Annuls of Oncology 12: 1423-1432. 2001.
© 2001 Kluwer Academic Publishers. Printed in the Netherlands.
Original article
Alterations of routine blood tests in adult patients with soft tissue sarcomas:
Relationships to cytokine serum levels and prognostic significance
W. Ruka,1 P. Rutkowski,1 J. Kaminska," A. Rysinska^ & J. Steffen
3
'Department oj Soft Tissue/Bone Sarcomas, 'Department of Tumor Markers. ' Department of Immunology. M. Sklodowska-Curie Memorial
Cancer Center and Institute oj Oncologv Warsaw, Poland
Summary
Background: It has been reported that malignancy is often
accompanied by hematological alterations and that such alterations may correlate with poor prognosis. It has also been
demonstrated that several cytokines may be synthesized by
many malignant tumors and that elevated serum levels of
some cytokines are associated with changes in blood cell
counts in cancer patients. However, so far little is known about
the prognostic significance and mechanism of hematological
changes in soft tissue sarcomas. The aim of the study was to
evaluate the routine blood tests of disturbances in patients with
malignant soft-tissue tumors prior to treatment and to correlate these results with selected cytokine serum levels, clinicopathological features of the tumors and patient survival.
Patients and methods: 145 patients (75 males, 70 females;
mean age 49.97 ± 16.9 yrs) with histologically confirmed soft
tissue sarcomas before treatment were enrolled into the study.
In all these patients we evaluated routine blood tests (hemoglobin level HGB, white blood cell count WBC, platelet count
PLT, white blood cell differential count-neutrocyte count NE,
lymphocyte count LY, monocyte count MN, eosinophile count
EO) and serum levels of 13 cytokines and soluble cytokine
receptors (IL-6, IL-8, IL-10, TNFa, G-CSF, M-CSF, bFGF,
VEGF, IL-lra, sIL-2R, sIL-6R, TNF RI.TNF R I I ) - ELISA
method. Peripheral blood samples from 50 healthy volunteers
served as control. Statistical analysis was performed using
Kolmogorov-Smirnov and Mann-Whitney U-tests, X~ test
(P < 0.05), where appropriate. For survival analysis the
Kaplan-Meier method, log-rank test and multivariate Cox
analysis were applied.
Introduction
Hematological changes occur quite frequently in patients
with various malignant solid tumors [1-6] and, as reported by some authors, may correlate significantly with
patients' prognosis [4, 7-9]. Although the mechanisms
responsible for these anomalies are still poorly understood, it is well known from in vitro and in vivo studies
that both colony-stimulating factors and proinflammatory cytokines are essential in hematopoiesis and can
also affect the number of neutrophils, lymphocytes and
monocytes in the blood [2, 10, 11].
Cytokines are a group of regulatory proteins, which act
Results: Alterations of at least one of the standard blood
tests were found in 43.4% of all cases. The most frequent
alterations were: neutrophilia (28.3% of cases), leukocytosis
(27.6%). decreased HGB (25.5%), monocytosis (19.3%) and
thrombocytosis (14.5%); they correlated strongly with elevated
serum levels of several cytokines and soluble cytokine receptors (particularly: sIL-2R, IL-6, IL-8, M-CSF, VEGF. T N F
Rl, TNF RII) (P < 0.001). Lymphocytopenia (LY < 1.0)
found in 10.3% of patients correlated strongly with increased
serum levels of IL-6, sIL-2R. TNF RI. In parallel, we found
a significant difference in serum levels of II of 13 cytokines
(IL-lra, sIL-2R. IL-6, IL-8. IL-10,TNF RI.TNF RII.TNFa,
M-CSF, bFGF, VEGF) (P < 0.001) in soft tissue sarcoma
patients compared to healthy controls. Hematological alterations were significantly more frequent in patients with advanced
tumors. In multivariate analysis we found no prognostic significance of any of the routine blood tests in soft tissue
sarcoma patients.
Conclusion: The results of this study demonstrate that
hematological alterations, which occur in over 40% of soft
tissue sarcoma cases, are found more frequently in patients
with advanced tumors. Strong correlations between the occurrence of hematological abnormalities and elevated serum levels
of several cytokines and soluble cytokine receptors, suggest
that the former may develop as a result of cytokine misbalance
frequently detected in soft tissue sarcoma patients. However,
the results of routine blood tests alone are no independent
prognostic factor for survival of soft-tissue sarcoma patients.
Key words: blood cell count, cytokines, prognosis, soft tissue
sarcomas
as mediators in the control of cell growth, survival and
differentiation in various tissues in normal renewal, and
repair of damage. They are produced by particular cells in
response to appropriate stimuli [12-14]. In malignancy,
various cytokines might be synthesized and released by
tumor and stromal cells as well as by immune cells
attracted to the tumor site [15-18]. Cytokines are involved
at every stage of tumor growth and dissemination, and
act (combined with other factors) as autocrine-paracrine tumor growth factors. Elevated cytokine serum
levels in cancer patients may have important prognostic
value [17, 19-24].
Moreover, it has been reported that high serum levels
1424
of interleukin 6 in cancer patients correlate with thrombocythemia [25, 26], increased polymorphonuclear leukocyte and monocyte counts [27], as well as with decreased
hemoglobin [28], and enhanced release of G-CSF from
tumor cells may result in peripheral neutrophilia [2].
There are only a few reports about hematological
changes and their relationship to cytokine serum levels
[2, 29-32] in such a rare group of malignant solid tumors
as soft tissue sarcomas.
We have previously reported [33] that serum levels of
several proinflammatory, proangiogenic and hematopoietic cytokines, as well as some of their soluble receptors, are frequently elevated in patients with various
non-lymphoid malignant tumors, both of epithelial and
mesenchymal origin. More than 50% of patients with
soft tissue sarcomas who were included in the above
study had increased serum levels of IL-6, IL-8, IL-10,
IL-lra, M-CSF and sIL-2R, and 20% to 50% had
increased serum levels of VEGF, bFGF, TNF RI,
TNFRII, TNFot and G-CSF. Thirteen cytokines and
their soluble receptors, which were selected for the
present analysis, represent the major subtypes of cytokines (interleukins, tumor necrosis factor family members, hematopoietic colony-stimulating factors, chemokines, angiogenic factors) [12, 34]. It is also significant
for the criteria of choice of analyzed cytokines that most
of the selected cytokines are suspected of playing a role
in the pathogenesis of other solid neoplasms [20-22,
35-37].
The aim of this study was to evaluate the pretreatment routine hematological parameters in patients with
untreated malignant soft tissue tumors and to correlate
these results with serum levels of selected cytokines and
clinicopathological features of the tumors. Special interest was paid to a possible prognostic significance of
hematological tests in soft tissue sarcomas. Correlations
between serum cytokine levels and clinicopathological
features and survival of soft tissue sarcoma patients will
be the subject of a separate report.
Patients and methods
Patients
Between April 1997 and December 1999, 145 patients with soft tissue
sarcomas who underwent treatment in the Soft Tissue/Bone Sarcomas
Department of the M. Sklodowska-Curie Memorial Cancer Center
and Institute of Oncology (MSCMCCI) Warsaw, Poland, were enrolled
into the study. There were 75 men and 70 women of mean age 49.9 ±
16.9 years (ranging from 18 to 80). All patients were selected on the
basis of the following criteria: histologically proven soft tissue sarcoma, including primary tumors before treatment (74 cases) and locally
recurrent tumors (71 cases) (all recurrences occurred more than three
months after primary surgery). None of these patients had received
prior radiotherapy or chemotherapy. At the time of presentation 113
patients had no distant metastases (M0) and 32 patients were in stage
Ml with lung metastases. None of the patients had evidence of a
primary myeloproliferative disorder. Patients had not undergone any
other preliminary selection. Only patients who met with these conditions were included in the study. All of the patients provided informed
consent in accordance with institutional regulatory requirements
Table I. Patient characteristics.
Mean value ± standard
deviation (percentage)
Age
Gender
Male
Female
Tumor pathology
Liposarcomas
Synovial sarcomas
MFH
Malignant schwannomas
Leiomyosarcomas
Others
Tumor status
Primary
Recurrent
Tumor size
< 10 cm
> 10 cm
Tumor grade
Gl
G2-3
Disease status
No metastases M0
Lung metastases M1
49.9116.9 years
75 (52%)
70 (48%)
31 (21 4%)
23 (15.8%)
22 (15 2%)
20(13.8%)
18 (12.4%)
31 (21.4%)
74(51%)
71 (49%)
59 (41%)
86 (59%)
30 (21%)
115 (79%)
113 (78%)
32 (22%)
Peripheral blood samples were collected from all patients before
treatment and from 50 healthy volunteers (26 male. 24 female; mean
age 38.9 ± 15.1 years; range 19-71 years) without hematological abnormalities served as controls. The control group was fully comparable
with the studied group of patients according to gender and age
distribution. For serum cytokine evaluations a 30 ml blood specimen
was obtained from each patient between 7.30 and 11.00 a.m. after
overnight fasting. Sera were stored at -70 °C until analysis.
In all patients blood samples for routine blood tests were collected
simultaneously.
Clinicopathological features of malignant soft tissue tumors were
determined by physical examination, pathological examination of the
primary lesion and routine imaging examination (chest X-ray, tumorcomputed tomography, ultrasonography of the abdominal cavity) at
the time of blood collection. According to pathological examination
there were 31 patients with liposarcomas (21.4%). 23 with synovial
sarcomas (15.8%), 22 with malignant fibrous histiocytomas (MFH)
(15 2%). 20 with malignant schwannomas (13.8%). 18 with leiomyosarcomas (12.4%) and 31 patients with other types of sarcomas (21.4%).
Patient characteristics are shown in Table I.
Patients were stratified on the basis of tumor size (less and more
than 10 cm according to the dichotomization suggested by Gustaflson
[38]) and tumor grade (Gl to G3 degree according to the standard
histological criteria). Patients were also divided into two groups according to age: 1) younger =£49 years and 2) older >49 years (based
on the mean age at the time of presentation in the studied group). The
median patients follow-up time was 26.1 months.
The patients'clinical status was not known by the persons running
cytokine and hematological assays, while assay results were not known
to the surgeons recording the disease status.
Cytokine asssays
Sera were assayed for levels of 13 cytokines and soluble cytokine
receptors with ELISA method using commercially available kits
(R&D Systems, Minneapolis; Minnesota) according to the recommen-
1425
dations of the manufacturer. The levels of following 13 cytokines and
cytokine receptors were measured: intcrleukin 6 (IL-6); interleukin 8
(1L-8); tumor necrosis factor a (TNFa); tumor necrosis factor receptor 1 (TNF RI); tumor necrosis factor receptor II (TNF R1I): soluble
interleukin 2 receptor (sIL-2R), soluble interleukin 6 receptor (si L6R); interleukin 1 receptor antagonist (lL-Ira); interleukin 10 (IL-10):
vascular endothelial growth factor (VEGF): basic fibroblast growth
factor (bFGF); granulocyte colony-stimulating factor (G-CSF); macrophage colony-stimulating factor (M-CSF). Reference values for each
cytokine level were established in the Department of Tumor Markers
from a study of 50 healthy volunteers and cut-off values were as follows
(established on the basis of 5 and 95 percentile in the control group):
IL-6 (0.7-2.4 pg/ml); IL-8 (10-12.5 pg/ml): TNFa (<4.4 pg/ml);TNF
RI (600-1200 pg/ml); TNF Rll (1577-3500 pg/ml); siL-2R (758-1753
pg/ml); slL-6R (<55070 pg/ml); IL-lra (98-473 pg/ml), 1L-I0 (2.07.3 pg/ml); VEGF (< 325 pg/ml): bFGF (<4.4 pg/ml): M-CSF (100600 pg/ml); G-CSF (7.0-34.0 pg/ml).
Hematological analyses
We have incorporated blood laboratory tests, which are a part of
routine admission tests for all patients in MSCMCC. We evaluated the
results of the following blood tests: hemoglobin level HGB (reference
value: 11.0-18.0 g/dl), platelet count PLT (140-450 x IO9/1). white
blood cell count WBC (4-10 x 109/l), white blood cell differential
count-neutrocyte count NE (2.3-6.6 x I09/]). lymphocyte count LY
(1.0-3.5 x |0 9 /l), monocyte count MN (0.1-0.8 x |0 9 /l), eosinophile
count EO (0-0.5 x 10'Vl).
Statistical analysis
All statistical analyses were performed using STATIST1CA Software
(StatSoft) with the Mann-Whitney U-test, Smirnov-Kolmogorov
(only significant differences for both tests were considered significant)
and x2 test for comparison between two groups, where appropriate. We
used the Kaplan-Meier method, log-rank test and multivariate Cox
analysis for overall survival (OS) analysis. Differences at P < 0.05 were
considered to be significant.
Results
Hematological test abnormalities
Of the 145 patients enrolled, one or more abnormal
blood tests were found in 63 cases (43.4%).
The following blood test alterations were most frequent: neutrophilia (41 cases - 28.3%), leucocytosis (40
cases - 27.6%), decreased HGB (37 cases - 25.5%),
monocytosis (28 cases - 19.3%) and thrombocytosis
(21 cases - 14.5%). Fifteen patients (10.3%) displayed
decreased complete lymphocyte counts LY < 1.0 x 10s*/!.
Eleven patients (7.6%) presented eosinophilia (EO >
0.5 x 1O9/1).
Table 2 summarizes hematological data by tumor
stage. We observed that PLT, WBC, NE and MN were
significantly increased in more advanced stages of the
tumors (size and grade) (Figure 4). The PLT and NE
counts were also significantly higher in Ml vs. M0
patients. HGB and LY correlated inversely only with the
tumor size, while EO correlated with none of the clinicopathological features of studied tumors.
Taking into account all blood tests alterations - the
percentage of patients with hematological disturbances
was significantly higher in the group of patients with
larger tumors > 10 cm (49 of 86 cases 57%) than in
patients with smaller tumors < 10 cm (14 of 59 cases
23.7%) (P < 0.001).
We also found a significant inverse relationship between PLT and HGB. In the group of patients with
thrombocytosis the mean hemoglobin level was: 10.14 ±
1.8 g/dl vs. 13.76 ± 2.7 g/dl {P < 0.001) in patients with
normal platelet counts.
Cvtokine asssay results
Significant differences of serum levels of 11 of 13 cytokines (IL-lra, sIL-2R, IL-6, IL-8, IL-10, TNF RI, TNF
Table 2 Relationship between essential climcopathological features and results of standard hematological parameters in patients with soft tissue
sarcomas.
Clinicopathological
feature
Tumor size
< 10cm
> 10 cm
Tumor grade
Gl vs.
G2/3
Disease status
MO
Ml
HGB (g/dl)
PLT (x 109/l)
WBC (x 109/l)
NE (x 109/l)
MO (x 109/l)
LY (x 109/l)
EO (x lO'Vl)
Mean value ± standard devi ation /"-value
14.34 ±3.19
12.56±2.53
P < 0.025
256 5 ±78 4
332.96 ±143.8
P < 0.005
7 89 ± 5 41
9.57±4.54
P < 0.001
4.48 ±2 17
6.64 ±3.78
P < 0.001
0.58 ±0.22
0.76±0.55
P < 0.01
2 02 ±0.81
1.78 ±0.69
P < 0.025
0.27 ±0.32
0.29 ±0.78
NS
14.16 ± 1.39
13.15 ±3.04
NS
249.36 ±67.87
313 02 ± 130.76
P < 0 025
6.72 ±2.02
9.11 ±4.87
P < 0.05
4.23 ±1.85
5.95 + 3.36
P < 0.05
0.52±0 18
0.72 ±0.48
P < 0.025
1.76 + 0.55
1 89 ±0.79
NS
0.2 ±0.13
0.24 ±0.26
NS
13.28 ±2.27
13.32±4.5I
NS
287.6 ± 115.0
353 6 ± 145.9
P < 0.025
8.54±4.9
9.98±4.5
NS
5.28±2 9
7 35 ± 4 4
P < 0 005
0.68 ±0.48
0.7 ±0.29
NS
1.9 ±0.79
1 7±0.55
NS
0.31 ±0.7
0.2 ±0.17
NS
Abbreviation. NS - not significant.
1426
1.0
Group 1 - LY >1.0 G/L
Group II - LY <1 0 G/L
p=0.03
0.9
univariate analysis
0,8
o
o
o
^ ^ « — T x 1 ' ' <b- Group 1
0,6
o0.5
ev-
0,4
(b"i t i t — t en- •
+
Group II
+
+
0,3
200
400
600
800
1000
1200
1400
Days of study
Figure I Overall survival by absolute lymphocyte count.
Table 3. Mean cytokine serum concentrations [pg/ml] in healthy
subjects and soft tissue sarcoma patients groups.
Cytokine type
IL-lr
slL-2R
IL-6
IL-8
II-10
slL6R
TNFRI
TNFRII
TNFa
G-CSF
M-CSF
VEGF
bFGF
Cytokine mean
value in healthy
subjects group
Cytokine mean
value in STS
patients group
/"-value
585.34
2095 32
17 04
25.36
7 12
33956.34
1885.96
3755 93
<0001
< 0 001
< 0.001
< 0.001
< 0.001
NS
0.61
471
21 66
23.51
703.58
668.16
5.46
< 0.001
NS
< 0.001
< 0.001
< 0.001
261.92
1231.31
0 94
2 18
3.1 I
32056.5
879.95
2364.15
452.65
131 05
2.60
<0 001
<0 001
RII, TNFa, M-CSF, bFGF, VEGF) (P < 0.001) were
found in patients vs. controls (Table 3). We found no
significant changes in serum levels of sIL-6R and GCSF
in patients vs. healthy subjects. Increased serum levels
of IL-6 (69%),TNF RI (68%), IL-8 (60%) and TNF RII
(58%) were most frequent in the analyzed group of
patients.
Elevated serum levels of complexity of several cytokines and soluble cytokine receptors (particularly:
sIL-2R, IL-6, IL-8, MCSF, VEGF, TNFRI, TNFRII)
correlated strongly with the most common abnormalities in hematological tests: neutrophilia, leukocytosis,
monocytosis, thrombocytosis and anaemia (P < 0.001)
(Table 4, Figure 2). It was also found that lymphopenia
correlated significantly with increased serum levels of
IL-6, sIL-2R and TNFRI (Figure 3). No correlations
were found between the most common hematological
abnormalities and serum levels of TNFa and GCSF.
Survival analyses
At the time of this analysis, 56 patients from the studied
group had died of the disease (38.6%). Median overall
survival (OS) was 28.5 months. Average three-year OS,
as calculated by the Kaplan-Meier method, was 49%.
In a univariate log-rank analysis we revealed that the
five following factors correlated significantly with OS:
tumor size (< and > 10 cm), tumor grade (Gl vs. G2-3),
presence of metastatic disease (M0 vs. Ml), leukocytosis
(< and > 10 x 109/l), lymphopenia (< and > 1.0 x 109/l)
(Figure 1) (P < 0.05).
However, in a multivariate survival analysis only the
presence of metastases and tumor grade showed statistical significance.
Discussion
Oncologists often come across clinical conditions which
are accompanied by changes in the number of blood
cells. Secondary hematological abnormalities are a
common feature of malignant disease [1-3, 7, 39, 40].
Analysis of differences in whole blood cell composition
between cancer patients and healthy subjects may provide a better understanding of neoplastic disease. The
mechanisms of peripheral hematological changes in patients with various malignant tumors may be complex
and are still poorly understood. Hematopoietic and
proinflammatory cytokines as well as their free receptors and antagonists may be primarily involved in these
changes, since their role in normal hematopoiesis has
been well documented both in in vino and in vivo studies
[11]. Cytokines, considered crucial factors in hematopoiesis, may act as positive (e.g., IL-6, IL-11, GM-CSF)
or negative (e.g., the transforming growth factor TGFfS)
regulators of differentiation, maturation and growth
of the hematopoietic cells [41, 42]. In a wide sense
Table 4. Relationship between hematological abnormalities and cytokine serum levels.
Hematological
parameter
IL-lra
sIL-2R
IL-6
IL-8
IL-10
HGB
> ll.Og/dl
< 11.0 g/dl
P < 0.005
456 4
979 4
P < 0.001
1560 6
3835.5
P < 0.001
8.29
43.04
P < 0.001
20.02
42 29
P < 0 005
6 02
10.35
NS
34072 5
33209.4
P < 0.001
1439.2
3405.3
P < 0.001
3182 5
5653 4
PLT
< 4 5 0 x 109/l
>450 x 109/l
P < 0 005
472.4
1320.6
P < 0 001
1769.1
4321.1
P < 0.001
10.08
61.04
P < 0.005
22.31
45 46
P < 001
6 35
11.89
NS
33218.5
38800 0
P < 0.001
1658.3
3647.7
WBC
<IO.Ox 109/l
>IOOxlO 9 /
P < 0 001
433.44
1002.65
P < 0.001
1719.5
3216.5
P < 0.001
10.0
35 8
P < 0 001
21.57
36.25
P < 0 005
6.28
9.31
NS
33863.9
33926 1
NE
< 6 . 6 x 1O9/I
> 6 . 6 x 1O9/I
P < 0 005
431.19
894.98
P < 0 001
1702.95
2976.95
P < 0.001
9.67
32.73
P < 0.001
21.0
33 32
P < 0 025
6.38
9.33
MO
< 0 . 9 x 109/l
> 0 . 9 x 109/l
P < 0 025
534.19
681.03
P < 0 025
1770.85
3042.38
P < 0.001
11.66
33.0
P < 0.005
22.33
32.2
LY
>1.0xl09/l
< l . 0 x 109/l
NS
581.21
416.07
P < 0.05
1980.4
2778.6
P < 0.05
13 78
39 51
EO
< 0 . 5 x 109/l
> 0 . 5 x 109/l
P < 0.025
552.29
751 75
P < 0.005
1896.36
4131.09
P < 0.01
14.46
42.65
Abbreviation. NS - not significant.
G-CSF
M-CSF
bFGF
VEGF
NS
4 59
5.27
NS
19.84
34.04
P < 0.001
545 8
1170.8
P < 0.005
4.69
7.86
P < 0.001
351.4
1660.4
P < 0.001
3406 9
6373.5
NS
4.69
50
NS
20.21
43.62
P < 0.001
611 87
1280.75
P < 0.025
4.79
9 75
P < 0.001
365 19
2558.6
P < 0.001
1566.5
2840.6
P < 0 001
3448.54
4630.67
NS
4.74
4 59
NS
19.55
33.73
P < 0.001
612.07
954.31
P < 0.001
4.3
861
P < 0.001
483.97
1197.66
NS
33194.3
35164.0
P < 0.001
1566.3
2719.3
P < 0.001
3442 7
4326.4
NS
4.74
4.67
NS
19.67
28.13
P < 0.001
614 39
934.87
P < 0.005
4.45
7.1
P < 0.001
494.98
973.13
NS
7.08
7.91
NS
33753 1
34342 5
P < 0.001
1553.53
3090.59
P < 0 05
3356 75
4829 12
NS
4 74
4.86
NS
22.71
18.48
P < 0.001
625.33
988 97
P < 0.05
4.58
7.07
P < 0.001
473.17
1225.38
NS
23.39
33.76
NS
7.17
7.49
NS
34490.3
30132.0
P < 0 05
1856.6
2208.0
P = 0.05
3578.1
5248.0
NS
4 74
44
NS
22.63
16.52
NS
689.28
846.93
NS
5.19
4.92
NS
578.59
1110.69
NS
23.19
41.82
NS
7.1
8.76
NS
34050 7
32392.0
P < 0.005
1732.7
3905.7
P < 0.01
3464 5
6703.6
NS
4.68
5.1
NS
22.36
20 81
P < 0.005
672 2
1132.75
NS
5.26
4.36
NS
642 86
609 73
sIL-6R
TNFRI
TNFRII
TNFa
Statistical significance (/'-value), mean value (pg/ml)
1428
24 C
.SO
-2C
-
•4C
cz
0
PLT- 4!>0 i*LT .
13900
COCO
1O0&O
soco
eooo
• oco
tooo
JOCQ
L l
20 00
. 1 —1—
.OCO
4300
"I
3900
ri
0
n
PL1V4 50 Pl,T>4 5
Figure 2. Relationships between platelet count and cytokine serum levels in patients with soft tissue sarcomas.
TNF RI
a uu ana
1OOOO
8000
eooo
4000
2000
_rh
•
0
LY>1.0
LY<1.0
Figure 3 Relationships between lymphocyte count and selected cytokine serum levels.
hematopoietic cytokines include not only typical growth
factors (e.g., erythropoietin, G-CSF, M-CSF), but also
various cytokines (e.g., IL-6,TNFcc, macrophage inflammatory protein MIP-lot) [43, 44], which were identified
by other activities not connected with hematopoiesis. The
function of hematopoietic growth factors is not only
limited to cell lines. These multiple cytokines involved
in hematopoiesis create an extremely complicated cytokine network and demonstrate multifunctional, overlapping activity [11, 44]. Important biological effects and
immunological implications of the particular cytokines
can be reflected in routine blood test abnormalities.
1429
The synthesis and secretion of cytokines has been well
documented in many human malignant solid tumors
and several reports have demonstrated their enhanced
levels in sera of patients with various malignancies [14,
18, 19, 24, 29, 37, 45]. However, only a few reports have
addressed this problem in soft tissue sarcomas [2, 3032]. To the best of our knowledge, our present report
is the first attempt to look for correlations between
various hematological abnormalities with serum levels
of a larger variety of cytokines and free cytokine receptors in a large series of patients with soft tissue sarcomas. Special interest was also paid to the possible prognostic value of routine blood tests in the soft tissue
malignant tumor group.
It is important to emphasize that many statistical
computations could distort the results of the study. In
order to improve the reliability of statistical analyses
only significant differences for both applied tests were
considered significant for comparisons between the two
groups.
We revealed the different serum levels of several
cytokines and cytokine receptors in healthy and sarcoma patients, thus supporting the thesis about their role
in the pathogenesis of sarcomas. Moreover, cytokine
levels and disease stage were strongly dependent (data
to be published elsewhere). We have found that the most
common hematological abnormalities seen in our patients, that is low hemoglobin level as well as high white
blood cell, neutrophile, monocyte and thrombocyte
counts, correlated very strongly with elevated serum
levels of several cytokines and cytokine receptors, in
particular of IL-6, IL-8, M-CSF, VEGF, sIL-2R, TNF
RI and TNF RII. We failed to identify any single cytokine responsible for most of the changes in whole blood
cell counts. It is worthwhile to stress that G-CSF levels,
which can physiologically increase the number and activity of neutrophils, had no relationship with neutrophil
count. This observation remains consistent with data of
other authors [46], who demonstrated that the neutrophils to lymphocytes (N/L) ratio was related to sIL-2R
levels, but not to G-CSF levels. Thrombocytogenesis
is regulated by multiple humoral factors [47, 48], but
IL-6 is thought to the most potent stimulator of platelet
formation, also in cancer patients [25, 28, 49, 50]. However, it has been proved that other factors, e.g., M-CSF
and erythropoietin, play a role in the stimulation of
thrombopoiesis. TNFoc is known to depress erythropoiesis anemia, and response of the host in the form of
increased erythropoietin production may have influence
on the increased platelet count [51]. All these cytokines,
with the notable exception of TNFoc, were found to be
related to thrombocytosis in our study.
Lymphocytopenia, which was seen less frequently in
soft tissue sarcoma patients, correlated with enhanced
IL-6, sIl-2R and TNF RI serum levels. Blay et al. found
a relationship between lymphopenia and elevated plasma levels of IL-6 and VEGF [52]. We also observed that
increased levels of IL-6 - multifunctional and pleiotropic cytokine - were related to decreased lymphocyte
count and noted a similar relationship for sIL-2R and
TNF RI serum levels. It was also previously suggested
that sIL-2R may suppress the function of T lymphocytes
[46].
In general, our data suggest that the imbalance involving several cytokines and cytokine receptors, frequently
occurring in cancer patients, may be one of the major
mechanisms of hematological abnormalities developing
in the course of tumor growth and progression. Since, in
most patients with soft tissue sarcomas the serum levels
of proinflammatory, hematopoietic and proangiogenic
cytokines are elevated in parallel, it is very likely that
they may synergize in influencing blood cell counts.
There is ample evidence that several cytokines of various
primary defined functions are in fact pleiotropic and are
frequently released sequentially in response to appropriate stimuli.
In our study we observed different hematological
abnormalities in about 40 % of patients with soft-tissue
sarcomas, without evidence of a concomitant primary
myeloproliferative disorder or any other condition (e.g.,
infection), except the malignant disease, to explain these
disturbances. Patients with more advanced disease
showed a tendency toward more frequent occurrence of
hematological anomalies. These differences reached
statistical significance. The most common anomaly was
an increased number of blood leukocytes, mainly neutrophils. It was noted by Satomi et al. [46] that in advanced
colorectal cancer the neutrophils to lymphocytes ratio
(N/L) was significantly increased. Our findings have
also shown that WBC and NE increased with malignant
soft tissue sarcoma stage, while the number of lymphocytes was diminished. Neutrophilia in disseminated cancer patients is commonly connected with decreased
marrow granulocyte reserve [4], It was also reported
that neutrophils might promote cancer growth and progression [8]. Leukocytosis was a univariate unfavorable
prognostic factor, as earlier reports demonstrated in
some other neoplasms [7, 40]. However, in multivariate
analyses this variable lost its significance. Similarly,
other routine blood tests used seem to be of no significant value in predicting the survival in soft tissue sarcoma patients. In multivariate analyses only well-established prognostic factors (tumor grade and presence of
lung metastases) proved their significance for predicting
the overall survival in soft tissue sarcoma patients.
Apart from the leukocytosis and neutrophilia, a few
other hematological changes were recorded in the
present study. Among these decreased hemoglobin level
and thrombocytosis are also rather frequently recorded
hematological disturbances [1, 8, 28, 39, 49, 53, 54] in
cancer patients. An inverse relationship between the
hemoglobin level and the platelet count has been found
in sarcoma patients, and these abnormalities appeared to
occur more frequently in patients with more advanced
sarcomas. It has been demonstrated that increased platelet count may indicate unfavorable prognosis [9, 55, 56].
A very valuable study was presented by van der Zee et al.
[28], who demonstrated that high pretreatment platelet
1430
WBC
•>
Hadlan
•00
700
SOD
400
a
300
|
300
100
1
o
1 '
_L
H
Figure 4 Relationship between tumor size and hemalological tests in soft tissue sarcomas.
count was a negative prognostic factor in patients with
ovarian cancer. However, there are also contrary reports
about the prognostic role of thrombocytosis in cancer
patients [57, 58]. As mentioned previously, we also failed
to find thrombocytosis as a reliable prognostic factor in
the presented study.
A significant increase in monocyte count was found
in our group of patients. Some authors recorded a
relationship between monocyte count elevation and
shorter survival in malignancy [40]. An increase in
eosinophil count was a very rare finding in our study we detected it in only about 7% of patients. It has been
suggested that disseminated cancer (independent from
the histologic type of the tumor) may be accompanied
by marked blood eosinophilia [59]. Eosinophilia is considered as an ominous sign reflecting malignancy aggressiveness. We found no support for these suggestions
in our data.
Special interest was directed to lymphocyte count.
Significantly lower values of lymphocytes were detected
in 10% of patients and were related to increased tumor
burden. This has also been seen in cancers of various
origin [60]. It was suggested that the low lymphocyte
count can indicate immunodeficiency status in cancer
patients. Increased numbers of circulating CD8+ (suppressor) T cells are a common feature in cancer patients
[61]. Blay et al. [62, 63] demonstrated the value of
lymphopenia as a predictor of the toxicity of chemotherapy. The same authors have recently reported that
pretreatment lymphocytopenia (defined as LY < 1.0 x
10y/l) [52] is related to more aggressive behavior of
tumor and may be an independent unfavorable prognostic factor for survival in soft-tissue sarcomas. In our
series we observed that lymphocyte count < l . 0 x 1O9/1
correlated with poorer survival but this relationship was
significant only in univariate analysis (Figure 1). This
discrepancy may be due in part to the higher incidence
of advanced stages of the disease in the series of Blay
et al. [52]. Lymphocytopenia - as a marker of cancer
progression - may only be used as an additional, complementary factor which helps to distinguish low and
high-risk soft tissue sarcoma patients.
In conclusion, our results demonstrate that significant alterations in whole blood cell counts and leukocyte
differentials are rather commonly found in soft tissue
sarcoma patients, particularly in more advanced stages.
However commonly recorded hematological analyses,
when compared to well-known prognostic factors, have
a limited predictive value in soft tissue sarcoma patients.
Strong correlations between the presence of several
hematological abnormalities and serum levels of several
cytokines, suggest that the former may develop as a
result of cytokine misbalance frequently occurring in
soft tissue sarcoma patients.
Acknowledgements
Preliminary results of this study were presented as a
proffered paper during the sixth Annual Scientific Meeting of Connective Tissue Oncology Society, Amsterdam,
2-4 November 2000. The study was supported by the
Polish State Committee for Scientific Research grant
No. 4P05C010 97C/3337.
1431
References
1. Banerjee RN, Narang RM Haematological changes in malignancy. Br J Haematol 1967; 13: 829-43.
2. Hocking W, Goodman J, Golde D. Granulocytosis associated
with tumor cell production of colony-stimulating activity. Blood
1983; 61: 600-3.
3. Johnson RA, Roodman GD. Hematologic manifestation of malignancy. Dis Mo 1989; 35, 11: 721-68
4. Liu Y, Griamaldi M, Dannaher C. Myelopoietic abnormalities in
patients with metastatic carcinoma. Cancer 1978; 411 1193-8
5. Norcross JW. Haematologic manifestations of malignant disease.
Med Clin North Am 1963; 47. 345-8
6. Speights VO, Johgnson MW, Stoltenberg P et al. Complete blood
indices in colorectal carcinoma. Arch Pathol Lab Med 1992, 116.
258-60.
7. Hannisdal E, Engan T. Blood analyses and survival in symptomand survey-detected lung cancer patients. J Intern Med 1991; 229.
4:337-41.
8. Yeu-Tsu NL, van Peenen HJ, Watson F. Chemical and hematological screening in patients with malignant and non-malignant
conditions. J Surg Oncol 1980; 15: 43-51.
9. Hernandez E, Heller PB, Whitney C et al. Thrombocytosis in
surgically treated stage IB squamous cell cervical carcinoma (a
Gynecologic Oncology Group study). Gynecol Oncol 1994; 55:
328-32.
10. Stock W, Hoffman R. White blood cells I: Non-malignant disorders. Lancet 2000; 355: 1351-7
11. Sheridan WP, Hunt P, Simonet S, Ulnch TR. Hematological
effects of cytokines. In Remick DG and Friedland J (eds). Cytokines in Health and Disease, Second edition, revised and expanded. New York/Basel/Hong Kong- Marcel Dekker 1997:
487-505.
12. Kelso A. Cytokines: Principles and prospects. Imm Cell Biol
1998; 76. 300-17.
13. Baraldi-Junkins CA, Beck AC. Rothstein G. Hematopoiesis and
cytokines. Relevance to cancer and aging. Hematol Oncol Clin
North Am 2000; 14, 1:45-61.
14. Patel PM. Collins MK. Cytokine modulation of cell growth and
role in tumour therapy. Eur J Cancer 1992, 28, 1: 298-302
15 ICurzrock R Cytokine deregulation in hematological malignancies: Clinical and biological implications Clin Cancer Res 1997;
3- 2581-4.
16. Panozzo MP, Basso D, De Paoli M et al. Cytokines may influence
tumor growth and spread. An in vitro study in two human cancer
cell lines lnt J Clin Lab Res 1996, 26. 240-4.
17. Plante M, Rubin S,Wong G et al. lnterleukin-6 level in serum and
ascites as a prognostic factor in patients with epithelial ovarian
cancer. Cancer 1994; 73: 1882-8
18. Negus RP, Balkwill FR. Cytokines in tumour growth, migration
and metastasis. World J Urol 1996, 14. 157-65.
19. Raziuddin S, Sheikha A, Abu-Eshy S, Al.-Janadi M. Circulating
levels of cytokines and soluble cytokine receptors in various T-cell
malignancies Cancer 1994: 73: 2426-31.
20. Kawashima O, Kamiyoshihara M, Sakata S et al. The clinicopathological significance of preoperative serum-soluble mterleukin-2 receptor concentrations in operable non-small-cell lung
cancer patients. Ann Surg Oncol 2000; 7. 239-45.
21. Belluco C, Nitti D, Frantz M et al Interleukin-6 blood level is
associated with circulating carcinoembryonic antigen and prognosis in patients with colorectal cancer Ann Surg Oncol 2000, 7.
133-8.
22. Kuraa H, Hecr K, Lee P et al Preoperative serum vascular
endothehal growth factor can predict stage in colorectal cancer.
Clin Cancer Res 1998; 4: 1279^85.
23. Belluco C, Nitti D, Frantz M et al Interleukin-6 blood level is
associated with circulating carcinoembryonic antigen and
prognosis in patients with colorectal cancer. Ann Surg Oncol
2000; 7- 133-8
24. Katsumata N, Eguchi K. Fukuda M et al Serum levels of
25.
26.
27
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
cytokines in patients with untreated primary lung cancer. Clinical
Cancer Res 1996: 2: 553-9
NagasawaT. Orita T. Matsuhita J et al Thrombopoietic activity
of human interleukin 6. FEBS Lett 1990; 260: 176-8.
Blay JY. Favrot M. Rossi JF. Wijdenes J. Role of intcrleukin-6 in
paraneoplastic thrombocytosis. Blood 1993; 82- 2261-2.
Blay JY, Rossi JP. Wijdenes J et al. Role of interleukin-6 in the
paraneoplastic inflammatory syndrome associated with renal-cell
cercinoma. lnt J Cancer 1997. 72: 424-30.
Van der Zee A. Cuyper E, Limburg P et al. Higher levels of
interleukin-6 in cystic fluids from patients with malignant
versus benign ovarian tumors correlate with decreased hemoglobin levels and increased platelet counts. Cancer 1995; 75:
1004-9.
Pekarek L. Weichselbaum, Beckett M et al. Footprinting of
individual tumors and their variants by constitutive cytokine
expression patterns. Cancer Res 1993; 53: 1978-81
Spellman JE. Gollnick SO. Zhang PJ. Tomasi TB. Cytokine
production by human soft tissue sarcomas: Implications for
immunosupression within the tumor bed. Surg Oncol 1996: 5:
237-44.
Verheul H, Hoekman K. Lupu F. Platelet and coagulation activation with vascular endothehal growth factor generation in soft
tissue sarcomas. Clin Cancer Res 2000; 6: 166-71.
Graeven U. Andre N. Achilles E et al Serum levels of vascular
endothehal growth factor and basic fibroblast growth factor in
patients with soft tissue sarcoma. J Cancer Res Clin Oncol 1999:
125: 577-81.
Kaminska J. Kowalska M. Rysinska A. Serum levels of cytokine
in patients with malignant neoplasms (in Polish). Nowotwory
1999; Suppl 1:21-5
Arai K. Lee F. Miyajima A. Miyatake S et al. Cytokines: Coordinators of inflammatory responses. Ann Rev lnc 1990; 59: 783836.
Fuhrmann-Benzakein E. Ma M. Rubbia-Brandt L et al. Elevated
levels of angiogenic cytokines in the plasma of cancer patients, lnt
J Cancer 2000: 85- 40-5
Aderka D. Engelmann H. Hornik Vet al Increased serum levels
of soluble receptors for tumor necrosis factor in cancer patients.
Cancer Res 1991:51 5602-7
De Vila F. Orditura M. Galizia G et al. Serum inlerleukin-10
levels as a prognostic factor in advanced non-small cell lung
cancer patients Chest 2000: 117: 365-73.
Gustafson P. Soft tissue sarcoma. Epidemiology and prognosis in
508 patients. Acta Orthop Scand 1994: Suppl 259.65: I -31.
Den Ouden M. Ubachs JM. Sloot JE. van Wersch JW. Whole
blood cell counts and leukocyte differentials in patients with
benign or malignant ovarian tumours. Eur J Obsl Gyn Rep Biol
1997: 72: 73-7
Bruckner HW, Lavin PT. Plaxe C et al Absolute granulocyle.
lymphocyte and monocyle counts JAMA 1982.247. 1004-6
Han ZC, Caen JP. Cytokines acting on committed haematopoietic progenitors. Bailheres Clin Hematol 1994: 71 65-89
Li WM, Huang WQ. Huang YH et al. Positive and negative
hematopoietic cytokines produced by bone marrow cndothelial
cells. Cytokine 2000; 12. 1017-23.
Broxmeyer H. Kim Ch. Regulation of hematopiesis m a sea of
chemokine family members with a plethora of redundant activities. Exp Haematol 1999: 27: 1113-23.
Metcalf D Hematopoietic regulators Redundancy or subtlety.
Blood 1993: 82: 3515-23.
Whatson JM. Sensintaffar JL. Berek JS. Martinez-Maza O. Constitutive production of interleukin-6 by ovarian cancer cell lines
and by primary ovarian tumor cultures Cancer Res 1990: 50:
6959-65
Satomi A. Murakami S. Ishida K. Significance of increased
neutrophils in patients with advanced colorectal cancer. Acta
Oncol 1995: 34: 69-73.
Williams N. Jackson H. Walker F. Oon SH. Multiple levels of
regulation of megakaryocytopoiesis. Blood Cells 1989: 15: 123 33.
1432
48. George J. Platelets. Lancet 2000; 355: 1531-9.
49. Nakano T. Chahinian AP. Shinjo M et al. lnterleukin 6 and its
relationship to clinical parameters in patients with malignant
pleural mesothehoma. Br J Cancer 1998; 77: 907-12.
50. Ishibashi T, Kimuru H. ShikamaTet al. lnterleukin-6 is a potent
thrombopoietic factor in vivo in mice. Blood 1989, 74: 1241-4.
51. Colibon U, Johnson R, Roodman GD. Erythropoiclin (rhEp)
maintains platelet (Pit) counts but fails to reverse the anemia in
mice exposed to continuous tumor necrosis factor (TNF) in viw.
Blood 1988;72(Suppl 1): 112a.
52 Blay JY. Ray-Coquard I. van Glabbeke M. Lymphopenia (LYP)
as an independent prognostic factor for survival in advanced soft
tissue sarcomas as well as in lymphomas. and metaslatic breast
and renal carcinoma Connective Tissue Oncology Society 6th
Annual Scientific Meeting 2000.
53. Chalas E, Wclshinger M, Englellener W. The clinical significance
of thrombocytosis in women presenting with a pelvic mass Am J
Obstet Gynecol 1992; 166: 974-7.
54 Constantini V. Zacharski L. Moritz T. Edwards R. The platelet
count in carcinoma of the lung and colon. Thromb Haemost 1990,
64.501 5.
55 Monreal M. Fernandez-Llamazares J. Pinol M et al Platelet
count and survival in patients with colorectal cancer - a preliminary study Thromb Haemost 1998; 79: 916-8.
56. Pedersen LM, Milnian N. The prognostic value of thrombocytosis in patients with primary lung cancer. Ugeskr Laeger 1998: 160:
3917-20.
57. Lopes A, Daras V. Cross P et al. Thrombocytosis as a prognostic
factor in women with cervical cancer. Cancer 1994: 74: 90-2.
58. Zeimct AG, Marth C, Muller-Holzner E et al. Significance of
59.
60.
61.
62
63.
thrombocytosis in patients with epithelial ovarian cancer. Am J
Obstet Gynecol 1994; 170: 549-54.
Slefamni M. Claustro J. Motos R. Bendigo L. Blood and bone
marrow eosinophilia in malignant tumors. Cancer 1991: 68: 543-8.
Zacharski LR, Linman JW. Lymphocytopenia: Its causes and
significance Mayo Clin Proc 1971; 46: 168-73.
Miyazaki K. Shimada K. Katabuchi H et al. Activated (HLADR+) T-lymphocyte subsets in early epithelial ovarian cancer and
malignant ovarian germ cell tumors. Gynecol Oncol 1995. 58:
362-7
Blay JY, Chauvin F, Le Cesne A et al. Early lymphopenia after
cytotoxic chemotherapy as a risk factor for febrile neutropenia. J
Clin Oncol 1996; 14, 2: 636-43.
Blay JY, Le Cesne A. Mermet C et al. A risk model for thrombocytopenia requiring platelet transfusion after cytotoxic chemotherapy. Blood 1998: 92: 2: 405-10
Received 15 January 2001, accepted 9 May 2001.
Correspondence to
P. Rutkowski, MD
M. Sklodowska-Curie Memorial Cancer Center and
Institute of Oncology
Department of Soft Tissue/Bone Sarcomas
W. Roentgena Str. 5
02-781 Warsaw
Poland
E-mail: [email protected]