The

n e w e ng l a n d j o u r na l

of

m e dic i n e

original article

Clinical Course and Prognosis of Smoldering

(Asymptomatic) Multiple Myeloma
Robert A. Kyle, M.D., Ellen D. Remstein, M.D., Terry M. Therneau, Ph.D.,
Angela Dispenzieri, M.D., Paul J. Kurtin, M.D., Janice M. Hodnefield, M.S.,
Dirk R. Larson, M.S., Matthew F. Plevak, B.S., Diane F. Jelinek, Ph.D.,
Rafael Fonseca, M.D., Lee Joseph Melton III, M.D.,
and S. Vincent Rajkumar, M.D.

A bs t r ac t
Background
From the Division of Hematology (R.A.K.,
A.D., R.F., S.V.R.), the Department of Laboratory Medicine and Pathology (E.D.R.,
P.J.K., J.M.H.), the Division of Biostatistics (T.M.T., D.R.L., M.F.P.), the Department of Immunology (D.F.J.), and the Department of Health Sciences Research
(L.J.M.), Mayo Clinic, Rochester, MN. Address reprint requests to Dr. Kyle at the
Division of Hematology, Mayo Clinic, 200
First St. SW, Rochester, MN 55905, or at
kyle.robert@mayo.edu.
N Engl J Med 2007;356:2582-90.
Copyright 2007 Massachusetts Medical Society.

Smoldering (asymptomatic) multiple myeloma is an asymptomatic plasma-cell proliferative disorder associated with a high risk of progression to symptomatic multiple
myeloma or amyloidosis. Prognostic factors for the progression and outcome of this
disease are unclear.
Methods

We searched a computerized database and reviewed the medical records of all patients
at Mayo Clinic who fulfilled the criteria of the International Myeloma Working Group
for the diagnosis of smoldering multiple myeloma between 1970 and 1995. Bone marrow aspirate and biopsy specimens were studied, and patients were followed throughout the course of disease.
Results

During the 26-year period, 276 patients fulfilled the criteria for smoldering multiple
myeloma. During 2131 cumulative person-years of follow-up, symptomatic multiple
myeloma or amyloidosis developed in 163 persons (59%). The overall risk of progression was 10% per year for the first 5 years, approximately 3% per year for the next
5 years, and 1% per year for the last 10 years; the cumulative probability of progression
was 73% at 15 years. At diagnosis, significant risk factors for progression included
the serum level and type of monoclonal protein, the presence of urinary light chain,
the extent and pattern of bone marrow involvement, and the reduction in uninvolved
immunoglobulins. The proportion of plasma cells in the bone marrow and the serum
monoclonal protein level were combined to create a risk-stratification model with three
distinct prognostic groups.
Conclusions

The risk of progression from smoldering multiple myeloma to symptomatic disease

is related to the proportion of bone marrow plasma cells and the serum monoclonal
protein level at diagnosis.

2582

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.

Clinical Course and Prognosis of Smoldering Multiple Myeloma

moldering multiple myeloma is an

asymptomatic proliferative disorder of plasma cells with a high risk of progression to
symptomatic, or active, multiple myeloma. Previous studies have used various definitions of the
disease, and this variability has resulted in important differences in the reported clinical course of
the disease.1-7 International consensus criteria for
smoldering multiple myeloma were adopted recently to rectify this problem.8
We report here on the prognosis and risk factors for progression of smoldering multiple myeloma in a large cohort of patients for whom longterm follow-up data were available and in whom
the disease was defined with the use of criteria
of the International Myeloma Working Group.
These criteria do not include karyotyping, fluorescence in situ hybridization (FISH), or magnetic
resonance imaging for detection of bone lesions.8
Instead, the group defined smoldering multiple
myeloma as a disorder in which the patient has
a serum monoclonal protein level of 3 g per deciliter or more or a proportion of plasma cells in the
bone marrow of 10% or more but no end-organ
damage. The diagnosis of active multiple myeloma
requires the presence of a monoclonal protein in
serum or urine, plasma cells in the bone marrow,
or plasmacytoma and end-organ damage related
to plasma-cell proliferation. End-organ damage
was defined as hypercalcemia, renal insufficiency,
anemia, bone lesions, or recurrent bacterial infections.8
The study cohort was stratified into three prognostic groups at initial diagnosis: group 1 had a
proportion of bone marrow plasma cells of 10%
or more and a serum monoclonal protein level of
3 g per deciliter or more; group 2, 10% or more
bone marrow plasma cells and a serum monoclonal protein level of less than 3 g per deciliter; and
group 3, less than 10% bone marrow plasma cells
and a serum monoclonal protein level of 3 g per
deciliter or more.

Me thods

or more or bone marrow containing 10% or more

plasma cells. Patients were evaluated from 1970 to
1995, an interval allowing for a minimum potential follow-up of 10 years. Patients with active multiple myeloma or primary amyloidosis were excluded, as were patients who had ever received
chemotherapy.
Bone Marrow Examination

We estimated the proportion of plasma cells in

bone marrow aspirates and in paraffin-embedded
bone marrow biopsy specimens stained with hematoxylin and eosin, with a differential count of
200 to 500 cells. In addition, immunoperoxidase
stains were applied to paraffin sections of the bone
marrow biopsy specimens according to previously
published methods9 with the use of antibodies directed against CD138, an antigen normally expressed on the surface of plasma cells (Dako Cytomation), multiple myeloma oncogene 1 (MUM1,
Dako Cytomation), and cyclin D1 (Biocare). Cyclin
D1 expression is thought to be a surrogate for the
translocation t(11;14)(q14;q32); CD138 is a cytoplasmic and Golgi stain, and MUM1 is a nuclear
stain. The bone marrowaspirate, biopsy, and immunoperoxidase stains were reviewed by one of
two hematopathologists.
Plasma-cell estimates by all three methods were
combined to provide the proportion of bone marrow plasma cells, categorized as 0 to 4%, 5 to 9%,
10 to 14%, and so on. Cases with discordance between the plasma-cell estimates in bone marrow
aspirate and biopsy specimens were reviewed individually, and the final plasma-cell estimate represented an average of the aspirate and biopsy estimates. The pattern of involvement of plasma cells
in bone marrow (singly distributed cells, small
clusters of cells, cells filling at least one interfatty
marrow space, or sheets of cells spanning the interfatty spaces)10 and the overall bone marrow
cellularity were recorded in each case. Monoclonal
proteins were identified with cellulose acetate or
agarose-gel electrophoresis in combination with
immunoelectrophoresis or immunofixation.11

Study Cohort

Follow-up and Study End Point

After the study was approved by the Mayo Clinics

institutional review board, we searched a computerized database and reviewed the medical records
of all patients who had been seen at the Mayo
Clinic within 30 days after detection of an IgG or
IgA monoclonal protein level of 3 g per deciliter

Follow-up included a review of the medical records

of patients at the Mayo Clinic and of death certificates for patients who had died. Patients were sent
letters of inquiry if they had not visited the Mayo
Clinic in the preceding year. The primary study end
point was progression to active multiple myeloma

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.

2583

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

Figure 1. Bone MarrowBiopsy Specimens from Two Patients with Smoldering Multiple Myeloma.
In Panel A, in which bone marrow cells from one patient have been stained with hematoxylin and eosin, plasma
cells are difficult to discern by morphologic analysis alone. In Panel B, cells from the same specimen have underRETAKE
AUTHOR
ICM
Kylenuclear staining is
gone immunohistochemical staining
for MUM1.
Strong
present 1st
in the plasma cells, but this pat2nd
REG
F
FIGURE
f
1
a_e
tern leads to a slight underestimation of the number of plasma cells. In Panel C, immunohistochemical
staining for
3rd
CASE
TITLE
CD138 on the same specimen shows
strong
cytoplasmic and Golgi stainingRevised
in plasma cells, but this pattern results
EMail
Line it is essential
4-C
in a slight overestimation of the number
of plasma cells. Thus,
to combine the estimates of all methSIZE
Enon
mleahyIn Panel
H/T D, a H/T
ods to estimate the number of plasma cellsARTIST:
accurately.
specimen33p9
from another patient with smolderFILL
Combo
ing multiple myeloma shows monotonous plasma-cell aggregates filling interfatty marrow spaces (hematoxylin and
AUTHOR,
PLEASE
NOTE:
eosin). In Panel E, immunohistochemical staining
for CD138
on the
same specimen shows plasma-cell aggregates
Figure has been redrawn and type has been reset.
filling interfatty marrow spaces.
Please check carefully.
JOB:

35625

(anemia, hypercalcemia, renal insufficiency, or

bone lesions) or amyloidosis (positive results on
Congo red staining and characteristic clinical features), with either one requiring therapy.
Statistical Analysis

Progression was calculated in terms of both the

cumulative probability and the cumulative incidence of progression. The cumulative probability
was calculated by means of a KaplanMeier analysis12 in which data from patients who had died
were censored; curves were compared with use of
2584

ISSUE:

6-21-07

the log-rank test.13 The cumulative incidence curve,

which explicitly accounted for death as a competing risk, was computed with the method of
Gooley et al.14 Effects of potential risk factors on
progression rates were examined in a Cox proportional-hazards model.15
The risk of progression to active multiple myeloma or amyloidosis, as compared with that in the
general population, was assessed with standardized incidence ratios,16 whereby observed cases
were compared with the number expected by applying age-specific and sex-specific incidence rates

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.

Clinical Course and Prognosis of Smoldering Multiple Myeloma

Table 1. Relative Risk of Progression among 276 Patients with Smoldering Multiple Myeloma, According to Prognostic Group.*
Progression

No. of Patients

Expected Rate of Disease

Relative Risk (95% CI)

Multiple myeloma
All patients

157

0.301

522 (443610)

Group 1

75

0.072

1038 (8171302)

Group 2

72

0.174

413 (323520)

Group 3

10

0.054

184 (88338)

All patients

0.100

50 (16117)

Group 1

0.022

136 (28402)

Group 2

0.060

33 (4121)

Group 3

0.019

0 (0198)

162

0.401

404 (344471)

Amyloidosis

Total

* Patients were divided into three prognostic groups: group 1 (bone marrow plasma cells, 10%; monoclonal protein level,
3 g per deciliter), group 2 (plasma cells, 10%; monoclonal protein level, <3 g per deciliter), and group 3 (plasma cells,
<10%; monoclonal protein level, 3 g per deciliter). The 162 patients with disease progression were followed for a total
of 641 person-years. CI denotes confidence interval.
The expected rate of multiple myeloma was obtained from the Iowa Surveillance, Epidemiology, and End Results program
(19732002).17 The expected rate of amyloidosis was calculated from data for patients in Olmsted County, Minnesota.18
The reference group for the relative risk of progression was persons without smoldering multiple myeloma.
One patient had progression, but the date of progression was not available; therefore, this patient was excluded from
the analysis of progression.

for multiple myeloma in the white cohort from the

Iowa Surveillance, Epidemiology, and End Results
program17 to the person-years of follow-up specific for age, sex, and calendar year in our study
cohort. The age-specific and sex-specific incidence
rates of amyloidosis were based on data from Olm
sted County, Minnesota.18

R e sult s
Characteristics of the Patients

Of the 3549 patients with myeloma diagnosed between 1970 and 1995, 276 (8%) fulfilled the criteria
for smoldering multiple myeloma. The median age
at diagnosis was 64 years (range, 26 to 90), and
only eight patients (3%) were younger than 40 years
of age. A total of 171 patients (62%) were men, and
105 (38%) were women.
Laboratory and Bone Marrow Findings

The initial hemoglobin level ranged from 10.0 to

16.8 g per deciliter (median, 13.0). The hemoglobin level was 12 g per deciliter or more in 76% of
patients.
The serum monoclonal protein level at the time
of diagnosis ranged from 0.5 to 5.4 g per deciliter (median, 2.9). The levels for 11% of the patients
were 4 g per deciliter or more (including 1% with

5 g per deciliter or more), those for 37% were 3 to

3.9 g per deciliter, and those for 52% were less
than 3 g per deciliter. The monoclonal protein
level in the 27 patients in group 3 (with a monoclonal protein spike of 3 g per deciliter and <10%
plasma cells) ranged from 3.0 to 3.9 g per deciliter (median, 3.1). Of these 27 patients, 26 had
monoclonal IgG present in serum, whereas only
2 had a urinary monoclonal protein level of more
than 200 mg per 24 hours (240 and 450 mg per
24 hours). Of the 276 patients, 74% had monoclonal IgG, 22.5% had IgA, 0.5% had IgD, and 3%
had biclonal immunoglobulins. The light-chain
type was in 67% and in 33%. Concentrations
of uninvolved (normal, polyclonal, or background)
immunoglobulins were reduced in 83% of 230
patients whose immunoglobulin levels were determined quantitatively. A single reduction in immunoglobulin was found in 31% of patients, whereas
52% had a reduction of both uninvolved immunoglobulins (e.g., reduced IgM and IgA in a patient
with a monoclonal IgG).
Immunoelectrophoresis or immunofixation was
performed on urine samples from 259 of the patients (94%). Of those patients, 92 (36%) had a
monoclonal light chain, 43 (17%) had a light
chain, and 123 (47%) had negative results for
monoclonal light chain. A urinary monoclonal

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.

2585

The

n e w e ng l a n d j o u r na l

of

m e dic i n e

of persons without smoldering multiple myeloma

who would be expected to have active disease, and
Smoldering Multiple Myeloma
the risk of amyloidosis was increased by a factor of
50 (Table 1). Of 16 patients with DurieSalmon
80
78
stage II disease, 15 had progression to active mul73
tiple myeloma; both patients with DurieSalmon
66
60
stage III had progression to active multiple my51
eloma (median time to progression, 13.8 months
40
and 15.1 months). Among patients with progresMGUS
sion of smoldering multiple myeloma, 97% had
progression to active multiple myeloma. Rates of
20
21
death owing to other diseases, including cardio16
10
4
vascular and cerebrovascular disease and non
0
plasma-cell cancers, were 18% at 5 years, 26% at
0
5
10
15
20
25
10 years, 30% at 15 years, and 35% at 20 years. The
Years since Diagnosis
overall rate of survival was 60% at 5 years, 34% at
Figure 2. Probability of Progression to Active Multiple Myeloma or Primary
10 years, and 20% at 15 years (median, 6.3). In the
RETAKE
AUTHOR:
Kyle
Amyloidosis inICM
Patients
with Smoldering
Multiple Myeloma
or 1st
Monoclonal
128 patients who did not have progression to ac2nd
4
F FIGURE: 2 ofSignificance
Gammopathy REG
of Undetermined
(MGUS).
3rd
tive disease after more than 10 years after diagCASE confidence intervals.
Revised
I bars denote 95%
nosis, the disease was of the IgG subtype in 81%,
Line
4-C
EMail
SIZE
ARTIST: ts
H/T
H/T
the baseline median level of plasma cells in the
22p3
Enon
Combo
bone marrow was 16%, the median spike in the
protein AUTHOR,
level of 0.1
g
or
less
per
24
hours
was
found
PLEASE NOTE:
in 84%;
only
fourandpatients
(1.5%)
Figure
has been
redrawn
type has been
reset. had a level of serum monoclonal protein level was 2.8 g per
Please check carefully.
deciliter, and uninvolved immunoglobulins were
more than
1.0 g per 24 hours.
reduced in 78%.
The
bone
marrow
aspirate
and
biopsy
speciJOB: 35625
ISSUE: 06-21-07
mens were examined in all 276 patients (Fig. 1).
In the plasma-cell category, the most common Risk Factors for Progression
proportion was 15 to 19%. Of the 276 patients, We evaluated baseline factors with respect to pro10% had less than 10% plasma cells in the mar- gression of smoldering multiple myeloma to active
row, and 10% had 50% or more plasma cells. Cy- disease or amyloidosis in 163 patients. These facclin D1 was expressed by the plasma cells in 18% tors included sex, hemoglobin level, a spike in the
serum monoclonal protein level of 4 g per deciliter
of the bone marrow specimens.
or more, the type of serum heavy chain, the serum
Outcome
albumin level, the presence and type of urinary
During 2131 cumulative person-years of follow-up light chain, a reduction in levels of uninvolved im(range, 0 to 29; median, 6.1), 85% of the patients munoglobulins, the expression of cyclin D1, the
with smoldering multiple myeloma died (median proportion of plasma cells in the bone marrow, infollow-up of those still living, 11.6 years). During volvement of the interfatty marrow space, decreased
this period, active multiple myeloma developed in proportion of normal hematopoietic elements (10%
158 patients (57%), who had a median survival af- or more below the expected level for age), and aster the time of diagnosis of 3.4 years; amyloidosis signment to prognostic group 1, 2, or 3 (as defined
developed in 5 (2%) (Table 1). The cumulative prob- in the Methods section). Significant baseline risk
ability of progression to active multiple myeloma factors for progression of smoldering multiple
or amyloidosis was 51% at 5 years, 66% at 10 years, myeloma to active disease or amyloidosis in the
and 73% at 15 years; the median time to progres- univariate analysis included the level of serum
sion was 4.8 years (Fig. 2). The overall risk of pro- monoclonal protein (P<0.001), the presence of IgA
gression was 10% per year for the first 5 years, ap- monoclonal protein (P=0.004), the presence of uriproximately 3% per year for the next 5 years, and nary light chain (P=0.04), the extent of bone marrow involvement (plasma cells, 20%; P<0.001),
1% per year for the last 10 years.
The number of patients with progression to ac- a reduction in levels of uninvolved immunoglobtive multiple myeloma was 522 times the number ulins (P=0.001), and the pattern of plasma-cell
Probability of Progression (%)

100

2586

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.

Clinical Course and Prognosis of Smoldering Multiple Myeloma

Table 2. Risk Factors for Disease Progression among 276 Patients with Smoldering Multiple Myeloma (19701995).*

Variable

Patients

Events

Median
Time to
Progression

P Value

Median Rate of Progression

Univariate Multivariate
Analysis
Analysis
no.

mo

Prognostic group
106

78

Group 2

143

Group 3

27

IgA
IgG

10 yr

15 yr

%
<0.001

Group 1

<0.001

27

69

77

87

74

93

43

64

70

10

228

15

33

39

62

46

27

66

77

80

205

112

75

46

62

71

Serum heavy chain

0.004

Bone marrow plasma cells

<20%

5 yr

<0.001

0.01

0.005

164

78

117

36

53

61

2050%

84

65

26

68

82

92

>50%

27

19

21

85

93

NA

47

64

71

80

80

90

Serum monoclonal protein

<0.001

<4 g/dl

244

138

75

4 g/dl

31

24

18

Urinary light chain

0.04

0.04

0.09

Negative

123

71

89

46

62

68

or

135

84

49

56

70

80

123

71

89

46

62

68

92

57

60

51

67

82

43

27

32

68

77

77

Type of urinary light chain

Negative

0.08

<0.001

Reduction of uninvolved immuno

globulins (no.)
0

0.12

0.003

40

16

159

24

33

59

71

35

89

38

59

62

119

85

32

68

81

84

Pattern of involvement of bone

marrow plasma cells

<0.001

0.02

Interfatty space or sheets

103

77

28

68

79

88

Singly distributed cells or

small clusters

160

78

103

40

57

63

276

162

58

51

66

73

Total

* Patients were divided into three prognostic groups: group 1 (bone marrow plasma cells, 10%; monoclonal protein level, 3 g per deciliter),
group 2 (plasma cells, 10%; monoclonal protein level, <3 g per deciliter), and group 3 (plasma cells, <10%; monoclonal protein level, 3 g
per deciliter). NA denotes not applicable.
The multivariate P values for the type of heavy chain represent the test for an additional significant contribution to the predictive ability of a
model already containing the prognostic groups; the multivariate P values for the remaining variables represent the test for an additional
significant contribution when added individually to a model already containing prognostic groups and the type of heavy chain.
One patient had progression, but the date of progression was not available; therefore, this patient was excluded from the analysis of progression.

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.

2587

The

Probability of Progression (%)

100

87

n e w e ng l a n d j o u r na l

Group 1

77

80

69

70

70

64
60

54
43

40

Group 2
Group 3

39
33

20

15
P<0.001

10

15

20

25

Years since Diagnosis

Figure 3. Probability of Progression to Active Multiple Myeloma or Primary

RETAKE
1st
AUTHOR:
Kyle
Amyloidosis inICM
Patients
with Smoldering
Multiple Myeloma
among
Three
2nd
FIGURE:
3
of
4
Risk Groups. REG F
3rd
CASE
Revised
The proportion
of plasma cells in bone marrow and the
serum monoclonal
4-C
EMail combined to create Line
SIZE
protein level were
a
risk-stratification
model with three
ARTIST: ts
H/T
H/T
22p3
Enon groups. At 15 years, the cumulative probability
distinct prognostic
of proCombo
gression was 87% for the 106 patients in group 1 (bone marrow plasma
AUTHOR, PLEASE NOTE:
cells, 10%; monoclonal
protein
level, 3
perhas
deciliter),
70% for the 142
Figure has
been redrawn
andgtype
been reset.
patients in group 2 (plasma cells,
monoclonal protein level, <3 g per
Please10%;
check carefully.
deciliter), and 39% for the 27 patients in group 3 (plasma cells, <10%;
monoclonalJOB:
protein
35625level, 3 g per deciliter).
ISSUE: 06-21-07

involvement in bone marrow (sheets of cells spanning the interfatty marrow spaces) (P<0.001) (Table 2).
On the basis of a fitted model containing the
proportion of plasma cells in the bone marrow and
the serum monoclonal protein level, the risk of
progression to active multiple myeloma or amyloidosis at 10 years was 55% for patients with an
initial plasma-cell level of 10 to 14%; progression
occurred in 70% of the 27 patients who had more
than 50% plasma cells in bone marrow (median
time to progression, 21 months) (Table 2). The risk
of progression to active multiple myeloma or amyloidosis at 10 years was 57% in patients with an
initial monoclonal protein level of 2 g per deciliter
and 70% in those whose initial level was 5 g per
deciliter. Such factors as hemoglobin, sex, serum
albumin level, proportion of normal hematopoietic elements, and expression of cyclin D1 were
not significantly associated with progression.
On multivariate analysis, the serum monoclonal protein level and the proportion of plasma cells
in the bone marrow emerged as significant independent risk factors for progression. On the basis
of these findings and the fact that these two variables were also the main components of the defi2588

of

m e dic i n e

nition of smoldering multiple myeloma, we constructed a risk-stratification model containing three

risk groups. The cumulative probability of progression at 15 years was 87% for the 106 patients in
group 1 (10% plasma cells and 3 g of monoclonal protein per deciliter), 70% for the 142 patients
in group 2 (10% plasma cells and <3 g of monoclonal protein per deciliter), and 39% for the 27
patients in group 3 (<10% plasma cells and 3 g
of monoclonal protein per deciliter) (Fig. 3). The
median time to progression was 2 years in group
1, 8 years in group 2, and 19 years in group 3
(P<0.001). The type of serum heavy chain added
significantly to the multivariate model containing the three prognostic risk groups (Table 2).

Dis cus sion

The currently accepted definition of smoldering
multiple myeloma requires a serum monoclonal
protein level of 3 g per deciliter or more, 10% or
more plasma cells in the bone marrow, or both in
the absence of end-organ damage. It would be interesting to correlate cytogenetic abnormalities
with the outcome of smoldering multiple myeloma,
but the bone marrow in patients with this disease
has few metaphases, and interphase FISH studies
are therefore necessary. FISH may identify patients
at increased risk, but since our cohort was studied before 1996, FISH was not available. Smoldering multiple myeloma resembles monoclonal gammopathy of undetermined significance (MGUS) in
that end-organ damage is absent, but clinically it
is far more likely to progress to active multiple myeloma or amyloidosis at 20 years (a 78% probability for smoldering multiple myeloma vs. 21% for
MGUS).19
MGUS is characterized by a serum monoclonal
protein level of less than 3 g per deciliter, a proportion of plasma cells in the bone marrow that is
less than 10%, and an absence of related organ or
tissue impairment (i.e., no end-organ damage such
as hypercalcemia, renal insufficiency, anemia, or
lytic bone lesions related to the plasma-cell disorder). MGUS and smoldering multiple myeloma are
asymptomatic and should not be treated. Active
multiple myeloma is characterized by a serum
monoclonal protein level that is usually more than
3 g per deciliter or a urinary monoclonal protein,
usually more than 10% plasma cells in the bone
marrow, and most important, end-organ damage
caused by plasma-cell proliferation (Fig. 4).
Estimates of the frequency of smoldering mul-

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.

Clinical Course and Prognosis of Smoldering Multiple Myeloma

Smoldering Multiple Myeloma

Multiple Myeloma

Serum Protein
Electrophoresis

MGUS

Alb 1 2

Bone Marrow
Clinical Picture
Therapy

Alb 1 2

Alb 1 2

<10% Plasma cells

10% Plasma cells

10% Plasma cells

Asymptomatic
No end-organ damage

Asymptomatic
No end-organ damage

Symptomatic
End-organ damage present

Observation only

Observation only

Therapy required

Figure 4. Characteristics of Active Multiple Myeloma and Its Precursors.

RETAKE
1st
AUTHOR: Kyle
ICM
Monoclonal gammopathy of undetermined
significance (MGUS) and smoldering multiple myeloma are asymp
2nd
FIGURE:
4
of
4
REG myeloma.
F
tomatic precursors of active multiple
Serum protein electrophoresis in each
3rd of these disorders shows a
CASE region that can be easily identified Revised
prominent monoclonal protein in the
as a dense band or spike. Other clinical
Line
4-C End-organ
EMail these disorders are
features that can be used to distinguish
also shown.
SIZE damage is defined as hypercalARTIST: ts
H/T to a plasma-cell
H/T
33p9
cemia, anemia, renal failure, or lytic
bone
lesions
attributable
disorder. Alb denotes albumin.
Enon
Combo

AUTHOR, PLEASE NOTE:

Figure has been redrawn and type has been reset.
tiple myeloma have been affected by the variabilmonoclonal
protein, although significant on uniPlease check
carefully.
ity in the criteria used to establish the diagnosis.3,6 variate analysis in our study, did not achieve staJOB: in
35625
Inconsistent diagnostic criteria
previous studies have resulted in the reporting of variable median times to progression.1-7 Our study shows that
the risk of progression is significantly affected by
the level of monoclonal protein, the proportion of
bone marrow plasma cells, or both. There were
also substantial differences in the median time to
progression among the three risk groups (2, 8, and
19 years).
Our study shows that the overall risk of progression in smoldering multiple myeloma is greatly influenced by the time elapsed since diagnosis,
in contrast to the risk of progression in MGUS,19
which remains constant over time. We found that
the overall risk of progression among patients with
smoldering multiple myeloma was approximately
10% per year in the first 5 years and 3% per year
in the next 5 years with a decrease to 1% per year
thereafter. No such time-dependent change in risk
occurs with MGUS.19
Other investigators have reported that the IgA
isotype and the presence of urinary monoclonal
protein are adverse prognostic factors for patients
with smoldering multiple myeloma.5,6,20 In contrast to these findings, the presence of urinary

ISSUE: 06-21-07in the multivariate analysis.

tistical significance
The roles of the serum free light-chain ratio and
magnetic resonance imaging in determining the
outcome of smoldering multiple myeloma are unknown and require investigation.
The age and sex of the patients in our study
and the distribution of heavy-chain and light-chain
types were similar to those in patients with active
multiple myeloma.21 However, other findings, such
as a reduction in the level of uninvolved immunoglobulins19,21 and the presence of monoclonal
urinary light chains, were intermediate between
those in patients with active multiple myeloma and
those in patients with MGUS.
On the basis of our experience, we suggest that
the standard of care for patients with smoldering
multiple myeloma should be close follow-up every
few months. Physicians should repeat the pertinent laboratory tests 2 to 3 months after the initial
recognition of the disease to rule out an early
active form; if the results are stable, the studies
should initially be repeated every 4 to 6 months.
However, given the high risk of progression among
patients in prognostic groups 2 and 3, along with
the availability of new active agents for the treat-

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.

2589

Clinical Course and Prognosis of Smoldering Multiple Myeloma

ment of active multiple myeloma, investigational

approaches may be considered for selected patients
in appropriate clinical trials. Ongoing trials are
testing the use of bisphosphonates, interleukin-1
inhibitors, clarithromycin, dehydroepiandrosterone,
and thalidomide in the treatment of smoldering
multiple myeloma in an attempt to delay progression to active multiple myeloma.22,23

Supported in part by grants (CA62242 and CA107476) from

the National Cancer Institute.
No potential conflict of interest relevant to this article was
reported.
We thank the Mayo Clinic Cancer Center, which provided hematoxylin and eosin staining and immunohistochemical processing of the bone marrowbiopsy specimens; Barbara A. Todd,
for obtaining the bone marrow material for evaluation; and the
Section of Scientific Publications at Mayo Clinic, for providing
assistance with the preparation of the manuscript.

References
1. Kyle RA, Greipp PR. Smoldering mul-

tiple myeloma. N Engl J Med 1980;302:

1347-9.
2. Alexanian R, Barlogie B, Dixon D.
Prognosis of asymptomatic multiple myeloma. Arch Intern Med 1988;148:1963-5.
3. Dimopoulos MA, Moulopoulos A,
Smith T, Delasalle KB, Alexanian R. Risk
of disease progression in asymptomatic
multiple myeloma. Am J Med 1993;94:5761.
4. Facon T, Menard JF, Michaux JL, et al.
Prognostic factors in low tumour mass
asymptomatic multiple myeloma: a report
on 91 patients. Am J Hematol 1995;48:71-5.
5. Weber DM, Dimopoulos MA, Moulopoulos LA, Delasalle KB, Smith T, Alexanian R. Prognostic features of asymptomatic multiple myeloma. Br J Haematol
1997;97:810-4.
6. Cesana C, Klersy C, Barbarano L, et
al. Prognostic factors for malignant
transformation in monoclonal gammopathy of undetermined significance and
smoldering multiple myeloma. J Clin Oncol 2002;20:1625-34.
7. Rosinol L, Blade J, Esteve J, et al.
Smoldering multiple myeloma: natural
history and recognition of an evolving
type. Br J Haematol 2003;123:631-6.
8. International
Myeloma Working
Group. Criteria for the classification of
monoclonal gammopathies, multiple myeloma and related disorders: a report of
the International Myeloma Working Group.
Br J Haematol 2003;121:749-57.

2590

9. Kurtin PJ, Hobday KS, Ziesmer S,

Caron BL. Demonstration of distinct antigenic profiles of small B-cell lymphomas

by paraffin section immunohistochemistry. Am J Clin Pathol 1999;112:319-29.
10. Sukpanichnant S, Cousar JB, Leelasiri
A, Graber SE, Greer JP, Collins RD. Diagnostic criteria and histologic grading in
multiple myeloma: histologic and immunohistologic analysis of 176 cases with
clinical correlation. Hum Pathol 1994;25:
308-18.
11. Kyle RA, Katzmann JA, Lust JA, Dispenzieri A. Immunochemical characterization of immunoglobulins. In: Rose
NR, Hamilton RG, Detrick B, eds. Manual
of clinical laboratory immunology. 6th
ed. Washington, DC: ASM Press. 2002:7191.
12. Kaplan EL, Meier P. Nonparametric
estimation from incomplete observations.
J Am Stat Assoc 1958;53:457-81.
13. Peto R, Peto J. Asymptotically efficient rank invariant test procedures. J R
Stat Soc [A] 1972;135:185-207.
14. Gooley TA, Leisenring W, Crowley J,
Storer BE. Estimation of failure probabilities in the presence of competing risks:
new representations of old estimators.
Stat Med 1999;18:695-706.
15. Cox DR. Regression models and lifetables. J R Stat Soc [B] 1972;34:187-220.
16. Berry G. The analysis of mortality by
the subject-years method. Biometrics 1983;
39:173-84.
17. Surveillance, Epidemiology, and End

Results (SEER) Program. SEER*Stat database: incidence SEER 9 Regs Public

Use, Nov 2004 Sub (1973-2002), National
Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch,
released April 2005, based on the November 2004 submission. (Accessed May 25,
2007, at http://www.seer.cancer.gov.)
18. Kyle RA, Linos A, Beard CM, et al. Incidence and natural history of primary
systemic amyloidosis in Olmsted County,
Minnesota, 1950 through 1989. Blood
1992;79:1817-22.
19. Kyle RA, Therneau TM, Rajkumar SV,
et al. A long-term study of prognosis in
monoclonal gammopathy of undetermined
significance. N Engl J Med 2002;346:5649.
20. Weber DM, Wang LM, Delasalle KB,
Smith T, Alexanian R. Risk factors for
early progression of asymptomatic multiple myeloma. Hematol J 2003;4:Suppl 1:S31.
abstract.
21. Kyle RA, Gertz MA, Witzig TE, et al.
Review of 1027 patients with newly diagnosed multiple myeloma. Mayo Clin Proc
2003;78:21-33.
22. Rajkumar SV, Dispenzieri A, Fonseca
R, et al. Thalidomide for previously untreated indolent or smoldering multiple
myeloma. Leukemia 2001;15:1274-6.
23. Rajkumar SV, Gertz MA, Lacy MQ, et
al. Thalidomide as initial therapy for early-stage myeloma. Leukemia 2003;17:7759.
Copyright 2007 Massachusetts Medical Society.

n engl j med 356;25 www.nejm.org june 21, 2007

The New England Journal of Medicine

Downloaded from nejm.org by JORGE TORRES on May 9, 2013. For personal use only. No other uses without permission.
Copyright 2007 Massachusetts Medical Society. All rights reserved.