Adoptive immunotherapy of CNS malignancies.

Although many patients with malignant brain tumors can be rendered free of bulk disease by current surgery and radiotherapy techniques, complete tumor eradication is extremely difficult to achieve, raising interest for T-cell adoptive immunotherapy. Conclusive experimental data generated by many investigators coupled with clinical experience have debunked many of the theoretical 'obstacles' to immunotherapy in the CNS. First, there does not appear to be a significant vascular barrier in brain tumors to prevent trafficking of systemically administered activated T cells. Moreover, T cells stimulated in the periphery by DC vaccination are able to mediate regression of established intracranial tumors. Second, brain tumor patients are able to mount an immune response against autologous tumor. Not surprisingly, patient factors such as tumor burden, corticosteroid use, advanced age, or recent chemotherapy can inhibit the immune response to tumor. Directing this type of therapy to patients without these factors may improve the likelihood of response. Third, therapeutic immune responses occurring within the CNS against tumors derived from CNS tissue have not been associated with clinical signs of autoimmune reactions against normal brain tissue. The general toxicity associated with the systemic adoptive transfer of ex vivo activated LN cells is very low. Toxicity of activated T cells delivered locally into the tumor resection cavity likewise is low and transient. Current phase II clinical trials of AI are in progress to determine the response rate for patients with newly diagnosed malignant gliomas. Future developments to characterize shared brain tumor antigens and develop more effective strategies for vaccination may lead to a more effective and broadly applicable therapy for CNS malignancies.

T-cell adoptive therapy of tumors: mechanisms of improved therapeutic performance.

The T cells of many cancer patients are naturally sensitized to tumor-associated antigens (Ag), or they can readily be sensitized with vaccine maneuvers. In melanoma patients, the adoptive transfer of such T cells can often be causally linked to the objective regression of established tumors. So far, few patients have shown sustained clinical benefit from such therapy, but preclinical mouse studies have now clearly delineated the hurdles that must be overcome to render T-cell-based antitumor therapy effective. Contrary to earlier expectations, it is now established that remarkably potent CD4+ and CD8+ pre-effector T cells are naturally sensitized even in mice bearing progressive, weakly immunogenic tumors. However, such T cells often display signal transduction impairments as a consequence of the tumor environment, which limit their acquisition of optimal effector function. Extracorporealization and culture of these tumor-sensitized T cells with appropriate activation stimuli not only restores normal signal transduction, but also confers resolute effector activity that can often sustain tumor rejection upon reinfusion. In mouse studies, the L-selectin(low) fraction of T cells in tumor-draining lymph nodes (TDLN) constitutes the potent pre-effector population and comprises both CD4+ and helper-independent CD8+ T cells. Appropriate in vitro activation confers an apparently unrestricted trafficking capacity to this fraction, and even the ability to proliferate within the tumor bed, leading to unprecedented tumor rejection at anatomic sites (e.g., subcutaneous and intracranial) that were historically refractory to such treatment. Such results underscore the surprising capacity of appropriately activated effector T cells to withstand the immunosuppressive, tolerogenic, and apoptotic influences of the typical tumor environment. Given the increasingly appreciated and critical communications between T cells and host Ag-presenting cells (APC), which cross-present tumor Ag, it is likely that dendritic cell-based vaccine maneuvers that promote sensitization of T1-committed L-selectin(low) antitumor T cells will play an increasingly important role in adoptive therapy strategies.

Potent effector function of tumor-sensitized L-selectin(low) T cells against subcutaneous tumors requires LFA-1 co-stimulation.

OBJECTIVE: Animal tumor models have demonstrated that adoptive transfer of tumor-draining lymph node (TDLN) T lymphocytes can cure established tumors in many anatomic sites. However, subcutaneous tumors are relatively refractory and have required maximally tolerated doses of cells. The goals of this study were to determine whether a subset of TDLN T lymphocytes varying in expression of the cell adhesion molecule L-selectin (CD62L) had augmented therapeutic efficacy and to determine the co-stimulatory requirements for trafficking and anti-tumor effector function. STUDY DESIGN: TDLNs were recovered from mice bearing progressive MCA 205 fibrosarcomas, and the T lymphocytes were segregated into CD62L(low) and CD62L(high) subsets and activated ex vivo with anti-CD3 mAb and IL-2. Mice bearing established subcutaneous MCA 205 tumors were treated with activated T cell subsets and in some experiments with additional mAb against cell adhesion molecules. RESULTS: Adoptive transfer of as few as 5 x 10(6) activated cells cured mice bearing 3-day subcutaneous MCA 205 tumors initiated with 6 x 10(6) cells, and the tumors demonstrated a dense infiltrate of CD62L(low) cells. In marked contrast, adoptive transfer of 10 times as many T cells derived from the reciprocal CD62L(high) compartment had no effect on tumor growth. The effector function of the CD62L(low) T cells was clearly dependent on co-stimulation through the cell adhesion molecule LFA-1, because anti-LFA-1 mAb completely abrogated the anti-tumor reactivity of the transferred cells against subcutaneous tumors and inhibited tumor infiltration. In contrast, blockade of ICAM-1, VLA-4, or VCAM-1 had no inhibitory effect on the anti-tumor function. CONCLUSION: These studies demonstrate the high therapeutic activity of the CD62L(low) subset of tumor-draining LN T cells against subcutaneous tumors, a relatively refractory site, and confirm the essential role of LFA-1 for effector T cell function. SIGNIFICANCE: Identification of the phenotype and requirements for effector function of T lymphocytes sensitized to tumor antigens has implications for clinical trials of adoptive immunotherapy for head and neck carcinoma using a similar approach.

T cell-mediated tumor rejection displays diverse dependence upon perforin and IFN-gamma mechanisms that cannot be predicted from in vitro T cell characteristics.

Experimental pulmonary metastases have been successfully treated by adoptive transfer of tumor-sensitized T cells from perforin knockout (KO) or Fas/APO-1 ligand(KO) mice, suggesting a prominent role for secretion of cytokines such as IFN-gamma. In the present study we confirmed that rejection of established methylcholanthrene-205 (MCA-205) pulmonary metastases displayed a requirement for T cell IFN-gamma expression. However, this requirement could be obviated by transferring larger numbers of tumor-sensitized IFN-gamma (KO) T cells or by immunosensitizing sublethal irradiation (500 rad) of the host before adoptive therapy. Extrapulmonary tumors (MCA-205 s.c. and intracranial) that required adjunct sublethal irradiation for treatment efficacy also displayed no requirement for host or T cell expression of IFN-gamma. Nonetheless, rejection of MCA-205 s.c. tumors and i.p. EL-4 tumors, but not MCA-205 pulmonary or intracranial tumors, displayed a significant requirement for T cell perforin expression (i.e., CTL participation). The capacity of T cells to lyse tumor targets and secrete IFN-gamma in vitro before adoptive transfer was nonpredictive of the roles of these activities in subsequent tumor rejection. Adoptive therapy studies employing KO mice are therefore indispensable for revealing a diversity of tumor rejection mechanisms that may lack in vitro correlation due to delays in their induction. Seemingly contradictory KO data from different studies are reconciled by the capacity of anti-tumor T cells to rely on alternative mechanisms when treated in larger numbers, the variable participation of CTL at different anatomic locations of tumor, and the apparent capacity of sublethal irradiation to provide a therapeutic alternative to host or T cell IFN-gamma production.

Helper-independent, L-selectinlow CD8+ T cells with broad anti-tumor efficacy are naturally sensitized during tumor progression.

We recently reported that the CD4(+) T cell subset with low L-selectin expression (CD62L(low)) in tumor-draining lymph nodes (TDLN) can be culture activated and adoptively transferred to eradicate established pulmonary and intracranial tumors in syngeneic mice, even without coadministration of IL-2. We have extended these studies to characterize the small subset of L-selectin(low) CD8(+) T cells naturally present in TDLN of mice bearing weakly immunogenic tumors. Isolated L-selectin(low) CD8(+) T cells displayed the functional phenotype of helper-independent T cells, and when adoptively transferred could consistently eradicate, like L-selectin(low) CD4(+) T cells, both established pulmonary and intracranial tumors without coadministration of exogenous IL-2. Whereas adoptively transferred L-selectin(low) CD4(+) T cells were more potent on a cell number basis for eradicating 3-day intracranial and s.c. tumors, L-selectin(low) CD8(+) T cells were more potent against advanced (10-day) pulmonary metastases. Although the presence of CD4(+) T cells enhanced generation of L-selectin(low) CD8(+) effector T cells, the latter could also be obtained from CD4 knockout mice or normal mice in vivo depleted of CD4(+) T cells before tumor sensitization. Culture-activated L-selectin(low) CD8(+) T cells did not lyse relevant tumor targets in vitro, but secreted IFN-gamma and GM-CSF when specifically stimulated with relevant tumor preparations. These data indicate that even without specific vaccine maneuvers, progressive tumor growth leads to independent sensitization of both CD4(+) and CD8(+) anti-tumor T cells in TDLN, phenotypically L-selectin(low) at the time of harvest, each of which requires only culture activation to unmask highly potent stand-alone effector function.

Systemic adoptive T-cell immunotherapy in recurrent and metastatic carcinoma of the head and neck: a phase 1 study.

OBJECTIVE: To evaluate the feasibility and toxic effects of systemic adoptive T-cell immunotherapy in patients with unresectable squamous cell carcinoma of the head and neck (SCCHN). DESIGN: Nonrandomized phase 1 clinical trial. SETTING: Academic tertiary care hospital. PATIENTS: Between April 1, 1996, and September 30, 1998, 17 patients with confirmed recurrent and metastatic SCC of the upper aerodigestive tract were enrolled. Two patients did not receive T cells because of poor vaccine response. Fifteen patients were successfully treated with T-cell immunotherapy. INTERVENTION: Patients were vaccinated on the thigh with irradiated autologous tumor cells admixed with granulocyte-macrophage colony-stimulating factor (GM-CSF) followed by 3 additional daily injections of GM-CSF at the vaccination site. Eight to 10 days later, tumor cell vaccine-draining inguinal lymph nodes were resected, and lymph node lymphocytes were activated with staphylococcal enterotoxin A and expanded in interleukin 2 in vitro. Resulting cultured cells were infused into patients peripherally on an outpatient basis. RESULTS: Toxic effects of infusion were limited to grade 2 reactions in 3 of 16 treatments. One patient required overnight hospitalization for fever and emesis. Median cell expansion was 37 times (range, 4-416 times), and median cell dose was 7.5 x 10(9) (range, 1.3 x 10(8) to 4.2 x 10(10)). Infused cells were predominantly CD3+ (>97%), being a mixture of CD4+ and CD8+ cells. Three patients demonstrated stabilization of previously progressive disease. Two patients experienced favorable clinical courses after adoptive T-cell transfer, including 1 patient with no evidence of disease 4 years after surgical resection of a vertebral body metastasis. CONCLUSIONS: Adoptive immunotherapy is a technically feasible and safe treatment with low toxicity and may demonstrate therapeutic activity in patients with unresectable SCCHN.

Cross-presentation of tumor antigens to effector T cells is sufficient to mediate effective immunotherapy of established intracranial tumors.

The systemic adoptive transfer of tumor-sensitized T cells, activated ex vivo, can eliminate established intracranial tumors. Regression of MHC class II negative MCA 205 fibrosarcomas occurs optimally following adoptive transfer of both CD4 and CD8 tumor-sensitized T cells, indicating an important function for tumor-infiltrating APC. Here, we demonstrate that during an effector response, indirect presentation of tumor Ags to transferred T cells is sufficient to mediate intracranial tumor regression. BALB/c --> CB6F1 (H-2bxd) bone marrow chimeras were challenged with the MCA 205 fibrosarcoma (H-2b). The tumor grew progressively in the H-2b-tolerant chimeras and stimulated an immune response in tumor-draining lymph nodes. Tumor-sensitized lymph node T cells were activated ex vivo with anti-CD3 and IL-2, then adoptively transferred to sublethally irradiated BALB/c or C57BL/6 recipients bearing established intracranial MCA 205 tumors. The transferred T cells eradicated MCA 205 tumors in BALB/c recipients and demonstrated tumor specificity, but had no therapeutic efficacy in the C57BL/6 recipients. These data establish that tumor-associated host cell constituents provide sufficient Ag presentation to drive effector T cell function in the complete absence of direct tumor recognition. This effector mechanism has an evident capacity to remain operative in circumstances of immune escape, where the tumor does not express the relevant MHC molecules, and may have importance even at times when direct CTL recognition also remains operative.

T cell adoptive immunotherapy of newly diagnosed gliomas.

Patients with newly diagnosed gliomas were treated with adoptive transfer of ex vivo activated T lymphocytes, derived from lymph nodes (LNs) draining autologous tumor vaccines, to determine the long-term toxicity of this treatment. Twelve consecutive patients were enrolled: 2 with grade II astrocytoma, 4 with anaplastic gliomas, and 6 with glioblastoma multiforme. Patients were injected intradermally with short-term cultured autologous irradiated tumor cells, admixed with granulocyte macrophage colony-stimulating factor, to stimulate draining LNs. The LN cells were activated with staphylococcal enterotoxin A for 48 h and then cultured in medium containing interleukin 2 for an additional 6-8 days and subsequently transferred i.v. to the patients. The number of cells obtained from the LNs ranged from 9 x 10(7) to 1.1 x 10(9), and the median cell proliferation was 41-fold. The dose of T cells infused ranged from 0.6 to 5.5 x 10(10) with a median of 1.1 x 10(10), the majority of which were CD 4+ (mean, 71%). The entire treatment was performed as outpatient therapy and was associated with a toxicity of grade 2 or less, consisting mainly of fever, nausea, and myalgias during the first 24 h. There were no indications of late adverse events from this treatment even among three patients with follow-up greater than 2 years post T cell transfer. Moreover, four patients demonstrated partial regression of residual tumor. This Phase I clinical trial of adoptive immunotherapy for patients with newly diagnosed malignant gliomas demonstrates feasibility, lack of long-term toxicity, and several objective clinical responses.

CD4+ T cells in adoptive immunotherapy and the indirect mechanism of tumor rejection.

Tumor-specific CD4+ effector T cells often play a decisive role in immunologic tumor rejection, in some cases without evident co-participation of CD8+ T cells. During such CD4+ T-cell-mediated rejection there is often no detectable direct contact between T cells and tumor cells. Optimally prepared, adoptively transferred CD4+ T cells can reject established tumors with great efficiency even when targeted tumor cells express no MHC Class II molecules, implying that recognition of tumor antigen (Ag) occurs via MHC Class II-expressing host antigen-presenting cells (APC) within the tumor. Because consequent rejection also excludes Ag-specific contact between CD4+ T cells and MHC Class IIneg tumor cells, the most critical CD4+ T-cell-mediated event is likely cytokine release, resulting in an accumulation and activation of accessory cells such as tumoricidal macrophages and lymphokine-activated killer cells. Although such an indirect rejection mechanism may appear antithetical to popular strategies centered on CD8+ cytotoxic T cell (CTL), current evidence suggest that even CD8+ T-cell-mediated recognition/rejection often bypasses direct tumor cell contact and is largely cytokine mediated. While CTL are likely to participate prominently in many models of tumor rejection, indirect mechanisms of recognition/rejection have the theoretical advantage of remaining operative even when individual tumor cells evade direct contact by down-regulating MHC and/or Ag expression.

Interleukin-13 receptor alpha chain: a novel tumor-associated transmembrane protein in primary explants of human malignant gliomas.

Human malignant glioma cell lines express high levels of interleukin-13 receptor (IL-13R). However, the subunit structure of this receptor in primary brain tumor cells is not known. Herein, we examined the subunit composition of IL-13R by analyzing the expression of four different putative subunits of IL-13R complex in 25 primary explants of malignant brain tumors. Reverse transcription-PCR (RT-PCR) of RNA from these tumor cells, normal astrocytes, and normal brain tissue showed that transcripts of IL-13R alpha chain were present in greater abundance in malignant glioma cells compared with normal astrocytes or normal brain tissues. The transcripts for two other chains (e.g., IL-13Ralpha' and IL-4Rbeta), on the other hand, yielded similar PCR positivity in brain tumors as well as in normal samples, whereas transcripts for gammac chain were absent in all brain tumor cells and normal tissues. The specificity of RT-PCR products for these genes was confirmed by oligo liquid hybridization analysis using a radiolabeled sequence-specific internal probe. Indirect immunofluorescence studies for different receptor chains confirmed the RT-PCR results and demonstrated a striking difference in the level of expression of IL-13Ralpha protein between normal astrocytes and malignant astrocytoma cells. These studies establish the IL-13Ralpha subunit as a novel tumor-specific protein that may be useful as a tumor marker, a target for cytotoxin/immunotoxin, or alternatively, a tumor-associated antigen for active, specific immunotherapy.

Infiltration of tumors by systemically transferred tumor-reactive T lymphocytes is required for antitumor efficacy.

The systemic transfer of ex vivo-activated tumor-sensitized T lymphocytes can mediate immunologically specific regression of established tumors. However, it has not been conclusively established whether the infiltration of systemically transferred T cells into metastases is required for their effector function. In this study, T cells from lymph nodes draining the murine fibrosarcoma MCA 205 cells were activated ex vivo with anti-CD3 monoclonal antibody and interleukin-2. During the final 24 h of culture, the T cells were treated with pertussis toxin (PTX) to inhibit signaling through G protein-coupled chemokine receptors required for diapedesis. Systemically transferred PTX-treated cells did not have any therapeutic efficacy against 3-day established pulmonary metastases. This lack of efficacy correlated with their failure to infiltrate the tumor parenchyma. However, PTX-treated cells responded to tumor antigen stimulation with IFN-gamma secretion in vitro. More importantly, PTX-treated effector T cells prevented tumor growth when they were admixed with tumor cells and inoculated s.c. These results demonstrate that systemically transferred tumor-reactive T lymphocytes need to infiltrate the tumor parenchyma through the endothelium to initiate tumor regression, but PTX-sensitive proteins are not required for either antigen recognition or effector functions.

T-cell adoptive immunotherapy of metastatic renal cell carcinoma.

OBJECTIVES: To determine the feasibility and toxicity of the adoptive transfer of ex vivo-activated T lymphocytes that have been sensitized to autologous tumor vaccine in vivo. METHODS: Twenty patients with extensive metastatic renal cell carcinoma received systemic adoptive immunotherapy with autologous T cells in the absence of conjunctional interleukin-2 (IL-2) administration. Patients were vaccinated intradermally with irradiated autologous tumor cells and granulocyte-macrophage colony-stimulating factor as an adjuvant to stimulate an immune response. Inguinal lymph nodes draining the vaccine site were surgically removed, and the cells were stimulated with staphylococcal enterotoxin A followed by expansion in 60 IU/mL IL-2, and in some cases additionally stimulated with anti-CD3 monoclonal antibody and IL-2, to obtain a large number of cells. RESULTS: The staphylococcal enterotoxin A/IL-2 activation induced vigorous proliferation of T cells (median expansion 26-fold) that were a mixture of CD4 and CD8 T lymphocytes. Activated cells were infused intravenously at doses ranging from 2x10(9) to 9.5x10(10). There was minimal toxicity consisting of grade 1 or 2 fever and nausea, and the entire treatment was delivered as outpatient therapy. One patient had a partial response, one had a mixed response, and 8 had stable disease lasting at least 5 months. CONCLUSIONS: Adoptive transfer of ex vivo-activated, tumor vaccine-primed lymph node cells is feasible and is associated with minimal toxicity when used alone. These results warrant further study in a Phase II trial.

Critical role of CD11a (LFA-1) in therapeutic efficacy of systemically transferred antitumor effector T cells.

The systemic adoptive transfer of activated T cells, derived from tumor-draining lymph nodes (LNs), mediates the regression of established tumors. In this study, the requirement of cell adhesion molecules, CD11a/CD18 (LFA-1), CD54 (ICAM-1), CD49d/CD29 (VLA-4), and CD106 (VCAM-1), for T cell infiltration into tumors and antitumor function was investigated. Administration of anti-CD11a mAb completely abrogated the efficacy of adoptive immunotherapy for both intracranial and pulmonary metastatic MCA 205 fibrosarcomas. In contrast, adoptive immunotherapy was effective in animals treated with anti-CD49d mAb, anti-CD106 mAb, anti-CD54 mAb, or in CD54 knockout recipients. Trafficking of transferred cells to the intracranial tumor was not affected by any of the mAb. However, the tumor-specific secretion of IFN-gamma by activated LN T cells was suppressed by anti-CD11a mAb or anti-CD54 mAb. To account for the different effects of CD11a and CD54 blockade in vivo, an additional CD11a/CD18 ligand, CD102 (ICAM-2), was demonstrated on tumor-associated macrophages but not on tumor cells. These results show that CD11a mediates a critical function in interactions between effector T cells, tumor cells, and host accessory cells in situ leading to tumor regression.

Allogeneic class I major histocompatibility complex gene transfer in murine neuroblastoma in vivo.

In vivo intratumoral gene transfer of allogeneic class I major histocompatibility complex (MHC) genes augments the immune response against weak tumor antigens. In this study, mice inoculated with the allogeneic MHC molecule (H-2Kb), had transduced-murine neuroblastoma C1300S3 cells showed prolonged survival relative to non-transduced or neo transduced tumors (p < 0.005). Interestingly, direct in vivo gene transfer of H-2Kb plasmid DNA complexed with HVJ-liposomes into S3 tumors was highly efficient, resulting in transduction of 8% of the interstitial cells within the tumor but rarely within tumor cells. Regression of established tumors and prolonged survival occurred in 50% of mice injected with H-2Kb, in contrast to no tumor regression in mice receiving control plasmid (p < 0.005). This study concludes that interstitial cells could serve as an important target of intratumoral gene transfer, and further that HVJ-liposome complexes could be a vehicle for in vivo gene transfer.

Ex vivo activation of tumor-draining lymph node T cells reverses defects in signal transduction molecules.

The adoptive transfer of tumor-draining lymph node (LN) T cells activated ex vivo with anti-CD3 and interleukin 2 (IL-2) mediates the regression of the poorly immunogenic murine melanoma D5. The efficacy of the activated LN cells is augmented when the sensitizing tumor is a genetically modified variant (designated D5G6) that secretes granulocyte/macrophage-colony-stimulating factor. In contrast to anti-CD3/IL-2-activated LN cells, adoptive transfer of freshly isolated tumor-draining LN T cells has no therapeutic activity. To determine whether the acquisition of antitumor function during ex vivo activation is associated with modifications in signal transduction capacity, the protein tyrosine kinases p56lck and p59fyn and proteins of the NF-kappaB family were analyzed in tumor-draining LN T cells. The levels of p56lck and p59fyn were lower in tumor-draining than in normal LN T cells and production of tyrosine-phosphorylated substrates was markedly depressed following anti-CD3 stimulation. After 5-day anti-CD3/IL-2 activation, levels of p56lck and p59fyn and protein tyrosine kinase activity increased. Interestingly, the levels of p56lck, p59fyn, and tyrosine kinase activity were higher in activated T cells derived from LN that drained D5G6 than they were in those from D5 tumors. In contrast, the cytoplasmic levels of c-Rel and Rel A were normal in freshly isolated tumor-draining LN, as was nuclear kappaB DNA-binding activity induced by anti-CD3 mAb or phorbol myristate acetate. Stimulation of activated LN cells with D5 tumor cells induced the nuclear translocation of NF-kappaB. These findings indicate that the recovery of proteins mediating signal transduction through the T cell receptor/CD3 complex in LN T cells activated ex vivo was associated with the acquisition of antitumor function.

Systemic T cell adoptive immunotherapy of malignant gliomas.

OBJECT: To determine the feasibility, toxicity, and potential therapeutic benefits of systemic adoptive immunotherapy, 10 patients with progressive primary or recurrent malignant glioma received this treatment. Adoptive immunotherapy, the transfer of immune T lymphocytes, is capable of mediating the regression of experimental brain tumors in animal models. In animal models, lymph nodes (LNs) that drain the tumor vaccine site are a rich source of tumor-immune T cells. METHODS: In this clinical study, patients were inoculated intradermally with irradiated autologous tumor cells and granulocyte macrophage-colony stimulating factor as an adjuvant. Cells from draining inguinal LNs, surgically resected 7 days after vaccination, were stimulated sequentially with staphylococcal enterotoxin A and anti-CD3, and a low dose of interleukin-2 (60 IU/ml) was used to expand the stimulated cells. The maximum cell proliferation was 350-fold over 10 days of culture. The activated cells were virtually all T cells consisting of various proportions of CD4 and CD8 cells. These cells were given to patients by intravenous infusion at doses ranging from 9 x 10(8) to 1.5 x 10(11). There were no Grade 3 or 4 toxicities associated with the treatment. Following T-cell transfer therapy, radiographic regression that lasted at least 6 months was demonstrated in two patients with recurrent tumors. One patient demonstrated stable disease that has lasted for more than 17 months. The remaining patients had progressive disease; however, four of the eight patients with recurrent tumor remain alive more than 1 year after surgery for recurrence. Three patients required intervention with corticosteroid agents or additional surgery approximately 1 month following cell transfer. CONCLUSIONS: These intriguing clinical observations warrant further trials to determine whether this approach can provide therapeutic benefits and improve survival.

Tumor immunology.

Malignant tumors express antigens that may stimulate and serve as targets for antitumor immunity. Virally induced tumors usually contain integrated proviral genomes in theircellulargenomes and often express viral genome-encoded proteins that may stimulate specific host immune responses. Antigens unique to individual tumors that stimulate specific rejection of transplanted tumors have been demonstrated only in experimental animals. Other tumor antigens that potentially can stimulate immune responses are shared by different tumors. These include products of mutated or rearranged oncogenes or tumor-suppressor genes. Tumors may also overexpress tissue differentiation antigens or embryonic antigens, which also have the potential to be recognized by the immune system. The recent identification of tumor antigens recognized by cytotoxic T cells opens up new possibilities for constructing chemically defined antigens for specific immunotherapy. Treatment of malignant tumors in humans by immunologic approaches, although theoretically attractive, has not yet succeeded on a large scale. Important progress in immunotherapy of cancer is emerging with several different treatment modalities.

Treatment of murine gliomas by adoptive transfer of ex vivo activated tumor-draining lymph node cells.

The adoptive transfer of tumor-reactive T lymphocytes has recently been demonstrated to be an effective means for mediating the regression of experimental intracranial fibrosarcomas. In this study, mice bearing syngeneic intracranial GL261 gliomas were cured by the combination of sublethal whole body irradiation followed by the intravenous transfer of tumor-draining lymph node (LN) T cells activated with anti-CD3 or staphylococcal enterotoxin C2 (SEC2). To further identify the functional effector T cel population in the adoptive immunotherapy, LN T cells were separated into two subsets, based on the level of expression of the cell adhesion molecule CD62L (L-selectin). As few as 5 x 10(5) CD62Llow cells could cure the majority of animals, whereas 2 x 10(6) CD62Lhigh cells were completely ineffective. Moreover, T cells isolated from advanced intracranial tumors were identified to be predominantly CD62Llow. In contrast, spleens contained a mixture of CD62L low and high cells similar to the transferred cell population. T cells in the glioma site were more actively proliferating than those isolated from the spleen. Mice cured of GL261 tumors demonstrated long-term immunologic memory by rejecting intracranial challenges of the original tumor but not an immunologically distinct tumor. Furthermore, despite infiltration of transferred cells into the intracranial tumors, cured mice did not exhibit any apparent neurologic abnormalities during treatment, prolonged follow-up, or after intracranial tumor rechallenge. This study demonstrates the effective treatment of an intracranial murine glioma by the systemic adoptive transfer of activated tumor-draining LN T cells and selective tumor infiltration by the therapeutically active CD62Llow T cells.

Isolation based on L-selectin expression of immune effector T cells derived from tumor-draining lymph nodes.

The ability to generate a large number of tumor-reactive T lymphocytes is the most critical requirement for adoptive immunotherapy. Our laboratory has previously demonstrated that cells from tumor-draining lymph nodes (LNs) are an excellent source of tumor-reactive T lymphocytes. After activation with anti-CD3, these cells readily proliferate in low concentrations of interleukin 2 and acquire effector functions. The adoptive transfer of these cells is capable of mediating the regression of tumors established in the lung as well as in the brain. Here, we analyzed several adhesion molecules on the tumor-draining LN T cells and separated them based on L-selectin expression. The homing receptor L-selectin mediates adhesion to the luminal surface of specialized high endothelial venules, thus regulating lymphocyte recirculation through peripheral LNs. In response to progressive tumor growth, a small population of draining LN T cells down-regulated L-selectin and increased the expression of CD44 and lymphocyte function-associated antigen 1. In adoptive immunotherapy, purified T cells with low L-selectin (L-selectin-) expression constituted all the in vivo antitumor reactivity, whereas isolated high L-selectin (L-selectin+) cells were ineffective. Furthermore, reverse transcription-PCR analysis revealed that L-selectin- cells expressed interleukin 2, IFN-gamma and tumor necrosis factor alpha mRNA upon in vitro stimulation with specific tumor cells. These results suggest that highly potent immune T cells can be isolated based on their pattern of adhesion molecule expression. The ability of the immune effector cells to transcribe cytokine genes when stimulated with tumor cells provides a basis for identifying similar cells for adoptive immunotherapy of cancer in humans.

Treatment of intracranial tumors by systemic transfer of superantigen-activated tumor-draining lymph node T cells.

Adoptive transfer of tumor-sensitized T lymphocytes has demonstrated therapeutic efficacy in animal tumor models and in some patients with melanoma and renal cell cancers. In experimental settings, T lymphocytes derived from lymph nodes (LNs) draining progressively growing tumors can be activated ex vivo to generate tumor-reactive lymphocytes with therapeutic efficacy. Despite the theoretical concern regarding inaccessibility of the central nervous system to systemically transferred T cells, our recent experiments demonstrated that anti-CD3-activated tumor-draining LN cells are capable of mediating the regression of established intracerebral tumors. In this study, several staphylococcal enterotoxins (SEs), including SEA, SEC2, and SEE, and exfoliating toxin, known to be superantigens, were tested for their ability to stimulate tumor-draining LN cells to acquire antitumor reactivity for the treatment of intracerebral tumors. SEs bind to the MHC class II molecule and provide an activating signal to T cells bearing particular T-cell receptor Vbeta chains. Tumor-draining LN cells activated with SEs demonstrated selective Vbeta T-cell expansion. In adoptive immunotherapy of intracranial (IC) tumors, SEA-and SEC2-activated cells had the highest efficacy, whereas SEE-activated cells were not therapeutic. Despite the antigen independence of SE activation, the T cells retained immunological specificity for the tumor, which provided the initial in vivo sensitization of the LN. During the ex vivo stimulation with superantigens, both CD4+ and CD8+ T cells proliferated, and both subsets were required to mediate regression of IC tumors. In contrast to the adoptive immunotherapy of visceral tumors, the systemic administration of exogenous interleukin 2 failed to support the antitumor reactivity in mice depleted of CD4 cells, and, in fact, it inhibited the therapeutic efficacy. Furthermore, mice cured of intracerebral tumors by the adoptive transfer of T cells were resistant to an IC tumor rechallenge. However, in contrast to the immunological specificity demonstrated during the primary adoptive transfer, cured mice were able to reject challenge with several immunologically distinct fibrosarcomas but not a melanoma. These results indicate that superantigen-activated LN cells can circulate to and interact with intracerebral tumors mediating tumor regression in an immunologically specific manner. Although conditions that optimize the treatment of intracerebral tumors appear to be different from those for visceral tumors, analysis of T-cell receptor Vbeta expression among cells activated with several superantigens does not reflect a preferential usage of Vbeta gene segments in the immune response to autochthonous tumors.