Traditionally, anti-myeloma therapy is initiated by presence of end-organ damage defined by CRAB criteria (hypercalcemia, renal failure, anemia, and/or lytic bone lesions). Recently, several clinical trials have been initiated to evaluate the role of treatment in smoldering myeloma (SMM) patients, which by diagnostic criteria lack end-organ damage. Several studies have indicated the prognostic significance of magnetic resonance imaging (MRI) for patients with smoldering (SMM) and multiple myeloma (MM). Given that the risk of transformation from SMM to MM varies greatly, we have conducted a large retrospective clinical study, using whole-body MRI, to characterize patterns of tumor infiltration in the bone marrow among patients diagnosed with SMM. These patients were then followed longitudinally for progression to MM. Furthermore, we have expanded MRI studies to patients with monoclonal gammopathy of undetermined significance (MGUS), and have analyzed the imaging findings within established prognostic parameters of this group of patients.
Methods: A total of 157 patients with SMM, 138 with monoclonal gammopathy of undetermined significance (MGUS), and 249 with MM were assessed by whole body MRI. Patients with second malignancies, amyloidosis, solitary plasmocytoma or preceding systemic treatment were excluded from the study. In patients with MGUS and SMM initiation of systemic treatment and in patients with MM, relapse after systemic therapy was defined as end point for progression free survival. For all patients we collected extensive clinical and diagnostic data. Using logistic regression, we evaluated the presence/absence of focal or diffuse signal abnormalities. Using log-rank test we defined treatment-free survival for SMM in relation to imaging results. Median follow-up was 4.0, 4.5 and 3.7 years for MGUS, SMM and MM patients, respectively.
Results: In MGUS patients, focal lesions were detectable in 23.9% and a diffuse infiltration in 53%. Diffuse and focal infiltration patterns appeared independently from each other. The presence and number of focal lesions as well as a severe diffuse infiltration were statistically significant adverse prognostic factors for progression free survival. In multivariate analysis, only the number of focal lesions remained statistically significant (p=0.0005). In SMM patients, focal lesions were present in 34.4% and a diffuse infiltration pattern in 45.9%. Plasma cell percentage, a moderate diffuse infiltration (but not focal lesions) and beta2-microglobulin were statistically significant prognostic parameters.
Whole-body MRI indicated that for 65.6% of the patients with SMM, patterns of tumor cell infiltration in the bone marrow were similar to MGUS patients; whereas 34.4% showed patterns similar to MM patients. Sensitivity and specificity of prediction performance of the classification model was 0.8 and 0.76, respectively. Using a log-rank test for prediction of treatment-free survival of SMM patients in relation to whole-body MRI patterns of tumor cell infiltration in the bone marrow, we found a borderline difference between MGUS-like SMM (n=124) and MM-like SMM patients (n=33) (p=0.08); whereas the difference between MM-like SMM patients and the 138 MGUS patients was significant (p=0.02).
Conclusions: In summary, in this first large clinical study including 544 patients, whole-body MRI was able to discern over 30% of patients with SMM presenting with patterns of tumor cell infiltration in the bone marrow similar to those of MM. Given that many patients with SMM develop symptomatic disease within 1-2 years, in the future, advanced imaging may play a major role in defining patients with SMM who should be candidates for early treatment.