Characterization of Human Multiple Myeloma Cells with High-Resolution Magic Angle Spinning Magnetic Resonance Spectroscopy

Advancements in the pathogenetic pathways in multiple myeloma have led to the identification of several primary and secondary genetic lesions and ultimately to a multiple myeloma genetic classification with prognostic implications. Although disregulation of cyclin activity has been recognized as a key event leading to the multiple myeloma phenotype, little is known about the metabolic consequences of this phenomenon. We have studied intact multiple myeloma cells by high resolution magnetic resonance spectroscopy to establish the metabolomic profiles of different native multiple myeloma cells as compared to other lymphoproliferative disorders.

Multiple myeloma cells obtained from bone marrow aspirates (n =15), blood (n =3) or other biologic tissues (n =2) from 20 multiple myeloma patients and separated by density gradient centrifugation were evaluated and metabolic profiles were correlated with cytogenetic characteristics of the disease and patients clinical characteristics.

Twelve patients were females (60%) with a median age of 65 years (range 50–82). Multiple myeloma monoclonal proteins were IgG (N=9), IgA (N=5) or BJ (N=6). Five of them (25%) had renal insufficiency. Nine patients (45%) had predominantly extramedullar diesase including four cases of plasma cell leukemia. IgH translocations were identified in 5 samples (25%), hyperploidy in 2 (10%), and other or no genetic lesions in 13 (65%), del13 was present in 9 samples (45%) and p53 alterations in 5 (25%).

Bone marrow samples from thirteen patients with conventional multiple myeloma presented a relatively constant metabolic pattern with predominantly lipidic signals and a metilen to metil ratio ranging from 1.9 to 4.9 (median 2.9). No differences in this pattern were observed among subgroups of primary translocations or involvement of Rb and p53 genes. Four patients with plasma cell leukemia and three with predominant extranodal disease presented either non detectable lipid signals (N=3) or a higher metilen to metil ratio ranging from 2.8 to 3.9 (median 3.5). In fact, extranodal or leukemic disease was significantly associated to undetectable lipids (P < 0.031) or the composite variable undetectable lipids or metilen to metil ratio > 3 (P < 0.043). Furthermore, after a median follow-up of 18 months, absence of lipids in the metabolic profile was also associated to a shorter survival (median 0.45 years, 95%CI 0–1.03 versus 3 years, 95%CI 0.95–5.06, P < 0.022).

These results suggest that metabolic profiles of different multiple myeloma genetic subtypes share common and reletively constant characteristics, while cells obtained from patients with plasma cell leukemia or predominantly extramedullar disease present a clearly distinct profile, probably reflecting the metabolic effect of clonal evolution at a genetic level.