T Cells Engrafted with a Recombinant Anti-CD30 Receptor Target Autologous CD30+ Cutaneous Lymphoma Cells

Home , Carcinoembryonic antigen

Gene Therapy (2001) 8, 891–895  2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00 www.nature.com/gt BRIEF COMMUNICATION T cells engrafted with a recombinant anti-CD30 receptor target autologous CD30+ cutaneous lymphoma cells

A Hombach1, JM Muche2, M Gerken1, S Gellrich2, C Heuser1, C Pohl3, W Sterry2 and H Abken1 1Department I of Internal Medicine, Laboratory Tumor Genetics, University of Cologne; 2St Elisabeth-Krankenhaus Ko¨ln-Hohenlind, Cologne; and 3Department of Dermatology and Allergy, University Hospital Charite´, Humboldt University Berlin, Berlin, Germany

T cells can be directed to antigen-specific, MHC-inde- Peripheral blood-derived T cells from the lymphoma patient pendent target cell lysis by grafting with a recombinant were retrovirally engrafted with a recombinant anti-CD30- receptor with antibody-like specificity. Here, we asked scFv-␥ receptor. Upon cocultivation with autologous CD30+ whether T cells from the peripheral blood of a patient with lymphoma cells, grafted T cells increase IFN-␥ secretion and cutaneous T cell lymphoma can be recruited for an immune lyse specifically lymphoma cells with high efficiency, even at response against autologous tumor cells. Lymphoma cells an effector to target cell ratio of as low as 1:20. Our data with a CD3− CD4+ CD30+ phenotype and clonal TCR-V␤7 demonstrate that the recombinant anti-CD30-␥ receptor rearrangement were isolated from a cutaneous lesion. The overcomes T cell tolerance for tumor cells and directs T cells lymphoma lesion additionally harbored CD3+ CD25+ acti- specifically against autologous lymphoma cells. Gene Ther- vated normal T cells despite ongoing tumor progression. apy (2001) 8, 891–895.

Keywords: cutaneous T cell lymphoma; adoptive immunotherapy; chimeric receptor, CD30

T cells can be equipped with pre-defined binding speci- CTLs, that were successfully generated by peptid vacci- ficity by grafting with a recombinant receptor that har- nation, did not induce significant tumor regression.7 On bors an antibody-derived binding domain and an intra- the other hand, the cellular anergy seems to be restricted cellular signalling domain for cellular activation. The to tumor-specific CTLs because viral antigens induce a chimeric receptor strategy combines the universal normal cellular immune response in these patients.6 properties of antibodies for specific antigen recognition From this situation arises the question of whether T with highly efficient cellular activation upon antigen-spe- cells from the peripheral blood of a tumor patient whose cific receptor crosslinking. To test this strategy for use immune functions are not yet substantially affected by in the cellular immunotherapy of malignant diseases, a the malignant disease can be recruited for an immune number of in vitro and in vivo systems have been applied. response against the autologous tumor cells. To address Most of these systems have in common that cells of estab- this issue we engrafted peripheral blood T cells of a lished tumor cell lines were utilized as a model system patient suffering from a CD30+ cutaneous T cell lym- (Refs 1–3, for review see Ref. 4). On the other hand, it phoma with a recombinant receptor molecule that has was recently demonstrated that primary, in situ growing binding specificity for the CD30 antigen. The antigen- tumors and transplanted tumor cell lines differ substan- binding site of the receptor is composed of an anti-CD30 tially in their ability to induce a cellular antitumor single chain antibody fragment (HRS3-scFv) that was response and that tumor cell lines may poorly reflect the fused to the transmembrane and intracellular domains of clinical situation.5 Particularily, in situ growing autolog- the high affinity IgE receptor signalling chain (Fc⑀RI␥). ous tumor cells are partially or completely tolerized by The CD30 target-antigen is a 120 kDa glycoprotein of the the immune system and an efficient antitumor response TNF/nerve growth factor receptor family8 that is pre- is frequently suppressed during tumor progression. dominantly expressed on Hodgkin- and some non-Hodg- Accordingly, in the case of melanoma substantial popu- kin lymphomas and on virus-infected lymphocytes. A lations of tumor-specific CD8+ cytotoxic T cells (CTLs) small subpopulation of normal activated lymphocytes in could be detected despite tumor progression. Tumor- lymphatic tissues, predominantly around B cell follicles specific CTLs, however, were demonstrated to be func- and the edge of germinal centers in lymph nodes and tionally anergized6 and, moreover, melanoma-specific spleen, express CD30 whereas resting lymphocytes do not.9 It was suggested that CD30 may be involved in the negative selection of thymocytes.10 However, CD30-lack- ing mice do not show an altered primary or secondary B Correspondence: H Abken, Lab. Tumor Genetics, Dept I of Internal Medi- 10 cine, University of Cologne, Josef-Stelzmann-Str. 9, D-50931 Cologne, cell and T cell response, respectively. Due to its highly Germany restricted expression pattern on malignant and virus- Received 21 December 2000; accepted 19 March 2001 infected cells and on a small proportion of activated T Autologous T cell targeting of cutaneous lymphoma A Hombach et al 892 and B cells, the CD30 antigen is a suitable target for an the CD3− CD4+ CD30+ phenotype could not be detected antibody-based immunotherapy. While the binding in the patient’s peripheral blood either by flow cytometry specificity and the functional activity of the recombinant or by clone-specific PCR (data not shown). We conclude anti-CD30-␥ immunoreceptor was previously shown in that the lymphoma lesion harbors one predominant cell detail,11 we demonstrate here that receptor-grafted per- clone specified by clonal V␤7–J␤2.7 TCR rearrangement ipheral blood T cells of a patient with a cutaneous T cell and a CD3− CD4+ CD30+ phenotype suggesting that this lymphoma are specifically activated against autologous cell population represents the malignant cell clone. More- tumor cells mediating a MHC-independent, CD30- over, due to the lack of CD3 expression, the lymphoma specific cellular immune response that is highly effective cells differ phenotypically from infiltrating normal CD3+ even at very low effector to target cell ratios. T lymphocytes. To determine whether the latter represent Suspension cells were isolated from a biopsy of a activated or resting T cells we monitored the number of cutaneous T cell lymphoma lesion. Flow cytometric IL-2 receptor (CD25) expressing CD3+ cells present within analysis revealed a predominant cell population with a the lymphoma lesion. Strikingly, about 23% of normal CD3− CD4+ CD30+ phenotype in addition to low numbers CD3+ T cells derived from the lymphoma biopsy express of normal cells (Figure 1a, b). These CD3− CD4+ CD30+ the IL-2 receptor (Figure 1c, d) indicating the presence of cells clonally express the TCR V␤7 chain as demonstrated a high number of activated T cells. by clone-specific V␤7-PCR (data not shown). Although We asked whether anti-CD30-␥ immunoreceptor- present in high numbers in the biopsy material, cells of grafted T cells from the lymphoma patient’s peripheral blood are specifically activated against autologous CD30+ lymphoma cells. Therefore, the patient’s peripheral blood lymphocytes and for comparison six healthy donors were engrafted by retroviral gene transfer with the recombinant anti-CD30 receptor HRS3-scFv-␥ and the anti-CEA receptor BW431/26-scFv-Fc-␥, respectively. Recombinant receptor-expressing T cells were identified by flow cytometry utilizing receptor-specific anti-idiotypic and anti- CD3 mAbs. Retroviral transduction of the lymphoma patient’s peripheral blood T cells resulted in about 5% of T cells expressing the recombinant receptor whereas the transduction rates of healthy donors were in the range of 4% to 40% (Table 1). To test whether anti-CD30-␥ receptor-expressing T cells are specifically activated by CD30+ lymphoma cells we cocultivated anti-CEA- and anti- CD30-␥ receptor-grafted allogeneic and autologous T cells, respectively, with lymphoma cells for 72 h at an initial cell ratio of 1:1. The cell culture supernatants were harvested and analyzed for the presence of IFN-␥ by ELISA. Cocultivation of both HRS3-scFv-␥ receptor- grafted allogeneic (Figure 2a) and autologous (Figure 2b) T cells with CD30+ lymphoma cells resulted in increased IFN-␥ secretion indicating CD30-specific cellular activation of grafted T cells. In contrast, neither cocultivation Figure 1 Immunophenotyping of cells isolated from the cutaneous lym- of nontransduced or anti-CEA-␥ receptor-grafted allo- phoma lesion and the peripheral blood. Cells of a cutaneous T cell lym- geneic or autologous T cells with CD30+ lymphoma cells phoma were derived from a 37-year-old male, who developed a tumor nod- ule on the left lower leg. The diagnosis of a primary cutaneous CD30- nor incubation of receptor-grafted T cells without lym- positive anaplastic large cell lymphoma was established according to the phoma cells resulted in increased IFN-␥ secretion. It is EORTC classification.12 Immunohistochemistry revealed large anaplastic noteworthy that cells obtained from the lymphoma tumor cells with a CD30+ CD4+ CD3− and V␤7+ phenotype (data not lesion, that represent predominantly CD3− CD4+ CD30+ shown). Staging by CT scan excluded an extracutaneous manifestation of lymphoma cells, secreted substantial amounts of IFN-␥ the lymphoma. X-ray therapy was administered as initial therapy (total (Figure 2). dose 40 Gy). After a total dose of 20 Gy, perilesional multiple tumor nod- We next analyzed whether anti-CD30-␥ receptor- ules with up to 2 cm in diameter occurred. Biopsies were taken from these + nodules and tumor cells were prepared as previously described.13 Primary grafted autologous T cells deplete CD30 lymphoma cells cells of the cutaneous T cell lymphoma biopsy were cultured in RPMI in a cocultivation assay. Therefore, receptor-grafted and 1640 medium supplemented with 10% (v/v) FCS, 100 U/ml penicillin, nontransduced peripheral blood T cells from the lym- ␮ ␮ + 100 g/ml streptomycin (all Life Technologies), 200 g/ml Meropenem phoma patient were co-incubated with autologous CD30 (Gru¨nenthal, Stolberg, Germany), and 400 U/ml IL-2 (Endogen). Cells lymphoma cells in initial cell ratios of 1:1 and 0.2:1, from the lymphoma lesion were viable for about 14 days in culture and, before addition of exogeneous IL-2, characterized by two-color immuno- respectively. After 72 h the cells were harvested and fluorescence utilizing FITC-conjugated anti-CD4, anti-CD25 or anti- stained with an anti-CD3 mAb in combination with anti- CD30 mAbs and phycoerythrin-(PE)-conjugated anti-CD4 or anti-CD3 CD4 or anti-CD3 mAbs and propidium iodide. The cells mAbs, respectively (all Dako, Hamburg, Germany). The cells were ana- were analyzed by flow cytometry for the presence of lyzed in a FACScan cytofluorometer (Becton Dickinson, Mountain View, viable cells with a CD3− CD30+ and CD3− CD4+ pheno- CA, USA). Dead cells were stained by propidium iodide (Sigma, Deisen- type. Figure 3 shows exemplarily a dot blot analysis of hofen, Germany) and excluded from analysis. (a–c) Two-color immunoreceptor-grafted T cells that were co-incubated with fluorescence analysis of cells from the lymphoma lesion. (d) Histogramm + analysis of CD3+ T cells from the lymphoma lesion. CD3+ T cells from initially equal numbers of autologous CD30 lymphoma (c) were gated and the percentage of CD25+ cells was determined. cells; Figure 4 summarizes the data of these experiments.

Gene Therapy Autologous T cell targeting of cutaneous lymphoma A Hombach et al 893 Table 1 Efﬁciency of recombinant receptor expression in peripheral blood T cells from various donors

BW431/26-scFv-Fc-␥a HRS3-scFv-␥a

Healthy donorb 1 ND 40.1d 2 10.1 13.3 3 12.9 17.0 4 8.4 6.0 5 32 20.9 6c 4.3 7.5 Lymphoma patientb 3.41 5.12 a,bThe generation of the CEA-specific BW431/26-scFv-Fc-␥ and the CD30-specific HRS3-scFv-␥ receptor was recently described.11,14 For retroviral transduction of peripheral blood T cells, the receptor ␥ ␥ expression cassette (HRS3-scFv- ; BW431/26-scFv-Fc- ) was ␥ inserted into the retroviral vector pSTITCH.15,16 To generate GALV- Figure 2 IFN- release of receptor-grafted peripheral blood T cells upon pseudotyped retrovirus, 293T cells were cotransfected with the co-incubation with lymphoma cells. Receptor-grafted allogeneic (a) and ␮ autologous (b) T cells (1 × 105 cells per well) were co-incubated with recombinant pSTITCH vector DNA (6 g) and with DNA of the + ␮ CD30 lymphoma cells (1 × 105 cells per well) in RPMI 1640 medium, retroviral helper plasmids pHIT and pCOLT (each 6 g DNA/ml), ␮ encoding the MMLV gag and pol genes and the GALV envelope 10% FCS in a final volume of 200 l per well for 72 h. For control reasons, gene, respectively.13 This procedure results in transient production grafted T cells and lymphoma cells, respectively, were individually incubated (1 × 105 cells in 200 ␮l RPMI 1640 medium, 10% FCS per well). of high titers of infectious retrovirus. Peripheral blood lymphocytes ␮ were isolated by density centrifugation and cultured for 48 h in Cell culture supernatants (100 l per well) were harvested and analyzed for IFN-␥ by ELISA utilizing a solid phase anti-human IFN-␥ antibody RPMI 1640 medium supplemented with 10% (v/v) FCS (both Life ␮ ␥ Technologies, Paisley, UK), 400 U/ml IL-2 (Endogen, Woburn, MA, (1 g/ml) for capture and a biotinylated anti-human IFN- antibody (both Endogen) for detection. The assay was done in triplicate; data are USA) and 100 ng/ml anti-CD3 mAb (OKT3; ATCC CRL 8001) that ± ␥ was purified from hybridoma supernatants by binding to protein presented as the mean s.e.m. The line indicates the amount of IFN- A. For retroviral transduction, cells of six randomly selected donors secreted by lymphoma cells. and of a patient with a cutaneous T cell lymphoma, respectively, were harvested, resuspended in medium with 400 U/ml IL-2, and co-cultured for 48 h with transfected 293T cells. cRecombinant receptor-grafted allogeneic T cells of this donor were cocultivated with primary cutaneous T lymphoma cells. dExpression of recombinant receptors was monitored utilizing anti- idiotypic mAbs as follows: receptor-grafted T cells were incubated with 10 ␮g/ml of the anti-idiotypic mAb 9G10 (anti-HRS3-scFv-␥), the anti-idiotypic mAb BW2064/36 (anti-BW431/26-scFv-Fc-␥) (both IgG1),17,18 and 2.5 ␮g/ml of the anti-CD3 mAb OKT3 (IgG2a). Bound antibodies were subsequently detected by a FITC-conju- ′ ′ gated F(ab )2 anti-mouse IgG1 antibody and a PE-conjugated F(ab )2 anti-mouse IgG2a antibody, respectively (each 1 ␮g/ml) (both Southern Biotechnology, Palo Alto, CA, USA).

CD3− CD4+ CD30+ lymphoma cells are specifically elimin- ated in the presence of autologous HRS3-scFv-␥ receptor- grafted T cells but not in the presence of nontransduced or anti-CEA-␥ receptor-grafted autologous T cells. Taking into account the transduction rate of about 5%, the cell ratio of 1:1 actually represents a ratio of about one anti- Figure 3 Two-color immunofluorescence analysis (FACS) of lymphoma ␥ cells upon incubation with receptor-grafted autologous T cells from the CD30- receptor-expressing T cell to 20 lymphoma cells. peripheral blood. Peripheral T cells from the lymphoma patient were Strikingly, this effector:target (E:T) ratio resulted in spe- grafted with the recombinant anti-CD30-␥ and an anti-CEA-␥ receptor cific lymphoma cell depletion of more than 95% of the for control, respectively (cf. Table 1). Recombinant receptor-grafted and control within 72 h. Moreover, even a cell ratio of 0.2:1 nontransduced peripheral blood T cells (1 × 105 cells per well) were co- + representing an E:T ratio of 1:100 still resulted in incubated in 96-well round-bottom microtiter plates for 72 h with CD30 × 5 depletion of lymphoma cells of about 30–40% of control autologous T lymphoma cells (1 10 cells per well) in a final volume of 200 ␮l per well RPMI 1640 + 10% FCS. The assay was performed in (Figure 4). The cytolysis is specific because nontrans- ␥ triplicate. The cells from three wells were pooled and stained for flow cyto- duced and anti-CEA- receptor-transduced autologous T metric analysis simultaneously with either a FITC-conjugated anti-CD4 cells had no substantial effect. In addition to depletion of or anti-CD30 mAb and a PE-conjugated anti-CD3 mAb, respectively (all CD3− CD30+ lymphoma cells, CD30-specific targeting from Dako). Dead cells were stained with propidium iodide (Sigma) and also resulted in depletion of normal CD3+ CD30+ T cells excluded from analysis. (a, d) Anti-CD30 receptor grafted T cells (cf. Table 1). (b, e) Anti-CEA receptor grafted T cells (cf. Table 1). (c, f) Non- (Figure 3d), further indicating the CD30 specificity of tar- − + − transduced T cells. The percent of viable CD3 CD4 (a–c) and CD3 get cell elimination. Since lymphoma cells are charac- CD30+ (d, e) cells is indicated. terized by a CD3− CD4+ and a CD3− CD30+ phenotype and cells with both phenotypes are depleted upon cocultivation with anti-CD30-␥ receptor-grafted T cells, depletion of CD30+ cells is unlikely to result from down-

Gene Therapy Autologous T cell targeting of cutaneous lymphoma A Hombach et al 894 an anti-CD30-␥ receptor resulted in highly efficient, MHC-independent targeting of the autologous CD30+ lymphoma cells. The recruitment of receptor-grafted peripheral blood T cells may substitute anergized tumor- specific CTLs that seem to persist in tumor patients.6 Moreover, the anti-CD30-␥ receptor-mediated cellular immune response against autologous tumor cells is highly efficient even at E:T cell ratios of as low as 1:20, implying serial cytolysis of autologous target cells without evidence of substantial T cell exhaustion. Evidence for serial cytolysis of target cells by receptor-grafted T cells was also reported for an allogeneic system.19 It is essential for therapeutic efficacy that grafted T cells are capable of repeated cytolysis of target cells since the indi- vidual transduction rates differ remarkably dependent on the T cell donor. Antigen-driven T cell proliferation will, at least partially, compensate for low transduction rates. On the other hand, we recently demonstrated that anti- CD30-␥ receptor grafted T cells can be substantially enriched after retroviral transduction utilizing an anti- idiotypic mAb with specificity for the receptor’s antigen binding domain.15 Enriched receptor-grafted T cells can be further expanded by short-term culture thereby sig- nificantly increasing the number of effector cells for clinical application. Based on the data presented in this report we conclude that anti-CD30-␥ receptor-grafted peripheral blood T cells can be recruited to direct a highly effective and specific antitumor response against autologous CD30+ lymphoma cells. This strategy may be promising Figure 4 Summary of FACS analyses: specific cytolysis of lymphoma cells to overcome T cell tolerance for autologous tumor cells. by autologous, receptor-grafted T cells. Peripheral T cells of the lymphoma patient were grafted with the recombinant anti-CD30-␥ and an anti-CEA- ␥ receptor for control, respectively (cf. Table 1). Recombinant receptor- Acknowledgements × 5 grafted and nontransduced peripheral blood T cells (1 10 cells per well We would like to thank Dr RL Bolhuis (Department of and 0.2 × 105 cells per well, respectively) were coincubated for 72 h in 96- well round-bottom microtiter plates with CD30+ autologous T lymphoma Clinical and Tumor Immunology, Daniel den Hoed Can- cells (1 × 105 cells per well) in a final volume of 200 ␮l per well RPMI cer Center, Rotterdam, The Netherlands) for providing 1640 medium + 10% FCS. The assay was performed twice in triplicate. us with the retroviral expression vector pSTITCH. This The cells of three wells were pooled, simultaneously stained with anti- work is supported by grants from the Deutsche For- CD3 and either anti-CD4 or anti-CD30 mAbs and analyzed by two-color − schungsgemeinschaft (DFG), Bonn, Germany, through flow cytometry as described above. The number of viable cells with a CD3 SFB 502, the Deutsche Krebshilfe, Bonn, through 70–2235- CD4+ CD30+ phenotype, indicative for the cutaneous lymphoma cells, was determined and expressed in percent of control, ie the number of viable Ab1 and 70–2091-Lu1, and the Ko¨ln Fortune Program. tumor cells present upon co-incubation for 72 h with nontransduced T cells. Each staining experiment was done twice and the results of anti- References CD30/CD3 and anti-CD4/CD3 staining, respectively, were summarized. Data are expressed as the mean ± s.e.m. 1 Eshhar Z, Waks T, Gross G, Schindler DG. Specific activation and targeting of cytotoxic lymphocytes through chimeric single chains consisting of antibody-binding domains and the gamma regulation of CD30 antigen expression upon binding to or zeta subunits of the immunoglobulin and T-cell receptors. the recombinant anti-CD30 immunoreceptor. Proc Natl Acad Sci USA 1993; 90: 720–724. Taken together, our data indicate that (1) CD25+ acti- 2 Hwu P et al. In vivo antitumor activity of T cells redirected with vated T cells are present within the cutaneous lymphoma chimeric antibody/T cell receptor genes. Cancer Res 1995; 55: 3369–3373. lesion (cf. Figure 1) and (2) cells derived from the lym- ␥ 3 Altenschmidt U, Klundt E, Groner B. Adoptive transfer of in phoma lesion secrete the Th1 cytokine IFN- . However, vitro-targeted, activated T lymphocytes results in total tumor clinical progression of the tumor occurred despite the regression. J Immunol 1997; 159: 5509–5515. presence of high numbers of activated T cells and a 4 Abken H, Hombach A, Reinhold U, Ferrone S. Can combined T tumor environment that seems to be favorable for a cellu- cell- and antibody-based immunotherapy outsmart tumor cells? lar immune response. Moreover, in vitro cocultivation of Immunol Today 1998; 19: 2–5. lymphoma cells with normal autologous lymphocytes 5 Willimsky G, Blankenstein T. Interleukin-7/B7.1-encoding from the peripheral blood that were pre-activated by adenoviruses induce rejection of transplanted but not nontrans- treatment with an anti-CD3 mAb (OKT3) plus IL-2 did planted tumors. Cancer Res 2000; 60: 685–692. not induce efficient lymphoma cell depletion (Figures 3, 6 Lee PP et al. Characterization of circulating T cells specific for tumor associated antigens in melanoma patients. Nat Med 1999; 4). In this experimental situation, IL-2-mediated cellular 5: 677–685. activation of normal blood lymphocytes is not sufficient 7 Lee KH et al. Increased vaccine-specific T cell frequency after to revert tolerance against autologous lymphoma cells peptide-based vaccination correlates with increased suscepti- and to induce a cellular antitumor response. In contrast, bility to in vitro stimulation but does not lead to tumor engraftment of autologous peripheral blood T cells with regression. J Immunol 1999; 163: 6292–6300.

Gene Therapy Autologous T cell targeting of cutaneous lymphoma A Hombach et al 895 8Du¨ rkop H et al. Molecular cloning and expression of a new 15 Hombach A et al. T cell activation by recombinant Fc⑀RI ␥-chain member of the nerve growth factor receptor family that is immune receptors: an extracellular spacer domain impairs anti- characteristic for Hodgkin’s disease. Cell 1992; 68: 421–427. gen dependent T cell activation but not antigen-recognition. 9 Ellis TM et al. CD30 is a signal transducing molecule that defines Gene Therapy 2000; 7: 1067–1075. a subset of human activated CD45RO+ T cells. J Immunol 1993; 16 Weijtens ME, Willemsen RA, Hart EH, Bolhuis RL. A retroviral 151: 2380–2389. vector system ‘STITCH’ in combination with an optimized sin- 10 Amakawa R et al. Impaired negative selection of T cells in Hodg- gle chain antibody chimeric receptor gene structure allows kin’s disease antigen CD30-deficient mice. Cell 1996; 84: 551–562. efficient gene transduction and expression in human T lympho- 11 Hombach A et al. A chimeric T cell receptor recognizing the cytes. Gene Therapy 1998; 5: 1195–1203. CD30 antigen converts cytotoxic T cells to specificity for Hodg- 17 Pohl C et al. CD30 specific AB1-AB2-AB3 internal image anti- kin and Reed Stemberg cells. Cancer Res 1998; 58: 1116–1119. body network: potential use as anti-idiotype vaccine against 12 Willemze R et al. EORTC classification for primary cutaneous Hodgkin’s lymphoma. Int J Cancer 1993; 54: 418–425. lymphomas: a proposal from the Cutaneous Lymphoma Study 18 Bruynk A, Seemann G, Bosslet K. Charcterization of a Group of the European Organization for Research and Treat- humanized bispecific monoclonal antibody for cancer therapy. ment of Cancer. Blood 1997; 90: 354–371. Br J Cancer 1993; 67: 436–440. 13 Trefzer U et al. Hybrid cell vaccination for cancer immune ther- 19 Weijtens ME et al. Single chain Ig/gamma gene-redirected apy: first clinical trial with metastatic melanoma. Int J Cancer human T lymphocytes produce cytokines, specifically lyse 2000; 85: 618–626. tumor cells, and recycle lytic capacity. J Immunol 1996; 157: 14 Hombach A et al. A chimeric receptor that selectively targets 836–843. membrane-bound carcinoembryonic antigen (mCEA) in the presence of soluble CEA. Gene Therapy 1999; 6: 300–304.

Gene Therapy