P-155: Evaluation of pseudo-CT MRI sequences (zero-echo-time (ZTE) and back bone (BB) gradient-echo) for the detection of lytic bone lesions in MM: comparison with PET/CT.

Introduction:
MRI is highly sensitive for assessing bone marrow involvement in multiple myeloma (MM), but does not allow assessment of mineral bone and detection of osteolysis.



In this prospective study, we evaluated the diagnostic accuracy, repeatability, and reproducibility of two pseudo-CT MRI sequences (zero-echo-time (ZTE) and back bone (BB) gradient-echo) in detecting osteolytic lesions in MM, using whole-body CT (WB-CT) as the reference standard.

Methods:
Between June 2021 and December 2022, we prospectively enrolled consecutive patients with either a newly diagnosed (ND) MM or MGUS, or MM with progressive disease (PD) after therapy. All participants first underwent a clinically indicated 18F FDG-PET/CT, including an optimized WB-CT. They then underwent ZTE and BB sequences covering the lumbar spine, pelvis, and proximal femurs embedded in a 3.0T whole-body MRI (WB-MRI) examination at maximum one-month interval. Ten bone regions and two scores (categorical score: presence/absence of lytic lesion; semi-quantitative score: number of lesions) were evaluated by three radiologists on ZTE, BB, and WB-CT images. Accuracy (ACC), repeatability and reproducibility of MRI scores (from Gwet’s agreement coefficients AC1/AC2) and agreement in semi-quantitative scoring were assessed per-sequence, per-region and per-patient.

Results:

A total of 47 participants were included (27 male, median age of 67 years (range 63-70), 21 NDMM, 26 PD, 5 biological progression, 21 progression under therapy with a median of 3 prior lines of therapy (range 1-8).



In the per-patient analysis, ACC^ZTE and ACC^BB were very high (=98%) for the 2 readers experienced in pseudo-CT reading, and high (≥ 0.85) for the external reader.



In the per-region analysis, both ACC^ZTE and ACC^BB were very high (at least ≥ 0.87) in the coxal bones and femurs; in the spine and sacrum, ACC^BB remained high (≥ 0.91) while ACC^ZTE ranged from 0.74 to 0.89 due to false-negative observations.



Regardless of sequences, per-region and per-patient analyses, repeatability was very good (all AC1 ≥ 0.87), while reproducibility was at least good (AC2 ≥ 0.63).

Regardless of region, the semi-quantitative scores were higher with BB compared to ZTE.





Conclusions:
Both MRI-based pseudo-CT sequences had a high diagnostic accuracy in detecting osteolytic lesions in MM. The BB sequence had a higher accuracy than ZTE in the spine and a higher agreement with WB-CT. The inclusion of these sequences in WB-MRI may strengthen the value of the technique for the assessment of skeletal involvement in MM. Further studies are needed to assess the added value of these sequences in conjunction with conventional MRI sequences, and compare the diagnostic accuracy of WB-MRI supplemented with these sequences with that of FDG-PET.