Synthesis of pyrophosphate-containing compounds that stimulate Vgamma2Vdelta2 T cells: application to cancer immunotherapy.

Human Vgamma2Vdelta2 T cells recognize nonpeptide antigens, such as isoprenoid pyrophosphomonoester intermediates, alkylamine compounds, and bisphosphonate drugs, as well as some tumor cells. Although attempts have been made to derive novel cancer immunotherapies based on the discovery of these unconventional antigens, effective therapies remain to be developed. Here, we synthesized a series of pyrophosphate-containing compounds and examined the chemical requirements for the recognition of pyrophosphomonoester antigens by gammadelta T cells. The structural analysis clearly demonstrated that a proximal methylene moiety plays a crucial role in the stimulatory activity of the antigens. For optimal gammadelta T cell proliferation, we find that the use of human serum albumin was preferred and that pyrophosphomonoesters were superior to nitrogen-containing bisphosphonate compounds. Using these techniques, we have successfully expanded gammadelta T cells from healthy donors as well as from cancer patients using one of the most active compounds, 2-methyl-3-butenyl-1-pyrophosphate (2M3B1PP). The resulting expanded gammadelta T cells exhibited potent, cytotoxic activity against a wide variety of tumor cell lines. Even gammadelta T cells from a patient with advanced liver carcinoma efficiently responded to 2M3B1PP and exhibited strong cytotoxic activity against tumor cells. The pretreatment of tumor cells with nonpeptide antigens was essential for efficient cytotoxicity via TCR-gammadelta. The present study suggests a novel strategy for cancer immunotherapy using synthetic small pyrophosphate-containing compounds and nitrogen-containing bisphosphonates.

CD95-mediated tumor recognition by CD4+ effector cells in a murine mammary model.

The authors examined cellular mechanisms involved in anti-tumor reactivity induced by the murine MT-9G1 mammary tumor line, which was transduced to secrete granulocyte macrophage-colony-stimulating factor (GM-CSF). Compared with the parental MT-901 tumor, MT-9G1 subcutaneous tumors elicited an influx of CD4+ cells and dendritic cells. Secondary in vitro activation of tumor-draining lymph node cells with anti-CD3 and interleukin-2 resulted in effector cells that can mediate regression of established pulmonary metastases after adoptive transfer. In vivo depletion of T-cell subsets showed that tumor regression required CD4+ tumor-draining lymph node cells rather than CD8+ cells. The activated CD4+ cells expressed CD95L and mediated lysis of CD95+ MT-901 tumor cells, which were major histocompatibility complex class II negative. The CD4+ cells also released GM-CSF in response to tumor stimulation. A Fas fusion protein inhibited tumor lysis and GM-CSF release by the CD4+ cells. These studies document an alternate pathway by which CD4+ immune cells may recognize major histocompatibility complex class II-deficient tumors in which CD95L-bearing T cells induced an anti-tumor response mediated via CD95L:CD95.

Enhanced adjuvant effect of granulocyte-macrophage colony-stimulating factor plus interleukin-12 compared with either alone in vaccine-induced tumor immunity.

Using the poorly immunogenic D5 murine melanoma, we examined the adjuvant effect of granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin-12 (IL-12) secretion by gene-modified tumor cells inoculated as a vaccine to prime tumor-draining lymph nodes (TDLNs). D5 transfectants that secreted IL-12 or GM-CSF alone were compared with a double transfectant that secreted equivalent amounts of both cytokines. TDLN cells harvested 9-10 days after subcutaneous tumor inoculation were cultured sequentially in anti-CD3 and IL-2 and assessed for antitumor reactivity against wild-type D5 tumor. The double transfectant-induced TDLN effector cells had greater cytotoxicity in a long-term assay than TDLN cells primed by single transfectants. In adoptive immunotherapy, the TDLN cells primed by the double transfectant were significantly better at mediating the regression of established tumors compared with the TDLN cells elicited by the single transfectants. Both IL-12 and GM-CSF had adjuvant effects in promoting tumor-reactive TDLN cells, but the combination was better than either alone. These observations suggest that the immunomodulation roles of IL-12 and GM-CSF are different and complementary.

Interferon-gamma-inducing factor elicits antitumor immunity in association with interferon-gamma production.

Interferon-gamma-inducing factor (IGIF) is a novel cytokine that stimulates T-cell proliferation, augments natural killer (NK) cell lytic activity, and induces interferon-gamma (IFN-gamma) production in established type 1 T-helper (Th1) cells in the presence of anti-CD3 antibody. The in vitro induction of IFN-gamma by recombinant murine IGIF in these cells was more potent than that induced by murine interleukin-12 (IL-12) and occurred apparently independent of murine IL-12. Here we report that subcutaneous injection into mice of tumor cells transfected with murine IGIF complementary DNA (cDNA) resulted in > or = 10-fold increase of mitogen-stimulated IFN-gamma production in cultured splenocytes. In addition, IGIF-transfected Renca and K1735 tumor cells can be rejected in vivo. The IGIF antitumor effect was abrogated in mice that were sublethally irradiated or depleted of both CD4+ and CD8+ T cells but not in mice depleted of either subpopulation alone. The antitumor effect mediated by IGIF appears to be dependent on IFN-gamma production, because in vivo neutralization of IFN-gamma was accompanied by growth of IGIF-transfected tumors in 100% of the animals. Taken together, our results show that murine IGIF can elicit T-cell-dependent antitumor immunity associated with IFN-gamma induction.

Cytokines as an adjuvant to tumor vaccines: efficacy of local methods of delivery.

BACKGROUND: We examined alternative methods of delivering cytokines as an adjunct for priming lymph node (LN) cells draining sites of vaccine inoculation for the purpose of generating immune cells for adoptive immunotherapy. METHODS: Using syngeneic murine tumors we examined the ability of IL-2, IL-4, or GM-CSF delivered locally to a site of tumor inoculum to induce antitumor reactive draining LN cells. Mice were inoculated subcutaneously with tumor cells transduced to secrete cytokine; tumor cells admixed with fibroblasts transduced to secrete cytokine; or intralesional inoculation of cytokine in established tumor to induce sensitized LN cells capable of mediating tumor regression in adoptive transfer. RESULTS: Both IL-4 and GM-CSF cytokines were effective in enhancing the antitumor reactivity of vaccine-primed LN cells compared to IL-2, which was ineffective. The local delivery of GM-CSF by autocrine or paracrine secretion of genetically engineered cells, as well as direct intratumoral delivery was capable of upregulating LN sensitization compared to systemic administration, which did not. CONCLUSIONS: The local delivery of GM-CSF as an adjuvant for tumor vaccination can be accomplished by various methods, including direct injection, which avoids the need for gene transfer.

Reduced efficacy of allogeneic versus syngeneic fibroblasts modified to secrete cytokines as a tumor vaccine adjuvant.

We examined the relative efficacy of allogeneic versus syngeneic fibroblasts admixed with tumor cells as a vaccine to induce antitumor T-cell reactivity. Allogeneic (3T3) or syngeneic (BLK) fibroblasts transfected to secrete equivalent amounts of GM-CSF were admixed with either D5 melanoma or MCA 207 sarcoma and inoculated s.c. into the flanks of C57BL/6 mice. Vaccine-primed lymph node (LN) cells were examined for in vivo antitumor reactivity in an adoptive transfer model. At fibroblast: tumor cell ratios of < or=1, allogeneic and syngeneic granulocyte macrophage colony-stimulating factor-secreting fibroblasts enhanced T-cell reactivity to tumor cells. However, at ratios of 2.4, the adjuvant effect induced by granulocyte macrophage colony-stimulating factor was not evident. Instead, we observed increased alloreactivity of primed LN cells against 3T3 targets as assessed by cytotoxicity and cytokine release assays, which was not observed with syngeneic fibroblasts. Moreover, with increasing numbers of allogeneic fibroblasts, there was a skewing of the T-cell Vbeta repertoire. These latter cells responded to tumor stimulation with the release of greater amounts of interleukin 10, which may account for the diminished antitumor reactivity observed in vivo. Allogeneic fibroblasts transduced to secrete interleukin 2 or IFN-gamma also induced diminished tumor reactivity of primed LN cells. Syngeneic fibroblasts are superior to allogeneic fibroblasts as vehicles to deliver cytokines in tumor vaccines.

Type 1 versus type 2 cytokine release by Vbeta T cell subpopulations determines in vivo antitumor reactivity: IL-10 mediates a suppressive role.

We have previously reported that CD8+ tumor-draining lymph node (TDLN) cells activated with anti-CD3 and IL-2-mediated tumor regression in adoptive immunotherapy. In this study, we examined the TCR Vbeta repertoire usage of TDLN cells with respect to cytokine release profiles and therapeutic efficacy in vivo. The majority of the whole population of TDLN cells after activation with anti-CD3 were composed of Vbeta3+, -5+, -7+, -8+, and -11+ cells. Enrichment of Vbeta subsets of TDLN cells by in vitro activation with anti-Vbeta mAb revealed Vbeta8+ cells released high amounts of IFN-gamma and granulocyte/macrophage-CSF (GM-CSF) with minimal amounts of IL-10 in response to tumor and mediated tumor regression in vivo. In contrast, enriched populations of Vbeta5+, Vbeta7+, and Vbeta11+ cells released low amounts of IFN-gamma and GM-CSF with high levels of IL-10 and had no in vivo antitumor reactivity. In vitro depletion of specific Vbeta subsets from the whole TDLN pool confirmed that the profile of cytokines released correlated with in vivo antitumor function. Therapeutic efficacy mediated by TDLN cells required the release of IFN-gamma and GM-CSF since in vivo neutralization of both cytokines inhibited tumor regression. The administration of anti-IL-10 mAb abrogated the suppressed antitumor response manifested by adoptively transferred TDLN cells, which elaborated increased levels of IL-10. Our study documents that type 1 cytokine release (i.e., IFN-gamma and GM-CSF) promotes in vivo tumor Ag recognition, in contrast to type 2 release (i.e., IL-10), which suppresses this interaction, and discriminates the functional activity of Vbeta subpopulations of effector cells.

Immune responsiveness to a murine mammary carcinoma modified to express B7-1, interleukin-12, or GM-CSF.

This report characterizes the immunological host response to a syngeneic murine mammary carcinoma along with variants genetically modified to express B7-1 or secrete GM-CSF and interleukin-12 (IL-12). MT-901 is a subline of a mammary adenocarcinoma that was chemically induced in the Balb/c host. It was found to be weakly immunogenic by immunization/ challenge experiments, and it induced tumor-specific T-cell responses in lymph nodes (LN) draining progressive subcutaneous tumors. Tumor clones expressing B7-1 or secreting GM-CSF exhibited reduced tumorigenicity without completely abrogating tumor growth, whereas IL-12 elaboration lead to complete tumor growth inhibition. In vivo subcutaneous inoculation of a transgenic cell clone secreting GM-CSF (240 ng/10(6) cells/24 hours) resulted in significantly enhanced T-cell reactivity of tumor-draining lymph node (TDLN) cells as compared to wild-type TDLN cells. This finding was obtained from observations assessed by several different methods, including: 1) in vitro cytotoxicity, 2) in vitro interferon-gamma release, and 3) adoptive transfer in mice with established tumor. Moreover, the transfer of activated LN cells derived from mice inoculated with GM-CSF-secreting tumor cells resulted in the prolonged survival of animals with macroscopic metastatic disease, which was not evident utilizing LN cells from mice inoculated with wild-type tumor. By contrast, clones that expressed B7-1 or IL-12 (4 ng/10(6) cells/24 hours) did not elicit enhanced tumor-reactive TDLN cells compared with wild-type tumor when assessed in the adoptive transfer model. The autocrine secretion of GM-CSF by transduced tumor cells was found to serve as an effective immune adjuvant in the host response to this weakly immunogenic tumor.

Different cytokine profiles released by CD4+ and CD8+ tumor-draining lymph node cells involved in mediating tumor regression.

We have previously demonstrated that the growth of weakly immunogenic murine sarcomas leads to the induction of immunologically specific pre-effector cells in tumor-draining lymph nodes (TDLN). The in vitro activation of TDLN cells with anti-CD3 monoclonal antibodies (mAbs) and interleukin-2 (IL-2) resulted in the acquisition of effector function as measured by tumor regression in the adoptive immunotherapy of pulmonary metastases. Further studies were performed to characterize the mechanisms associated with in vivo tumor reactivity mediated by activated TDLN cells. By positive selection, CD4+ and CD8+ T cells were purified and activated by the anti-CD3/IL-2 method. CD8+, but not CD4+, cells manifested tumor-specific granulocyte-macrophage colony-stimulating factor (GM-CSF) and interferon-gamma (IFN-gamma) release in vitro, and elicited tumor regression in vivo. By contrast, only activated CD4+ were found to release significant amounts of IL-2 in response to tumor antigen but did not mediate tumor regression in vivo. Mixing the two purified populations enhanced the antitumor activity of the CD8+ T cells. In culture, IL-2 was found to augment the relative amount of tumor-specific release of GM-CSF and IFN-gamma by activated TDLN cells. We found that the tumor-specific release of GM-CSF and IFN-gamma by activated lymphocytes was strongly associated with the in vivo therapeutic efficacy of these cells. Evidence in support of this included the following: (1) cytokine release of TDLN derived after different durations of tumor growth correlated with tumor reactivity in adoptive transfer studies, (2) cytokine release of T cells derived from different lymphoid organs corresponded with tumor reactivity in adoptive transfer, and (3) in vivo administration of neutralizing mAb to IFN-gamma and GM-CSF significantly inhibited the antitumor reactivity of TDLN cells. These studies document the contributory roles of IFN-gamma, GM-CSF, and IL-2 released by activated CD4+ and CD8+ T cells involved in tumor regression.

Concurrent induction of CD4+ and CD8+ T cells during tumor growth with antitumor reactivity in adoptive immunotherapy.

We have previously reported that the poorly immunogenic D5 melanoma transduced to secrete granulocyte-macrophage colony-stimulating factor (GM-CSF) will elicit immunity in tumor-draining lymph node (TDLN) cells after subcutaneous inoculation. After in vitro activation with anti-CD3 and interleukin-2 (IL-2), these cells acquire in vivo antitumor reactivity to wild-type tumor in the adoptive immunotherapy of pulmonary metastases. Using monoclonal antibodies, depletion of CD4+ or CD8+ T cells immediately after the adoptive transfer of activated TDLN cells revealed that both subsets could mediate the regression of tumor in the absence of exogenous IL-2 administration. CD8+ cells were more potent than CD4+ cells in mediating tumor regression on a per cell basis. We found that the exogenous administration of IL-2 enhanced the antitumor efficacy of CD4+ T cells. Purified CD4+ and CD8+ TDLN cells that were activated separately in culture released GM-CSF and interferon-gamma in response to wild-type tumor in vitro and mediated tumor regression in vivo. Last, the induction of either immune CD4+ or CD8+ T-cell subset during growth of the GM-CSF-secreting melanoma was found to be unaffected by the depletion of the alternate T-cell subset before tumor inoculation. These findings demonstrate that both CD4+ and CD8+ T cells can independently acquire therapeutic reactivity and presumably recognize two separate epitopes involved in tumor rejection.

Adoptive immunotherapy with vaccine-primed lymph node cells secondarily activated with anti-CD3 and interleukin-2.

PURPOSE: In preclinical studies, we have reported the ability to induce immune T cells in lymph nodes (LN) primed by in vivo vaccination with tumor cells admixed with a bacterial adjuvant. These LN cells can be activated and expanded ex vivo for the successful immunotherapy of established tumors. We have applied these methods to generate vaccine-primed LN in patients with advanced melanoma and renal cell cancer (RCC) for therapy. MATERIALS AND METHODS: Irradiated autologous tumor cells admixed with bacille Calmette-Guérin (BCG) were used to vaccinate patients. Seven days later, draining LN were removed for activation with anti-CD3 monoclonal antibody (mAb) followed by expansion in interleukin-2 (IL-2). Activated LN cells were administered intravenously (IV) with the concomitant administration of IL-2. RESULTS: A total of 23 patients were evaluated (11 melanoma and 12 RCC). Vaccine-primed LN were expanded ex vivo with a mean of 8.4 x 10(10) cells administered per patient. Among 20 patients assessed, 15 demonstrated minimal cytotoxicity of autologous tumor cells by the activated LN cells, with the remaining mediating nonspecific cytotoxicity. By contrast, a majority of the activated LN cells showed highly specific release of granulocyte-macrophage colony-stimulating factor (GM-CSF) and interferon gamma (IFN-gamma) to autologous but not allogeneic tumor stimulation. This tumor-specific cytokine release was found to be major histocompatibility complex (MHC) class I-restricted, which indicates the involvement of CD8+ cells. Among 11 melanoma patients, one had a partial tumor response. Among 12 RCC patients, two had complete and two partial responses. A trend (P = .066) between the enhancement of delayed-type hypersensitivity (DTH) reactivity to autologous tumor after therapy and tumor regression was observed. CONCLUSION: Tumor vaccines can be used to induce immunologically specific T-cell responses against melanoma and RCC in draining LN. Anti-CD3/IL-2 activation of primed LN cells can be reliably performed for clinical therapy and appears to have activity in patients with metastatic RCC.

Usefulness of bone metabolic markers in the diagnosis and follow-up of bone metastasis from lung cancer.

Ninety-one lung cancer patients were evaluated to determine the usefulness of bone metabolic markers in the diagnosis and follow-up of bone metastases and also to investigate their clinical usefulness as an adjunct to bone scintigraphy. Both bone resorption markers, ICTP and fDPD, and bone formation markers, Al-p, BAL, PICP and BGP, were evaluated in 47 patients with and 44 without bone metastasis. The patients with bone metastasis were classified according to the bone metastatic burden, and they were also separately classified into groups according to the course of the bone metastasis. ICTP, fDPD, Al-p and BAL were significantly elevated (P < 0.001) in patients with bone metastasis, but PICP and BGP were not. Receiver-operating characteristic (ROC) curves of these markers revealed that ICTP was most highly correlated with the diagnosis of bone metastasis. The sensitivity of ICTP (71.4%) and fDPD (61.0%) were good with high specificity. T scores of ICTP, fDPD and BAL tended to be higher at higher grades of bone metastasis. T-scores of ICTP, fDPD and BAL were elevated in the newly diagnosed cases and progressed cases, but the T-scores of ICTP and fDPD in those cases were higher than that of BAL. In the follow-up study, ICTP was well correlated with uncontrolled or controlled bone metastasis. Thus, bone resorption markers, especially ICTP, could be a good indicator of the progression and multiplicity of disease, and it could help in the follow-up and in the monitoring of therapy for bone metastasis from lung cancer.

Immune response in human melanoma after transfer of an allogeneic class I major histocompatibility complex gene with DNA-liposome complexes.

Analysis of the antitumor immune response after gene transfer of a foreign major histocompatibility complex class I protein, HLA-B7, was performed. Ten HLA-B7-negative patients with stage IV melanoma were treated in an effort to stimulate local tumor immunity. Plasmid DNA was detected within treated tumor nodules, and RNA encoding recombinant HLA-B7 or HLA-B7 protein was demonstrated in 9 of 10 patients. T cell migration into treated lesions was observed and tumor-infiltrating lymphocyte reactivity was enhanced in six of seven and two of two patients analyzed, respectively. In contrast, the frequency of cytotoxic T lymphocyte against autologous tumor in circulating peripheral blood lymphocytes was not altered significantly, suggesting that peripheral blood lymphocyte reactivity is not indicative of local tumor responsiveness. Local inhibition of tumor growth was detected after gene transfer in two patients, one of whom showed a partial remission. This patient subsequently received treatment with tumor-infiltrating lymphocytes derived from gene-modified tumor, with a complete regression of residual disease. Thus, gene transfer with DNA-liposome complexes encoding an allogeneic major histocompatibility complex protein stimulated local antitumor immune responses that facilitated the generation of effector cells for immunotherapy of cancer.

Primary lymphomas of the breast detected by Ga-67 scintigraphy.

Three patients with primary breast lymphoma were reported. All patients had rapidly growing breast masses and these masses showed nonspecific findings in mammography and on ultrasound. They resembled phyllodes tumor or invasive ductal carcinomas. Ga-67 scintigraphy showed intense accumulation was seen in all of the lymphomas. In one patient, a breast carcinoma was detected in the contralateral breast which did not show Ga-67 accumulation. In the other two patients, axillary involvement was also detected. Scintigraphy was helpful to confirm the diagnosis of breast lymphoma.

Analytical study of the clinical response to two distinct adoptive immunotherapies for advanced hepatocellular carcinoma: comparison between LAK cell and CTL therapy.

To evaluate the effect of two distinct adoptive immunotherapies, tumor-specific cytotoxic T-cell (CTL) therapy and lymphokine-activated killer (LAK) cell therapy, the clinical responses of patients with stage IV primary hepatocellular carcinoma (HCC) treated with these therapies were studied. Of 18 patients treated with CTL, 3 had complete regression (CR), 2 had partial regression, and 3 had minor regression (MR). Their median survival was 21 months after the end of therapy, and 1 CR patient survived for > 6 years. On the other hand, in the LAK-cell-treated group of eight patients, four had MR and their median survival was only 2 months. No survival was observed 27 months after the end of LAK cell therapy. These results indicate that tumor-specific CTL therapy is more effective than LAK cell therapy and that it might be a promising therapeutic tool for advanced HCC patients.

Therapeutic efficacy of T cells derived from lymph nodes draining a poorly immunogenic tumor transduced to secrete granulocyte-macrophage colony-stimulating factor.

We examined the host immune response to the poorly immunogenic B16-BL6 melanoma, which was transduced to secrete granulocyte-macrophage colony-stimulating factor (GM-CSF) (450 ng/10(6)/24 h). Tumor growth after subcutaneous inoculation was not significantly altered, although an influx of neutrophils and monocytes/macrophages was evident within tumors and draining lymph nodes (LNs). Immunization with irradiated transduced cells did not induce systemic immunity to the parental tumor. However, vaccination with transduced tumors significantly augmented in vivo sensitization of draining LN cells. These tumor-draining LN (TDLN) cells, when secondarily stimulated in vitro with anti-CD3 monoclonal antibodies and expanded in interleukin-2 (10 U/ml), exhibited greater release of GM-CST and interferon-gamma against tumor compared with TDLN cells from animals with parental tumor. In adoptive immunotherapy, activated LN cells draining transduced tumors mediated significant reductions of the numbers of established pulmonary metastases compared with LN cells draining parental tumor, which were ineffective. In addition, the therapeutic efficacy of LN cells draining transduced tumors was significantly better than LN cells primed in vivo with tumor cells admixed with Corynebacterium parvum, which we have previously described as an approach to generate immune cells. Thus, GM-CSF appears to be an important adjuvant in the induction of tumor immunity.

Tumor-specific granulocyte/macrophage colony-stimulating factor and interferon gamma secretion is associated with in vivo therapeutic efficacy of activated tumor-draining lymph node cells.

In this study, cytokine release by tumor-draining lymph node cells sensitized in vitro (IVS-TDLN) was examined and correlated with therapeutic efficacy in adoptive immunotherapy. Mice bearing immunologically distinct MCA 207 and MCA 205 sarcoma tumors were utilized in criss-cross experiments. IVS-TDLN obtained from mice bearing 10-day subcutaneous (s.c.) tumors mediated immunologically specific regression of established 3-day pulmonary metastases, but demonstrated non-specific cytolytic reactivity against both tumors in a 4-h 51Cr-release assay. By contrast, these IVS-TDLN cells were found specifically to secrete granulocyte/macrophage colony-stimulating factor (GM-CSF) and interferon gamma (IFN gamma) when restimulated in vitro with irradiated tumor cells. To determine the predictive value of tumor-specific cytokine release with in vivo therapeutic efficacy, a kinetic analysis of antitumor activities of TDLN obtained from animals bearing MCA 207 tumors for increasing lengths of time was performed. IVS-TDLN cells from mice bearing day-7, -10 and -14 s.c. tumors manifested tumor-specific release of GM-CSF and IFN gamma, and mediated significant antitumor reactivity in vivo. In contrast IVS-LN cells from day-0 and day-21 tumor-bearing animals did not release significant amounts of GM-CSF and IFN gamma, and were not therapeutically efficacious in vivo. Day-4 IVS-TDLN released high levels of GM-CSF and IFN gamma non-specifically, and were not therapeutic in adoptive immunotherapy at doses effective for day-7 and day-14 IVS-TDLN cells. In other experiments, IVS cells generated from different lymph node groups in animals bearing 10-day established s.c. tumors were examined and found to have unique profiles of cytokine release. In these studies, the ability of IVS cells to release specifically both cytokines as opposed to one was associated with greater therapeutic efficacy on a per cell basis. Our findings suggest that the tumor-specific releases of GM-CSF and IFN gamma are useful parameters to assess the in vivo therapeutic efficacy of immune lymphocytes.

[An evaluation of perioperative and continuous infusion of tegafur as surgical adjuvant chemotherapy of resected gastric cancer].

Perioperative continuous intravenous infusion of Tegafur (FT) seems theoretically an effective adjuvant chemotherapy for patients with gastric carcinoma. Then we estimated the clinical effect of this therapy on the survival of patients who had undergone curative (Stage I, II, III) and non-curative (Stage IV) resection of a primary gastric carcinoma. One hundred fifty patients received chemotherapy and two hundred thirty-seven were observed. Results showed a significant difference in survival curves of Stage I (p less than 0.01) and Stage III (p less than 0.01). The 5-year survival rate was 100% for the adjuvant therapy arm and 82% for the observation arm in Stage I. In stage III, the survival curve of both arms was similar from the beginning to the 5th year, but the 7-year survival rate was 53% and 39%, respectively. There was no serious side effect of chemotherapy. These results demonstrate that this chemotherapy regimen is beneficial and recommendable as a perioperative adjuvant chemotherapy for resected gastric cancer in which the cancer cells were almost removed.

Induction of autologous tumor-specific cytotoxic T cells in patients with liver cancer. Characterizations and clinical utilization.

To generate autologous-tumor-specific cytotoxic T cells (CTL), peripheral blood mononuclear cells (PBMC) obtained from cancer patients were cultivated with autologous tumor cells for 5 days, and restimulated with interleukin-2 for another 5 days. Subsequently, their cytotoxic activity was examined by an in vitro cytotoxic test as well as by Winn's assay utilizing nude mouse transplanted autologous tumors. The present results demonstrated that these in vitro-stimulated cells were able to kill autologous tumor cells but not allogeneic tumors, and that they also inhibited the growth of transplanted autologous tumors in the nude mouse. Their cytotoxic activity was completely abrogated by pre-treatment with either anti-CD3 or anti-CD8, but not with anti-CD4, plus complement. Based on these studies, we injected these CTL via the hepatic artery into patients having either nonresected tumors or recurrent tumors in the liver. Among 15 treated patients (13 with hepatocellular carcinoma and 2 with metastatic liver cancer) 2 complete responses, 3 partial responses and 4 minor responses were observed. During the 6 to 25 months following injection of CTL, no definite signs of tumor recurrence or regrowth were demonstrated in these 5 responding patients (complete plus partial).

[Cancer cells in the portal and hepatic veins of resected specimens of hepatocellular carcinomas].

After injecting 100 ml of a saline solution into the hepatic artery of 27 livers evidencing a hepatocellular carcinoma, samples from the resected livers were collected through the portal and the hepatic vein. Cancer cells were detected in 4 out of 27 samples from the portal vein and in 1 out of 19 samples from the hepatic vein, and the total number of positive cases were 4. Further, tumors over 5 cm in diameter showed a tendency of flowing cancer cells into the portal vein. Four positive cases had an intrahepatic and a distant metastasis within one year of their operation and two of these patients died.