Published Ahead of Print on February 22, 2010 as 10.1200/JCO.2009.25.5356
The latest version is at http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2009.25.5356
JOURNAL OF CLINICAL ONCOLOGY
ORIGINAL
REPORT
Prognostic Significance of Focal Lesions in Whole-Body
Magnetic Resonance Imaging in Patients With
Asymptomatic Multiple Myeloma
Jens Hillengass, Kerstin Fechtner, Marc-Andre´ Weber, Tobias Ba¨uerle, Sofia Ayyaz, Christiane Heiss,
Thomas Hielscher, Thomas M. Moehler, Gerlinde Egerer, Kai Neben, Anthony D. Ho, Hans-Ulrich Kauczor,
Stefan Delorme, and Hartmut Goldschmidt
From the Department of Hematology
ABSTRACT
and Oncology; and Department of Diag-
nostic and Interventional Radiology,
University of Heidelberg; Department of
Purpose
With whole-body magnetic resonance imaging (wb-MRI), almost the whole bone marrow compart-
Radiology, German Cancer Research
Center; Department of Medical Physics
ment can be examined in patients with monoclonal plasma cell disease. Focal lesions (FLs) detected
in Radiology, German Cancer Research
by spinal MRI have been of prognostic significance in symptomatic multiple myeloma (sMM). In
Center; Department of Biostatistics,
this study, we investigated the prognostic significance of FLs in wb-MRI in patients with
German Cancer Research Center; and
asymptomatic multiple myeloma (aMM).
National Center for Tumor Diseases,
Heidelberg, Germany.
Patients and Methods
Wb-MRI was performed in 149 patients with aMM. The prognostic significance of the presence
Submitted August 7, 2009; accepted
December 3, 2009; published online
and absence, as well as the number, of FLs for progression into sMM was analyzed.
ahead of print at www.jco.org on
Results
February 22, 2010.
FLs were present in 28% of patients. The presence per se of FLs and a number of greater than
Supported in part by a grant from
one FL were the strongest adverse prognostic factors for progression into sMM (P
.001) in
Dietmar-Hopp-Stiftung.
multivariate analysis. A diffuse infiltration pattern in MRI, a monoclonal protein of 40 g/L or greater,
Presented in part at the 12th Interna-
and a plasma cell infiltration in bone marrow of 20% or greater were other adverse prognostic
tional Myeloma Workshop, February
factors for progression-free survival in univariate analysis.
26-March 1, 2009, Washington, DC.
Conclusion
Authors' disclosures of potential con-
We recommend use of wb-MRI for risk stratification of patients with asymptomatic multiple myeloma.
flicts of interest and author contribu-
tions are found at the end of this
article.
J Clin Oncol 28. © 2010 by American Society of Clinical Oncology
Corresponding author: Jens Hillengass,
MD, Department of Internal Medicine,
Among imaging methods, magnetic resonance
University Hospital Heidelberg, Im
INTRODUCTION
Neuenheimer Feld 410, Heidelberg,
imaging (MRI) has the highest sensitivity for detect-
Germany; e-mail: Jens.Hillengass@med
Asymptomatic multiple myeloma (aMM) is a disor-
ing bone marrow involvement in multiple my-
.uni-heidelberg.de.
der that originates from monoclonal plasma cells. It
eloma. Spinal MRI is recommended for clinical
© 2010 by American Society of Clinical
is defined by a serum concentration of a monoclonal
work-up of patients with monoclonal gammopathy
Oncology
protein greater than 30 g/L and/or 10% or greater
of undetermined significance, solitary plasmacy-
0732-183X/10/2899-1/$20.00
clonal plasma cells in bone marrow.1 Patients with
toma, and aMM.6 In sMM, greater than seven focal
DOI: 10.1200/JCO.2009.25.5356
aMM do not need the systemic treatment that pa-
lesions (FLs) detected by axial MRI (ie, spine and
tients with symptomatic multiple myeloma (sMM)
sacral bone) has been of adverse prognostic signifi-
would require.2 The annual progression rate into
cance.7 With fast sequences and a rolling table device
sMM is 10% per year for the first 5 years; the cumu-
available, MRI can be performed as a whole-body
lative probability of progression is 73% at 15 years,
protocol (wb-MRI).8 This is important, because not
but there is great interindividual variability.3 There-
examining the appendicular skeleton will miss ap-
fore, the identification of predictors of progression
proximately 50% of FLs,9 of which those will also be
into sMM is of great importance.
occult to plain film radiographs that have not yet
Until now, the most widely accepted clinical risk
caused destruction of mineralized bone. We hy-
factors for aMM for progression into sMM have been
pothesized that the presence of FLs as well as the
concentration and type of serum M protein, amount of
number would be of prognostic significance for
Bence Jones Protein excretion in urine, pattern and
progression-free survival in patients with aMM in a
percentage of bone marrow plasma cells, and reduc-
similar fashion as they have already been shown to
tion of uninvolved immunoglobulins (Igs).3-5
be in sMM.7
© 2010 by American Society of Clinical Oncology
1
Information downloaded from jco.ascopubs.org and provided by Universitaet Heidelberg on February 23, 2010 from
129.206.90.2.
Copyright
Copyright ©
2010
2010 by
by
the
American
American
Society
Society of
of
Clinical
Clinical
Oncology. AllOncology
rights reserved.

---

Hillengass et al
years. The start of systemic therapy was defined as the date of event for the
PATIENTS AND METHODS
analysis of progression-free survival.
Statistical Analysis
Patients
The search for an optimal cutoff point in number of FLs with respect to
From November 2003 to April 2008, a total of 149 individuals with aMM
progression-free survival into symptomatic disease was performed by using
(73 women and 76 men) were examined with wb-MRI. The median age was 58
maximally selected rank statistics.12-14
years (range, 25 to 81 years). Staging followed the classification proposed by
the International Myeloma Working Group, but this classification did not
Univariate Analysis
include MRI findings.1 In 16 patients, a localized plasma cell tumor plus
The significance of the number of focal FLs in wb-MRI greater than the
evidence of systemic disease according to other factors, such as elevated mono-
cutoff point, the presence of a diffuse infiltration pattern in MRI, and the
clonal protein or percentage of plasma cells in bone marrow, were present.
presence of established adverse prognostic markers for aMM for the progres-
None of the patients had signs of bone disease besides one with osteolysis
sion into sMM was analyzed. Time to progression into sMM was estimated by
caused by the plasma cell tumor, and none of the patients were treated system-
using the Kaplan-Meier method. Group comparisons were made by using a
ically until sMM occurred later. The only therapy was local irradiation with a
log-rank test. The prognostic value of a number of FLs greater than the cutoff
dose of 25 to 50 Gy in 14 patients and a tumor extirpation in two patients. Like
point for the probability of progression was analyzed in a Cox proportional
the other patients in this study, those patients would not have had an indica-
hazard regression model.
tion for a systemic treatment after definitive local therapy, such as irradiation
Multivariate Analysis
and/or surgery; hence, we decided to include them into this analysis. Patients
Risk factors analyzed in multivariate analysis were M protein concentra-
with true solitary plasmacytoma were excluded from this analysis, because
tion of
40 g/L; presence of an IgA monoclonal protein; reduction of unin-
other survival rates and prognostic factors for this entity have been described in
volved Igs; presence of urinary Bence Jones Protein and a plasma cell
the literature.
infiltration in bone marrow of
20%.3-5 MRI-derived risk factors were the
The median time of follow-up for the whole patient group was 23.7
number of FLs greater than the cutoff point and the presence or absence of
months, and information about patients was last updated in January 2009.
diffuse bone marrow involvement. Multivariate analysis was performed with a
Retrospective evaluation of wb-MRI data of patients with MM in our clinic
Cox proportional hazards regression model.
had been approved by the institutional ethics review board.
Backward variable selection was performed for the multivariate model
MRI Protocol
with a significance level for staying in the model of 0.2 to select the most
Wb-MRI was performed with two 1.5-Tesla, whole-body systems (Mag-
important risk factors for progression-free survival into symptomatic dis-
netom Avanto, Siemens Medical Solutions, Erlangen, Germany) with phased-
ease.15 Correlation analysis between the different risk factors was done by
array, body-matrix surface coils (Siemens Medical Solutions, Erlangen,
using the Kendall test.
Germany) that had the following parameters: T1-weighted turbo-spin echo
sequence (TR
627 milliseconds [ms]; TE
11 ms) of the head (voxel size,
1.25
1.25
5mm3; scan time, 2 minutes 4 seconds), thorax and abdomen
RESULTS
(voxel size, 1.25
1.25
5mm3; scan time, 2 minutes 4 seconds), pelvis
(1.25
1.25
5mm3; scan time, 2 minutes 4 seconds), and leg (voxel size,
Optimal Cutoff Point for the Number of FLs
1.25
1.25
5mm3; scan time, 2 minutes 4 seconds), all in coronal orienta-
Search for an optimal cutoff point of the number of FLs revealed
tion; T2-weighted short- inversion recovery (STIR) sequence (TR
3,340
that patients with greater than one FL had significantly shorter
ms; TE
109 ms; TI
160 ms) of the head (voxel size, 1.25
1.25
5mm3;
scan time, 1 minute 20 seconds), thorax and abdomen (voxel size, 1.25
progression-free survival than those without or with only one FL
1.25
5mm3; scan time, 1 minute 20 seconds), and pelvis (voxel size, 1.25
(P
.001). The Kaplan-Meier plot for progression into symptomatic
1.25
5mm3; scan time, 1 minute 20 seconds), all in coronal orientation;
disease is shown in Figure 1.
T1-weighted turbo spin echo sequence (TR
400 ms; TE
11 ms) of the
spine in sagittal orientation (voxel size, 1.836
1.836
3.5 mm3; scan time, 1
minute 16 seconds); T2-weighted FLASH 2D sequence (TR
402 ms; TE
12 ms) of the spine in sagittal orientation (voxel size, 0.84
0.84
5mm3;
scan time, 1 minute 38 seconds). The patients were positioned with arms along
1.0
1
their bodies, and the series covered the region between the skull vertex and the
aliv
> 1
midcalf. Depending on the body height of the patient, the distal calves and the
0.8
feet were not included.
Surv
Image Analysis
0.6
All MRI scans were read by two experienced radiologists (K.F., M.A.W.)
Free
in consensus who were blinded to the diagnosis of the patients and who
Log-rank P < .001
0.4
counted the number of FLs in the axial skeleton (ie, spine and sacral bone), the
extra-axial skeleton (ie, all other parts of the skeleton), and in soft tissue
0.2
separately. Furthermore, diffuse bone marrow infiltration in the axial skeleton
was recorded according to the criteria proposed by Staebler and Baur.10,11 FLs
Progression-
presented with decreased signal intensity in T1-weighted images and with
0
612
18
24
30
36
42
48
54
60
increased signal intensity in T2-weighted images. Diffuse infiltration was char-
acterized as homogeneous signal decrease in T1- and increase in T2-weighted
Time Since MRI Treatment (months)
images compared with the signal intensity of the vertebral disk. Lesions in
0 or 1 FL
126
106
81
64
49
36
20
11
3
1
1
typical locations for degenerative changes were not counted.
More than 1 FL 23
19
10
5
3
2
Evaluation of Progression-Free Survival
Follow-up took place every 3 to 6 months and included clinical history
Fig 1. Kaplan-Meier plots for progression into symptomatic myeloma of patients
who had no or one focal lesion (FL) compared with patients who had greater than
and examination as well as serum and urine markers. A radiologic skeletal
one FL. The median time to progression was not reached (last event at 43
survey (RSS) was obtained for occurrence of symptoms or an increase in
months) for the patient group with no or one FL and 13 months for the patient
monoclonal protein levels; the average interval between x-ray studies was 2
group with greater than one FL, respectively. MRI, magnetic resonance imaging.
2
© 2010 by American Society of Clinical Oncology
JOURNAL OF CLINICAL ONCOLOGY
Information downloaded from jco.ascopubs.org and provided by Universitaet Heidelberg on February 23, 2010 from
129.206.90.2.
Copyright © 2010 by the American Society of Clinical Oncology. All rights reserved.

---

Whole-Body MRI in Asymptomatic Multiple Myeloma
and pelvis only, the skeletal involvement by MM would have been
Table 1. Comparison of the Number of FLs and Bone Marrow Patterns
completely missed.
in wb-MRI
Bone Marrow Pattern
Disease Progression
Patterns by No.
of Focal Lesions
Normal
Diffuse
Comparison of the 16 patients with additional solitary plasma
No FL
69
38
cell tumor-- of which 14 also had corresponding osteolysis--with the
One FL
133 patients without any plasma cell tumor revealed no significant
Axial
6
3
differences in progression-free survival. Therefore, these patients were
Extra-axial
4
6
included in the analysis.
Greater than one FL
Axial
2
1
Univariate Analysis
Extra-axial
5
4
The univariate analysis revealed that the presence of FLs as well as
Axial
extra-axial
3
8
greater than one FL, diffuse bone marrow infiltration in MRI, a mono-
Abbreviations: FL, focal lesion; wb, whole body; MRI, magnetic reso-
clonal protein of greater than 40 g/L and a plasma cell infiltration in
nance imaging.
bone marrow of greater than 20% were adverse prognostic factors for
progression into symptomatic disease. Results for all analyzed param-
Incidence of FLs
eters are shown in Table 2.
In the whole patient group, FLs were detected in 42 (28%) of 149
Multivariate Analysis
patients (number of FL range, one to
20). The number of patients
Multivariate analysis of the MRI and non-MRI parameters re-
with FLs in the axial skeleton only was 12 (8%); in the extra-axial
vealed that the presence and the number of FLs as well as a diffuse bone
skeleton only, 19 (13%); and in both locations, 11 (7%). In the group
marrow infiltration in MRI remained the only significant adverse
with merely extra-axial lesions, 10 of 19 patients had just one FL. In the
prognostic factors for progression into symptomatic disease in our
study sample, individuals who had FLs in the extra-axial skeleton only
cohort when analysis was adjusted for the other prognostic parameters
were more frequent than those in whom FLs were only located in the
mentioned in the Patients and Methods section. Results of this analysis
axial skeleton (19 v 12 patients). Of 23 patients with greater than one
are listed in Table 3 for all risk factors (upper part) and for the most
focal lesion, nine patients had extra-axial lesions only, which would
significant risk factors determined by backward selection (lower part).
have been missed by axial MRI alone.
Clinical characteristics of the 23 patients with greater than one FL who
In 16 patients with initial plasma cell tumor, nine had merely one
developed necessity of systemic treatment are listed in Table 4.
of these lesions in wb-MRI, and seven showed additional FLs. In this
group, seven patients developed signs of disease progression with need
Correlations of Prognostic Parameters
for systemic treatment. From those patients, one had only a single
Analysis of the relationship of the MRI parameters and the
lesion, and the others had two or more FLs.
prognostic factors listed in Table 1 revealed no significant correla-
tions. Comparison of serum levels of 2 microglobulin or albumin,
FLs and Diffuse Infiltration Pattern
which are standard prognostic factors for sMM, in patient groups with
The groups of patients with different numbers of FLs and with or
or without FLs did not reveal a significant difference (P
.6 and
without diffuse infiltration pattern in bone marrow are listed in Table
P
.06, respectively).
1. In 60 patients (40%), a diffuse bone marrow infiltration was de-
tected in MRI whether or not FLs were present. Of 107 patients
without any FLs, 38 patients (36%) presented with a diffuse bone
DISCUSSION
marrow infiltration on MRI.
This study demonstrates, for the first time to our knowledge, the
Notably, nine of the 19 patients (6% of the whole patient group)
prognostic significance of FLs detected by wb-MRI in patients with
who had only extra-axial FLs had entirely normal spines and pelvises
(ie, no visible diffuse bone marrow involvement). Of these nine pa-
tients, four had only one FL. Thus, had MRI been limited to the spine
Table 3. Results of the Multivariate Analysis of All Variables and of
Selected Variables for Progression-Free Survival
Variable by Multivariate Analysis Type
Hazard Ratio
P
Full model
MRI-FL above cutoff point of one FL
3.01
.002
Table 2. Results of the Univariate Analysis for Progression-Free Survival
Diffuse bone marrow infiltration in MRI
2.37
.03
Univariate
Hazard Ratio
P
M protein concentration
40 g/L
1.87
.44
MRI-FL greater than cut-off point of one FL
4.05
.001
Presence of IgA
0.84
.71
Diffuse bone marrow infiltration in MRI
3.14
.001
Reduction of uninvolved Ig
1.03
.95
M-protein concentration
40 g/L
5.71
.005
Presence of urinary Bence Jones protein
0.94
.87
Presence of IgA
0.61
.26
Plasma cell infiltration in bone marrow
20%
1.30
.53
Reduction of uninvolved Ig
0.66
.35
Final model after backward selection
Presence of urinary Bence Jones Protein
1.34
.38
MRI-FL cutoff point
3.25
.001
Plasma cell infiltration in bone marrow
20%
2.14
.03
Diffuse bone marrow infiltration in MRI
2.64
.006
Abbreviations:
MRI,
magnetic
resonance
imaging;
FL,
focal
lesion;
Abbreviations:
MRI,
magnetic
resonance
imaging;
FL,
focal
lesion;
Ig, immunoglobulin.
Ig, immunoglobulin.
www.jco.org
© 2010 by American Society of Clinical Oncology
3
Information downloaded from jco.ascopubs.org and provided by Universitaet Heidelberg on February 23, 2010 from
129.206.90.2.
Copyright © 2010 by the American Society of Clinical Oncology. All rights reserved.

---

Hillengass et al
In all recent studies investigating the prognostic significance of
Table 4. Initial Clinical and Demographic Characteristics of the 23 Patients
MRI in plasma cell disorders, the vertebral column only, including
With FLs of 37 Patients Developing Necessity of Systemic Treatment
partly the sacral bone or pelvis as region where most MM lesions can
Patients
be expected, was scanned. Although this technique is easy to perform
Characteristic
No.
%
within relatively short examination times, it misses a significant num-
Non-MRI prognostic factor
ber of FLs.9 In our cohort, extra-axial FLs would have been missed in
M protein
40 g/L
2
9
30 (20%) of 149 patients if only an axial MRI had been performed. In
IgA
5
22
the 42 patients with proven FLs, 45% actually had exclusively extra-
Reduction of univolved Ig
18
78
axial lesions. In 1997, first results on the use of wb-MRI were pub-
Presence urinary Bence Jones Protein
13
57
Plasma cell infiltration in bone marrow
12
52
lished.8 Modern MRI scanners with affordable technical extensions
No. of non-MRI risk factors per patient
(ie, body matrix coil and moving table device) allowed scanning of the
02
9
whole body within reasonable time (ie, 30 to 45 minutes) and with
15
22
sufficient diagnostic resolution. The feasibility of this technique in
26
26
patients with sMM recently has been demonstrated.21,22
37
30
Especially for aMM, the degree of plasma cell infiltration detected
43
13
by unilateral bone marrow histology or aspirate has been of prognostic
50
0
Patient characteristic
significance.3,23 However, MM often involves the skeleton in a patchy,
Type of immunoglobulin
rather than homogeneous, fashion. Therefore, bone marrow histology
IgG
16
70
is potentially subject to sampling errors, if the histology is intended to
IgA
5
22
estimate the severity of the disease and the true burden of malignant
Bence Jones
2
9
plasma cells.
Age
50 years
19
83
In our study, 40% of patients had a diffuse infiltration and also
Female sex
14
61
had only a few FLs. Therefore, it was expected that the bone marrow
Abbreviations: FL, focal lesion; MRI, magnetic resonance imaging;
samples would be representative and that the bone marrow diagnostic
Ig, immunoglobulin.
Multiple mention is possible, as one patient can have more than one risk
also would be significant with regard to the prognosis.
factor and more than one characteristic, respectively.
The multivariate analysis, including the MRI parameters, re-
vealed, however, that monoclonal protein and plasma cell infiltration
were not independent (ie, not significant) when analysis was adjusted
aMM for progression into symptomatic disease for the first time. We
for the other parameters. Correlation analysis of the MRI parameters
found that the detection of more than one FL had the highest adverse
with the established prognostic factors revealed no significant rela-
prognostic significance in our cohort.
tionship. However, the P value for the correlation of M protein and FL
First investigations with MRI in MM were published in 1987 by
was .05 and, therefore, nearly missed the significance level. Because of
Ludwig et al16 in Lancet. Later studies with this technique in mono-
this relationship, the loss of significance of M protein may be caused by
clonal plasma cell disease revealed focal, diffuse, or variegated growth
a statistical effect.
patterns in bone marrow.10,17,18 Although most of those studies com-
End organ damage currently is the most important factor for the
pared normal versus abnormal bone marrow patterns, the investiga-
classification and the decision to treat systemically in monoclonal
tion by Moulopoulos et al18 was the only one to find that a diffuse
plasma cell disease. Therefore, serum calcium, renal damage, anemia,
infiltration is a bad prognostic sign. In this study, a diffuse bone
and bone destruction (ie, osteoporosis or focal lytic bone lesions) are
marrow infiltration was associated with a shorter progression-free
the most important parameters (ie, CRAB criteria1). Because the bone
survival. However, this kind of pattern is assessed subjectively as either
involvement often leads to fractures, pain, and immobilization with all
present or absent,11 and the signal intensity of normal vertebral disks
their sequelae, it can be considered the major clinical manifestation
was the reference. Any attempt at grading diffuse infiltration (eg, mild,
with respect to quality of life as well as survival.24
moderate, or severe) is, in our view, inadequate. Furthermore, the
For the detection of myeloma bone disease, RSS is in general still
aspect of bone marrow in MRI is influenced by several other factors,
considered the gold standard imaging method in sMM according to
like age or bone marrow activation, because of other causes like an
the recent consensus panel of the International Myeloma Workshop.
infectious disease.19,20 Despite these concerns, we found that the pres-
This technique displays osteolyses if a reduction of greater than 30% to
ence of a diffuse bone marrow infiltration in MRI was, indeed, relevant
50% of mineralized bone has occurred.25 The obvious limitations of
for prognosis. Additional developments, like diffusion-weighted im-
RSS and biopsy led to attempts to better assess the true extent of MM,
aging or other techniques, may contribute to a better quantification of
particularly by using modern cross-section imaging methods. Among
diffusely distributed plasma cells in bone marrow.
these MRI, computed tomography (CT), and positron emission to-
It was impressively demonstrated by Walker et al7 in 611 patients
mography (PET)/PET-CT have contributed to the assessment not
with sMM treated on the Arkansas Total Therapy 2 protocol that the
only of lytic bone lesions but also of bone marrow involvement with-
presence of seven or more FLs is an independent adverse prognostic
out bone destruction, and even of extramedullary disease.26-28 We
factor. These results raised the question if a nodular manifestation of
used MRI in this study, because it has been proven that MRI findings
MM in MRI is accompanied by a worse prognosis compared with
correlate highly significant with plasma cell infiltration in bone mar-
diffusely infiltrated or normal bone marrow. In our study, sample a
row.29 Furthermore, comparison of the different techniques demon-
cutoff point of more than one FL in wb-MRI was the best discrimina-
strated that MRI is superior to plain x-ray films, CT, and PET.21,22,30
tor between high and low risk for progression into sMM.
CT has a higher resolution than wb-MRI and may be better suitable to
4
© 2010 by American Society of Clinical Oncology
JOURNAL OF CLINICAL ONCOLOGY
Information downloaded from jco.ascopubs.org and provided by Universitaet Heidelberg on February 23, 2010 from
129.206.90.2.
Copyright © 2010 by the American Society of Clinical Oncology. All rights reserved.

---

Whole-Body MRI in Asymptomatic Multiple Myeloma
detect small lytic lesions in mineralized bone, but MRI has the highest
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS
sensitivity for bone marrow involvement in MM,22,26 even if the min-
OF INTEREST
eralized bone is still intact. Furthermore, the occurrences of a signifi-
cant lytic bone lesion that escapes detection by MRI will be rare.
The author(s) indicated no potential conflicts of interest.
It is unclear how to handle those patients in whom FLs in MRI are
detected but no other criteria supporting the initiation of systemic
therapy are found. In our study population, several patients had FLs
AUTHOR CONTRIBUTIONS
but did not develop any sign of end organ damage within a median
observation period of 23.7 months. Dinter et al31 examined the impact
Conception and design: Jens Hillengass, Tobias Ba¨uerle,
of wb-MRI findings on the decision process of the responsible physi-
Thomas M. Moehler
cian.31 However, the cohort was relatively small, and no prospective
Financial support: Hartmut Goldschmidt
Administrative support: Kai Neben, Anthony D. Ho,
comparison of treatment initiation versus a watch-and-wait strategy
Hartmut Goldschmidt
was made. Obviously, there is a need for prospective studies to inves-
Provision of study materials or patients: Jens Hillengass, Kerstin
tigate additional characteristics to support the initiation of systemic
Fechtner, Marc-Andre´ Weber, Gerlinde Egerer, Kai Neben, Anthony D.
antineoplastic therapy in this group of patients.
Ho, Hans-Ulrich Kauczor, Hartmut Goldschmidt
In conclusion, the presence of FLs detected in wb-MRI, as well as
Collection and assembly of data: Jens Hillengass, Kerstin Fechtner,
their number, are highly significant adverse prognostic factors for
Marc-Andre´ Weber, Tobias Ba¨uerle, Sofia Ayyaz
patients with aMM. We recommend performing wb-MRI for risk
Data analysis and interpretation: Jens Hillengass, Christiane Heiss,
Thomas Hielscher, Stefan Delorme
stratification in this group of patients. However, it needs to be exam-
Manuscript writing: Jens Hillengass, Stefan Delorme,
ined prospectively whether patients in whom MRI shows lesions will
Hartmut Goldschmidt
benefit from a treatment that they would not receive if only standard
Final approval of manuscript: Jens Hillengass, Marc-Andre´ Weber,
imaging protocols were used.
Stefan Delorme, Hartmut Goldschmidt
10. Sta¨bler A, Baur A, Bartl R, et al: Contrast
infiltration in patients with plasma cell neoplasms in
REFERENCES
enhancement and quantitative signal analysis in MR
comparison to the radiological skeletal survey. Eur
imaging of multiple myeloma: Assessment of focal
Radiol 16:1005-1014, 2006
1. Criteria for the classification of monoclonal
and diffuse growth patterns in marrow correlated
22. Baur-Melnyk A, Buhmann S, Becker C, et al:
gammopathies, multiple myeloma and related disor-
with biopsies and survival rates. AJR Am J Roent-
Whole-body MRI versus whole-body MDCT for stag-
ders: A report of the International Myeloma Working
genol 167:1029-1036, 1996
ing of multiple myeloma. AJR Am J Roentgenol
Group. Br J Haematol 121:749-757, 2003
11. Baur A, Stabler A, Nagel D, et al: Magnetic
190:1097-1104, 2008
2. Dispenzieri A, Rajkumar SV, Gertz MA, et al:
resonance imaging as a supplement for the clinical
23. Cavo M, Baccarani M, Gobbi M, et al: Prognostic
Treatment of newly diagnosed multiple myeloma
staging system of Durie and Salmon? Cancer 95:
value of bone marrow plasma cell infiltration in stage I
based on Mayo Stratification of Myeloma and Risk-
1334-1345, 2002
multiple myeloma. Br J Haematol 55:683-690, 1983
Adapted Therapy (mSMART): Consensus state-
12. Lausen B, Schumacher M: Maximally Se-
24. Berenson JR: Myeloma bone disease. Best
ment. Mayo Clin Proc 82:323-341, 2007
lected Rank Statistics. Biometrics 48:73-85, 1992
Pract Res Clin Haematol 18:653-672, 2005
3. Kyle RA, Remstein ED, Therneau TM, et al: Clini-
13. Hothorn T, Lausen B: On the Exact Distribu-
25. Edelstyn GA, Gillespie PJ, Grebbell FS: The radio-
cal course and prognosis of smoldering (asymptomatic)
tion of Maximally Selected Rank Statistics. Compu-
logical demonstration of osseous metastases: Experi-
multiple myeloma. N Engl J Med 356:2582-2590, 2007
tional Statistics and Data Analysis 43:121-137, 2003
mental observations. Clin Radiol 18:158-162, 1967
4. Weber DM, Dimopoulos MA, Moulopoulos
14. Lausen B, Hothorn T: Optimally Selected Prognos-
26. Mahnken AH, Wildberger JE, Gehbauer G, et
tic Factors. Biometrical Journal 2004:364-374, 2004
al: Multidetector CT of the spine in multiple my-
LA, et al: Prognostic features of asymptomatic mul-
15. Lawless JF, Singhal K: Efficient screening of
eloma: Comparison with MR imaging and radiogra-
tiple myeloma. Br J Haematol 97:810-814, 1997
nonnormal regression models Biometrics 34:318-
phy. AJR Am J Roentgenol 178:1429-1436, 2002
5. Cesana C, Klersy C, Barbarano L, et al: Prog-
327, 1978
27. Horger M, Claussen CD, Bross-Bach U, et al:
nostic factors for malignant transformation in mono-
16. Ludwig H, Fruhwald F, Tscholakoff D, et al:
Whole-body low-dose multidetector row-CT in the
clonal gammopathy of undetermined significance
Magnetic resonance imaging of the spine in multiple
diagnosis of multiple myeloma: An alternative to con-
and smoldering multiple myeloma. J Clin Oncol
myeloma. Lancet 2:364-366, 1987
ventional radiography. Eur J Radiol 54:289-297, 2005
20:1625-1634, 2002
17. Lecouvet FE, Vande Berg BC, Michaux L, et
28. Durie BG, Waxman AD, D'Agnolo A, et al:
6. Dimopoulos M, Terpos E, Comenzo RL, et al:
al: Stage III multiple myeloma: Clinical and prognos-
Whole-body (18)F-FDG PET identifies high-risk my-
International myeloma working group consensus
tic value of spinal bone marrow MR imaging. Radi-
eloma. J Nucl Med 43:1457-1463, 2002
statement and guidelines regarding the current role of
ology 209:653-660, 1998
29. Baur A, Bartl R, Pellengahr C, et al: Neovas-
imaging techniques in the diagnosis and monitoring of
18. Moulopoulos LA, Gika D, Anagnostopoulos A,
cularization of bone marrow in patients with diffuse
multiple Myeloma. Leukemia 23:1545-1556, 2009
et al: Prognostic significance of magnetic resonance
multiple myeloma: A correlative study of magnetic
7. Walker R, Barlogie B, Haessler J, et al: Mag-
imaging of bone marrow in previously untreated
resonance imaging and histopathologic findings.
netic resonance imaging in multiple myeloma: Diag-
patients with multiple myeloma. Ann Oncol 16:
Cancer 101:2599-2604, 2004
nostic and clinical implications. J Clin Oncol 25:
1824-1828, 2005
30. Hur J, Yoon CS, Ryu YH, et al: Comparative
1121-1128, 2007
19. Moore SG, Dawson KL: Red and yellow mar-
study of fluorodeoxyglucose positron emission to-
8. Johnson KM, Leavitt GD, Kayser HW: Total-
row in the femur: Age-related changes in appear-
mography and magnetic resonance imaging for the
body MR imaging in as little as 18 seconds. Radiol-
ance at MR imaging. Radiology 175:219-223, 1990
detection of spinal bone marrow infiltration in un-
ogy 202:262-267, 1997
20. Ricci C, Cova M, Kang YS, et al: Normal
treated patients with multiple myeloma. Acta Radiol
9. Ba¨uerle T, Hillengass J, Fechtner K, et al: The
age-related patterns of cellular and fatty bone mar-
49:427-435, 2008
importance of whole body as compared with spinal
row distribution in the axial skeleton: MR imaging
31. Dinter DJ, Neff WK, Klaus J, et al: Comparison
magnetic resonance imaging in patients with multi-
study. Radiology 177:83-88, 1990
of whole-body MR imaging and conventional X-ray
ple myeloma and monoclonal gammopathy of unde-
21. Ghanem N, Lohrmann C, Engelhardt M, et al:
examination in patients with multiple myeloma and
termined significance. Radiology 252:477-485, 2009
Whole-body MRI in the detection of bone marrow
implications for therapy. Ann Hematol 88:457-464, 2009

www.jco.org
© 2010 by American Society of Clinical Oncology
5
Information downloaded from jco.ascopubs.org and provided by Universitaet Heidelberg on February 23, 2010 from
129.206.90.2.
Copyright © 2010 by the American Society of Clinical Oncology. All rights reserved.

---

Hillengass et al
Acknowledgment
We thank Axel Benner and Lutz Edler from the Department of Biostatistics of the German Cancer Research Center for their
statistical supervision.
Appendix
In our cohort of 149 patients during the follow-up time, 37 patients developed symptoms that led to an initiation of systemic
treatment. In detail, 33 patients were treated because of occurrence of CRAB symptoms (renal insufficiency, n
4; anemia, n
4; bone
disease in conventional radiography, n
25). The other patients were treated because of clinical symptoms, as pain caused by
extra-osseous tumor parts (n
2), development of systemic AL-amyloidosis (n
1), and increasing immunoglobulin in serum greater
than 50 g/L (n
1) to prevent complications.
6
© 2010 by American Society of Clinical Oncology
JOURNAL OF CLINICAL ONCOLOGY
Information downloaded from jco.ascopubs.org and provided by Universitaet Heidelberg on February 23, 2010 from
129.206.90.2.
Copyright © 2010 by the American Society of Clinical Oncology. All rights reserved.

---