Augmentation of host antitumor immunity by low doses of cyclophosphamide and mafosfamide in two animal tumor models

Critical Review of Preclinical Approaches to Evaluate the Potential of Immunosuppressive Drugs to Influence Human Neoplasia

Many immunosuppressive drugs are associated with an increased risk of B-cell lymphoma, squamous cell carcinoma, and Kaposi sarcoma. Thirteen immunosuppressive drugs have been tested in 2-year carcinogenicity studies (abatacept; azathioprine; busulfan; cyclophosphamide; cyclosporine; dexamethasone; everolimus; leflunomide; methotrexate; mycophenolate mofetil; prednisone; sirolimus; and tacrolimus) and in additional models including neonatal and genetically modified mice; chemical, viral, ultraviolet, and ionizing radiation co-carcinogenesis, and in models with transplanted tumor cells. The purpose of this review is to outline the mechanisms by which immunosuppressive drugs can influence neoplasia, to summarize the available preclinical data on the 13 drugs, and to critically review the performance of the models. A combination of primary tumor and metastasis assays conducted with transplanted cells may provide the highest value for hazard identification and can be applied on a case-by-case basis. However, for both small molecules and therapeutic proteins, determining the relative risk to patients from preclinical data remains problematic. Classifying immunosuppressive drugs based on their mechanism of action and hazard identification from preclinical studies and a prospective pharmacovigilance program to monitor carcinogenic risk may be a feasible way to manage patient safety during the clinical development program and postmarketing.

Cytosine arabinoside induces costimulatory molecule expression in acute myeloid leukemia cells

Chemotherapeutic drugs kill cancer cells mainly by direct cytotoxicity, but they might also induce a stronger host immune response by causing the tumor to produce costimulatory cell surface molecules like CD80. We previously reported that in myeloid leukemic cells, gamma-irradiation induced CD80 expression. In this study, we show that cytosine arabinoside (Ara-C), even at low doses, induced CD80 expression in vitro in mouse DA1-3b leukemic cells, by a mechanism that involved reactive oxygen species. In vivo experiments in the mouse DA1-3b/C3H whole-animal acute myeloid leukemia (AML) model showed that injection of Ara-C induced expression of CD80 and CD86, and decreased expression of B7-H1, indicating that chemotherapy can modify costimulatory molecule expression in vivo, in a way not necessarily observed in vitro. Mouse leukemic cells exposed in vivo to Ara-C were more susceptible to specific cytotoxic lymphocyte (CTL)-mediated killing. Ara-C also induced CD80 or CD86 expression in 14 of 21 primary cultured human AML samples. In humans being treated for AML, induction chemotherapy increased CD86 expression in the leukemic cells. These findings indicate possible synergistic strategies between CTL-based immunotherapy and chemotherapy for treatment. They also suggest an additional mechanism by which chemotherapy can eradicate AML blasts.

Synergistic anti-tumor activity of a novel immunomodulator, BCH-1393, in combination with cyclophosphamide

N,N-dimethylaminopurine pentoxycarbonyl D-arginine (BCH-1393) is a novel low molecular weight synthetic immunomodulator that has been shown to significantly stimulate cytotoxic T-lymphocyte responses both in vitro and in vivo (Zacharie B, Gagnon L, Attardo G, Connolly TP, St-Denis Y, Penney CL. Synthesis and activity of 6-substituted purine linker amine immunostimulants. J. Med. Chem. 1997;40:2883-94). Prompted by this evidence, we extended evaluation of BCH-1393 for anticancer activity in syngeneic mouse experimental tumor models. Consistent with previous findings, in vitro assessment of BCH-1393 activity demonstrated a significant increase in the CTL responses in the range of 10(-9)-10(-5) M. Treatment of mice with four consecutive daily intraperitoneal injections at 25 and 50 mg/kg resulted in a significant increase of the relative percentage of blood CD4+, CD8+, NK and monocyte subsets without any evidence of toxicity. In vivo anti-tumor activity of BCH-1393 was evaluated, either alone or in combination with subtherapeutic doses of cyclophosphamide (Cy), against weakly immunogenic mouse breast carcinoma DA-3 and strongly immunogenic colon adenocarcinoma MC38. Daily intraperitoneal injection of BCH-1393 at 50 mg/kg alone was well tolerated but produced a relatively weak anti-tumor effect in both tumor models. However, a significant inhibition of tumor outgrowth and suppression of established tumor growth was observed when BCH-1393 was administered in combination with subtherapeutic doses of Cy. Combination treatment of 50 mg/kg BCH-1393 with 100 mg/kg Cy (given as single intravenous bolus injection) starting 2 days prior to DA-3 tumor cell inoculation prevented tumor outgrowth in 70-80% of treated mice. In the remaining 20-30% of mice that had developed tumors, a nearly complete (90%) tumor growth inhibition was observed at days 22-24 post tumor implant. In the MC38 tumor model, combination treatment of established tumors with BCH-1393 and Cy (CTX) at 50 mg/kg resulted in a significant delay in tumor growth compared to CTX treatment alone. The observed concomitant anti-tumor activity of BCH-1393 with cyclophosphamide warrants further investigation of this immunomodulator as an adjunctive treatment of cancer.

Cyclophosphamide induces type I interferon and augments the number of CD44hi T lymphocytes in mice: implications for strategies of chemoimmunotherapy of cancer

In a previous study, we reported that a single injection of cyclophosphamide (CTX) in tumor-bearing mice resulted in tumor eradication when the animals were subsequently injected with tumor-sensitized lymphocytes. Notably, CTX acted by inducing bystander effects on T cells, and the response to the combined CTX/adoptive immunotherapy regimen was inhibited in mice treated with antibodies to mouse interferon (IFN)–/β. In the present study, we have investigated whether CTX induced the expression of type I IFN, and we have characterized the CTX effects on the phenotype of T cells in normal mice. CTX injection resulted in an accumulation of type I IFN messenger RNA in the spleen of inoculated mice, at 24 to 48 hours, that was associated with IFN detection in the majority of the animals. CTX also enhanced the expression of the Ly-6C on spleen lymphocytes. This enhancement was inhibited in mice treated with anti–type I IFN antibodies. Moreover, CTX induced a long-lasting increase in in vivo lymphocyte proliferation and in the percentage of CD44hiCD4+ and CD44hiCD8+T lymphocytes. These results demonstrate that CTX is an inducer of type I IFN in vivo and enhances the number of T cells exhibiting the CD44hi memory phenotype. Since type I IFN has been recently recognized as the important cytokine for the in vivo expansion and long-term survival of memory T cells, we suggest that induction of this cytokine may explain at least part of the immunomodulatory effects observed after CTX treatment. Finally, these findings provide a new rationale for combined treatments with CTX and adoptive immunotherapy in cancer patients.

Importance of cyclophosphamide-induced bystander effect on T cells for a successful tumor eradication in response to adoptive immunotherapy in mice

Cyclophosphamide (CTX) increases the antitumor effectiveness of adoptive immunotherapy in mice, and combined immunotherapy regimens are now used in some clinical trials. However, the mechanisms underlying the synergistic antitumor responses are still unclear. The purpose of this study was (a) to evaluate the antitumor response to CTX and adoptive immunotherapy in mice bearing four different syngeneic tumors (two responsive in vivo to CTX and two resistant); and (b) to define the mechanism(s) of the CTX-immunotherapy synergism. Tumor-bearing DBA/2 mice were treated with a single injection of CTX followed by an intravenous infusion of tumor-immune spleen cells. In all the four tumor models, a single CTX injection resulted in an impressive antitumor response to the subsequent injection of spleen cells from mice immunized with homologous tumor cells independently of the in vivo response to CTX alone. Detailed analysis of the antitumor mechanisms in mice transplanted with metastatic Friend leukemia cells revealed that (a) the effectiveness of this combined therapy was dependent neither on the CTX-induced reduction of tumor burden nor on CTX-induced inhibition of some putative tumor-induced suppressor cells; (b) the CTX/immune cells' regimen strongly protected the mice from subsequent injection of FLC, provided the animals were also preinoculated with inactivated homologous tumor together with the immune spleen cells; (c) CD4(+) T immune lymphocytes were the major cell type responsible for the antitumor activity; (d) the combined therapy was ineffective in mice treated with antiasialo-GM1 or anti-IFN-alpha/beta antibodies; (e) spleen and/ or bone marrow cells from CTX-treated mice produced soluble factors that assisted in proliferation of the spleen cells. Altogether, these results indicate that CTX acts via bystander effects, possibly through production of T cell growth factors occurring during the rebound events after drug administration, which may sustain the proliferation, survival, and activity of the transferred immune T lymphocytes. Thus, our findings indicate the need for reappraisal of the mechanisms underlying the synergistic effects of CTX and adoptive immunotherapy, and may provide new insights into the definition of new and more effective strategies with chemotherapy and adoptive immunotherapy for cancer patients.

IL-15 mediates anti-tumor effects after cyclophosphamide injection of tumor-bearing mice and enhances adoptive immunotherapy: the potential role of NK cell subpopulations

The daily administration of IL-15 to cyclophosphamide (CY)-injected mice bearing the 76-9 rhabdomyosarcoma was shown to prolong the period of remission induced by CY. In addition, IL-15 was shown to enhance the efficacy of adoptive immunotherapy. Cytotoxicity assays using spleens from normal and tumor-bearing mice indicated that IL-15 enhanced NK cell activity but there was no evidence for class I-restricted cytolytic T cell activity. To determine whether IL-15 was likely to induce different cytotoxic effectors at the tumor site compared with the spleen, tumors were removed after CY injection and cell suspensions were incubated with IL-15 in parallel with isolated spleen cells. Both populations were seen to expand to yield predominantly cells coexpressing NK1.1 and B220 antigens. However, tumor-associated NK cells were shown to differ from expanded spleen NK cells in terms of the proportions of LGL-1+ cells and cells expressing early and late NK cell differentiation antigens. Both expanded populations expressed high NK cell cytotoxic activity but only the spleen cells expressed lymphocyte-activated killer cell activity. It was apparent that the expanded tumor-associated NK cells expressed low-level class I-restricted lytic activity. The potential of activated NK cells in the circulation to exert anti-tumor effects was shown by the adoptive transfer of expanded NK cells to tumor-bearing mice after CY injection when significant prolongation of life was seen in all cases. The data indicate that IL-15 may serve as a useful anti-cancer adjuvant by activating initially the NK cell arm of the immune network.

A cell-surface epitope associated with liver-preferential metastasis detected by the new monoclonal antibody 3H4 in the murine tumor model ER 15-P

In the tumor model ER 15-P, a chemically induced pleomorphic myofibrosarcoma of the C57/Bl6J mouse, cell lines with liver-preferential metastatic tumor spread were selected in vivo. In order to describe cell-surface molecules relevant for hepatic metastasis, monoclonal antibodies were raised against the liver-preferential variants. In a syngeneic immunization with viable tumor cells cyclophosphamide was used for augmentation of the humoral antitumor immunity. The monoclonal antibody mAb 3H4, an IgG2b isotype, reacted with a cell-surface epitope exclusively detected on the liver-preferential metastatic phenotype (Me) of the tumor model ER 15-P; no reactivity with the non-organ-specific metastatic phenotype (P) was observed. Regarding the morphological heterogeneity of different Me and P tumor cell populations, mAb 3H4 antigen expression was consistently associated with liver-preferential metastasis, not with different morphological stages of differentiation. It showed no cross-reaction with other tumor cell lines tested except MethA murine fibrosarcoma. The antibody was unreactive with normal tissue cells in C57/Bl6J mice. mAb 3H4 antigen expression was not dependent on the cell cycle. In an experimental assay of hematogenous metastasis, preincubation with mAb 3H4 significantly reduced the number of liver metastases of the liver-preferential tumor cells. Although no crossreaction of the primary ER 15-P with mAb 3H4 was observed, the antibody also significantly reduced the number of renal metastases of the P tumor cell population. The syngeneic IgG2b monoclonal antibody mAb 3H4 identified a new tumor-associated cell-surface antigen correlating with liver-preferential metastasis. mAb 3H4 antigen expression was a stable property of the liver-preferential tumor cells regardless of morphological diversity or functional cell status. In an in vivo blocking assay mAb 3H4 reduced liver colonization in vivo.

Restoration of interleukin-2 production in tumor-bearing rats through reducing tumor-derived transforming growth factor beta by treatment with bleomycin

We studied mechanisms of immunosuppression caused by tumor-derived transforming growth factor-beta (TGF beta) and restoration of the immune response by treatment with bleomycin in rats bearing KDH-8 hepatoma. Interleukin-2 (IL-2) production from splenocytes of KDH-8-tumor-bearing rats progressively decreased as the KDH-8 tumor grew. IL-2 production from concanavalin-A-stimulated normal rat splenocytes was significantly inhibited by in vitro cultured KDH-8-tumor-cell-conditioned medium; this inhibition could be blocked by neutralizing the conditioned medium with anti-TGF beta antibody. TGF beta activities were found in KDH-8-tumor-tissue-conditioned medium without acid treatment and were found in tumor-cell-conditioned medium after acid treatment; TGF beta mRNA and TGF beta protein were found in cultured KDH-8 tumor cells. These results suggested that the KDH-8-tumor-derived TGF beta might be involved in the inhibition of IL-2 production from splenocytes. To determine whether bleomycin chemotherapy could reduce tumor-derived TGF beta and restore the immune responses, we treated KDH-8 tumor-bearing rats with bleomycin (5 mg/kg, one shot) at an appropriate time (before the occurrence of immunosuppression) resulting in a significant reduction of TGF beta activity in KDH-8 tumor tissues and restoration of IL-2 production from splenocytes of tumor-bearing rats; KDH-8 tumor growth ultimately regressed. In vitro experiments also showed that TGF beta activity, mRNA expression, and protein synthesis in KDH-8 tumor cells were reduced by bleomycin treatment, and that bleomycin-treated-KDH-8-tumor-cell-conditioned medium did not inhibit IL-2 production from normal rat splenocytes. These results suggest that bleomycin treatment restored IL-2 production in tumor-bearing rats through reducing the tumor-derived TGF beta.

Correlation between the sensitivity or resistance to IL-2 and the response to cyclophosphamide of 4 tumors transplantable in the same murine host

We have studied the anti-tumor response to cyclophosphamide (CTX) in DBA/2 mice transplanted s.c. with 4 tumors exhibiting different responses to IL-2: ESb lymphoma and Friend leukemia cells (non-responsive or poorly responsive, respectively), p11-R-Eb and Eb lymphoma cells (both highly responsive to IL-2). CTX injections on days 7, 14 and 21 resulted in a significant anti-tumor response in mice transplanted s.c. with Friend leukemia cells or ESb cells, whereas no anti-tumor effect was observed in mice injected with Eb or p11-R-Eb cells. All 4 tumor cell lines were equally sensitive to the cytotoxic effects of mafosfamide, an in vitro active analogue of CTX. To define the host mechanisms responsible for the lack of an anti-tumor effect of CTX in mice transplanted with IL-2-responsive tumors, we studied several aspects of the spontaneous or IL-2-induced anti-tumor response in mice transplanted with p11-R-Eb cells. Injection of monoclonal antibodies (MAbs) to IFN-gamma completely abolished the anti-tumor effects of IL-2. Using a Winn assay, clear-cut anti-tumor activity was found in spleen cells from mice transplanted with the IL-2-responsive tumors. This activity was abolished by CTX, which also abrogated the anti-tumor response to IL-2 in mice injected with p11-R-Eb cells. Our results indicate an inverse correlation between sensitivity to IL-2 and response to CTX and emphasize the importance of initial host-tumor interaction in determining the type of response to IL-2 or CTX.

Mafosfamide induces DNA fragmentation and apoptosis in human T-lymphocytes. A possible mechanism of its immunosuppressive action

Cyclophosphamide, an alkylating agent belonging to the family of nitrogen mustards, is commonly used to treat progressive autoimmune diseases in humans. At the molecular level, its cytotoxicity results from DNA double strand crosslinks and, at higher concentrations, from DNA strand breaks. At the cellular level, cyclophosphamide may selectively affect mature lymphocytes with relative sparing of the respective precursor cells. In this study, we show that 4-hydroxycyclophosphamide (4-OH-CP), the active metabolite of cyclophosphamide, induces apoptosis in mature human lymphocytes at concentrations that are achieved in vivo. Since cyclophosphamide requires enzymatic conversion in the liver to yield its active metabolite, 4-OH-CP was generated in vitro by non-enzymatic hydrolysis of mafosfamide. Apoptotic cell death of lymphocytes was characterized by typical morphological changes, nucleosomal DNA fragmentation, and quantified by 3'-OH end labeling of fragmented DNA. The percentage of apoptotic cells both depended on drug concentration and time of exposure. Cycloheximide or ZnSO4 did not suppress 4-OH-CP induced apoptosis. Etoposide, a topoisomerase II inhibitor known to induce apoptosis in human tumor cell lines like 4-OH-CP, did induce detectable DNA fragmentation in only a minor proportion of T-lymphocytes but suppressed T-cell proliferation.

Combination of immunotherapy with cyclophosphamide/actinomycin D chemotherapy potentiates antileukemic effect and reduces toxicity in a L1210 leukemia model in mice

The therapeutic effects of the combination of chemotherapy (cyclophosphamide and actinomycin D) and immunotherapy (TNF-alpha and macrophages) were evaluated on L1210 leukemia in mice. When given as single agents, both cyclophosphamide (CY), administered intraperitoneally 2 days after subcutaneous inoculation of leukemic cells, and actinomycin D (Act D), injected intratumorally (i.t.) 4 days following injection of leukemic cells, exerted therapeutic effects and prolonged mice survival. Unexpectedly, combination of CY and Act D did not result in prolongation of mice survival, due mainly to substantial cumulative toxic effects that led to death in several cases. Immunotherapy with TNF-alpha and M phi, injected i.t. on day 4 following inoculation of leukemic cells, did not give significant therapeutic effect, either when used alone or when used in conjunction. However, combination of chemotherapy and immunotherapy, including all four agents, produced a beneficial effect resulting in significant prolongation of the survival of leukemia-bearing mice. This study indicates the potential of appropriate combinations of cytotoxic drugs with immunotherapy against neoplasia.

Some approaches to improve the therapeutic effectiveness of adoptive chemoimmunotherapy with spleen cells from melphalan-treated BALB/c mice bearing a large MOPC-315 tumor

Spleen cells from BALB/c mice that are in the process of eradicating a large MOPC-315 tumor following low-dose L-PAM therapy (L-PAM TuB spleen cells) were previously shown to be capable of bringing about the complete regression of a large (15 to 20 mm) s.c. MOPC-315 tumor in a substantial percentage of T-cell-deficient (athymic nude) mice that had been treated with low-dose L-PAM (adoptive chemo-immunotherapy; ACIT). Here we show that aggressive depletion of CD4+T-cells by treatment both of spleen-cell donors and of recipients with anti-L3T4 monoclonal antibody (MAb) greatly improved the therapeutic effectiveness of L-PAM TuB spleen cells in ACIT. In the light of the apparent importance of cytotoxic T-lymphocytes (CTLs) for tumor eradication in low-dose L-PAM-treated MOPC-315-tumor bearers, it is interesting that treatment of L-PAM TuB spleen-cell donors with anti-L3T4 MAb was found to result in the generation of enhanced splenic anti-MOPC-315 cytotoxicity. Although most athymic nude mice in which the tumor had apparently completely regressed following ACIT remained tumor-free, approximately 1/3 of the mice relapsed. However, a substantial percentage of the relapsing mice were rescued by a low dose of L-PAM, which was not effective in causing tumor regression in athymic nude mice bearing a comparably large tumor if the mice had not been subjected previously to ACIT. Almost all athymic nude mice that had been "cured" of a large MOPC-315 tumor by ACIT but did not resist a subsequent MOPC-315 tumor challenge were rescued by low dose L-PAM. Thus, the therapeutic effectiveness of L-PAM TuB spleen cells in ACIT may be improved by aggressive depletion of CD4+ T-cells, suggesting that a low dose of L-PAM, which leads to the acquisition of potent splenic-tumor-eradicating immunity in BALB/c mice bearing a large MOPC-315 tumor, does not eliminate completely (or possibly not at all) the inhibitory activity of CD4+ T-cells. In addition, athymic nude mice that are not endowed with fully protective tumor-eradicating immunity following ACIT still have a substantial residual anti-tumor immune potential that can be exploited to bring about eradication of a large tumor burden following low-dose L-PAM therapy.

Enhancement of cytotoxic T lymphocyte growth from spleens of P815-tumor-bearing host mice with mafosfamide

Mafosfamide (Mafo) is an analog of cyclophosphamide that does not require hepatic activation and therefore has in vitro activity. The present study was conducted to determine the effects of in vitro treatment with Mafo on the generation and growth of cytotoxic T lymphocytes (CTL) from tumor-bearing host mice (TBH). In contrast to early (day-11) TBH splenocytes, splenocytes from late (days 18-20) P815 TBH mice suppress the in vitro generation of CTL. Treatment of late TBH splenocytes in vitro with 5-15 microM Mafo resulted in a reduced ability of these cells to suppress in vitro CTL generation. Treatment of late TBH splenocytes with 10 microM Mafo also inhibited their ability to suppress adoptive immunotherapy of intradermal tumors with immune splenocytes. These doses of Mafo were selectively toxic to the suppressive effects of late TBH splenocytes, since treatment of early TBH splenocytes with 1-10 microM Mafo did not significantly inhibit CTL generation. Spleen cells from early (days 10-12) TBH mice, carried in long-term in vitro sensitization cultures in the presence of tumor cells and 20 U/ml human recombinant interleukin-2, did not increase in cell number over time. However, when pretreated with 3 microM Mafo, this population of tumor-sensitized lymphocytes demonstrated 450-fold growth over 6 weeks as compared to the static cell numbers for the untreated controls. High levels of tumor-specific cytolytic activity were maintained in these expanded cells. These results suggest that Mafo pretreatment markedly and selectively inhibits suppressor cells that limit long-term expansion of splenic CTL in culture and inhibit adoptive immunotherapy of solid tumors.

Enhanced expansion of the thymic CD8+ cell subset as a potential mechanism for the generation of enhanced antitumor cytotoxicity by thymocytes from low-dose melphalan-treated MOPC-315 tumor bearers

We have previously shown that thymocytes from low-dose melphalan (L-phenylalanine mustard)-treated MOPC-315-tumor-bearing mice (melphalan TuB) are able to generate an enhanced level of anti-MOPC-315 cytotoxicity, as compared to thymocytes from untreated MOPC-315-tumor-bearing mice or thymocytes from untreated or low-dose melphalan-treated normal mice, upon in vitro stimulation with MOPC-315 tumor cells in the presence of a low concentration of recombinant interleukin-2 (rIL-2). Here we show that the generation of enhanced anti-MOPC-315 cytotoxicity by melphalan TuB thymocytes depends on the ability of the thymocytes to proliferate. In addition, the ability of melphalan TuB thymocytes to generate an enhanced level of anti-MOPC-315 cytotoxicity correlated with their ability to proliferate more readily than thymocytes from untreated tumor-bearing mice and thymocytes from untreated or melphalan-treated normal mice in response to stimulation with MOPC-315 tumor cells plus a low concentration of rIL-2. Moreover, although fresh melphalan TuB thymocytes do not contain a higher percentage of phenotypically mature cells (i.e., CD4-/CD8+ or CD4+/CD8-) than do thymocytes from normal mice or untreated tumor-bearing mice, after a 5-day culture with both MOPC-315 tumor cells and a low concentration of rIL-2, cultures of thymocytes from melphalan TuB contained a much higher percentage of CD4-/CD8+ (but not CD4+/CD8-) cells than did cultures of thymocytes from the other two sources. Since CD4-/CD8+ cells were previously shown to be responsible for the exertion of antitumor cytotoxicity by thymocytes stimulated with MOPC-315 in vitro, our results indicate that the enhanced antitumor cytotoxicity exerted by melphalan TuB thymocytes following in vitro stimulation with MOPC-315 tumor cells in the presence of a low concentration of rIL-2 is due, at least in part, to an expansion of the pool of CD4-/CD8+ effector cells.

Eradication of a large MOPC-315 tumor in athymic nude mice by chemoimmunotherapy with Lyt2+ splenic T cells from melphalan-treated BALB/c mice bearing a large MOPC-315 tumor

We have previously shown that spleen cells from BALB/c mice that are in the process of eradicating a large MOPC-315 tumor following low-dose (2.5 mg/kg) melphalan (L-phenylalanine mustard) therapy are effective in preventing tumor progression upon adoptive transfer into BALB/c mice bearing a barely palpable tumor that had been treated with a subcurative dose of melphalan [Mokyr et al. (1989) Cancer Res 49:4597]. Here we show that such spleen cells in conjunction with a subcurative dose of drug (adoptive chemoimmunotherapy, ACIT) can cause the complete regression of a large (15-20 mm) s.c. MOPC-315 tumor in a large percentage of T-cell-deficient (athymic nude) tumor-bearing mice. Spleen cells that were effective in ACIT of athymic nude mice displayed in vitro a substantial direct lytic activity against MOPC-315 tumor cells, and the lytic activity was greatly enhanced when the spleen cells were cultured for 5 days with or without mitomycin-C-treated MOPC-315 stimulator tumor cells. The cells responsible for the therapeutic effectiveness of the spleen cells in ACIT of athymic nude mice, as well as the cells responsible for the direct in vitro anti-MOPC-315 lytic activity of the spleen cells, were of the Lyt 2 and not the L3T4 phenotype. Most of the athymic nude mice that completely eradicated a large MOPC-315 tumor as a consequence of ACIT were capable of rejecting a challenge with 30-100 times the minimal lethal tumor dose for 100% of normal BALB/c mice administered more than 1 month after the ACIT. The ability of these athymic nude mice to resist the tumor challenge was associated with the presence of a greatly elevated percentage of cells expressing T cell surface markers in their spleens. Thus, it is conceivable that splenic Lyt 2+T cells from melphalan-treated BALB/c mice bearing a large MOPC-315 tumor mediate their therapeutic effectiveness in ACIT of athymic nude mice bearing a large MOPC-315 tumor, at least in part, through direct cytotoxicity for MOPC-315 tumor cells. In addition, eradication of a large MOPC-315 tumor through cooperation between antitumor immunity and melphalan toxicity endues the athymic nude mice with an elevated percentage of T cells in their secondary lymphoid organs, and these T cells are probably responsible for the long-lasting protective antitumor immunity exhibited by these mice.

Possible role of acrolein in oxazaphosphorine-induced enhancement of immunological reactivity

The aim of the present study was to analyze further the immunopotentiating effects of low doses of oxazaphosphorines. We examined 4-hydroperoxycyclophosphamide (4-HC) and mafosfamide, which degrade spontaneously in water without requiring liver enzymes to become active. Both drugs, at concentrations ranging from 0.01 microM to 1 microM, enhanced mitogenic responses of human lymphocytes. Higher concentrations were toxic. Acrolein, which is one of the degradation products of oxazaphosphorines, had similar effects. Immunopotentiation was not monocyte-dependent. Attempts to inactivate released acrolein with human serum reduced toxicity but the immunostimulating property of the drugs remained Similar effects were noted when lymphocytes were exposed to acrolein dissolved in serum. 2-Mercaptoethane-sulfonate (mesna), which is highly reactive with acrolein, reduced the toxicity of solutions of both oxazaphosphorines and acrolein. Immunopotentiation was not clearly demonstrable since mesna itself enhanced the responses. Pretreatment of lymphocytes with 4-HC or mafosfamide did not reduce the capacity of concanavalin A to induce suppressor cells. It is speculated that acrolein may play a role in oxazaphosphorine-induced enhancements of immune responses.

Evidence of a role for NK cells in oxazaphosphorine-mediated tumor regression

The present studies showed that nude mice xenotransplanted with L5222 leukemia responded as did syngeneic BD IX rats to low doses of mafosfamide or cyclophosphamide. Unlike rats, nude mice rarely showed resistance to a second tumor challenge. The observation that concurrent treatment of rats with cyclosporin A did not alter the rate of survival clearly indicated a T-cell-independent mechanism of tumor defense. The incidence of lung colonies from i.v. injected Lewis lung-tumor cells could be enhanced by a high dose pretreatment with mafosfamide or cyclophosphamide, whereas pretreatment at low doses was inhibitory. Since identical experiments carried out in NK-cell-deficient C57Bl/6 "beige" mice did not show such an effect, NK cells appeared to represent a possible effector cell in oxazaphosphorine-mediated antitumor effects. This assumption was further supported by the fact that enhanced NK cell activity could be observed in the 51Cr release assay using spleen cells from mafosfamide-treated L5222-bearing rats. The transplantation of the unrelated syngeneic ovarian carcinoma OV-342 to animals that had previously been cured of L5222 leukemia did not lead to the rejection of this tumor. This indicates that a specific resistance against L5222 leukemia had developed. In contrast, a T-cell-dependent antitumor effect was demonstrated for mafosfamide in the MOPC-315 mouse plasmocytoma. Therefore, we conclude that the effector cell for tumor rejection depends on the type of tumor. This, of course, does not exclude a common target cell for the immunopharmacological activity of oxazaphosphorines.