Migration activity of peritoneal exudate cells in Syrian hamsters in normal conditions and during suppression of their natural resistance to tumors

Migration activity (MA) of peritoneal exudate cells (PEC) was studied in Syrian hamsters in normalcy and under intraperitoneal injection into the animals of inactivated normal and tumor cells including those capable of inhibiting the natural animals' resistance to tumor. No significant individual differences were found in MA of PEC of intact hamsters. MA of PEC of hamsters treated with inactivated normal or spontaneously transformed in vitro cells of hamster embryo did not differ from the means of MA in the control. MA of PEC of hamsters treated with inactivated tumor cells was found to be appreciably enhanced. The ability of inactivated tumor cells to induce the enhancement of MA of PEC correlates with their ability to suppress natural tumor resistance.     

The chemiluminescence reactions of the blood neutrophils and of peritoneal exudate cells in Syrian hamsters to the intraperitoneal administration of cellular and microbial materials]

The luminol-dependent chemiluminescence (CL) activity of peritoneal exudate cells and blood neutrophils of Syrian hamsters inoculated intraperitoneally with heat-inactivated microbial particles of Candida albicans, (C. albicans), heated irradiated normal cells and native or heated irradiated malignant tumor cells was studied. The inoculation with particles of C. albicans and heated normal cells induced significant activation of CL of peritoneal exudate cells, but did not influence the CL reaction of blood neutrophils. The inoculation of animals with nonheated irradiated tumor cells led to increase of CL response of both peritoneal exudate cells and blood neutrophils. The inoculation with heated irradiated tumor cells did not activate CL of peritoneal exudate cells and led to slight, but long-lasting decrease of CL response of blood neutrophils.     

Correlation between increase in Ia-bearing macrophages and induction of T cell-dependent antitumor activity by Lactobacillus casei in mice

Summary When Lactobacillus casei YIT 9018 (LC 9018) or Corynebacterium parvum, known to be immunomodulators possessing antitumor activity, were injected i.p. into BALB/c mice, peritoneal exudate macrophage Ia antigen detected by indirect immunofluorescence method was expressed on their cell surface, but it was not expressed following the injection of 10% proteose peptone, an inflammatory agent, or Lactobacillus fermentum YIT 0159 (LF 0159), which have no antitumor activity. The percentage and absolute number of Ia-positive peritoneal macrophages were maximum on the 7th day after the injection of LC 9018. Immunization by injection of Meth A fibrosarcoma cells treated with mitomycin C (MMC-Meth A) 7 days after LC 9018 injection suppressed the growth of Meth A implanted i.p. 14 days after MMC-Meth A injection. A shorter interval between the injections of LC 9018 and MMC-Meth A did not allow suppression of Meth A growth. These results showed that the increase in Ia-positive macrophages in the peritoneal cavity coincided with the effective interval for induction of the antitumor activity by LC 9018. The antitumor activity induced by injections of LC 9018 and MMC-Meth A did not affect the growth of RL l leukemic cells, syngeneic to BALB/c mice. Neutralization (Winn type) tests showed that peritoneal T lymphocytes possessed tumor cytotoxicity and that the antitumor capacity was reduced by in vivo treatment with anti I-A^d monoclonal antibody simultaneously with and 1 day prior to MMC-Meth A injection. These results indicate that LC 9018-induced Ia-positive macrophages, which first encounter a tumor antigen in the peritoneal cavity, play an important role in the in vivo induction of tumor specific T cell-mediated antitumor immunity.     

Superoxide production by phagocytes in myeloid Graffi tumor-bearing hamsters

A progressive suppression of the phagocytic ability of peritoneal macrophages and polymorphonuclears (PMNs) in hamsters with transplanted myeloid tumors was previously established. The i.p. application of Cu/Zn SOD, isolated from the fungal strain Humicola lutea (HLSOD) (2 injections before and 5 injections after tumor transplantation) induced the mean survival time of the animals as well as a temporally stimulating action on the macrophage and PMNs phagocyting indices. In the present work, the superoxide production of peritoneal macrophages and PMNs during 30 days of tumor progression was followed. Effects of the application of HLSOD in an optimal protective dose on the superoxide production in peritoneal macrophages and blood PMNs were examined. The spontaneous and phorbol-myristate acetate (PMA)-inducible O2- production in both types of phagocytes was 4-5-fold increased in tumor-bearing hamsters (TBH), as compared to the controls, at day 14 after tumor transplantation (the day of tumor appearance in transplanted animals). Furthermore, O2- production was also similar to the control values for the following days of observation. HLSOD treatment of TBH induced a normalization of superoxide production in macrophages and PMNs. Therefore, the established decrease of superoxide anions in phagocyting cells of TBH indicates possible effects of HLSOD on the host antioxidant defense.     

Anti-tumor cytostatic mechanism and delayed-type hypersensitivity against a syngeneic murine tumor. Comparison between neonatally thymectomized mice and congenitally athymic nude mice

We studied the anti-tumor mechanism against a syngeneic tumor using a BALB/c-MA tumor system by cytolysis and cytostasis assays in vitro comparing mice neonatally thymectomized at 1 day or 7 days after birth (NTx-1, NTx-7), sham-operated (sham) mice, and congenitally athymic nude BALB/c mice. NTx-1 mice showed more rapid tumor growth and a slightly lower degree of strong cytostatic activity in peritoneal exudate cells (PEC) than NTx-7 or sham mice. Nude mice showed more rapid MA growth than NTx-1 mice and no cytostatic activity in PEC. After immunization with mitomycin C-treated MA (MMC-MA), NTx-1 mice acquired an immunoprophylactic capacity against MA and showed cytostatic activity and delayed footpad reaction (DFR) to MA, however, nude mice showed no acquisition of such an immunity, or cytostatic activity, or DFR to MA. These differences between NTx-1 and nude mice could be well-explained by less capacity of nude mice to produce a macrophage-activating factor, which activates macrophages to exert cytostasis and DFR. However, NTx-1 mice could not reject MA by immunization with MMC-MA in CFA (MMC-MA/CFA), although such immunized sham mice could eliminate MA completely. Both PEC and spleen cells from Sham mice immunized with MMC-MA/CFA showed cytostatic activity, whereas NTx-1 mice showed cytostatic activity of the same level in PEC and less in spleen cells compared to Sham mice. Cytolytic activity was never detected throughout this study in a BALB/c-MA system. These data suggest that cytostasis plays an important role in antitumor immunity against a syngeneic MA tumor and that two types of cytostasis is included from the standpoint of thymus-dependency of ontogenic development, relatively low and high.     

Antitumor and immunologic effects of a pyridine-extracted fraction of Propionibacterium acnes

Summary The antitumor and immunological effects of a pyridine extractable fraction of Propionibacterium acnes were tested in a murine ovarian teratocarcinoma (MOT) model. Previous studies have demonstrated that tumor rejection in this model depends upon sequential activation of tumoricidal neutrophils (PMNs) followed by cytostatic macrophages. The pyridine extract significantly prolonged the survival of mice challenged with 10³ or 10⁴ MOT cells but had little impact on a 10⁵ tumor inoculum. In vivo cytoreduction occurred within the first 24 h following IP treatment which correlated temporally with the influx of tumoricidal PMNs into the peritoneal cavity. Immunotherapy failure in mice challenged with 10⁵ MOT cells occurred between 48 and 72 h after treatment when macrophage chemotaxis into the peritoneal cavity was initiated. Although injection of unfractionated bacteria activated MOT-cytostatic macrophages, the pyridine extract was deficient in this regard. Intraperitoneal injection of the pyridine extract resulted in an early (day +1) depression and late (day +5) enhancement of peritoneal NK cytotoxicity. These data suggest that the retention of neutrophil-activating moieties in the pyridine extract are sufficient for antitumor effects against low tumor inocula while the depletion of macrophage-activating determinants results in diminished effects against larger tumor cell challenges.     

Immune reactivity in SL2 lymphoma-bearing mice compared with SL2-immunized mice

Summary We have studied the rather paradoxical phenomenon of the growth of an antigenic tumor in an immunocomponent host. This phenomenon was studied by comparing (a) the lymphocyte reactivity and (b) the macrophage cytotoxicity, during SL2 growth in DBA/2 mice (SL2-bearing mice) and in DBA/2 mice immunized against SL2 tumor cells (SL2-immune mice). Immune mice rejected a challenge of tumor cells. The immune T-lymphocytes rendered macrophages cytotoxic (arming) and were able to transfer tumor resistance to naive animals. Nonimmunized mice did not reject a challenge of SL2 cells. In these tumor-bearing mice various forms of immune reactivity were tested. Lymphocytes with the capacity to arm macrophages could not be found in the lymphoid organs. However, lymphocytes isolated from the tissue directly surrounding the subcutaneous SL2 tumor could arm macrophages in vitro.Shortly after subcutaneous tumor grafting cytotoxic macrophages were found in the peritoneal cavity. In the serum macrophage arming factors were detected that rendered macrophages cytotoxic in vitro. This cytotoxicity of the peritoneal macrophages and the presence of macrophage arming factors in the serum showed a similar biphasic pattern. The first phase of cytotoxicity between day 3 and 8 after tumor grafting was tumor (SL2) specific. The second phase from day 12 and onwards was not tumor specific. During the first 4 days after SL2 grafting the DBA/2 mice expressed a specific concomitant immunity to a second tumor graft. Then 7 or more days after grafting the first SL2 tumor, the concomitant immunity was nonspecific as the growth of a second SL2 tumor graft and a L5178Y (DBA/2) tumor graft were inhibited. In addition, the immune suppressive activity of serum and lymphocytes was tested. Neither serum nor lymphocytes from SL2-bearing mice suppressed the macrophage arming capacity of SL2 immune lymphocytes. Lymphocytes from tumor-bearing mice did not inhibit the capacity of SL2-immune lymphocytes to transfer resistance to naive animals. On the contrary, lymphocytes obtained from SL2-bearing mice 14 days after SL2 grafting transfered tumor resistance in a Winn-type assay. These data suggest that the growth of an antigenic tumor is due to the inability of the immune system to mount an effective antitumor effector cell population during tumor growth, rather than an immune suppression of the antitumor reactivity, as a limited immune reactivity could be detected in tumor-bearing mice, whereas immune suppression could not be detected.     

The thalamic intergeniculate leaflet mediates locomotor activity-induced reversal of phenotype in photoperiod nonresponsive Siberian hamsters

The role of the intergeniculate leaflet of the thalamus (IGL) in photoperiod responsiveness was examined in a laboratory-selected line of photoperiod nonresponsive (NR) Siberian hamsters. NR hamsters fail to exhibit typical winter-type responses (i.e., gonadal regression and development of winter-type pelage) when exposed to short day lengths (e.g., 10 h of light/day). Earlier studies revealed that NR hamsters will exhibit winter-type responses when exposed to short photoperiod if they are given free access to a running wheel. The present study tested the hypothesis that this locomotor activity-induced reversal of phenotype is dependent on the IGL. Male NR hamsters underwent destruction of the IGL prior to being housed in short day lengths in cages equipped with running wheels. Activity rhythms were monitored for 8 weeks, after which time pelage response and paired testes weights were obtained. In contrast to sham-operated NR animals given access to running wheels, IGL-ablated animals showed no increase in the duration of nocturnal running wheel activity and became active later in the night than sham-lesioned animals. Lesioned animals also failed to exhibit the typical short photoperiod-induced gonadal regression and pelage molt. The results implicate the IGL in the mechanism by which running wheel activity can influence photoperiodic responses.     

Destruction of tumor cells by BCG-activated alqolar macrophages.

Alveolar macrophages obtained from Syrian golden hamsters were tested for their ability to destroy tumor cells. Only macrophages obtained from BCG immune animals rechallenged intratracheally with BCG five days before assay exhibited cytotoxic activity. Maximum destruction of tumor cells occurred after 5 days of incubation. Immunologic activation of macrophages was required to attain cytotoxic alveolar macrophages. Induction of inflammatory lung exudates by a variety of nonspecific irritants did not result in tumor cell destruction by macrophages. These observations may prove useful in designing an approach for immunotherapy of lung cancer.     

The role of host lymphocytes and host macrophages in antitumor reactions after injection of sensitized lymphocytes and tumor target cells into naive mice

Summary DBA/2 mice were immunized i.p. against syngeneic SL2 lymphosarcoma cells. At various days after the last immunization peritoneal and spleen lymphocytes were collected. The lymphocyte suspensions were enriched for T-cells by nylon wool filtration.The peritoneal T-cells from immunized mice (a) expressed direct specific antitumor cytotoxicity in vitro, (b) induced macrophage cytotoxicity in vitro, and (c) exerted tumor neutralization measured in a Winn-type assay. Spleen T-cells from these immunized mice (a) expressed no direct specific antitumor cytotoxicity in vitro, (b) only induced moderate macrophage cytotoxicity in vitro, but (c) exerted tumor neutralization in a Winn assay.For effective tumor neutralization in vivo effector target cell ratios of 1000:1 were required. When the effector/target ratio of 1000:1 was maintained but the absolute numbers of effector and target cells were lowered from 10⁶ to 10⁵ lymphocytes and 10³ to 10² target cells respectively, no tumor neutralization was obtained.The major effect of the sensitized-transferred T-lymphocytes seemed to be the induction of cytotoxic macrophages in the (naive) recipient mice, as the peritoneal macrophages collected from the recipient mice 7 days after i.p. injection of a mixture of sensitized T-cells and tumor cells were cytotoxic. Purified peritoneal T-lymphocytes collected from these recipient mice were able to induce macrophage cytotoxicity in vitro but expressed no cytotoxic T-cell activity.In conclusion, our results show that in the tumor system used, tumor neutralization after transfer of sensitized lymphocytes is not dependent on the presence of cytotoxic T-lymphocytes. Lymphocytes with the strongest potency to render macrophages cytotoxic (in vitro and in vivo) also induce the best tumor neutralization in vivo, suggesting an important role for host macrophages as antitumor effector cells.     

小鼠腹腔巨噬细胞被S-O_2—1激活后的抑瘤作用 Ⅰ.小鼠腹腔巨噬细胞的体外细胞毒作用及扫描电镜观察

在六十年代Mathe就用卡介苗来治疗小儿白血病,并获得了良好的效果。以后人们开始重视如何利用微生物佐剂刺激机体的防御机制,增强机体对肿瘤的排斥能力,来达到提高肿瘤治疗效果的目的。     

Activated macrophages migrate to the subcutaneous tumor site via the peritoneum: a novel route of cell trafficking

Spontaneous regression of AK-5 tumor in syngeneic hosts reported earlier involves the interplay of Th1-type cytokines and cell-mediated immunity. Upon subcutaneous transplantation of AK-5 cells, there was accumulation of immune cells in the peritoneum, of which macrophages were the predominant type and were found to be in a hyperactive state. They released macrophage-derived tumoricidal mediators like NO, O2(-), and ONOO(-) which exhibited potent cytotoxic activity against AK-5 cells in vitro. Interestingly, there was a dramatic disappearance of these hyperactive cells from the peritoneal cavity which correlated well with the onset of tumor regression at the subcutaneous site. Direct labeling of these cells in the peritoneum by the tracking dye PKH26 showed their migration to the tumor site. Similarly, frozen tumor sections when scanned under confocal microscope clearly exhibited fluorescent macrophages embedded into the tumor. Immunohistochemical sections of these intratumoral macrophages showed nitrotyrosine residues, indicating their contribution in the free-radical-mediated AK-5 cell death, thereby leading to successful tumor remission. These observations suggest a directional migration of the hyperactivated peritoneal population to the tumor site. We have also confirmed the influx of macrophages and other immune cells into the peritoneum after sc transplantation of Meth A tumor cells in Balb/c mice. Our studies suggest a role for the peritoneal compartment in imparting appropriate stimulus to the immune cells prior to their participation in the antitumor immune response. These studies suggest a novel route of macrophage trafficking via the peritoneum.     

The rate of in vivo selection of highly metastatic and resistance-inhibiting variants of tumor cells

The rate of in vivo selection of tumour cells of STHE strain (spontaneously transformed in vitro hamster embryo cells, previously not selected in vivo) after their intravenous inoculation to normal Syrian hamsters was studied. At different time intervals after inoculation tumour cells were isolated from the lung tissue of individual animals, cultured in vitro and the their metastatic activity (MA) and natural resistance-depressing activity (RDA) were studied. It was demonstrated that selection of highly metastatic variants of tumour cells begins immediately after inoculation; on days 5-6 after inoculation the presence of such variants is observed in the majority of animals. Selection of resistance-depressing variants of tumour cells (determined in a less sensitive test) was first noticed on days 9-10 after inoculation of tumour cells. The correlation between MA and RDA of tumour cells was observed. The data obtained give evidence in favour of preexistence of genetic MA and RDA variants of tumour cells in the original population of tumour cells.     

Effects of microwave exposure on the hamster immune system. III. Macrophage resistance to vesicular stomatitis virus infection

Exposure of hamsters to microwave (MW) energy (2.45 GHz, 25 mW/cm2, 1 h) resulted in activation of peritoneal macrophages (PM) to a viricidal state restricting the replication of vesicular stomatitis virus (VSV). The PM from MW-exposed hamsters were viricidal as early as 1 day after exposure and remained active for 5 days. Immunization of hamsters with vaccinia virus induced viricidal PM by 3 to 4 days and they remained active for 7 days. To test the hypothesis that thermogenic MW exposure results in the release of endotoxin across the intestinal epithelium which subsequently activates PM, hamsters were injected with lipopolysaccharide (LPS) and their viricidal activity was studied. Lipopolysaccharide in vitro (0.2 microgram) and in vivo (0.5 microgram) activated macrophages to a viricidal state. When administered in vivo, LPS (0.5 microgram) activated macrophages as early as 1 day and the activity remained for 3 days. While MW exposure of PM in vitro failed to induce viricidal activity, exposure of PM to LPS in vitro induced strong viricidal activity. This suggests that the in vivo response of PM to MW is an indirect one, which is consistent with the hypothesis that MW-induced PM viricidal activity may be mediated via LPS. In preliminary experiments, MW exposure resulted in extended survival time for hamsters challenged with a lethal dose of vesicular stomatitis virus, supporting the concept that MW-activated PM may be a useful therapeutic modality.     

Functional heterogeneity among peritoneal macrophages: I. Effector cell activity of macrophages against syngeneic and xenogeneic tumor cells

Peritoneal exudate cells from immunized and nonimmunized animals were separated into subpopulations by centrifugation on discontinuous bovine serum albumin (BSA) density gradients. Cells in the several subpopulations were then tested for their cytostatic or cytotoxic activity against syngeneic and xenogeneic tumor cells. Nonimmune macrophages isolated at the 8 to 11% BSA interface were highly inhibitory to the growth of syngeneic and xenogeneic tumor cells during coculture for 24 to 48 hr. A second macrophage subpopulation of heavier density was not as effective in preventing tumor growth and frequently augmented it. Cytotoxic activity against (C58NT) D tumor cells could not be detected with macrophages or subpopulations of macrophages from immune as well as nonimmune animals, as determined by a 4-hr chromium release assay. The cytotoxic activity of the immune peritoneal exudate cells observed by this assay could be accounted for by the small percentage of lymphocytes present.     

Activity of adenosine metabolism enzymes and spontaneous chemiluminescence in macrophages in the tumor growth process

Adenosine deaminase and 5'-nucleotidase activities as well as chemiluminescence emission were measured in peritoneal macrophages of Syrian hamsters in the growth process of tumours with different grade of malignancy. The adenosine deaminase activity was established to decrease, while the 5'nucleotidase activity--to increase in macrophages after the subcutaneous injection of tumour cells with high level of malignancy as compared with these values in normal cells. This is accompanied by a decrease of the macrophage chemiluminescence during the whole experimental period. At the same time adenosine deaminase and 5'-nucleotidase activities as well as chemiluminescence emission in peritoneal macrophages of hamsters treated with low-malignancy cells do not differ from these values in the control group.     

Effects of microwave exposure on the hamster immune system. II. Peritoneal macrophage function

Acute exposure to hamsters to microwave energy (2.45 GHz; 25 mW/cm2 for 60 min) resulted in activation of peritoneal macrophages that were significantly more viricidal to vaccinia virus as compared to sham-exposed or normal (minimum-handling) controls. Macrophages from microwave-exposed hamsters became activated as early as 6 h after exposure and remained activated for up to 12 days. The activation of macrophages by microwave exposure paralleled the macrophage activation after vaccinia virus immunization. Activated macrophages from vaccinia-immunized hamsters did not differ in their viricidal activity when the hamsters were microwave- or sham-exposed. Exposure for 60 min at 15 mW/cm2 did not activate the macrophages while 40 mW/cm2 exposure was harmful to some hamsters. Average maximum core temperatures in the exposed (25 mW/cm2) and sham groups were 40.5 degrees C (+/- 0.35 SD) and 38.4 degrees C (+/- 0.5 SD), respectively. In vitro heating of macrophages to 40.5 degrees C was not as effective as in vivo microwave exposure in activating macrophages to the viricidal state. Macrophages from normal, sham-exposed, and microwave-exposed hamsters were not morphologically different, and they all phagocytosed India ink particles. Moreover, immune macrophage cytotoxicity for virus-infected or noninfected target cells was not suppressed in the microwave-irradiated group (25 mW/cm2, 1 h) as compared to sham-exposed controls, indicating that peritoneal macrophages were not functionally suppressed or injured by microwave hyperthermia.     

Changes of expression of peritoneal macrophages Fc and C3 receptors induced by transplantable melanomas

Expression of Fc and C3 receptors was studied in the rosette tests on isolated peritoneal macrophages of control and melanoma-bearing hamsters. In hamsters with transplanted melanomas an increase of the percentage of macrophages with Fc and C3 receptor expression was observed. The increase was prominent among macrophages from animals with transplanted amelanotic melanoma, a tumor line with greater malignancy and changed antigenicity and immunogenicity.     

Interleukin-8 has antitumor effects in the rat which are not associated with polymorphonuclear leukocyte cytotoxicity

The antitumor effect of lipopolysaccharides (LPS) has been observed in several experimental models and is likely to be mediated by macrophages. Stimulation of macrophages with LPS results in the release of several cytokines, including tumor necrosis factor, interleukin-1 and neutrophil-activating peptide-1/interleukin-8 (IL-8), which activates polymorphonuclear leukocytes (PMN) in vitro. Since PMN have an antitumor activity, we tested the in vivo effect of IL-8 on the growth of peritoneal carcinomatoses induced by PROb colon cancer cells in syngeneic rat. IL-8 induced a significant regression of tumors measuring 1–5 mm, and a complete regression was observed in 8 out of 40 rats in four independent experiments. IL-8 was not directly cytotoxic in vitro for tumor cells and was effective in vivo in a narrow range of doses. IL-8 had a significant chemotactic effect for peritoneal PMN in both normal and tumor-bearing rats. PMN taken from the peritoneum of tumor-bearing rats during IL-8 treatment had the same cytotoxic activity against PROb tumor cells as PMN from untreated control rats. Microscopic examinations of tumors during the treatment showed poor infiltrating by PMN. We conclude that the antitumor activity of IL-8 in this model is not mediated by PMN cytotoxicity.     

Locoregional therapy with polyethylene-glycol-modified interleukin-2 of an intradermally growing hepatocellular carcinoma in the guinea pig induces T-cell-mediated antitumor activity

Therapy with repeated intratumoral and perilymphatic administration of relatively low doses of polyethylene-glycol(PEG)-modified interleukin-2 (IL-2) in the syngeneic guinea pig line 10 (L10) hepatocarcinoma results in significant local tumor growth inhibition and a delay in development of regional lymph node metastases of more than 3 weeks when compared to controls. Occasionally animals are cured of tumor. The mechanism of this antitumor activity was studied. The antitumor activity of locoregionally administered PEG-IL-2 was abrogated by pretreatment with polyclonal anti-thymocyte serum, indicating that the observed tumor growth inhibition was a T-cell-mediated phenomenon. Besides the locoregional tumor growth inhibition, a systemic effect was recorded as the growth of a second tumor cell inoculum at the contralateral side was inhibited as well. Furthermore, those animals cured after PEG-IL-2 therapy developed specific immunity against the L10 tumor and this immunity could be transferred to naive animals by spleen cells. Immunohistological observations of the tumor site revealed a slight increase of helper and cytotoxic T cell subpopulations after PEG-IL-2 therapy. More pronounced, however, was the rise in number of eosinophilic granulocytes present in the stroma surrounding the tumor cells. Involvement of cytotoxic cells in the antitumor effects of PEG-IL-2 could not be demonstrated: regional lymph node cells and spleen cells obtained immediately after therapy (day 15) or on day 21 showed no cytotoxic activity in vitro against L10, K562, Daudi and line 1 (L1) target cells.In conclusion, locoregional therapy with PEG-IL-2 induced a systemic T-cell-mediated antitumor response. As no cytotoxic T cell activity was measured, however, the underlying mechanism is most likely a T-helper response. Eosinophils at the tumor site may be tumoricidal but further experiments must reveal the role of these cells in the PEG-IL-2-induced tumor regression.