RSNA 2011 

Abstract Archives of the RSNA, 2011


MSVN51-08

Diffusion-weighted Imaging Facilitates Detection of Spinal Multiple Myeloma and Assists in Diagnosing Equivocal Lesions

Scientific Formal (Paper) Presentations

Presented on December 1, 2011
Presented as part of MSVN51: Neuroradiology Series: Spine

Participants

James M. Kessler MD, MPH, Presenter: Director of Business Development, FrontRad Technologies, Inc.
Puneet Pawha MD, Abstract Co-Author: Nothing to Disclose
Katya Shpilberg MD, Abstract Co-Author: Nothing to Disclose
Lawrence N. Tanenbaum MD, Abstract Co-Author: Speakers Bureau, General Electric Company Speakers Bureau, Bracco Group Speakers Bureau, Vital Images, Inc Medical Advisory Board, General Electric Company Medical Advisory Board, Bracco Group Medical Advisory Board, Vital Images, Inc Medical Advisory Board, Olea Medical Medical Advisory Board, Nocimed, LLC

PURPOSE

Traditional multiple myeloma (MM) imaging employs T1, T2, STIR and occasionally contrast enhanced (CE) sequences. Disease detection and characterization remain difficult. Elderly patients with heterogeneous marrow, degenerative changes, and posterior element and vertebral endplate disease complicate detection and characterization of individual lesions. MM presents unique challenges since traditional approaches often lack sensitivity. Recent investigations support improved metastatic spine disease evaluation using diffusion weighted imaging (DWI). Composed of densely packed cells, MM impedes water diffusivity leading to focal hyperintensity on DWI. We sought to evaluate DWI benefits in spinal MM lesion detection and characterization.

METHOD AND MATERIALS

We retrospectively reviewed spine surveys from 15 pathology proven MM patients performed from January 2010 - December 2010 [1-19 lesions per patient (total=105)].  7 patients (45 lesions) received contrast. Routine sequences included sagittal T1, STIR and T2 weighted images. When feasible, CE sequences were performed. Sagittal echo-planar DWI sequences were added (3-6 directions, B value 500-800). Initially, T1, STIR and (if available) CE sequences were evaluated.  All sequences were consensus reviewed twice by 2 experienced neuroradiologists.  Diagnostic confidence for each lesion was graded 1-4 (1=benign; 2=equivocal; 3=probable; 4=definite). Lesions were regraded after DWI supplementation. Lesion conspicuity differences were recorded as "more," "less," or "same." A non-normal data distribution mandated use of nonparametric methods including Kruskall-Wallis, Wilcoxon Signed Rank and Sign tests (α=0.05).  Bowker's test for symmetry and Kappa coefficient estimations evaluated the sigificance and magnitude of diagnostic confidence score differences for routine and DWI techniques.

RESULTS

Diagnostic confidence was greater with DWI (μ=3.4) than without (μ=3.0) [p=0.0029], was upgraded in 35%, identical in 49% and downgraded in 16% of lesions.  11 lesions (10%) were missed without DWI. 1 lesion (1%) was visible only on DWI.  On DWI, 51 lesions (49%) were more conspicuous, 32 (30%) equally conspicuous, and 22 (21%) less conspicuous.

CONCLUSION

Fast and easy to perform, DWI improves MM lesion detection and characterization, especially in problematic locations.

CLINICAL RELEVANCE/APPLICATION

Improved MM tumor burden estimation facilitates accurate prognostication and treatment response evaluation.

Cite This Abstract

Kessler, J, Pawha, P, Shpilberg, K, Tanenbaum, L, Diffusion-weighted Imaging Facilitates Detection of Spinal Multiple Myeloma and Assists in Diagnosing Equivocal Lesions.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11001777.html