1. No category

From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
1993 82: 1204-1211
Tumor-infiltrating lymphocytes derived from select B-cell lymphomas
secrete granulocyte-macrophage colony-stimulating factor and tumor
necrosis factor-alpha in response to autologous tumor stimulation
DJ Schwartzentruber, M Stetler-Stevenson, SA Rosenberg and SL Topalian
Updated information and services can be found at:
http://www.bloodjournal.org/content/82/4/1204.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
Tumor-Infiltrating Lymphocytes Derived From Select B-Cell Lymphomas
Secrete Granulocyte-Macrophage Colony-Stimulating Factor and Tumor
Necrosis Factor-a in Response to Autologous Tumor Stimulation
By Douglas J. Schwartzentruber, Maryalice Stetler-Stevenson, Steven A. Rosenberg, and Suzanne L. Topalian
Tumor infiltrating lymphocytes (TIL)were cultured from 17
B-celllymphoma specimens derived from patients with predominantly low-grade malignancies. Specimens included
1 5 lymph-node biopsies, 1 malignant pleural effusion, and
PBL from 1 patient with circulating lymphoma cells. The
phenotypic and proliferative characteristics of TIL cultured
in interleukin-2 (IL-2) were studied, as well as cytolysis
and cytokine secretion in response t o autologous tumor.
Flow cytometry of fresh tumor suspensions showed that
50% of cells (median) were malignant B cells and 36%
were infiltrating T lymphocytes. After culture for approximately 1 month, TIL were 75% f 8% CD3' (mean f
SEM), 47% f 8%CD4+ and 35% f 7% CD8'. TIL proliferation was modest in most cases: the median maximum
expansion was 32-fold in 25 days. Lysis of autologous tu-
mor in 4-hour ''Cr release assays was mediated by 2 of 12
TIL studied, but was nonspecific. However, these same
two TIL, when cocultured with various tumor stimulators,
preferentially secreted tumor necrosis factor-a and granulocyte-macrophage colony-stimulating factor after autologous tumor stimulation; unstimulated TIL secreted undetectable or barely detectable levels of these cytokines. In
one TIL culture, cytokines were secreted by purified CD4+
TIL but not by CD8' cells, and secretion was completely
abrogated by the anti-major histocompatibility complex
(MHC) class II antibody l V A l 2 . Thus, although specific cytokine secretion by lymphoma TIL in response t o autologous tumor was observed, it occurred in fewer than 20% of
patients studied.
0 1993 by The American Society of Hematology.
I
creased when compared with peripheral blood cells from
normal donors.' In a separate study, when CD3+ TIL were
cocultured with autologous tumor for 6 days, 41% of cultures showed increased proliferation compared with nontumor-stimulated cultures.' The specificity of this response to
tumor was not further studied.
The ability of lymphoma TIL to kill autologous tumor
has not been well defined, although tumor-specific cytolytic
lymphocytes have been identified in the peripheral blood
(PB) of one patient with follicular l y m p h ~ m aTIL
. ~ may
also interact with malignant B cells by suppressing their
function. For example, in one study, Ig secretion by malignant B cells was inhibited by T lymphocytes infiltrating
these lymphomas." Hence, suggestive evidence exists for
host immune reactivity toward lymphoma.
Our group has previously shown that TIL can be expanded under the influence of IL-2 from a variety of solid
tumors, including melanomas,"-I3 breast cancers,I4 colon
cancers," and renal cell carcinomas.I6In these studies, the
majority of melanoma TIL were cytolytic for autologous
tumor, while TIL from other histologieswere generally noncytolytic. However, some noncytolytic cultures showed autologous tumor recognition when assayed for secretion of
granulocyte-macrophage colony-stimulating factor (GMCSF), tumor necrosis factor-a (TNF-a), and/or interferon-y
(IFN-y) in 24-hour cultures. In the current study, using similar culture techniques, we show that CD3+ TIL grown from
fewer than 20%of lymphoma patients are capable of autologous tumor recognition, manifested by secretion of GMCSF and TNF-a. In one instance, cytokine secretion was
mediated by CD4+ T cells in an major histocompatibility
complex (MHC) class I1 restricted manner.
MMUNOTHERAPY consisting of interleukin-2 (IL-2)
alone or in combination with lymphokine activated
killer (LAK) cells for patients with refractory B-cell nonHodgkin's lymphoma (NHL) has resulted in significant
clinical regressions.'s2The mechanisms of regression are not
fully understood but are presumably immune-mediated.
Thus, our interest has focused on the T lymphocytes infiltrating these tumors. Previous studies from other laboratories have associated a favorable prognosis for B-cell lymphoma patients with a high content of tumor infiltrating T
cells,3 a high percentage of infiltrating CD8+ cytotoxic/sup
pressor cells,4and in vitro proliferation of tumor infiltrating
lymphocytes (TIL) in response to autologoustumor stimulation.' Some authors have compared the immune cells infiltrating B-cell lymphoma lymph nodes with those in reactive
lymph node hyperplasia. The finding of greater numbers of
chronically activated T cells and fewer naive cells in the
malignant nodes, suggested an accumulation of memory
cells within tumor involved
Despite numerous reports characterizing the phenotype
of lymphocytes infiltrating B-cell lymphomas and speculating on their possible prognostic significance, limited studies
analyzing their in vitro functions are available. In particular, little experience exists with retrieving and culturing
CD3+TIL from patients with lymphoma. In one study, TIL
were isolated and cultured under limiting dilution conditions, and their proliferation frequency was noted to be de-
From the Surgery Branch and Department ofPathology, National
Cancer Institute, National Institutes of Health, Bethesda, MD.
Submitted October 5, 1992; accepted April 8, 1993.
Address reprint requests to Douglas J. Schwartzentruber, MD,
National Cancer Institute, 9000 Rockville Pike, Bldg 10, R m 2B04,
Bethesda, MD 20892.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C.section 1734 solely to
indicate this fact.
0 1993 by The American Society of Hematology.
0006-4971/93/8204-0009$3.00/0
1204
MATERIALS AND METHODS
Lymphocyte cultures. Lymphocytes were cultured from specimens derived from I 7 B-cell lymphoma patients. Fifteen specimens
were lymph-node biopsies, one was a malignant pleural effusion,
and one was PB from a patient with circulating malignant cells. For
descriptive purposes, lymphocytes cultured from the pleural effusion and PB lymphocytes (PBL)are referred to as TIL in this manuBlood, V 0 l 8 2 . N 0 4 ( A ~ g ~ 15),
s t 1993:pp1204-1211
From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
1205
LYMPHOMA TUMOR INFILTRATING LYMPHOCYTES
script. Cultures were initiated and maintained as previously
described.'ls1'Solid lymph-node tissues were minced and enzymatically digested (hyaluronidasetype V 0.01%, DNAse type I0.002%,
collagenasetype IV 0.1%; Sigma, St Louis, MO) for 4 hours at 37°C
or up to 24 hours at room temperature. The resultant single-cell
suspensions were passed over a Ficoll-Hypaque (Organon Teknika,
Durham, NC) density gradient if viability was poor or if there were
numerous red blood cells (RBCs). The malignant pleural effusion
was obtained by thoracentesis and the cell pellet washed twice in
Hanks' balanced salt solution (HBSS). The cell suspensions, containing tumor cells and normal lymphocytes, were cultured in 24well tissue culture plates (Costar, Cambridge, MA) at a density of
0.5 x lo6to 1.O X IO6 cells/mL in media consisting of RPMI 1640 +
10% heat-inactivated human AB serum, 50 IU/mL penicillin, 50
pg/mL streptomycin, 50 pg/mL gentamicin, 250 ng/mL amphotericin, 10 mmol/L HEPES buffer and 2 mmol/L L-glutamine. This
media was supplemented with supernatants from 4-day-old allogeneic LAK cell cultures at a concentration of 20% (vol/vol). Recombinant IL-2 (rIL2, specific activity 3.6 X lo6 Cetus U/mg, determined by CTLL assay; 1 Cetus U = 6 IU; kindly provided by the
Cetus Corp, Emeryville, CA) was added at a final concentration of
6,000 IU/mL. Cultures were maintained at 37"C, in 5% CO,, and
passaged or media replenished every 7 to 10 days. In addition to
establishing cultures, samples of the initial tumor-cell suspensions
were cryopreserved for later use in various assays.
LAK cell cultures were established from normal donor PBL by
methods previously
In brief, lymphocytes were retrieved from the interface of a Ficoll-Hypaque density gradient.
They were cultured at 1 X 106/mLin RPMI 1640 media containing
2% heat-inactivated human AB serum, antibiotics, HEPES buffer
glutamine and 6,000 IU/mL rIL-2. Cultures were maintained in
175 cm2 flasks (Falcon; Becton Dickinson, Oxnard, CA) for 4 to 7
days. The supernatants from 4-day cultures were used, as described
above, to supplement TIL cultures. LAK cells were used as effectors
in cytolysis assays.
Tumor lines. Daudi, a LAK-sensitive B-cell lymphoma line,
was cultured by serial passage in RPMI 1640 + 10%fetal calf serum
(FCS). Two Epstein-Barr virus (EBV)-transformedB-cell lines were
previously generated from PBL by standard techniques. Cultured
melanoma lines were established as describedI7 and serially passaged in the above media.
Phenotype analysis. Flow cytometry was performed on fresh
tumor suspensions as well as on growing TIL cultures using methods previously described." Antibodies used were anti-Leu-4 (CD3,
pan T cell), anti-Leu-3 (CD4, T helper/inducer cells), anti-Leu-2
(CD8, T cytotoxic/suppressor cells), anti-Leu-7 (CD57, natural
killer (NK) cells, T-cell subsets), anti-Leu-19 (CD56, NK cells,
LAK cells, T-cell subsets), anti-HLA-DR (MHC class 11), antiLeu- 12 (CDI 9, B cells), anti-Leu- 16 (CD20, B cells), anti-Leu- 14
(CD22, B cells), anti-K (IgG light chain), anti4 (IgG light chain),
anti-IL-2 receptor (CD25), and anti-Thy 1.2 (murine T cells) (Becton Dickinson, Mountain View, CA).
Cytotoxicity assays. The cytotoxicity of cultured TIL against
fresh cryopreserved and cultured tumor targets was measured in
standard 4-hour "Cr release assays, as previously described." Cryopreserved tumors were thawed by rapid rewarming and passed over
Ficoll-Hypaquedensity gradients if viability was less than 50%.The
number of tumor cells in each target was estimated by counting all
viable cells and multiplying by the percent tumor cells contained in
similarly cryopreserved tumor-cell suspensions that had been analyzed by flow cytometry. The ratio of spontaneous/maximum 51Cr
release was less than or equal to 30%for all tumor targets except for
two (35% for tumor 1103 and 36% for an allogeneic lymphoma in
the same experiment).
Measurement of cytokine secretion by TIL. TIL were cocultured with various tumor stimulators, and cytokineswere measured
in the culture supernatants as previously
Briefly, TIL
that had received fresh media 5 to 8 days before, were washed twice
and plated (1 X 106/well)in 24-well tissue culture plates in 1 mL of
media containing 120 IU/mL rIL-2. TIL were either cultured alone
or with 5 X lo5 tumor stimulators. In addition, TIL were cultured
in the same media containing 120 IU/mL of IL-2 on anti-CD3
monoclonal antibody (MoAb) (OKT3, Ortho Pharmaceuticals,
Raritan, NJ) coated tissue-culture plates as a positive control for
cytokine production.14Fresh cryopreservedtumor stimulators were
thawed by rapid rewarming and passed over a Ficoll-Hypaque density gradient if viability was less than 50%.A 137Cssource was used
to deliver 15,000rad to fresh tumors and Daudi, and 30,000 rad to a
cultured melanoma line. Culture supernatants were collected after
24 hours (or 6 hours for some experiments)and after centrifugation
were stored at -70°C. Cytokine measurements were performed on
thawed supernatants in duplicate using commerciallyavailable enzyme-linked immunosorbent assays(ELISA). Standard curves were
generated for each assay and experimental values were computed
with the use of regression analysis. If detectable, background cytokine values produced by tumors alone were subtracted from the
coculture values when reporting results. Cytokine secretion by TIL
was considered relevant when tumor-stimulated values were twofold greater than unstimulated values, and preferential for autologous tumor when twofold greater than allogeneictumor-stimulated
values. GM-CSF was measured with Factor-Test Human GM-CSF
ELISA Test Kit (Genzyme, Boston, MA); lowest detectableconcentration = 4 pg/mL. TNF-a was measured with Quantikine Human
TNF-a Immunoassay (R & D Systems, Minneapolis, MN); lowest
detectable concentration = 4.8 pg/mL or with Factor-Test Human
TNF-a ELISA Test Kit (Genzyme); lowest detectable concentration = 12.5 pg/mL. IFN--y was measured with Inter Test-y Human
IFN-7 ELISA Kit (Genzyme); lowest detectable concentration =
100 pg/mL.
TIL separation into CD4+ and CD8+subsets. Positive selection
of CD4' and CD8+TIL was performed using tissue culture flasks to
which anti-CD4 or CD8 MoAb were covalently bound (Applied
Immune Sciences, Menlo Park, CA) as previously des~ribed.",'~
Antibody blocking of TIL cytokine secretion. In an attempt to
inhibit TIL cytokine release, tumor stimulators were pretreated
with either anti-Thy 1.2, W6/32 (anti-HLA-A, -B, -C, Sera Lab
Accurate Chemical and Scientific Corporation, Westbury, NY)or
IVA12 (anti-HLA-DR, -DP, -DQ, purified from hybridoma culture supernatants; American Type Culture Collection, Rockville,
MD). Tumor cells ( 5 X lo5)in 0.4 mL of media were incubated with
0.1 mL of MoAb for 30 to 45 minutes at room temperature before
the addition of TIL (final concentrations of MoAb were 25 to 50
pg/mL). Culture supematants were then collected and cytokines
measured as described above.
RESULTS
Tumor specimens and TIL culture characteristics. Seventeen patients with B-cell lymphomas were studied (Table
1). Most had stage IV disease and had received multiagent
chemotherapy. Their histologic classification according to
the Working Form~lation'~
was 12 low, 2 intermediate and
3 high grade. Segments of lymph nodes were obtained from
surgical diagnostic biopsies from 15 patients, and a 400-mL
thoracentesis specimen was obtained from 1 patient with a
malignant pleural effusion. In addition, one patient with
circulating malignant cells underwent phlebotomy to obtain PB cells for study (Table 2). All but one of the solid
From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
1206
SCHWARTZENTRUBER ET AL
lymph node specimens were enzymatically digested as described in the Materials and Methods (tumor 1206 was mechanically disaggregated); the majority of node segments
were 1 to 2 cm in greatest dimension. The median number
of viable cells retrieved from lymph node specimens was 4.8
X IO8 (range, 1.1 to I7 X 10'); the median viability was 95%
(range, 67% to 100%).
Flow cytometry was performed on fresh cryopreserved
tumor cell suspensions and the median percentage of tumor
was determined to be 50% from the measurements of
various B-cell markers (Table 2). Infiltrating T cells comprised 36% of cells. A more detailed phenotypic analysis
(Table 3) showed that 58% of T cells were CD4' (helper
lymphocytes) and 3 I % of T cells were CD8' (cytotoxic/suppressor lymphocytes). In 10 fresh specimens studied, a
mean of 67% of CD3+ cells were activated as shown by their
coexpression of HLA-DR. Cells expressing the B-cell
markers CD 19, CD20 and/or CD22 accounted for 48% to
52% of the total cell number. Monoclonality of these B-cell
populations was determined with anti-IgG h or K antibodies
and this information was used in differentiating normal
or
from malignant B cells. Normal B cells were rare (4%)
nonexistent except in tumors I 103, I 13I , and I I57 (containing 6%, 2576, and 6% normal B cells, respectively). As
expected, all B cells expressed HLA-DR. Tumor suspensions contained very few NK cells (CD56+, mean 6%) with
the exception of specimen 1048 in which these cells were the
predominant normal lymphocyte population present.
Seventeen tumor cell suspensions were placed in culture
with IL-2 (Table 2). Five cultures were established from
cryopreserved cells and this did not appear to influence the
subsequent lymphocyte proliferation. The median maximum cell proliferation for all cultures was 32-fold (range, 0
Table 1. Characteristics of 17 Patients With B-Cell Lymphoma
Patient
No.
Staqe'
986
1048
1062
1084
1094
1096
1103
1115
1124
1131
1139
1150
1157
1159
1198
1206
1218
IV
111
111
IV
IV
IV
IV
IV
IV
Ill
IV
Ill
IV
IV
IV
IV
IV
Histologic
Gradet
Intermediate
High
Low
Low
Low
Low
Low
Low
Low
Low
Low
High
Low
High
Low
Intermediate
Low
Prior
Theraovt
Chemo.
Chemo
None
Chemo
Chemo
Chemo,
Chemo
Chemo
Chemo,
None
Chemo,
Chemo,
Chemo,
Chemo,
Chemo,
Chemo,
Chemo
Rad
Rad
dFN
Rad, IL-2
IL-2
IL-2
Rad
IL-2
IL-2
Abbreviations: Chemo, multiagent chemotherapy; Rad, radiation therapy.
Ann Arbor staging classification.
t Working formulation."
Prior therapy at least 30 days before surgical biopsy.
+
Table 2. Characteristicsof 17 B-Cell Lymphoma Specimens and
Lymphocyte Proliferation in Culture
Lymphocyte Proliferation
Specimen
Cell
Yield
No.
(XlO-')
986
1048
1062
1084
1094
1096
1103
1115
1124
1131
11397
1150
1 157#
1159
1198
1206"'
1218
Median
Range
8.4
10
1.6
17
4.0
11
2.9
10
4.4
4.1
6.1
1.9
98
1.1
4.8
9.5
10
6.1
1.1-98
% Tumor'
% T Cells'
49
71
18
50
33
90
57
74
50
25
94
8
42
70
91
34
69
50
8-94
46
6
69
41
20
5
47
26
50
55
2
83
36
22
7
35
28
36
2-83
Day of
Maximum
Expansion
Maximum
Expansion
lndext
18
20
25
14
70
23
24
106
NA
161
77
23
105
22
104
23
32
25
14-161
5.0
32t
1.9
4.4
4.1 X I O 2
65
l.l§
1.5 x 105
0
7.4 x 10611
1.5 X 10'
6.5t
2.7 x 104$
1.ot
5.4 x 103t
1.7
93
32
0-7.4 X 10'
Abbreviation: NA, not applicable.
Determined by flow cytometry of single-cell tumor suspensions. Tumor cells represent monoclonal populations expressing IgG h or K.
t Maximum expansion index = extrapolated maximum no. of lymphocytes/no. of lymphocytes at start of culture.
Culture initiated from cryopreserved tumor cell suspension.
§ Culture electively terminated.
I1Terminated because of contamination.
ll Tumor source: PBL, 2 1 cc.
# Tumor source: malignant pleural effusion, 400 cc.
** Tumor mechanically disaggregated (no enzymes).
*
to 7.4 X lo6). Maximum lymphocyte proliferation occurred
on day 25 of culture (range, 14 to 161). One culture was
electively terminated when its growth curve had reached a
plateau and one culture was terminated because of contamination. The remainder of cultures were maintained beyond
the day of their maximum expansion.
After a mean of 33 days of incubation, 16 TIL cultures
could be phenotyped and were found to contain virtually no
tumor cells (CD20-) or monocytes/macrophages (CD 14-)
(Table 4). Most cultures were composed predominantly of
T cells (mean 75% CD3+). Nine cultures were predominated by CD4+ cells and seven cultures by CD8+ cells. Four
cultures (1048, 1084, 1 150, and 1206) contained only 5% to
39% CD3+ cells and greater than 80%CD56+ cells. Interestingly, while tumors 1048 and I 150 contained significant
CD56+ cell populations at the outset, tumors 1084 and 1206
contained only 0% and 1 % CD56' cells before culture, respectively.
Cytolysis by TIL. Cytolysis by 12 B-cell lymphoma TIL
after 24 to 76 days in culture was tested in standard 4-hour
51Crrelease assays (Table 5). The cytolytic activity of each
TIL was tested once and repeated twice for TIL I 13 I . Only
2 of 12 cultures, TIL 1048 and 1 131, lysed fresh autologous
From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
1207
LYMPHOMA TUMOR INFILTRATING LYMPHOCYTES
Table 3. Cell-Surface Phenotvoes of Fresh Tumor SusDensions
% Positive Cells
Tumor No.
CD3
CD4
CD8
CD3+/HLA-DRt
CD56
CD19
CD20
CD22
986
1048
1062
1084
1094
1096
1103
1115
1124
1131
1139
1150
1157
1159
1198
1206
1218
46
6
69
41
20
5
47
26
50
55
2
83
36
22
7
65
28
37
0
58
28
12
4
13
4
37
37
2
35
25
4
5
42
16
4
3
12
12
6
1
38
9
5
11
0
37
10
15
1
21
9
30
ND
ND
ND
ND
ND
ND
ND
42
5
20
0
7
2
1
3
1
12
7
3
0
1
1
38
34
12
49
19
81
63
81
49
50
95
8
48
71
89
33
72
49
62
29
24
33
67
65
20
46
46
94
7
61
69
89
34
72
50
36
14
52
5
51
58
32
52
37
95
18
53
75
91
34
63
Mean f SEM
Range
36 f 6
2-83
21 f 4
0-58
11 23
24 f 6
1-69
6*3
0-42
52 f 6
8-95
51 ? 6
7-94
48 6
5-95
0
ND
19
28
30
1
69
14
24
6
ND
0-38
CD3-/HLA-DRf
52
ND
ND
ND
ND
ND
82
73
50
43
95
18
59
74
92
25
68
*
61 f 7
18-95
Abbreviation: ND, not done.
tumor targets to a significant degree (greater than 10%lysis
at an effector: target ratio of 40:l). TIL 1048 lysed Daudi
equally as well as autologous tumor, as might be expected
from their phenotype that was predominantly CD56'. TIL
1131, which expressed T-cell markers but not LAK cell
markers, displayed low levels of lysis of the autologous as
well as other tumor targets, making it difficult to comment
on tumor specificity. Despite T-cell subset separation on
antibody-coated flasks, lysis by purified CD8' TIL 1 131 did
not exceed lysis by the bulk culture (two separate experiments, data not shown). Even though lysis of autologous
tumor targets was limited in this series, 6 of 12 TIL were
capable of lysing Daudi. This LAK-like lysis is consistent
with the high numbers of CD56+ cells (29% to 93%)seen in
4 of these 6 TIL. Though only 2 of 12 autologous tumors
were lysed by TIL, 9 of these 12 tumor targets were lysed by
allogeneic LAK cells at levels greater than 10%.
Cytokine secretion by B-cell lymphoma TIL. Cytokine
Table 4. Phenotypes of 6-Cell Lymphoma TIL Cultures
% Positive Cells
TIL No.
986
1048
1062
1084
1094
1096
1103
1115
1131
1139'
1150
1157t
1159
1198
1206
1218
Mean ? SEM
Range
Culture
Day
CD3
CD4
CD8
CD57
CD56
CD20
CD14
HLA-DR
20
26
25
28
34
33
25
42
25
75
29
27
62
37
36
23
81
19
93
26
96
95
68
97
98
98
39
97
99
99
5
91
60
2
62
11
29
71
22
91
68
56
65
21
98
27
2
67
11
71
14
51
74
21
35
10
37
44
4
81
2
68
7
33
21
11
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
ND
ND
ND
14
9
9
5
0
6
3
8
70
93
5
82
15
13
29
5
3
39
82
5
11
4
94
9
ND
ND
ND
ND
87
21
25
85
73
97
29
88
75 f 8
5-99
47 f 8
2-98
35 i 7
2-81
922
0-26
35 2 9
3-94
0
0-1
0
0-1
57* 10
9-97
*
33 4
20-75
ND
3
26
7
11
1
Abbreviation: ND, not done.
Lymphocyte culture established from PEL of a patient with malignant circulating cells.
t Lymphocyte culture established from a malignant pleural effusion.
0
ND
ND
0
ND
0
0
0
1
60
9
ND
ND
ND
ND
From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
SCHWARTZENTRUBER ET AL
1208
Table 5. Tumor Lysis by Lymphoma TIL Cultures
Tumor Targets (% Lysis, E:T = 40: 1)'
Effectors
No.
Culture
Day
Autot
Allo$
Daudi
Auto Tumor
Lysis by
LAK Cells
1048
1084
1094
1096
1103
1115
1131
11397
1150
1 157"
1198
1218
27
31
36
35
24
41
26
76
30
30
37
24
43
7
3
8
-1
-4
14
4
7
5
-7
7
2
10
15
-5
7
1
-4
3
15
-3
-4
11
52
68
6
2
52
38
7
-1
58
-1
1
36
35
13
59
33
12
25
8
12
7
12
-6
12
TIL
* Composite results of 4-hour 5'Cr release assays.
t Fresh autologous tumor target.
t Allogeneic tumor target is fresh B-cell lymphoma (tumor 1084 for all
experiments except: TIL 1048, allo is tumor 1 131; 1084, allo is 1048;
1094, allo is 1096; 1096, allo is 1094; and 1150, allo is 1 157).
§ Lymphocyteculture established from PEL of a patient with malignant
circulating cells.
'I Lymphocyte culture established from a malignant pleural effusion.
secretion by TIL was measured after cultures were established for 24 to 76 days. TIL (1 X IO6 per/mL) were cultured
alone, with irradiated tumor cells (5 X lo5 per/mL) or with
immobilized anti-CD3 MoAb for 6 to 24 hours, and GMCSF, TNF-a, and IFN-7 were measured in the supernatants. GM-CSF secretion is shown in Table 6. Detectable
levels of this cytokine were secreted by unstimulated TIL in
8 of 12 cultures. When TIL were cocultured with autologous tumor, 3 of 12 TIL secreted amounts of GM-CSF more
than two-fold greater than unstimulated TIL (TIL 1048,
1084, and 1 131). However, when allogeneic tumor targets
were cocultured with TIL, preferential cytokine secretion
after autologous tumor stimulation was noted in only 2 of
12 cultures, TIL 1048 and 1 131. TIL 1048 secreted 289
pg/mL/24 hours of GM-CSF when stimulated with autologous tumor, but only 8 pg/mL when cultured alone; TIL
I131 secreted 243 pg/mL/24 hours when stimulated with
autologous tumor, but 13 pg/mL alone. Autologous tumorstimulated cytokine secretion was at least twofold greater
than that elicited by any allogeneic tumor stimulator. However, GM-CSF secretion was also observed in response to
some allogeneic stimulators (twofold greater than secretion
by TIL alone). While it is possible that TIL were reacting to
shared tumor antigens presented by common HLA molecules on these allogeneic stimulators, limited amounts of
reagents precluded further investigations along these lines.
TIL 1048 and 1131 were simultaneously tested in one
criss-cross experiment to further evaluate the specificity of
this TIL-tumor interaction. GM-CSF (Table 7) and TNF-a
(data not shown) were preferentially secreted by TIL 1048
and 1 131 after stimulation by their respective autologous
tumors. Little or no secretion was elicited by these tumors
when used as allogeneic stimulators, or by tumors from two
Table 6. GM-CSF Secretion by Lymphoma TIL in Response
t o Tumor Stimulation [pg/mL/24 h)
TIL Responder Cells
Stimulators
TILNo.
Culture
Day
None
Auto'
Allot
Daudi
Anti-CD3
1048
1084
1094
1096
1103
1115
1131
1139t
1150
11575
1198
1218
28
30
36
35
24
41
26
76
28
26
37
24
8
<4
100
91
52
>512
13
33
<4
<4
<4
104
289
12
101
157
48
368
243
43
<4
<4
<4
65
13
42
121
71
55
286
58
43
t 4
<4
<4
70
29
ND
<4
292
83
7512
112
76
<4
<4
<4
ND
62
110
>512
502
475
>512
>512
>512
285
>512
>512
,512
TIL (1 x 10e/mL)were culturedalone or with irradiatedtumor cells (5 X
105/mL) for 24 hours and GM-CSF was measured in the supernatants by
ELISA. All tumor alone controls produced t 4 pg/mL of GM-CSF with the
exception of tumor 1048 = 61 pg/mL and the allogeneic tumor in exp.
1084 = 41 pg/mL. These values were subtracted from the experimental
values.
Abbreviation: ND, not done.
Fresh irradiated autologous tumor.
t Allogeneic tumor is fresh 6-cell lymphoma (tumor 1084for all experiments except: TIL 1084, allo is tumor 1048; 1094, allo is 1096; 1096,
allo is 1094; and 1 150, allo is 1 1 57).
Lymphocyteculture established from PBL of a patient with malignant
circulating cells.
§ Lymphocyte culture established from a malignant effusion.
*
other patients. TIL 1048 were tested in two separate experiments (Tables 6 and 7) in which a total of four different
allogeneic fresh lymphomas, one allogeneic EBV-B-cell
line, and Daudi were used as stimulators. TIL 1 13 1 were
tested in three separate experiments (Tables 6, 7, and 8) in
which a total of six different allogeneic lymphomas, one
allogeneic EBV-B-cell line, and Daudi were used as stimulators. One experiment with TIL 1131 was performed with
purified CD4+ and CD8+ T-cell subsets (Table 8, discussed
below). Of note, all TIL except 1048 secreted large amounts
of GM-CSF when cultured on immobilized anti-CD3
MoAb.
Table 7. Specificity of GM-CSF Secretion by Lymphoma TIL 1048
and 113T
GM-CSF Secretion (pg/mL)'
~
TIL
No.
1048
1131
TIL
Alone T l L + T u 1 0 4 8 TlL+Tu1131
t4
<4
>512
t4
13
56
TIL+Tu1103 TlL+Tu1115
<4
14
<4
<4
Abbreviation: Tu, tumor.
TIL (1 X 106/mL)were cultured alone or with irradiated tumor cells (5
x 105/mL)for 24 hours and GM-CSF was measured in the supernatants
by ELISA. All tumor alone controls secreted s 5 pg/mL of GM-CSF and
were subtracted from the experimental values.
From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
1209
LYMPHOMA TUMOR INFILTRATING LYMPHOCYTES
Table 8. Antibody Blocking of Cytokine Secretion
by Lymphoma TIL After Autologous Tumor Stimulation
Cytokine Secretion
Ipg/mL/24 h)
Antibody
Experiment 1
1 13 1 Unseparated TIL
Tumor alone
TIL alone
TIL tumor
+
None
None
None
Anti-Thy 1.2
W6/32
IVA12
Experiment 2
1131 CD4* TIL
Tumor alone
TIL alone
TIL + tumor
1131 CD8'TIL
TIL alone
TIL +tumor
None
None
None
Anti-Thy 1.2
W6/32
IVA12
None
None
GM-CSF
TNF-a
<4
<4
56
46
27
<4
<12
<4
5
100
109
70
5
<12
<12
45
63
54
<12
14
5
<I2
<I2
29
49
27
4 2
14
TIL (1 X 10s/mL)were cocultured with irradiatedtumor cells (5 x 105)
for 24 hours and cytokines measured in culture supernatants by ELISA.
Tumors were incubated with HBSS or MoAbs for 3 0 minutes at room
temperature before coculture with TIL.
Same experiment as in Table 7.
TNF-a secretion was measured simultaneously with GMCSF in all TIL culture supernatants. As for GM-CSF, only
TIL 1048 and I 131 preferentially secreted TNF-a after autologous tumor stimulation. However, alloreactivity was
minimal. These results are depicted in Fig 1. For both TIL
1048 and 1131, autologous tumor-stimulated TIL produced much higher amounts of TNF-a (97 and 87 pg/mL,
respectively) than unstimulated TIL (9 and <5 pg/mL respectively) or TIL stimulated by a panel of allogeneic tumors. All 12 TIL cultures secreted high amounts of TNF-a
when stimulated with anti-CD3 MoAb ( 121 to > 1,000 pg/
mL/24 hours).
In addition to measuring GM-CSF and TNF-a, IFN-7
100
secretion was measured in all TIL culture supernatants. No
TIL secreted IFN-7 either when unstimulated or stimulated
with autologous tumor (data not shown). However, 10 of 12
cultures secreted IFN-7 after anti-CD3 MoAb stimulation
(260 to >6,400 pg/mL/24 hours).
Antibody blocking of cytokine secretion. To investigate
the immunologic specificity of the interactions of TIL 1048
and 1131 with autologous tumor, we attempted to inhibit
cytokine secretion with anti-MHC MoAb.
Antibody blocking studies with TIL I I3 1 are shown in
Table 8. In the first experiment, unseparated bulk-cultured
TIL, containing 76% CD4+ cells and 28% CD8+ cells, secreted GM-CSF and TNF-a in response to autologous tumor stimulation. This secretion was abrogated by the MoAb
IVA 12 (anti-MHC class II), indicating a role for CD4+ cells
in cytokine secretion. No inhibition was mediated by the
irrelevant anti-Thy 1.2 MoAb. While the anti-MHC class I
MoAb W6/32 failed to inhibit TNF-a secretion compared
with the control without MoAb, it did mediate a 52% reduction in GM-CSF release, suggesting a possible role for CD8+
cells as well. This was further explored with purified CD4+
and CD8+ subsets of TIL 1 131. As shown in the second
experiment in Table 8, CD8+ TIL failed to secrete cytokines, while purified CD4+ TIL secreted more cytokine than
the unseparated TIL and were markedly inhibited by the
IVAl2 MoAb. No cytokine secretion was elicited by four
allogeneic lymphomas in these two experiments (data not
shown). Thus, these data suggest that CD4+ TIL derived
from this lymphoma can recognize autologous tumor in a
specific, MHC class 11-restricted manner.
Secretion of GM-CSF and TNF-a by autologous tumorstimulated TIL 1048 could not be blocked by anti-MHC
class I or class I1 antibodies (data not shown). This finding is
consistent with the surface phenotype of this culture (19%
CD3+, 93% CD56') which is characteristic of non-MHC
restricted LAK cells.
DISCUSSION
TIL have been selectively cultured from several types of
solid human tumors and their in vitro properties studied.
Specific autologous antitumor immune reactivity, as measured by cytolysis and/or cytokine secretion, has been described in CD3+ TIL cultured from select melanoma^,'^,'^
-
-
Stimulators
0None
80
Fig 1. Specific secretion of TNF-u by lymphoma
TIL in response t o autologous tumor. TIL (1 X 1Os/
mL) were cultured alone or with irradiated tumor
cells (5 X 105/mL), for 24 hours (TIL 1048) or 6
hours (TIL 1 131). TNF-a was measured in the culture supernatants by ELISA. Background values of
TNF-a produced by tumor alone (1048 = 6 1 pg/
mL. 1131 = 7 pg/mL, EBV-B cells r 2 7 pg/mL, and
Daudi ~ 1 pg/mL)
0
have been subtracted from the
experimental values. Not depicted is TNF-u production by TIL when stimulated by anti-CD3 MoAb: TIL
1048 = 121 pg/mL and TIL 1131 greater than
1,000 pg/mL.
-
I..._..I
=
1
EBV-BLine
DID3 Daudi
E
2
Autologous Tumor
All0 Lymphoma
EXXl All0 Lymphoma
60-
G
U
+
z
40
-
20
-
0
TIL 1048 (Day 28)
TIL 1131 (Day 26)
From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
1210
breast cancers,l4and colon cancers.I5 However, little information is available regarding the in vivo function and in
vitro properties of B-cell lymphoma TIL. Recent studies
showing the clinical responsiveness of some lymphoma patients to IL-2-based immunotherapies's2suggest that antitumor immune reactivity may be present in such patients.
This study of B-cell lymphoma TIL was undertaken to
search for in vitro evidence of such immune responsiveness.
We found that fresh tumor-cell suspensionsderived from
17 NHL patients were composed of 50% tumor cells, similar
to another study.20Thirty-six percent of cells were T lymphocytes, with CD4' cells predominating over CD8+ cells
(mean CD4/CD8 ratio = 2.9). This finding is consistent
with other lymphoma studies, in which T lymphocytes accounted for 23%to 4 1% of cells, with CD4/CD8 ratios of 1.4
to 2.4.3,7,20-22
Of note, 67% of T lymphocytes in our study
were activated, as shown by their cell-surface expression of
HLA-DR. This is slightly higher than the 23% to 48% reported by
Fifteen of 17 TIL specimens proliferated when cultured
long-term in the presence of IL-2, resulting in overall maximum expansion of 32-fold on day 25 of culture. This maximum expansion was less and occurred sooner than that reported for TIL from a variety of histologies including
melanoma," breast,14 renal cell,I6 and colon ~arcinoma,'~
which had proliferated 2,700-, 6,700-, 68,000-, and 606fold, respectively, on days 6 1,65, 33, and 49. No such data
exist for lymphoma TIL, because this is the first attempt, to
our knowledge, to establish long-term lymphoma TIL cultures. Based on their projected expansion, 4 of 17 cultures
could have reached the numerical target of 1 X 10" cells
that is used in the immunotherapy of melanoma patients.23
Because the majority of specimens obtained were segments
of routine lymph-node biopsies, it is likely that greater cell
numbers could be generated from larger planned biopsies.
Long-term culture of fresh tumor cell suspensions in IL-2
resulted in the complete disappearance of malignant B cells
and the emergence of activated T-cell cultures with a predominance of CD4+ cells. This emergence of CD4+ cells in
long-term culture is similar to TIL-derived from renal ce11,I6
breast,I4 and colon1scarcinomas, but contrasts with melanoma TIL cultures that become predominantly CD8+ over
time." Interestingly, four lymphoma TIL cultures became
predominantly CD56' with low expression of CD3, consistent with a LAK cell phenotype, despite the very small numbers of these cells present in the initial tumor-cell suspensions.
The particular interest of this study was to functionally
characterize T lymphocytes grown from lymphoma specimens. The lytic capacity of TIL was tested in 4-hour 51Cr
release assays. Significant lysis of autologous tumor was detected in only 2 of 12 cultures tested. One of these, TIL
1048, expressed a LAK cell phenotype and was capable of
lysing the MHC class I negative Daudi lymphoma as well as
autologous tumor. TIL 1 131, which expressed a T-cell phenotype, lysed the autologous tumor at low levels and consequently it was impossible to discern preferential lysis towards the tumor of origin. Even though most fresh
lymphoma targets were lysed by allogeneic LAK cells, levels
SCHWARTZENTRUBER ET AL
of lysis were generally low and this is comparable with the
lysis reported from autologous LAK cells in seven lymphoma patients.24Hence, the poor cytolytic capacity of Bcell lymphoma TIL, or the intrinsic resistance of tumor targets to lysis, did not allow for the identification of specific
MHC-restricted antitumor interactions. This lack of cytolytic activity by TIL is similar to that observed in bread4
and colon cancerT1L.I' In contrast, about one third ofmelanoma TIL cultures exhibit tumor-specific MHC class I-restricted cytolysis of autologous tumor.I2
In searching further for antitumor immune reactivity, cytokine secretion by B-cell lymphoma TIL in response to
stimulation with a panel of tumor targets was measured.
TIL 1048, which displayed a LAK cell phenotype, preferentially secreted GM-CSF and TNF-a after autologous tumor
Stimulation. As expected, this TIL-tumor interaction was
not MHC mediated, because cytokine secretion could not
be abrogated by anti-MHC class I or class I1 antibodies. It
has been reported that LAK cells can secrete TNF and IFNy when stimulated by the erythroleukemia cell line K562.25
In the same study, a panel of leukemia and lymphoma cell
lines stimulated LAK cells to secrete cytotoxic factors into
the culture supernatants that were capable of killing MCF-7
breast cancer cells. These factors were subsequently found
to include TNF and IFN-y, suggesting a direct antitumor
role of cytokine secretion by lymphocytes. The specificityof
TIL 1048 for autologous tumor in the context ofa LAK cell
phenotype also raises the possibility that nonspecific cytokine secretion might be elicited by testing additional stimulators.
TIL 1 131 also preferentially secreted GM-CSF and TNFa when stimulated by autologous tumor, which was at least
twofold greater than secretion stimulated by a panel of allogeneic tumor stimulators. By isolating subpopulations of T
cells, we demonstrated that CD4+ 1 131 TIL (and not CD8'
TIL) were responsible for this cytokine secretion. These TIL
interacted with tumor in a classic MHC-restricted manner,
because anti-MHC class I1 MoAb completely abrogated cytokine secretion. The possibility of an autologous mixed
lymphocyte reaction could not be completely excluded, because autologous PBL and EBV-transformed B-cell lines
were not available for use as stimulators in these cytokine
experiments.
Evidence for antitumor immune reactivity by T lymphocytes in B-cell lymphoma, previously unreported for this
disease, is not unique to this tumor. Specific cytokine secretion has been described in TIL from melanoma," breast
cancer,I4 and colon carcinoma15; experiments with TIL
from the first two histologies have included EBV-transformed autologous B cell stimulators as specificity controls.
The lower but present reactivity of TIL 1 131 with some
allogeneic tumors, as evidenced by GM-CSF secretion (but
not by TNF-a secretion), raises the possibility of recognition
of shared tumor antigens. The phenomenon of specific cytokine release in response to shared tumor antigens on
MHC-compatible cells has been shown in TIL from melanoma patients.26The implications of these findings require
- M h e r study.
The in vitro observation of antitumor immune reactivity
From www.bloodjournal.org by guest on December 3, 2014. For personal use only.
LYMPHOMA TUMOR INFILTRATING LYMPHOCYTES
in B-cell lymphoma TIL suggests a mechanism for the clinical regression of tumor after immunotherapy of select lymphoma patients. Nearly 20% of NHL patients treated with
IL-2 (with and without LAK cells) experienced significant
tumor regressions. Similar clinical responses have been
A greater underdocumented in patients with melan~ma.~’
standing of TIL-tumor interactions will be helpful in developing new immunotherapies for NHL patients.
‘2’
ACKNOWLEDGMENT
We thank Arnold Mixon, Kim Condron, and Julie McClanahan
for performing flow cytometry, and Etta Owens for her expert assistance in preparing this manuscript.
REFERENCES
I . Weber JS, Yang JC, Topalian SL, Schwartzentruber DJ,
White DE, Rosenberg SA: The use of interleukin-2 and lymphokine-activated killer cells for the treatment of patients with nonHodgkin’s lymphoma. J Clin Oncol 10:33, 1992
2. Weber J, Yang JC, Topalian SL, Schwartzentruber DJ, Parkinson D, White DE, Rosenberg SA: Interleukin-2 and LAK cells
for the treatment of patients with non-Hodgkin’s lymphoma. Proc
Am SOCClin Oncol 11:258, 1992 (abstr)
3. Ponvit-Ksiazek A, Christensson B, Lindemalm C, Mellstedt
H, Tribukait B, Biberfeld G, Biberfeld P: Characterization of malignant and non-neoplastic cell phenotypes in highly malignant nonHodgkin’s lymphomas. Int J Cancer 32:667, 1983
4. Lippman SM, Spier CM, Miller TP, Slymen DJ, Rybski JA,
Grogan TM: Tumor-infiltrating T-lymphocytes in B-cell diffuse
large cell lymphoma related to disease course. Mod Pathol 3:361,
I990
5. Jacob M-C, Piccinni M-P, Bonnefoix T, Sotto M-F, Couderc
P, Bensa J-C, Sotto J-J: T lymphocytes from invaded lymph nodes
in patients with B-cell-derived non-Hodgkin’s lymphoma: Reactivity toward the malignant clone. Blood 75: 1 154, 1990
6. Diaz JI, Edinger MG, Stoler MH, Tubbs RR: Activated T-cell
subsets in benign lymphoid hyperplasias and B-cell non-Hodgkin’s
lymphomas. Am J Pathol 139:503, 1991
7. Jacob M-C, Favre M, kmarc, Hadour F, Sotto M-F, Bonnefoix T, Sotto J-J, Bema J-C: CD45RA expression by CD4 T lymphocytes in tumors invaded by B-cell non-Hodgkin’s lymphoma
(NHL) or Hodgkin’s disease (HD). Am J Hematol 39:45, 1992
8. Bonnefoix T, Claret E, Piccinni M-P, Jacob M-C, Zheng X,
Sotto J-J: Impaired clonogenic potential ofCD25 positive T cells in
lymph nodes involved by B cell non-Hodgkin’s lymphomas. Immunol Lett 27:135, 1991
9. Luna-Fineman S, Lee JE, Wesley PK, Clayberger C, Krensky
AM: Human cytotoxic T-lymphocytes specificfor autologous follicular lymphoma recognizeimmunoglobulin in a major histocompatibility complex restricted fashion. Cancer 70:2 18 1, 1992
10. Braziel RM, Sussman E, Jaffe ES, Neckers LM, Cossman J:
Induction of immunoglobulin secretion in follicular non-Hodgkin’s lymphomas: Role of immunoregulatory T cells. Blood 66: 128,
1985
1 1. Topalian SL, Muul LM, Solomon D, Rosenberg SA: Expansion of human tumor infiltrating lymphocytes for use in immunotherapy trials. J Immunol Methods 102:127, 1987
12. Topalian SL, Solomon D, Rosenberg SA: Tumor specific
cytolysis by lymphocytes infiltrating human melanomas. J Immunol 142:3714, 1989
1211
13. Schwartzentruber DJ, Topalian SL, Mancini M, Rosenberg
SA: Specific release of granulocyte-macrophage colony-stimulating
factor, tumor necrosis factor-a, and IFN-y by human tumor-infiltrating lymphocytes after autologous tumor stimulation. J Immunol 146:3674, 1991
14. Schwartzentruber DJ, Solomon D, Rosenberg SA, Topalian
SL: Characterization of lymphocytes infiltrating human breast
cancer: Specific immune reactivity detected by measuring cytokine
secretion. J lmmunother l2:l, 1992
15. Hom SS, Rosenberg SA, Topalian SL: Specific immune recognition of autologous tumor by lymphocytes infiltrating colon carcinomas: Analysis by cytokine secretion. Cancer Immunol Immunother 36:1, 1993
16. Belldegrun A, Muul LM, Rosenberg SA: Interleukin-2 expanded tumor-infiltrating lymphocytes in human renal cell cancer:
Isolation, characterization, and antitumor activity. Cancer Res
48:206, 1988
17. Topalian SL, Kasid A, Rosenberg SA: Immunoselection of a
human melanoma resistant to specific lysis by autologous tumorinfiltrating lymphocytes. J Immunol 144:4487, 1990
18. Morecki S, Topalian SL, Myers WW, Okrongly D, Okarma
TB, Rosenberg SA: Separation and growth of human CD4+ and
CD8+ tumor infiltrating lymphocytes and peripheral blood mononuclear cells by direct positive panning on covalently attached
monoclonal antibody-coated flasks. J Biol Response Modif 9:463,
I990
19. The non-Hodgkin’slymphomapathologicclassificationproject: National Cancer Institute sponsored study of classification of
non-Hodgkin’s lymphomas: Summary and description of a working formulation for clinical usage. Cancer 49:2112, 1982
20. Bonnefoix T, Piccinni M-P, Jacob M-C, Pegourie B, Sotto
J-J: Limiting dilution analysis of the frequency of IL-2 responsive T
cells in lymph nodes involved by B-cell nomHodgkin’s lymphomas. Leuk Res 13:323, 1989
2 1. Knowles DM, Halper JP, Jakobiec FA: T-lymphocyte subpopulations in B-cell-derived non-Hodgkin’s lymphomas and
Hodgkin’s disease. Cancer 54544, 1984
22. Ruco LP, Stoppacciaro A, Valtieri M, Uccini S, Vitolo D,
Baroni CD: T-lymphocyte subsets in human lymph nodes: Relative
increase of OKT-8+ cells in neoplastic and reactive B-cell proliferation. Clin Immunol Immunopathol30:337, 1984
23. Rosenberg SA, Packard BS, Aebersold PM, Solomon D, Topalian SL, Toy ST, Simon P, Lotze MT, Yang JC, Seipp CA, Simpson C, Carter C, Bock s, Schwartzentruber D, Wei JP, White DE:
Use of tumor-infiltrating lymphocytes and interleukin-2 in the immunotherapy of patients with metastatic melanoma. N Engl J Med
319:1676, 1988
24. Oshimi K, Oshimi Y, Akutsu M, Takei Y, Saito H, Okada
M, Mizoguchi H: Cytotoxicity of interleukin 2-activated lymphocytes for leukemia and lymphoma cells. Blood 68:938, 1986
25. Chong AS-F, Scuderi P, Grimes WJ, Hersh EM: Tumor targets stimulate IL-2 activated killer cells to produce interferon-y and
tumor necrosis factor. J Immunol 142:2133, 1989
26. Hom SS, Schwartzentruber DJ, Rosenberg SA, Topalian SL:
Specificrelease ofcytokines by lymphocytes infiltrating human melanomas in response to shared melanoma antigens. J Immunother
13:18, 1993
27. Rosenberg SA, Lotze MT, Yang JC, Aebersold PM, Linehan
WM, Seipp C, White DE: Experience with the use of high-dose
interleukin-2 in the treatment of 652 cancer patients. Ann Surg
210:474, 1989