Leukemia (2011) 25, 1502–1509
& 2011 Macmillan Publishers Limited All rights reserved 0887-6924/11
www.nature.com/leu
ORIGINAL ARTICLE
The absolute monocyte and lymphocyte prognostic score predicts survival and identifies
high-risk patients in diffuse large-B-cell lymphoma
RA Wilcox1,2, K Ristow1, TM Habermann1, DJ Inwards1, INM Micallef1, PB Johnston1, JP Colgan1, GS Nowakowski1, SM Ansell1,
TE Witzig1, SN Markovic1,2 and L Porrata1
1
Department of Internal Medicine, Division of Hematology, Mayo Clinic, Rochester, MN, USA and 2Department of Medical
Oncology, Mayo Clinic, Rochester, MN, USA
Despite the use of modern immunochemotherapy regimens,
almost 50% of patients with diffuse large-B-cell lymphoma will
relapse. Current prognostic models, including the International
Prognostic Index, incorporate patient and tumor characteristics. In contrast, recent observations show that variables
related to host adaptive immunity and the tumor microenvironment are significant prognostic variables in non-Hodgkin
lymphoma. Therefore, we retrospectively examined the absolute monocyte and lymphocyte counts as prognostic variables
in a cohort of 366 diffuse large-B-cell lymphoma patients who
were treated between 1993 and 2007 and followed at a single
institution. The absolute monocyte and lymphocyte counts in
univariate analysis predicted progression-free and overall
survival when analyzed as continuous and dichotomized
variables. On multivariate analysis performed with factors
included in the IPI, the absolute monocyte and lymphocyte
counts remained independent predictors of progression-free
and overall survival. Therefore, the absolute monocyte and
lymphocyte counts were combined to generate a prognostic
score that identified patients with an especially poor overall
survival. This prognostic score was independent of the IPI and
added to its ability to identify high-risk patients.
Leukemia (2011) 25, 1502–1509; doi:10.1038/leu.2011.112;
published online 24 May 2011
Keywords: monocyte; absolute lymphocyte count; diffuse largeB-cell lymphoma; prognosis
Introduction
The International Prognostic Index (IPI) is currently the standard
prognostic tool used to predict clinical outcomes for patients
with diffuse large-B-cell lymphoma (DLBCL). Developed in the
pre-rituximab era, the IPI was able to stratify patients into four
discrete groups with 5-year overall survival (OS) ranging from
approximately 25 to 75%. However, with improved outcomes
attributed to the addition of rituximab, the identification of a
high-risk subset of patients with an anticipated 5-year survival of
less than 50% remains a challenge with use of the IPI alone.1,2
Gene-expression profiling,3–7 immunohistochemistry-based detection of prognostic biomarkers8–16 and early interim analysis
with positron emission tomography17,18 following the initiation
of immunochemotherapy have all been explored as predictors
that may identify high-risk patients. Although promising, many
of these methods are costly, difficult to obtain, not easily
Correspondence: Dr RA Wilcox, Department of Internal Medicine,
Division of Hematology, Mayo Clinic, 200 First Street SW, Rochester,
MN 55905, USA.
E-mail: [email protected]
Received 5 October 2010; revised 26 April 2011; accepted 27 April
2011; published online 24 May 2011
interpreted and require further validation. Therefore, the
identification of prognostic factors that are inexpensive, widely
available and easily interpreted by clinicians engaged in the
care of patients with DLBCL is needed.
Most of the proposed prognostic factors utilized in DLBCL
include solely patient or tumor characteristics. However, geneexpression profiling studies in non-Hodgkin lymphomas (NHL)
show that gene expression by tumor-infiltrating lymphocytes
and myeloid-derived cells predict clinical outcomes.3,19 Lymphocytes have an important role in immune surveillance in
NHL, a view supported by the observation that lymphopenia is
an adverse prognostic factor in NHL of various subtypes,
including DLBCL,20–23 and is widely considered an important
surrogate marker of immunological reconstitution following
stem cell transplantation in NHL.24 Monocytes and lymphomaassociated macrophages (LAMs) contribute to the suppression of
host anti-tumor immunity and the promotion of tumor angiogenesis. Furthermore, monocytes may also provide trophic
factors that directly promote the growth and survival of
malignant lymphocytes, as recently shown in both B- and
T-cell-derived NHL.25,26
Therefore, the aim of this study was to devise a prognostic
model that is comprised of immunologically relevant host
factors that is simple, applicable for routine clinical use, and
sufficient to identify high-risk patients. As myeloid cells,
including monocytes and their progeny, promote lymphomagenesis both directly, via the provision of trophic factors, and
indirectly via the suppression of host immunity and stimulation
of tumor angiogenesis, we examined the peripheral blood
monocyte count at diagnosis as a prognostic factor for survival
in DLBCL patients.
Patients and methods
Patients
Consecutive patients with DLBCL who were evaluated and
treated with CHOP (cyclophosphamide, hydroxydaunorubicin,
oncovin, prednisone) or R-CHOP (rituximab-cyclophosphamide, hydroxydaunorubicin, oncovin, prednisone) at the Mayo
Clinic, Rochester between 1993 and 2007 were considered for
study participation. HIV-positive patients were excluded from
this study. The laboratory values were obtained from routine
automated complete blood count (CBC) determination at the
time of diagnosis.27 Patient data was collected and entered into
the Mayo Clinic Lymphoma Database. Our study sample
included 366 DLBCL patients, none of whom declined to
participate in this retrospective review. An additional 26 patients
were eligible for study participation, but were excluded owing
to missing laboratory data. Study approval was granted by the
AMC and ALC predict survival in DLBCL
RA Wilcox et al
1503
Mayo Clinic Institutional Review Board, in accordance with US
federal regulations and the Declaration of Helsinki.
Table 1
Patient characteristics (n ¼ 366)
Characteristic
Study objective
The primary objective of the study was to determine the
prognostic significance of the absolute monocyte and lymphocyte counts at diagnosis with both OS and progression-free
survival (PFS). The absolute monocyte count (AMC) and
absolute lymphocyte count (ALC) were obtained from a standard
CBC obtained at the time of diagnosis. The secondary study
objective was to determine whether the AMC or ALC are
independently significant prognostic factors when compared
with the IPI. Patient and disease characteristics included in the
IPI (age (X60 vs o60 years), Ann Arbor stage (III/IV vs I/II),
Eastern Cooperative Oncology Group performance status (41
vs p1), lactate dehydrogenase (4normal vs normal for age/sex)
and the number of extranodal sites of disease (41 vs p1)) were
utilized.28
Statistical analysis
PFS and OS were estimated using the Kaplan–Meier method and
two-tailed log-rank test.29 Response criteria were based on the
criteria from the International Harmonization Project.30 PFS was
calculated from the date of diagnosis to the date of disease
progression or last follow-up. OS was calculated from the date
of diagnosis to the date of death or last follow-up. The AMC was
analyzed as both a continuous and dichotomized variable using
a median split. Receiver operating characteristics analysis was
also performed (Cox’s regression analysis) to determine the
optimal cut-point for both the AMC and ALC. However, the
optimal AMC cut-point determined by receiver operating
characteristics analysis for the AMC (610/ml) was nearly
identical to the median AMC (data not shown). Therefore, a
median split was selected for subsequent analyses using the
AMC. The optimal cut-point for the ALC by receiver operating
characteristic analysis was 1470/ml. However, as an ALC cutpoint of 1000/ml has been utilized in previous studies,23,31,32
and failed to yield a significant difference in the ability to risk
stratify patients by OS when compared with the cut-point
selected by receiver operating characteristic analysis (data not
shown), we chose to dichotomize the ALC using a cut-point of
1000/ml. The Cox proportional hazards model was used to
evaluate the AMC and ALC as prognostic factors for PFS and OS
and to adjust for other known prognostic variables included in
the IPI.33 P-values were not adjusted for multiple comparisons.
All two-sided P-values o0.05 were determined to be statistically significant.
Results
Patient characteristics
The median age at diagnosis for this cohort of 366 DLBCL
patients was 64 years (range, 20–92 years), as summarized in
Table 1. Of these, 58% of patients were men. Median follow-up
following diagnosis was 50 months for the entire cohort (range,
o1 month to 146 months) and 5.5 years for censored patients.
At diagnosis, the median AMC was 630/ml (range 30–4040,
interquartile range 470–833) and the median ALC was 1230/ml
(range 140–9200, interquartile range 870–1770). All patients
were treated initially with CHOP or R-CHOP. The estimated
5-year OS for the entire cohort was 64%.
No.
Age (years)
Median (range)
%
64 (20–92)
Gender
Male
214
58
Ann Arbor stage
I
II
III
IV
85
96
38
147
23
26
10
40
LDH
4Normal
pNormal
151
215
41
59
Extranodal sites of disease
41
p1
73
293
20
80
ECOG PS
41
p1
44
322
12
88
IPI
0
1
2
3
4
5
47
111
94
75
34
5
13
30
26
20
9
1
Treatment
CHOP
R-CHOP
Radiation
110
256
88
30
70
24
Absolute monocyte count
Median (interquartile range)
630/ml (470–833)
Absolute lymphocyte count
Median (interquartile range)
1230/ml (870–1770)
Abbreviations: CHOP, cyclophosphamide, hydroxydaunorubicin,
oncovin, prednisone; R-CHOP, rituximab-cyclophosphamide, hydroxydaunorubicin, oncovin, prednisone; ECOG, Eastern Cooperative
Oncology Group; IPI, International Prognostic Index; LDH, lactate
dehydrogenase; PS, performance status.
Elevation in the AMC and relative lymphopenia are
adverse prognostic factors
To determine the prognostic significance of the AMC and ALC,
both values were obtained from a CBC at the time of initial
diagnosis. On univariate analysis, the AMC predicted OS as a
continuous variable, whereas the relationship between the ALC
and OS approached, but did not reach, statistical significance
(Table 2). An elevated AMC, analyzed as a dichotomized
variable using a median split, was associated with inferior OS on
univariate analysis (hazard ratio 2.77, 95% confidence interval
1.92–4.09, Po0.0001). Lymphopenia defined as an ALC
p1000/ml, was associated with inferior OS (hazard ratio 1.74,
95% confidence interval 1.23–2.43, P ¼ 0.002). For comparison, each of the five factors that comprise the IPI was included
in the analysis. Of these, performance status (X2), lactate
dehydrogenase (4normal), number of extranodal sites (41) and
Ann Arbor stage (stage III/IV) were of prognostic significance on
univariate analysis (Table 2).
Leukemia
AMC and ALC predict survival in DLBCL
RA Wilcox et al
1504
Table 2
Univariate and multivariate analyses for overall survival
Prognostic factors
Univariate analysis
HR (95% CI)
AMC (continuous variable)
AMC X630/ml
ALC (continuous variable)
ALC p1000/ml
Age X60 years
ECOG PS X2
LDH 4normal
Number of extranodal sites 41
Ann Arbor stage III/IV
1.0008
2.77
0.80
1.74
1.25
1.64
2.14
1.60
2.04
(1.0005–1.001)
(1.92–4.09)
(0.62–1.01)
(1.23–2.43)
(0.88–1.81)
(1.02–2.54)
(1.52–3.00)
(1.07–2.33)
(1.44–2.90)
Multivariate analysis
P-value
o0.0001
o0.0001
0.06
0.002
0.2
0.04
o0.0001
0.02
o0.0001
HR (95% CI)
P-value
2.65 (1.82–3.92)
o0.0001
1.56
1.21
1.07
1.61
1.12
1.45
(1.10–2.22)
(0.85–1.76)
(0.65–1.70)
(1.12–2.33)
(0.72–1.70)
(0.98–2.15)
0.01
0.30
0.77
0.01
0.61
0.06
Abbreviations: AMC, absolute monocyte count; ALC, absolute lymphocyte count; CI, confidence interval; ECOG, Eastern Cooperative Oncology
Group; HR, hazard ratio; LDH, lactate dehydrogenase; PS, performance status.
As components of the IPI are important and commonly used
prognostic factors in DLBCL, we included components of the IPI
in a multivariate analysis with both the AMC and ALC as
dichotomized variables. As summarized in Table 2, only the
AMC, ALC and lactate dehydrogenase were independently
significant prognostic factors for OS, with hazard ratios of 2.65
(95% confidence interval 1.82–3.92, Po0.0001), 1.56 (95%
confidence interval 1.10–2.22, P ¼ 0.01) and 1.61 (95%
confidence interval 1.12–2.33, P ¼ 0.01), respectively. Both
the AMC and ALC remained independent predictors of survival
when adjusted for the IPI alone (high risk vs low/intermediate
risk) on multivariate analysis, with hazard ratios of 2.61
(95% confidence interval 1.83–3.80, Po0.0001) and 1.76
(95% confidence interval 1.25–2.47, P ¼ 0.001), respectively.
Similarly, the AMC and ALC were independently significant
predictors of PFS when adjusted for components of the IPI on
multivariate analysis, with hazard ratios of 2.5 (95% confidence
interval 1.6–4.0, Po0.0001) and 3.2 (95% confidence interval
2.0–5.1, Po0.0001), respectively.
An AMC X630/ml was associated with 4.8 years median OS
(94 events observed; 95% confidence interval 3.7–6.8 years)
and an estimated 5-year OS of 49% (95% confidence interval
42–57%). In contrast, median OS has not been reached (43
events observed; 95% confidence interval 8.3–12.2 years) for
patients with an AMC o630/ml, with an estimated 5-year OS of
80% (95% confidence interval 74–86%; Figure 1a). Similarly, an
elevated AMC was associated with inferior PFS (89 vs 46 events
observed; median PFS 5.0 years vs not reached; 5-year PFS 50 vs
74%, Po0.0001), as shown in Figure 1b. Lymphopenia (ALC
p1000/ml) was associated with inferior survival, with a median
OS of 6.3 years (59 events observed; 95% confidence interval
2.9–11.6 years) and PFS of 2.4 years (66 events observed; 95%
confidence interval 1.3–11.6 years) with estimated 5-year OS
and PFS of 53% (95% confidence interval 44–62%) and 44%
(95% confidence interval 35–54%), respectively. In contrast, the
median OS (78 events observed) and PFS (69 events observed)
had not been reached, with estimated 5-year OS and PFS of 70%
(95% confidence interval 64–76%) and 71% (95% confidence
interval 66–78%), respectively, for patients with an ALC 41000/
ml (Figures 1c and d).
The AMC/ALC prognostic score identifies poor-risk
patients and provides additional prognostic information
when superimposed on the IPI
Like the IPI, use of the AMC or ALC as single variables was
unable to identify a subset of truly high-risk patients, even
Leukemia
though each variable alone was able to identify poor-risk
patients with 5-year OS comparable to poor-risk patients
identified by the IPI. Furthermore, factors included in the IPI
may be co-dependent variables. For example, patients in this
cohort with advanced stage disease were more likely to have an
elevated lactate dehydrogenase (Po0.0001). In contrast, the
AMC and ALC are each independent from the other (data not
shown). Therefore, we combined the dichotomized AMC and
ALC to generate the AMC/ALC prognostic score and stratified
patients into three risk groups: low- (AMC o630/ml and ALC
41000/ml), intermediate- (AMC X630/ml or ALC p1000/ml) and
high-risk (AMC X630/ml and ALC p1000/ml) populations.
PFS and OS were analyzed using either the AMC/ALC
prognostic score or the IPI. The estimated 5-year PFS among
low- and intermediate-risk patients stratified by the AMC/ALC
prognostic score was 83 and 59%, respectively (Figure 2a).
Similarly, the estimated 5-year PFS among low-, low-intermediate- and high-intermediate-risk patients was 80, 53 and 46%,
respectively, using the IPI (Figure 2b). The AMC/ALC prognostic
score identified a group of poor-risk patients with a median PFS
of 14 months (95% confidence interval 10–23 months) and an
estimated 5-year PFS of 30%. Similarly, the median PFS was 14
months (95% confidence interval 7–95 months) and the
estimated 5-year PFS was 37% among high-risk patients
identified using the IPI. For both low- and intermediate-risk
patients, identified by either the AMC/ALC prognostic score or
the IPI, median OS either had not been reached or exceeded 6
years (in low-intermediate-risk IPI). In contrast, the median OS
was 2.2 years (95% confidence interval 1.7–3.8 years), with an
estimated 5-year OS of 32%, among high-risk patients identified
using the AMC/ALC prognostic score (Figure 2c). By comparison, a median OS of 2.1 years (95% confidence interval 1.2–7.9
years) was observed in high-risk patients identified by the IPI,
with a 38% 5-year OS (Figure 2d). Clearly, the AMC/ALC
prognostic score was able to risk-stratify patients in a manner
comparable to the IPI.
We next examined the AMC/ALC prognostic score’s performance in the rituximab era by confining our analysis to only
those patients treated with R-CHOP (n ¼ 256). As shown in
Figure 3, the AMC/ALC prognostic score was able to identify
low-, intermediate- and high-risk patients with 5-year PFS
ranging from 26 to 88% and 5-year OS ranging from 28 to 88%
(Figures 3a and c). In contrast, high-risk patients identified by the
IPI experienced a 44% 5-year PFS and OS (Figures 3b and d).
The AMC/ALC prognostic score not only remained prognostically significant in the rituximab era, but was at least as
proficient as the IPI in identifying high-risk patients.
AMC and ALC predict survival in DLBCL
RA Wilcox et al
1505
Figure 1 Kaplan–Meier estimates of overall survival (a, c) and progression-free survival (b, d) in 366 diffuse large-B-cell NHL patients stratified by
the absolute monocyte count (X630/ml (dotted line) vs o630/ml (solid line), in a and b) or the absolute lymphocyte count (p1000/ml (dotted line)
vs 41000/ml (solid line), in c and d) observed at the time of diagnosis are shown.
Figure 2 Kaplan–Meier estimates of progression-free (a, b) and overall (c, d) survival for the entire cohort of patients stratified by the AMC/ALC
prognostic score (a, c) and the International Prognostic Index (b, d) are shown.
The AMC/ALC score remains an independently significant
prognostic factor when adjusting for the IPI. Therefore, we
sought to determine whether it may provide additional prog-
nostic information when combined with the IPI. To test this
possibility, both low- (Figure 4a) and intermediate-risk
(Figure 4b) patients were segregated by the IPI. There were too
Leukemia
AMC and ALC predict survival in DLBCL
RA Wilcox et al
1506
Figure 3 Kaplan–Meier estimates of progression-free (a, b) and overall (c, d) survival for R-CHOP-treated patients stratified by the AMC/ALC
prognostic score (a, c) and the International Prognostic Index (b, d) are shown.
Figure 4 Patients identified by the IPI as either low- (a) or low-intermediate/high-intermediate risk (b) were further stratified into low- (solid line),
intermediate- (dotted line) or high-risk (dashed line) groups by the AMC/ALC prognostic score (Kaplan–Meier estimates of overall survival are
shown).
few high-risk patients to make a similar analysis in this subgroup
meaningful. The 158 low-risk and 169 intermediate-risk (highintermediate and low-intermediate were combined) patients
identified by the IPI were subsequently risk stratified using the
AMC/ALC prognostic score. In both cases, the AMC/ALC score
was able to further risk-stratify these patients, although this
was most significant in patients identified as intermediate risk by
the IPI (Figure 4b). Among these patients, 5-year OS ranged from
64 to 77% among low- and intermediate-risk patients identified
by the AMC/ALC prognostic score. In contrast, 21% of patients
identified by the IPI as ‘intermediate-risk’, upon further riskstratification by the AMC/ALC prognostic score, were found to
have dismal outcomes, with a median OS of only 19 months
Leukemia
(95% confidence interval 12–32 months) and 5-year OS of 21%
(Figure 4b). Similar results were obtained when intermediaterisk patients treated with R-CHOP were risk-stratified by the
AMC/ALC prognostic score (Figure 5). In this case, the high-risk
subgroup experienced a median OS of 17 months (95%
confidence interval 11–26 months) and 5-year OS of 17%.
The AMC/ALC prognostic score, derived from values easily
obtained from a CBC, is comparable to the IPI in its ability to risk
stratify patients in the pre- and post-rituximab era. Furthermore,
it provides additional prognostic information when used to
sub-stratify risk groups obtained from the IPI, leading to the
identification of high-risk patients otherwise classified as
intermediate risk by the IPI alone.
AMC and ALC predict survival in DLBCL
RA Wilcox et al
1507
Figure 5 Patients treated with R-CHOP identified by the IPI as lowintermediate/high-intermediate risk were further stratified into low(solid line), intermediate- (dotted line) or high-risk (dashed line) groups
by the AMC/ALC prognostic score (Kaplan–Meier estimates of overall
survival are shown).
Discussion
Recent work, based on large gene-expression profiling and
immunohistochemistry-based studies in a variety of histological
subtypes of NHL, shows that the quality of tumor-infiltrating
lymphocytes and myeloid-derived cells provide valuable prognostic information that, in some cases, may exceed that
provided by consideration of the tumor characteristics
alone.3,19,34 This implies that a prognostic system that considers
features of the tumor-bearing host and the tumor microenvironment may provide prognostic information that is independent
from conventional prognostic models. Therefore, we sought to
investigate whether the AMC and ALC, as biomarkers for host
immunity and the tumor microenvironment, affect survival in
DLBCL.
Myeloid-lineage cells, including monocytes and their progeny, promote tumorigenesis.35–38 These myeloid-derived cells
contribute to the suppression of host antitumor immunity and
have an important role in tumor angiogenesis. Not surprisingly
then, development of peripheral blood neutrophilia or monocytosis are adverse prognostic factors in multiple solid
tumors.39–46 Multiple mechanisms, none of which are mutually
exclusive, may be invoked to explain the prognostic significance
of a relative monocytosis in DLBCL. First, monocytes and
monocytic myeloid-derived suppressor cells suppress host
antitumor immunity by a variety of different mechanisms.47,48
For example, we have previously shown that both peripheral
blood monocytes and their progeny within the tumor microenvironment inducibly express the T-cell co-inhibitory ligand
B7-H1 (PD-L1).47 B7-H1 expression was found to directly
impair the expansion and effector functions of tumor-specific
T cells, and when expressed by myeloid-lineage cells within the
tumor microenvironment led to the expansion of suppressive
regulatory T cells. Next, a considerable body of evidence
highlights the role of myeloid-lineage cells in supporting tumor
angiogenesis, including a subpopulation of monocytes expressing the angiopoietin-2 receptor Tie2.49,50 Given the role of
macrophages in lymphomagenesis, peripheral blood monocytes
may represent an important reservoir of macrophage precursors.
To address this possibility, the density of CD68 þ lymphomaassociated macrophages in lymph nodes was examined in this
study (data not shown). Although a direct correlation between
the AMC and the density of LAMs may have been anticipated,
no such correlation was observed. This suggests that the
heterogeneity of LAM density observed may be explained by
differences in monocyte recruitment and the retention of tissueresident macrophages, as opposed to the AMC. Finally,
monocytes are an important source of soluble mediators, like
B-lymphocyte stimulator, that support the growth and survival of
both normal and malignant B cells.51–53 Not surprisingly then,
monocytes have been convincingly shown to support the growth
of malignant lymphocytes in both T- and B-cell NHL.26,54
The vast majority of DLBCL originate from either germinalcenter B cells (that is, GCB type) or from B cells with a geneexpression profile resembling activated B cells (that is, ABC
type). Each DLBCL subtype may preferentially exploit a
particular pathway during lymphomagenesis. For example, both
the STAT3 and nuclear factor-kB pathways have been implicated in the pathogenesis of ABC-DLBCL and chemokine and
cytokine genes targeted by STAT3 and nuclear factor-kB
regulate monocyte recruitment and ultimately LAM function
and polarity. Similarly, inflammatory and hematopoietic cytokines that contribute to monocyte expansion (and the ‘leukemoid reaction’) may be differentially expressed in DLBCL,
raising the possibility that differences in the AMC may, at least in
part, be explained by differences in the cell of origin (either GCB
or non-GCB). To examine this possibility, DLBCL specimens, in
which immunohistochemical staining for CD10, BCL-6 and
MUM-1 was performed at diagnosis (n ¼ 42), were classified as
either GCB or non-GCB in origin using the Hans algorithm (data
not shown).14 No significant difference in the mean AMC was
observed between GCB and non-GCB DLBCL and the AMC
remained an independent prognostic factor when adjusting for
the cell of origin on multivariate analysis (data not shown).
Although this observation, limited by the small sample size,
suggests that the cell of origin in DLBCL may not be an
important determinant of monocyte expansion in DLBCL, it
certainly does not exclude the real possibility that genes
differentially expressed in GCB- or non-GCB-type DLBCL
regulate the biology of monocytes and their progeny within
the tumor microenvironment.
The contribution of immune suppression in lymphomagenesis, illustrated by post-transplant lymphoproliferative
disease and HIV-associated lymphomas, highlights the important
contribution of T and NK cells in immune surveillance.
Therefore, lymphopenia is considered a surrogate marker of
host immunological incompetence.24,55 In addition, lymphocytes (including natural killer cells) are important mediators of
antibody-dependent cell-mediated cytotoxicity, and may be
required for rituximab-mediated, antibody-dependent cellmediated cytotoxicity-dependent destruction of malignant B
cells.56 Not surprisingly then, lymphopenia is an adverse
prognostic factor in indolent and aggressive NHL, including
DLBCL. For example, lymphopenia is a poor-risk feature in
DLBCL patients treated with either CHOP or R-CHOP and in
GCB- and non-GCB-type DLBCL.23,31 More recently, the
development of new-onset lymphopenia following autologousperipheral blood stem cell transplantation was associated with a
significant increase in the risk of disease relapse.32 Patients with
an ALC o1000/ml following autologous-peripheral blood stem
cell transplantation had an incidence of disease relapse of 92%,
compared with a relapse rate of 19% among patients without
lymphopenia.
The evidence linking both myeloid-lineage cells and lymphopenia with lymphomagenesis, and in this study with poor
PFS and OS, raises the possibility that strategies which deplete
tumorigenic myeloid-lineage cells or reverse lymphopenia may
Leukemia
AMC and ALC predict survival in DLBCL
RA Wilcox et al
1508
represent novel therapeutic strategies in DLBCL. Therapeutic
strategies targeting myeloid-lineage cells, including myeloidderived suppressor cells, are currently being investigated in solid
tumors and may warrant further investigation in NHL.57 In
contrast, manipulation of the lymphocyte content administered
with autologous stem cells during autologous-peripheral blood
stem cell transplantation, in hopes of raising the ALC and
improving survival, is the subject of an ongoing phase III clinical
trial in relapsed NHL (NCT00566228).
Given the need for improved prognostic models in DLBCL,
other predictors have been investigated. Some of these, initially
identified before the introduction of rituximab, no longer
provide useful prognostic information in the rituximab era.8,9
Other predictive models based on gene-expression profiling,
while prognostically useful, are cumbersome, difficult to
interpret, require the availability of suitable biopsy material
and are cost prohibitive.3,5 Although the introduction of interim
positron emission tomography analysis and risk-adapted therapy
is being explored, this will require further validation, as the
interpretation of interim positron emission tomography scans is
not always reproducible and the rate of false-positive interim
scans is not insignificant.17,18 Despite the growing body of
evidence highlighting the relationship between the tumor
microenvironment and lymphomagenesis, the incorporation of
prognostic variables associated with the tumor microenvironment into routine clinical practice has remained a challenge, as
most studies exploring this relationship are based on geneexpression profiling or immunohistochemistry data. In contrast,
the AMC/ALC prognostic score, derived from a CBC at
diagnosis, is widely available and may be easily incorporated
into clinical practice. Furthermore, the AMC/ALC prognostic
score provided prognostic information independently of that
included in the IPI and was able to provide additional prognostic
information when used in conjunction with the IPI. This was
particularly true for low-intermediate- and high-intermediaterisk patients, which collectively accounted for nearly 50% of the
patients in our series. When patients identified as intermediate
risk by the IPI were further risk stratified by the AMC/ALC
prognostic score, approximately 20% of these patients were
identified as high risk, as evidenced by the fewer than 20% longterm survivors, even in the rituximab era.
Given the limited number of patients included in this
retrospective study, the AMC/ALC prognostic score will require
validation in an independent cohort of patients. Despite the
need for independent validation, the AMC/ALC prognostic score
demonstrates that variables related to the host immune response
and the tumor microenvironment are worth consideration when
generating prognostic indices in NHL, as these variables may
provide additional prognostic information independently from
conventional patient- and tumor-specific factors.
Conflict of interest
The authors declare no conflict of interest.
Author contributions
RAW designed study, performed research, analyzed data and
wrote the manuscript; KR performed research, provided
technical assistance and reviewed the manuscript; TMH, DJI,
INMM, PBJ, JPC, GSN, SMA, TEW and SNM assisted with data
analysis and manuscript preparation; LP assisted with study
design, data analysis and manuscript preparation.
Leukemia
References
1 Sehn LH, Berry B, Chhanabhai M, Fitzgerald C, Gill K, Hoskins P
et al. The revised International Prognostic Index (R-IPI) is a better
predictor of outcome than the standard IPI for patients with diffuse
large B-cell lymphoma treated with R-CHOP. Blood 2007; 109:
1857–1861.
2 Ziepert M, Hasenclever D, Kuhnt E, Glass B, Schmitz N,
Pfreundschuh M et al. Standard International Prognostic Index
remains a valid predictor of outcome for patients with aggressive
CD20+ B-cell lymphoma in the rituximab era. J Clin Oncol 2010;
28: 2373–2380.
3 Lenz G, Wright G, Dave SS, Xiao W, Powell J, Zhao H et al.
Stromal gene signatures in large-B-cell lymphomas. N Engl J Med
2008; 359: 2313–2323.
4 Alizadeh AA, Eisen MB, Davis RE, Ma C, Lossos IS, Rosenwald A
et al. Distinct types of diffuse large B-cell lymphoma identified by
gene expression profiling. Nature 2000; 403: 503–511.
5 Lossos IS, Czerwinski DK, Alizadeh AA, Wechser MA, Tibshirani
R, Botstein D et al. Prediction of survival in diffuse large-B-cell
lymphoma based on the expression of six genes. N Engl J Med
2004; 350: 1828–1837.
6 Rosenwald A, Wright G, Chan WC, Connors JM, Campo E, Fisher
RI et al. The use of molecular profiling to predict survival after
chemotherapy for diffuse large-B-cell lymphoma. N Engl J Med
2002; 346: 1937–1947.
7 Shipp MA, Ross KN, Tamayo P, Weng AP, Kutok JL, Aguiar RC
et al. Diffuse large B-cell lymphoma outcome prediction by geneexpression profiling and supervised machine learning. Nat Med
2002; 8: 68–74.
8 Hermine O, Haioun C, Lepage E, d’Agay MF, Briere J, Lavignac C
et al. Prognostic significance of bcl-2 protein expression in
aggressive non-Hodgkin’s lymphoma. Groupe d’Etude des Lymphomes de l’Adulte (GELA). Blood 1996; 87: 265–272.
9 Mounier N, Briere J, Gisselbrecht C, Emile JF, Lederlin P, Sebban C
et al. Rituximab plus CHOP (R-CHOP) overcomes bcl-2-associated
resistance to chemotherapy in elderly patients with diffuse large Bcell lymphoma (DLBCL). Blood 2003; 101: 4279–4284.
10 Ansell SM, Stenson M, Habermann TM, Jelinek DF, Witzig TE.
Cd4+ T-cell immune response to large B-cell non-Hodgkin’s
lymphoma predicts patient outcome. J Clin Oncol 2001; 19:
720–726.
11 Tzankov A, Meier C, Hirschmann P, Went P, Pileri SA,
Dirnhofer S. Correlation of high numbers of intratumoral FOXP3+
regulatory T cells with improved survival in germinal center-like
diffuse large B-cell lymphoma, follicular lymphoma and classical
Hodgkin’s lymphoma. Haematologica 2008; 93: 193–200.
12 Lee NR, Song EK, Jang KY, Choi HN, Moon WS, Kwon K et al.
Prognostic impact of tumor infiltrating FOXP3 positive regulatory T
cells in diffuse large B-cell lymphoma at diagnosis. Leuk Lymph
2008; 49: 247–256.
13 Hasselblom S, Sigurdadottir M, Hansson U, Nilsson-Ehle H, Ridell
B, Andersson PO. The number of tumour-infiltrating TIA-1+
cytotoxic T cells but not FOXP3+ regulatory T cells predicts
outcome in diffuse large B-cell lymphoma. Br J Haematol 2007;
137: 364–373.
14 Hans CP, Weisenburger DD, Greiner TC, Gascoyne RD, Delabie J,
Ott G et al. Confirmation of the molecular classification of diffuse
large B-cell lymphoma by immunohistochemistry using a tissue
microarray. Blood 2004; 103: 275–282.
15 Nyman H, Jerkeman M, Karjalainen-Lindsberg ML, Banham AH,
Leppa S. Prognostic impact of activated B-cell focused classification in diffuse large B-cell lymphoma patients treated with
R-CHOP. Mod Pathol 2009; 22: 1094–1101.
16 Nyman H, Adde M, Karjalainen-Lindsberg ML, Taskinen M,
Berglund M, Amini RM et al. Prognostic impact of immunohistochemically defined germinal center phenotype in diffuse large
B-cell lymphoma patients treated with immunochemotherapy.
Blood 2007; 109: 4930–4935.
17 Mikhaeel NG. Interim fluorodeoxyglucose positron emission tomography for early response assessment in diffuse large B cell
lymphoma: where are we now? Leuk Lymph 2009; 50: 1931–1936.
18 Duhrsen U, Huttmann A, Jockel KH, Muller S. Positron emission
tomography guided therapy of aggressive non-Hodgkin lymphomas
Fthe PETAL trial. Leuk Lymph 2009; 50: 1757–1760.
AMC and ALC predict survival in DLBCL
RA Wilcox et al
1509
19 Dave SS, Wright G, Tan B, Rosenwald A, Gascoyne RD, Chan WC
et al. Prediction of survival in follicular lymphoma based on
molecular features of tumor-infiltrating immune cells. N Engl J
Med 2004; 351: 2159–2169.
20 Oki Y, Yamamoto K, Kato H, Kuwatsuka Y, Taji H, Kagami Y et al.
Low absolute lymphocyte count is a poor prognostic marker in
patients with diffuse large B-cell lymphoma and suggests patients’
survival benefit from rituximab. Eur J Haematol 2008; 81:
448–453.
21 Behl D, Ristow K, Markovic SN, Witzig TE, Habermann TM,
Colgan JP et al. Absolute lymphocyte count predicts therapeutic
efficacy of rituximab therapy in follicular lymphomas. Br J
Haematol 2007; 137: 409–415.
22 Cox MC, Nofroni I, Laverde G, Ferrari A, Amodeo R, Tatarelli C
et al. Absolute lymphocyte count is a prognostic factor in diffuse
large B-cell lymphoma. Br J Haematol 2008; 141: 265–268.
23 Kim DH, Baek JH, Chae YS, Kim YK, Kim HJ, Park YH et al.
Absolute lymphocyte counts predicts response to chemotherapy
and survival in diffuse large B-cell lymphoma. Leukemia 2007; 21:
2227–2230.
24 Fruehauf S, Tricot G. Comparison of unmobilized and mobilized
graft characteristics and the implications of cell subsets on
autologous and allogeneic transplantation outcomes. Biol Blood
Marrow Transplant 2010; 16: 1629–1648.
25 Shivakumar L, Ansell S. Targeting B-lymphocyte stimulator/B-cell
activating factor and a proliferation-inducing ligand in hematologic
malignancies. Clin Lymph Myel 2006; 7: 106–108.
26 Wilcox RA, Wada DA, Ziesmer SC, Elsawa SF, Comfere NI,
Dietz AB et al. Monocytes promote tumor cell survival in
T-cell lymphoproliferative disorders and are impaired in their
ability to differentiate into mature dendritic cells. Blood 2009; 114:
2936–2944.
27 Cox CJ, Habermann TM, Payne BA, Klee GG, Pierre RV.
Evaluation of the Coulter Counter model S-Plus IV. Am J Clin
Pathol 1985; 84: 297–306.
28 Anonymous. A predictive model for aggressive non-Hodgkin’s
lymphoma. The International Non-Hodgkin’s Lymphoma Prognostic Factors Project. N Engl J Med 1993; 329: 987–994.
29 Kaplan EL, Meier P. Nonparametric estimation from incomplete
observations. J Am Stat Assoc 1958; 53: 457–481.
30 Cheson BD, Pfistner B, Juweid ME, Gascoyne RD, Specht L,
Horning SJ et al. Revised response criteria for malignant
lymphoma. J Clin Oncol 2007; 25: 579–586.
31 Song MK, Chung JS, Seol YM, Kim SG, Shin HJ, Choi YJ et al.
Influence of low absolute lymphocyte count of patients with
nongerminal center type diffuse large B-cell lymphoma with
R-CHOP therapy. Ann Oncol 2010; 21: 140–144.
32 Porrata LF, Inwards DJ, Ansell SM, Micallef IN, Johnston PB,
Hogan WJ et al. New-onset lymphopenia assessed during routine
follow-up is a risk factor for relapse postautologous peripheral
blood hematopoietic stem cell transplantation in patients with
diffuse large B-cell lymphoma. Biol Blood Marrow Transplant
2010; 16: 376–383.
33 Cox DR. Regression models and life tablets. J R Statist Soc Ser C
1972; 34: 187–202.
34 Ansell SM, Stenson M, Habermann TM, Jelinek DF, Witzig TE.
Cd4+ T-cell immune response to large B-cell non-Hodgkin’s
lymphoma predicts patient outcome. J Clin Oncol 2001; 19:
720–726.
35 Young MR, Newby M, Wepsic HT. Hematopoiesis and suppressor
bone marrow cells in mice bearing large metastatic Lewis lung
carcinoma tumors. Cancer Res 1987; 47: 100–105.
36 Young MR, Ellis NK, Young ME, Wepsic HT. Stimulation of
hematopoiesis and bone marrow suppressor cells by the subcutaneous injection of linoleic acid. Cell Immunol 1987; 107:
238–248.
37 Gabrilovich DI, Nagaraj S. Myeloid-derived suppressor cells as
regulators of the immune system. Nat Rev Immunol 2009; 9: 162–174.
38 Seung LP, Rowley DA, Dubey P, Schreiber H. Synergy between
T-cell immunity and inhibition of paracrine stimulation causes
tumor rejection. Proc Natl Acad Sci USA 1995; 92: 6254–6258.
39 Mandrekar SJ, Schild SE, Hillman SL, Allen KL, Marks RS, Mailliard
JA et al. A prognostic model for advanced stage nonsmall cell lung
cancer. Pooled analysis of North Central Cancer Treatment Group
trials. Cancer 2006; 107: 781–792.
40 Schmidt H, Bastholt L, Geertsen P, Christensen IJ, Larsen S, Gehl J
et al. Elevated neutrophil and monocyte counts in peripheral blood
are associated with poor survival in patients with metastatic
melanoma: a prognostic model. Br J Cancer 2005; 93: 273–278.
41 Schmidt H, Suciu S, Punt CJ, Gore M, Kruit W, Patel P et al.
Pretreatment levels of peripheral neutrophils and leukocytes as
independent predictors of overall survival in patients with
American Joint Committee on Cancer Stage IV Melanoma: results
of the EORTC 18951 Biochemotherapy Trial. J Clin Oncol 2007;
25: 1562–1569.
42 Michael M, Goldstein D, Clarke SJ, Milner AD, Beale P, Friedlander
M et al. Prognostic factors predictive of response and survival to a
modified FOLFOX regimen: importance of an increased neutrophil
count. Clin Colorectal Cancer 2006; 6: 297–304.
43 Choueiri TK, Garcia JA, Elson P, Khasawneh M, Usman S,
Golshayan AR et al. Clinical factors associated with outcome in
patients with metastatic clear-cell renal cell carcinoma treated
with vascular endothelial growth factor-targeted therapy. Cancer
2007; 110: 543–550.
44 Donskov F, von der Maase H. Impact of immune parameters on
long-term survival in metastatic renal cell carcinoma. J Clin Oncol
2006; 24: 1997–2005.
45 Paesmans M, Sculier JP, Lecomte J, Thiriaux J, Libert P, Sergysels R
et al. Prognostic factors for patients with small cell lung carcinoma:
analysis of a series of 763 patients included in 4 consecutive
prospective trials with a minimum follow-up of 5 years. Cancer
2000; 89: 523–533.
46 Takasaki Y, Iwanaga M, Tsukasaki K, Kusano M, Sugahara K,
Yamada Y et al. Impact of visceral involvements and blood cell
count abnormalities on survival in adult T-cell leukemia/lymphoma
(ATLL). Leuk Res 2007; 31: 751–757.
47 Wilcox RA, Feldman AL, Wada DA, Yang ZZ, Comfere NI, Dong
H et al. B7-H1 (PD-L1, CD274) suppresses host immunity in T-cell
lymphoproliferative disorders. Blood 2009; 114: 2149–2158.
48 Wilcox RA. Cancer-associated myeloproliferation: old association,
new therapeutic target. Mayo Clin Proc 2010; 85: 656–663.
49 De Palma M, Venneri MA, Galli R, Sergi Sergi L, Politi LS,
Sampaolesi M et al. Tie2 identifies a hematopoietic lineage of
proangiogenic monocytes required for tumor vessel formation and
a mesenchymal population of pericyte progenitors. Cancer Cell
2005; 8: 211–226.
50 Venneri MA, De Palma M, Ponzoni M, Pucci F, Scielzo C, Zonari E
et al. Identification of proangiogenic TIE2-expressing monocytes
(TEMs) in human peripheral blood and cancer. Blood 2007; 109:
5276–5285.
51 He B, Chadburn A, Jou E, Schattner EJ, Knowles DM, Cerutti A.
Lymphoma B cells evade apoptosis through the TNF family
members BAFF/BLyS and APRIL. J Immunol 2004; 172: 3268–3279.
52 Novak AJ, Grote DM, Stenson M, Ziesmer SC, Witzig TE,
Habermann TM et al. Expression of BLyS and its receptors in
B-cell non-Hodgkin lymphoma: correlation with disease activity
and patient outcome. Blood 2004; 104: 2247–2253.
53 Oki Y, Georgakis GV, Migone TS, Kwak LW, Younes A. Elevated
serum BLyS levels in patients with non-Hodgkin lymphoma. Leuk
Lymph 2007; 48: 1869–1871.
54 Seiffert M, Schulz A, Ohl S, Dohner H, Stilgenbauer S, Lichter P.
Soluble CD14 is a novel monocyte-derived survival factor for
chronic lymphocytic leukemia cells, which is induced by CLL cells
in vitro and present at abnormally high levels in vivo. Blood 2010;
116: 4223–4230.
55 Porrata LF, Markovic SN. Timely reconstitution of immune
competence affects clinical outcome following autologous stem
cell transplantation. Clin Exp Med 2004; 4: 78–85.
56 Weiner GJ. Rituximab: mechanism of action. Semin Hematol
2010; 47: 115–123.
57 Ko JS, Bukowski RM, Fincke JH. Myeloid-derived suppressor cells:
a novel therapeutic target. Curr Oncol Rep 2009; 11: 87–93.
Leukemia