Simplified response monitoring criteria for multiple myeloma in patients undergoing therapy with novel agents using
Christoph Schabel1, Marius Horger1, Sara Kum1
1Department of Diagnostic and Interventional Radiology, Eberhard-Karls-University Tuebingen, Hoppe-Seyler-Str. 3, 72076 Tuebingen, Germany.
Introduction:
Multiple myeloma is a malignant hematological disorder of the mature B-cell lymphocytes originating in the bone marrow. While therapy monitoring is still mainly based on laboratory biomarkers, the additional use of imaging has been advocated due to inaccuracies of serological biomarkers or in a-secretory myelomas. Non-enhanced CT and MRI have similar sensitivities for lesions in yellow marrow-rich bone marrow cavities with a favourable risk and cost-effectiveness profile of CT. Nevertheless, these methods are still limited by frequently high numbers of medullary lesions and its time consumption for proper evaluation.
Objective:
To establish simplified response criteria by correlating size and CT attenuation changes of medullary multiple myeloma lesions in the appendicular skeleton with the course of lytic bone lesions in the entire skeleton. Furthermore to evaluate these criteria with respect to established hematological myeloma-specific parameters for the prediction of treatment response to bortezomib or lenalidomide.
Materials And Methods:
Non-enhanced reduced-dose whole-body CT examinations of 78 consecutive patients (43 male, 35 female, mean age 63.69±9.2years) with stage III multiple myeloma were retrospectively re-evaluated. On per patient basis, size and mean CT attenuation of 2-4 representative lesions in the limbs were measured at baseline and at a follow-up after a mean of 8 months. Results were compared with the course of lytical bone lesions as well with that of specific hematological biomarkers. Myeloma response was assessed according to the International Myeloma Working Group (IMWG) uniform response criteria. Testing for correlation between response of medullary lesions (Respmed) and response of all myeloma manifestations including osteolyses (Resptotal) was performed using the corrected contingency coefficient (Ccorr).
Results:
The correlation between Respmed based on length diameter and transverse diameter and Resptotal was perfect (Ccorr=1.0; p<0.0001) whereas the correlation based on density was moderate (Ccorr=0.54; p<0.0001). The evaluation of simplified response criteria with a measurement of only 2 medullary lesions yielded the best sensitivity and specificity valued for treatment-induced changes for the length diameter evaluation with 94.4%/95.7% for prediction of progressive disease and 78.6%/93.3% for prediction of therapy response. There were no significant differences between patients treated with bortezomib and lenalidomide (p>0.05).
Conclusion:
Measurements of size of a minimum of two medullary lesions is sufficient for response assessment and correlates very well with the course of lytic bone lesions and that of hematologic parameters.
