2354
Prediction of early treatment response in patients with newly diagnosis multiple myeloma using WB-MRI: comparative study with and without anemia
Huazheng Dong1, Wenyang Huang2, Xiaodong Ji3, Shuang Xia3, Zhiwei Shen4, Dehui Zou2, Wen Shen3, and Zhiyi Song1
1Tianjin Academy of Traditional Chinese Medicine Affiliated Hospital, Tianjin, China, Tianjin, China, 2State Key Laboratory of Experimental Hematology, Institute of Hematology and Blood Diseases Hospital, CAMS and PUMC, Tianjin 300020, China, Tianjin, China, 3Tianjin First Central Hospital Affiliated To Nankai University, Tianjin, China, Tianjin, China, 4Philips healthcare，Beijing， China, Beijing, China
1Tianjin Academy of Traditional Chinese Medicine Affiliated Hospital, Tianjin, China, Tianjin, China, 2State Key Laboratory of Experimental Hematology, Institute of Hematology and Blood Diseases Hospital, CAMS and PUMC, Tianjin 300020, China, Tianjin, China, 3Tianjin First Central Hospital Affiliated To Nankai University, Tianjin, China, Tianjin, China, 4Philips healthcare，Beijing， China, Beijing, China
Synopsis
For patients with newly diagnosed multiple myeloma, red marrow reconversion caused by anemia is common,its hyperintense signal on DWI image could affect the evaluation , because it is very similar with the diffusion signal in focal lesions. We aim to evaluate the efficacy prediction of early treatment response in MM patients with and without anemia based on the tumor burden score (TBS), ADC value and signal fat fraction (sFF). We found that early treatment response of MM could be predicted in patients without anemia using TBS, ADC and sFF, anemia is a factor which must be taken into consideration in evaluation.
Synopsis
For patients with newly diagnosed multiple myeloma (MM), red marrow reconversion caused by anemia[1] is very common. Due to anemia, hyperintense signal on DWI image[2] could affect the evaluation of MM, because it is very similar with the diffusion signal in focal lesions. In this study, we aim to evaluate the efficacy prediction of early treatment response in MM patients with and without anemia based on the tumor burden score (TBS), ADC value and signal fat fraction (sFF). We found that early treatment response of MM could be predicted in patients without anemia using TBS, ADC and sFF acquired by whole body DWI and mDIXON MRI, and anemia is a factor which must be take into consideration in efficacy evaluation.Introduction
For patients with newly diagnosed multiple myeloma (MM), red marrow reconversion caused by anemia[1][L1] is very common. Due to anemia, hyperintense signal on DWI image[2] could affect the evaluation of MM, because it is very similar with the diffusion signal in focal lesions. In this study, we aim to explore the clinical value in prediction of early treatment response with the baseline WB-MRI quantitative parameters (whole body tumor burden score, ADC value and fat-signal fraction) in newly diagnosis MM with and without anemia based on Messiou C’s guideline[3].Materials and Methods
56 patients with multiple myeloma (40 males of aged 40-75 years, 16 females of aged 36-74 years) confirmed pathologically were collected from July 2015 to February 2019. All patients were performed MR examination on a 3T MR scanner (Ingenia, Philips, Best, The Netherland). The protocol included whole body DWI and T1 weighted-mDIXON examination before treatment. The tumor burden score (TBS) of each patient was evaluated based on Messiou C’s guideline[3]. 173 involved sites of diffuse lesion and 158 focal lesions of the whole body were detected and the corresponding ADC value, fat signal intensity (Mfat) and water signal intensity (Mwater) were extracted. The signal fat fraction (sFF) was also calculated by dividing the signal intensity in the fat image by the corresponding in-phase mDIXON image. The average values of ADC and sFF for each patient were calculated. The baseline clinical data (BMPC, β2-microglobulin, ALB, Hb, Cr, ISS-stage and so on) were also recorded. After 4 cycles of induction, treatment response was assessed at the start of each cycle and at the end of induction. Deep response group and non-deep response group were divided according to Stringent complete remission (sCR), complete response (CR), very good partial response (VGPR) , partial response (PR), minimal response (MR) and stable disease (SD) . According to whether the patient’s hemoglobin is less than 100g/l, we performed subgroup analysis. Student’s t test and Mann-Whitney U test were used to compare the differences between two response groups among clinical data and imaging parameters . Chi-square test or Fisher’s exact test was used to compare categorical variables of patients’ characteristics. The relationship between treatment response and TBS, ADC and sFF was assessed using Spearman’s correlation analysis. Receiver operating characteristics (ROC) curves were plotted for each MRI parameter and optimal cutoff points were determined based on the maximum Youden index.Result
For non-anemic patients, there was moderate negative correlation between TBS and depth of response (r=-0.405), strong negative correlation between ADC value and depth of response (r=-0.707), and positive high correlation between sFF and depth of response (r=0.635). Receiver Operator Characteristic analysis indicated that baseline TBS, ADC value and sFF could be treat as the biomarkers for an superior response(≥VGPR) with AUC of 0.773, 0.908 and 0.867. The corresponding cut-off value was TBS of 16.0, ADC of 0.844 ×10−3 mm2/s and sFF of 0. 148. However, none relationship was found above three parameters with the response depth for anemia patients.Discussion and Conclusion
The combined effects of plasma cells infiltration, red bone marrow and yellow bone marrow may affect TBS, ADC and SFF value. For anemia patients, in spite of the less plasma cell bone marrow infiltration, a wide range of abnormal signals can still be seen on the whole-body MRI image, which may interference a successful prediction of early treatment response using baseline WB-MRI quantitative parameter.Our result indicates that the baseline TBS, ADC value and sFF have potential as a noninvasive imaging biomarker for predicting the early treatment response of newly diagnosis MM without anemia. Anemia is a factor which must be take into consideration in efficacy evaluation.Acknowledgements
Funding: This study was supported by the National NaturalScience Fund (81871342) (Shuang Xia) and the NationalKey Research and Development (2019YFC0120903)(Shuang Xia).References
[1] Ormond Filho AG, Carneiro BC, Pastore D, et al. Whole-Body Imaging of Multiple Myeloma: Diagnostic Criteria. Radiographics. 2019. 39(4): 1077-1097.
[2] Padhani AR, van Ree K, Collins DJ, D'Sa S, Makris A. Assessing the relation between bone marrow signal intensity and apparent diffusion coefficient in diffusion-weighted MRI. AJR Am J Roentgenol. 2013. 200(1): 163-70.
[3] Messiou C, Hillengass J, Delorme S, et al. Guidelines for Acquisition, Interpretation, and Reporting of Whole-Body MRI in Myeloma: Myeloma Response Assessment and Diagnosis System (MY-RADS). Radiology. 2019. 291(1): 5-13.
Figures
Figure 1 44-year-old
man with newly diagnosed multiple myeloma (BMPC 20% and Hb 118g/L). The MRI
pattern of bone marrow infiltration was diffuse involvement
A. the baseline DWI
inverted coronal 3D b 800 MIP; B. the T2-STIR image; C[L1] . the fat-only Dixon T1WI image; D. the water-only Dixon T1WI
image; Only one diffuse lesion (high signal intensity on DWI MIP ,
T2-STIR and water-only Dixon T1WI images and low signal intensity on fat-only
Dixon T1WI images) can be seen on the right humerus. No abnormal signal lesions
can be seen in other parts. The TBS (tumor burden score) is recorded as D1.
Figure 2 53-year-
old man with newly diagnosed multiple myeloma(BMPC 24%
and Hb 74g/L). A the baseline
inverted coronal 3D b800 MIP;B the T2-STIR image; C the
fat-only Dixon T1WI image; D the water-only Dixon T1WI image; E the schematic diagram of the same patient which can show the tumor burden score clearly. This patient had diffuse signal in 5 anatomic
sites(including dorsal spine,lumbar spine, pelvis,long bones,ribs ),the diffuse
lesion score would be D5;he also had focal lesions: >10 lesions in long bong
= 3,5-15mm = 2.The focal lesion score would be
F5.Finally,the TBS is 10.
Figure 3 shows the ROI delineated in the focal
lesions in lumbar spine and pelvis. The focal lesions pointed by the arrow are
greater than 1 cm in diameter. The outline of the lesion was manually draw on
the DWI image(Fig.4A), then the ROI outline was copied to the ADC map(Fig.4B),
fat-only image(Fig. 4C)and water only image(Fig. 4D)respectively.
Figure 4 The arrow indicates that the diffuse lesion in
lumbar spine . the contour of ROI includes the whole lumbar vertebrae. The
outline of whole lumbar vertebrae was manually draw on the DWI image(Fig.5A),
then it was copied to the ADC map(Fig.5B),fat-only image(Fig. 5C)and water only
image(Fig. 5D)respectively.
Figure 5 Receiver operating characteristic(ROC) curve for TBS,ADC
and sFF for predicting achievement of deep response (CR/VGPR).Area under ROC
curve are 0.733 for TBS,0.908 for ADC and 0.867 for sFF;
Figure 6 Representative
baseline WB-MRI images of a 45-year-old male patient with a Hb value of 136g/L,
after four cycles of induction chemotherapy, he obtained VGPR. B. coronal DWI image shows high signal in lumbar
vertebrae;C. coronal ADC map shows iso-signal in the same position, ADC
value=0.81×10-3mm2/s;D. coronal fat phase image of
T1-Dixon image shows low- signal, SIFO value=152.01; E.coronal water phase
image of T1-Dixon image shows high-signal, SIWO=355.00.The sFF value is 0.3
Figure 7 Representative
baseline WB-MRI images of a 59-year-old female patient with a Hb value of
122g/L, after four cycles of induction chemotherapy, she obtained PR. A.
inverted coronal 3D b800 MIP image shows focal lesions in lumbar vertebrae; B.
coronal DWI image shows high signal in lumbar vertebrae; C. coronal ADC map
shows iso-signal in the same position, ADC value=0.99×10-3mm2/s; D. coronal fat
phase image of T1-Dixon image shows low- signal, SIFO value=107.73; E. coronal
water phase image of T1-Dixon image shows high-signal, SIWO=570.65. The sFF value is
0.16.
Baseline
characteristics and demographics of patients
Baseline characteristics and demographics of patients
Differentiation
of lesion distribution between deep response and non-deep response groups
Detailed
clinical information in the deep response and non-deep response groups
Spearman correlation analysis between
response depth and MRI parameters
Comparison of baseline MR parameters
between the two response groups