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
© Copyright 2024