Liver Metastatic Models Of Human Primary Gastric Lymphoma Constructed Using Surgical Orthotopic Implantation Of Histologically Intact Tissue In Nude Mice And Adoptive Immunotherapy Study Of Cytokine-Induced Killer Cells

Primary gastric lymphoma (PGL) accounts for 2.5%-3.4% of gastric tumors, 5%-11% of malignant tumors of the stomach, and 48%-63% of gastrointestinal malignancies. PGL is the second frequent tumor with gastric carcinoma being first in the stomach. Up till now the etiology and pathogenesis of PGL have not been found. The symptoms, signs and laboratory examinations of PGL lack specificity, which results misdiagnosis with gastric carcinoma. Generally, preoperative diagnosis and staging are difficult. Compared with other frequent tumors, the deepness and width of the basic and clinical researches on PGL are greatly restricted mainly due to the lack of knowledge on the biological characteristics of PGL, the lack of tumor tissue specimen for direct researches, and the lack of animal tumor models for experimental use.In order to investigate pathogenesis, metastasis biology and experimental therapies of PGL, it is necessary to establish an ideal animal model of PGL. The present study first successfully established a liver metastatic model of human primary gastric non-Hodgkin's B cell lymphoma (the liver metastatic lesion) in nude mice and a model of human primary gastric non-Hodgkin's B cell lymphoma (the primary lesion) in nude mice using advanced methods of surgical orthotopic implantation of histologically intact metastatic tissue and in vivo orthotopic continuous screening, termed HGBL-0304 and HGBL-0305, respectively.HGBL-0304 model: The fresh liver metastatic tumor fragments from the patient with PGL were orthotopically implanted into the stomachs of five nude mice for the first passage. Up to the present time, human primary gastric lymphoma has been maintained for 45 generations by orthotopic passage (stomach to stomach). Totally, 263 nude mice were implanted. During the passage, the latency of transplanted tumors obviously shortened with the average of 7.6 days. The average passage interval was 28 days. Histological examination revealed that tumor growth was observed in the sinus ventriculi, lesser curvature of stomach, or pylorus of all the transplanted mice, and the whole stomach was invaded by tumors in some animal. Metastasis rates of liver, spleen, lymph nodes and peritoneal planting were 100%, 94.3%, 62.6% and 43.5%, respectively. Usually, bearing-cancer mice approached to moribund for exhaustion 6-7 weeks after transplantation. The longest survival was 76.5 days with a median survival of 53 days after tumor implantation.HGBL-0305 model: The fresh primary tumor fragments from the same patient were orthotopically implanted into the stomachs of seven nude mice for the first passage and successful implantation were achieved in five nude mice (71.4%). Up to the present time, human primary gastric lymphoma has been maintained for 45 generations by orthotopic passage (stomach to stomach). Totally, 156 nude mice were implanted. During the passage, the latency of transplanted tumors obviously shortened with the average of 7.4 days. The average passage interval was 27 days. Histological examination revealed that tumor growth was observed in the sinus ventriculi, lesser curvature of stomach, or pylorus of all the transplanted mice, and the whole stomach was invaded by tumors in some animal. Metastasis rates of liver, spleen, lymph nodes and peritoneal planting were 69.5%, 55.6%, 45.7% and 30.5%, respectively. Usually, bearing-cancer mice approached to moribund for exhaustion 7-9 weeks after transplantation. The longest survival was 142 days with a median survival of 108 days after tumor implantation.There was significant difference in metastatic biology between two models. In HGBL-0304 model, more earliness in liver metastatic time (2 weeks after implantation), higher liver metastatic rate (100%), more serious extent of liver metastasis (diffusely involving the left and right lobes of the liver), and shorter survival with a mean survival of 54.3 days were found, which were indicative of more malignancy than HGBL-0305 model. Simultaneously, the result also proved that the metastatic lesion implantation was a effective method for improving metastatic rate to establish high metastatic animal model of human cancers.After continuous passage for a long time, not only the transplanted tumors were consistent with original human PGL in histopathology, immunophenotype, ultrastructure, DNA ploid level and karyotype, but also growth rate of the transplanted tumors in nude mice and resuscitation rate of liquid nitrogen cryopreservation were both 100% in HGBL-0304 and HGBL-0305 models. The transplanted tumors revealed no spontaneous regression, and proliferated in accordance with one invariable growth rule, i.e. in the order of latent period, slow growth period and rapid growth period. Dyscrasia status was always seen when bearing-cancer mice approached dying. Those indicated that two models had steady biological characteristics.The two models completely simulated the natural clinicopathological course and retained original biological features of PGL. The establishment of those models has a significant value in the researches concerning metastatic biology, histogenesis, experimental therapy, tumor immunity, and molecular biology of PGL.Since Schmidt-Wolf IG, a professor in Stanford University Medical Center, reported an improved protocol for the generation of high numbers of effector cells (cytokine-induced killer cells, CIK) that had increased cytotoxicity against human B lymphoma cell lines, CIK cell therapy has greatly developed and becomes an optimal project in the field of adoptive immunotherapy. To generate CIK cells, human peripheral blood mononuclear cells (PBMCs) are separated and cultured ex vivo in the presence of many cytokines stimulation (anti-CD3 McAb, IL-2, IFN-γ, IL-1α, etc.) for the specified time. CIK cells are a integrated group of heterogeneous cells, in which the main effector cell type simultaneously expresses two membrane proteins, i.e. CD3 and CD56, and thus it is termed natural killer (NK)-like T lymphocyte that has both high antitumor activity of T cell and non-major histocompatibility complex (MHC) restricted cytotoxicity of NK cell against tumor cells. CIK cells overcome the limitations of lymphokine-activated killer (LAK) and tumor-infiltration lymphocyte (TIL) previously used that include low proliferation, inherent low cytotoxic activity, difficult preparation, and possible alterations in function during extraction from human tissue. Compared with LAK cells and TIL cells, CIK cells have several advantages in antitumor cell acitivity. Those advantages are as follows: (1) high proliferation; (2) high cytotoxic activity; (3) large-scale antitumor spectrum; (4) sensitivity of multidrug-resistant tumor cell lines to CIK cells; (5) immunosuppressive agents have no effects on antitumor activity of CIK cells; (6) minimal cytotoxicity against normal hematopoietic precursor cells; (7) potent resistance to apoptosis of effector cells through Fas-FasL pathway by tumor cells. Antitumor activity of CIK cells is carried out by three main mechanisms, which are as follows: (1) to directly recognize and kill tumor cells; (2) to secrete inflammatory cytokines and enhance immune function thus to inhibit tumor cell growth; (3) to induce apoptosis of tumor cells. In the present, domestic and overseas researchers have widely performed animal experiments, preclinical tests as well as phaseⅠandⅡtrials concerning CIK cells against many tumor cell lines and in vivo tumors.It is the low frequency of PGL in clinic practice that restricts the basic and clinical researches on PGL. Based on the establishment of liver metastasis model of human primary gastric lymphoma in nude mice, we first further investigate inhibitory effect of CIK cells on growth and liver metastasis of PGL, which will provide experimental data for applying CIK cells to treat PGL in clinic. In the therapy experiment, antitumor activities of three effector cell types, i.e. LAK cells, CIK cells from healthy donors and CIK cells from the patients with PGL, are comparatively observed.The experiment included five treatment groups, which were saline control group (0.3ml), LAK cell group (6×10~9/0.3ml), healthy donor CIK cells group (6×10~9/0.3ml), PGL patient CIK cells group (3×10~9/0.3ml), PGL patient CIK cells group (6×10~9/0.3ml). After treatment, tumor weight of every group was (1.40±0.17)g, (0.85±0.23)g, (0.65±0.16)g, (0.83±0.22)g, and (0.61±0.20)g, respectively; and the difference in tumor weight was significant (p<0.01), compared with saline control group. Inhibitory rate of tumor growth was 0, 39.28%, 53.57%, 40.38%, and 56.42%, respectively. Liver metastasis rate was 100%, 62.5%, 50%, 62.5%, and 37.5%, respectively; compared with saline control group, the difference in liver metastasis rate was significant (p<0.01).The experiment revealed that antitumor effect on PGL was seen in three effector cell types, i.e. LAK cells, CIK cells from healthy donors and CIK cells from the patients with PGL, presenting inhibition of PGL xenograft growth, inhibition of liver metastasis of PGL xenograft in mice, and alleviation of dyscrasia. To be emphasized, CIK cells from the patients with PGL demonstrated more remarkable antitumor activity with dose-effect relationship, and increase of numbers of effector cells could inhibit metastasis to liver more effectively. Valuable experimental evidence was supplied for adoptive immunotherapy of PGL with CIK cells.