Therapeutic and Specific Antitumor Immunity Induced by IL12 and IL18 in Fibroblasts and DCs: Preclinical Studies and Application to Patients
Michael T. Lotze 1,2 , Walter Storkus, 1,2, Laurence Zitvoge1,2, John Kirkwood3, Theresa Whiteside4, Elaine Elder4, Pawel Kalinski, 1, Robbie Mailliard1, Fumiaki Tanaka1,2,4, Wataru Hashimoto1,2, Paul D. Robbins2, Hideaki Tahara1,2,3
1 Departments of Surgery, 2 Molecular Genetics and Biochemistry, 3Medicine, and 4Pathology, University of Pittsburgh School of Medicine, 300 Kaufmann Building, 3471 Fifth Avenue, Pittsburgh, PA 15213, USA, 5Department of Surgery and Bioengineering, Institute of Medical Science, The University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo, 108-8639, Japan
We investigated in vivo antitumor effects of sytemic and intratumoral (i.t.) administration of cytokines in murine tumor systems. Potent antitumor effects were observed when established ovalbumin transfected B16 melanoma, methylcholanthrene induced sarcoma, or 3LL lung adenocarcinoma.were treated with peptide or cytokine transfected DC. Clinical trials have completed with 3/28 melanoma patients treated with multivalent peptides pulsed onto autologous DCs derived from GM/IL4 cultured macrophages and in 7/29 patients treated with autologous fibroblasts transfected with IL12. We have extended these observations to demonstrate that IL12 and now IL18 transfected DC are competent to mediate antitumor effects when administered intratumorally in mouse. The MCA205 fibrosarcoma was treated in syngeneic immunocompetent mice with either peptide pulsed immature dendritic cells i.t. injection of Ad.PTH.IL-18 (p=0.0025 vs. control vector treatment), and potent cytotoxic T lymphocytes (CTLs) were generated in culture. The in vivo antitumor effect and cytotoxic activity was significantly less (p=0.021) in gld mice (Fas ligand deficient). We co-injected immature DC and Ad.PTH.IL-18 i.t. into MCA205 fibrosarcoma and MC38 adenocarcinoma with complete abrogation of injected and distant noninjected tumors. The induced cytolytic activity was tumor specific and MHC class I restricted. We believe that the initial antitumor effect of locally expressed IL-18 is mediated by NK activity. Co-injection of DC further enhances this antitumor effect by strong CTL induction, consistent with our recent report using a co-culture system (Cancer Res 2000 60:4838-44). This study suggests that cancer patients with multiple distant metastasis could be treated using this strategy.
Megan A. Cooper, Todd A. Fehniger, and Michael A. Caligiuri.
The Ohio State University Comprehensive Cancer Center, Columbus, OH
IL-15 is a pleiotropic cytokine expressed by a wide variety of cell types including bone marrow stroma and antigen-presenting cells such as monocyte/macrophages. IL-15 shares common signaling receptors with IL-2, but due to its distinct tissue distribution and the presence of its private high-affinity receptor, has very unique actions in vivo. Some of the best-studied effects of IL-15 are its ability to stimulate differentiation and activation of natural killer (NK) cells. During the early immune response, NK cells are a first-line defense against a number of pathogens through their cytolysis of infected or transformed cells and early provision of cytokines and chemokines for activation of the innate and adaptive immune response. IL-15 is requisite for NK cell differentiation and survival in vivo, and thus is crucial for maintaining homeostasis of these innate immune lymphocytes. In humans, NK cells can be divided into two distinct subsets characterized by their surface density expression of the CD56 antigen. The CD56bright NK subset represents 10% of NK cells and only 1% of circulating lymphocytes, yet is a powerhouse of rapid cytokine production compared to its CD56dim cytolytic NK counterpart. IL-15 drives CD34+IL-15R+ human NK precursors through differentiation into CD56bright NK cells. Production of IL-15 is rapidly up regulated following certain infections and, in combination with other monocyte-derived cytokines, co-stimulates interferon gamma (IFN-g) production from resting CD56bright NK cells. In turn, NK cell-derived IFN-g is requisite for the activation of macrophages and elimination of intracellular pathogens. In vivo, IL-15 is a critical co-factor for NK cell IFN-g production following administration of IL-12 or LPS. Thus, IL-15 participates in the first line of host innate immune defense, before an adaptive immune response can be mounted. Expression of IL-15 is very tightly-controlled at the post-transcriptional level and chronic over expression of IL-15 can lead to the development of a fatal T-NK leukemia in a transgenic mouse model, providing some insight into the importance of IL-15’s regulation in vivo. Recent evidence has shown that IL-15 also is important for the survival and expansion of memory CD8+ T cells, in distinct contrast to IL-2, which promotes activation-induced cell death (AICD) of these lymphocytes. Thus, this unique immunoregulatory cytokine participates in both innate and adaptive immunity, providing a bridge between these two arms of the immune response.
Paul Sondel MD PhD
The University of Wisconsin Comprehensive Cancer Center, Madison WI
Humanized tumor reactive mAbs linked to IL2, are designated “immunocytokines” (IC). KS-IL2 recognizes the human epithelial cell adhesion molecule (EpCAM) on adenocarcinomas; hu14.18-IL2 recognizes the GD2 disialoganglioside on neuroectodermal tumors. These were created by our collaborators, Ralph Reisfeld (Scripps, La Jolla CA), and Steve Gillies (Lexigen, Lexington MA) and INDs are held by EMD (Durham NC). Prior studies have shown i.v. treatment of mice with IC induces dramatically greater antitumor effects than treatment with the mAb, treatment with IL2, or combined treatment with the mAb and IL2. We have shown that KS-IL2 induces protection against 2 sub-clones of EpCAM transfected CT-26 tumor. Remarkably, the anti-tumor effect induced by KS-IL2 in mice syngeneic to CT-26 is mediated by T cells against one CT-26 sub-clone, and primarily by NK cells against the other sub-clone. These 2 tumor sub-clones differ for their level of MHC-class I expression. In a syngeneic neuroblastoma model, hu14.18-IL2 mediates potent antitumor effects via NK cells. When sub-optimal regimens of hu14.18-IL2 are given, some animals show an initial antitumor response followed by tumor recurrence. These recurrent tumors have up-regulated their MHC-class I antigens, to facilitate escape from NK induced destruction. In this same model, those neuroblastomas that escape from T cell mediated anti-tumor effects show a down-regulation of MHC-class-I antigens. Thus, in both the CT26 and the neuroblastoma murine models, the MHC class-I expression influences whether IC controls tumor growth in vivo via NK cells or via T cells. We have begun clinical testing with hu14.18-IL2. A phase I trial has shown acceptable tolerance at doses up to 6.0 mg/M2/d for 3 daily injections in 27 patients. Serum levels of hu14.18-IL2 peak at over 3 mcg/ml with a half life of 3.5 hours. The circulating hu14.18-IL2 in patients’ sera maintains its GD2 binding capacity, its IL2 stimulatory ability and its ability to induce ADCC, using patients PBMCs as effectors. Once the MTD is determined, we plan to begin a phase II trial in patients with melanoma.
JM Wigginton1, ME Edgerly1, SE Bates2, P Choyke3, RH Wiltrout4, and J Janik5.
1Pediatric Oncology Branch, 2Cancer Therapeutics Branch, and 5Metabolism Branch, Center for Cancer Research, NCI, and 3Department of Radiology, Clinical Center, NIH, Bethesda, MD; 4Laboratory of Experimental Immunology, NCI at Frederick, Frederick, MD.
We previously demonstrated that systemic administration of the combination of IL-12 and intermittent pulse IL-2 possesses synergistic antitumor activity in models of well established primary and/or metastatic murine renal cell carcinoma (Renca). We subsequently have reported that this regimen can induce complete regression of autochthonous tumor in a transgenic mouse model of spontaneous mammary carcinoma. Further, in this model, IL-12/pulse IL-2 delays the genetically-programmed progression of preneoplastic lesions to overt carcinoma, and does so in conjunction with marked enhancement of local T cell infiltration, Fas/Fas-L gene expression, and the induction of mammary epithelial apoptosis. In the Renca model, IL-12/pulse IL-2 induces vascular endothelial injury, and inhibits tumor neovascularization and mediates tumor regression via mechanisms which share a common dependency on IFN-( and the Fas/Fas-L pathway. These and other studies provided impetus for our subsequent execution of a primate toxicology assessment of this combination, and the initiation of phase I investigation of IL-12/IL-2 combinations, including IL-12/pulse IL-2 in adults with advanced solid tumors. In this study, cohorts of 3-6 patients are being treated with IL-2 given every 8 hours on days 1 and 9 of each 35 day treatment cycle, and intravenous IL-12 once daily on days 2, 4, 6, 10, 12 and 14. The respective doses of IL-2 and IL-12 are as follows: I (30,000 IU/kg, 30ng/kg), II (30,000IU/kg, 100 ng/kg), III (100,000 IU/kg, 100ng/kg), IV (100,000 IU/kg, 200ng/kg), V (200,000IU/kg, 200ng/kg), and VI (200,000 IU/kg, 300ng/kg). During the initial cycle of therapy, serial samples are obtained for investigation of the serum cytokine cascade and functional immunoregulatory effects of this combination in treated patients. Similarly, sequential dynamic enhanced MRI scans are performed to assess the impact of therapy on tumor neovascularization. Where possible, tumor biopsies are obtained both pre- and post-cycle 1. As of September 2001, seven patients have been enrolled on this study, and patients are now being enrolled on dose level 3. Tumor types among enrolled patients include renal cell carcinoma (4), hepatocellular carcinoma (1), colon carcinoma (1) and malignant fibrous histiocytoma (1). To date, therapy has been well tolerated overall with no dose-limiting toxicities. Common toxicities have included fever, chills, fatigue, and reversible cytopenias among others. Updated information will be presented regarding the spectrum of toxicities and outcome of these patients as well as results from concurrent studies to monitor relevant biologic endpoints.
Phase I Trial of Twice-Weekly IV rhIL-12 Plus Low-Dose SC IL-2 in Patients with Melanoma and Renal Cell Cancer
Gollob JA1, Veenstra KG1, Mier JW1, McDermott DF1, Atkins MB1
1Beth Israel Deaconess Medical Center, Boston, MA
Interleukin (IL)-12 induces tumor regression in murine models of melanoma and renal cell cancer, and that activity is enhanced through combination therapy with IL-12 + IL-2. In early trials of rhIL-12, there were few tumor responses. This paucity of responses was associated with the rapid downmodulation of rhIL-12-induced IFN-g production in patients repeatedly dosed with rhIL-12. In a subsequent phase I trial of twice weekly IV rhIL-12, antitumor responses were seen in the small number of patients who were able to maintain the induction of IFN-g and IL-15 by rhIL-12 over an extended period of time. In those patients who could not maintain IFN-g induction, it was found that lymphocyte IFN-g production could be restored in vitro through combined stimulation with IL-12 + IL-2. These findings provided the rationale for adding low-dose IL-2 to twice-weekly IV rhIL-12 in order to maintain IL-12-induced immune activation and thereby enhance the antitumor activity of rhIL-12. A phase I trial of twice-weekly IV rhIL-12 plus low-dose SC IL-2 opened at the Beth Israel Deaconess Medical Center in March 2000. A cycle of therapy consists of IV rhIL-12 given twice-weekly (doses 3-4 days apart) for 6 weeks. During the first cycle, rhIL-12 is given alone until the end of the third week, at which time the SC IL-2 is begun, with doses given 1 hour before and 20 hours after each dose of rhIL-12. Stable or responding patients are able to receive subsequent cycles of therapy, all of which consist of rhIL-12 + IL-2 given together throughout the 6 weeks. Serial blood draws over a 24-hour period are performed for immune monitoring during cycle 1 after the first dose of rhIL-12 (start of week 1), last dose of rhIL-12 (start of week 3), first doses of rhIL-12 + IL-2 (end of week 3), and last doses of rhIL-12 + IL-2 (end of week 6). The rhIL-12 (ng/kg)/IL-2 (MIU/m2) dose levels are 300/0.5, 500/0.5, 500/1.0, 500/3.0, and 500/6.0. As of September 2001, 16 patients have been treated on the first four dose levels (7 melanoma, 8 renal cell cancer, 1 transitional cell cancer). The addition of IL-2 to rhIL-12 has been well-tolerated, with some patients at the 500/1.0 and 500/3.0 dose levels experiencing a recrudescence of mild-moderate flu-like symptoms. There have been no dose-limiting toxicities attributable to the rhIL-12 + IL-2 combination. Although no major responses were seen at the first three dose levels, one melanoma patient treated at 500/0.5 had regression of multiple cutaneous metastases, a patient with renal cell cancer treated at 500/1.0 had resolution of a malignant pleural effusion, and a patient with ocular melanoma and extensive hepatic metastases treated at 500/1.0 has had stable disease through 6 cycles of therapy. At the 500/3.0 dose level, three patients (2 melanoma, 1 renal cell cancer) are currently being treated, and two have been evaluated for response. One patient with melanoma has had a PR after 2 cycles, with >50% regression of a large left hilar lung mass. Another patient with melanoma and in-transit skin metastases over the chest wall has had complete flattening of multiple hyperpigmented nodular lesions after 2 cycles. An examination of in vivo cytokine production showed the downregulation of rhIL-12-induced IFN-g and IP-10 production at the start of the third week of cycle 1. While the addition of IL-2 to rhIL-12 at the end of week 3 modestly augmented IFN-g and IP-10 production at the 500/1.0 dose level, it nearly fully restored IFN-g and IP-10 production at the 500/3.0 dose level. This effect of IL-2 was evident after the first dose, and was still maintained at the start of week 6. These findings demonstrate that low-dose IL-2 can be safely administered concurrently with rhIL-12, and can augment rhIL-12-induced IFN-g and IP-10 production without significant toxicity. The antitumor activity observed at the 500/3.0 dose level is encouraging, and suggests that this strategy for maintaining immune activation by rhIL-12 may indeed augment its antitumor activity. Escalation to the final dose level of 500/6.0 will allow for the further examination of the tolerability and efficacy of this regimen, and will determine which dose of SC IL-2 in combination with twice-weekly IV rhIL-12 is most appropriate to take into phase II testing.
Interleukin 7 is well known as a lymphopoietic cytokine required for T and B cell development in mice and for T cell development in humans. Within the thymus, IL-7 plays a critical role in the survival and proliferation of developing T cells and is likely to be involved in T cell receptor rearrangement. In addition, IL-7 has potent effects on mature T cells. These effects are multiple and include co-stimulating for T cell activation, enhancement of cytolytic function and inhibition of programmed cell death. The net result is an enhanced proliferative response to high-affinity antigens and the potential recruitment of T cells bearing low-affinity receptors. Recently, IL-7 has been shown to be a critical cytokine for mature T cell homeostasis, based upon its requirement for homeostatic peripheral expansion following T cell depletion and the discovery that T cell depleted humans have elevated levels of circulating IL-7 which return to normal upon recovery of normal T cell numbers. Thus, therapeutic administration of IL-7 holds promise as both an immunorestorative agent following T cell depleting therapies and, potentially, as a vaccine adjuvant. In murine models, IL-7 enhances thymopoiesis following bone marrow transplantation. In athymic mice, IL-7 enhances the peripheral homeostatic expansion of mature T cells and can restore immunocompetence to a stringent antigen in the absence of a thymus. Therefore, the pharmacologic administration of IL-7 can potently enhance T cell regeneration by increasing T cell numbers and by inducing immune responses following T cell depletion in mice. Emerging data has now confirmed that IL-7 potently increases T cell numbers in T cell replete and T cell depleted non-human primates. As a vaccine adjuvant, IL-7 increases the number of antigen-specific CD8 T cells following dendritic cell vaccination. In this regard, IL-7 appears to be more potent than either IL-2 or IL-15. In summary, pre-clinical models suggest that therapeutic administration of IL-7 will improve T cell immune reconstitution in clinical settings associated with T cell depletion. In addition, IL-7 is predicted to enhance responses to T cell based vaccines by increasing the number and, potentially, the function of antigen-specific effector cells induced in vivo.