Anti-CD19 immunotoxin enhances the activity of chemotherapy in severe combined immunodeficient mice with human pre-B acute lymphoblastic leukemia.

The anti-CD19 immunotoxin (IT) (HD37-dgRTA) is effective in killing B-lineage leukemia cells and in curing severe combined immunodeficient mice with acute lymphoblastic leukemia. The present study aimed to identify effective combinations of HD37-dgRTA and chemotherapeutic agents. The in-vitro cytotoxicity assays demonstrate that the combination of HD37-dgRTA and either daunorubicin or vincristine is effective. The in-vivo experiments using HD37-dgRTA with vincristine prolonged the survival of mice compared to the chemotherapeutic agent or IT (90.7 vs. 147.1 days). Also, 80% of the mice treated with IT plus vincristine were long-term survivors.

A comparison of an anti-CD25 immunotoxin, Ontak and anti-CD25 microbeads for their ability to deplete alloreactive T cells in vitro.

Ex vivo depletion of alloreactive CD25(+) T cells from a stem cell transplant (SCT) can reduce the incidence of graft-versus-host disease (GVHD) while preserving antimicrobial and perhaps antileukemia activity. However, the most effective methods for allodepleting T cells prior to transplant have not been determined. In this study, we have compared three agents that deplete CD25(+) activated, alloreactive T cells. These included Ontak (Denileukin Diftitox), an IL-2 fusion toxin, anti-CD25 microbeads (MACS), an anti-CD25 immunotoxin (IT) and a combination of the IT and MACS. Peripheral blood mononuclear cells (PBMCs) activated in a primary mixed lymphocyte reaction (MLR) were allodepleted using optimal amounts of each agent, and the cells were then analyzed by flow cytometry. The treated cells were examined both for remaining alloreactivity and for the preservation of third party reactivity by testing them in a secondary MLR. Our data demonstrate that both the anti-CD25 IT and the anti-CD25 MACS were equally effective in depleting CD4(+)CD25(+) cells and in sparing T cells that were reactive with third party cells. The anti-CD25 IT was, however, superior in depleting alloreactive CD8(+)CD25(+) cells. In contrast, Ontak did not eliminate alloreactive cells and the Ontak-treated cells retained significant reactivity against the original stimulator cells.

Add-back of allodepleted donor T cells to improve immune reconstitution after haplo-identical stem cell transplantation.

Poor immune reconstitution after haplo-identical stem cell transplantation results in high mortality from viral infections and relapse. One approach to overcome this problem is to deplete alloreactive cells selectively by deleting T cells activated by recipient stimulators, using an immunotoxin directed against the activation marker CD25. However, the degree of depletion of alloreactive cells is variable following stimulation with recipient PBMC, and this can result in GvHD. We have shown that using recipient EBV-transformed LCL as stimulators to activate donor alloreactive T cells results in more consistent depletion of in vitro alloreactivity while preserving T-cell responses to viral and potential myeloid tumor Ag. Based on these data, we have embarked on a phase I clinical dose escalation study of add-back of allo-LCL-depleted donor T cells in the haplo-identical setting, to determine if the allodepletion we achieve to allow infusion of sufficient T cells to restore useful antiviral/anti-leukemic responses without causing GvHD. Fifteen patients have so far been treated. The incidence of significant acute or chronic GvHD has been low (2/15), as has mortality from infection (1/15). Preliminary data show accelerated immune reconstitution in dose level 2 patients. Infused allodepleted donor T cells appear able to expand significantly in the face of viral reactivations, and doses as low as 3 x 10(5)/kg may be sufficient to confer useful antiviral immunity in this setting. At a median follow-up of 19.5 months, nine of 15 patients are alive and disease-free. Five patients have relapsed, all of whom have died.

An anti-CD45RO immunotoxin kills HIV-latently infected cells from individuals on HAART with little effect on CD8 memory.

CD4+ CD45RO+ T cells are the major latent viral reservoir in HIV-infected individuals and hence a major obstacle in curing the disease. An anti-CD45RO immunotoxin (IT) can decrease the number of both productively and latently infected CD4+ T cells obtained from HIV-infected individuals with detectable viremia. In this study, we determined whether this IT could also kill latently infected replication-competent CD4+ T cells obtained from infected individuals without detectable plasma viremia. Our results demonstrate that ex vivo treatment with the anti-CD45RO IT significantly reduced the frequency of these cells. In contrast, the IT had only a modest effect on the cytomegalovirus-specific memory responses of CD8+ T cells. These results suggest that purging latent cells from infected individuals on highly active antiretroviral therapy with the anti-CD45RO IT might reduce the HIV latent reservoir without seriously compromising CD8+ T cell memory responses.

Clinical-scale selective depletion of alloreactive T cells using an anti-CD25 immunotoxin.

Allogeneic hematopoietic stem cell transplantation is the treatment of choice for many hematological malignancies. Its efficacy is limited by graft-versus-host disease (GVHD), the leading cause of post-transplant morbidity and mortality. GVHD is mediated by a subpopulation of T cells in the stem cell graft. Ex vivo T cell depletion of all T cells of the graft can prevent development of GVHD but can lead to a delay in immune reconstitution and an increase of potentially lethal opportunistic infections and leukemic relapses. Hypothetically, an approach that enables a selective depletion of the alloreactive donor T cells that cause GVHD while preserving third party (anti-leukemic and anti-microbial) reactivity would be optimal for recipients of HSCT. Our preliminary data demonstrated that an anti-CD25 immunotoxin, which reacts with a cell surface activation antigen, can selectively deplete alloreactive donor T cells activated by non-leukemic recipient white blood cells while preserving the beneficial third-party reactivity in vitro. In this report we describe a method for clinical-scale ex vivo selective depletion of alloreactive donor T cells using the anti-CD25 immunotoxin, RFT5-SMPT-dgRTA. Two logs of alloreactive T cells could be selectively depleted while preserving third party reactivity. This method was reproducible in 10 pre-clinical experiments with 8 HLA-mismatched healthy volunteer pairs and 2 HLA-matched sibling donor/patient pairs.

Clinical evaluation of ricin A-chain immunotoxins in patients with Hodgkin's lymphoma.

BACKGROUND: Immunotoxins (ITs) consist of cell binding ligands coupled to toxins or their subunits. Hodgkin's lymphoma (HL) is an excellent target for ITs since lymphocyte activation markers such as CD25 and CD30 are expressed in large numbers. The ITs RFT5.dgA (anti CD25) and Ki-4.dgA (anti CD30) were constructed by linking the monoclonal antibodies RFT5 and Ki-4 to deglycosylated ricin A-chain (dgA). Both ITs showed potent specific activity against HL cells in vitro and in vivo in animal models, and were subsequently evaluated in phase I/II clinical trials in humans. PATIENTS AND METHODS: In two separate trials, the ITs were administered i.v. four times every other day over 4 h. The objectives of the phase I trials included the determination of the maximum tolerated dose (MTD), dose-limiting toxicities (DLTs), pharmacokinetics, antitumor activity and immune response against the IT. RESULTS: Twenty-seven patients with refractory HL were included in the phase I/II study of RFT5.dgA and 17 patients were included in the phase I study of Ki-4.dgA. The MTD of RFT5.dgA was 15 mg/m(2), whereas that of Ki-4.dgA was 5 mg/m(2). DLTs were related to vascular leak syndrome, consisting of edema, tachycardia, dyspnea, weakness and myalgia. Measurement of serum levels of RFT5.dgA demonstrated a C(max) of 0.2-9.7 micro g/ml with a half-life (t()) varying from 4 to 10.5 h. Peak serum concentration of Ki-4.dgA ranged from 0.23 to 1.7 micro g/ml. In both trials approximately 60% of patients developed human anti-mouse and/or anti-dgA antibodies. Seventeen of 18 patients treated at the MTD of RFT5.dgA were evaluable for clinical response. Responses included two partial remissions (PR), one minor response (MR) and five stable diseases (SD). Fifteen of 17 patients treated with Ki-4.dgA were evaluable for clinical response. Responses included one PR, one MR and two SD. CONCLUSIONS: RFT5.dgA and Ki-4.dgA showed moderate efficacy in heavily pretreated refractory patients with HL. Ki-4.dgA was less well tolerated than RFT5.dgA. This might be due, at least in part, to the formation of Ki-4.dgA/sCD30 complexes.

Treatment of SCID/human B cell precursor ALL with anti-CD19 and anti-CD22 immunotoxins.

The anti-CD19 (HD37-dgRTA) and anti-CD22 (RFB4-dgRTA) immunotoxins (ITs) are murine IgG(1) monoclonal antibodies (Mabs) conjugated to a deglycosylated ricin A chain (dgRTA). They are effective in killing B-lineage non-Hodgkin's lymphoma (NHL) cells in vitro, in vivo and in adult patients with B-lineage NHL. The potential of these agents for the treatment of childhood B-precursor acute lymphoblastic leukemia (ALL) is unknown. The anti-CD19 and anti-CD22 ITs should have anti-tumor activity against childhood B-lineage ALL since both target antigens are expressed on the surface of these cells. We have previously shown that, in vitro these two ITs selectively kill leukemia cells obtained from children with leukemia. To evaluate the efficacy of our ITs in an in vivo model we injected the human pre-B ALL cell line, NALM-6-UM1, into severe combined immunodeficient (SCID) mice. We tested the ability of two ITs to prolong survival or cure mice of both early and advanced tumors. In early disease, treatment with HD37-dgRTA, RFB4-dgRTA, or Combotox (an equimolar concentration of the two ITs) significantly improved their survival. In advanced disease, treatment with RFB4-dgRTA or Combotox significantly improved survival. Overall there were 10 long-term survivors who were cured, as determined by survival beyond 150 days with no evidence of disease as determined by polymerase chain reaction (PCR) analysis.

Definitive separation of graft-versus-leukemia- and graft-versus-host-specific CD4+ T cells by virtue of their receptor beta loci sequences.

Although graft-versus-host (GVH) disease (GVHD) is usually associated with graft versus leukemia (GVL), GVL can occur in the absence of clinical GVHD. There is evidence to suggest that GVL and GVH are mediated by different clones of T cells. The objective of this study was to identify the two types of T cells based on their receptor sequences. To this end we used irradiated nonleukemic cells from recipients as stimulator cells in a primary mixed leukocyte reaction (MLR). The activated CD4(+) donor T cells that expressed CD25 were purified by cell sorting. To prepare GVL-specific T cells, alloreactive T cells in the primary MLR were first depleted with an anti-CD25 immunotoxin. The remaining T cells had negligible alloreactivity in a secondary MLR. The allodepleted cells were then stimulated by using purified leukemia cells from the same individual as stimulator cells, and the CD25(+)-activated cells were purified by cell sorting. The GVL- and GVH-specific T cells were analyzed for their T cell receptor (TCR) clonality by using anchored RT-PCR of all the TCRbeta locus complementarity-determining region 3 (CDR3) sequences. By comparing TCRbeta CDR3 sequences from transformed bacterial colonies, we were able to demonstrate that T cells mediating GVH were different from those mediating GVL in each of the eight HLA-mismatched and one HLA-matched donor/recipient pairs. By using the appropriate TCRbeta CDR3-specific primers and probes, the GVH- and GVL-specific clones were monitored in a recipient undergoing an allogeneic stem cell transplant from her HLA-matched related donor.

Optimized clinical-scale culture conditions for ex vivo selective depletion of host-reactive donor lymphocytes: a strategy for GvHD prophylaxis in allogeneic PBSC transplantation.

BACKGROUND: Ex vivo selective depletion (SD) is a strategy to prevent GvHD, in which host-reactive donor lymphocytes are selectively eliminated from a PBSC allograft while conserving useful donor immune function. Prior to testing this strategy in patients, our goal was to develop a clinical-scale SD process, which involves co-culture of donor lymphocytes and irradiated recipient cells, followed by the addition of an immunotoxin (IT) directed against the alpha-chain of the IL-2 receptor (CD25), expressed on activated donor T cells. METHODS: Stimulator cells were generated from immunomagnetically selected and expanded recipient T lymphocytes. Donor PBMCs from G-CSF-mobilized peripheral blood were co-cultured for 72 h with irradiated stimulator cells. Alloreactive T cells were targeted for elimination by the addition of the anti-CD25 IT, RFT5-SMPT-dgA, and the IT enhancer, NH(4)Cl. RESULTS: Stimulator-cell selection/expansion yielded > 2 x 10(10) highly enriched CD3(+) cells (98.9 +/- 2.2%). After SD, cell recovery was 68.5 +/- 23.3% and viability was 84.6 +/- 6.4%. This permitted a potential T-cell dose >/= 1 x 10(8) CD3(+) cells kg(-1) to transplant recipients. Although SD donor lymphocytes retained little proliferative capacity against the original stimulator cells (2.6 +/- 0.6%), responses were conserved against third party cells (107.6 +/- 18.6%), the bacterial superantigen staphylococcus enterotoxin B (108.2 +/- 4.2%), and CMV Ag (72.1 +/- 3.8%). DISCUSSION: We have demonstrated that ex vivo SD is feasible in clinical-scale culture conditions. The ability of this strategy to prevent GvHD is the subject of an ongoing clinical trial, in which the SD lymphocyte product is transplanted in conjunction with a T cell-depleted PBSC allograft.

Monitoring expression of HER-2 on circulating epithelial cells in patients with advanced breast cancer.

Nineteen breast cancer patients with measurable metastatic disease who were starting an initial or new line of therapy were evaluated for circulating epithelial cells (CECs) a minimum of 4 times over the course of treatment. In 7 of the 10 CEC+ patients, HER-2 expression was detected on the CECs. CECs expressing HER-2 varied among patients and in serial samples from the same patient including a shift from HER-2- to HER-2+ CECs. These results demonstrate that it is possible to quantify receptors essential for rationally designed therapy using CECs and that reliance on the immunophenotype of the primary tumor can be misleading.

Chemical construction of immunotoxins.

Immunotoxins are chimeric proteins consisting of an antibody linked to a toxin. The antibodies most frequently used for the preparation of immunotoxins are murine monoclonal antibodies belonging to IgG isotype. The most used toxins for the chemical construction of immunotoxins are Ricin toxin A chain in its deglycosylated form and recombinant Pseudomonas endotoxin with the cell-binding domain deleted. The linkage of the antibody to the toxin can be accomplished by chemical methods using reagents that crosslink antibody to toxin. The usual crosslinkers attach disulfide groups into the antibody molecule to form a disulfide bond between the antibody and the toxin. Disulfide bonds are susceptible to reduction in the cytoplasm of the targeted cells thereby releasing the toxin so that it can exert its cytotoxic activity only into the cells (e.g., tumor cells) binding the antibody moiety. This article describes various methods to obtain antibodies and toxins and several procedures for their crosslinking as well as "in vitro" and "in vivo" testing of the immunotoxins efficacy.

The toxicity of deglycosylated ricin A chain-containing immunotoxins in patients with non-Hodgkin's lymphoma is exacerbated by prior radiotherapy: a retrospective analysis of patients in five clinical trials.

A retrospective analysis of 102 patients with relapsed, non-Hodgkin's lymphoma treated with two different ricin A chain-containing immunotoxins (ITs) in five Phase I clinical trials indicates that the dose-limiting toxicity, vascular leak syndrome, was more frequent and more severe in patients who had undergone prior radiotherapy (RT). Excluding patients with prior RT from the calculations of the maximum tolerated dose indicates that the maximum tolerated doses of these ITs had not been reached in any trial and are clearly higher than reported previously. Excluding patients with prior RT from future clinical trials may increase the dose of ITs that can be given in the absence of severe vascular leak syndrome.

A role for intracellular pH in membrane IgM-mediated cell death of human B lymphomas.

We show that anti-IgM-induced cell death in a human B lymphoma cell line, B104, is associated with early intracellular acidification and cell shrinkage. In contrast, another human B cell lymphoma line, Daudi, less susceptible to B cell antigen receptor-mediated cell death, responded to anti-IgM with an early increase in intracellular pH (pH(i)). The anti-IgM-induced changes of pH(i) were associated with different levels of activation of the Na(+)/H(+) exchanger isoform 1 (NHE1) as judged by its phosphorylation status. Prevention of anti-IgM-induced cell death in B104 cells by the calcineurin phosphatase inhibitor, cyclosporin A, abrogated both intracellular acidification and cell shrinkage and was associated with an increase in the phosphorylation level of NHE1 within the first 60 min of stimulation. This indicates a key role for calcineurin in regulating pH(i) and cell viability. The potential role of pH(i) in cell viability was confirmed in Daudi cells treated with an Na(+)/H(+) exchanger inhibitor 5-(N,N-hexamethylene)amiloride. These observations indicate that the outcome of the anti-IgM treatment depends on NHE1-controlled pH(i). We suggest that inactivation of the NHE1 in anti-IgM-stimulated cells results in intracellular acidification and subsequently triggers or amplifies cell death.

Homodimers but not monomers of Rituxan (chimeric anti-CD20) induce apoptosis in human B-lymphoma cells and synergize with a chemotherapeutic agent and an immunotoxin.

In 1997, a chimeric anti-CD20 monoclonal antibody (mAb) (Rituxan) was approved for the treatment of low-grade/follicular B-cell lymphoma. Rituxan has a long half-life and low immunogenicity, and it mediates effector function. Rituxan induces apoptosis in some tumor cell lines in vitro. Previous studies with mAbs that react with neoplastic B cells have demonstrated that homodimers of immunoglobulin G ([IgG](2)) often inhibit cell growth more effectively than their monomeric (IgG)(1) counterparts. In this study, the ability of IgG or F(ab')(2) homodimers vs monomers of Rituxan were compared for their ability to inhibit the growth of several different B-lymphoma cell lines in vitro. It was found that homodimers of Rituxan had superior antigrowth activity in vitro and that F(ab')(2) homodimers were the most active. Homodimers, but not monomers, of Rituxan induced both apoptosis and necrosis of several B-cell lymphoma lines in vitro; the inhibition of cell growth was not dependent upon the presence of Fc receptors or upon 10-fold or greater differences in the density of CD20 on the target cells. Rituxan homodimers, compared with monomers, also rendered drug-resistant CD20(+) B-lymphoma cells more sensitive to chemotherapeutic agents and synergized with an anti-CD22 immunotoxin in vitro.

The growth and metastasis of human, HER-2/neu-overexpressing tumor cell lines in male SCID mice.

HER-2/neu is overexpressed on a variety of human adenocarcinomas and overexpression has been associated with a poor prognosis. For this reason, HER-2 has become an attractive target for immunotherapy. To facilitate testing of anti-HER-2-monoclonal antibodies (MAbs) and immunotoxins (ITs), we have evaluated the in vivo growth and metastatic spread of three HER-2-overexpressing human breast cancer cell lines (BT474, MDA-MB-453 and HCC1954) and one ovarian cancer cell line (SKOV3.ip1) in pre-irradiated male SCID mice using subcutaneous (s.c.), intravenous (i.v.) and intraperitoneal (i.p.) routes of injection. All the cell lines tested grew as s.c. tumors and the growth of BT474 and MDA-MB-453 cells after s.c. injection was improved by co-inoculation with Matrigel. Metastases to the lungs were detectable by PCR or histopathology after s.c. injection of BT474 and to a much lesser extent after s.c. injection of HCC1954, MD-MB-453 and SKOV3.ip1 cells. I.p. injection of HCC1954 and SKOV3.ip1 cells produced fatal ascites while i.v. injection of SKOV3.ip1, but not BT474 or MDA-MB-453 cells, resulted in infiltration of lungs and death within 9-11 weeks.

The effect of a monoclonal antibody coupled to ricin A chain-derived peptides on endothelial cells in vitro: insights into toxin-mediated vascular damage.

Immunotoxins (ITs) containing plant or bacterial toxins have a dose-limiting toxicity of vascular leak syndrome (VLS) in humans. The active A chain of ricin toxin (RTA), other toxins, ribosome-inactivating proteins, and the VLS-inducing cytokine IL-2 contain the conserved sequence motif (x)D(y) where x = L, I, G, or V and y = V, L, or S. RTA-derived LDV-containing peptides attached to a monoclonal antibody, RFB4, induce endothelial cell (EC) damage in vitro and vascular leak in two animal models in vivo. We have now investigated the mechanism(s) by which this occurs and have found that (1) the exposed D75 in the LDV sequence in RTA and the C-terminal flanking threonine play critical roles in the ability of RFB4-conjugated RTA peptide to bind to and damage ECs and (2) the LDV sequence in RTA induces early manifestations of apoptosis in HUVECs by activating caspase-3. These data suggest that RTA-mediated inhibition of protein synthesis (due to its active site) and apoptosis (due to LDV) may be mediated by different portions of the RTA molecule. These results suggest that ITs prepared with RTA mutants containing alterations in LDVT may kill tumor cells in vivo in the absence of EC-mediated VLS.

Immunotoxins against CD19 and CD22 are effective in killing precursor-B acute lymphoblastic leukemia cells in vitro.

Monoclonal antibodies (Mabs) conjugated to toxins or their subunits (immunotoxins or ITs) are undergoing clinical testing in adults with a variety of malignancies. The potential impact of this form of therapy in pediatric precursor B-lineage acute lymphoblastic leukemia (pre-B ALL) has yet to be determined. Mabs directed against the cell surface antigens, CD19 and CD22 conjugated to deglycosylated ricin A chain (dgRTA) have been tested in patients with non-Hodgkin's lymphoma (NHL), but not in patients with pre-B ALL. Because of the encouraging performance of these ITs in phase I trials, we evaluated the specific cytotoxicity of anti-CD19 (HD37-dgRTA) and anti-CD22 (RFB4-dgRTA) ITs or their combination (Combotox) on patient-derived pre-B ALL cells maintained in vitro on a stromal feeder layer. After 48 h in culture, cytotoxicity to tumor cells was determined by flow cytometry using propidium iodide (PI) and fluorescein isothiocyanate (FITC)-conjugated anti-CD10, 19, and 22. Both RFB4-dgRTA and HD37-dgRTA induced a statistically significant reduction in the number of viable leukemic cells, and Combotox was even more effective. Our results demonstrate that these ITs are specifically cytotoxic to primary pre-B ALL cells and that they should be further evaluated for the therapy of B-lineage ALL.

A phase I study of combination therapy with immunotoxins IgG-HD37-deglycosylated ricin A chain (dgA) and IgG-RFB4-dgA (Combotox) in patients with refractory CD19(+), CD22(+) B cell lymphoma.

This study used an 8-day continuous infusion regimen of a 1:1 mixture of two immunotoxins (ITs) prepared from deglycosylated ricin A chain (dgA) conjugated to monoclonal antibodies directed against CD22 (RFB4-dgA) and CD19 (HD37-dgA; Combotox) in a Phase I trial involving 22 patients with refractory B cell lymphoma to determine the maximum tolerated dose, clinical pharmacology, and toxicity profile and to characterize any clinical responses. Adult patients received a continuous infusion of Combotox at 10, 20, or 30 mg/m2/192 h. No intrapatient dose escalation was permitted. Patients with > or =50 circulating tumor cells (CTCs)/mm3 in peripheral blood tolerated all doses without major toxicity. The maximum level of serum IT (Cmax) achieved in this group was 345 ng/ml of RFB4-dgA and 660 ng/ml of HD37-dgA (1005 ng/ml of Combotox). In contrast, patients without CTCs (<50/mm3) had unpredictable clinical courses that included two deaths probably related to the IT. Additionally, patients exhibited a significant potential for association between mortality and a history of either autologous bone marrow or peripheral blood stem cell transplants (P2 = 0.003) and between mortality and a history of radiation therapy (P2 = 0.036). In patients with CTCs, prior therapies appeared to have little impact on toxicity. Subsequent evaluation of the ITs revealed biochemical heterogeneity between two lots of HD37-dgA. In addition, HD37-dgA thawed at the study site tended to contain significant particulates, which were not apparent in matched controls stored at the originating site. This suggests that a tendency to aggregate may have resulted from shipping, storage, and handling of the IT that occurred prior to preparation for administration. It is not clear to what extent, if any, the aggregation of HD37-dgA IT was related to the encountered clinical toxicities; however, the potential to aggregate does suggest one possible basis for problems in our clinical experience with HD37-dgA and leads us to the conclusion that non-aggregate-forming formulations for these ITs should be pursued prior to future clinical trials.

The combined use of an immunotoxin and a radioimmunoconjugate to treat disseminated human B-cell lymphoma in immunodeficient mice.

Immunoconjugates (ICs) consist of a targeting moiety and a toxic moiety and have the specificity that traditional cancer therapy lacks. At appropriate doses, ICs are safe and effective in treating various cancers in experimental animals and in humans. However, because cures are rarely achieved using single agents, regimens involving combinations of agents with different mechanisms of action must be evaluated. In this study, we explored the efficacy and toxicity of a combination of two IC therapies, radioimmunotherapy (RIT) and immunotoxin (IT) therapy, to treat advanced, disseminated human lymphoma in immunodeficient mice. We proposed to use the bystander effect of RIT to reduce large tumor burdens, followed by an IT to eliminate residual tumor cells. Our results indicate that, when used alone, both RIT and IT therapy were safe and effective, but not curative. When the two therapies were combined, efficacy and toxicity became dependent on the temporal order of administration. Thus, with the doses used in this study, when RIT was administered after IT therapy, the regimen was curative. In contrast, when RIT was administered before IT therapy, the combination was highly toxic or even lethal. Both RIT and IT therapy induced pulmonary vascular leak, but with different kinetics. When RIT was given prior to IT therapy, the pulmonary vascular leak became life-threatening but not when the two agents were administered in the reverse order.