant lesions and investigate the combined prognostic value of these imaging modalities with other prognostic markers.

Methods: We included 170 NDMM patients who were diagnosed between the years of 2010-2015 to ensure sufficient median follow-up time, which was 10 years. Median age of the study cohort was 60 years (range: 34-76), 104 were male (61%), and 31 patients (18%) had high-risk according to gene-expression profiles. All patients underwent PET-CT and DW-MRI evaluation at three pre-defined timepoints: baseline, post autologous stem-cell transplantation (ASCT) and at relapse. Minimal residual disease testing (MRD) by 8 color flow cytometry at a sensitivity level of 10^-5 post ASCT was available for 70 patients.

Results: At baseline, 144/170 patients (85%) presented with focal lesions. Both imaging modalities, PET-CT and DW-MRI, showed positive results in 116 patients (68%). However, 21 patients (12%) had focal lesions detectable only by DW-MRI, while 7 patients (4%) showed focal lesions solely in PET-CT. After ASCT, only 12 patients had PET-CT positive lesions, whereas DW-MRI detected focal lesions in 53 patients, indicating a much higher detection rate with DW-MRI. Overall, 55 out of 170 patients (32%) showed residual lesions at this time point. The 26 patients w/o focal lesions in both PET-CT and DW-MRI at baseline remained negative post ASCT. Except for one patient, focal lesions after ASCT were persistent from initial diagnosis. Of note, two-thirds of the patients had residual focal lesions in the humeri and/or femora, which is surprising given that myeloma is considered a disease of the axial skeleton. Next, we correlated the presence of residual lesions with baseline patient characteristics, including age, gender, stage, and cytogenetic risk factors. Compared to patients w/o residual lesions, PET-positive patients showed an enrichment of rISS stage III (7 vs 50%, p<0.001) and deletion 17p (10 vs 33%, p=0.04). In contrast, patients who were DW-MRI positive only showed no significant association with any of the investigated variables. Both PET- and DW-MRI positive lesions post ASCT were associated with poor PFS (median 1.5 and 2.4 years, respectively vs. 9.4 years in imaging-negative patients, p<0.001) and OS (median 3.7 and 5.7 years, respectively vs. 11.3 years, p<0.001). Importantly, in a multivariate model the adverse prognostic effect of residual lesions was independent of baseline risk factors. To increase the power of the following combined imaging/MRD analysis, we defined an “Imaging+” group, which included the patients with PET+ and/or DW-MRI+ focal lesions. Among the 20 patients who achieved MRD-negative and imaging-negative status, an excellent PFS was observed, with only three events occurring in an 8-year follow-up. Residual lesions despite MRD negativity were only seen in two patients, and one of them progressed after 1.5 years. Dismal PFS was seen for double-positive patients (median 1.4 years), significantly worse compared to patients with MRD positivity only (8.9 years).

Conclusion: Taken together, DW-MRI captures more patients with prognostically relevant residual focal lesions compared to PET-CT and should therefore be considered as a preferable imaging modality post ASCT. The high number of residual lesions in the extremities suggests a role of humeri and femora as reservoirs for treatment-resistant cells, deserving further investigation. For patients with MRD and imaging double positivity post ASCT, prognosis is dismal and alternate therapeutic approaches should be considered.