Published Ahead of Print on April 4, 2005 as 10.1200/JCO.2005.04.242
VOLUME
23
NUMBER
15
MAY
20
2005
JOURNAL OF CLINICAL ONCOLOGY
ORIGINAL
REPORT
From the Mayo Clinic College of
International Staging System for Multiple Myeloma
Medicine and Eastern Cooperative
Oncology Group, Rochester, MN;
Philip R. Greipp, Jesus San Miguel, Brian G.M. Durie, John J. Crowley, Bart Barlogie, Joan Bladé,
Grupo Espanol di Mieloma and Program
Mario Boccadoro, J. Anthony Child, Jean-Luc Harousseau, Robert A. Kyle, Juan J. Lahuerta,
de Estudio y Tratamiento de las
Heinz Ludwig, Gareth Morgan, Raymond Powles, Kazuyuki Shimizu, Chaim Shustik, Pieter Sonneveld,
Hemapatias Malignas, University of
Patrizia Tosi, Ingemar Turesson, and Jan Westin
Salamanca, Salamanca; PETHEMA,
Hospital Clinic, Barcelona; Grupo Espa-
ABSTRACT
nol di Mieloma, Hospital Universitario,
Madrid, Spain; Southwest Oncology
Group, International Myeloma Founda-
Purpose
tion and Cedars Sinai Comprehensive
There is a need for a simple, reliable staging system for multiple myeloma that can be applied
Cancer Center, Los Angeles, CA;
internationally for patient classification and stratification.
University of Arkansas for Medical
Patients and Methods
Sciences, Little Rock, AR; Cancer
Clinical and laboratory data were gathered on 10,750 previously untreated symptomatic
Research and Biostatistics, Seattle,
WA; Italian Cooperative Group,
myeloma patients from 17 institutions, including sites in North America, Europe, and Asia.
University of Torino, Torino; Italian
Potential prognostic factors were evaluated by univariate and multivariate techniques. Three
Cooperative Group, Istituto di
modeling approaches were then explored to develop a staging system including two nontree
Ematologia Seragnoli, Bologna, Italy;
and one tree survival assessment methodologies.
Medical Research Council, The General
Infirmary at Leeds, Leeds; Medical
Results
Research Council, Royal Mardsen
Serum beta -microglobulin (S
M), serum albumin, platelet count, serum creatinine, and age
2
2
Hospital, Sutton; Myeloma Unit,
emerged as powerful predictors of survival and were then used in the tree analysis approach.
Parkside Oncology Clinic, Wimbledon,
A combination of S
M and serum albumin provided the simplest, most powerful and
2
United Kingdom; Intergroup Francais
reproducible three-stage classification. This new International Staging System (ISS) was
Myelome, Institut de Biologie, Nantes,
validated in the remaining patients and consists of the following stages: stage I, S
M less
France; Wilhelminenspital Der Stat
2
Wien, Vienna, Austria; Japan Myeloma
than 3.5 mg/L plus serum albumin
3.5 g/dL (median survival, 62 months); stage II, neither
Study Group, Nagoya City Higashi
stage I nor III (median survival, 44 months); and stage III, S
M
5.5 mg/L (median survival,
2
General Hospital, Nagoya, Japan;
29 months). The ISS system was further validated by demonstrating effectiveness in
National Cancer Institute of Canada,
patients in North America, Europe, and Asia; in patients less than and
65 years of age; in
and McGill University, Montreal,
patients with standard therapy or autotransplantation; and in comparison with the Durie/
Quebec, Canada; Dutch-Belgi`an
Salmon staging system.
Haematology-Oncology Cooperative
Group Data Center, Erasmus MC
Conclusion
Hospital, Rotterdam, the Netherlands;
The new ISS is simple, based on easy to use variables (S
M and serum albumin), and
2
Nordic Myeloma Study Group, Malmo,
recommended for early adoption and widespread use.
Sweden; and Lund University Hospital,
Lund, Sweden.
J Clin Oncol 23. © 2005 by American Society of Clinical Oncology
Submitted November 15, 2004; accepted
February 17, 2005.
Supported by the International
Studies conducted in the 1960s and
INTRODUCTION
Myeloma Foundation, Los Angeles, CA.
early 1970s identified a number of clinical
Authors' disclosures of potential con-
The outcome for patients with multiple my-
and laboratory parameters that are indepen-
flicts of interest are found at the end of
this article.
eloma is highly variable. Although the median
dent predictors of survival duration includ-
overall survival time is 3 to 4 years, the range is
ing hemoglobin level, serum calcium, serum
Address reprint requests to International
Myeloma Foundation, International Head-
from less than 6 months to greater than 10
creatinine, and severity of bone lesions.1,2
quarters, 12650 Riverside Dr, Suite 206,
years. This variability derives from heteroge-
Subsequently, combinations of prognostic
North Hollywood, CA 91607; e-mail:
lpaik@myeloma.org.
neity in both myeloma cell biology and multi-
factors were suggested for staging classifica-
© 2005 by American Society of Clinical
ple host factors. Knowledge of tumor and host
tion of myeloma patients.3-5 In 1975, Durie
Oncology
factors associated with prognosis is critical for
and Salmon6 introduced a staging system, the
0732-183X/05/2315-1/$20.00
understanding disease outcome, identifying
Durie/Salmon (DS) system, using commonly
DOI: 10.1200/JCO.2005.04.242
risk groups, and optimizing patient treatment.
available clinical parameters that predicted
1
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright
Copyright ©
2005
2005 by
by
the
American
American
Society
Society of
of
Clinical
Clinical
Oncology. AllOncology
rights reserved.

---

Greipp et al
myeloma cell tumor burden. Factors in the DS classification
included the site of data submission, date of initial treatment, and
included the level and type of monoclonal protein, hemoglo-
date of death or last follow-up. Investigators provided the patient's
bin, calcium level, and number of bone lesions. Creatinine
age at initiation of treatment and also the patient's sex; ethnicity
and race; performance status; hemoglobin level; platelet count;
level (substage A: serum creatinine
2 mg/dL; and substage B:
level and type of M-protein; serum levels of calcium, creatinine,
serum creatinine
2 mg/dL) further defined lower versus
and albumin; DS stage and substage (A or B); number of bone
higher risk patients in each of the three tumor mass stages. The
lesions, compression fractures, and pathologic fractures; bone
DS system was widely adopted as the standard for prognosti-
marrow plasma cell percentage; and levels of lactate dehydroge-
cation in myeloma. The number of lytic lesions on routine
nase, S
M, and C-reactive protein. We also gathered data on
2
radiographs (skeletal survey), an important element of the DS
standard cytogenetics, FISH, and proliferative activity of plasma
system, is unfortunately observer dependent. In an effort to
cells (labeling index or S phase) where available.
Myeloma was diagnosed using standard criteria.22 Patients with
ensure a more objective pretreatment classification of patients,
asymptomatic (smoldering) myeloma were not included. Patients
several staging methods were proposed.7-9 In the 1980s, serum
with immunoglobulin (Ig) M­related disorders or with primary amy-
beta2-microglobulin (S 2M) emerged as the single most pow-
loidosis were not included. Only patients about to start chemotherapy
erful prognostic factor and was considered a simple reliable
were included. Data used for analysis were gathered within 1 month
predictor of survival duration.10-13
before initiation of treatment. Treatment included standard and
Subsequently, other prognostic factors were introduced,
high-dose chemotherapy with stem-cell transplantation. Prior radia-
including serum levels of C-reactive protein,14 albumin,7 and
tion therapy was permitted. Survival was measured from the onset of
nonradiation therapy to time of death or last contact.
the proliferative activity of bone marrow plasma cells assessed
Patient records were randomly selected to create a training
by labeling index15 or flow cytometry cell-cycle (S phase) anal-
sample data set. This data set was balanced by institution or group
ysis.16 Combining these factors with S 2M provided improved
submitting patient data. All submitted clinical and laboratory data
prognostic stratification compared with S 2M alone.7,14-16
were initially assessed for completeness of submission and to priori-
However, there was no consensus about which factors should
tize candidate prognostic factors. Using the training set, models were
be combined with S
developed using the following three different methods: a Cox regres-
2M, and there was no consensus on cutoff
values for S
sion model23 using continuous variables where available, called the
2M or other variables.
weighted variable model; a Cox regression model using dichotomized
Recently, conventional cytogenetics by karyotyping has
variables, called the number of risk factors model; and a survival tree
emerged as a relevant prognostic factor in myeloma patients.
model. These three methods are described in detail in Crowley et al.24
Deletion of chromosome 13 (del 13) is the most common and
The natural log transformation was performed for creatinine and
the most significant prognostic abnormality observed.17-20
S M before inclusion in the weighted variable model, based on
2
Conventional cytogenetics is able to identify abnormalities in
examining separate nonparametric plots of log relative risk by creat-
the myeloma clone in 20% to 30% of patients. Fluorescent in
inine and S M, respectively.25 For the number of risk factors model,
2
situ hybridization (FISH) techniques may offer more sensitive
each continuous variable was dichotomized based on finding the
optimal cut point based on the log-rank statistic.26 Survival tree meth-
and specific identification of such critical abnormalities. Al-
odology extends the recursive partitioning methods to a censored
though prognostically important correlations have emerged,
survival data setting.26
practical application of these techniques has been hampered
In this case, the survival tree model proved to be the most
by lack of standardization, costs, and restricted availability.
efficient methodology and was used to develop the International
The favorable experience of international cooperative efforts
Staging System (ISS). It was validated using the randomly selected
for the design of a prognostic index in non-Hodgkin's lym-
validation set; survival differences in the staging system were ex-
phoma21 led us to embark on a project to design a staging
amined in key subsets. Survival was estimated using the Kaplan-
Meier method,27 with differences in survival examined using the
system for multiple myeloma that would be based on widely
log-rank test.
available, objective parameters used around the world.
An additional statistical technique called extreme regres-
sion28 was used to assess patients with very poor survival (median
survival time,
12 months). This is a type of forward stepwise
PATIENTS AND METHODS
regression analysis from which multiple potential models can be
derived and compared for utility and statistical significance.
We gathered data on 10,750 patients from 15 Asian, European,
and North American institutions and groups. Data were col-
lected from 1981 through 2002. All patients had survival status
RESULTS
and date of last follow-up recorded within 6 months of the data
submission. At the time of analysis, 66% of patients had died. A
total of 7,942 patients received standard therapy, and 2,808 patients
Patient Characteristics
received high-dose therapy as initial treatment. Patients who received
Median age of the 10,750 patients was 60 years; 57%
delayed high-dose therapy beyond 9 months after initial treatment
of patients were male, and 60% had IgG isotype, 24% had
were included in the standard therapy group. Of the 10,750 patients,
7,430 (69.1%) came from clinical trial data. The median age at the
IgA isotype, 11% had light-chain isotype only, 3% had IgD
time of initial chemotherapy for the clinical trial patients and the non-
isotype, and 2% had biclonal or other isotype. Median serum
clinical trial patients was 60 and 63 years, respectively. Data collected
M-protein level was 3.9 g/dL, hemoglobin level was 10.5 g/dL,
2
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved.

---

ISS for Multiple Myeloma
platelet level was 222
100/ L, creatinine level was 1.1 mg/dL,
are were as follows: IgG versus IgA, P
.001; IgG versus
S 2M level was 3.8 g/mL, albumin level was 3.6 g/dL, and
light chain only, P
.001; and IgA versus light chain only, P
bone marrow plasma cell percent was 40.0%. Forty-three per-
.009. Regarding sex, there were 4,597 female patients and
cent of patients had three or more bone lesions, and 25% had
6,153 male patients with median survival times of 45 and 44
pathologic fractures. Overall survival time for the entire group
months, respectively (slight advantage for female patients).
was 44 months.
Development of a Myeloma Staging System
Preliminary Prognostic Factor Analysis
The information from univariate and multivariate analy-
In preparation for developing a staging system, half of
ses was used to explore three modeling approaches. The most
the total myeloma patients (5,383 patients) were randomly
significant prognostic factors were assessed using the following
selected as a training sample.
three methods: (1) the weighted variable model; (2) the model
Univariate and Multivariate Survival Analysis
based on the number of risk factors occurring in an individual
Table 1 allows comparison of variables in both univar-
patient; and (3) the survival tree model in which risk factors
iate and multivariate models and lists the 10 most impor-
present at each branch point are sequentially reassessed.
tant prognostic factors in univariate analyses. The variables
The weighted variable model was derived from the Cox
are ranked by hazard ratio, with all being significant at the
regression approach using continuous variables. The re-
P
.001 level. The numbers and percentages of patients in
gression equation was used to derive a prediction of risk,
each risk category are also listed. S 2M and serum albumin
which was then stratified by tertiles to form risk groups.
were the most consistent, broadly applicable, prognostic
Similarly, for a model based on the number of risk factors,
factors correlated with survival duration. Attention is
the risk groups were based on the five factors identified in
drawn to low platelet count (ranked 2) and high serum
Table 1 (right side) as the major prognostic factors in a Cox
creatinine (ranked 4) results; although these parameters
multivariate regression analysis using, in this case, dichoto-
rank highly, they identify small patient subsets of 12% and
17%, respectively. Conversely, high S
mous variables. S 2M and serum albumin were the domi-
2M (ranked 1) and
low serum albumin (ranked 8) identify larger patient
nant independent prognostic factors in all three models.
subsets of 56% and 40%, respectively. Other relevant prog-
With the survival tree model, a three-stage system us-
nostic factors were age, hemoglobin, calcium, lactate dehy-
ing S 2M and albumin provided the most highly statisti-
drogenase, and bone marrow plasma cell infiltration (Table
cally significant results. From now on, this system is called
1). Additional, although weaker, prognostic factors (ranked
ISS (Table 2). Median survivals were as follows: stage I, 62
11 to 15) on univariate analysis were C-reactive protein, Ig
months; stage II, 45 months; and stage III, 29 months
isotype, size of M-component, and extent of bone lesions
(P
.0001 for differences). Patient numbers were well
(data not shown). There were 5,894 patients with IgG iso-
distributed across the three stages (stage I, 28.9%; stage
type, 2,375 with IgA isotype, and 1,035 with light-chain only
II, 37.5%; and stage III, 33.6%). Among the three meth-
isotype. The median survival time for the IgG patients was
ods for developing a staging system, we chose the survival
longer (49 months) compared with the IgA patients (40
tree approach as being the simplest, most effective and
months) and light-chain patients (35 months). The P values
readily understood method.
www.jco.org
3
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved.

---

Greipp et al
the characteristics of patients according to the new ISS, we
Table 2. New International Staging System
observed that the more advanced the stage, the higher was the
Median
proportion of patients with advanced age, anemia (hemoglo-
Survival
Stage
Criteria
(months)
bin
10 g/dL), thrombocytopenia (platelet count
130,000/
I
Serum
-microglobulin
3.5 mg/L
62
L), high bone marrow infiltration, and poor performance
2
Serum albumin
3.5 g/dL
status (Table 3). The frequency of patients with advanced DS
II
Not stage I or III
44
stage III (A or B) progressively increased from stage ISS I
III
Serum
-microglobulin
5.5 mg/L
29
2
(38%) to stage ISS II (54%) and stage ISS III (70%) categories.
There are two categories for stage II: serum
-microglobulin
3.5
2
For DS stage IIIB, all patients (100%) were ISS stage III.
mg/L but serum albumin
3.5 g/dL; or serum
-microglobulin 3.5 to
2
5.5 mg/L irrespective of the serum albumin level.
As an adjunct to the new staging system, a search was
undertaken to identify simple predictors of very poor prog-
nosis (eg, overall survival
12 months). A type of forward
stepwise regression analysis was performed to identify fac-
Validation of the New Staging System (ISS) in the
tors associated with very poor risk, as noted in Patients and
Remaining Half of Patients (Validation Sample)
Methods.28 This technique is called extreme regression.
The ISS system was applied to the remaining patients,
Using this technique, four factors emerged as being helpful
who constituted half of the whole sample. The discrimina-
in identifying very poor risk patients; these factors were
tion and survival durations were almost identical (stage I, 62
S 2M more than 10 mg/L, serum creatinine more than 4
months; stage II, 44 months; and stage III, 29 months) to the
mg/dL, serum albumin less than 2.5 g/dL, and platelet count
training sample (Fig 1 and Table 2).
less than 130,000/ L. It was possible to fit several different
Because the training and validation samples had almost
models with this regression technique. With all of the mod-
identical outcome, the two patient populations were com-
els, the very poor risk group was approximately 5% of the
bined into a single group for further analysis. It is important
total population. Although significant P values of less than
to note that serum albumin added consistent prognostic dis-
.0001 were obtained, it must be noted that the worst median
crimination versus S 2M alone. Thus, of the 3,157 patients
overall survival time for a poor risk group was 17 months.
with a low S 2M less than 3.5 mg/L, 1,020 (32% of these
Thus, there was somewhat limited ability to accurately pre-
patients; 12.5% of the total population) were classified as stage
dict very poor survival with these routine test parameters.
II because of a low albumin less than 3.5 mg/L. On analyzing
Half of the patients had a survival time of more than 17
months and did not fall into the less than 12 months very
poor risk group that was being sought.
Cytogenetic data (Table 3) were available in a subset of
390 patients. No strong correlations with stage were ob-
served. The translocation t(4;14) occurred with a lower
incidence (P
.035) in stage I patients than in stage II and
III patients (6% v 16% and 11%, respectively). Although
there was a slight trend for less frequent complex karyotypic
abnormalities, deletion 13 by FISH and deletion 13 by cy-
togenetics in stage I disease, these trends were not statisti-
cally significant (P
.075 and P
.162, respectively).
Considering the impact of cytogenetic abnormalities over-
all, patients with and without cytogenetic abnormalities of
any type were compared. The median overall survival for
the 113 patients with cytogenetic abnormalities was 42
months v 69 months for the 277 patients with no cytoge-
netic abnormalities. The P value for no cytogenetic abnor-
malities versus cytogenetic abnormalities is P
.03.
Other Assessments of the New ISS System
Geographic region.
In proposing an international
system, it was important to validate the system by geo-
graphic region. As illustrated in Figure 2, there was compara-
ble utility in patients from North America, Europe, and Asia.
Discriminatory efficacy was also excellent comparing individ-
Fig 1. Training versus validation datasets. ISS, International Staging Sys-
tem. A is training dataset; B is validation dataset.
ual institutions versus cooperative groups (data not shown).
4
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved.

---

ISS for Multiple Myeloma
Table 3. Distribution of Clinical and Laboratory Variables by ISS Stages I, II, and III
ISS Stage
Stage I
Stage II
Stage III
No. of Patients/
No. of Patients/
No. of Patients/
Factor
Total No.
%
Total No.
%
Total No.
%
P
Age 65
years
605/2,303
26
1,118/3,152
35
1,061/2,685
40
.001
S
M
3.5
g/mL
0/2,307
0
2,137/3,157
68
2,693/2,693
100
.001
2
Albumin
3.5 g/dL
0/2,307
0
1,873/2,985
63
1,239/2,494
50
.001
HGB
10 g/dL
414/2,295
18
1,264/3,139
40
1,772/2,672
66
.001
Creatinine
2 mg/dL
43/2,291
2
201/3,129
6
1,138/2,662
43
.001
PLT
130
103/ L
119/2,126
6
308/2,989
10
491/2,535
19
.001
Calcium
10 mg/dL
513/2,111
24
860/2,920
29
1,139/2,501
46
.001
3 lytic lesions
866/1,911
45
1,226/2,619
47
1,184/2,293
52
.001
CRP
0.8 mg/dL
222/1,208
18
464/1,473
32
445/1,134
39
.001
LDH above normal
186/923
20
264/1,113
24
348/939
37
.001
BMPC
33%
930/2,199
42
1,782/2,992
60
1,877/2,532
74
.001
PS 2
649/2,128
30
1,215/2,999
41
1,348/2,578
52
.001
Durie/Salmon stage III (A or B)
782/2,046
38
1,471/2,748
54
1,638/2,356
70
.001
Any clonal CA
33/144
23
40/132
30
40/114
35
.093
Complex karyotype
9/66
14
21/88
24
22/85
26
.162
Del13 by cytogenetics
10/125
8
18/113
16
14/87
16
.112
T11; 14
18/123
15
22/145
15
18/109
17
.921
T4; 14
8/125
6
23/140
16
11/104
11
.035
Del13 by FISH
55/125
44
76/134
57
48/107
45
.075
Abbreviations: ISS, International Staging System; S
M, serum beta -microglobulin; HGB, hemoglobin; PLT, platelets; CRP, C-reactive protein; LDH, lactate
2
2
dehydrogenase; BMPC, bone marrow plasma cells; PS, performance status; CA, cytogenetic abnormalities; FISH, fluorescent in situ hybridization.
Number with factor for group level/number known with or without factor for group level.
Fisher exact test, otherwise 2 test.
Age. Age is not only an important prognostic factor, but
IIA reflects a similar patient population, with a median
it also critically influences treatment options, such as high-
survival time of 58 months. ISS stage II patients correspond
dose therapy. Accordingly, we wanted to analyze whether the
to DS stage IIIA patients, with median survival times of 44
ISS system applies equally to young and older patients. Thus,
and 45 months, respectively. Of particular note, ISS stage III
although older patients (eg,
65 years) have poorer survival
identifies DS substage B (serum creatinine
2 mg/dL) for
than younger patients, it is important to note, as illustrated in
DS stages I, II, and III (ie, the poor-risk B subset), irrespec-
Figure 3, that the ISS system applies to both groups.
tive of tumor burden. It is obviously helpful to have such
Treatment type. As far as treatment type is concerned,
patients categorized collectively in ISS stage III. In Table 3, it
7,920 patients were treated with standard-dose therapy as
can be seen that of the 1,382 total patients with a serum
the primary modality, whereas 2,807 patients received high-
creatinine
2 mg/dL, 82% had a S 2M value of more than
dose therapy with autologous marrow or stem-cell rescue
5.5 mg/L and, therefore, were classified as ISS stage III.
on an intent-to-treat basis (defined as within 9 months of
Importantly, the S 2M values (ranked number 1 in both
start of therapy). Again, the ISS system discriminated sim-
univariate and multivariate analyses) are much more pow-
ilarly for the two groups, as shown in Figure 4.
erful versus serum creatinine values (ranked number 4 in
Comparisons with the DS system. The survival dura-
univariate and number 5 in multivariate analyses; Table 1).
tion comparisons of the DS system versus the ISS system are
listed in Table 4 and Figure 5. Compared with the DS
DISCUSSION
classification, the ISS provides more equal distribution of
patients across the three stages. DS stage I patients are
underrepresented in these data sets. The 8% value is lower
The DS clinical staging system has remained the most
than the typical 20% level, most likely because DS stage I
widely used staging system for over 25 years. Although a
patients are excluded from many protocols involved in the
few prognostic parameters, such as S 2M, have emerged
data sets analyzed. Nonetheless, the survival of ISS stage I
as better predictors of survival duration, there has been
corresponds exactly to the DS stage IA patients who are
no consensus as to the optimal use of single or multiple
incorporated in these analyses. Both groups of patients have
prognostic factors. The current large international data
median survival times of 62 months. Interestingly, DS stage
set of patients with symptomatic myeloma offers the
www.jco.org
5
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved.

---

Greipp et al
Fig 3. International Staging System (ISS); staging by age. A is patient's age
65
years; B is patient's age
65 years.
datasets. The broad applicability of the ISS system was
further illustrated with validation by geographic area, pa-
tient age, and treatment type and in comparison with the
DS staging system (Figs 1 and 2, Table 4). Of particular note,
ISS stage III is clearly delineated as a poor-risk group (39% of
patients), with a median survival time of 29 months (Table 4)
versus the more mixed and numerous DS stage IIIA (49% of
Fig 2. International Staging System (ISS); staging by geographic region. A
is Asia; B is Europe; C is North America.
patients; median survival time, 45 months) and stage IIIB
(17% of patients; median survival time, 24 months).
Early attempts to improve on the DS staging system
opportunity to establish a statistically superior and
were not widely adopted.7 But now, S 2M is widely recog-
widely accepted new staging system.
nized as the single most important variable predicting
The new ISS (Table 2) was developed using univariate
survival.13 When added to S 2M, serum albumin level
and multivariate analyses (Table 1) and three types of mod-
was known to add significantly to prognostication.7 There
eling approaches. S 2M and serum albumin were selected
was much debate as to whether these were sufficient prog-
from the various potential prognostic factors both because
nostic factors or whether better prognostic factors were
of the statistical power in various models as well as the
required.29-33 However, in the absence of any additional,
known wide availability of these two simple inexpensive
powerful prognostic factors, further analyses using S 2M
laboratory tests. The inclusion of serum albumin as the
and serum albumin were conducted. This led to a S 2M
second parameter added significantly in defining 1,020 pa-
and serum albumin staging system29 developed by the
tients (12.5% of total population), now identified as ISS
Southwest Oncology Group. The newly developed and
stage II, who would otherwise have been classified as stage I
proposed ISS system thus extends and validates these
based on low (
3.5 mg/L) S 2M alone. The large data set
prior observations.
afforded the opportunity to establish clear cutoff values to
The following question emerges: why are S 2M and
identify the three stages in the new ISS system (summarized
serum albumin such powerful prognostic factors? S 2M
in Table 2 and displayed in Fig 5). The survival differences
reflects not only tumor mass and renal function but also
were reproducibly demonstrated in the test and validation
other as yet unknown parameters, possibly including
6
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved.

---

ISS for Multiple Myeloma
Fig 4. International Staging System (ISS) stage by treatment type. A is
Fig 5. International Staging System (ISS) stage by Durie/Salmon stage. A
standard-dose chemotherapy; B is high-dose therapy on an intent-to-treat
is overall survival by ISS stage; B is overall survival by Durie/Salmon stage
basis. See text for discussion.
(I-III; A/B).
immune function.7,34,35 The specific cause of decreased al-
displayed). Because the levels of S 2M and albumin are now
bumin in some multiple myeloma patients is not certain; how-
specified by the ISS, it is critical that laboratory variation be
ever, a lower albumin may reflect effects on the liver by
minimized by standardizing methods used to determine
interleukin-6 produced by the microenvironment of myeloma
their levels, specifically in multiple myeloma. That work
cells.34,35 The strong correlations between serum levels of
and standardization is underway by the Nordic Myeloma
S 2M and serum albumin and myeloma patient survival imply
Study Group, who are members of the International My-
connections to important underlying mechanisms. There are
eloma Working Group.
several clues in the published literature,36-41 but to date, the
We conclude that the ISS staging system is broadly
underlying biology remains to be explored.
useful and that it will provide a sound base for more ad-
The ISS provides useful prognostic groupings in a va-
vanced studies in the future. Identification of highest risk
riety of situations (in patients aged greater or less than 65
patients was achieved in only a small number of patients
years, Fig 3; with conventional v high-dose transplantation
(5% to 9%) using standard variables. Better identification of
therapy, Fig 4; in Europe, Asia, and North America, Fig 2;
such patients may require a more refined cytogenetic and
and in single institutions v cooperative groups, data not
molecular genetic classification. As more data and follow-up
Table 4. Comparison Between Durie/Salmon and ISS Staging Systems: Survival Duration by Stage in Months
Durie/Salmon
ISS
Stage
% of Patients
Median Survival (months)
Stage
% of Patients
Median Survival (months)
IA
7.5
62
I
28
62
IB
0.5
22
IIA
22
58
II
33
44
IIB
4
34
IIIA
49
45
III
39
29
IIIB
17
24
Abbreviation: ISS, International Staging System.
Percentage of patients falling into each staging category.
www.jco.org
7
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved.

---

Greipp et al
become available, the International Myeloma Working Group
Medi-Clinic, Plumstead, South Africa;Douglas Joshua, Royal
plans to develop a second staging system using conventional
Prince Alfred Hospital, Australia; Sevgi Kalayoglu-Besisik,
and FISH cytogenetics, molecular genetics, proteomics, and
Istanbul University, Istanbul, Turkey; Henk Lokhorst, Univer-
S-phase analysis for use by reference centers and eventually for
sity Hosptial Utrecht, Utrecht, Netherlands; Philippe Moreau,
all patients with myeloma.
Institut de Biologie, Nantes, France; Hirokazu Murakami,
Gunma University, Japan; Eiichi Nagura, National Chubu

Hospital, Japan; Martin Oken, Humphrey Cancer Clinic,
Appendix
Robbinsdale, MN; Santiago Pavlovsky, Fundaleu, Buenos
The journal of Clinical Oncology limits author lists to 20.
Aires, Argentina; Eric Rasmussen, Cancer Research and Bio-
The authors listed for this manuscript are therefore the mem-
statistics, Seattle, WA; Paul Richardson, Dana-Farber Cancer
bers of the International Myeloma Working Group who rep-
Center, Boston, MA; Angelina Rodriquez-Morales, Banco
resent institutions and groups contributing patient data for the
Municipal de Sangre de La Region Capital, Venezuela; David
analyses. Meral Beksac, Ankara University, Ibni Sina Hospi-
Roodman, University of Pittsburgh, Pittsburgh, PA; David
tal, Ege and Inonu Universities, Ankara and Istanbul, Turkey
Siegel, Hackensack University Cancer Center, Hackensack, NJ;
also contributed patient data. Other members of the Interna-
Bhawna Sirohi, Royal Marsden Hospital, United Kingdom;
tional Myeloma Working Group who contributed to this study
Keith Stewart, University of Toronto, Toronto, Canada;
include the following: Raymond Alexanian, University of
Guido Tricot, University of Arkansas, Little Rock, AR; Brian
Texas, Houston, TX; Kenneth Anderson, Dana-Farber Cancer
Van Ness, University of Minnesota, Minneapolis, MN; David
Center, Boston, MA; Michel Attal, Institut de Biologie, Nantes,
Vesole, Medical College of Wisconsin, Madison, WI; Donna
France; Hervé Avet-Loiseau, Institut de Biologie, Nantes,
France; Ismet Aydogdu, Inonu University, Turkey; Regis
Weber, M.D. Anderson Cancer Center, Houston, TX; and
Bataille, University of Nantes, France; William Bensinger,
Keith Wheatley, University of Birmingham, Birmingham,
Fred Hutchinson Cancer Research Center, Seattle, WA;
United Kingdom.
Peter Bergsagel, Cornell Medical Center, New York, NY;
Seckin Cagirgan, Ege University, Turkey; Michele Cavo,
Authors' Disclosures of Potential
Istituto di Ematologia, University of Bologna, Bologna,
Conflicts of Interest
Italy; Ray Comenzo, Sloan-Kettering Cancer Research Cen-
The following authors or their immediate family mem-
ter, New York, NY; William Dalton, University of South
bers have indicated a financial interest. No conflict exists
Florida, Tampa, FL; Meletios Dimopoulous, University of
for drugs or devices used in a study if they are not being
Athens, Athens, Greece; Mark Drayson, University of Bir-
evaluated as part of the investigation. Consultant/Advisory
mingham, Birmingham, United Kingdom; Thierry Facon,
Role: Robert A. Kyle, Celgene, Millennium, Neo Rx; Jan
Institut de Biologie, Nantes, France; Dorotea Fantl, So-
Westin, Celgene, Millennium. Stock Ownership: Jan Westin,
ciedad Argentinade Hematolgia, Argentina; Rafael Fonseca,
AstraZeneca. Honoraria: Robert A. Kyle, Millennium,
Mayo Clinic College of Medicine and Eastern Cooperative
Novartis; Heinz Ludwig, Auipeu, Ortho Biotech, Roche.
Oncology Group, Rochester, MN; Gosta Gahrton, Karolin-
Research Funding: Heinz Ludwig, Schering. For a detailed
ska Institute, Stockholm, Sweden; Hartmut Goldschmidt,
description of these categories, or for more information
University of Heidelberg, Heidelberg, Germany; Morie
about ASCO's conflict of interest policy, please refer to the
Gertz, Mayo Clinic, Rochester, NY; Vania Hungria, Clinica
Author Disclosure Declaration and Disclosures of Potential
Sao Germano, Sao Paulo, Brazil; Mohamad Hussein, Cleve-
Conflicts of Interest found in Information for Contributors
land Clinic, Cleveland, OH; Peter Jacobs, Constantiaberg
in the front of each issue.
mia: Report on the first myelomatosis trial: Part I.
9. Merlini G, Waldenström JG, Jayakar SD: A
REFERENCES
Analysis of presenting features of prognostic
new improved clinical staging system for multi-
importance. Br J Haematol 24:123-139, 1973
ple myeloma based on analysis of 123 treated
1. Carbone PP, Kellerhouse LE, Gehan EA:
5. Chronic Leukemia Myeloma Task Force,
patients. Blood 55:1011-1019, 1980
Plasmacytic myeloma: A study of the relation-
National Cancer Institute: Proposed guidelines
10. Cassuto JP, Krebs BJ, Viot G, et al: Beta 2
ship of survival to various clinical manifestations
for protocol studies: II. Plasma cell myeloma.
microglobulin, a tumor marker of lymphoprolif-
and anomalous protein type in 112 patients.
Cancer Chemother Rep 4:145-158, 1973
erative disorders. Lancet 2:108-109, 1978
Am J Med 42:937-948, 1967
6. Durie BGM, Salmon SE: A clinical staging
11. Norfolk D, Child JA, Cooper EH, et al:
2. Costa G, Engle RL, Schilling A, et al:
system for multiple myeloma. Cancer 36:842-
Serum beta 2 microglobulin in myelomatosis:
Melphalan and prednisone: An effective combi-
854, 1975
Potential value in stratification and monitoring.
nation for the treatment of multiple myeloma.
7. Bataille R, Durie BGM, Grenier J, et al:
Br J Cancer 39:510-515, 1979
Am J Med 54:589-599, 1973
Prognostic factors and staging in multiple my-
12. Bataille R, Durie BGM, Grenier J: Serum
3. Dawson AA, Ogston D: Factors influenc-
eloma: A reappraisal. J Clin Oncol 4:80-87, 1986
beta 2 microglobulin and survival duration in
ing the prognosis in myelomatosis. Postgrad
8. Medical Research Council's Working Party
multiple myeloma: A simple reliable marker for
Med J 47:635-638, 1971
on Leukemia in Adults: Prognostic features in
staging. Br J Haematol 55:439-447, 1983
4. Report of the Medical Research Council's
the third MRC myelomatosis trial. Br J Cancer
13. Durie BGM, Stock-Novack D, Salmon SE,
Working Party for Therapeutic Trials in Leuke-
42:831-840, 1980
et al: Prognostic value of pretreatment serum
8
JOURNAL OF CLINICAL ONCOLOGY
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved.

---

ISS for Multiple Myeloma
beta
microglobulin in myeloma: A Southwest
Lymphomas: Report of the jury. J Clin Oncol
Oncology Group Study 8229. J Clin Oncol 14:
2
Oncology Group study. Blood 75:823-830, 1990
17:423-429, 1999
974-977, 1996
14. Bataille R, Boccadoro M, Klein B, et al:
22. Durie BGM, Salmon SE. Recent Advances
34. Jacobson J, Hussein M, Barlogie B, et al:
C-reactive protein and bta
microglobulin pro-
in Haematology. Edinburgh, United Kingdom,
2
A new staging system for multiple myeloma
duce a sample and powerful myeloma staging
Churchill Livingstone, 1977
patients based on the Southwest Oncology
system. Blood 80:733-737, 1992
23. Cox DR: Regression models and life ta-
Group (SWOG) experience. Br J Haematol 122:
15. Greipp PR, Lust JA, O'Fallon WM, et al:
bles. J R Stat Soc B 34:187-220, 1972
441-450, 2003
Plasma
cell
labeling
index
and
beta M-
24. Crowley J, McCoy J, LeBlanc M, et al:
2
35. Durie BGM, Kyle R, Belch A, et al: My-
microglobulin predict survival independent of
Extreme regression: A statistical technique for
eloma management guidelines: A consensus
thymidine-kinase and C-reactive protein in multi-
finding good or poor prognostic groups, illustrated
report from the Scientific Advisors of the Inter-
ple myeloma. Blood 81:3382-3387, 1993
using myeloma patient data from Intergroup trial
national Myeloma Foundation. Hematol J 4:379-
16. San Miguel JF, Garcia-Sanz R, Gonzalez
S9321. Blood 104:3856, 2004 (abstr 5202)
398, 2003
M, et al: A new staging system for multiple
25. Gentleman R, Crowley J: Graphical meth-
36. Johnson DR, Pober JS: Tumor necrosis
myeloma based on the number of S-phase
ods for censored data. J Am Stat Assoc 86:678-
factor and immune interferon synergistically in-
plasma cells. Blood 85:448-455, 1995
683, 1991
crease transcription of HLA class I heavy-and
17. Dewald GW, Kyle RA, Hicks GA, et al: The
26. LeBlanc M, Crowley J: Survival trees by
light-chain genes in vascular endothelium. Proc
clinical significance of cytogenetic studies in 100
goodness of split. J Am Stat Assoc 88:457-467,
Natl Acad Sci USA 87:5183-5187, 1990
patients with multiple myeloma, plasma cell leu-
1993
37. Beersma MFC, Bijlmakers MJE, Ploegh
kemia, or amyloidosis. Blood 66:380-390, 1985
27. Kaplan EL, Meier P: Nonparametric esti-
HL: Human cytomegalovirus down-regulates
18. Tricot G, Barlogie B, Jagannath S, et al:
mation from incomplete observations. J Am Stat
HLA class I expression by reducing the stability
Poor prognosis in multiple myeloma is associ-
Assoc 53:457-481, 1958
of class I H chains. J Immunol 151:4455-4464,
ated with partial or complete deletions of chro-
28. Crowley JJ, McCoy J, LeBlanc M, et al:
1993
mosome 13 or abnormalities involving 11q and
Extreme regression: A statistical technique for
38. Perosa F, Luccarelli G, Prete M, et al:
not with other karyotype abnormalities. Blood
finding good or poor prognostic groups. Blood
Increased serum levels of beta2m-free HLA
86:4250-4256, 1995
104:385b, 2004 (abstr 5202)
class I heavy chain in multiple myeloma. Br J
19. Tricot G, Sawyer JR, Jagannath S, et al:
29. Kyle RA: Why better prognostic factors for
Unique role of cytogenetics in the prognosis of
multiple myeloma are needed? Blood 83:1713-
Haematol 106:987-994, 1999
patients with myeloma receiving high-dose ther-
1716, 1994
39. Christianson GJ, Brooks W, Vekasi S, et al:
apy and autotransplants. J Clin Oncol 15:2659-
30. Kyle RA: Prognostic factors in multiple
Beta -microglobulin-deficient mice are protected
2
2666, 1997
myeloma. Stem Cells 13:56-63, 1995
from hypergammaglobulinemia and have defec-
20. Avet-Loiseau H, Andree-Ashley LE, Moore
31. San Miguel JF: Overview of prognostic
tive antibody responses because of increased
D, et al: Molecular cytogenetic abnormalities in
factors in multiple myeloma. Cancer Res Ther-
IgG catabolism. J Immunol 159:4781-4792, 1997
multiple myeloma and plasma cell leukemia
apy Control 6:97-99, 1998
40. Bologa RM, Levine DM, Parker TS, et al:
measured using comparative genomic hybridiza-
32. Turesson I, Abildgaard N, Ahlgren T: Prog-
Interleukin-6 predicts hypoalbuminemia, hypo-
tion. Genes Chromosomes Cancer 19:124-133,
nostic evaluation in multiple myeloma: An analy-
cholesterolemia, and mortality in hemodialysis
1997
sis of the impact of new prognostic factors. Br J
patients. Am J Kidney Dis 32:107-114, 1998
21. Shipp MA, Abeloff MD, Antman KH, et al:
Haematol 106:1005-1012, 1999
41. Maire P, Wuarin J, Schibler U: The role of
International Consensus Conference on High-
33. Modiano MR, Villar-Werstler P, Crowley J,
cis-acting promoter elements in tissue-specific
Dose Therapy With Hematopoietic Stem Cell
et al: Evaluation of race as a prognostic factor in
albumin gene expression. Science 244:343-346,
Transplantation in Aggressive Non-Hodgkin's
multiple myeloma: An ancillary of Southwest
1989
www.jco.org
9
Downloaded from www.jco.org at CEDARS SINAI MEDICAL CENTER MEDICAL LIBRARY on April 4, 2005 .
Copyright © 2005 by the American Society of Clinical Oncology. All rights reserved.

---