Only Concomitant Presence of Cytogenetic Abnormalities in Randomly Sampled Bone Marrow and in MRI-Defined Focal Lesions Imparts Poor Prognosis and Is Associated with Gene Expression Profiling (GEP)-Based High-Risk Myeloma.

Abstract

Cytogenetic abnormalities (CA) can be detected in one-third of newly diagnosed patients with multiple myeloma (MM), reaching virtually universal presence at the terminal disease stage. According to both histopathological and radiological findings, bone marrow growth patterns are either diffusely infiltrative or characterized by focal lesion (FL) growth, around which typically osteolytic bone destruction develops. The detection of CA at diagnosis confers a poor prognosis, and the sustained suppression of CA is critical for long-term survival. Applying MRI examinations in virtually all newly diagnosed patients prior to protocol-based therapy, FL have been detected in ~80% of patients and impart, along with the detection of CA on random bone marrow examination from the posterior iliac crest, shorter event-free and overall survival. Such MRI-FL harbor viable MM cells often with CA, persist in clinical CR for a median of 1–2 years, eventually resolve in 60% and constitute sites of relapse often without M-protein secretion, collectively suggesting an important role of FL in both disease manifestation and progression. We reviewed our data base of 1202 patients enrolled in Total Therapy (TT) protocols for entries of CT-guided fine needle aspirations (FNA) from MRI-defined FL submitted to cytogenetic analysis at baseline and on any occasion prior to first transplant. We identified 320 patients with cytogenetic information on both randomly sampled (RS) and FNA from FL. Congruency between FL and RS examinations was documented in 71% including 53% without detectable CA and 18% with CA in both sites; 14% had RS-CA without FL-CA and 16% had FL-CA without RS-CA. The overall RS-based CA frequency of 31% was identical to the 31% when all 1202 RS prior to transplant were considered. The relative distribution of standard prognostic factors was similar among the 4 RS-FL constellations in terms of B2M (>5.5mg/L), albumin (<3.5g/dL) and creatinine (>=2mg/dL) levels. The frequency of gene expression profiling (GEP)-defined risk (determined on RS) was 15% among196 patients with concurrent RS and FL sampling and thus virtually identical to the 14% incidence among all 620 subjects with RS information only. However, 53% of the subgroup exhibiting both RS-CA and FL-CA had high-risk disease compared to only 6% in the remainder (p < 0.001). Analysis of overall survival according to the 4 RS-FL CA constellations revealed an adverse impact only of the concomitant presence of RS-CA and FL-CA (Figure), whereas the presence of either RS-CA or FL-CA individually was prognosis-neutral, an observation confirmed by multivariate analysis (HR of 3.27, p<0.001). The much higher frequency of GEP-defined high-risk in patients with concomitant RS-CA and FL-CA requires further study, including the examination by GEP of MM-cells procured from both RS and FL. [Display omitted]