Premium
Whole body magnetic resonance imaging in newly diagnosed multiple myeloma: early changes in lesional signal fat fraction predict disease response
Author(s) -
Latifoltojar Arash,
HallCraggs Margaret,
Rabin Neil,
Popat Rakesh,
Bainbridge Alan,
Dikaios Nikolaos,
Sokolska Magdalena,
Rismani Ali,
D'Sa Shirley,
Punwani Shonit,
Yong Kwee
Publication year - 2017
Publication title -
british journal of haematology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.907
H-Index - 186
eISSN - 1365-2141
pISSN - 0007-1048
DOI - 10.1111/bjh.14401
Subject(s) - medicine , magnetic resonance imaging , imaging biomarker , multiple myeloma , effective diffusion coefficient , diffusion mri , biomarker , nuclear medicine , prospective cohort study , progressive disease , radiology , receiver operating characteristic , chemotherapy , biochemistry , chemistry
Summary Cross‐sectional imaging techniques are being increasingly used for disease evaluation in patients with multiple myeloma. Whole body magnetic resonance imaging ( WB ‐ MRI ) scanning is superior to plain radiography in baseline assessment of patients but changes following treatment have not been systematically explored. We carried out paired WB ‐ MRI scans in 21 newly diagnosed patients prior to, and 8‐weeks after, starting chemotherapy, and analysed stringently selected focal lesions ( FL s) for parametric changes. A total of 323 FL s were evaluated, median 20 per patient. At 8 weeks, there was a reduction in estimated tumour volume ( eTV ), and an increase in signal fat fraction ( sFF ) and apparent diffusion coefficient ( ADC ) in the group as a whole ( P < 0·001). Patients who achieved complete/very good partial response ( CR / VGPR ) to induction had a significantly greater increase in sFF compared to those achieving ≤ partial response ( PR ; P = 0·001). When analysed on a per‐patient basis, all patients achieving CR / VGPR had a significant sFF increase in their FL 's, in contrast to patients achieving ≤ PR . sFF changes in patients reaching maximal response within 100 days (fast responders) were greater compared to slow responders ( P = 0·001). Receiver Operator Characteristic analysis indicated that sFF changes at 8 weeks were the best biomarker (area under the Curve 0·95) for an inferior response (≤ PR ). We conclude that early lesional sFF changes may provide important information on depth of response, and are worthy of further prospective study.