Adoptive transfer of T-cell precursors enhances T-cell reconstitution after allogeneic hematopoietic stem cell transplantation

Immunoincompetence after allogeneic hematopoietic stem cell transplantation (HSCT) affects in particular the T-cell lineage and is associated with an increased risk for infections, graft failure and malignant relapse. To generate large numbers of T-cell precursors for adoptive therapy, we cultured mouse hematopoietic stem cells (HSCs) in vitro on OP9 mouse stromal cells expressing the Notch-1 ligand Delta-like-1 (OP9-DL1). We infused these cells, together with T-cell-depleted mouse bone marrow or purified HSCs, into lethally irradiated allogeneic recipients and determined their effect on T-cell reconstitution after transplantation. Recipients of OP9-DL1-derived T-cell precursors showed increased thymic cellularity and substantially improved donor T-cell chimerism (versus recipients of bone marrow or HSCs only). OP9-DL1-derived T-cell precursors gave rise to host-tolerant CD4+ and CD8+ populations with normal T-cell antigen receptor repertoires, cytokine secretion and proliferative responses to antigen. Administration of OP9-DL1-derived T-cell precursors increased resistance to infection with Listeria monocytogenes and mediated significant graft-versus-tumor (GVT) activity but not graft-versus-host disease (GVHD). We conclude that the adoptive transfer of OP9-DL1-derived T-cell precursors markedly enhances T-cell reconstitution after transplantation, resulting in GVT activity without GVHD.     

Allogeneic tumor-specific cytotoxic T lymphocytes

Summary We report the development of cytotoxic T lymphocytes specific for an allogeneic brain tumor in a rat model. DA strain cytotoxic T cell precursors stimulated by an allogeneic tumor (9L gliosarcoma) from the Fischer rat could generate a population of cytotoxic T lymphocytes that lysed the allogeneic 9L tumor but failed to lyse other targets, including Fischer concanavalin-A(ConA)-stimulated lymphoid blast targets. DA T cells depleted of reactivity to the Fischer haplotype (DA-f) retained reactivity to the 9L tumor, demonstrating that T cell precursors with specificity for normal Fischer alloantigens were not required for the generation of a response to the 9L Fischer tumor. The preferential lysis of the tumor target did not simply reflect a higher density of Fischer target antigens on the tumor than that found on normal Fischer ConA blast targets. First, the relative densities of class I antigen on the 9L tumor and normal Fischer ConA blasts were comparable. Second, cytotoxic T cells could not be generated from DA-f precursors when Fischer ConA blasts were used as stimulators. If DA-f T cells were simply responding to the higher density of Fischer antigen found on 9L tumor, it would have been expected that the ConA blasts expressing comparable levels of antigen to that found on the tumor would have generated cytotoxicity for both the 9L and ConA targets. We conclude that the cytotoxic T cells are specific for a determinant expressed only by the tumor. Such tumor-specific cytotoxic T cells could be useful in vivo for adoptive immunotherapy of brain tumors.     

GMP production of anti-tumor cytotoxic T-cell lines for adoptive T-cell therapy in patients with solid neoplasia

BACKGROUND: The adoptive transfer of ex vivo-induced tumor-specific T-cell lines provides a promising approach for cancer immunotherapy. We have demonstrated previously the feasibility of inducing in vitro long-term anti-tumor cytotoxic T-cell (CTL) lines directed against different types of solid tumors derived from both autologous and allogeneic PBMC. We have now investigated the possibility of producing large amounts of autologous anti-tumor CTL, in compliance with good manufacturing practices, for in vivo use. METHODS: Four patients with advanced solid tumors (two sarcoma, one renal cell cancer and one ovarian cancer), who had received several lines of anticancer therapy, were enrolled. For anti-tumor CTL induction, patient-derived CD8-enriched PBMC were stimulated with DC pulsed with apoptotic autologous tumor cells (TC) as the source of tumor Ag. CTL were then restimulated in the presence of TC and expanded in an Ag-independent way. RESULTS: Large amounts of anti-tumor CTL (range 14-20 x 10(9)), which displayed high levels of cytotoxic activity against autologous TC, were obtained in all patients by means of two-three rounds of tumor-specific stimulation and two rounds of Ag-independent expansion, even when a very low number of viable TC was available. More than 90% of effector cells were CD3(+) CD8(+) T cells, while CD4(+) T lymphocytes and/or NK cells were less than 10%. DISCUSSION: Our results demonstrate the feasibility of obtaining large quantities of anti-tumor specific CTL suitable for adoptive immunotherapy approaches.     

RNA interference targeting programmed death receptor-1 improves immune functions of tumor-specific T cells

Adoptive cell transfer (ACT), either using rapidly expanded tumor infiltrating lymphocytes or T-cell receptor transduced peripheral blood lymphocytes, can be considered one of the most promising approaches in cancer immunotherapy. ACT results in the repopulation of the host with high frequencies of tumor-specific T cells; however, optimal function of these cells within the tumor micro-environment is required to reach long-term tumor clearance. We and others have shown that ongoing anti-tumor immune responses can be impaired by the expression of ligands, such as PD-L1 (B7-H1) on tumor cells. Such inhibitory molecules can affect T cells at the effector phase via their receptor PD-1. PD-L1/PD-1 interaction has indeed been shown crucial in inducing T-cell anergy and maintaining peripheral tolerance. In order to maximize anti-tumor responses, antibodies that target the PD-1/PD-L1 axis are currently in phase I/II trials. Alternatively, a more refined approach could be the selective targeting of PD-1 in tumor-specific T cells to obtain long-term resistance against PD-1-mediated inhibition. We addressed whether this goal could be achieved by means of retroviral siRNA delivery. Effective siRNA sequences resulting in the reduction of surface PD-1 expression led to improved murine as well as human T-cell immune functions in response to PD-L1 expressing melanoma cells. These data suggest that blockade of PD-1-mediated T-cell inhibition through siRNA forms a promising approach to achieve long-lasting enhancement of tumor-specific T-cell function in adoptive T-cell therapy protocols.     

The anti-tumor efficacy of lymphokine-activated killer cells and recombinant interleukin 2 in vivo

We showed previously that adoptive immunotherapy with the combination of LAK cells and recombinant IL 2 (RIL 2) can markedly reduce pulmonary micrometastases from multiple sarcomas established 3 days after the i.v. injection of syngeneic tumor cells in C57BL/6 mice. In this report, we analyzed the factors required for successful therapy. Titration analysis in vivo revealed an inverse relationship between the number of pulmonary metastases remaining after treatment and both the number of LAK cells and the amount of RIL 2 administered. Fresh or unstimulated splenocytes had no anti-tumor effect; a 2- to 3-day incubation of splenocytes in RIL 2 was required. LAK cells generated from allogeneic DBA (H-2d) splenocytes were as effective in vivo as syngeneic, C57BL/6 (H-2b) LAK cells. The anti-metastatic capacity of LAK cells was significantly reduced or eliminated when irradiated with 3000 rad before adoptive transfer. The combined therapy of LAK cells plus RIL 2 was shown to be highly effective in mice immunosuppressed by 500 rad total body irradiation and in treating macrometastases established in the lung 10 days after the i.v. injection of sarcoma cells. Further, reduction of both micrometastases and macrometastases could also be achieved by RIL 2 alone when administered at higher levels than were required with LAK cells. The value of LAK cell transfer and of IL 2 administration for the treatment of tumors established at other sites is currently under investigation.     

Recipient leukocyte infusion enhances the local and systemic graft-versus-neuroblastoma effect of allogeneic bone marrow transplantation in mice

Allogeneic hematopoietic stem cell transplantation and donor leukocyte infusion (DLI) may hold potential as a novel form of immunotherapy for high-risk neuroblastoma. DLI, however, carries the risk of graft-versus-host disease (GvHD). Recipient leukocyte infusion (RLI) induces graft-versus-leukemia responses without GvHD in mice and is currently being explored clinically. Here, we demonstrate that both DLI and RLI, when given to mixed C57BL/6 → A/J radiation chimeras carrying subcutaneous Neuro2A neuroblastoma implants, can slow the local growth of such tumors. DLI provoked full donor chimerism and GvHD; RLI produced graft rejection but left mice healthy. Flow cytometric studies showed that the chimerism of intratumoral leukocytes paralleled the systemic chimerism. This was associated with increased CD8/CD4 ratios, CD8⁺ T-cell IFN-γ expression and NK-cell Granzyme B expression within the tumor, following both DLI and RLI. The clinically safe anti-tumor effect of RLI was further enhanced by adoptively transferred naïve recipient-type NK cells. In models of intravenous Neuro2A tumor challenge, allogeneic chimeras showed superior overall survival over syngeneic chimeras. Bioluminescence imaging in allogeneic chimeras challenged with luciferase-transduced Neuro2A cells showed both DLI and RLI to prolong metastasis-free survival. This is the first experimental evidence that RLI can safely produce a local and systemic anti-tumor effect against a solid tumor. Our data indicate that RLI may provide combined T-cell and NK-cell reactivity effectively targeting Neuro2A neuroblastoma.     

Alteration of allospecific t-cell receptors after differentiation from prethymic precursor cells in semiallogeneic environments

Murine bone marrow cells (strain A) have been allowed to differentiate in vivo in syngeneic (A) or semiallogeneic hosts (A × B) to produce mature splenic T lymphocytes. After stimulation of these cells with irradiated allogeneic (C) spleen cells in tissue cultures, the cytotoxic T-cell blasts (CTL) were purified by velocity sedimentation and used to immunize (A × C) F₁ hybrid mice, to produce antisera recognizing the receptor structure (for C) on the relevant A cytotoxic cells (and their precursors). Using these sera we have been able to show that the T-cell receptor for alloantigen C on strain A cytotoxic precursor lymphocytes (CTLp) seems to differ according to the host environment in which those T cells differentiate from immature bone marrow precursors.     

Lysis of fresh human solid tumor cells by autologous lymphocytes activated in vitro by allosensitization

Summary We have previously demonstrated that cancer patients' peripheral blood lymphocytes (PBL) allosensitized against single or pool normal donor PBL are capable of lysing fresh autologous tumor cells in a 4-h ⁵¹Cr-release assay. In this report, we present further investigations into this phenomenon. These alloactivated killer cells (A-AK cells) lysed autologous and allogeneic tumors and allogeneic but not autologous PBL. Furthermore, cold target inhibition studies demonstrated that autologous and allogeneic tumors were lysed by the same effector cells. Multiple metastases from the same patient were equivalently lysed by these A-AK cells. The presence of adherent cells and proliferation of the precursors were necessary to generate A-AK cells, although the effector cell itself was radioresistant and nonadherent. The effects of allosensitization were enhanced by the addition of lectin-free interleukin-2 preparations to the in vitro culture by partial depletion of adherent cells prior to sensitization. The A-AK effector cell was OKT3+, OKT8+, OKT4–, OKM1– and could be generated by just 3 days of allosensitization. The precursors for A-AK cells could be separated from NK cells on percoll gradients and lysis could be generated from thoracic duct lymphocytes, a population devoid of NK cells. The phenotype of the majority of the precursor cells was OKT3+, OKT4–. These allocatived PBL could be expanded in crude or lectin-free interleukin-2 without loss of cytotoxicity for fresh autologous tumor cells. Activated T cells represent a population of non-NK cells with broad lytic specificity for fresh tumor cells. Such cells may be of value in the adoptive immunotherapy of human solid tumors and may play a role in immune surveillance.     

Sclerodermatous chronic graft-versus-host disease induced by host T-cell-mediated autoimmunity

Despite a long-standing hypothesis that chronic graft-versus-host disease (cGVHD) is an autoimmune disorder, most mouse models of cGVHD have been developed on the assumption that donor T cells are essential for its development. Here we show that cGVHD may be caused by autoreactive host T cells in mice that have been lethally irradiated and grafted with T-cell-depleted allogeneic bone marrow cells. In this chimera, host T cells derived from radioresistant intrathymic T-cell precursors caused dermal fibrosis and periportal inflammation, without the requirement for donor T cells. The lack of host DCs within the thymus after high-dose irradiation allowed autoreactive host T cells to escape thymic negative selection. Moreover, the homeostatic expansion of these T cells may augment their autoreactivity. These findings indicate that host T-cell-mediated cGVHD is an autoimmune process that occurs following the grafting of T-cell-depleted BM cells into hosts with functioning thymuses. We propose, based on the present data, that host T-cell-dependent autoimmunity is a potential mechanism by which cGVHD is induced.     

GM-CSF-secreting cancer immunotherapies: preclinical analysis of the mechanism of action

Granulocyte-macrophage colony-stimulating factor (GM-CSF)-secreting tumor cell immunotherapies have demonstrated long-lasting, and specific anti-tumor immune responses in animal models. The studies reported here specifically evaluate two aspects of the immune response generated by such immunotherapies: the persistence of irradiated tumor cells at the immunization site, and the breadth of the immune response elicited to tumor associated antigens (TAA) derived from the immunotherapy. To further define the mechanism of GM-CSF-secreting cancer immunotherapies, immunohistochemistry studies were performed using the B16F10 melanoma tumor model. In contrast to previous reports, our data revealed that the irradiated tumor cells persisted and secreted high levels of GM-CSF at the injection site for more than 21 days. Furthermore, dense infiltrates of dendritic cells were observed only in mice treated with GM-CSF-secreting B16F10 cells, and not in mice treated with unmodified B16F10 cells with or without concurrent injection of rGM-CSF. In addition, histological studies also revealed enhanced neutrophil and CD4+ T cell infiltration, as well as the presence of apoptotic cells, at the injection site of mice treated with GM-CSF-secreting tumor cells. To evaluate the scope of the immune response generated by GM-CSF-secreting cancer immunotherapies, several related B16 melanoma tumor cell subclones that exist as a result of genetic drift in the original cell line were used to challenge mice previously immunized with GM-CSF-secreting B16F10 cells. These studies revealed that GM-CSF-secreting cancer immunotherapies elicit T cell responses that effectively control growth of related but antigenically distinct tumors. Taken together, these studies provide important new insights into the mechanism of action of this promising novel cancer immunotherapy. Electronic supplementary material The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Dendritic cell maturation and subsequent enhanced T-cell stimulation induced with the novel synthetic immune response modifier R-848.

Agents that enhance dendritic cell maturation can enhance T-cell activation and therefore may improve the efficiency of vaccines or improve cellular immunotherapy. Previously, we demonstrated that a novel low-molecular-weight synthetic immune response modifier, R-848, induces IL-12 and IFN-alpha secretion from monocytes and macrophages. Here we report that R-848 induces the maturation of human monocyte-derived dendritic cells. Characteristic of dendritic cell maturation, R-848 treatment induces cell surface expression of CD83 and increases cell surface expression of CD80, CD86, CD40, and HLA-DR. Additionally, R-848 induces cytokine (IL-6, IL-12, TNF-alpha, IFN-alpha) and chemokine (IL-8, MIP-1alpha, MCP-1) secretion from dendritic cells. Most significantly, R-848 enhances dendritic cell antigen presenting function, as measured by increased T-cell proliferation and T-cell cytokine secretion in both allogeneic and autologous T-cell systems. Consequently, low-molecular-weight synthetic molecules such as R-848 and its derivatives may be useful as vaccine adjuvants or as ex vivo stimulators of dendritic cells for cellular immunotherapy.     

Dendritic cells crosspresent antigens from live B16 cells more efficiently than from apoptotic cells and protect from melanoma in a therapeutic model

Dendritic cells (DC) are able to elicit anti-tumoral CD8(+) T cell responses by cross-presenting exogenous antigens in association with major histocompatibility complex (MHC) class I molecules. Therefore they are crucial actors in cell-based cancer immunotherapy. Although apoptotic cells are usually considered to be the best source of antigens, live cells are also able to provide antigens for cross-presentation by DC. We have recently shown that prophylactic immunotherapy by DC after capture of antigens from live B16 melanoma cells induced strong CD8(+) T-cell responses and protection against a lethal tumor challenge in vivo in C57Bl/6 mice. Here, we showed that DC cross-presenting antigens from live B16 cells can also inhibit melanoma lung dissemination in a therapeutic protocol in mice. DC were first incubated with live tumor cells for antigen uptake and processing, then purified and irradiated for safety prior to injection. This treatment induced stronger tumor-specific CD8(+) T-cell responses than treatment by DC cross-presenting antigens from apoptotic cells. Apoptotic B16 cells induced more IL-10 secretion by DC than live B16 cells. They underwent strong native antigen degradation and led to the expression of fewer MHC class I/epitope complexes on the surface of DC than live cells. Therefore, the possibility to use live cells as sources of tumor antigens must be taken into account to improve the efficiency of cancer immunotherapy.     

Endogenous production of prothymocyte differentiating activity by phytohemagglutinin-stimulated T-cell-depleted human marrow

The PHA responsiveness of marrow T-cell precursors remains a matter of controversy. We have investigated the capacity of human marrow to proliferate under phytohemagglutinin (PHA) stimulation following extensive removal of mature T cells by complement-dependent cytotoxicity with MBG6 and RFT8 monoclonal antibodies. PHA-induced thymidine uptake by marrow cells occurred with a peak on Days 6-8 of incubation instead of Day 3 for PBL. This peak was observed 48 hr earlier in the presence of PHA-stimulated T-depleted marrow cell supernatants. These supernatants can also promote the growth of mature T-cell colonies from MBG6-, RFT8-, T11-, T3- marrow. However, full colony development requires exogenous interleukin 2 (IL-2). IL-2 could be detected in marrow supernatants but only at very low levels and beyond Days 3 and 4. In contrast Days 1-6 marrow supernatants were equally effective in promoting MBG6-RFT8- marrow cell responsiveness to PHA. We conclude that marrow T-cell precursors are not PHA responsive and that PHA induces the production by marrow non-T cells of a prothymocyte-differentiating activity (PTDA); PTDA can differentiate marrow T-cell progenitors into PHA-responsive T cells; following activation by PHA, these cells undergo limited proliferation induced by IL-2 endogenously released from de novo differentiated T cells. It is suggested that this mechanism may account for extrathymic differentiation of the T-cell lineage in heavily irradiated marrow transplantation recipients.     

Characterization of a CD4-positive T-cell line derived from an athymic (nu/nu) mouse.

We have isolated a Thy-1+, CD3+, CD4+ T-cell line from the spleen of a 12-week-old nu/nu (nude) BALB/c mouse. The cell line is clonal, and it expresses an alpha beta T-cell antigen receptor. Upon activation, these cells secrete IL-2 but not IL-4, putting them in the Th1 category. The cells can be triggered to proliferate and secrete lymphokines in the presence of irradiated syngeneic or allogeneic splenic feeder cells that express a variety of MHC haplotypes. This response is MHC class II-specific, because it can be blocked by either anti-Ia or anti-CD4 antibodies. From the response pattern of this T-cell line, we conclude that it recognizes a common determinant on class II MHC antigens. This nude mouse T-lymphocyte presumably has not undergone thymic selection. Therefore its unique specificity may reflect both the bias of T-cell antigen receptor genes for encoding receptors that recognize MHC molecules and the requirement for functional thymic epithelial cells for the efficient education of a self-MHC-restricted repertoire.     

Engineered drug-resistant immunocompetent cells enhance tumor cell killing during a chemotherapy challenge

Establishment of immunocompetent cell mediated anti-tumor immunity is often mitigated by the myelosuppressive effects during administration of chemotherapy. We hypothesized that protecting these immune cells from drug induced toxicities may allow for the combined administration of immunotherapy and chemotherapy. Using a SIV-based lentiviral gene transfer system we delivered the drug-resistant variant P140KMGMT into the immunocompetent cell lines NK-92 and TALL-104, and the myelogenous leukemia cell line, K562, which is a target for both NK-92 and TALL-104 cells. Genetically engineered immunocompetent cells developed significant resistance to temozolomide compared to non-modified cells, and genetic modification of these cells did not affect their ability to kill K562 cells. We then evaluated the effectiveness of drug-resistant immunocompetent cell mediated killing of tumor cells in the presence and absence of chemotherapy. During a chemotherapy challenge the cytotoxic activity of non-modified immunocompetent cells was dramatically impaired. However, when combined with chemotherapy, genetically-modified immune cells retained their cytotoxic activities and efficiently killed non-modified target cells. These results show that engineering immunocompetent cells to withstand chemotherapy challenges can enhance tumor cell killing when chemotherapy is applied in conjunction with cell-based immunotherapy.     

Repertoire enhancement with adoptively transferred female lymphocytes controls the growth of pre-implanted murine prostate cancer

Background

In prostate cancer, genes encoding androgen-regulated, Y-chromosome-encoded, and tissue-specific antigens may all be overexpressed. In the adult male host, however, most high affinity T cells targeting these potential tumor rejection antigens will be removed during negative selection. In contrast, the female mature T-cell repertoire should contain abundant precursors capable of recognizing these classes of prostate cancer antigens and mediating effective anti-tumor immune responses.

Methodology/Principal Findings

We find that syngeneic TRAMP-C2 prostatic adenocarcinoma cells are spontaneously rejected in female hosts. Adoptive transfer of naïve female lymphocytes to irradiated male hosts bearing pre-implanted TRAMP-C2 tumor cells slows tumor growth and mediates tumor rejection in some animals. The success of this adoptive transfer was dependent on the transfer of female CD4 T cells and independent of the presence of CD25-expressing regulatory T cells in the transferred lymphocytes. We identify in female CD4 T cells stimulated with TRAMP-C2 a dominant MHC II-restricted response to the Y-chromosome antigen DBY. Furthermore, CD8 T cell responses in female lymphocytes to the immunodominant MHC I-restricted antigen SPAS-1 are markedly increased compared to male mice. Finally, we find no exacerbation of graft-versus-host disease in either syngeneic or minor-antigen mismatched allogeneic lymphocyte adoptive transfer models by using female into male versus male into male cells.

Conclusions/Significance

This study shows that adoptively transferred female lymphocytes, particularly CD4 T cells, can control the outgrowth of pre-implanted prostatic adenocarcinoma cells. This approach does not significantly worsen graft-versus-host responses suggesting it may be viable in the clinic. Further, enhancing the available immune repertoire with female-derived T cells may provide an excellent pool of prostate cancer reactive T cells for further augmentation by combination with either vaccination or immune regulatory blockade strategies.     

β-defensin 2 as an adjuvant promotes anti-melanoma immune responses and inhibits the growth of implanted murine melanoma in vivo

β-defensin 2 is a small antimicrobial peptide of the innate immune system and has been thought to regulate anti-tumor immunity. However, little is known on whether β-defensin 2 could modulate melanoma-specific NK and T cell responses. In this study, we first cloned the murine β-defensin 2 gene by RT-PCR and generated the β-defensin 2 stably expressing B16 cells (B16-mBD2). Subsequently, we evaluated whether vaccination with irradiated B16-mBD2 could modulate the growth of implanted B16 cells and determined the potential mechanisms underlying the action of B16-mBD2 vaccine in modulating the growth of B16 tumors in C57BL/6. We found that vaccination with irradiated B16-mBD2, but not with control B16-p or parental B16, inhibited the development and progression of B16 tumors, and prolonged the survival of tumor-bearing mice. However, vaccination with irradiated B16-mBD2 failed to inhibit the development of B16 tumors in the CD4(+)- or CD8(+)-depleted recipients. Furthermore, vaccination with irradiated B16-mBD2 stimulated strong NK activity and promoted potent B16-specific CTL responses, accompanied by augmenting IFN-γ and IL-12, but not IL-4, responses in the recipient mice. Moreover, vaccination with irradiated B16-mBD2 promoted the infiltration of CD8(+) and CD4(+) T, NK cells and macrophages in the tumor tissues. These data suggest β-defensin 2 may act as a positive regulator, promoting anti-tumor NK and T cell responses in vivo. Therefore, β-defensin 2 may be used for the development of immunotherapy for the intervention of melanoma.     

Various uses of BCG and allogeneic acute leukemia cells to treat patients with acute myelogenous leukemia

Summary Ninety-six remission patients with acute myelogenous leukemia have been treated with various forms of immunotherapy and chemotherapy in three distinct studies and the clinical outcome of these patients has been reported. In the first study 22 patients were maintained on chemotherapy alone and 28 patients were given the same chemotherapy and additional immunotherapy consisting of BCG and irradiated allogeneic AML cells given at separate sites weekly. It was found that there was a significant increase in survival time of the patients who received immunotherapy (median 510 days) compared with the chemotherapy alone patients (270 days). The p value for this was 0.03. The reason for this prolongation of survival was mainly due to a markedly increased survival time of immunotherapy patients after they relapsed when compared with the chemotherapy patients (165 days compared with 75 days median, p equal to 0.0005). In the second sequential study 24 patients were given immunotherapy alone consisting of irradiated allogeneic AML cells and BCG given at separate sites, and this was compared with unirradiated allogeneic cells and BCG given to 22 patients. There was no difference in the remission length or survival between these two groups. In the third study 13 patients received irradiated cells and BCG as in Study 1 and a further 11 patients received the same immunotherapy but also received a mixture of cells and BCG given during the first three months. There was no difference in the remission and survival of these two groups. The significance of these results is discussed.