MRI and Metastatic Bone Survey (MBS) Examinations in Multiple Myeloma (MM): Laboratory Correlates and Prognostic Implications for 611 Patients Receiving Total Therapy 2.

Background: The Durie-Salmon (DS) staging system represented the first clinically useful classification of MM. While distinguishing patients with differing event-free (EFS) and overall survival (OS), considerable outcome heterogeneity persisted.

Patients and Methods: As part of a high-dose melphalan-based tandem transplant trial, Total Therapy 2 (TT2), 611 patients had both MBS and MRI performed at baseline, along with standard prognostic factors and cytogenetics.

Results: In a comparison of all areas imaged by both techniques, at least one focal lesion (FL) was detected in 451 (74%) patients by MRI and 344 (56%) by MBS; 128 (21%) had no FL by either technique; 312 (51%) showed FL by both MBS and MRI; of 267 patients without FL on MBS, 139 (52%) had FL on MRI; of 160 without FL on MRI, 32 (20%) had FL on MBS. Mean FL number among those with FL was 13.4 by MRI and 7.8 by MBS. Significantly higher proportions of patients had FL on MRI than MBS in spine (78% vs. 16% p<0.001); pelvis (64% vs. 28%, p<0.001) and sternum (24% vs. 3%, p<0.001). Higher FL number on both MRI and MBS and heterogeneity of diffuse marrow signal on MRI STIR images (indicating micro-nodular disease) were prognostically harmful along with the presence of cytogenetic abnormalities (CA), elevated serum levels of B2M, CRP and lactate dehydrogenase (LDH), presence of hypo-albuminemia and advanced age. On multivariate analysis (MVA), MRI- but not MBS-defined FL number was an independent adverse baseline feature for overall survival (HR=1.73; p<.001), independent of CA (HR=2.11; p<.001) and elevations of LDH (HR=1.56; p=.006) and B2M levels (HR=1.43; p=.026). Hypo-albuminemia and increased CRP levels were prognosis-neutral once the afore-mentioned parameters were accounted for. Clinical complete response (CCR) (MVA: HR, 0.60; p=0.007) and MRI-CR (univariately: HR, 0.58; p=0.004)) were additional favorable features when examined as time-dependent variables along with baseline features. Higher MRI-FL number (>7) was significantly associated with higher serum levels of LDH, CRP and creatinine as well as with hypo-albuminemia, whereas levels of B2M, bone marrow plasmacytosis, hemoglobin, CA, age, gender, race and Ig isotype were not.

Conclusion: Our data justify routine application of MRI in MM:

1. as the appropriate imaging tool permitting detection of eventually devastating FL long before osteolytic disease is recognized on MBS,

2. as an independent staging tool with prognostic implications (that should replace MBS-FL used in the Durie-Salmon staging system) after accounting for LDH, B2M and presence of CA;

3. to document a superior state of complete remission (MRI-CR) conferring survival advantage, especially evident in patients with a high focal lesion number at baseline;

4. as a tool for detecting and staging non-secretory and macro-focal MM, with the latter often having minimal or no myelomatous involvement on random bone marrow examination; and

5. as a method of detecting non-secretory or macro-focal relapse that is becoming more common in advanced stages of intensely treated patients.