Radiol Oncol 2024; 58(1): 43-50. doi: 10.2478/raon-2024-0003
43
research article
Quantitative assessment of bone marrow 
infiltration and characterization of tumor 
burden using dual-layer spectral CT in patients 
with multiple myeloma 
Xing Xiong1, Rong Hong1, Xu Fan1, Zhengmei Hao1, Xiaohui Zhang2, Yu Zhang3,  
Chunhong Hu1
1 Department of Radiology, The First Affiliated Hospital of Soochow University, Suzhou, Jiangsu, China
2 Department of Clinical Science, Philips Healthcare Greater China, Shanghai, China
3 Department of Radiology, Dushu Lake Hospital Affiliated to Soochow University, Suzhou, China
Radiol Oncol 2024; 58(1): 43-50.
Received 18 August 2023 
Accepted 31 October 2023
Correspondence to: Yu Zhang M.D., Department of Radiology, Dushu Lake Hospital Affiliated to Soochow University, Suzhou, 215163, China. 
E-mail: zhangyusdfyy@163.com and Chunhong Hu M.D., Department of Radiology, The First Affiliated Hospital of Soochow University, 
Suzhou,215006, Jiangsu, China. E-mail: sudahuchunhong@163.com
Xing Xiong and Rong Hong contributed equally to this work.
Disclosure: No potential conflicts of interest were disclosed.
This is an open access article distributed under the terms of the CC-BY license (https://creativecommons.org/licenses/by/4.0/).
Background. The aim of the study was to evaluate whether virtual calcium subtraction (VNCa) image extracted 
from dual-layer spectral CT could estimate bone marrow (BM) infiltration with MRI as the reference standard and 
characterize tumor burden in patients with multiple myeloma (MM).
Patients and methods. Forty-seven patients with newly diagnosed MM were retrospectively enrolled. They had 
undergone whole-body low-dose dual-layer spectral CT (DLCT) and whole-body MRI within one week. VNCa images 
with calcium-suppressed (CaSupp) indices ranging from 25 to 95 at an interval of 10 and apparent diffusion coef-
ficient (ADC) maps were quantitatively analyzed on vertebral bodies L1−L5 at the central slice of images. The optimal 
combination was selected by correlation analysis between CT numbers and ADC values. Then, it was used to char-
acterize tumor burden by correlation analysis and receiver operating characteristic (ROC) curves analysis, including 
plasma cell infiltration rate (PCIR), high serum-free light chains (SFLC) ratio and the high-risk cytogenetic (HRC) status. 
Results. The most significant quantitative correlation between CT numbers of VNCa images and ADC values could 
be found at CaSupp index 85 for averaged L1−L5 (r = 0.612, p < 0.001). It allowed quantitative evaluation of PCIR (r = 
0.835, p < 0.001). It could also anticipate high SFLC ratio and the HRC status with area under the curve (AUC) of 0.876 
and 0.760, respectively.
Conclusions. The VNCa measurements of averaged L1−L5 showed the highest correlation with ADC at CaSupp 
index 85. It could therefore be used as additional imaging biomarker for non-invasive assessment of tumor burden if 
ADC is not feasible. 
Key words: bone marrow; tumor burden; virtual non-calcium; dual energy CT; multiple myeloma
Introduction
Multiple myeloma (MM) is one of the malignant 
hematological diseases with monoclonal prolifera-
tion of plasma cells which primarily involves bone 
marrow (BM).1 “Myeloma bone disease” forms 
when malignant proliferation of plasma cells dis-
places the healthy BM, and then results in activa-
tion of osteoclasts and inhibition of osteoblastic 
activity.2,3 The characterization of BM tumor bur-
den has important indications for treatment regi-
mens, treatment response and surveillance. It has 
Radiol Oncol 2024; 58(1): 43-50.
Xiong X et al. / Tumor burden and dual-layer spectral CT in multiple myeloma44
been exclusively accomplished by BM biopsy and 
serologic/urine markers such as plasma cell infil-
tration rate (PCIR), serum-free light chains (SFLC) 
ratio, paraproteins (M-protein) in serum/urine and 
cytogenetic status.4-6 However, these biomarkers 
examinations suffer unavoidable deficits such as 
invasive, painful and expensive.
As first introduced by Durie and Salmon in 1975, 
conventional radiographic survey of the skeleton 
was applied to stage MM bone disease.7 However, 
owing to low sensitivity in detecting osteolytic 
lesions and unable to evaluate therapy response, 
it calls for more practical techniques to be used. 
With the development of imaging techniques such 
as monoenergetic computed tomography (MECT), 
magnetic resonance imaging (MRI), and fluoro-
deoxyglucose positron-emission-tomography CT 
(FDG PET/CT), direct evaluation of BM infiltration 
has become possible.8 MECT is widespread avail-
able and economic efficient, so MM patients are 
commonly first assessed with whole body MECT 
scans.9 The major limitation of MECT is low sen-
sitivity for detecting nonlytic BM infiltration in 
the axial skeleton, which is more common for MM 
patients. MRI is confirmed to be “imaging golden 
standard” for BM infiltration which has proven 
higher sensitivity in detecting MM lesions than 
any other modality.10 Whereas, it takes long time 
to accomplish examination for patients which may 
cause unbearable pain and claustrophobia.11,12 FDG 
PET/CT has been lately recommended to evalu-
ate response and residual activity in treated pa-
tients as it could respond to BM changes quickly.13 
However, the associated radiation and economic 
cost should be considered.
Dual-layer spectral CT (DLCT) is a novel CT 
technique with two different detector layers atop 
each other to absorb different parts of the poly-
chromatic-attenuated X-ray spectrum. It could 
construct various parameter images e.g., uric acid, 
iodine, or calcium according to the aim of research 
retrospectively. Recent studies showed that DLCT, 
especially virtual non-calcium (VNCa) image, 
shows significant improvements in comparison to 
MECT and comparable to FDG PET/CT and MRI 
in the evaluation of MM.14-16 Hence, our study had 
two objectives: firstly, to explore the potential of 
VNCa image in estimating BM infiltration with 
MRI as the reference standard in MM patients. 
Secondly, to identify if VNCa image could charac-
terize tumor burden by correlate with established 
biomarkers (PCIR, SFLC ratio and cytogenetic sta-
tus).
Patients and methods
Patient characteristics 
The study was approved by ethics committee of lo-
cal institution and the need for written informed 
consent was waived due to retrospective nature 
of the study (registration number: 000/2021). All 
scans were performed for conventional clinical re-
quirements. 
We have collected the information of MM pa-
tients from 6/2021 to 10/2022 admitted to our in-
stitution consecutively. The inclusion criteria were 
as follows: (1) histologically confirmed diagno-
sis of MM; (2) the interval between clinical data, 
whole-body low-dose DLCT and whole-body MRI 
examination no more than two weeks; (3) received 
no specific therapy for MM before. The exclusion 
criteria were as follows: (1) patient’s age below 18 
years; (2) no complete clinical data, neither DLCT 
nor MRI examination; (3) obvious metal or motion 
artifacts affecting the lumbar vertebral segmenta-
tion. 
Imaging acquisition and post-processing
All scans were performed on a commercially avail-
able spectral detector DLCT scanner (IQon Spectral 
CT, Philips Healthcare), following the most recent 
recommendations of the International Myeloma 
Working Group (IMWG).17 Patients were placed in 
a head-first supine position. The scan ranges from 
vertex of the skull to the knees. No contrast agent 
was given. Scan parameters were as follows: tube 
voltage, 120 kV; tube current, 70 mAs; collimation, 
64×0.625 mm; pitch, 0.990; rotation time, 0.75 s; 
volumetric computed tomography dose index, 7.4 
mGy. Mean dose length product was 1069.2 ± 205.9 
mGy*cm. The field of view (FOV) was adjusted de-
pending on patient body volume.
The corresponding MRI examination was per-
formed on a 3.0 T scanner (Magnetic Verio, Siemens 
Healthcare, Erlangen Germany). The patients were 
also placed in a head-first supine position. Phased-
array surface coils were installed to cover from the 
head to the upper femur. No contrast medium was 
given. The protocol parameters were as follows: 
T2 turbo inversion recovery magnitude (TIRM) 
sequence [echo time (TE), 84 ms; repetition time 
(TR), 7110 ms; slice thickness, 5 mm; slice gap, 1.5 
mm; FOV, 480 mm] was acquired on the coronal 
plane from the head to the upper femur. On the 
same coverage area, axial DWI sequences were ac-
quired using two values (b = 50, 700 s/mm2) with 
Radiol Oncol 2024; 58(1): 43-50.
Xiong X et al. / Tumor burden and dual-layer spectral CT in multiple myeloma 45
the following parameters: TR, 4000 ms; TE, 46 ms; 
slice thickness, 5 mm; slice gap, 0; FOV, 450 mm.
Post-processing of spectral-based image (SBI) 
data was performed with the vendor’s software 
(IntelliSpace Portal Version 11, Philips Healthcare). 
First, all SBI images were reconstructed in a 512 × 
512 matrix, slice thickness 2 mm with an overlap of 
1 mm. Then, VNCa images were created from SBI 
data by exploiting the material specific attenuation 
of X-rays in different energy levels to simulate each 
voxels attenuation in Hounsfield units without the 
calcium-specific contribution. The intelligent post-
processing vendor allows calcium suppression in 
seamlessly adjustable factors. In our study, VNCa 
images were reconstructed with calcium-sup-
pressed (CaSupp) indices ranging from 25 to 95 in 
steps of 10. Among them, CaSupp indice 25 means 
images has minimum visibility of bony structures 
and 95 means maximum visibility. 
Segmentation of the bone marrow 
Although the MM lesions were scattered, it in-
volved typical location such as lumbar vertebra, 
pelvis and ribs. So, we chose to focus on L1−L5 due 
to the large size of those vertebrae with maximized 
reliable measurement, typical sites of BM infiltra-
tion and minimally affected by the intrauterine de-
vice.18,19 Using the same software, regions of inter-
ests (ROIs) were positioned manually in the sagittal 
vertebral bodies L1−L5 to measure the respective CT 
numbers and basivertebral vein was avoided from 
the ROIs. Since the lumbar vertebra were wide, a 
standard circular ROI which size set to 100 mm2 
was placed at the central slice. To ensure compara-
bility, ROIs were copied between different CaSupp 
indices. At the same time, the corresponding loca-
tion was contoured manually on the axial apparent 
diffusion coefficient (ADC) map (Figure 1). The im-
ages were analyzed by two radiologists with more 
than 5 years of experience who were blinded to any 
patient information. The intraclass correlation co-
efficient (ICC) was calculated for determining the 
interrater reliability of the quantitative assessment. 
The final CT number and ADC values were aver-
aged. The analysis of VNCa and ADC images was 
conducted for 10 min per person.
Assessment of established biomarkers
PCIR was obtained through BM biopsy on the 
wing of ilium and assessed by our in-house pa-
thologists. Immunoturbidimetry was used to 
detect the expression levels of SFLC kappa and 
FIGURE 1. An example of bone marrow (BM) segmentation in multiple myeloma (MM) patient. An oval regions of interests (ROIs) 
of 100 mm2 was drawn at the central slice of sagittal vertebral bodies L1−L5 in different virtual calcium subtraction (VNCa) 
images and the corresponding location was manually displayed on the axial apparent diffusion coefficient (ADC) map. (A) 
calcium-suppressed (CaSupp) index 25, (B) CaSupp index 55, (C) CaSupp index 85, (D) CaSupp index95, (E) magnification of 
(C), (F) ADC map.
A B C D
E
F
Radiol Oncol 2024; 58(1): 43-50.
Xiong X et al. / Tumor burden and dual-layer spectral CT in multiple myeloma46
lambda. SFLC ratio were classified as high (<0.01 
or >100) or low (0.01−100) according to the IMWG 
criteria and the practical experience of our institu-
tion.20,21 Cytogenetic status was performed by fluo-
rescence in situ hybridization (FISH) in interphase 
cells to overcome the problem of karyotyping. MM 
patients were divided into high-risk cytogenetic 
(HRC) and standard risk cytogenetic (SRC) groups 
on the basis of FISH results. Patients who pre-
sented with any of the following cytogenetic ab-
normalities (CAs) were categorized into the HRC 
group: del(17p), t(4;14), t(14;16), t (14;20), gain(1p), or 
p53 mutation. Other MM patients were allocated 
into the SRC group.
Statistic assessment
Statistical analysis was performed by either SPSS 
22.0 software (Chicago, IL, USA) or MedCalc sta-
tistical software version 16.4.3 (Ostend, Belgium). 
Correlations between different VNCa CT numbers 
(combined L1−L5 with different CaSupp indices) 
and ADC values were calculated. Since VNCa CT 
numbers and ADC values were normally distrib-
uted, Pearson’s correlation analysis was applied. 
Then the optimal combination was used to char-
acterize PCIR by correlation analysis and receiver 
operating characteristic curves (ROC) analysis 
was carried out to predict binary outcomes “high 
SFLC ratio” and “HRC status”. Statistical signifi-
cance was defined as p ≤ 0.05. 
Results
Patient characteristics
A total of 382 MM patients were admitted at the 
hematology center in our institution for whole-
body DLCT. Of these, 5 patients had to be excluded 
because they were under 18 years old. 239 patients 
had to be excluded because they have received 
anti-myeloma treatment. 61 patients had no com-
plete clinical data, neither DLCT nor MRI exami-
nation. Another 30 patients had obvious metal or 
motion artifacts that affected the lumbar verte-
bral segmentation. Consequently, 47 MM patients 
were included. Enrollment results of MM patients 
after exclusion were shown in the Figure 2. The 
average interval between the clinical examina-
tion and DLCT scan was 10 days [interquartile 
range 3.0−13.5 days]. The clinical characteristics 
are shown in Table 1. The interrater reliability of 
all quantitative measurements was very high with 
ICC ranged from 0.824−0.970.
Correlation analysis
1880 and 235 ROIs were derived from VNCa im-
ages for different CaSupp indices and ADC maps, 
respectively. Table 2 shows the mean ADC values 
and CT numbers (combination of different ver-
tebral bodies and CaSupp indices). Regardless 
of the measured location, CT numbers in VNCa 
FIGURE 2. Flow chart of patients’ selection.
DLCT = low-dose dual-layer spectral CT; MM = multiple myeloma
TABLE 1. Patient characteristics
Characteristics n
Age* 57.9 ± 8.1
Sex#
Males 26 (55.3%)
Females 21 (44.7%)
Myeloma subtypes#
IgG 32 (68.1%)
IgA 10 (21.3%)
Light chain 5 (10.6%)
Plasma cell infiltration ratio obtained from the wing of 
ilium* 0.54 ± 0.30
Kappa/lambda SFLC ratio#
High (< 0.01 or > 100) 27 (57.4%)
Low (0.01−100) 20 (42.6%)
Cytogenetic status#
HRC 26 (55.3%)
SRC 21 (44.7%)
*represented as Mean ± SD; # represented as number (percentage)
HRC = high-risk cytogenetic; SD = Standard deviation; SFLC = serum-free light chains; SRC = 
standard risk cytogenetics
Radiol Oncol 2024; 58(1): 43-50.
Xiong X et al. / Tumor burden and dual-layer spectral CT in multiple myeloma 47
images at CaSupp indices from 75 to 95 were sig-
nificantly correlated with ADC (Pearson’s r ranges 
from 0.342−0.612, with all p < 0.05). Inversely, CT 
numbers in VNCa images at CaSupp indices from 
35 to 45 showed no correlation with ADC for all 
locations. The highest correlation of VNCa-CT 
numbers and ADC values was averaged L1−L5 at 
CaSupp indices 85 (Pearson’s r = 0.612, p < 0.001). 
Figure 3 provides the statistical results regarding 
the correlation between CT numbers (combined 
different CaSupp indices with measured locations) 
and ADC values. 
Characterize tumor burden with optimal 
combination of CaSupp index and 
vertebral body
The CT number of averaged L1−L5 at CaSupp in-
dex 85 showed significant correlation with the 
PCIR (r = 0.835, p < 0.001) confirmed by BM biopsy 
(Figure 4). It showed a mean infiltration ratio of 
54% (range, 10%−95%; median 60%).
We performed ROC analysis with the predic-
tor binary outcome “SFLC ratio” and “cytogenetic 
status” using the CT number of averaged L1−L5 at 
CaSupp index 85. Expectedly, it exhibited satisfy-
ing performance for discriminating high and low 
SFLC ratio with area under the curve (AUC) of 
0.876 (0.736−0.958). The corresponding sensitivity, 
specificity and cutoff value were 0.952, 0.800, 10.66, 
respectively. Also, AUC for prediction of the “cy-
togenetic status” was 0.760 (0.603−0.878). The cor-
responding sensitivity, specificity and cutoff value 
were 0.714, 0.762, 20.43, respectively (Figure 5A, B).
Discussion
Our results showed that VNCa images derived 
from DLCT could estimate BM infiltration with 
FIGURE 3. Heat map of Pearson’s correlation r and p value between CT numbers (combined different calcium-suppressed [CaSupp] indices with 
measured locations) and apparent diffusion coefficient (ADC) values.
TABLE 2. Means and standard deviations of MRI apparent diffusion coefficient (ADC)  and CT numbers in virtual calcium subtraction (VNCa) 
images for all measured locations
L1 L2 L3 L4 L5 Averaged L1−L5
ADC 554.12±177.42
520.02
±171.74
546.19
±179.75
523.20
±175.14
524.3
±173.17
536.30
±163.93
CaSupp 25 -236.58±77.07 -227.60±73.00
-217.83
±83.35
-225.40
±82.36
-244.67
±79.27
-221.19
±78.89
CaSupp 35 -138.15±46.58 -131.15±43.86
-127.11
±51.4
-133.38
±49.30
-143.10
±48.23
-128.84
±47.37
CaSupp 45 -82.04±30.19
-77.33
±28.86
-75.38
±34.45
-81.01
±31.32
-85.24
±31.31
-76.42
±30.10
CaSupp 55 -45.12±20.87
-41.925
±20.70
-41.32
±24.80
-46.57
±20.95
-47.08
±21.44
-41.92
±19.91
CaSupp 65 -18.50±16.35
-16.38
±17.05
-16.79
±19.89
-21.71
±15.83
-19.71
±16.14
-17.07
±14.58
CaSupp 75 2.03±15.39
3.29
±16.53
2.12
±18.3
-2.58
±14.9
1.42
±14.54
2.08
±13.26
CaSupp 85 18.63±16.60
19.22
±17.75
17.45
±18.93
12.93
±16.52
18.55
±15.37
17.61
±14.59
CaSupp 95 32.46±19.15
32.46
±20.12
30.25
±20.75
25.98
±19.16
33.00
±17.41
30.59±
17.20
CaSupp = calcium-suppressed index
Radiol Oncol 2024; 58(1): 43-50.
Xiong X et al. / Tumor burden and dual-layer spectral CT in multiple myeloma48
MRI as the reference, especially the CT number 
of averaged L1−L5 at CaSupp index 85 showed the 
highest correlation with ADC. What’s more, it al-
lowed quantitative evaluation of tumor burden by 
correlating with PCIR and anticipating high SFLC 
ratio and the HRC status.
Instead of activating a second X-ray tube or rap-
id-voltage switching tube before performing the 
examination, DLCT adopts two different detector 
layers to decrease X-ray dose. For postprocess-
ing, the flexible vendor could construct different 
parameters. VNCa image is a common parameter 
in musculoskeletal system9,22,23, in which the osse-
ous component is removed from the spectral base 
data in order to improve visualization of BM. The 
degree of calcium suppression depends on the 
CaSupp index, which defines the calcium compo-
sition level. Several documents have confirmed 
the importance of VNCa image. Fervers et al. as-
sumed that the pathologic BM was defined as vox-
els >0 HU and concluded that it could significantly 
predict BM infiltration, osteolytic lesions and the 
clinical diagnosis of MM.14 However, there is no 
consensus for the CT cutoff number of pathologic 
BM. Brandelik et al. assessed the potential of VNCa 
images to reflect BM infiltration.16 They evaluated 
the different regions (C7, T12, L1−L5) and infiltra-
tion patterns (non-diffuse and diffuse). However, 
C7 is not the typical region for BM infiltration and 
could be influenced by beam hardening artifacts 
easily as far as we know.24 Fervers et al. also inves-
tigated if VNCa images might discriminate meta-
bolically vital, focal lesions from avital lesions in 
MM patients with FDG PET/CT as the standard 
of reference.15 Best result was yielded by high cal-
cium suppression, followed by medium and low 
calcium suppression. However, the median inter-
val time between DECT and FDG PET/CT was 53 
days which was so long to leave time window for 
possible change in tumor biology between two 
images. In our study, the CaSupp indices ranged 
from 25 to 95 with an interval of 10 to search for 
the optimal CaSupp index, which may be more 
scientific and comprehensive. There is a growing 
tendency of the importance for increased CaSupp 
index that high CaSupp index could provide more 
information for BM infiltration and tumor burden 
than low CaSupp index. This might due to gradual 
exposure of underlying plasma cell cluster by in-
creasing calcium suppression, which further vali-
dates VNCa images as a measurement tool for tu-
mor burden. The averaged L1−L5 seems to be more 
representative than single lumbar vertebra due to 
the large size of those vertebrae with maximized 
reliable measurement avoiding sclerosis, fractures, 
or disc herniations. We did not divide the infiltra-
tion pattern according to MRI performance and 
we believe that this “agnostic” approach provides 
a more reliable marrow sample for evaluation of 
BM infiltration.25
We have included laboratory biomarkers to 
evaluate MM tumor burden. Among them, PCIR 
was obtained through BM biopsy on the iliac crest 
clinically, which is painful and uncomfortable for 
most patients. Despite IMWG recommendation26, a 
recent large-scale clinical analysis was performed 
to explore whether BM biopsy is necessary in all 
patients diagnosed with monoclonal protein since 
in some cases it did not contribute to the diagno-
sis. In our study, PCIR was correlated well with 
CT number of averaged L1−L5 at CaSupp index 
85. Thus, it’s promising to obtain PCIR results by 
measuring CT number noninvasively. Due to dif-
ferent thresholds for the serum paraproteins of 
MM subtypes (e.g., IgA, IgG, IgM), only SFLC ra-
tio was taken into consideration which is also an 
important indicator of tumor burden. In 2014, the 
IMWG included the SFLC ratio in the diagnostic 
criteria for MM, and SFLC ratio >100 is consid-
ered as a biomarker for ultrahigh-risk smolder-
ing MM patient.4 However, some MM patients are 
non-secretory or hypo-secretory and are therefore 
difficult to surveille by means of serologic/urine 
markers alone which influences patient manage-
ment at primary diagnosis and during therapy.27 
FIGURE 4. Bivariate correlation between CT number (averaged L1−L5 at calcium-
suppressed [CaSupp] index 85) and plasma cell infiltration rate (PCIR) confirmed 
by bone marrow biopsy. The Pearson’s r yields 0.835 with p value < 0.001.
Radiol Oncol 2024; 58(1): 43-50.
Xiong X et al. / Tumor burden and dual-layer spectral CT in multiple myeloma 49
What’s more, myeloma may escape hematologic 
diagnosis if it extends outside the marrow cavities 
(extramedullary).28 Similarly, ROC analysis indi-
cates satisfactory performance for VNCa images to 
discriminate high and low SFLC ratio with AUC 
0.876. Some studies have found that CAs are sig-
nificantly associated with the proliferation and se-
cretion of tumor cells.29-31 It was obtained through 
different invasive methods such as FISH. However, 
this technique suffers some drawbacks. For exam-
ple, the patients may experience the pain of biopsy 
and bear the expensive expenses. What’s more, 
the BM results may be influenced by intratumoral 
heterogeneity and poor sample quality.32-33 So, de-
veloping a convenient and noninvasive method to 
predict cytogenetic status is critical for clinicians 
and patients. The results showed that VNCa im-
ages could anticipate HRC status with preferable 
AUC, sensitivity and specificity of 0.760, 0.714 and 
0.762. Since the above specific situations may exist 
in clinical practice, such as painful and unbearable 
biopsy for some patients, non-secretory or hypo-
secretory M protein, extramedullary infiltration 
et al., DLCT could be employed to evaluate tumor 
burden additionally. 
There are some limitations that needed to be 
discussed. First, the number of patients was rath-
er small. Since the incidence rate of MM is lower 
than other diseases and is complex to deal with, so 
patients are usually admitted to specialized hos-
pitals. Second, it was validated in the lumbar ver-
tebra which were considered as the representative 
region of BM infiltration and minimally affected 
by the intrauterine device. But this needs to be up-
scaled across the body and also has more robust 
measurement of technique accuracy. Third, corre-
lation with PCIR was possible only for the pelvic 
bones, but this reflects the deficit of daily practice. 
Finally, this study investigated the ability of DLCT 
acquired by specific scanner, imaging protocols, 
and post-processing tools which may not widely 
applied in other institutions. In the future, more 
studies are needed for definitive evaluation of this 
powerful technological equipment.
Conclusions
Quantitative assessment of VNCa images in DLCT 
is a potential determination of BM infiltration ex-
tent in MM for radiologists and would be prom-
ising incorporated into the daily clinical practice, 
especially when the gold standard MRI is not ac-
cessible. Therefore, VNCa images could be used 
as additional imaging biomarkers for non-invasive 
assessment of tumor burden. 
Acknowledgement
This study has received funding by Gusu health tal-
ent project of Suzhou (GSWS2020003) and Suzhou 
Science and Technology Project (SKJY2021025).
FIGURE 5. Receiver operating characteristic curves for CT number (averaged L1−L5 at calcium-suppressed [CaSupp] index 85) 
to predict “high serum-free light chains (SFLC) ratio” (A) and “cytogenetic status” (B).
A B
Radiol Oncol 2024; 58(1): 43-50.
Xiong X et al. / Tumor burden and dual-layer spectral CT in multiple myeloma50
References
1. Kazandjian D. Multiple myeloma epidemiology and survival: a unique 
malignancy. Semin Oncol 2016; 43: 676-81. doi: 10.1053/j.seminon-
col.2016.11.004
2. Panaroni C, Yee AJ, Raje NS. Myeloma and bone disease. Curr Osteoporos 
Rep 2017; 15: 483-98. doi: 10.1007/s11914-017-0397-5
3. O’Donnell EK, Raje NS. Myeloma bone disease: pathogenesis and treat-
ment. Clin Adv Hematol Oncol 2017; 15: 285-95. PMID: 28591104
4. Rajkumar SV, Dimopoulos MA, Palumbo A, Blade J, Merlini G, Mateos MV, 
et al. International Myeloma Working Group updated criteria for the diag-
nosis of multiple myeloma. Lancet Oncol 2014; 15: e538-48. doi: 10.1016/
S1470-2045(14)70442-5
5. Duvauferrier R, Valence M, Patrat-Delon S, Brillet E, Niederberger E, 
Marchand A, et al. Current role of CT and whole body MRI in multi-
ple myeloma. Diagn Interv Imaging 2013; 94: 169-83. doi: 10.1016/j.
diii.2012.12.001
6. Dimopoulos M, Terpos E, Comenzo RL, Tosi P, Beksac M, Sezer O, et al. 
International Myeloma Working Group consensus statement and guidelines 
regarding the current role of imaging techniques in the diagnosis and moni-
toring of multiple myeloma. Leukemia 2009; 23: 1545-56. doi: 10.1038/
leu.2009.89
7. Durie BG, Salmon SE. A clinical staging system for multiple myeloma. 
Correlation of measured myeloma cell mass with presenting clinical fea-
tures, response to treatment, and survival. Cancer 1975; 36: 842-54. doi: 
10.1002/1097-0142(197509)36:3<842::aid-cncr2820360303>3.0.co;2-u
8. Portet M, Owens E, Howlett D. The use of whole-body MRI in multiple 
myeloma. Clin Med 2019; 19: 355-6. doi: 10.7861/clinmedicine.19-4-355
9. Kosmala A, Weng AM, Heidemeier A, Krauss B, Knop S, Bley TA, et al. 
Multiple myeloma and dual-energy CT: diagnostic accuracy of virtual non-
calcium technique for detection of bone marrow infiltration of the spine 
and pelvis. Radiology 2018; 286: 205-13. doi: 10.1148/radiol.2017170281
10. Hillengass J, Usmani S, Rajkumar SV, Durie BGM, Mateos MV, Lonial S, et 
al. International Myeloma Working Group consensus recommendations on 
imaging in monoclonal plasma cell disorders. Lancet Oncol 2019; 20: e302-
e312. doi: 10.1016/S1470-2045(19)30309-2
11. Shah LM, Hanrahan CJ. MRI of spinal bone marrow: part I, techniques and 
normal age-related appearances. AJR Am J Roentgenol 2011; 197: 1298-
308. doi: 10.2214/AJR.11.7005
12. Padhani AR, Koh DM, Collins DJ. Whole-body diffusion-weighted MR imag-
ing in cancer: current status and research directions. Radiology 2011; 261: 
700-18. doi: 10.1148/radiol.11110474
13. Cavo M, Terpos E, Nanni C, Moreau P, Lentzsch S, Zweegman S, et al. Role of 
(18)F-FDG PET/CT in the diagnosis and management of multiple myeloma 
and other plasma cell disorders: a consensus statement by the International 
Myeloma Working Group. Lancet Oncol 2017; 18: e206-e217. doi: 10.1016/
S1470-2045(17)30189-4
14. Fervers P, Fervers F, Kottlors J, Lohneis P, Pollman-Schweckhorst P, Zaytoun 
H, et al. Feasibility of artificial intelligence-supported assessment of bone 
marrow infiltration using dual-energy computed tomography in patients 
with evidence of monoclonal protein - a retrospective observational study. 
Eur Radiol 2022; 32: 2901-11. doi: 10.1007/s00330-021-08419-2
15. Fervers P, Glauner A, Gertz R, Täger P, Kottlors J, Maintz D, et al. Virtual cal-
cium-suppression in dual energy computed tomography predicts metabolic 
activity of focal MM lesions as determined by fluorodeoxyglucose positron-
emission-tomography. Eur J Radiol 2021; 135: 109502. doi: 10.1016/j.
ejrad.2020.109502
16. Brandelik SC, Skornitzke S, Mokry T, Sauer S, Stiller W, Nattenmüller J, et al. 
Quantitative and qualitative assessment of plasma cell dyscrasias in dual-
layer spectral CT. Eur Radiol 2021; 31: 7664-73. doi: 10.1007/s00330-021-
07821-0
17. Moulopoulos LA, Koutoulidis V, Hillengass J, Zamagni E, Aquerreta JD, Roche 
CL, et al. Recommendations for acquisition, interpretation and reporting 
of whole body low dose CT in patients with multiple myeloma and other 
plasma cell disorders: a report of the IMWG Bone Working Group. Blood 
Cancer J 2018; 8: 95. doi: 10.1038/s41408-018-0124-1
18. Reinert CP, Krieg E, Esser M, Nikolaou K, Bösmüller H, Horger M. Role of 
computed tomography texture analysis using dual-energy-based bone 
marrow imaging for multiple myeloma characterization: comparison with 
histology and established serologic parameters. Eur Radiol 2021; 31: 2357-
67. doi: 10.1007/s00330-020-07320-8
19. Hu C, Zhang Y, Xiong X, Meng Q, Yao F, Ye A, et al. Quantitative evaluation of 
bone marrow infiltration using dual-energy spectral computed tomography 
in patients with multiple myeloma. J Xray Sci Technol 2021; 29: 463-75. doi: 
10.3233/XST-200811
20. Mosebach J, Thierjung H, Schlemmer HP, Delorme S. Multiple myeloma 
guidelines and their recent updates: implications for imaging. Rofo 2019; 
191: 998-1009. doi: 10.1055/a-0897-3966
21. Cowan AJ, Green DJ, Kwok M, Lee S, Coffey DG, Holmberg LA, et al. 
Diagnosis and management of multiple myeloma: a review. JAMA 2022; 
327: 464-77. doi: 10.1001/jama.2022.0003
22. Thomas C, Schabel C, Krauss B, Weisel K, Bongers M, Claussen CD, et al. 
Dual-energy CT: virtual calcium subtraction for assessment of bone marrow 
involvement of the spine in multiple myeloma. AJR Am J Roentgenol 2015; 
204: W324-31. doi: 10.2214/AJR.14.12613
23. Ekert K, Hinterleitner C, Baumgartner K, Fritz J, Horger M. Extended texture 
analysis of non-enhanced whole-body MRI image data for response assess-
ment in multiple myeloma patients undergoing systemic therapy. Cancers 
2020; 12: 761. doi: 10.3390/cancers12030761
24. Yu Z, Leng S, Jorgensen SM, Li Z, Gutjahr R, Chen B, et al. Evaluation of con-
ventional imaging performance in a research whole-body CT system with 
a photon-counting detector array. Phys Med Biol 2016; 61: 1572-95. doi: 
10.1088/0031-9155/61/4/1572
25. Koutoulidis V, Terpos E, Papanikolaou N, Fontara S, Seimenis I, 
Gavriatopoulou M, et al. Comparison of MRI features of fat fraction and 
ADC for early treatment response assessment in participants with multiple 
myeloma. Radiology 2022; 304: 137-44. doi: 10.1148/radiol.211388
26. Sidiqi MH, Aljama M, Kumar SK, Jevremovic D, Buadi FK, Warsame R, et 
al. The role of bone marrow biopsy in patients with plasma cell disorders: 
should all patients with a monoclonal protein be biopsied? Blood Cancer J 
2020; 10: 52. doi: 10.1038/s41408-020-0319-0
27. Dupuis MM, Tuchman SA. Non-secretory multiple myeloma: from biology 
to clinical management. Onco Targets Ther 2016; 9: 7583-90. doi: 10.2147/
OTT.S122241
28. Wale A, Pawlyn C, Kaiser M, Messiou C. Frequency, distribution and clinical 
management of incidental findings and extramedullary plasmacytomas in 
whole body diffusion weighted magnetic resonance imaging in patients 
with multiple myeloma. Haematologica 2016; 101: e142-4. doi: 10.3324/
haematol.2015.139816
29. Hamdaoui H, Benlarroubia O, Ait Boujmia OK, Mossafa H, Ouldim K, 
Belkhayat A, et al. Cytogenetic and FISH analysis of 93 multiple myeloma 
Moroccan patients. Mol Genet Genomic Med 2020; 8: e1363. doi: 10.1002/
mgg3.1363
30. Saxe D, Seo EJ, Bergeron MB, Han JY. Recent advances in cytogenetic char-
acterization of multiple myeloma. Int J Lab Hematol 2019; 41: 5-14. doi: 
10.1111/ijlh.12882
31. Zhao XQ, Zhao SY, Chen WX, Liu XW, Yan HX, Lou YJ. Correlation between 
clinical factors and prognosis in newly diagnosed multiple myeloma. J Coll 
Physicians Surg Pak 2020; 30: 601-5. doi: 10.29271/jcpsp.2020.06.601
32. Usmani SZ, Crowley J, Hoering A, Mitchell A, Waheed S, Nair B, et al. 
Improvement in long-term outcomes with successive total therapy trials 
for multiple myeloma: are patients now being cured? Leukemia 2013; 27: 
226-32. doi: 10.1038/leu.2012.160
33. Ross FM, Avet-Loiseau H, Ameye G, Gutiérrez NC, Liebisch P, O’Connor S, 
et al. Report from the European Myeloma Network on interphase FISH in 
multiple myeloma and related disorders. Haematologica 2012; 97: 1272-7. 
doi: 10.3324/haematol.2011.056176