(CANCER RESEARCH 48. 2596-2603, May I. 1988)
Modulation of Natural Killer Cell Activity by Serum from Cancer Patients:
Preliminary Results of a Study of Patients with Colorectal Adenocarcinoma
or Other Types of Cancer1
Madalina Pislarasu, A. Oproiu, Doina Taranu, R. B. Herberman,2 and A. Sulica
Department of Immunology, Babes Institute [M. P., A. S.J, and Clinic of Gastroenterologe, Fundeni Hospital [A. O., D. T.J, Bucharest, Romania, and Pittsburgh Cancer
Institute and Departments of Medicine and Pathology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania 15213-2592 [R. B. H.J
INTRODUCTION
ABSTRACT
NK3 cells are a class of nonadherent, nonphagocytic, rapidly
As previously reported for natural killer (NK) cells of normal individ
uals, prior incubation of peripheral blood lymphocytes from cancer
patients with human normal serum or monomeric immunoglobulin G
reduced their subsequent capacity to kill K562 target cells in a 4-h 5lCr
the putative stimulatory or protective factor may contribute to more
efficient function of their circulating NK cells.
cytotoxic LGL which can efficiently lyse a wide variety of tumor
cells, virally infected cells, and immature cell types of normal
origin (1, 2). Besides their cytolytic function, NK cells can
secrete cytokines (3, 4) and appear to participate in the control
of cell differentiation, in particular of hematopoietic precursors
and thymocytes (5, 6). Their contribution to antitumor resist
ance has been indicated mainly by the considerable evidence
regarding the ability of NK cells to eliminate metastatic cells
and thereby resist tumor spread (7-9). Nevertheless, knowledge
of the potential role of these cells in the control of human
neoplasia is largely limited to observations that somewhat de
pressed NK activity is usually found in the peripheral blood of
patients bearing advanced solid tumors (10,11). Such decreased
functional NK levels were not attributable to a decreased num
ber of circulating NK cells, and additional regulatory factors
appear to be involved in the determination of the cytotoxic
activity of the cells isolated from cancer patients.
The basis for the physiological regulation of NK activity is
almost unknown in spite of the large number of studies dealing
with the identification and characterization of humoral and
cellular factors acting on NK cells and modulating their killing
function as assessed mainly by in vitro experimentation (12).
Such soluble factors regulating positively or negatively the NK
cells are interferon (13), interleukin 2 (14), and prostaglandin
E2(15).
We have previously demonstrated (16, 17) that mlgG isolated
from human serum can act as an additional negative regulatory
factor of NK activity. Prior treatment of PBL with normal
autologous serum induced inhibition of NK activity which could
be reduced by selective depletion of IgG from the serum. More
recent data indicated that the mlgG-induced reduction of the
killing capacity of NK cells was caused by the inhibition of a
postbinding event, mediated at least partially by cyclic AMP
(18) and triggered by a direct interaction between mlgG and
LGL (Ref. 19; Footnote 4) through the Cy3 domain of IgG
and the Fc receptors of NK cells (20).
In order to evaluate the possible mechanisms underlying the
low NK reactivity in cancer patients, in the present study we
have analyzed the regulatory effects of untreated or heat-treated
serum isolated from tumor-bearing individuals on autologous
PBL and on cells harvested from healthy donors. The results
of this study indicate that either of two patterns of seruminduced regulation of NK activity is present in malignant dis-
Received 4/23/87; revised 1/7/88; accepted 1/20/88.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work was performed under the US-Romania Agreement on Cancer
Research.
2To whom requests for reprints should be addressed.
3 The abbreviations used are: NK, natural killer, LGL, large granular lympho
cytes; mlgG, monomeric IgG; PBL. peripheral blood lymphocytes; CEA, carcinoembryonic antigen; TNM, tumor-nodes-metastases; PCS, fetal calf serum;
CCM, complete culture medium; LU, lytic units.
4 A. Sulica, A. C. Bancu, C. Galatiuc, M. Manciulea, and R. B. Herberman.
Regulation of human natural cytotoxicity by IgG. IV. Association between direct
binding or monomeric IgG to the Fc receptors of large granular lymphocytes and
inhibition of the cytotoxic function of natural killer cells, manuscript in prepa
ration.
release assay. The NK activity of such treated effector cells was signifi
cantly inhibited only by 58% of sera from patients with colorectal
adenocarcinoma (21 of 36 cases) and by 67% of sera from patients with
other lymphoid or nonlymphoid solid tumors (22 of 33 cases). The
cytotoxic activity of cells previously incubated with eight noninhibitory
sera was even augmented relative to medium-treated peripheral blood
lymphocytes (control). The 26 untreated cancer sera which did not inhibit
significantly the NK activity (/ ) always developed significant inhibitory
capacity upon heating at 56°Cfor 30 min (A*). An additional seven (21%)
patients with colorectal carcinoma and four (27%) patients with other
cancers were identified as having type II NK regulation, defined as sera
with untreated inhibitory capacity (/*) but with appreciably more inhi
bition after heating (A*). The sera of the last group of patients with
colorectal adenocarcinoma (14 of 36 cases) defined as having type III
NK regulation were not different from control sera isolated from normal
individuals (/*A~) except that they induced an inhibition greater than
that caused by normal sera.
The modulatory characteristics of sera from the first two categories of
patients appear to be cancer associated, since the patterns /A* or /*A +
were observed with sera from only one of 30 patients with benign digestive
diseases and two of 100 apparently healthy individuals. Preliminary
results of longitudinal investigations of patients with colorectal adenocar
cinoma revealed also that these patterns disappeared several months
after resection of their tumor in all five tested patients, whereas the type
HI NK regulation found in patients with poor prognostic factors was
unchanged after surgery in the other five of six patients. The three
different categories of cancer sera identified by the functional assay of
NK regulation indicated differences among our group of patients which
were not paralleled by differences in levels of cytotoxic reactivity of their
NK cells assayed in vitro in the absence of autologous serum. The
abnormal behavior of sera from 37 (54%) patients with different types of
malignant disease suggests the presence of a thermolabile serum factor
which can either stimulate the activity of NK cells isolated from both
cancer patients and normal donors, or protect them against the serum
inhibitory factor which, in fact, are in higher amounts or have higher
binding affinities as compared to those in sera of healthy donors. Since
the type I of NK regulation appears to be correlated with both early stage
of malignancy and a better clinical course postoperative!}', the results of
this study suggest that the in vivoeffect of cancer patients' sera containing
2596
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.
MODULATION
OF NK CELL ACTIVITY BY SERUM FROM CANCER PATIENTS
ease. The modulatory characteristics of serum from 63% of
cases with colorectal adenocarcinoma appear to be cancer as
sociated, since this pattern has practically not been observed
with serum from normal individuals or patients with benign
digestive diseases and seems to disappear after tumor resection.
MATERIALS AND METHODS
Populations Studied. The patients studied consisted of a first group
of 38 adults with a variety of lymphoreticular or solid malignant tumors
(4 Hodgkin's or non-Hodgkin's lymphoma, 4 breast, 2 skin, 1 thyroid,
6 lung, 12 gastrointestinal, and 9 genitourinary carcinomas) and a
second group of 31 cases with colorectal carcinoma. The immunological
investigations (see below) were performed on hospitalized patients
usually as soon as their clinical diagnosis was established and always
before any antitumor therapy, except one patient with rectal cancer
who was irradiated 2 yr before. In the case of patients with colorectal
carcinoma, all diagnoses were histopathologically confirmed on biopsy
specimens obtained by colonoscopy. Patients in the second group had
routine physical and laboratory examinations, including scinti- and
echography and determination of serum CEA and ferritin. The Dukes
and TNM staging, available only in surgically treated cases, were
determined at certain periods of time after immunological investiga
tions (3 to 160 days). Before surgical treatment, 9 patients had under
gone radiotherapy. Some of the patients with colorectal cancer were
again clinically examined and immunologically tested at various times
following surgical resection of their tumor. Control subjects consisted
of 30 patients with benign digestive diseases (8 chronic hepatitis, 3
chronic pancreatitis, 3 Crohn's disease, 3 multiple colon polyposus, 9
rectocolitis, and 4 rectal polyposus). The 100 apparently healthy indi
viduals used as controls in our experiments were blood donors at the
Haematology Center, Bucharest, Romania.
Preparation of Lymphocytes. PBL were isolated from heparinized
blood by sedimentation in Sepcel (Comasim, Bucharest, Romania),
washed in K '-65 solution (Cantacuzino Institute, Bucharest, Romania),
and resuspend in RPMI-1640 medium (Flow Laboratory, McLean,
VA) supplemented with 10% PCS (Biofluids, Inc., Rockville, MD), 20
mM 4-(2-hydroxyethyl)-l-piperazineethanesulfonicacid
buffer, 1% glutamine, and antibiotics used as (CM. For washing the cells, medium
with only 2% FCS was used.
Treatment of Human Serum. On the day of the experiments, an
aliquot of the serum from patients with malignant or benign diseases
or from normal donors was inactivated at 56°Cfor 30 min. Aliquots of
some unheated sera from cancer patients were kept frozen at —
15°C
and used after several weeks or months when additional determinations
were performed with PBL from normal donors (see below).
Treatment of PBL. Lymphocytes (2 x 10' cells/ml) were incubated
in Petri dishes for 2 h at 37°Cin CCM, in the presence or absence of
20% untreated or heat-inactivated human serum. The medium-treated
cells were used as control. In most of the experiments, aliquots of cell
suspension were similarly exposed to 20% serum isolated from a normal
donor and, in some experiments, the cells were treated with 0.5 mg of
mlgG isolated, as previously reponed, from normal human sera (16).
These different types of serum were usually added also to PBL from
the individual used as the donor of normal serum, and therefore the
effects of autologous and homologous sera on both cell preparations
were compared under the same technical conditions.
Following incubation, the nonadherent lymphocytes were harvested,
washed twice in culture medium with 2% FCS, and finally resuspended
in CCM.
NK Assay. A 4-h 51Crrelease assay was performed in round-bottomed
microtiter plates as described (16). Four 3-fold dilutions of treated
lymphocytes (0.1 ml) were mixed with equal volumes of "Cr labeled
K562 targets (3 to 4 x IO3cells). LU at 30% specific "Cr release were
Serum-induced inhibition of NK activity was arbitrarily dotinoli as
levels at least 20% below the control levels. Differences in the range of
—20to +20% between the inhibition determined by each tested serum
before and after its treatment at 56°Cwere considered nonsignificant,
and therefore the heat-induced difference (A) of the regulatory capacity
of a certain serum was considered as significant when it was equal to
or higher than 20%. This value of ±20%was selected as the cut-off
since it is the limit of the mean ±2SD calculated for all control sera
used in our present study.
Statistical Calculation. The data were analyzed by using a paired t
test.
RESULTS
Investigation on Patients with Various Types of Cancer. Ad
dition of 20% normal serum or 0.5 mg of monomeric IgG to
PBL harvested from 19 patients with various types of malignant
disease caused in all experiments significant inhibition of NK
activity (at least 20%), whereas prior incubation of the PBL in
the presence of 20% patients' serum produced a significant
reduction of the NK activity in only 10 of the 19 cases (Table
1). In addition, it was noted that the cytotoxic activity of cells
incubated with serum of patients with either Hodgkin's disease
(2 cases), breast cancer (1 case), or colon cancer (2 cases) was
even higher than that of medium-treated PBL (control). The
inability of these cancer sera to inhibit NK activity of PBL was
seen when sera were tested at concentrations of 5 to 20% (data
not shown).
The effect of cancer patients' serum on NK activity, expressed
as the percentage of the control (PBL preincubated in medium
alone), of 19 additional cancer patients is summarized in Table
2. In addition, it was found that all 6 noninhibitory sera (i.e.,
O. V., M. A., D. N., M. S., V. E., M. C.) acquired the capability
to down-regulate the cytotoxic reactivity of their NK cells upon
heating at 56°Cfor 30 min. Such heat-induced differences (A)
between the percentage of NK activity of PBL preincubated
with untreated autologous serum and that obtained with heated
serum, defined as significant when higher than 20%, were
Table 1 NK activity of PBL isolated from cancer patients when incubated in vitro
with autologous or normal homologous serum or with human monomeric IgG
prior to the cytotoxic assay
PBL (2 x 10' cells/ml) from patients with various types of cancer were
incubated in Petri dishes for 2 h at 37°Cin the presence of 20% autologous or
normal serum or 0.5 mg of human monomeric IgG or medium alone (control).
The nonadherent cells were harvested, extensively washed with cold medium, and
used as effector cells in the NK assay.
NK activity (LU/107 cells) of PBL when
treated with
Patient's Normal Monomeric
calculated from the curves obtained at different effector/target cell
ratios. The value of LU/107 cells of nonadherent PBL previously
incubated at 37"C in medium alone was considered as the control for
NK activity of each cell preparation, and the value of different serumtreated effector cells was expressed as the percentage of the control.
2597
Patient
C.
D.M.
A.M.
CS.
Clinical diagnosis
Medium
diseaseHodgkin's
diseaseNon-Hodgkin'slymphomaBreast
serum
serum
IgG
cancerBreast
E.D.
M.C.I.S. cancerThyroid
cancerBronchial
carcinomaBronchial
N.V.
I.I231.5431.067.228.546.020.8116.2166.6250.031.2100.0280
l
R.M. carcinomaBronchial
D.C.
carcinomaBronchial
carcinomaLiver
I.Z.V.C.V.C.
cancerLiver
cancerAbdominal
R.P.
tumorColon
V.A.
carcinomaColon
carcinomaBladder
P.Z.
A.T.
cancerBladder
P.G.
cancerBladder
H.R.
cancerProstatic
M.Hodgkin's carcinoma100.0142.81818.2153.828.527.3333.3II
" ND, not determined.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.
MODULATION OF NIC CELL ACTIVITY BY SERUM FROM CANCER PATIENTS
Table 2 Effect of inactivation at 56'Cfor 30 min on the capacity of serum from
cancer patients to regulate the NK activity ofPBL isolated from either patients or
normal donors
NK activity (% of the control)"
of PBL incubated with patient
serum
Donor of
used in
NK
assayPatientNormalPatient
PatientM. locationNon-Hodgkin's
inac
tivated26.4
E.S.
42.81.6
lymphomaSkinSkinBreastBreastLungLungColonColonRectumPBL
75.010.0
M.0.
40.0157.0
NormalPatientNormalPatient
38.063.0
2.095.0
104.711.4
125.73.418.017.4
21.0-8.0
V.I.V.D.
NormalPatient
5.48.7
12.68.7
one inhibitory normal serum, treatment at 56°Cdidn't signifi
cantly modify the percentage of inhibition caused by normal
sera, since the value of the heat-induced difference was in the
range of -20 to +20%, even with 2 noninhibitory untreated
sera.
The behavior of some sera from patients with different types
of malignant disease suggested the presence of a thermolabile
serum factor which could either stimulate the activity of NK
effector cells isolated from both normal donors and cancer
patients or protect them against the serum inhibitory effects.
Thus, it seems that removal by heat inactivation of the stimu
lator or protective factor(s) which might be present in these
cancer patients, but not in normal individuals, reveals the
presence of the inhibitory factors in cancer sera. In fact, the
inhibitory factors in some cancer patients' sera were found to
be in even higher amounts or to have higher binding affinities
as compared to those in sera of healthy donors. Thus, the mean
19.2132.664.927.8109.0
10.072.540.641.64.0
9.260.124.3-13.8105.0
NormalPatientPatientPatientPatient
±SEM of the percentage of inhibition of NK activity in a total
A.M.I.I.E.D.M.M.S.Tumor
of 85 experiments performed with different normal or cancer
PBL, each of them being treated under the same experimental
conditions, was significantly greater (P < 0.001) after preincubation with heated serum from cancer patients (72.0 ±2.3) as
compared to the percentage of inhibition induced by normal
59.047.8 41.061.2 serum (53.4 ±3.0) whose regulatory effect of NK activity was
NormalPatient 100.0117.0
rather unchanged upon heating.
61.2
Investigations on Patients with Colorectal Adenocarcinoma. In
NormalUntreated57.7
129.0Heat
67.8A*31.332.28.4
our initial studies, unheated sera from 4 of 5 patients with
E.T.
V.
colon or rectal tumors had a noninhibitory effect on the NK
E.N.
activity of PBL (Tables 1 and 2). Therefore, it was of interest
53.071.4 -12.55.629.3-8.3
NormalPatient 40.577.0
to study a large group of untreated (by either surgery, radio-,
or chemotherapy) patients with colorectal adenocarcinoma in
A.S.
45.643.5
NormalPatient 75.035.2
different stages of disease. Some relevant clinical, histológica!,
and laboratory data on this group of 31 patients are shown in
M.O.I.M.
Table 3. The blood specimens for immunological investigations
NormalPatient 72.430.6
82.053.5 -7.6-22.0
were obtained as soon as the clinical diagnosis was established,
and therefore there are some patients in which the diagnosis of
NormalPatient 14.448.4
53.951.0 -19.5-2.5
maligancy was not followed for some reasons by surgical treat
N.M.
ment, and consequently the stage could not be determined by
-0.981.2
NormalPatient 5.6114.6
6.553.5
the criteria shown in Table 3.
C.P.E.P.
The pattern observed in 4 of 5 colorectal cancer patients and
NormalPatient 38.575.0
10.073.5 28.51.5
in 46% of all cancer sera tested in our initial studies (Tables 1
and 2), namely, the failure of unheated serum to inhibit NK
NormalPatient113.062.8
91.061.076.051.0
86.019.337.011.8
5.041.7
activity and the detection of inhibitory capacity after heat
M.AbdomenAbdomenLiverLiverKidneyBladderBladderBladderOvaryPatientPatient treatment, was looked for in the sera of the patients listed in
Table 3. Untreated and heat-treated sera at a final concentration
" In each experiment, the NK activity calculated as LU/107 cells for mediumtreated PBL (control) obtained from either patient or normal donor was consid
of 20% were usually incubated in the same or in separate
ered 100%.
experiments with PBL from both patient and normal donors,
* A, differences between the percentage of inhibition of NK activity of PBL by
and as above, the NK activities were determined and expressed
untreated serum and that by heat-treated serum.
as the percentage of medium-treated cells.
Table 4 shows that the sera of 11 of the patients (36%) had
observed not only in these 6 cases but also with 3 additional
sera in which the inhibitory effects by untreated serum were the following characteristics, which have been defined as type I
NK regulation: (a) failure of untreated serum to reduce signifi
augmented by heat treatment. In 14 experiments, PBL isolated
from healthy donors were used in the NK assays in parallel to cantly the NK activity of at least one effector cell population;
(b) inhibitory effects on NK activity after heating at 56°C,with
cells obtained from cancer patients. In all but two of these
parallel experiments, the cancer sera tested as unheated or heat-induced differences generally higher than 50%. In addi
tion, the majority of treated sera from this first group of patients
heated behaved in a similar fashion on both patient and normal
expressed an inhibitory capacity equal to or greater than that
PBL.
of a normal serum added to the same effector cell population,
In contrast to the effects of heating on the inhibitory activity
either before or after its heat treatment. These results suggested
of sera from cancer patients, results of experiments performed
with 31 sera obtained from normal donors and tested on au- that the untreated type I serum contained either a heat-labile
factor interfering with NK inhibitory activity or a rather high
tologous PBL and, additionally in 9 experiments, on PBL
amount of an NK stimulatory factor able to counteract or mask
isolated from cancer patients, indicated that the NK activity
was consistently inhibited by all but three of these untreated
almost completely the effect of inhibitory factors.
A second group of 7 patients (23%) was identified as having
normal sera. Except for one noninhibitory normal serum and
2598
F.M.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.
MODULATION OF NK CELL ACTIVITY BY SERUM FROM CANCER PATIENTS
Table 3 Clinical, histological, and laboratory data regarding the group of patients with colorectal adenocarcinoma
stagingAnatomical
dataTypeTubularColloidalMixedTubularTubularTubularMixedTubularTub
locationColonColonColonRectumRectumColonColonRectumColonRectumColonRectumColonRectumRectumColonColonRectumRectumColonColonColonRectum
siteSigmoidSigmoidConcomitant
Case12345678910111213141516171819202122232425262728293031Age76685151696256505530596666515240584630327660627071416438547375SexMMFMFMMMMFMMMMFFMFM
(ng/ml)7.460.024.04.8>
hepaticAngle
(T,N„M„)la
descendingRectumRectumDescendingDescendingRectumRectosigmoidRectumAscending
+
(T.NoMo)II
(T,N0M„)II
(TjN„M„)II
(TjNoMo)II
(TjNoMo)III
(TjN.^Mo)IV
(T.N.M,)II
100.056.0>
100.04.7>
transverseRectumSigmoidRectumRectumConcomitant
+
(T,N„M„)Ib
(TjNoMo)III
(TjN.L.M,,)II
(T3N„M„)IV
100.027.0>
100.016.076.06.43.962.0>
descending+
sigmoidRectosigmoidRectumRectumSigmoidSigmoidRectosigmoidRectumDescendingRectumRectumRectumRectumSigmoidCecum
(T.N.M,)IV
(TjN.M,)II
(TjNoMo)IV
(T2N0M,)II
(T,N„M„)II
(T,NoM„)III
(T3N„;)M„)II
(TjNoMo)II
(TjNoMo)II
(T,N„M„)III
(TjN^.Mo)IV
ascendingSigmoidDukesDB,AB,B,BjB,C,DB,B,C,B,DDB,DB,B,C,B,B,B2C,DDTNMIV(T3N„M,)II
+
(T.N.M,)IV
(T.N.M,)Histological
' PD, poorly differentiated; WD, well differentiated; MD, medium differentiated; ND, nondifferentiated.
100.0>
100.086.025.0<2.537.025.056.
100.042.0
type II NK regulation, defined as sera with untreated inhibitory
category of sera was only changed in 2 cases (Patients 17 and
activity but with appreciably more inhibition (>20% difference)
18) whose first determined A values were, in fact, near the
after heating, exhibited with at least one of the PBL populations
borderline for categorization.
used as effector cells. This pattern suggested that these sera
The type I or II patterns of NK regulations in patients with
probably possess the stimulatory or protective factor in a lower colorectal adenocarcinoma (Table 4) were found in only one of
amount than that found in the sera with type I NK regulation.
30 cases with inflammatory digestive diseases or with benign
The sera of a third group of 13 patients, defined as having colorectal polyposus. The mean ±SEM of NK activity (per
centage of the control) of patients' PBL treated with their
type III NK regulation, had the following pattern: untreated
serum inhibited strongly the NK activity; and this effect was autologous serum (32.1 ±5.0) was very similar to that obtained
not significantly changed by heat treatment (A < 20%). Almost
upon exposure to normal serum (32.2 ±4.6) and, except for
one case, heating of these patients' sera didn't induce significant
all of these patients had poor prognostic factors, such as distant
métastasesor tumor invasion into neighbouring organs, young elevation of their inhibitory capacity.
age, or a personal (previous noncolonic cancer) or familial
Table 6 shows a summary of the results of our studies on 36
patients with colorectal adenocarcinoma (presented in Tables
history of cancer.
Tests of the majority of cancer sera on both autologous PBL
1, 2, and 4) together with the data of 30 patients with benign
and cells from a normal donor revealed a good correlation
digestive disease and those of 100 normal individuals. As com
between the data provided with each effector cell population in pared to the results obtained with PBL isolated from healthy
regard to the percentage of the control value of NK activity (r donors when incubated with untreated or heat-treated autolo
= 0.62) and A value (r = 0.78). Since the particular type of gous serum, sera from the first group of cancer patients (42%
reactivity of each serum could be defined irrespective of the of the total cases) were highly significantly (P < 0.001 ) different
donor of cells used in the NK assay, the reproducibility of the from normal sera with regard both to their inability to inhibit
modulatory effects of sera from 18 cancer patients on the NK NK activity and their high A values, whereas the sera from the
7 type II cancer patients (19% of cases) were significantly (P <
activity of normal PBL was tested several times after the initial
determination. Although for about 25% of the investigated sera 0.001) different only in terms of the heat-induced difference.
there were appreciable differences in their ability to modulate
The sera of cancer patients of the third group (59% of cases)
at different times the NK activity of 2 to 4 different PBL, the were not different from control sera isolated from healthy
overall interassay variability was not significant, as the mean ± individuals with regard to A value, but a significantly higher (P
SEM of NK activity (percentage of the control) of the first < 0.005) mean ±SEM inhibition (67.0 ±4.1) of NK activity
investigation (68.9 ±10.5) was similar to that obtained when of normal and cancer PBL was found in 24 determinations in
the serum-induced regulation of NK activity was tested the comparison to that induced by normal sera (45.8 ±5.4). No
second (64.7 ±11.8) or the third time (80.9 ±35.0). Addition
differences were also observed between sera of patients with
ally, Table 5 shows the reproducibility of the A values of 12 benign digestive diseases and those isolated from apparently
cancer sera tested 2 to 3 times in separate assays on normal
healthy donors. However, of all of these sera, there were 1 of
PBL. Although there was a certain degree of variation, the the patients with benign diseases (3% of investigated cases) and
2599
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.
MODULATION OF NK CELL ACTIVITY BY SERUM FROM CANCER PATIENTS
Table 4 Distinct types of NK regulation by serum from patients with colorectal carcinoma
withCase12345678910111213141516171819202122232425262728293031Tumor
NK activity of PBL incubated
observationsDukesNDDB/ABBBBCDNDBBCBNDNDDNDD
serum%
of
serum (% of
of NK
cells used in
ofcontrolND°90.3100.061.0144.0100.090.0108.089.050.0100.0144.472.0116.0100.093.3101.0113.5145.0ND67.0128.023.068.554.387.74
control)ND28.048.6ND25.746.330.046.340.046.338.065.034.065.029.421.0129.1105.760.0ND73.060.
regulationIIIIIIIIIIIIInnnnnnHIininminHIininilimHImHIClinical
locationColonColonColonRectumColonColonColonRectumColonRectumColonRectumColonRectumRectumColonColonRectumRectumColonColonRectumRectumColonRectum
N
assayPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientN
K
&irradiated
yr old, diagnosed
2 yr be
foreExtended
process, ra
diotherapyRadiotherapyRadiotherapyRadiotherap
previouslyRadiotherapy30
yr old, diffuse co
lon polyposus, ra
diotherapy32
oldExtended
yr
processRadiotherapyMultiple
colon poly
posusRadiotherapyRadiotherapyRadiotherapyDe
process
* ND, not determined.
2 from normal donors (2% of tested individuals) which pre
sented the typical type I or II patterns of NK regulation. It
must be mentioned that 4 other normal sera were also found in
our present investigation to be noninhibitory, but after heating
they did not acquire the ability to inhibit NK activity, and
consequently they were not considered to belong to type II NK
regulation.
Of the 25 colorectal cancer patients who have been treated
surgically, 11 were subsequently tested at various times after
resection of their tumor (Table 7). Five type I patients (Cases
3, 5, 7, 8, and 11) with improving clinical courses, as assessed
by weight, Karnofsky scores, and freedom from detectable
tumor, were found to have changed their type of NK regulation
to III when determinations several months after surgery were
again performed. In the case of Patient 11 with an initially
extended process of malignancy and postoperative residual
tumor tissue, the serum characteristics defining the type I NK
regulation were again found about 1 yr later, when resection of
additional tumor was not possible to be performed. Sequential
immunological determinations of patients with initial type III
2600
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.
MODULATION
OF NK CELL ACTIVITY BY SERUM FROM CANCER PATIENTS
Table 5 The value of serum from patients with coloréela!
adenocarcinoma
assayed at various time intervals on normal PBL isolated from different donors
eters tested together allowed the identification of 3 distinct NK
regulation types characterized, in general, as /~A+/?* (type I),
/+A+/T (type II), or r\~R+ (type III). Lack of inhibition of
A value provided by cancer serum
tested at various times on different
PBL of normal
NK activity by serum from 36% of all our investigated patients
(25
of 69 cases) with various lymphoid or nonlymphoid solid
location)ColonColonColonColonColonRectumColonColonRectumColonColonRectum179.4
malignant
tumors, a percentage similar to that identified in the
N.6791112131718212226Donor's (If79.4
more homogeneous group of patients with colorectal carcinoma
(I)81.6(1)73.6
(15 of 36 cases) (Table 6), was found in only 3 of the 100
(I)95.0
normal donors and 1 of the 30 patients with benign digestive
(I)20.3
(I)31.8(11)29.6
diseases. It should be mentioned that 4 noninhibitory sera of
(II)20.9
(II)21.0(11)21.0(11)17.2(111)12.6(111)0.0(111)241.0(1)52.0(1)49.6(1)145.8(1)62.1
(II)6.7(111)1.3(111)19.4(111)-5.6(111)0.0
patients with type I NK regulation (Table 4) produced even an
augmentation of NK activity relative to medium-treated PBL.
The noninhibitory characteristics (/") of these cancer sera,
donorsCaseD.
serum(tumor
(III)3-1.2(111)17.4(111)
" Type of NK regulation indicated in parentheses was characterized by taking
into account, besides the A value, the other 2 parameters, namely the regulatory
effects of untreated cancer serum as compared to medium-treated and to normal
serum-treated PBL.
of NK regulation revealed that the sera of 5 of 6 cases had
similar inhibitory effects as those found before surgery (Table
7). The results with sera from two additional treated patients
not studied before surgery are also presented in Table 7. The
serum of Case M. T., also in good clinical condition and free
of any detectable tumor, showed a type III pattern of NK
regulation. A different pattern was observed with Case A. G.,
who had sigmoid colon cancer resected 1 yr before our first
investigation when the recurrence was not detected clinically
but whose serum showed a significant A value. Nine mo later,
the patient presented radiologically detectable paracolic lymphadenopathy which was correlated with the appearance of type I
NK regulation.
DISCUSSION
Our study provides evidence regarding the ability of serum
from cancer patients (a) to induce negative /// vitro modulation
of NK activity (/, inhibition) expressed as the percentage of the
activity of medium-treated control cells, (b) to acquire or not
an appreciable difference (A) in its regulatory effects upon heat
treatment, and (c) to inhibit or not the NK function more
strongly relative (R) to normal control serum. These 3 param
indicated by a lack of negative modulation of NK activity or by
even positive modulatory effects, were always reversed by their
testing at 56°Cfor 30 min, a treatment which endowed them
with the ability to reduce NK activity. This heat-induced differ
ence in the regulatory capacity was also observed with 10 /*
cancer sera (20% of tested cases), defined as having the type II
NK regulation. The A* characteristic of 56% (28 of 50) of the
cancer sera tested was observed with only 3 of the 60 control
sera from either patients with nonmalignant digestive diseases
(1 of 23) or apparently healthy donors (2 of 30). The last
pattern of NK regulation (type III) was provided by those cancer
sera which usually possessed only the characteristic to more
strongly inhibit NK activity as compared to normal sera.
These 3 different categories of cancer sera revealed differ
ences among our group of patients with colorectal cancer which
were not paralleled by differences in levels of cytotoxic reactivity
of their NK cells. Thus, the values for the mean ±SEM of LU/
IO7 cells of medium-treated PBL isolated from patients with
type I, II, or III NK regulation were 124.5 ±51.8, 94.5 ±19.1,
and 99.3 ±46.1, respectively, not different among them or
from the NK activity expressed by PBL of normal individuals
(data not shown). Interestingly the reported decrease in relative
number of circulating HNK-1* cells in colon cancer patients
was associated with an increased cytolytic function compared
to the population of HNK-1* cells purified from healthy sub
jects (21), accounting perhaps for the "normal" base-line NK
activity expressed by medium-treated PBL of patients investi
gated in the present study. Additionally, the effector cells of
almost all cancer patients tested responded normally by inhi-
Table 6 Summary of the results obtained with untreated and heat-treated serum isolated from patients with colorectal adenocarcinoma or with benign digestive
diseases
Statistical differences with respect to control values were calculated by Student's ( test for paired data. The control values of NK activity, expressed both as
percentage of the control (41.6 ±2.3, n = 85) and as heat-induced difference (A) (3.1 ±2.2, n = 31), are means of the results furnished by all experiments performed
in this study with normal PBL incubated with normal serum before or after its heat treatment. Incubation of some of these unheated and heated normal sera with
patient PBL in parallel experiments provided mean ±SEM values of NK activity very similar to control values, e.g., 41.6 ±2.5 (n = 100) for the percentage of the
control, and -0.7 ±6.2 (n = 9) for A.
withDonors
serumColorectal
of
I)Colorectalcancer (type
NK activity of PBL treated
of
of cells used
cases1571430Donor inassayPatientNormalPatient
NK
control104.5
of
±6.0°'*(14)'
102.8
(12)43.6
±7.9*
serum
control)50.3
(% of
±7.7 (14)
(11)21.6
54.2
±6.8
NormalPatient
±6.0 (5)
(7)33.7
52.0 ±9.9
±6.611(5)
20.3
(7)4.0
±3.8*
±10.8(5)
(7)52.5
50.9 ±5.0
NormalPatient
±5.7 (14)
(11)32.1
37.5 ±8.2
±3.2(14)
1.5(7)4.8
4.8 ±
±6.6 (14)
(10)32.2
57.3 ±9.8
Normal%
±5.0(29)
28.2 ±4.3 (30)Patient
±1.3(29)
6.8 ±2.6(17)Normal
±4.6 (29)
34.7 ±4.4 (27)
II)Colorectalcancer (type
cancer (type
HI)Benign
serum
(A)67.1
±6.6*(12)
67.0
(8)24.9
±6.2*
digestive diseasesNo.
" Mean ±SEM.
'/>< 0.001.
' Numbers in parentheses, number of experiments.
d P<0.01.
2601
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.
MODULATION OF NK CELL ACTIVITY BY SERUM FROM CANCER PATIENTS
Table 7 Preliminary results of the longitudinal study of serum-induced NK regulatory effects of patients with colorectal adenocarcinoma at various times after tumor
resection
withCase3S7811202122242628A.C.M.
NK activity of PBL incubated
serum%of
of NK regula
tionBefore
of
serum (% of
After
after tumor
cells used in
control0.10.258.018.861.648.341.332.771.510.769.042.83.832.4116.7100.053.887.536.638.733.735.40.127.660.040.015.326.818.08
control)0.150.070.840.467.733.332.646.194.357.787.846.145.226.891.30.121.864.825.066.741.440
surgery
surgeryI
observationsTumor
resection1716815412411153914571645il71221IS1923Donor
NK
assayPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNormalPatientNorma
IIII
freeTumor
IIII
recurrenceTumor
HImi
freeTumor
freeTumor
HIini
freeTumor
freeResidual
iniiHI
tissueExtended
tumor
recurrenceExtended
recurrenceTumor
inHIinin
freeTumor
freeHepatic
métastases.tumor
recurrenceHepatic
inin
métastases.deceased
laterResidual2 mo
inin
inin
tissueHepatic
tumor
métastasesTumor
inIII
recurrenceTumor
HIIIIHI
freeParacolon
fistulaTumor
IIIND
IIIND
recurrenceTumor
freeParacolon
ade-nopathyTumor
IIIIIIinHIClinical
freeTumor
T.Mo
freeTumor
freeTumor
free
' ND, not determined.
bition upon incubation in the presence of mlgG or normal
serum (Tables 1 and 4). These results raise the possibility that
the usual measurements of NK activity of cancer patients'
lymphocytes, at least in regard to patients with colorectal
carcinoma, may not be as indicative of clinical course of disease
or host resistance as the assessment of the patients' reactivity
in the presence of their serum, with the latter more closely
approximating the in vivo situation.
The distinctive patterns of NK regulation by cancer sera were
also observed when the in v/iro-treated cells were PBL isolated
from cancer patients as well as from normal donors. Thus, the
detection of / and A characteristics of sera obtained from
patients with malignant (Tables 2 and 4) or benign diseases or
from healthy individuals and tested comparatively in 70 exper
iments on both patient and normal PBL was concordant in
86% and 88% of experiments, respectively. This rather good
reproducibility, in addition to the small interassay variability
shown in Table 5, strongly supports the conclusion that the /~
and/or A+ characteristics of the substantial number of sera
tested were cancer associated. In fact, these characteristics of
type I pattern of 5 of 5 patients disappeared after tumor
resection, when the patients' sera were obtained again several
months later (Table 7).
An examination of the correlation between the type of serum
2602
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.
MODULATION OF NK CELL ACTIVITY BV SERUM FROM CANCER PATIENTS
regulatory activity and the extent of tumor in a given patient
should take into account that the Dukes and TNM stagings
were determined in some cases at intervals greater than 1 mo,
when the surgery was performed, and they were very likely
inaccurately evaluated in those cases treated by radiotherapy
before tumor resection. In spite of these difficulties, it seems
that the type III NK regulation in tumor-bearing patients can
be correlated with more advanced tumor staging, as Dukes C
or D stages were found in 50% of patients with this pattern (6
cases) and in all but one of the other 6 cases in whom radio
therapy was administered or poor prognostic factors were iden
tified (Table 4). Follow-up of 6 patients of this category revealed
in 5 cases tumor recurrence and even métastases(Table 7).
Conversely, the apparent correlation between the early stage of
malignancy and the type I NK regulation suggests that the 7~A*
serum characteristics of these cancer patients may reflect the
more efficient function in vivo of their circulating NK cells,
refractory to the negative regulatory factors actually present in
the serum. The mechanism underlying the noninhibitory capac
ity of these cancer sera may represent either a block in the
inhibition of cytophilic IgG or other serum inhibitory factors
or, more likely, a stimulatory factor which just regulates in the
opposite direction as mlgG.
It should be noted that the clinical correlations with type of
NK regulation by cancer patients' serum varied considerably
for specimens obtained prior to treatment versus those obtained
after treatment. In the untreated tumor-bearing patients, type
III regulation was associated with poor prognosis. In contrast,
after tumor removal, the shift from type I or II to type III was
associated with a tumor-free state, i.e., the development of a
pattern essentially the same as that seen with normal donors.
The nature of the postulated stimulatory factor in type I sera,
which has been identified functionally in this study, is virtually
unknown except for its thermolability. One possibility is 7interferon, which can be inactivated at 56°C,but recent obser
vations made by us5 and others (22) reveal that this lymphokine
in highly purified form is, in fact, unable to augment within a
period of 4 h the activity of human NK cells. Another major
candidate as potent activator of NK activity might be interleukin 2, but this molecule is thermostable. Experiments are pres
ently under way to identify the nature of this putative stimula
tory factor. Whether its measurement will prove to be valuable
in the assessment of a malignant transformation or the course
of disease in a given patient needs to await further longitudinal
NK studies and clinical follow-up.
ACKNOWLEDGMENTS
We would like to thank Mariana Feher, Elena /.men, and T. Regalia
for excellent technical assistance. The authors gratefully acknowledge
Mariana Pavlovski, Laura Bates, Lori Novak, and Barbara Klewien for
secretarial assistance.
' A. Sulica, M. Gherman, M. Manciulea, and R. B. Herberman. Regulation of
human natural cytotoxicity by IgG. III. Ability of i-interferon to restore the NK
activity of cells negatively regulated by cytophilic IgG, manuscript in preparation.
REFERENCES
1. Herberman, R. B., and Ortaldo, J. R. Natural killer cells: their role in
defenses against disease. Science (Wash. DC), 214: 24-27, 1981.
2. Timonen, T., Ortaldo, J. R., and Herberman. R. B. Characteristics of human
large granular lymphocyte and relationship to natural killer cells and K cells.
J. Exp. Med., 153: 569-576, 1981.
3. Kasahara, T., Djeu. J. T., Dougherty, S. F., and Oppenheim, J. J. Capacity
of human large granular lymphocytes to produce multiple lymphokines:
interleukin-2, interferon, and colony stimulating factor. J. luminimi., 131:
2379-2385,1983.
4. Scala, G., Allavena, P., Djeu, J. Y., Kasahara, T., Ortaldo, J. R., Herberman,
R. B.. and Oppenheim, J. J. Human large granular lymphocytes are potent
producers of interleukin-1. Nature (Lond.), 309: 56-58, 1984.
5. Toder, V., Nebel, L., Elrad, H., Blauk, M., Durdane, A., and Gleicher, N.
Studies of natural killer cells in pregnancy. II. The immunoregulatory effect
of pregnancy substances. J. Clin. Lab. Immun., 14: 129-136, 1984.
6. Tursz, T. Les lymphocytes tuers naturels. Leur role en pathologie. Presse
Med., 14: 219-222, 1985.
7. Gorelik, E., and Herberman, R. B. Role of natural killer cells in the control
of tumor growth and metastatic spread. In; R. B. Herberman (ed.). Cancer
Immunology: Innovative Approaches to Therapy, pp. 151-176. Boston:
Marl inns Nijhoff, 1986.
8. Herberman, R. B. Biological response modifiers as activators of natural killer
cells: potential role in treatment of metastasis. In: K. Honn, B. F. Sloane, J.
Crissman, and W. E. Powers (eds.), Potential Therapeutic Implications, pp.
227-236. New York: Martinus Nyhoff, 1986.
9. Barlozzari. T., Leonhardt, J., Wiltrout, R. V., Herberman, R. B., and Rey
nolds, C. W. Direct evidence for the role of LGL in the inhibition of
experimental tumor métastases.J. Iumiliimi,, ¡34:2783-2790, 1985.
10. Sibbitt, W. L., Bankhurst, A. D.. .Inmm ilio. A. J., Saiki, J. M., Seiers, J. H.,
and Dobernek, R. C. Defects in natural killer cell activity in human lung
carcinoma and malignant melanoma. Cancer Res., 44: 852-857, 1984.
11. Perez, M., Virelizier, J. L., Arenzana-Seisdedos, F., Fisher, A., and Griscelli,
C. Impaired natural killer activity in lymphohistiocytosis syndrome. J. IV
diatr., 104: 569-576, 1984.
12. Herberman, R. B., and Santoni, A. Regulation of natural killer cell activity.
In: E. Mihich (ed.). Biological Response in Cancer, pp. 121-144. New York:
Plenum Publishing Corp., 1984.
13. Trinchieri, G., Santoli. D., and Koprowski, H. Spontaneous cell-mediated
cytotoxicity in humans: role of interferon and immunoglobulins. J. Immunol.,
120: 849-856, 1978.
14. Kuribayashi, K., Gillis, S., Kern, D. E., and Henney, C. S. Murine NK cell
cultures: effect of interleukin-2 and interferon on cell growth and cytotoxic
reactivity. J. Immunol., 126: 2321-2328, 1981.
15. Droller, E. J., Schneider, M. V., and Perlmann, P. A possible role of
prostaglandins in the inhibition of natural and antibody-dependent cellmediated cytotoxicity against tumor cells. Cell Immunol., 29:165-172, 1978.
16. Sulica, A., Gherman. M.. Galatiuc, C., Manciulea, M.. and Herberman, R.
B. Inhibition of natural killer cell activity by cytophilic immunoglobulin G.
J. Immunol., ¡28:1051-1058, 1982.
17. Sulica, A., Gherman, M., Manciulea, M., Galatiuc, C., and Herberman, R.
B. Negative regulation of human NK activity by monomeric IgG. In: R. B.
Herberman (ed.), NK Cells and Other Natural Effector Cells, pp. 621-628.
New York: Academic Press, 1982.
18. Bancu, A. C., Gherman, M., Sulica, A., Gato, T., Parrar, W., and Herberman,
R. B. Regulation of human natural cytotoxicity by IgG. II. Cyclic AMP as a
mediator of monomeric IgG-induced inhibition of natural killer cell activity.
Cell. Immunol., in press. 1988.
19. Wilson, A. B., and Coombs, R. R. A. Fc -y-receptor bearing-lymphocytes in
human peripheral blood: cytophilic immunoglobulin binds almost exclusively
to large granular lymphocytes. Cell. Immunol., 90: 196-202, 1985.
20. Gherman, M., Manciulea, M., Bancu, A. C., Sulica, A., Stanworth, D. R.,
and Herberman, R. B. Regulation of human natural cytotoxicity. I. Charac
terization of the structural site on monomeric IgG responsible for inhibiting
natural killer cell activity. Mol. Immunol., 24: 743-750, 1987.
21. Balch, C. M., Tilden, A. B., Dougherty, P. A., Cloud, H. A., and Abo, T.
Heterogeneity of natural killer lymphocyte abnormalities in colon cancer
patients. Surgery (St. Louis), 95: 63-70, 1984.
22. Sayers, T. J., Mason, A. T., and Ortaldo, J. R. Regulation of human natural
killer cell activity by interferon: lack of role in interleukin-2-mediated aug
mentation. J. Immunol.. 136: 2176-2180, 1986.
2603
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.
Modulation of Natural Killer Cell Activity by Serum from Cancer
Patients: Preliminary Results of a Study of Patients with
Colorectal Adenocarcinoma or Other Types of Cancer
Madalina Pislarasu, A. Oproiu, Doina Taranu, et al.
Cancer Res 1988;48:2596-2603.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/48/9/2596
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1988 American Association for Cancer Research.