Methods for determining the toxicity of immunomodulatory agents for use in humans - Patents.com

Abstract

To provide a humanized mouse model and method for determining whether administration of an immunomodulator can induce severe cytokine release syndrome in humans.SOLUTION: Provided are a humanized mouse model and method for determining the immunotoxicity of drug candidates or drug combinations in humans. The present invention provides a method for determining whether an immunomodulator can induce severe cytokine release syndrome in humans after administration of the immunomodulator. In another example embodiment, provided is a method for determining whether a combination of a first immunomodulator and a second immunomodulator is likely to induce severe cytokine release syndrome in humans after administration of the combination of immunomodulators.SELECTED DRAWING: None

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of priority to U.S. Provisional Application No. 62 / 486,441, filed April 17, 2017, and U.S. Provisional Application No. 62 / 521,617, filed June 19, 2017, the contents of both of which are incorporated by reference in their entireties herein.

[0002] The present invention relates generally to methods for determining whether an immunomodulatory agent induces a cytokine release syndrome response in a human to which the immunomodulatory agent is administered. The present invention also provides an in vivo mouse method that has predictive value for use in the pharmaceutical safety evaluation of drug candidates. [Background technology]

[0003] Monoclonal antibodies (mAbs) have been used as therapeutics in the treatment of cancer and autoimmune diseases. Many of these therapeutic mAbs are targeted against proteins on the surface of immune cells, particularly T and B cells. However, mAbs can have various adverse effects upon infusion, such as cytokine release syndrome (CRS) or systemic inflammatory response syndrome (SIRS), which can be fatal. CRS becomes clinically manifest when a large number of lymphocytes (B cells, T cells, and / or natural killer (NK) cells) and / or myeloid cells (macrophages, dendritic cells, and monocytes) are activated by the administered mAb and release inflammatory cytokines. The timing of symptom onset and severity of CRS depend on the type of mAb and the amplitude of immune cell activation.

[0004] In general, there are two existing methods to test toxicity prior to clinical trials of drugs: in vivo tests in animal models and in vitro whole blood or peripheral blood mononuclear cell (PBMC) assays. Unfortunately, these two methods cannot adequately predict or determine toxicity in humans, especially immunological toxicity. In vitro tests cannot mimic the human patient's body; the systemic response to potential drug toxicity cannot be modeled in any model other than in vivo. Genomic responses in rodent and non-human primate models cannot mimic the human response. From a toxicity perspective, there is a substantial gap between preclinical trials and clinical trials.

[0005] Methods have been attempted to transplant human stem cells into non-human mammals for testing. However, such methods suffer from significant challenges in obtaining non-embryonic stem cells from patients (e.g., obtaining bone marrow from patients), and are disadvantageous in that they require too long a time to wait for the stem cells to grow and differentiate into various cells. Thus, such methods are invasive (if possible), cumbersome, and impractical.

[0006] TGN1412, developed by the now defunct TeGenero AG, Wurzburg, is a humanized monoclonal antibody (mAb) of the IgG4 subclass specific for the costimulatory molecule CD28 expressed by human T cells. It is called a "CD28 superagonist" (CD28SA) because, unlike classical CD28-specific mAbs, it can activate T lymphocytes without simultaneous engagement of the T cell antigen receptor (TCR) (Hunig, 2007, Adv Immunol, 95:111-148).

[0007] In a first-in-man trial conducted by the independent Parexel Clinical Trial Unit at Northwick Park Hospital, London on March 13, 2006, intravenous application of 100 μg / kg body weight TGN1412 to healthy human volunteers resulted in a life-threatening cytokine release syndrome that was only controlled after the volunteers were transferred to the hospital's intensive care unit (Suntharalingam et al., 2006, Engl J Med 355:1018-1028). Thus, the anti-CD28 immunomodulatory drug TGN1412 failed badly in its Phase I trial due to a cytokine storm occurring in the subjects under study, which severely threatened the lives of six healthy volunteers who participated, all six of whom suffered multiple organ failure.

[0008] However, preclinical studies in this study showed no evidence of such a "cytokine storm" in a similar rat model using a rat CD28-specific superagonist, and in cynomolgus monkeys (Macaca fascicularis) that received TGN1412 itself at doses up to 50-fold higher than in human volunteers (Duff, 2006, Expert Scientific Group on Phase One Clinical Trials Final Report., Norwich, UK:Stationary Office). Furthermore, addition of TGN1412 to cultures of human PBMCs did not result in cytokine release. All the crucial monkey and PBMC culture experiments were repeated by the British National Institute for Biological Standards and Control (NIBSC) acting on behalf of the government's Expert Scientific Group on Phase One Clinical Trials, which confirmed the innocuous behavior of TGN1412 in these systems (Duff, 2006). Thus, these rat, cynomolgus monkey, and cultured human PBMC assays were not adequate to warn against the cytokine storm experienced by human volunteers. The inability of rodents and cynomolgus monkeys to release toxic systemic cytokines after injection of CD28SA may be due to species differences in the reactivity of the intact immune system to such agents, and specific proposals have been made regarding such differences (Gogishvili et al., 2009, PLoS ONE, 4(2):e4643, https: / / doi.org / 10.1371 / journal.pone.0004643; Nguyen et al., 2006, Proc Natl Acad Sci US A., 103:7765-7770; Schraven and Kalinke, 2008, Immunity, 28:591-595).

[0009] Humans have approximately 1×10 1210 T lymphocytes, with less than 1 percent of such cells circulating in the blood at any given time. It is currently unclear whether the failure of cultured PBMCs to respond to TGN1412 is due to a functional defect in such cells compared with cells resident in lymphoid tissues (which responded clearly by cytokine release in volunteers), or to a requirement for TGN1412-mediated activation of T lymphocytes of a cell type present in lymphoid organs but not in blood. The failure of known human PBMC cultures to respond to soluble TGN1412 with cytokine release indicates that this system does not respond to all lymphocyte activators in the same manner as the intact human immune system inside the body. Correction of this defect may not only allow detailed analysis of the effects of human CD28 superagonists (SAs), such as TGN1412, but may also reveal the reactivity of other seemingly innocuous drugs in preclinical development. [Prior art documents] [Non-patent literature]

[0010] [Non-Patent Document 1] Hunig, 2007, Adv Immunol, vol. 95: pp. 111-148. [Non-Patent Document 2] Suntharalingam et al., 2006, Engl J Med, 355:1018-1028 [Non-Patent Document 3] Duff, 2006, Expert Scientific Group on Phase One Clinical Trials Final Report. Norwich, UK: Stationary Office. [Non-Patent Document 4] Gogishvili et al., 2009, PLoS ONE, Volume 4(2): Page e4643, https: / / doi.org / 10.1371 / journal.pone.0004643 [Non-Patent Document 5] Nguyen et al., 2006, Proc Natl Acad Sci U S A, 103:7765-7770 [Non-Patent Document 6] Schraven and Kalinke, 2008, Immunity, 28:591-595. Summary of the Invention [Means for solving the problem]

[0011] Thus, there is a continuing need for new in vivo humanized animal models that are useful for toxicity testing of immunomodulatory drugs and for determining toxicities such as cytokine release syndrome in individuals prior to patient treatment and clinical trials.

[0012] In a non-limiting example embodiment, the invention provides a method for determining whether an immunomodulatory agent is likely to induce severe cytokine release syndrome in a human following administration of the immunomodulatory agent. In an example embodiment, the method comprises: (a) preparing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; (b) 1.5–3.0 × 10 isolated from humans 7 engrafting the mouse with peripheral blood mononuclear cells (PBMCs); (c) administering an immunomodulatory agent to the mouse 5 to 7 days after engraftment of the PBMCs; (d) determining blood levels of a plurality of cytokines in said mouse, including IFN-γ and IL-10; blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse; (e) determining that the immunomodulatory agent is capable of inducing severe cytokine release syndrome in the human; the presence of severe cytokine release syndrome in the mouse indicates that administration of the immunomodulatory agent is likely to induce severe cytokine release syndrome in the human. Includes.

[0013] In another example embodiment, a method is provided for determining whether a combination of a first immunomodulatory agent and a second immunomodulatory agent is likely to induce severe cytokine release syndrome in a human following administration of said combination of immunomodulatory agents. The method comprises: (a) preparing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; (b) 1.5–3.0 × 10 isolated from humans 7 engrafting the mouse with peripheral blood mononuclear cells (PBMCs); (c) administering to the mouse a first immunomodulatory agent and a second immunomodulatory agent 5 to 7 days after engraftment of the PBMCs; (d) determining blood levels of a plurality of cytokines in said mouse, including IFN-γ and IL-10; blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse; (e) determining that said combination of immunomodulatory agents is likely to induce severe cytokine release syndrome in said human; the presence of severe cytokine release syndrome in the mouse indicates that administration of the combination of immunomodulatory agents is likely to induce severe cytokine release syndrome in the human. Includes.

[0014] In another exemplary embodiment, the present invention provides a method for determining a safe dosage of an immunomodulatory agent that does not induce cytokine release syndrome in a human following administration of the immunomodulatory agent. In an embodiment, the method comprises: (a) providing an immunomodulatory drug having a first dosage, wherein the first dosage of the immunomodulatory drug is determined to induce mild or severe cytokine release syndrome in a first humanized irradiated immunodeficient mouse following administration thereof; (b) providing a second immunodeficient mouse, the second mouse being irradiated with 75-125 cGy of X-rays; (c) 1.5–3.0 × 10 isolated from humans 7 engrafting the second mouse with peripheral blood mononuclear cells (PBMCs); (d) administering an immunomodulatory agent to the second mouse 5-7 days after engraftment of the PBMCs, wherein the immunomodulatory agent is administered at a second dosage that is less than the first dosage; (e) determining blood levels of a plurality of cytokines in the second mouse, including IFN-γ and IL-10; (f) determining a safe dosage of the immunomodulatory agent for administration in the human, the safe dosage being a dosage that results in blood levels of IFN-γ <300 pg / ml and IL-10 <25 pg / ml following administration of the immunomodulatory agent to the second mouse; blood levels of IFN-γ<300 pg / ml and IL-10<25 pg / ml in the second mouse, indicating that administration of the safe dosage of the immunomodulatory agent is likely not to induce cytokine release syndrome in the human; Includes.

[0015] In another exemplary embodiment, the present invention provides a method for determining the immunotoxicity of a drug candidate for use in humans. In an embodiment, the method comprises: (a) preparing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; (b) 4.5–5.5 × 10 7 (b) engrafting the mouse with human PBMCs; and (b) administering a drug candidate to the mouse 4-7 days after engraftment. (c) determining a cytokine concentration in the blood of the mouse, wherein the cytokine is at least one cytokine selected from the group consisting of IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNF; (d) determining the immunotoxicity of the drug candidate, IFN-γ ≥ 300 pg / ml, IL-2 ≥ 15 pg / ml, IL-4 ≥ 10 pg / ml, IL-6 ≥ 10 pg / ml, IL-10 ≥ 25 pg / ml, and TNF ≥ 5 pg / ml wherein the blood concentration in said mouse of at least one cytokine selected from the group consisting of is indicative of the immunotoxicity of said drug candidate in humans. Includes.

[0016] In another embodiment, a method is provided for determining the likelihood that administration of an immunomodulatory agent to a human will induce severe cytokine release syndrome in the human, the method comprising: (a) Human derived 1.5–3.0 × 10 7 providing a blood sample from a humanized irradiated immunodeficient mouse that has been administered an immunomodulatory drug 5-7 days after engraftment of isolated peripheral blood mononuclear cells (PBMCs); (b) detecting in vitro the concentration of a plurality of cytokines, including IFN-γ and / or IL-10, present in a blood sample from the mouse, wherein a concentration of IFN-γ≧1,800 pg / ml or IL-10≧120 pg / ml indicates that administration of the immunomodulatory agent to a human may induce severe cytokine release syndrome; Includes.

[0017] In another embodiment, a method is provided for determining the likelihood that administration of a combination of a first immunomodulatory agent and a second immunomodulatory agent to a human will induce severe cytokine release syndrome in the human, the method comprising: (a) Human derived 1.5–3.0 × 10 7providing a blood sample from a humanized irradiated immunodeficient mouse administered the combination of a first immunomodulatory agent and a second immunomodulatory agent 5-7 days after engraftment of isolated peripheral blood mononuclear cells (PBMCs); (b) detecting in vitro the concentration of IFN-γ and / or IL-10 present in a blood sample from the mouse, wherein a concentration of IFN-γ≧1,800 pg / ml or IL-10≧120 pg / ml indicates that administration of the combination of the first immunomodulatory agent and the second immunomodulatory agent to a human is likely to induce severe cytokine release syndrome; Includes.

[0018] In another embodiment, a method for determining the immunotoxicity of a drug candidate in a human is provided, the method comprising: (a) 4.5-5.5 × 10 7 providing a blood sample from a humanized irradiated immunodeficient mouse that has been administered the drug candidate 4-7 days after engraftment with isolated human peripheral blood mononuclear cells (PBMCs); (b) detecting in vitro the concentration of at least one human cytokine present in the mouse blood sample to determine the human immunotoxicity of the drug candidate, wherein the at least one human cytokine is selected from the group consisting of IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNF, and the drug candidate has low human immunotoxicity when a low human cytokine concentration is detected in the mouse blood sample; Includes. The present invention provides, for example, the following items. (Item 1) 1. A method for determining whether an immunomodulatory agent is likely to induce severe cytokine release syndrome in a human following administration of said immunomodulatory agent, comprising: (a) preparing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; (b) 1.5–3.0 × 10 isolated from humans 7 engrafting the mouse with peripheral blood mononuclear cells (PBMCs); (c) administering an immunomodulatory agent to the mouse 5 to 7 days after engraftment of the PBMCs; (d) determining blood levels of a plurality of cytokines in said mouse, including IFN-γ and IL-10; blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse; (e) determining that the immunomodulatory agent is capable of inducing severe cytokine release syndrome in the human; the presence of severe cytokine release syndrome in the mouse indicates that administration of the immunomodulatory agent is likely to induce severe cytokine release syndrome in the human. The method includes: (Item 2) The method of claim 1, wherein the mouse is an NSG, NSG-IL-6 or NSG-CSF-1 mouse. (Item 3) 2. The method of claim 1, wherein the mouse is an NSG mouse. (Item 4) 2. The method according to item 1, wherein the mouse in step (a) is irradiated with 100 cGy of X-rays. (Item 5) The engraftment step (b) is performed at a concentration of 2×10 7 The method of item 1, wherein the method is performed on PBMCs. (Item 6) The method according to item 1, wherein the administration step (c) is performed 6 days after engraftment. 2. The method of claim 1, wherein the plurality of cytokines in step (d) further comprises IL-2, IL-4, IL-6 and TNF. (Item 8) 2. The method of claim 1, wherein the blood levels of the plurality of cytokines are determined 2 to 6 hours after administration of the immunomodulatory agent. (Item 9) 2. The method of claim 1, wherein the blood levels of the cytokines are determined 6 hours after administration of the immunomodulatory agent. (Item 10) The immunomodulatory agent is selected from the group consisting of anti-CD28 monoclonal antibody (mAb), anti-CD3 mAb, anti-CD20 mAb, anti-CD52 mAb; granulocyte colony-stimulating factor (G-CSF); interferon; imiquimod; thalidomide, lenalidomide, pomalidomide, apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; talimogene laherparepvec; adalimumab, catumaxomab, ibritumomab tiuxetan, tositumomab-I 131 , brentuximab vedotin, cetuximab, rituximab, alemtuzumab, bevacizumab, pertuzumab, trastuzumab, trastuzumab emtansine, denosumab, ofatumumab, panitumumab, pembrolizumab, nivolumab, ipilimumab, atezolizumab, avelumab, durvalumab, daratumumab, ceritinib, elotuzumab and antithymocyte globulin. (Item 11) The method of claim 10, wherein the anti-CD28 mAb is TGN1412. (Item 12) The method of claim 10, wherein the anti-CD3 mAb is OKT3. (Item 13) The method of claim 10, wherein the anti-C20 mAb is rituximab. (Item 14) 11. The method of claim 10, wherein the anti-CD52 mAb is alemtuzumab. (Item 15) 1. A method for determining whether a combination of a first immunomodulatory agent and a second immunomodulatory agent is likely to induce severe cytokine release syndrome in a human following administration of said combination of immunomodulatory agents, comprising: (a) preparing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; (b) 1.5–3.0 × 10 isolated from humans 7 engrafting the mouse with peripheral blood mononuclear cells (PBMCs); (c) administering to the mouse a first immunomodulatory agent and a second immunomodulatory agent 5 to 7 days after engraftment of the PBMCs; (d) determining blood levels of a plurality of cytokines in said mouse, including IFN-γ and IL-10; blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse; (e) determining that said combination of immunomodulatory agents is likely to induce severe cytokine release syndrome in said human; the presence of severe cytokine release syndrome in the mouse indicates that administration of the combination of immunomodulatory agents is likely to induce severe cytokine release syndrome in the human. The method includes: (Item 16) The method of claim 15, wherein the mouse is an NSG, NSG-IL-6 or NSG-CSF-1 mouse. (Item 17) The method of claim 15, wherein the mouse is an NSG mouse. (Item 18) 16. The method of claim 15, wherein the mouse in step (a) is irradiated with 100 cGy of X-rays. (Item 19) The engraftment step (b) is performed at a concentration of 2×10 7 The method of item 15, wherein the method is performed on PBMCs. (Item 20) Item 16. The method of item 15, wherein the administration step (c) occurs 6 days after engraftment. (Item 21) 16. The method of claim 15, wherein the plurality of cytokines in step (d) further comprises IL-2, IL-4, IL-6 and TNF. (Item 22) 16. The method of claim 15, wherein the blood levels of the cytokines are determined 2 to 6 hours after administration of the combination of immunomodulatory agents. (Item 23) 16. The method of claim 15, wherein the blood levels of the cytokines are determined 6 hours after administration of the combination of immunomodulatory agents. (Item 24) the first immunomodulatory agent and the second immunomodulatory agent are selected from the group consisting of anti-CD28 monoclonal antibody (mAb), anti-CD3 mAb, anti-CD20 mAb, anti-CD52 mAb; granulocyte colony-stimulating factor (G-CSF); interferon; imiquimod; thalidomide, lenalidomide, pomalidomide, apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; talimogene laherparepvec; adalimumab, catumaxomab, ibritumomab tiuxetan, tositumomab-I 131 , brentuximab vedotin, cetuximab, rituximab, alemtuzumab, bevacizumab, pertuzumab, trastuzumab, trastuzumab emtansine, denosumab, ofatumumab, panitumumab, pembrolizumab, nivolumab, ipilimumab, atezolizumab, avelumab, durvalumab, daratumumab, ceritinib, elotuzumab, and antithymocyte globulin. (Item 25) 25. The method of item 24, wherein the first immunomodulatory agent is pembrolizumab or nivolumab and the second immunomodulatory agent is lenalidomide, pomalidomide, epacadostat, talimogene laherparepvec, ipilimumab, atezolizumab, avelumab, rituximab, alemtuzumab, ceritinib, daratumumab or durvalumab. (Item 26) 25. The method of item 24, wherein the first immunomodulatory agent is ipilimumab and the second immunomodulatory agent is lenalidomide, pomalidomide, pembrolizumab, atezolizumab, avelumab, rituximab, alemtuzumab, ceritinib, daratumumab or durvalumab. (Item 27) 25. The method of item 24, wherein the first immunomodulatory agent is atezolizumab, avelumab, or durvalumab, and the second immunomodulatory agent is lenalidomide, pomalidomide, pembrolizumab, ipilimumab, rituximab, ceritinib, daratumumab, or alemtuzumab. (Item 28) 25. The method of claim 24, wherein the anti-CD52 mAb is alemtuzumab, the anti-C20 mAb is rituximab, the anti-CD3 mAb is OKT3, or the anti-CD28 mAb is TGN1412. (Item 29) 1. A method for determining a safe dosage of an immunomodulatory agent that does not induce cytokine release syndrome in a human following administration of the agent, comprising: (a) providing an immunomodulatory drug having a first dosage, wherein the first dosage of the immunomodulatory drug is determined to induce mild or severe cytokine release syndrome in a first humanized irradiated immunodeficient mouse following administration thereof; (b) providing a second immunodeficient mouse, the second mouse being irradiated with 75-125 cGy of X-rays; (c) 1.5–3.0 × 10 isolated from humans 7 engrafting the second mouse with peripheral blood mononuclear cells (PBMCs); (d) administering an immunomodulatory agent to the second mouse 5-7 days after engrafting the PBMCs, wherein the immunomodulatory agent is administered at a second dosage that is less than the first dosage; (e) determining blood levels of a plurality of cytokines in the second mouse, including IFN-γ and IL-10; (f) determining a safe dosage of the immunomodulatory agent for administration in the human, the safe dosage being a dosage that results in blood levels of IFN-γ <300 pg / ml and IL-10 <25 pg / ml following administration of the immunomodulatory agent to the second mouse; blood levels of IFN-γ<300 pg / ml and IL-10<25 pg / ml in the second mouse, indicating that administration of the safe dosage of the immunomodulatory agent is likely not to induce cytokine release syndrome in the human; The method includes: (Item 30) 30. The method of claim 29, wherein the second mouse is an NSG, NSG-IL-6 or NSG-CSF-1 mouse. (Item 31) 30. The method of claim 29, wherein the second mouse is an NSG mouse. (Item 32) 30. The method of claim 29, wherein the second mouse in step (b) is irradiated with 100 cGy of X-rays. (Item 33) The engraftment step (c) is performed at a concentration of 2×10 7 30. The method of item 29, wherein the method is performed on PBMCs. (Item 34) 30. The method of claim 29, wherein the administering step (d) occurs 6 days after engraftment. (Item 35) 30. The method of claim 29, wherein the plurality of cytokines in step (d) further comprises IL-2, IL-4, IL-6 and TNF. (Item 36) 30. The method of item 29, wherein the blood levels of the plurality of cytokines are determined 2 to 6 hours after administration of the immunomodulatory agent. (Item 37) 30. The method of claim 29, wherein the blood levels of the plurality of cytokines are determined 6 hours after administration of the immunomodulatory agent. (Item 38) The immunomodulatory agent is selected from the group consisting of anti-CD28 monoclonal antibody (mAb), anti-CD3 mAb, anti-CD20 mAb, anti-CD52 mAb; granulocyte colony-stimulating factor (G-CSF); interferon; imiquimod; thalidomide, lenalidomide, pomalidomide, apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; talimogene laherparepvec; adalimumab, catumaxomab, ibritumomab tiuxetan, tositumomab-I 131 30. The method of claim 29, wherein the IL-16 receptor agonist (IL-16 receptor agonist) is selected from the group consisting of: brentuximab vedotin, cetuximab, rituximab, alemtuzumab, bevacizumab, pertuzumab, trastuzumab, trastuzumab emtansine, denosumab, ofatumumab, panitumumab, pembrolizumab, nivolumab, ipilimumab, atezolizumab, avelumab, durvalumab, daratumumab, ceritinib, elotuzumab, and antithymocyte globulin. (Item 39) 39. The method of claim 38, wherein the anti-CD28 mAb is TGN1412. (Item 40) 39. The method of claim 38, wherein the anti-CD3 mAb is OKT3. (Item 41) 39. The method of claim 38, wherein the anti-C20 mAb is rituximab. (Item 42) 39. The method of claim 38, wherein the anti-CD52 mAb is alemtuzumab. (Item 43) 1. A method for determining the immunotoxicity of a drug candidate for use in humans, comprising: (a) preparing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; (b) 4.5–5.5 × 10 7 (c) engrafting the mouse with human PBMCs; and (c) administering a drug candidate to the mouse 4-7 days after engraftment. (d) determining a cytokine concentration in the blood of the mouse, wherein the cytokine is at least one cytokine selected from the group consisting of IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNF; (e) determining the immunotoxicity of the drug candidate, IFN-γ ≥ 300 pg / ml, IL-2 ≥ 15 pg / ml, IL-4 ≥ 10 pg / ml, IL-6 ≥ 10 pg / ml, IL-10 ≥ 25 pg / ml, and TNF ≥ 5 pg / ml wherein the blood concentration in said mouse of at least one cytokine selected from the group consisting of is indicative of the immunotoxicity of said drug candidate in humans. The method includes: (Item 44) 43. The method of claim 42, wherein the mouse is an NSG, NSG-IL-6 or NSG-CSF-1 mouse. (Item 45) 43. The method of claim 42, wherein the mouse is an NSG mouse. (Item 46) 43. The method of claim 42, wherein the mouse in step (a) is irradiated with 100 cGy of X-rays. (Item 47) The engraftment step (b) is 5×10 7 The method of claim 42, wherein the method is performed on PBMCs. (Item 48) Item 43. The method of item 42, wherein the administering step (c) occurs 6 days after engraftment. (Item 49) 43. The method according to item 42, wherein the blood concentration of the cytokine is determined 2 to 6 hours after administration of the drug candidate. (Item 50) 43. The method of claim 42, wherein the blood concentration of the cytokine is determined 6 hours after administration of the drug candidate. (Item 51) 1. A method for determining the likelihood that administration of an immunomodulatory agent to a human will induce severe cytokine release syndrome in said human, comprising the steps of: (a) Human derived 1.5–3.0 × 10 7 providing a blood sample from a humanized irradiated immunodeficient mouse that has been administered an immunomodulatory drug 5-7 days after engraftment of isolated peripheral blood mononuclear cells (PBMCs); (b) detecting in vitro the concentration of a plurality of cytokines, including IFN-γ and / or IL-10, present in the blood sample of the mouse, a concentration of IFN-γ≧1,800 pg / ml or IL-10≧120 pg / ml in the mouse blood sample indicates that administration of the immunomodulatory agent to the human is likely to induce severe cytokine release syndrome. The method includes: (Item 52) 1. A method of determining the likelihood that administration of a combination of a first immunomodulatory agent and a second immunomodulatory agent to a human will induce severe cytokine release syndrome in said human, comprising: (a) Human derived 1.5–3.0 × 10 7 providing a blood sample from a humanized irradiated immunodeficient mouse administered the combination of a first immunomodulatory agent and a second immunomodulatory agent 5-7 days after engraftment of isolated peripheral blood mononuclear cells (PBMCs); (b) detecting in vitro the concentration of a plurality of cytokines, including IFN-γ and / or IL-10, present in the blood sample of the mouse, wherein a concentration of IFN-γ≧1,800 pg / ml or IL-10≧120 pg / ml indicates that administration of the combination of the first immunomodulatory agent and the second immunomodulatory agent to the human is likely to induce severe cytokine release syndrome; The method includes: (Item 53) 1. A method for determining the immunotoxicity of a drug candidate in a human, comprising: (a) 4.5-5.5 × 10 7providing a blood sample from a humanized irradiated immunodeficient mouse that has been administered the drug candidate 4-7 days after engraftment with isolated human peripheral blood mononuclear cells (PBMCs); (b) detecting in vitro the concentration of at least one human cytokine present in the mouse blood sample to determine the human immunotoxicity of the drug candidate, wherein the at least one human cytokine is selected from the group consisting of IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNF, and wherein the drug candidate has low human immunotoxicity when a low human cytokine concentration is detected in the mouse blood sample. The method includes:

[0019] Non-limiting example embodiments are described herein with reference to the following accompanying drawings: [Brief description of the drawings]

[0020] [Figure 1] Figure 1A depicts NSG and NSG-CSF-1 mouse body weight measurements after engraftment of human peripheral blood mononuclear cells (hPBMCs), 2x107 hPBMCs / mouse. There were 10 mice per group and data are presented as mean ± SEM.

[0021] Figure 1B shows the 2 x 10 7 FIG. 1 depicts NSG mouse body weight measurements after hPBMC / mouse engraftment. Each line represents one mouse.

[0022] [Diagram 2] Figure 2A depicts NSG, NSG-IL-6 and NSG-CSF-1 mouse reconstitution of human PBMCs 5 days after engraftment of 2x107 hPBMCs / mouse from donor 331. There were 5 mice per group and data are presented as mean ± SEM.

[0023] Figure 2B shows 2 × 10 IgG from four different donors, 362, 213, 309 and 364, in NSG mice. 7Depicting different cell populations at 5 or 10 days after engraftment of hPBMCs / mouse. 2-5 mice per group for each time point and data are presented as mean ± SEM.

[0024] Figure 2C shows 3 × 10 IgG from donor 358 in NSG mice. 7 Different cell populations are depicted at days 5 or 10 post-engraftment of hPBMCs / mouse. There were 4 mice per group per time point and data are presented as mean ± SEM.

[0025] [Figure 3-1] Figure 3A depicts body weight measurements of five NSG mice engrafted with donor 4692, 5x107 hPBMCs / mouse. Data are mean ± SEM.

[0026] Figure 3B shows donor 362, 5 x 10 7 Body weight measurements of five NSG mice engrafted with hPBMCs / mouse are depicted. Data are mean ± SEM.

[0027] [Figure 3-2] Figure 3C depicts body weight measurements of five NSG mice engrafted with donor 309, 2x107 hPBMCs / mouse. Data are mean ± SEM.

[0028] Figure 3D shows donor 358, 3 x 10 7 Body weight measurements of four NSG mice engrafted with hPBMCs / mouse are depicted. Data are mean ± SEM.

[0029] [Figure 4-1]Figure 4A depicts a comparison between PBMC reconstitution of 10 donors in humanized mice on day 5 (2x107 hPBMCs / mouse engraftment). Humanized NSG mice were tested for the indicated immune cell subset reconstitution by flow cytometry. Shown are human CD45+ cells as a percentage of total cells, as well as CD3, CD19, CD14 and CD56 as a percentage of CD45+ cells (gated on CD45+ cells) (only donors A4692, A4625 and A4668 showed CD45, CD3, CD19 and CD14, but not CD56).

[0030] [Figure 4-2] FIG. 4B depicts de-identified patient-donor information prior to engraftment (excluding donors A4625 and A4668).

[0031] [Figure 5-1] Figures 5A-F depict the induction of cytokines after injection of mAbs into humanized NSG mice with hPBMCs (2x107 PBMCs / mouse) from 10 different donors. Mice were injected iv with 0.5mg / kg OKT3 or 1mg / kg anti-CD28 (ANC28.1 / 5D10 mAb) or PBS (control). Mice were bled at 2 and 6 hours and circulating cytokine concentrations were measured by BD CBA Th1 / Th2 II kit. Cytokine concentrations are shown in Figures 5A-F. The number of mice per group was 2-5 and data are presented as mean ± SEM.

[0032] FIG. 5A depicts IFN-γ levels after induction following mAb challenge: severe / high response: ≧1,800 pg / ml; moderate / mild response: 300 pg / ml≦IFN-γ<1,800 pg / ml; and low / no response: <300 pg / ml.

[0033] FIG. 5B depicts IL-10 levels: severe / high response: ≧120 pg / ml; moderate / mild response: 25 pg / ml≦IL-10<120 pg / ml; and low / no response: <25 pg / ml.

[0034] [Figure 5-2] FIG. 5C depicts IL-6 levels in humanized mice after cytokine induction following mAb loading.

[0035] FIG. 5D depicts IL-2 levels in humanized mice after cytokine induction following mAb loading.

[0036] [Figure 5-3] FIG. 5E depicts IL-4 levels in humanized mice after cytokine induction following mAb loading.

[0037] FIG. 5F depicts TNF levels in humanized mice after cytokine induction following mAb loading.

[0038] [Figure 6] Figure 6 depicts the different responses among 10 donors receiving OKT3 or anti-CD28 mAb. After anti-CD28 mAb administration, donors A4692, A4668 and 362 responded with severe / high levels of IFN-γ and IL-10; donors A4625, 266, 345, 309 and 213 responded with moderate / mild levels of IFN-γ and IL-10; donors 364 and 353 responded with low / none levels of IFN-γ and IL-10.

[0039] [Figure 7] Figure 7 depicts the change in rectal temperature in humanized NSG mice engrafted with hPMBCs from 10 donors and subsequently injected with control PBS, OKT3 or anti-CD28 mAb. The number of mice per group ranged from 2 to 5 and data are presented as mean ± SEM.

[0040] [Figure 8]Figure 8 depicts the clinical scores of 10 donor humanized mice after injection of immunomodulatory drugs. Two and six hours after PBS, OKT3 or anti-CD28 mAb injection, the clinical score of each mouse was assessed according to the following criteria: score: 0 = normal activity; 1 = normal activity, piloerection, walking on tiptoes; 2 = hunched, less active but still mobile; 3 = less mobile but mobile when prompted; 4 = moribund (unresponsive to touch). The number of mice per group ranged from 2 to 5, and data are presented as mean ± SEM.

[0041] [Figure 9-1] 9A-9F depict a comparison of cytokine responses in donor 213 humanized mice with 2×10 7 PBMCs / mouse or 5×10 7 PBMCs / mouse.

[0042] Figure 9A shows the 5×10 7 PBMC / mouse, but not more than 1,800 pg / ml. 7 IFN-γ levels are not elevated in 10 PBMC / mouse.

[0043] Figure 9B shows the 5×10 7 PBMC / mouse, but not above 120 pg / ml. 7 IL-10 levels are not elevated in PBMC / mouse.

[0044] [Figure 9-2] FIG. 9C depicts IL-6 levels from 2×10 7 PBMCs / mouse and 5×10 7 PBMCs / mouse.

[0045] Figure 9D shows the 2 x 10 7 PBMCs / mouse and 5 x 10 7 IL-2 levels by PBMC / mouse are depicted.

[0046] [Figure 9-3]FIG. 9E depicts IL-4 levels from 2×10 7 PBMCs / mouse and 5×10 7 PBMCs / mouse.

[0047] Figure 9F shows the 2 x 10 7 PBMCs / mouse and 5 x 10 7 6 depicts TNF levels by PBMC / mouse.

[0048] [Figure 10-1] Figures 10A-F depict induction of cytokines after injection of antibodies in humanized NSG mice with different human PBMC concentrations (2x107, 3x107 and 4x107 per mouse from donor 309). Mice were injected iv with 0.5mg / kg OKT3, 1mg / kg anti-CD28, 10mg / kg KEYTRUDA® (pembrolizumab) or PBS as control. Mice were bled at 2 and 6 hours and circulating cytokine concentrations were measured by BD CBA Th1 / Th2 II kit. The number of mice per group was 3-5 and data are presented as mean ± SEM.

[0049] Figure 10A shows the 2 x 10 7 PBMCs / mouse, 3 x 10 7 PBMCs / mouse and 4 x 10 7 IFNγ levels by PBMC / mouse are depicted.

[0050] Figure 10B shows the 2 x 10 7 PBMCs / mouse, 3 x 10 7 PBMCs / mouse and 4 x 10 7 IL-10 levels by PBMC / mouse are depicted.

[0051] [Figure 10-2] FIG. 10C depicts IL-6 levels from 2×10 7 PBMCs / mouse, 3×10 7 PBMCs / mouse and 4×10 7 PBMCs / mouse.

[0052] Figure 10D shows the 2 x 107 PBMCs / mouse, 3 x 10 7 PBMCs / mouse and 4 x 10 7 IL-2 levels by PBMC / mouse are depicted.

[0053] [Figure 10-3] FIG. 10E depicts IL-4 levels from 2×10 7 PBMCs / mouse, 3×10 7 PBMCs / mouse and 4×10 7 PBMCs / mouse.

[0054] Figure 10F shows the 2 x 10 7 PBMCs / mouse, 3 x 10 7 PBMCs / mouse and 4 x 10 7 6 depicts TNF levels by PBMC / mouse.

[0055] [Figure 11] FIG. 11A-FIG. 11D depict temperature and clinical score changes in response to OKT3, anti-CD28 and KEYTRUDA® (pembrolizumab) in donor 309 PBMC humanized mice.

[0056] FIG. 11A depicts the donor 309 mouse cell population on day 5. Five NSG mice were inoculated with 2×10 7 Engraft 3 x 10 hPBMCs / mouse; 5 NSG mice, 3 x 10 7 hPBMCs / mouse were engrafted into five NSG mice, with 4 × 10 7 hPBMCs / mouse were engrafted. All mice were engrafted 4 hours after 100 cGy X-ray irradiation. Mice were bled on day 5 for CD45, CD3, CD19, CD14 and CD56 cell population examination by flow cytometry.

[0057] Figure 11B depicts the clinical scores of humanized mice 2 and 6 hours after injection of drugs: control PBS, 0.5 mg / kg OKT3, 1 mg / kg anti-CD28 or 10 mg / kg KEYTRUDA® (pembrolizumab). Clinical scores were assessed by the following criteria: score: 0=normal activity; 1=normal activity, piloerection, walking on tiptoes; 2=hunched, less active but still mobile; 3=less mobile but mobile when prompted; 4=moribund. 5 mice / group, data presented as mean±SEM.

[0058] FIG. 11C shows a 2×10 7 Measured rectal temperatures are shown using hPMBCs / mouse. Mice were injected with control PBS, 0.5 mg / kg OKT3, 1 mg / kg anti-CD28 or 10 mg / kg KEYTRUDA® (pembrolizumab). Rectal temperatures were measured 2 and 6 hours after drug injection. There were 3-5 mice / group and data are presented as mean ± SEM.

[0059] FIG. 11D shows a 3×10 7 Measured rectal temperatures are shown using hPMBCs / mouse. Mice were injected with control PBS, 0.5 mg / kg OKT3, 1 mg / kg anti-CD28 or 10 mg / kg KEYTRUDA® (pembrolizumab). Rectal temperatures were measured 2 and 6 hours after drug injection. There were 3-5 mice / group and data are presented as mean ± SEM.

[0060] FIG. 11E shows a 4×10 7 Measured rectal temperatures are shown using hPMBCs / mouse. Mice were injected with control PBS, 0.5 mg / kg OKT3, 1 mg / kg anti-CD28 or 10 mg / kg KEYTRUDA® (pembrolizumab). Rectal temperatures were measured 2 and 6 hours after drug injection. There were 3-5 mice / group and data are presented as mean ± SEM.

[0061] [Figure 12-1]FIG. 12A-FIG. 12F depict the dose effect of KEYTRUDA® (pembrolizumab) on induction of different cytokine levels by different doses of KEYTRUDA®. NSG mice were engrafted with 3×107 hPBMCs / mouse from donor 358 6 days prior to KEYTRUDA® dosing. On the day of dosing, mice were injected iv with PBS, 0.5 mg / kg OKT3, 2.5 mg / kg, 5 mg / kg, and 10 mg / kg KEYTRUDA®. Mice were bled at 2 and 6 hours and circulating cytokine concentrations were measured by BD CBA Th1 / Th2 II kit. 5 mice / group, data presented as mean±SEM.

[0062] Figure 12A shows the 3×10 7 IFNγ levels by PBMC / mouse are depicted.

[0063] Figure 12B shows the 3×10 7 IL-10 levels by PBMC / mouse are depicted.

[0064] [Figure 12-2] FIG. 12C depicts IL-6 levels with 3×10 7 PBMC / mouse.

[0065] Figure 12D shows a 3 x 10 7 IL-2 levels by PBMC / mouse are depicted.

[0066] [Figure 12-3] FIG. 12E depicts IL-4 levels with 3×10 7 PBMC / mouse.

[0067] Figure 12F shows the 3×10 7 6 depicts TNF levels by PBMC / mouse.

[0068] [Figure 13]FIG. 13A-FIG. 13C depict body temperature and clinical score changes in response to KEYTRUDA® (pembrolizumab) in donor 358 PBMC humanized mice.

[0069] Figure 13A depicts the donor 358 mouse cell population on day 5. Four (4) NSG mice were injected with 3 x 10 7 hPBMCs were engrafted and mice were bled on day 5 for CD45, CD3, CD19, CD14 and CD56 cell population examination by flow cytometry.

[0070] Figure 13B depicts the humanized mice clinical scores after drug injection. After 2 and 6 hours of PBS, 0.5 mg / kg OKT3, 2.5 mg / kg, 5 mg / kg and 10 mg / mg KEYTRUDA® (pembrolizumab), the clinical scores were evaluated according to the following criteria: score: 0=normal activity; 1=normal activity, piloerection, toe walking; 2=hunched, less active but still mobile; 3=less mobile but mobile when prompted; 4=moribund. 5 mice / group, data presented as mean±SEM.

[0071] Figure 13C depicts rectal temperature measured in humanized mice at different time points. Mice were injected with control PBS, 0.5 mg / kg OKT3, 2.5 mg / kg, 5 mg / kg and 10 mg / kg KEYTRUDA® (pembrolizumab) and rectal temperature was measured 2 and 6 hours after drug injection. There were 5 mice / group and data are presented as mean ± SEM.

[0072] [Figure 14-1] 14A-B depict the differential results of in vitro and in vivo measurements of cytokine release following drug exposure.

[0073] Figure 14A presents histograms comparing cytokine release levels in response to anti-CD28 treatment obtained by in vitro (panels a and c) or in vivo (panels b and d) experiments by four different PBMC donors for IFN-γ (panels a and b) or IL-10 (panels c and d). Data are presented as mean ± SEM.

[0074] [Figure 14-2] Figure 14B presents histograms comparing cytokine release levels in response to anti-CD28 treatment obtained by in vitro (panels e and g) or in vivo (panels f and h) experiments with four different PBMC donors for IL6 (panels e and f) or IL-4 (panels g and h). Data are presented as mean ± SEM.

[0075] [Figure 15] Figure 15 presents histograms of cytokine release levels in response to anti-CD28 treatment obtained by in vitro or in vivo experiments with PBMC donor 213 for a) IFN-γ; b) IL-10; c) IL-6; and d); IL-4. Data are presented as mean ± SEM. The dotted line in each graph is the control level for the anti-CD28 experiment.

[0076] [Figure 16] FIG. 16 presents histograms of cytokine release levels in response to treatment with KEYTRUDA® (pembrolizumab), REVLIMID® (lenalidomide) or both drugs for PBMC donor 213 or donor 364 for a) IFN-γ; b) IL-6; c) IL-4 d) IL-10; e) IL-2; and f) TNF. Treatment with PBS was the control. Data are presented as mean ± SEM.

[0077] [Figure 17]Figure 17 presents histograms of cytokine release levels in response to treatment with KEYTRUDA® (pembrolizumab), ATG or both drugs for PBMC donor 213 or donor 364 for a) IFN-γ; b) IL-10; c) IL-6 d) IL-2; e) IL-4; and f) TNF. Treatment with PBS was the control. Data are presented as mean ± SEM.

[0078] [Figure 18] Figure 18 presents histograms of cytokine release levels in response to treatment with anti-CD28, ATG or both drugs for PBMC donor 213 or donor 364 for a) IFN-γ; b) IL-6; c) IL-4 d) IL-2; e) IL-4; and f) TNF. Treatment with PBS was the control. Data are presented as mean ± SEM.

[0079] [Figure 19] FIG. 19 illustrates day 5 and day 10 cell populations in whole blood of humanized mice with or without irradiation prior to engraftment of human PBMCs; mice were humanized with PBMCs from one of six different donors: 362, 345 and 2785 (top graph) and 213, 364 and 3251 (bottom graph).

[0080] [Figure 20] FIG. 20 shows weight loss as a function of days after PBMC engraftment for mice humanized with PBMC from one of six different donors: 362, 345 and 2785 (panel a) and 213, 364 and 3251 (panel b) after irradiation (IR) and without irradiation (non-IR).

[0081] [Figure 21] FIG. 21 shows cytokine levels in PBMC humanized mice without any drug treatment, 10 days after PBMC engraftment.

[0082] [Figure 22-1] Figures 22A-B present graphs of the levels of cytokines released after drug treatment (by irradiated or non-irradiated mice humanized with PBMCs from either donor 362 or 213). Drugs (OKT3, KEYTRUDA, or ATG) or negative control (PBS) were administered on day 6 after PBMC engraftment.

[0083] FIG. 22A presents graphs of the levels of IFN-γ (panel a), IL-10 (panel b) and IL-6 (panel c).

[0084] [Figure 22-2] FIG. 22B presents graphs of the levels of IL-2 (panel d), IL-4 (panel e) and TNF (panel f). DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0085] Definition: In describing example embodiments, specific terminology is used for the sake of clarity, but the embodiments are not intended to be limited to this specific terminology. Unless otherwise noted, technical terms are used according to conventional usage.

[0086] As used herein, "a" or "an" can mean one or more. As used herein, "another" can mean at least a second or more. Further, unless otherwise required by context, singular terms include plurals and plural terms include the singular.

[0087] The term "NSG" refers to the immunodeficient mouse model of NOD scid gamma (i.e., NOD.Cg-Prkdc scid Il2rg tm1WjlThis refers to the NOD / ShiLtJ mouse (Jackson laboratory strain number: 005557). This mouse carries two mutations on the NOD / ShiLtJ genetic background; severe combined immunodeficiency (scid) and a complete null allele of the interleukin-2 (IL2) receptor common gamma chain (IL2rg null). These mice are severely immunodeficient.

[0088] The term "NSG-CSF-1" refers to the NSG mouse model, whose genome contains the human CSF-1 (macrophage colony-stimulating factor-1) gene, and transgenic NSG-CSF-1 mice express the human CSF-1 cytokine (Jackson Laboratory line number: 028654).

[0089] The term "NSG-IL-6" refers to the NSG mouse model, whose genome contains the human interleukin-6 (IL-6) gene and expresses the human IL-6 cytokine (Jackson Laboratory strain number: 028655).

[0090] The term "CD" refers to cluster of differentiation.

[0091] The term "CD3" refers to cluster of differentiation 3 and represents an antigen that is part of the T cell receptor (TCR) complex on mature T lymphocytes.

[0092] The term "CD4" refers to cluster of differentiation 4, which is a glycoprotein found on the surface of immune cells such as helper T cells, monocytes, macrophages, and dendritic cells.

[0093] The term "CD8" refers to cluster of differentiation 8, a co-receptor expressed primarily on the surface of cytotoxic T cells, but can also be found on natural killer cells, cortical thymocytes and dendritic cells.

[0094] The term "CD14" refers to cluster of differentiation 14, an antigen expressed primarily by macrophages and dendritic cells.

[0095] The term "CD19" refers to cluster of differentiation 19, an antigen found on B cells.

[0096] The term "CD28" refers to cluster of differentiation 28, a protein expressed in T cells that provides a costimulatory signal required for T cell activation and survival. T cell stimulation via CD28 in addition to the T cell receptor (TCR) can provide a strong signal for the production of various interleukins (e.g., IL-6).

[0097] The term "CD45 cell" refers to cluster of differentiation 45, an antigen present on human lymphocytes, monocytes and other myeloid cells.

[0098] The term "CD56" refers to cluster of differentiation 56, a homophilic binding glycoprotein expressed on the surface of natural killer cells (NK cells).

[0099] The term "peripheral blood mononuclear cells (PBMCs)" refers to peripheral blood cells with round nuclei. Such mononuclear blood cells recirculate between tissues and blood and are a critical component of the immune system to fight infection and adapt to invaders. There are two major types of mononuclear cells: lymphocytes and monocytes. The lymphocyte population of PBMCs typically consists of T cells, B cells and NK cells. PBMCs can be isolated from whole blood samples by methods well known in the art (e.g., Ficoll gradient).

[0100] The term "cytokine" refers to members of a class of small proteins (approximately 5-20 kDa) that are important in cell signaling. Cytokines include chemokines, interferons, interleukins, lymphokines and tumor necrosis factors. Examples of cytokines include IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNFα. Cytokines are produced by a wide range of cells, including immune cells such as macrophages, B lymphocytes, T lymphocytes and mast cells; a given cytokine may be produced by more than one type of cell. Its release has an effect on the behavior of cells in its surroundings. Cytokines have been identified as immunomodulators involved in autocrine, paracrine and endocrine signaling.

[0101] The term "cytokine release syndrome" ("CRS") is used herein interchangeably with "systemic inflammatory response syndrome" ("SIRS"), "cytokine cascade", "hypercytokinemia" and "cytokine storm". "Cytokine storm", also known in the art as "hypercytokinemia", defines a systemic inflammatory response in a patient that is characterized inter alia by hypotension, fever and / or chills, potentially resulting in death. Cytokine storm is likely due to an uncontrolled positive feedback loop between cytokines and immune cells that results in highly elevated levels of various cytokines. Although these terms may vary somewhat in severity, all of them are the result of an unacceptably high release of cytokines by a subject as a result of administration of a certain antibody to the subject. The subject responds to the treatment by releasing unacceptably high levels of cytokines. Reference herein to one of these terms is intended to encompass all terms.

[0102] The term "grading of cytokine release syndrome (CRS)" was introduced by Daniel W. Lee et al., "Current Concepts in the Diagnosis and Management of Cytokine Release Syndrome (CRS)." Grade 1: symptoms are not life-threatening and require only symptomatic measures (e.g., fever, nausea, fatigue, headache, myalgia, malaise); Grade 2: symptoms require and respond to moderate intervention, oxygen requirement <40%, or hypotension responsive to fluids or low doses of one vasopressor, or grade 2 organ toxicity; Grade 3: symptoms require and respond to aggressive intervention, oxygen requirement ≥40%, or hypotension requiring high doses or multiple vasopressors, or grade 3 organ toxicity or grade 4 hypertransaminasemia; Grade 4: life-threatening symptoms, requiring ventilator support, or grade 4 organ toxicity (except hypertransaminasemia); Grade 5: death. For purposes of this application, severe CRS refers to grades 4-5, and mild CRS includes grades 1-3.

[0103] The inventors have determined that the concentrations of certain cytokines released in a humanized mouse model in response to an administered drug or drug candidate can be used to predict the relative severity of cytokine release syndrome expected in humans in response to an administered drug or drug candidate. The inventors have determined that the concentrations of certain cytokines released in a humanized mouse model in response to an administered drug or drug candidate can be used to predict the relative severity of cytokine release syndrome expected in humans in response to an administered drug or drug candidate. 7 For immunodeficient mice engrafted with PBMCs, thresholds for mouse blood concentrations of certain cytokines were determined to determine the severity of cytokine release induced in mice by drugs administered to the mice 5-7 days after engraftment. The thresholds are summarized in the following table. [Table 1]

[0104] The mouse cytokine threshold concentration values ​​in Table 1 were determined from experimental data in conjunction with available literature reports and have a variability of ±10%.

[0105] The mouse cytokine threshold concentration values ​​for a severe / severe response correspond to human CRS grades 4-5; the mouse cytokine threshold concentration values ​​for a moderate / mild response correspond to human CRS grades 1-3; and the mouse cytokine threshold concentration values ​​for a moderate / mild response correspond to human CRS grades <1.

[0106] 1.5–3.0 × 10 of PBMC in mice 7 In the disclosed method of engraftment, the inventors have determined that if a severe / high response is observed in mice, e.g., IFN-γ ≥ 1,800 pg / ml or IL-10 ≥ 120 pg / ml, then there is a likelihood that the human may have severe cytokine release syndrome following administration of the drug or drug candidate. Similarly, 1.5-3.0 x 10 PBMC in mice 7 In the disclosed method of engraftment, the inventors have determined that if low / no response is observed in mice, e.g., IFN-γ<300 pg / ml or IL-10<25 pg / ml, then there is a chance that a human will not experience severe cytokine release syndrome following administration of the drug or drug candidate, but may instead have at most low levels of cytokine release. Between these thresholds, 1.5-3.0 x 10 7 The concentration of IFN-γ or IL-10 in mice engrafted with PBMCs is designated as moderate / mild response, indicating that humans may experience moderate / mild cytokine release syndrome after administration of a drug or drug candidate. The thresholds of induced IL-6, IL-4, IL-2 or TNFα in mice can be used similarly to assess the severity of the cytokine response to a drug or drug candidate. The thresholds of induced IFN-γ, IL-10, IL-6, IL-4, IL-2 and TNFα in mice can be used alone or in any combination to assess the severity of the cytokine response to a drug or drug candidate in humans.

[0107] The terms "donors", "individuals", "humans", "subjects" and "patients" (and the singular forms of these terms) are used somewhat interchangeably herein. In the studies performed by the inventors, "donor" PBMCs were used. In the present methods, rather than a "donor", PBMCs from the particular human, individual, subject or patient to whom the immunomodulatory agent is being considered for possible administration will be used. The use of one of these terms herein is intended to encompass each of these terms. In the present methods, the individuals, subjects and patients are human. However, it is contemplated that the present methods may be applied to other mammals using the methods and techniques described herein, perhaps with some modifications to the methods that can be determined by one of skill in the art.

[0108] The term "immunomodulator" refers to any therapeutic agent (e.g., mAb) that can activate or suppress the immune system by activating or inhibiting lymphocyte function, e.g., T-cell function, such as, for example, T-cell inhibition or activation. Immunomodulatory drugs or agents, or immunomodulatory substances, or immunotherapeutic drugs, can include, for example, interleukins, cytokines, chemokines, immunomodulatory imide drugs, or other agents that can be used in immunotherapy. As a non-limiting example, cancer immunotherapy attempts to stimulate the immune system to destroy tumors. Thus, immunomodulatory drugs or agents can be used to attempt to treat cancer in patients, but the use of immunomodulatory drugs is not limited to the treatment of cancer. Immunotherapy can be used alone or in combination with other treatment methods. An example embodiment of an immunomodulatory drug is an immunostimulatory drug, such as an antibody, preferably a monoclonal antibody (mAb). For example, the monoclonal antibody can be a human CD28-specific superagonist monoclonal antibody.

[0109] Examples of immunomodulatory agents include granulocyte colony stimulating factor (G-CSF); interferons; imiquimod; thalidomide and its derivatives or analogs, lenalidomide (REVLIMID®), pomalidomide (IMNOVID®), and apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; IMLYGIC™ (talimogene laherparepvec), Genetically modified oncolytic virus therapy; daratumumab (DARZALEX®), anti-CD38 antibodies; adalimumab (HUMIRA®), EMPLICITI™ (elotuzumab), epacadostat, an orally available hydroxyamidine inhibitor of indoleamine 2,3-dioxygenase (IDO1), catumaxomab (REMOVAB®), ibritumomab tiuxetan (ZEVALIN®), tositumomab-I 131 (BEXXAR®), brentuximab vedotin (ADCETRIS®), cetuximab (betuximab) (ERBITUX®), rituximab (MABTHERA® or RITUXAN®), alemtuzumab (CAMPATH-1H®), bevacizumab (AVASTIN®), pertuzumab (PERJETA®), trastuzumab (HERCEPTIN®), trastuzumab emtansinen (KADCYLA™), denosumab (PROLIA® or XGEVA®), ipilimumab (YERVOY®), ofatumumab (ARZERRA®), and panitumumab (VECTIBIX®).

[0110] Checkpoint inhibitors are drugs that block certain proteins made by some types of immune system cells, such as T cells, and some cancer cells. Such proteins can help to suppress immune responses and prevent T cells from killing cancer cells. When such proteins are blocked by checkpoint inhibitors, a "brake" is released on the immune system. This release of immune "brakes" can allow T cells to kill cancer cells more effectively, but can also result in CRS as an adverse effect.

[0111] An exemplary class of immunomodulatory drugs are checkpoint inhibitors, which are often monoclonal antibodies, such as the FDA-approved cancer drugs ipilimumab (YERVOY®), pembrolizumab (KEYTRUDA®), and nivolumab (OPDIVO®). A key aspect of the immune system is its ability to distinguish between normal cells in the body and cells that the immune system considers "foreign." This allows the immune system to attack foreign cells while sparing normal cells. To do this, the immune system uses "checkpoints," which are molecules on certain immune cells that need to be activated (or inactivated) to mount an immune response. Exemplary FDA-approved checkpoint inhibitors include the CTLA-4 inhibitor ipilimumab (YERVOY®), the PD-1 inhibitors pembrolizumab (KEYTRUDA®) and nivolumab (OPDIVO®), and the PD-L1 inhibitors atezolizumab (TECENTRIQ®), avelumab (BAVENCIO®), and durvalumab (IMFINZI®).

[0112] The term "T cell activation" preferably designates the activation mechanism of T cells, which may vary slightly between different types of T cells. However, the "two-signal model" in CD4+ T cells is applicable to most types of T cells. More specifically, activation of CD4+ T cells typically occurs through the engagement of both T cell receptor and CD28 on the T cell surface by major histocompatibility-encoded antigen-presenting molecules with their bound antigenic peptides and B7 family members on the surface of antigen-presenting cells (APCs), respectively. Both cell-cell contacts are generally required for the production of an effective immune response. For example, in the absence of CD28 costimulation, T cell receptor signaling alone can result in T cell anergy. Further signaling pathways downstream from both CD28 and T cell receptor involve many additional proteins known to those skilled in the art. Activation of T cells can be determined by cytokine release and / or cell proliferation, particularly proliferation of T cells, as described herein below.

[0113] The term "clinical score" is adopted as defined by Jamie L Brady et al. "Preclinical Screening for Acute Toxicity of Therapeutic Monoclonal Antibodies in a hu-SCID Model," Clinical & Translational Immunology (2014) 3, e29; doi:10.1038 / cti.2014.28; published online December 19, 2014: 0=normal activity; 1=normal activity, piloerection, walking on tiptoes; 2=hunched, reduced activity but still mobile; 3=reduced mobility but mobile when prompted; 4=moribund (unresponsive to touch).

[0114] The term "dose" or "dosage" means the amount of a drug to be taken by a patient at one time.

[0115] A "safe dosage" of an immunomodulatory agent for humans is defined herein as 1.5-3.0×10 7 It refers to the dosage that results in cytokine blood concentrations in mice that correspond to low / no response, e.g., IFN-γ <300 pg / ml and IL-10 <25 pg / ml, following administration of an immunomodulatory agent to irradiated immunodeficient mice engrafted with PBMCs. An immunologically safe dosage may or may not correspond to a dosage that has therapeutic efficacy.

[0116] "Efficacy" refers to the ability of a drug or active agent administered to a patient to produce a therapeutic effect in the patient.

[0117] A "therapeutically effective amount" or "effective amount" is an amount of a pharmaceutical agent to achieve a pharmacological effect. The term "therapeutically effective amount" includes, for example, a prophylactically effective amount. An "effective amount" of a drug is the amount required to achieve a desired pharmacological effect or therapeutic improvement without undue adverse side effects. The effective amount of a drug will be selected by those skilled in the art depending on the specific patient and disease. It is understood that an "effective amount" or "therapeutically effective amount" may vary from subject to subject due to variations in drug metabolism, age, weight, general condition of the subject, the condition being treated, the severity of the condition being treated, and the judgment of the prescribing physician.

[0118] As used herein, the term "immunotoxicity" refers to the propensity of a drug or drug candidate to produce deleterious immune activation, such as cytokine release syndrome.

[0119] The term "drug candidate" means any potential drug or composition, including one or more active agents, such as antibodies, small molecules and / or other compounds, that have been identified through drug discovery screening as potentially having a therapeutic effect in ameliorating, treating and / or curing a disease, illness, injury, illness or condition.

[0120] In one aspect, the present inventors have devised a humanized mouse model for screening and determining drug immunotoxicity, particularly cytokine release syndrome (CRS), in individuals for preclinical trials, clinical trials and / or individual treatment of individuals with drugs. The present invention is useful for determining an individual's responsiveness to an immunomodulatory drug and for determining a safe dosage for administration of an immunomodulatory drug to an individual.

[0121] In another aspect, the humanized mouse method is also useful in pharmaceutical safety evaluation of drug candidates. In particular, the disclosed method provides improved preclinical animal testing of immunotoxicity of drug candidates in humans.

[0122] The methods for determining drug immunotoxicity disclosed herein have the advantages of the ability to detect individual variability in response to a given immunomodulatory agent or combination of immunomodulatory agents, better response prediction accuracy than prior art methods such as in vitro PBMC cell culture methods, requiring only moderate amounts of blood (100 ml or less) to be drawn from individuals for testing, commercial availability of suitable immunodeficient mice, and a turnover time of less than two weeks.

[0123] With respect to preclinical immunotoxicity testing of drug candidates, the present method has all of the advantages described above and further provides a highly sensitive method for assessing the level of immune stimulation that a drug candidate will induce following administration, allowing early elimination of drug candidates that induce unacceptably high levels of immune stimulation (e.g., cytokine release induced by administration of the drug candidate in humans), thereby saving time and costs in drug development.

[0124] Advantageously, as further described herein, the inventors have found that this assay system predicts in vivo responses and represents a powerful tool in research and drug safety evaluation. Current clinical trials of new drug candidates in volunteer human subjects often result in drug failure. The cause of this failure is in large part due to the lack of toxicity exposure in preclinical trials due to the inadequacy of existing in vivo animal models. There is a long-standing unmet need for in vivo animal models that can accurately predict the adverse effects of potential drug candidates. This humanized mouse model is useful as a drug screening platform with a high degree of accuracy to identify potential drug candidates that induce cytokine release from a large number of clinically relevant drug candidates. Thus, this method represents a robust predictive assay for drug immunotoxicity testing, providing a necessary link between preclinical and clinical trials. Integration of this assay into drug development programs should accelerate the drug approval process, such as pre-FDA, for therapeutic drug development.

[0125] In one aspect, the present invention is directed to a method for determining whether an immunomodulatory drug causes immune toxicity in humans. In other words, the method can serve as a screening assay for individual patients to receive an immunomodulatory drug if the patient is safe to receive the immunomodulatory drug. The method can include collecting peripheral blood mononuclear cells (PBMCs) from a human subject, administering an immunomodulatory drug (e.g., mAb) to an irradiated, immunodeficient non-human mammal (e.g., a humanized immunodeficient mouse) engrafted with the collected PBMCs from the individual, and detecting one or more cytokines (exemplified by IFN-γ and / or IL-10) released, thereby determining whether the immunomodulatory drug causes immune toxicity, and if the immunomodulatory drug induces a severe cytokine storm in the non-human mammal, the immunomodulatory drug causes immune toxicity in the individual.

[0126] There is a long-standing unmet need for a reliable in vivo mouse method to help physicians determine potential CRS in human subjects prior to administration of immunomodulatory drugs. Ideally, the method should accurately determine which human subjects receiving immunomodulatory drugs will suffer from severe CRS adverse events. As reported in March 2006, described above, in a human clinical trial at Northwick Park Hospital involving mAb TGN1412, six volunteers in the trial were hospitalized for adverse events. Many of the patients suffered from angioedema, swelling of the skin and mucous membranes, followed by multiple organ dysfunction, but preclinical studies of this trial, whether in rat models, cynomolgus monkeys (Macaca fascicularis), or in vivo human PBMC cell culture assays, showed no indication of CRS. However, the inventors have discovered that certain patients may or may not have severe CRS adverse events to a given immunomodulatory drug, and have devised a method to predetermine whether a particular patient is likely to suffer from CRS if administered a particular drug.

[0127] Cytokine release syndrome occurs due to activation of T cells and natural killer (NK) cells as well as other immune cell populations (e.g., macrophages, etc.). With the addition of immunomodulators, activation of T cells and natural killer cells can lead to the release of high levels of cytokines and potential downstream injury and death. The role of T cell activation has received much attention due to the TGN1412 clinical trial, but NK cell activation has also been shown to be a source of cytokine release syndrome in response to certain immunomodulators. For example, treatment with the anti-CD52 antibody CAMPATH1-H has been shown to involve NK cells, resulting in the release of high levels of TNF, IFN-γ, and IL-6 in vivo (Wing MG et al. (1995) Ther. Immunol., 2:183-190) as well as IL-2 and I The toxicity resulting from L12 combination treatment has been shown to be the result of activation of natural killer cells, rather than B or T cells (Carson WE, (1999) J Immunol 162:4943-4951). With different immunomodulators and activation of various immune cell populations, cytokine release syndrome can be manifested by high levels of cytokine release that can vary with different activated immune cell populations.

[0128] The present inventors have realized that many variables in this complex biological system must be carefully selected and optimized before mouse models can provide accurate prediction of drug-induced CRS.Unfortunately, at present, there is little guidance in the art on which variables should be selected or how they should be optimized.Given the highly intricate and complex nature of this system, a delicate balance between these variables is required for mouse models to provide useful and accurate prediction of drug-induced CRS.

[0129] The inventors have discovered that the number of human cells (i.e., the number of PBMC administered to the mouse) and the distribution of human cell types present in the mouse (which changes with time after engraftment) are critical variables in this system. It was discovered that when human PBMCs are injected into the mouse, only human T cells can expand; other human cell types, such as NK cells and B cells, will begin to die off. It was further discovered that if the human T cell count becomes too high in the mouse, it will cause GVHD, which is manifested by weight loss, hunched posture, hair loss, reduced mobility, tachypnea, and ultimately death. If the mouse exhibits severe GVHD, the accuracy of the test results is significantly compromised.

[0130] Without being bound by theory, it was unexpectedly discovered that adult human T cells in PBMCs recognize the mouse as foreign and launch an attack on the mouse, causing health problems and possibly death to the mouse, with accompanying cytokine release When this occurs, the mice begin to suffer significant weight loss and exhibit pathological symptoms that make the mouse model inaccurate in determining CRS.

[0131] Without being bound by theory, it is believed that engraftment of too many PBMCs increases the cytokine release profile in mice such that the method provides inaccurate predictions (i.e., prone to providing false positives). 7 It is observed that when the number of engrafted PBMCs / mouse exceeds 1×10, some mice suffer from significant weight loss very rapidly after engraftment, presumably due to GVHD (graft versus host disease). If the mice have GVHD, the cytokine release response in such mice induced by the immunomodulatory drug cannot accurately determine the response of the human subject to the immunomodulatory drug. Alternatively, the number of engrafted PBMCs may be increased beyond a certain threshold (e.g., <1×10 7 If the number of human PBMCs injected is less than 100 / mouse, the method will also not be able to determine the response of human subjects to immunomodulatory drugs with optimal sensitivity and will provide false negatives. Additionally, Applicants irradiate mice with X-rays to destroy the mouse immune system prior to engraftment, minimizing rejection and maximizing survival of the injected human PBMCs.

[0132] Thus, the inventors determined that the number of PBMCs engrafted and the time of PBMC engraftment prior to immunomodulatory drug administration are critical to optimal accuracy of the assay by balancing the presence of appropriate populations for GVHD pathogenesis in mice and the distribution of human immune cells in the mice.

[0133] The PBMC humanized mouse model, unlike earlier BLT (bone marrow / liver / thymus) or stem cell humanized mouse models, has been considered in the art to be a T cell only model because when human PBMCs are injected into the mouse body, over time other human cell types die off while only the T cell population can expand. In the art, PBMC humanized mice are typically used after 10 days of PBMC engraftment with PBMC engraftment numbers typically between 1 million and 10 million cells per mouse, which requires studies to wait for the T cell population to expand to a number large enough to perform experiments. Typically 10% human CD45 or human CD3 T cells present in the mouse cell population have been used as a standard for the minimum number of human cells. However, the inventors have determined that in order for the PBMC humanized mouse model to provide optimal drug toxicity testing, not only T cells but also other cell types, particularly NK cells and monocytes, are required in the mouse.

[0134] Thus, in certain embodiments herein, the invention fulfills a long-standing unmet need, providing an in vivo method for determining whether an immunomodulatory drug induces a cytokine storm response (i.e., severe cytokine release syndrome) in an individual, such as a human, comprising the steps of: (a) collecting or isolating PBMCs from the individual human being considered to receive the immunomodulatory drug; and (b) injecting 1.5×10 PBMCs into irradiated (e.g., immune-deficient mice (NSG, NSG-IL-6 or NSG-CSF-1) mice with 1.5×10 PBMCs to functionally suppress mouse immune cells from attacking the engrafted PBMCs. 7 ~3.0×10 7(c) engrafting collected / isolated PBMCs; (d) administering an immunomodulatory agent to the engrafted immunodeficient mice on days 5-7, preferably day 6, following engraftment; (e) detecting the presence of one or more cytokines released following administration; and (f) evaluating the cytokine response to the immunomodulatory agent relative to a control agent (e.g., a control mAb) to determine whether cytokine release in the mice is at a level indicative of severe cytokine release syndrome by determining blood levels of multiple cytokines, including IFN-γ and IL-10, in the mice.

[0135] In one embodiment, the method comprises administering to the subject an amount of 1.5×10 IgG isolated from a human. 7 ~3.0×10 7The method includes engrafting PBMCs into an irradiated immunodeficient mouse (e.g., NSG, NSG-IL-6 or NSG-CSF-1); administering an immunomodulatory agent to the engrafted immunodeficient mouse 5-7 days, preferably 6 days, after engraftment; detecting the amount of one or more cytokines released after administration; and determining that the immunomodulatory agent is capable of inducing severe cytokine release syndrome in a human, the presence of severe cytokine release syndrome in the mouse indicating that administration of the immunomodulatory agent is capable of inducing severe cytokine release syndrome in the human. The method may further include providing an irradiated immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays, or isolating PBMCs from the human, or comparing the amount of cytokine released after drug administration to the amount released after administration of a negative control agent. Examples of suitable negative control agents include a buffer or an isotype control mAb. The cytokines detected can be IFN-γ, IL-10, IL-6, IL-2, IL-4, TNF or a combination of the foregoing, preferably the cytokine is IFN-γ or IL-10. The assay allows a physician to distinguish (or identify) which human patient(s) are likely to suffer from severe CRS compared to other patients who will not suffer from severe CRS.

[0136] In a preferred example embodiment, prior to engraftment, the immunodeficient mice are irradiated with 75-125 cGy of X-rays or 100 cGy of X-rays. In a further example embodiment, irradiation is administered at least 4 hours prior to engraftment.

[0137] The inventors have established that PBMCs can be collected from individuals (likely to receive an immunomodulatory drug) and used for engraftment in immunodeficient mice (e.g., NSG, NSG-IL-6 or NSG-CSF-1) to obtain humanized mice to test for immunotoxicity (i.e., cytokine release) induced by one or more drugs using the assays disclosed herein. Advantageously, the inventors have surprisingly found that this assay system predicts in vivo responses in human patients who will receive an immunomodulatory drug.

[0138] There are various human cytokines that can be used in the method of the present invention. Many inflammatory cytokines are known to be released in cytokine release syndrome, including IFN-γ, IL-Ιβ, TNF, IL-2, IL-6, IL-8, IL-10 and IL-12. Some cytokines are believed to have a significant importance in determining cytokine release syndrome over others. In the present invention, it may be particularly suitable to include IFN-γ and / or IL-10 as predictive cytokines for determining CRS response (Teachey DT et al., Cancer Discov., June 2016; Vol. 6(6): pp. 664-79. doi: 10.1158 / 2159-8290. CD-16-0040. Epub, April 13, 2016).

[0139] The inventors have found that IFN-γ and / or IL-10 in humanized mouse blood can be used to reliably predict the severity of CRS that may be expected in subjects. IFN-γ and IL-10 levels in PBMC humanized mice are higher than other cytokine levels after induction with immunomodulatory drugs and compared to humans. It is contemplated that other antibodies can be used to help determine the severity of potential responses.

[0140] The inventors found that donors whose IFN-γ levels were at or above 1,800 pg / ml and / or whose IL-10 levels were at or above 120 pg / ml in the blood of humanized mice 6 hours after administration of an immunomodulatory drug such as anti-CD28 (ANC28.1 / 5D10) to the mice were likely to develop severe CRS (e.g., grades 4-5) if the drug was administered to the human subject. Donors whose IFN-γ levels were between 300 pg / ml and 1,800 pg / ml and / or whose IL-10 levels were between 25 pg / ml and 120 pg / ml in the blood of humanized mice were likely to develop moderate CRS (developing grades 1-3 CRS) (but not severe CRS) after receiving the immunomodulatory drug. Donors whose IFN-γ levels fell below 300 pg / ml and / or IL-10 levels fell below 25 pg / ml in the blood of humanized mice were unlikely to develop CRS after receiving immunomodulatory drugs (Teachey DT et al., 2016; Weissmuller et al., PloS One, DOI:10.1371 / journal.pone.0149093, March 2016).

[0141] In certain embodiments, the invention is directed to the use of 15-30 million harvested / isolated human PBMCs per mouse from a donor / subject and engraftment of the PBMCs in immune-deficient mice to obtain humanized immune-deficient mice, as described below. In certain preferred embodiments, 15-25 million harvested / isolated human PBMCs per mouse are used. In certain preferred embodiments, 20 million harvested / isolated human PBMCs per mouse are used. Five to seven days after engraftment, an immunomodulatory drug is administered to the mice and human cytokine concentrations in the mice are measured to determine whether cytokine release of IFN-γ and IL-10 is a "severe or high response," "moderate or low response," or "non-response" as predictors or determinants of whether a particular donor / subject will exhibit cytokine release syndrome (CRS) following administration of an immunomodulatory drug. A "severe / high response" is measured as an IFN-γ level equal to or greater than 1,800 pg / ml and an IL-10 level equal to or greater than 120 pg / ml 6 hours after administration of an immunomodulatory drug to humanized mice. IFN-γ and IL-10 are measured because they have been identified as the best predictors of whether or not a cytokine release syndrome or cytokine storm will occur. Using this in vivo assay system, the inventors have demonstrated that different human donor PBMCs engrafted in immunodeficient mice respond differently to CRS-inducing drugs.

[0142] In this method, if a subject is a "low / non-responder" in the humanized mouse model method, it is unlikely that the subject will induce a cytokine response after administration of an immunomodulatory drug, and in this case, administration of the immunomodulatory drug to this individual is safe.Therefore, if a particular individual is determined to be a "low / non-responder" using this method, an immunomodulatory drug that was previously unavailable for use in treatment due to its potential to have a highly harmful cytokine storm effect can actually be used for this individual.If this individual is determined to have a severe / high IFN-γ and IL-10 response to an immunomodulatory drug using a mouse humanized by this individual's PBMCs using this method, this individual may induce a cytokine storm response, and therefore administration of the immunomodulatory drug to this individual is not safe.

[0143] In the method, immune-deficient mice (also known as immunodeficient mice) are used. NSG mice are a line of pure-bred laboratory mice that are immune-deficient. NSG mice lack mature T cells, B cells, and natural killer (NK) cells, or a combination of these. NSG mice also lack multiple cytokine signaling pathways and have many deficiencies in innate immunity. The combined immune-deficient state in NSG mice allows the engraftment of a wide range of primary human cells to obtain humanized mice, allowing sophisticated modeling of many areas of human biology and disease. The inventors have found that NSG mice have similar responses to NSG-CSF-1 and NSG-IL-6 mice. In this specification, the terms "humanized mice," "humanized immune-deficient mice," "humanized immune-deficient mice," and plural versions thereof are used interchangeably. Thus, the use of one of these terms should be interpreted as inclusive of all.

[0144] In certain embodiments, the non-human mammal is a genetically modified mouse that lacks an immune system (i.e., a humanized immunodeficient mouse). An example of an immunodeficient mouse is an NSG (i.e., NOD scid gamma (NOD.Cg-Prkdc scid Il2rg tm1WjlThe immunodeficient mice include, but are not limited to, human IgG1, IgG2, IgG3, IgG4, IgG5, IgG6, IgG7, IgG8, IgG9, IgG10, IgG11, IgG12, IgG13, IgG14, IgG15, IgG16, IgG17, IgG18, IgG19, IgG20, IgG21, IgG22, IgG23, IgG24, IgG25, IgG30, IgG41, IgG42, IgG43, IgG44, IgG45, IgG45, IgG46, IgG47, IgG48, IgG49, IgG48, IgG49, IgG49, IgG48, IgG49, IgG49, IgG41, IgG45, IgG47, IgG48, IgG49, IgG49, IgG41, IgG44, IgG45, IgG47, IgG48, IgG49, IgG49, IgG41 ...7, IgG48, IgG49, IgG47, IgG47,

[0145] Although the previous in vivo assays can be used to demonstrate drug-induced cytokine signals, the inventors hypothesize that the previous assays are susceptible to producing false positive and false negative responses because they cannot take into account the significance of the dynamic immunological changes (e.g., sufficient circulating CD3 and NK cells, and the presence of GVHD) along with the number of donor cells in the test animal. The former may be due to too many donor cells (i.e., CD3 and NK cells), which makes the assay too sensitive and fails to obtain accurate predictions of cytokine release. The latter may be due to too few donor cells (i.e., CD3 and NK cells), which also makes the assay too insensitive and fails to obtain accurate predictions of cytokine release. Complicating matters, too many CD3 cells may trigger graft-versus-host defense mechanisms to operate in the test animal, making the mouse model unsuitable for screening cytokine release.

[0146] The present invention relieves the long-standing unmet need to provide an in vivo screening assay that accurately determines severe CRS in human subjects, which is achieved by adjusting the amount of PBMC engrafted and the timing of drug administration after PBMC engraftment.

[0147] As one of skill in the art will appreciate, there are many suitable ways to collect and isolate PBMCs. There are also many suitable ways to introduce PBMCs into immunodeficient mice, including, but not limited to, intravenous and intracardiac.

[0148] The inventors have established that PBMCs can be collected from individual humans (who will likely receive an immunomodulatory drug) and then used for engraftment in immunodeficient mice (e.g., NSG, NSG-IL-6 or NSG-CSF-1) to test for toxicity (i.e., cytokine release) using the assays disclosed herein. Advantageously, the inventors have found that this assay system predicts in vivo responses in human patients providing PBMCs who will receive an immunomodulatory drug.

[0149] Without being bound to theory, it is believed that too many PBMCs may enhance the sensitivity of the cytokine release profile such that it provides too many inaccurate predictions (i.e., too many false positives). 7 When the number of PBMCs administered exceeds 10, it is observed that some mice suffer from significant weight loss, presumably due to GVHD (graft-versus-host disease). When this occurs, the cytokine release response in such mice induced by immunomodulatory drugs cannot accurately determine CRS in human subjects. Alternatively, the number of PBMCs administered may be increased beyond a certain threshold (e.g., <1×10 7 If the number of infected individuals falls below 100 / mouse), the method similarly fails to determine the CRS with optimal sensitivity (i.e., provides many false negatives).

[0150] In an example embodiment of the method, at least 4 hours prior to engrafting the isolated PBMCs into the immune-deficient mice, the mice are irradiated with 100 cGy of X-rays (or 75 cGy to 125 cGy of X-rays). Irradiation can occur after, before, or simultaneously with the steps of collecting and isolating PBMCs from the human. It is believed that by using irradiation, T cells will expand faster in the NSG mice. Without being bound by theory, it is believed that irradiation enhances the engraftment of human cells. Although the exact mechanism remains unclear, irradiation can result in myeloablation, which destroys mouse immune cells and increases human PBMC survival factors, speeding up human T cell expansion. Irradiation also induces cell death (apoptosis) of mouse immune cells in the peripheral blood, spleen, and bone marrow, directing the expanded human immune cells to the bone marrow. Without irradiation, it would take longer to obtain enough human immune cells to perform the method. However, over time, human PBMCs will lose some cell types, e.g., NK cells and monocytes, because these cell types cannot grow in mice and simply survive for a few days before dying. NK and myeloid cells generally turn over more quickly than CD3 T cells, but the short survival of NK and myeloid cells in mice is due to the lack of factors such as IL15 that stimulate their survival. With or without irradiation, NK cells should be lost by day 10 after PBMC engraftment due to the short survival of NK cells (and myeloid cells). Only T cells can expand in NSG mice. The assay requires as many human immune cell types as possible to test for toxicity. Therefore, to have an optimal population of engrafted human immune cells and to have a period before developing severe GVHD, e.g., 1.5×10 7 ~3.0×10 7 Days 5-7 of this method with engraftment of individual PBMCs are critical to achieving optimal sensitivity while minimizing false negatives.

[0151] In certain exemplary embodiments, the present invention provides a method for the delivery of a specific range of human PMBCs (1.5-3.0×10 7In a particular preferred embodiment, the number of human PBMCs engrafted is 2×10 7 Pieces / mouse.

[0152] The inventors have also discovered that if PBMCs from a particular individual are engrafted in irradiated immunosuppressed mice and then an immunomodulatory drug is administered to the mice at a critical period following such engraftment (e.g., 5-7 days later), preferably 6 days later, the cytokine response can be measured using IFN-γ and / or IL-10 to determine whether that particular human is a "severe" or "high responder," meaning that the subject may induce a massive cytokine release response, such as cytokine release syndrome, which may be potentially fatal to humans, and therefore should not be administered the immunomodulatory drug.

[0153] Certain T cell activators, particularly monoclonal antibodies (mAbs) that address the T cell antigen receptor (TCR), such as ORTHOCLONE OKT® 3 ("OKT3"), a murine monoclonal antibody (mAb) that was the first mAb used in the clinic for immunosuppression, can induce systemic release of proinflammatory cytokines (Abramowicz D. et al., Transplantation, April 1989;47(4):606-613). (p. 8). The most dangerous of these are TNF, interferon-gamma (IFN-gamma) and IL-2. In patients receiving mAb therapy, control of such cytokine release syndrome or "cytokine storm" is routinely achieved by high-dose corticosteroid treatment. ORTHOCLONE OKT®3 is the trade name for muromonab-CD3, an immunosuppressant drug given intravenously to reverse acute rejection of transplanted organs, including heart, kidney and liver. OKT3 works by blocking the function of T cells, which play a major role in acute graft rejection. OKT3 reacts with a molecule called CD3 on the membrane of T cells and blocks their function. Binding of OKT3 to T lymphocytes results in their initial activation, followed by cytokine release and blockade of T cell function. It is an immunosuppressant that is a strong inducer of CRS. The anti-CD28 antibody ANC28.1 / 5D10 is a weaker inducer of CRS. Other antibodies and immunomodulators, including drug candidates in development, can be used in the method.

[0154] Examples of immunomodulatory agents include anti-CD28 monoclonal antibodies (mAbs), anti-CD3 mAbs, anti-CD20 mAbs, anti-CD52 mAbs; granulocyte colony-stimulating factor (G-CSF); interferons; imiquimod; thalidomide and its derivatives or analogs, such as lenalidomide (REVLIMID®), pomalidomide (IMNOVID®), and apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; IMLYGIC™ (talimogene laherparepvec); adalimumab (HUMIRA®), catumaxomab (REMOVAB®), ibritumomab tiuxetan (ZEVALIN®), tositumomab-I 131(BEXXAR®), brentuximab vedotin (ADCETRIS®), cetuximab (ERBITUX®), rituximab (MABTHERA® or RITUXAN®), alemtuzumab (CAMPATH-1H®, CAMPATH® or LEMTRADA®), bevacizumab (AVASTIN®), pertuzumab (PERJETA®), trastuzumab (HERCEPTIN®), trastuzumab emtansine (KADCYLA™), denosumab (PROLIA® or XGEVA®), Ofat These include, but are not limited to, tumab (ARZERRA®), panitumumab (VECTIBIX®), pembrolizumab (KEYTRUDA®), nivolumab (OPDIVO®), ipilimumab (YERVOY®), atezolizumab (TECENTRIQ®), avelumab (BAVENCIO®), durvalumab (IMFINZI®), daratumumab (DARZALEX®), ceritinib (ZYKADIA®) and antithymocyte globulin (THYMOGLOBULIN® (rabbit) or ATGAM® (horse)).

[0155] Thus, in certain example embodiments, the immunomodulatory agent is a therapeutic antibody. The antibody may be monoclonal or polyclonal. Monoclonal antibodies (mAbs) can include, but are not limited to, TGN1412 (anti-CD28 mAb) (TAB08), OKT3 (anti-CD3 mAb), RITUXAN® (rituximab) (anti-CD20 mAb), LEMTRADA® (alemtuzumab, also sold as CAMPATH®) (anti-CDS2 mAb), KEYTRUDA® (pembrolizumab), OPDIVO® (nivolumab), YERVOY® (ipilimumab), ZYKADIA® (ceritinib), TECENTRIQ® (atezolizumab), BAVENCIO® (avelumab), IMFINZI® (durvalumab), and others. In other embodiments, the immunomodulatory agent can be a biologic drug, such as a small molecule drug, such as REVLIMID® (lenalidomide); a polyclonal antibody, such as anti-thymocyte globulin; or a protein, such as interferon.

[0156] As will be appreciated by those skilled in the art, immunomodulatory agents can be administered to non-human immunodeficient mammals using a variety of administration routes. Exemplary administration routes include, but are not limited to, intravenous, intrafemoral, intraventricular, intracardiac, intraperitoneal, and others. The preferred administration route is intravenous infusion.

[0157] Without being committed to a particular theory, it is believed that the predominant cells present in the humanized immunodeficient mice 5-7 days after mouse engraftment with human PBMCs can be tested for cytokine release syndrome or cytokine storm toxicity following administration of an immunomodulatory drug. The test depends on factors including the number of PBMC cells after injection and the ratio of human PBMCs. The balance of cell types and cell number (content) of the cells administered is believed to be important in determining cytokine release syndrome. For example, as shown in Figure 2B, NK cells (CD56) were approximately 20-30% of the CD45 cells present, decreasing to 1-5% by day 10 after engraftment. Thus, by day 10, the predominant cell type in this PBMC-humanized mouse is T cells, and the PBMC-humanized mouse does not have many other human cell types and is not a good model for toxicity testing.

[0158] Non-limiting example embodiments of the invention include a method of determining whether an immunomodulatory agent is likely to induce cytokine release syndrome (CRS), SIRS, or cytokine storm (a severe form of CRS) in an individual human following administration of the immunomodulatory agent to the individual. In an example embodiment, the method comprises: (a) preparing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; (b) 1.5–3.0 × 10 isolated from humans 7 engrafting the mouse with peripheral blood mononuclear cells (PBMCs); (c) administering an immunomodulatory agent to the mouse 5 to 7 days after engraftment; (d) determining blood levels of a plurality of cytokines in the mouse, including IFN-γ and IL-10, wherein blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse; (e) determining that the immunomodulatory agent is capable of inducing severe cytokine release syndrome in the human; the presence of severe cytokine release syndrome in the mouse indicates that administration of the immunomodulatory agent is likely to induce severe cytokine release syndrome in the human. Includes.

[0159] In another non-limiting example embodiment, the invention includes a method for determining whether an immunomodulatory agent is likely to induce cytokine release syndrome (CRS), SIRS, or cytokine storm (a severe form of CRS) in an individual human following administration of the immunomodulatory agent to the individual. In an example embodiment, the method comprises: (a) isolating peripheral blood mononuclear cells (PBMCs) from a human; (b) 2 × 10 7 engrafting said isolated PBMCs into irradiated immunodeficient mice; (c) administering an immunomodulatory agent to the mouse 5 to 7 days after engraftment; (d) determining blood levels of a plurality of cytokines in said mouse, including IFN-γ and IL-10; blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse, and the presence of severe cytokine release syndrome in the mouse indicates that administration of the immunomodulatory agent is likely to induce severe cytokine release syndrome in the human. Includes.

[0160] In an example embodiment, the method comprises: (e) determining whether the human is suitable for immunomodulatory therapy, Induced severe cytokine release syndrome in mice indicates humans are unsuitable for immunomodulatory drug therapy, Step It may further include:

[0161] If the concentration of cytokines IFN-γ and IL-10 in response to the immunomodulatory drug in the mouse model is measured as "severe / high", the immunomodulatory drug may induce severe cytokine release syndrome or cytokine storm, and the human is not suitable for therapy. If the human is not suitable for therapy with this particular immunomodulatory drug, another antibody or chemotherapy or other drug can be attempted to treat the individual / patient.

[0162] If the concentrations of both IFN-γ and IL-10 indicate a "low / no response" or "low response," then the individual is suitable for therapy with an immunomodulatory drug. The determination of the lower end of "no response" vs. "low response" can be difficult to distinguish and may depend on the sensitivity of the assay, but since both "no response" and "low response" indicate that the individual is suitable for therapy, it does not appear necessary to set sharp boundaries at these ends of the response spectrum.

[0163] As described herein, humanized engrafted mice were injected with different immunomodulatory drugs, such as various therapeutic antibodies, including anti-CD28 mAb, OKT3 (anti-CD3 mAb), and the effect of the antibodies in the humanized mice was observed. Both anti-CD28 and OKT3 treated mice showed a significant increase in key inflammatory cytokines for a given individual / sample. Administration to non-humanized mice did not induce a cytokine response. These findings indicate that humanized mice engrafted with PBMCs can determine the immunotoxicity of a given drug for a given subject / individual.

[0164] The method includes administering an immunomodulatory agent to a non-human mammal engrafted with human PBMCs and determining whether the immunomodulatory agent causes toxicity in the non-human mammal, where if the agent causes toxicity in the non-human mammal, the agent is deemed to cause immunotoxicity in humans.

[0165] Immunotoxicity refers to the unwanted / unintended effect of a drug on the function of an individual's immune system. See, e.g., Weir, A, Journal of Immunotoxicology, 5:3-10 (2008); Gribble, E J. et al., Expert Opinion Drug Metab Toxicol, 3(2) (2007).

[0166] In some instances, immunotoxicity may result in a cytokine storm in humans. Cytokine storm, cytokine release syndrome, or infusion reaction is an adverse event that is commonly seen after the first exposure to a drug (e.g., a therapeutic mAb). It is characterized by the systemic release of several inflammatory cytokines. Symptoms range from mild to severe and include fatigue, headache, urticaria, pruritus, bronchospasm, dyspnea, sensation of tongue or pharynx swelling, rhinitis, nausea, vomiting, flushing, fever, chills, hypotension, tachycardia, and asthenia. See, for example, Wing, M. et al., Journal of Immunotoxicology, 5:11-15 (2008) and Wang, H. et al. , American Journal of Emergency Medicine, 26:711-715 (2008).

[0167] Thus, in yet another aspect, the present invention is directed to a method of determining whether administration of an agent(s) causes cytokine release syndrome in an individual (e.g., a human) in need thereof, the method comprising administering the agent to a non-human mammal engrafted with a certain number of PBMCs after a certain number of days, determining the level of one or more human cytokines in the non-human immunodeficient mammal induced by the agent, and determining whether the agent causes cytokine release syndrome in the non-human mammal, where if the agent causes cytokine release syndrome in the non-human mammal, the agent will cause cytokine release syndrome in a human.

[0168] The inventors have found that the number of days to test immunomodulatory agents after engraftment of PBMCs in NSG, NSG-IL-6 or NSG-CSF-1 mice is critical, as long as the number of days results in a suitable cell composition. This number of days may vary slightly depending on the number of PBMCs initially engrafted in the mice. That is, if more PBMCs are engrafted (within the present range), the number of days before testing immunomodulatory agents may be slightly shorter (also within the present range) than if fewer PBMCs are engrafted (within the present range). When the engrafted PBMCs begin to kill mouse cells due to immune cell allogeneic recognition of adult human T cells against the mouse, the mouse is believed to develop graft-versus-host disease (GVHD). When this occurs, the mouse begins to suffer significant weight loss and exhibits severe morbidity that makes the mouse model inaccurate in determining CRS. In certain embodiments, the present invention provides for administration of immunomodulatory agents 4-7 days or 5-7 days after PBMC engraftment. In certain preferred embodiments, the present invention provides for administration of the immunomodulatory agent 6 days after PBMC engraftment.

[0169] We compared the IFN-γ and IL-10 response levels as a function of the number of cells injected into each mouse (20 million cells per mouse vs. 50 million cells per mouse) (see Figures 9A-B). The number of PBMCs injected into the mouse is a determining factor. When a higher number of PBMCs (e.g., 50 million PBMCs per mouse) are injected, the mice exhibited higher released cytokines, with IFN-γ and IL-10 all increasing to ≧1,800 pg / ml and ≧120 pg / ml, respectively. However, this donor (213) (used in Figure 9) released less cytokines when previously injected with 20 million PBMCs per mouse. This data indicates that 20 million PBMCs per mouse (or a range of 15-30 million PBMCs per mouse) is within the preferred range for patient screening. In accordance with the present invention, the range of PBMCs used in detecting a cytokine storm for individuals responding to an immunotherapeutic agent is between 15 and 30 million PBMCs. Preferably, the PBMC range is between 20 and 25 million PBMCs. More preferably, the concentration is 20 million PBMCs per mouse.

[0170] The present invention provides improved in vivo methods for determining the potential of immunomodulatory drugs to cause deleterious cytokine release syndrome (CRS) in human patients.

[0171] Advantageously, as further described herein, the inventors have also found that this assay system determines in vivo responses and represents a powerful tool in research and drug safety evaluation. Current clinical trials of new drug candidates in volunteer human subjects often result in drug failure. This failure is due in large part to toxicity that was not exposed in preclinical trials due to the inadequacy of existing in vivo animal models. There is a long-standing unmet need for in vivo animal models that can accurately predict the adverse effects of potential drug candidates. Screening of drug candidates for development into therapeutic candidates requires passing both in vitro and in vivo preclinical trials.

[0172] As will also be appreciated by those skilled in the art, there are a variety of ways to introduce PBMCs into a non-human mammal. Non-limiting examples of such methods include intravenous, intrafemoral, intraventricular, and intracardiac routes of administration. Preferably, the PBMCs are introduced via the intravenous route.

[0173] The inventors have discovered that the time of PBMC engraftment is critical, in particular, it is observed that immunomodulatory agents can be administered within a specific time window following engraftment of the subject's PBMCs into the mouse.

[0174] At day 3 after engraftment, there are insufficient cell numbers. We found that at day 5, there are sufficient human cell types, and the number of such human cells is optimal for testing (human CD45%>10%). However, at day 10, the cell numbers (percentage of total live cells) of many cell types decrease. For example, NK cells decreased from 20-30% at day 5 to 1-5% at day 10 (Figures 2B and 2C). Thus, the method includes administering a drug to the humanized mouse between days 4-7 or between days 5-7 or on day 6 after engraftment.

[0175] Many lymphoid, myeloid and potentially other human immune cell types are expected to participate in the immunotoxicity response, including T cells and NK cells, which play key roles. In this PBMC humanized mouse model, we found that different human cell types are present in the mouse at early time points. Human T cells and NK cells are the predominant cell populations in such mice at day 5 of engraftment (Figure 2 and Figure 4A).

[0176] For optimal toxicity testing, one may require significant levels of human cells in mice. When mice are injected with human PBMCs, only human T cells can expand; other cell types will begin to die off over time. A complex balance between cell number (i.e., number of PBMCs administered to mice) and cell type over time is required.

[0177] Moreover, the inventors found that the experiment needs to be performed before the mice develop graft-versus-host disease GVHD, which is manifested by weight loss. As described further below, after 8 days, the inventors observed significant weight loss in many of the PBMC-engrafted humanized mice, which showed severe graft-versus-host disease GVHD. In general, it is believed that human T cells in mice cannot grow forever; this means that if the number of T cells is high, they will attack the mouse and cause GVHD due to the release of cytokines. If the mouse has GVHD, this will affect the accuracy of the test results because cytokines are released when human T cells attack mouse cells (Ju XP et al., Transplantation., 1997;63(9):1307-1313). The mouse will eventually die due to GVHD.

[0178] Data from an example of GVHD is depicted in Figures 3A-3D. In Figures 3A and 3B, engraftment of PMBC from donors 4629 and 362 in mice was 5×10 7 PBMC engraftment caused significant weight loss as early as day 6. Figure 3C shows that for donor 309, 2 × 107 Figure 3D shows that mice receiving 3 x 10 PMBCs from donor 358 started to lose weight after day 8, whereas mice receiving 3 x 10 PMBCs from donor 358 started to lose weight after day 8. 7 1 shows that mice receiving PMBCs began to lose weight after day 7.

[0179] In addition, these mice with GVHD were morbid after 8 days and therefore were no longer suitable for testing. We observed weight loss, as well as hunched posture, hair loss, reduced mobility and tachypnea. After 20% weight loss, the mice had to be euthanized.

[0180] Therefore, it is critical that the immunomodulatory drug is administered no later than day 8 after PBMC engraftment to ensure that the mice are not suffering from GVHD or are otherwise unill. Furthermore, the drug cannot be administered too early, for example as early as day 3, as there may not be enough cells present. There must be enough circulating cells present for the test to be accurate.

[0181] Thus, taking into consideration the above factors, in certain embodiments, the immunomodulatory agent is administered to the mouse 5-7 days after the mouse is engrafted with PBMCs. In some preferred embodiments, the immunomodulatory agent is administered to the mouse 6 days after the mouse is engrafted with PBMCs. During this period, the mouse should be maintained under suitable conditions, including providing basic mammalian needs (e.g., food, water, light) as known to those of skill in the art.

[0182] In the example presented below, we selected day 6 for the experiment with 10 donors. Methods for measuring increased expression of one or more proinflammatory cytokines (human or mouse) are also known to those skilled in the art. Proinflammatory human cytokines include IL-2, IL-6, IL-8, IL-1β, IL-4, gamma interferon (IFN-γ), tumor necrosis factor alpha (TNF-α or "TNF"), IL-10, or combinations thereof. Increased expression of proinflammatory cytokines can be determined using flow cytometry as described herein.

[0183] In another aspect, a method is disclosed for determining whether a combination of immunomodulatory agents is likely to induce severe cytokine release syndrome (CRS) in a human following administration of the combination of immunomodulatory agents. In one embodiment, the combination comprises a first immunomodulatory agent and a second immunomodulatory agent. In another embodiment, the method includes the steps of providing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; and administering 1.5-3.0×10 IgG antibodies isolated from a human to a human. 7 The method includes engrafting peripheral blood mononuclear cells (PBMCs) into an immunodeficient mouse to produce a humanized mouse; administering a first immunomodulatory agent and a second immunomodulatory agent to the humanized mouse 5-7 days after engraftment of the PBMCs; determining blood levels of a plurality of cytokines, including IFN-γ and IL-10, in the humanized mouse, wherein blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse; and determining that the combination of immunomodulatory agents is capable of inducing severe cytokine release syndrome in humans, wherein the presence of severe cytokine release syndrome in the mouse indicates that administration of the combination of immunomodulatory agents is capable of inducing severe cytokine release syndrome in humans.

[0184] The number of PBMCs engrafted per mouse was 2 × 10 7The immunodeficient mouse may be an NSG mouse, an NSG-IL-6 mouse or an NSG-CSF-1 mouse, preferably an NSG mouse. The immunodeficient mouse may be irradiated with 100 cGy of X-rays prior to engraftment. Administration of the drug may be performed 6 days after engraftment. The cytokines may further comprise IL-2, IL-4, IL-6 or TNF. Cytokine concentrations may be determined for each of IFN-γ, IL-10, IL-6, IL-2, IL-4 and TNF. Blood concentrations of the cytokines are determined 2 to 6 hours, preferably 6 hours, after administration of the combination of immunomodulators.

[0185] In one exemplary combination of immunomodulatory agents, the immunomodulatory agents are selected from the group consisting of anti-CD28 monoclonal antibodies (mAbs), anti-CD3 mAbs, anti-CD20 mAbs, anti-CD52 mAbs; granulocyte colony-stimulating factor (G-CSF); interferons; imiquimod; thalidomide and its derivatives or analogs, such as lenalidomide (REVLIMID®), pomalidomide (IMNOVID®), and apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; IMLYGIC™ (talimogene laherparepvec); adalimumab (HUMIRA®), catumaxomab (REMOVAB®), ibritumomab tiuxetan (ZEVALIN®), tositumomab-I 131(BEXXAR®), brentuximab vedotin (ADCETRIS®), cetuximab (ERBITUX®), rituximab (MABTHERA® or RITUXAN®), alemtuzumab (CAMPATH-1H®, CAMPATH® or LEMTRADA®), bevacizumab (AVASTIN®), pertuzumab (PERJETA®), trastuzumab (HERCEPTIN®), trastuzumab emtansine (KADCYLA™), denosumab (PROLIA® or XGEVA®), Ofat and antithymocyte globulin (THYMOGLOBULIN® (rabbit) or ATGAM® (horse)).

[0186] In another exemplary combination, the first and second immunomodulatory agents are anti-CD28 mAb, anti-CD3 mAb, anti-CD20 mAb, anti-CD52 mAb, granulocyte colony-stimulating factor (G-CSF); interferon; imiquimod; thalidomide, lenalidomide, pomalidomide, apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; talimogene laherparepvec; adalimumab, catumaxomab, ibritumomab tiuxetan, tositumomab-I 131, brentuximab vedotin, cetuximab, rituximab, alemtuzumab, bevacizumab, pertuzumab, trastuzumab, trastuzumab emtansine, denosumab, ofatumumab, panitumumab, pembrolizumab, nivolumab, ipilimumab, atezolizumab, avelumab, durvalumab, daratumumab, ceritinib, elotuzumab and anti-thymocyte globulin. The anti-CD52 mAb can be alemtuzumab, the anti-C20 mAb can be rituximab, the anti-CD3 mAb can be OKT3, and the anti-CD28 mAb can be TGN1412.

[0187] In a preferred embodiment, the first immunomodulator is pembrolizumab or nivolumab, and the second immunomodulator is lenalidomide, pomalidomide, epacadostat, talimogene laherparepvec, ipilimumab, atezolizumab, avelumab, rituximab, alemtuzumab, ceritinib, daratumumab, elotuzumab, or durvalumab. In another preferred embodiment, the first immunomodulator is pembrolizumab, and the second immunomodulator is lenalidomide. In another preferred embodiment, the first immunomodulator is pembrolizumab, and the second immunomodulator is pomalidomide.

[0188] In a preferred embodiment, the first immunomodulator is nivolumab and the second immunomodulator is lenalidomide. In another preferred embodiment, the first immunomodulator is nivolumab and the second immunomodulator is pomalidomide. In another preferred embodiment, the first immunomodulator is nivolumab and the second immunomodulator is elotuzumab. In another preferred embodiment, the first immunomodulator is nivolumab and the second immunomodulator is daratumumab. In a preferred embodiment, the first immunomodulator is nivolumab and the second immunomodulator is ipilimumab.

[0189] In another preferred embodiment, the first immunomodulatory agent is ipilimumab and the second immunomodulatory agent is lenalidomide, pomalidomide, pembrolizumab, atezolizumab, avelumab, rituximab, alemtuzumab, ceritinib, daratumumab, or durvalumab.

[0190] In another preferred embodiment, the first immunomodulator is atezolizumab, avelumab or durvalumab, and the second immunomodulator is lenalidomide, pomalidomide, pembrolizumab, ipilimumab, rituximab, ceritinib, daratumumab or alemtuzumab. In another preferred embodiment, the first immunomodulator is durvalumab and the second immunomodulator is lenalidomide. In another preferred embodiment, the first immunomodulator is durvalumab and the second immunomodulator is rituximab. In another preferred embodiment, the first immunomodulator is durvalumab and the second immunomodulator is pomalidomide. In another preferred embodiment, the first immunomodulator is durvalumab and the second immunomodulator is daratumumab. In yet another preferred embodiment, the first immunomodulatory agent is durvalumab and the second immunomodulatory agent is ibrutinib.

[0191] In one preferred embodiment, the first immunomodulatory agent is elotuzumab and the second immunomodulatory agent is pomalidomide. In another preferred embodiment, the first immunomodulatory agent is atezolizumab and the second immunomodulatory agent is pomalidomide. In another preferred embodiment, the first immunomodulatory agent is atezolizumab and the second immunomodulatory agent is lenalidomide.

[0192] In another aspect, the present invention provides a method for determining a safe dosage of a disclosed immunomodulatory agent that does not induce cytokine release syndrome in humans following administration of the immunomodulatory agent, comprising the steps of: providing an immunomodulatory agent having a first dosage, the first dosage of which is determined to induce mild or severe cytokine release syndrome in a first irradiated humanized immunodeficient mouse following administration of the immunomodulatory agent; providing a second immunodeficient mouse, the second mouse being irradiated with 75-125 cGy of X-rays; and administering 1.5-3.0×10 IgG isolated from a human. 7 and administering an immunomodulatory agent to the second mouse 5-7 days after engrafting the PBMCs, wherein the immunomodulatory agent is administered at a second dosage less than the first dosage. Determining blood concentrations of a plurality of cytokines in the second mouse, including IFN-γ and IL-10. Determining a safe dosage of the immunomodulatory agent for administration in the human, wherein the safe dosage is a dosage that results in blood concentrations of IFN-γ <300 pg / ml and IL-10 <25 pg / ml following administration of the immunomodulatory agent to the second mouse, wherein blood concentrations of IFN-γ <300 pg / ml and IL-10 <25 pg / ml in the second mouse indicate that administration of the safe dosage of the immunomodulatory agent is likely not to induce cytokine release syndrome in the human.

[0193] In another embodiment, the method provides for optimizing the safe dosage of an immunomodulatory drug suspected of inducing mild or severe cytokine release syndrome in humans. The method includes the steps of providing an immunodeficient mouse, the mouse being irradiated with 75-125 cGy of X-rays; and administering 1.5-3.0×10 7and administering an immunomodulatory agent to the mouse 5-7 days after engraftment of the PBMCs, wherein the immunomodulatory agent is administered at a dosage lower than a dosage suspected of inducing mild or severe cytokine release; determining blood levels of a plurality of cytokines in the mouse, including IFN-γ and IL-10; and determining a safe dosage of the immunomodulatory agent for administration in the human, wherein the safe dosage is a dosage that results in blood levels of IFN-γ <300 pg / ml and IL-10 <25 pg / ml following administration of the immunomodulatory agent to the mouse, wherein blood levels of IFN-γ <300 pg / ml and IL-10 <25 pg / ml in the mouse indicate that administration of the safe dosage of the immunomodulatory agent is likely not to induce cytokine release syndrome in the human.

[0194] The number of PBMCs engrafted per mouse was 2 × 10 7 The immunodeficient mice may be PBMCs. Exemplary immunodeficient mice include NSG mice, NSG-IL-6 mice, NSG-CSF-1 mice, and others. Preferably, the immunodeficient mice are NSG mice. In a preferred embodiment, the immunodeficient mice are irradiated with 100 cGy of X-rays prior to engraftment. Administration of the drug may occur 6 days after engraftment. The cytokines may further include IL-2, IL-4, IL-6, or TNF. Blood levels of the cytokines may be determined 2-6 hours after administration of the immunomodulatory agent, preferably 6 hours after administration of the immunomodulatory agent. The immunomodulatory agent may be any of those disclosed herein. For example, the immunomodulatory agent can be an anti-CD28 mAb, an anti-CD3 mAb, an anti-CD20 mAb, an anti-CD52 mAb, pembrolizumab, ipilimumab, atezolizumab, avelumab, durvalumab, epacadostat, talimogene laherparepvec, nivolumab, lenalidomide, ceritinib, or antithymocyte globulin.

[0195] In another aspect, the invention provides an in vitro method for determining the likelihood that administration of an immunomodulatory agent to a human will induce cytokine release syndrome in the human, comprising: 7 The present invention provides a method for detecting a cytokine release syndrome (CDC) in a mouse blood sample, the method comprising: providing a blood sample from a humanized irradiated immunodeficient mouse administered an immunomodulatory agent 5-7 days after engraftment of isolated peripheral blood mononuclear cells (PBMCs); and detecting in vitro the concentration of a plurality of cytokines, including IFN-γ and / or IL-10, present in the mouse blood sample, wherein a concentration of IFN-γ≧1,800 pg / ml or IL-10≧120 pg / ml in the mouse blood sample indicates that administration of the immunomodulatory agent to a human may induce severe cytokine release syndrome.

[0196] In another aspect, the invention provides a method for determining the likelihood that administration of a combination of a first immunomodulatory agent and a second immunomodulatory agent disclosed herein to a human will induce severe cytokine release syndrome in the human. The method comprises determining the likelihood that administration of a combination of a first immunomodulatory agent and a second immunomodulatory agent to a human will induce severe cytokine release syndrome in the human. 7 The method may include providing a blood sample from a humanized irradiated immunodeficient mouse administered the combination of a first immunomodulatory agent and a second immunomodulatory agent 5-7 days after engraftment of isolated peripheral blood mononuclear cells (PBMCs); and detecting in vitro the concentration of a plurality of cytokines including IFN-γ and / or IL-10 present in the mouse blood sample, wherein a concentration of IFN-γ≧1,800 pg / ml or IL-10≧120 pg / ml indicates that administration of the combination of the first immunomodulatory agent and the second immunomodulatory agent to a human is likely to induce severe cytokine release syndrome.

[0197] In a further aspect, the present invention provides a humanized mouse model as a drug screening platform with the accuracy to identify from a large number of clinically relevant drug candidates for clinical evaluation. The assay eliminates potential drug candidates that induce cytokine release in humans, thus representing a robust predictive tool for drug immunotoxicity testing.

[0198] The present assay represents a drug testing assay for a drug candidate(s) that may adversely affect the immune system in humans. The present assay may also provide drug testing for a drug candidate or combination of drug candidates. The model provides a necessary link between preclinical and clinical trials. Integration of the present assay into a drug development program should accelerate the FDA approval process for therapeutic drug development. The drug candidates or combinations of drug candidates in these methods are not limited to the immunomodulatory drugs mentioned with respect to other embodiments herein, but may include any drug candidate that may have a therapeutic effect in treating, alleviating and / or curing a disease, illness, illness, injury or other condition.

[0199] Thus, in another aspect, the invention is directed to a method of determining whether a drug candidate causes immunotoxicity in humans, comprising administering the drug to a non-human immunodeficient mammal (e.g., an NSG, NSG-CSF-1 or NSG-IL-6 mouse) engrafted with human peripheral blood mononuclear cells (hPBMCs) and determining whether the drug causes immunotoxicity in the non-human mammal, thereby determining whether the drug causes immunotoxicity in humans.

[0200] Thus, the present invention provides an in vivo method for determining the immunotoxicity of drug candidates for use in humans. The in vivo method includes the steps of: (a) providing an immunodeficient mouse, the mouse being irradiated with 75 to 125 cGy of X-rays; and (b) irradiating 4.5 to 5.5 × 10 7 ~5.5×10 7 Human PBMCs, preferably 5.0 x 10 7(c) administering a drug candidate to the mouse 4-7 days after engraftment; (d) determining a cytokine concentration in the blood of the mouse, wherein the cytokine is at least one cytokine selected from the group consisting of IFN-γ≧300pg / ml, IL-2≧15pg / ml, IL-4≧10pg / ml, IL-6≧10pg / ml, IL-10≧25pg / ml or TNF≧5pg / ml; and (e) determining the immunotoxicity of the drug candidate, wherein a blood concentration of at least one cytokine selected from the group consisting of IFN-γ≧300pg / ml, IL-2≧15pg / ml, IL-4≧10pg / ml, IL-6≧10pg / ml, IL-10≧25pg / ml or TNF≧5pg / ml in the mouse is indicative of the immunotoxicity of the drug candidate in humans. The immunodeficient mouse may be an NSG mouse, an NSG-IL-6 mouse or an NSG-CSF-1 mouse, preferably an NSG mouse. In a preferred embodiment, the immunodeficient mouse is irradiated with 100 cGy of X-rays. Cytokine release may be determined in the blood of the mouse 2-6 hours after administration of the drug candidate, preferably 6 hours after administration of the drug candidate.

[0201] In one embodiment, the method for determining human immunotoxicity of a drug candidate comprises administering to a patient a 4.5×10 7 ~5.5×10 7 The method includes engrafting irradiated immunodeficient mice with human PBMCs; administering a drug candidate to the mice 4-7 days, preferably 5-7 days, more preferably 6 days after engraftment; determining cytokine release in the blood of the mice, where the cytokine is at least one cytokine selected from the group consisting of IFN-γ, IL-2, IL-4, IL-6, IL-10, and TNF; and identifying the drug candidate as having low human immunotoxicity if low cytokine release is detected in the blood of the mice. In one embodiment, the method for determining human immunotoxicity of a drug candidate comprises administering 4.5-5.5×10 human PBMCs to the mice 4-7 days, preferably 5-7 days, more preferably 6 days after engraftment. 7The method includes the steps of: preparing a blood sample from a humanized irradiated immunodeficient mouse administered with the drug candidate 4-7 days, preferably 5-7 days, more preferably 6 days after engraftment of isolated human peripheral blood mononuclear cells (PBMCs); and detecting in vitro the concentration of at least one human cytokine present in the mouse blood sample to determine the human immunotoxicity of the drug candidate, wherein the at least one human cytokine is selected from the group consisting of IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNFα, and the drug candidate has low human immunotoxicity if a low human cytokine concentration is detected in the mouse blood sample. The immunodeficient mouse may be an NSG mouse, an NSG-IL-6 mouse or an NSG-CSF-1 mouse, preferably the immunodeficient mouse is an NSG mouse. The immunodeficient mouse may be irradiated with 75-125 cGy of X-rays, preferably the immunodeficient mouse is irradiated with 100 cGy of X-rays. Cytokine release can be determined in the blood of the mice 2-6 hours after administration of the drug candidate, preferably 6 hours after administration of the drug candidate. Low cytokine release in the blood of the mice can include IFN-γ<300pg / ml, IL-10<25pg / ml, IL-2<15pg / ml, IL-4<10pg / ml, IL-6<10pg / ml or TNF<5pg / ml. Low cytokine release in the blood of the mice can include amounts of cytokines equal to or less than the amount of cytokine induced by administration of a negative control.

[0202] The present invention further provides a step for identifying whether a drug candidate(s) and / or drug combination has a safety profile suitable for FDA approval, where low cytokine release indicates a safety profile suitable for FDA approval. In a non-limiting example embodiment, 4.5×10 7 ~5.5×10 7 In another non-limiting embodiment, 5.0×10 human PBMCs are engrafted into irradiated immunodeficient mice. 7Human PBMCs are engrafted into the irradiated immunodeficient mice. The PBMCs engrafted in the method of determining the safety profile can be derived from a single individual or from a pool of humans. In a preferred embodiment, the immunodeficient mice are irradiated with 75cGy to 125cGy of X-rays at least 4 hours before the PBMCs are engrafted into the immunodeficient mice. In another preferred embodiment, the immunodeficient mice are irradiated with 1100cGy of X-rays at least 4 hours before the PBMCs are engrafted into the immunodeficient mice. The response to the drug can be evaluated, for example, in comparison to a control agent. The assay allows for the determination that the drug candidate passes the pharmaceutical safety evaluation.

[0203] In another exemplary embodiment, the present invention provides a humanized irradiated immunodeficient mouse engrafted with human peripheral blood mononuclear cells, comprising NOD.Cg-Prkdc scid Il2rg tm1Wjl The present invention provides a humanized irradiated immunodeficient mouse, which is a 1.5-3.0×10 7 PBMCs are engrafted. The NSG mice can further comprise the human macrophage colony stimulating factor-1 gene (NSG-CSF-1) or the human interleukin-6 gene (NSG-IL-6). More preferably, the mice are engrafted with 2×10 7 In another preferred embodiment, mice are engrafted with 4.5-5.5×10 PBMCs. 7 Engraft 10 PBMCs.

[0204] In another exemplary embodiment, the present invention provides a humanized immunodeficient mouse engrafted with human peripheral blood mononuclear cells, the NOD.Cg-Prkdc mouse carrying the human macrophage colony-stimulating factor-1 gene (NSG-CSG-1). scid Il2rg tm1Wjl The humanized immunodeficient mouse is preferably a 1.5 to 3.0 × 10 7 More preferably, mice are engrafted with 2×10 PBMCs. 7In another preferred embodiment, mice are engrafted with 4.5-5.5×10 PBMCs. 7 Engraft 10 PBMCs.

[0205] Additional aspects, advantages and / or other features of example embodiments of the present invention will become apparent from consideration of the following detailed description in conjunction with the accompanying drawings. It should be apparent to those skilled in the art that the described embodiments provided herein are merely exemplary and explanatory, and not limiting. Numerous embodiments of modifications thereof are contemplated as falling within the scope of the present disclosure and its equivalents.

[0206] The following examples are provided to further illustrate various non-limiting embodiments and techniques of the method, including the experiments carried out in the development of the method. However, it should be understood that these examples are meant to be illustrative and do not limit the scope of the claims. As will be apparent to those skilled in the art, many variations and modifications are intended to be encompassed within the spirit and scope of the present invention. EXAMPLES

[0207] Example 1 Engraftment of human PBMCs in humanized mice In this study, two lines of humanized immunodeficient mice were used: (i) 6-week-old female NOD.Cg-Prkdc scid Il2rg tm1Wjl / SzJ (NSG, JAX line number 005557) mice and (ii) NSG-CSF-1 mice (Jackson Laboratory line number: 028654). Mice were irradiated with 100 cGy of X-rays at least 4 hours prior to human PBMC engraftment. Purified / isolated human PBMCs (Astarte Biologics or Allcells) from the same donor were added at 1–5 × 10 7 NSG or NSG-CSF-1 mice were injected intravenously (iv) with 100 cells / mouse. After hPBMC injection, mice were observed daily for body weight, overall appearance of fur, and mobility.

[0208] Figure 1A shows the mean ± standard error of the mean (SEM) of 2 × 10 7 Figure 1B shows daily weight changes in 10 NSG or NSG-CSF-1 mice after hPBMC engraftment at 10 hPBMCs / mouse. 7 Figure 1 depicts weight measurements of five individual NSG mice (hPBMCs / mouse). Each line represents one mouse. It was observed that the majority of the mice exhibited significant weight loss, beginning at day 8. Furthermore, in addition to losing weight after engraftment, mice began to exhibit graft-versus-host disease (GVHD), as described further below. Example 2 Dynamics in human engrafted cell type and cell number at different time points post-engraftment

[0209] In this experiment, cell types and cell numbers of engrafted human mononuclear cells from different human donors were examined after hPBMC engraftment of three different mouse lineages. Mice were bled on days 5 and / or 10 after engraftment and human cells (type and percentage) were analyzed by flow cytometry. Specifically, mouse PBMCs (in mixture with hPBMCs) were stained with human antibodies: anti-CD45, anti-CD3, anti-CD14, anti-CD19 and anti-CD56.

[0210] To test human cell reconstitution (i.e., human cells showed different percentages of total live cells at different time points after engraftment), NSG, NSG-IL-6 (Jackson Laboratory line number 028655) and NSG-CSF-1 mice were used. Figure 2A shows that NSG, NSG-IL-6 and NSG-CSF-1 mouse reconstitution showed comparable human cells at day 5 after engraftment of hPBMC from donor 331. The results show that there was no significant difference between these three mouse lines. All three mouse lines have about 20% human CD45 cell reconstitution at day 5 after engraftment. Of the human CD45 cells, the majority were CD3 T cells and NK cells (Figure 2A).

[0211] In another study, in NSG, a different donor (Figure 2B, 2 × 10 7 PBMCs / mouse) or donor 358 (Figure 2C, 3 × 10 7 Different cell populations were examined at two different time points after engraftment of hPBMCs from humanized mice (1 PBMC / mouse). In these experiments, 2-5 mice per group per time point were used and data are presented as mean ± SEM. As shown in Figure 2B and Figure 2C, there were approximately 10-15% human CD45+ at day 5 and 35-50% CD45+ at day 10. These data indicate that engrafted human CD45 cells increased over time in humanized mice. Interestingly, it was observed that at day 5, 20-30% CD56 (NK) cells were present in the population of CD45+ cells in humanized mice. However, at day 10, CD56 (NK) cells were significantly reduced to 1-5%. These data suggest that the increase in percent CD45 correlates with the loss of certain mononuclear cell types (e.g., CD56 NK cells). In summary, there is a dynamic change in cell type and cell number of engrafted mononuclear cells in humanized immune-deficient mice after engraftment of hPMBCs. T cells and NK cells are the predominant cell type on day 5, but on day 10, only T cells are the predominant cell type, and NK cells die off over time. For the method to work, a balance needs to exist so that not all NK cells die off during drug testing (e.g., on day 10).

[0212] The hPBMC humanized mouse model is considered a T cell model because human T cells become the predominant cell population over time after human cell reconstitution. The results of this example confirmed this. Figures 2B and 2C show that among human CD45 cells, which increased from 10-15% on day 5 to 35-50% on day 10 after engraftment, the CD3 T cell percentage increased from 65-80% on day 5 to 90-95% on day 10.

[0213] There are many cell types involved in immunotoxicity response. T cells, NK cells and monocytes all play very important roles. In the hPBMC humanized mouse model, we found that different human cell types still exist in mice at early time points after engraftment. Human T cells and NK cells are predominant in the cell population in hPBMC humanized mice at day 5 after engraftment, Figure 2B and Figure 2C. This finding prompted us to test the immunotoxicity response in hPBMC humanized mice. We chose an early time point after engraftment, e.g., day 6, to test immunotoxicity and CRS.

[0214] Mice were irradiated on day 0 and engrafted with hPBMCs on the same day. The weight of the mice dropped from day 1 to day 3 due to irradiation. After day 4, the mice started to gain weight. At first glance, the NSG mice have more weight compared to NSG-CSF-1, as shown in FIG. 1A. The NSG mice also appeared healthier and more active. Therefore, the NSG mice were used for subsequent studies. Example 3 The importance of engraftment time

[0215] In this study, we investigated the basis of weight loss in some humanized mice after hPBMC engraftment. Significant weight loss was observed in many mice, indicating graft-versus-host disease (GVHD) in these mice. For example, GVHD data is depicted in Figure 3A, Figure 3B, Figure 3C, and Figure 3D.

[0216] Figures 3A and 3B show 5×10 IgG in NSG mice for donor 4692 and donor 362, respectively. 7 Figure 1 depicts weight measurements of five humanized NSG mice as a function of time after engraftment of PBMCs / mouse. This level of engrafted cells caused significant weight loss (approximately 10%) after 7 days after engraftment of PBMCs from either donor. Approximately 20% weight loss represents a severe state of GVHD in the mice and requires euthanasia of the mice.

[0217] Figure 3C shows donor 309 engraftment (2 × 10 7 FIG. 1 depicts body weight measurements of five humanized NSG mice as a function of time after PBMC engraftment (10 PBMCs / mouse). After day 8 post-PBMC engraftment, a significant weight loss (nearly 10%) is observed.

[0218] FIG 3D shows donor 358 engraftment (3 × 10 7 Body weight measurements of four humanized NSG mice as a function of time after PBMCs (10 PBMCs / mouse) are shown. Mice receiving PBMCs from donor 358 began to lose significant weight (nearly 10%) on day 9 post-engraftment.

[0219] It is believed that GVHD explained the weight loss observed. As the engrafted human T cells in the mice begin to grow, they attack the mouse cells. As described in Example 2 above, an increase in the percentage of CD45 cells was seen over time after engraftment, and at high T cell numbers, mice with GVHD became ill after 8 days and were no longer suitable for testing.

[0220] GVHD mice not only showed weight loss but also showed signs of hunched posture, hair loss, decreased mobility and tachypnea. After 20% weight loss, mice had to be euthanized and the experiment was terminated. The high incidence of GVHD after 8 days of engraftment highlights another aspect of the criticality of the timing of human PBMC engraftment prior to administration of immunomodulatory drugs for toxicity testing.

[0221] Immunomodulatory drugs cannot be administered too early after engraftment (e.g., on days 2-3) because there may not be sufficient circulating cell numbers and types in the mice to produce reliable and reproducible results.

[0222] However, if the mice had GVHD, this would affect the study results, therefore, day 6 after cell engraftment was selected for testing the effect of drug administration in additional donors in subsequent experiments.

[0223] At different time points after drug administration, different cytokines may be released. The earliest cytokine released should be TNF, which always peaks at or before 1 hour after drug administration. Furthermore, in the absence of organ failure, most cytokines will return to normal levels after 24 hours. Therefore, 2 and 6 hours after drug administration were selected as the time points for bleeding mice and serum testing for cytokine concentrations. Example 4 Human T and NK cells represent the predominant cell populations in hPBMC-humanized NSG mice at day 5 of engraftment

[0224] In this study, mononuclear cell reconstitution from ten (10) donors was examined and compared in humanized mice. Each mouse received 2 × 10 7 hPBMCs were engrafted. Anonymized patient information for 7 of the 10 donors is shown in Figure 4B.

[0225] Five humanized NSG mice per donor were tested by flow cytometry for the indicated immune cell subset reconstitution. In hPBMC-injected NSG mice, on day 5 of reconstitution, whole blood was analyzed by flow cytometry. Human CD45, CD3, CD19, CD14 and CD56 were measured. Human CD45+ cells as a percentage of total cells, as well as CD3, CD19, CD14 and CD56 as a percentage of CD45+ cells (gated on CD45+ cells), are shown in Figure 4A for 10 donors. Only donors A4692, A4625 and A4668 showed CD45, CD3, CD19 and CD14. Humanized mice showed an average of 10-25% human CD45+ cells in peripheral blood (Figure 4A). Among human CD45 cells, there were 30-80% T cells and 10-40% CD56 (NK) cells, showing diversity among different donors.

[0226] In subsequent experiments, a day within the range of days 5-7 after engraftment was selected as the optimal time to administer immunomodulatory drugs to mice for toxicity testing. Example 5 Cytokine release induced by immunomodulatory drugs in humanized mice

[0227] For preclinical and clinical trials, a positive control is required for all patients to establish a humanized mouse model for screening and determining drug immunotoxicity, cytokine release syndrome (CRS). ORTHOCLONE OKT3, also called muromonab-CD3, is a mouse monoclonal antibody (mAb) (anti-CD3 mAb) that was used as an immunosuppressant drug to immunosuppress graft recipients. OKT3 binds to the CD3 receptor, which can activate T cells to release cytokines and cause cytokine release syndrome (CRS). OKT3 was used as a positive control for all patients. To test the specificity and sensitivity of the method, a target drug with few methods to test its immunotoxicity is required to be selected. Anti-CD28 mAb was selected as the target drug for evaluation of the specificity and sensitivity of the method.

[0228] PBMCs from nine (10) different donors were used to generate humanized NSG mice for these experiments. hPBMC engraftment (2 × 10 7 Mice were induced for human cytokine release by iv injection of antibodies OKT3 (anti-CD3 mAb; BioLegend, Cat. No. 317302) or ANC28.1 / 5D10 (also referred to as "ANC28", "anti-CD28 mAb" or "anti-CD28"; Ancell, Cat. No. 177-824) on day 6. PBS buffer served as a negative control. Mice were bled at 2 and 6 hours and serum was collected and analyzed for cytokine concentrations using the BD Cytometric Bead Array (CBA) Human Th1 / Th2 Cytokine Kit II (BD, Cat. No. 551809) (see Figures 5A-F).

[0229] Figure 5A-5F depict multiple graphs of concentrations of different cytokines (i.e., IFN-γ, IL-6, IL-2, IL-10, IL-4 and TNF, respectively) measured at 2 and 6 hours after injection of antibodies OKT3 (anti-CD3 mAb) and ANC28 (anti-CD28) into a set of humanized mice for nine (9) different donors. Mice were injected iv with 0.5 mg / kg OKT3 or 1 mg / kg anti-CD28 and 5 ml / kg PBS (negative control). Mice were bled at 2 and 6 hours and circulating cytokine concentrations were measured by BD CBA Th1 / Th2 II kit. The number of mice per group was 2-5 and data are presented as mean ± SEM. Example 6 Enhanced circulating cytokine concentrations following drug administration

[0230] To determine whether a cytokine storm was induced, cytokines (human IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNF) were assayed in the serum of mice 2 and 6 hours after antibody injection (see Figures 5A-F). A significant induction of human IFN-γ, IL-6, IL-10, IL-2, IL-4 and TNF was found after injection of OKT3 in all 10 donors at both 2 and 6 hours.

[0231] However, anti-CD28 mAb-treated mice induced significant release of cytokines in only a portion of donors, and not all donors had a cytokine release response after anti-CD28 injection.

[0232] As shown in FIG. 5A, the 10 donors exhibited different release profiles for IFN-γ cytokines 6 hours after anti-CD28 injection: Severe / high response: donors A4692, A4668 and 362 (IFN-γ ≥ 1,800 pg / ml); Moderate / mild response: Donors A4625, 366, 345, 309, and 213 (IFN-γ ≥ 300 pg / ml to < 1,800 pg / ml) and Low / non-responder: donors 364 and 353 (IFN-γ<300 pg / ml).

[0233] Figure 5B shows the 2 x 10 7 Figure 1 shows anti-CD28 mAb IL-10 cytokine responses observed at 2 or 6 hours in NSG mice engrafted with 10 PBMCs / mouse. To make it easier to determine if a response is a severe / high response, a line has been drawn at the IL-10 level of 120 pg / ml, which is the cutoff between a severe / high response and a low response. The responses at 6 hours were as follows: Severe / high response: donors A4692, A4668 and 362 (IL-10 ≥ 120 pg / ml); Moderate / mild response: Donors A4625, 366, 345, 213 and 309 (IL-10 ≥ 25 pg / ml to < 120 pg / ml) and Low / non-response: donors 364 and 353 (<25 pg / ml).

[0234] Applicants observed that donors whose IFN-γ levels were ≧1,800 pg / ml (either with OKT3 or anti-CD28 mAb) also exhibited increased IL-10 levels (i.e., ≧120 pg / ml). If both IFN-γ and IL-10 levels are increased above ≧1,800 pg / ml and ≧120 pg / ml, respectively, the donor is highly likely to develop CRS if injected with the drug.

[0235] 5C-5F depict cytokine release of IL-6, IL-2, IL-4 and TNF at 2 or 6 hours, respectively. Figure 5C shows the 2 x 10 7 Figure 1 shows anti-CD28 mAb IL-6 cytokine responses observed at 2 or 6 hours in NSG mice engrafted with (PBMCs / mouse). Responses at 6 hours were as follows: Severe / high response: donors A4692, A4668 and 362 (IL-6 ≥ 25 pg / ml); Moderate / mild response: Donors A4625, 366, 345, 213 and 309 (IL-6 ≥ 10 pg / ml to < 25 pg / ml) and Low / non-responder: donors 364 and 353 (IL-6 <10 pg / ml).

[0236] Figure 5D shows the 2 x 10 7 Figure 1 shows anti-CD28 mAb IL-2 cytokine responses observed at 2 or 6 hours in NSG mice engrafted with (PBMCs / mouse). Responses at 6 hours were as follows: Severe / high response: donors A4692, A4668 and 362 (IL-2 ≥ 80 pg / ml); Moderate / mild response: donors A4625, 366, 345, 213 and 309 (15 pg / ml IL-2 < 80 pg / ml) and Low / non-responder: donors 364 and 353 (IL-2 <15 pg / ml).

[0237] Figure 5E shows the 2 x 10 7 Figure 1 shows anti-CD28 mAb IL-4 cytokine responses observed at 2 or 6 hours in NSG mice engrafted with (PBMCs / mouse). Responses at 6 hours were as follows: Severe / high response: donors A4692, A4668 and 362 (IL-4 ≥ 25 pg / ml); Moderate / mild response: donors A4625, 366, 345, 213 and 309 (10 pg / ml IL-4 < 25 pg / ml) and Low / non-responder: donors 364 and 353 (IL-4 < 10 pg / ml).

[0238] Figure 5F shows the 2 x 10 7 Figure 1 shows anti-CD28 mAb TNF cytokine responses observed at 2 or 6 hours in NSG mice engrafted with (PBMCs / mouse). Responses at 6 hours were as follows: Severe / high response: donors A4692, A4668 and 362 (TNF ≥ 20 pg / ml); Moderate / mild response: donors A4625, 366, 345, 213 and 309 (5 pg / ml ≤ TNF < 20 pg / ml) and Low / non-responder: donors 364 and 353 (TNF<5pg / ml).

[0239] Figure 6 summarizes the responses of 10 donors following administration of OKT3 or anti-CD28 in terms of their ability to induce IFN-γ and IL-10. Note that following anti-CD28 mAb administration, donors A4692, A4668 and 362 are severe / high responders; donors A4625, 366, 345, 309 and 213 are moderate / low responders; and donors 364 and 353 are low / non-responders. These data clearly indicate that the NSG / NSG-CSF-1 / NSG-IL-6 mouse by PBMC engraftment model is useful for distinguishing (or as a predictor of) whether humans are likely to induce severe cytokine release syndrome responses following administration of immunomodulatory drugs to humans.

[0240] As shown in Figures 5A-5F and 6, mice engrafted with PBMCs from all 10 donors / patients showed significant cytokine release after OKT3 injection. However, only a portion of mice engrafted with donor PBMCs showed significant induction of these cytokines by administration of anti-CD28 mAb. Not all donors had high levels of cytokine release in response to anti-CD28 injection. These results are similar to the diversity in humans. While all donors had responses to strong inducers of CRS such as OKT3, no other method can detect the enormous diversity among donors as observed in the present method for weak inducers such as anti-CD28. IFN-γ responses can be used as an example: mice engrafted with PBMCs from 3 donors / patients showed high responses to anti-CD28, mice engrafted with PBMCs from 5 donors / patients showed intermediate responses to anti-CD28, and mice engrafted with PBMCs from 2 donors / patients showed low / no responses to anti-CD28. Mice also exhibit hypothermia and increased clinical scores when cytokine release is present. This method can be used to detect new drug immunotoxicity and to screen drug toxicity for individual patients. Example 7 Body temperature changes in NSG mice following immunomodulatory drug treatment

[0241] Rectal temperatures of mice from Examples 5 and 6 were measured before treatment and immediately before bleeding at each time point. Temperature data are shown in Figure 7. Humanized NSG mice showed slightly decreased body temperature in some animals after drug treatment. It was observed that body temperature decreased more frequently in mice with high IFN-γ release (Figure 7). For mice in the OKT3 and anti-CD28 groups, body temperature decreased from 37-38°C to below 36°C at 6 hours. Figure 7 shows hypothermia induction after injection of drugs. Rectal temperatures were measured in 10 donor hPMBC humanized mice injected with control PBS, OKT3 and anti-CD28. The number of mice per group was 2-5 and data are presented as mean ± SEM. Example 8 Evaluation of the clinical scores of mice after injection of drugs

[0242] Clinical scores in mice were monitored by grading the signs and scores as follows: Score: 0=normal activity; 1=normal activity, piloerection, walking on tiptoes; 2=hunched, less active but still mobile; 3=less mobile but mobile when prompted; 4=moribund (critically ill). Mice with a clinical score of 4 were euthanized. The number of mice per group ranged from 2 to 5, and data are presented as mean ± SEM.

[0243] The majority of mice with cytokine release in the OKT3 or anti-CD28 mAb treated groups were scored as 1 at 6 hours. Mice with no or low cytokine release had no clinical score. The clinical scores of mice after injection of the drugs are depicted in FIG. 8. The majority of mice had a clinical score of 1. Example 9 PBMCs (5 × 10 7 Engraftment of 1000 cells / mouse provides a humanized mouse model for drug candidate toxicity screening

[0244] The applicants have developed a humanized mouse model that is useful for toxicity testing to screen potential drugs in discovery. In the early phase of drug candidate development, it is required to screen whether the potential drug candidate may retain toxic activity. In this mouse model, the cytokine response sensitivity has been intentionally enhanced so that any toxicity (e.g., cytokine release syndrome) associated with the potential drug candidate can be screened.

[0245] To do this, a large number of PBMCs (i.e., 5 × 10 7 2 × 10 cells / mouse) were engrafted into humanized NSG mice. On day 6, mice received drug and the cytokine release profile was determined. 7 PBMCs / mouse were used to assess IFN-γ and Il-10 levels.

[0246] Figures 9A and 9B show the effect of 2 × 10 on IFN-γ or IL-10 levels after drug injection. 7 vs.5×10 7 1 depicts donor 213 comparing PBMCs / mouse. IFN-γ and IL-10 levels were observed to be increased with higher PBMC engraftment after administration of OKT3 and anti-CD28 mAb on day 6 using the same amount of mAb (i.e., OKT3 mAb=0.5 mg / kg and anti-CD28 mAb=1 mg / kg). 5×10 7 This enhanced cytokine release by 2 × 10 PMBCs / mouse is considered suitable for screening assays in preclinical drug development. 7 PBMCs / mouse and 5 x 10 7 Figure 1 depicts a comparison of cytokine responses (IL-6, IL-2, IL-4, and TNF, respectively) in donor 213 with 1 PBMC / mouse. These data indicate that large numbers of PBMC provide a reliable and sensitive testing method for drug toxicity screening. Example 10 Comparison of cytokine release with varying PBMC engraftment concentrations

[0247] In this experiment, we used humanized mice engrafted with PBMCs at one of three concentrations to compare cytokine release and determine the effect of cell concentration. 7 PBMCs / mouse, 3 x 10 7 PBMCs / mouse or 4 x 10 7 Cytokine levels produced in humanized NSG mice engrafted with 10 PBMCs / mouse were compared following treatment of such mice with immunotherapeutics (i.e., mAb OKT3, anti-CD28 or KEYTRUDA® (pembrolizumab)).

[0248] Six days after engraftment, mice received immunotherapy (mAb) to determine the cytokine release profile in these three groups of mice. Mice were bled at 2 and 6 hours, and circulating cytokine concentrations were measured by BD CBA Th1 / Th2 II kit. Figures 10A-10F show the cytokine release profiles of 2×10 7 , 3×10 7 or 4×10 7 FIG. 10 depicts cytokine levels after drug injection in donor 309 humanized NSG mice engrafted with 10 PBMCs / mouse. FIG. 10A depicts INFγ levels for each group of mice. FIG. 10B depicts IL-10 levels for each group of mice. FIG. 10C depicts IL-6 levels for each group of mice. FIG. 10D depicts IL-2 levels for each group of mice. FIG. 10E depicts IL-4 levels for each group of mice. FIG. 10F depicts TNF levels for each group of mice.

[0249] Applicants found that under these experimental conditions, cytokine levels were higher at lower engraftment levels, 2×10 per mouse. 7 or 3×10 7 We observed that PBMCs in the two mouse groups showed similar increases after administration of OKT3 and anti-CD28 mAb and KEYTRUDA® (pembrolizumab) on day 6 using the same amounts of mAb (i.e., OKT3 mAb = 0.5 mg / kg, anti-CD28 mAb = 1 mg / kg and KEYTRUDA® = 10 mg / kg). However, the 4 × 10 per mouse 7 With engraftment of 2 × 10 PBMCs per mouse, under these experimental conditions, the cytokine response to each of the drug injections proved to be too high to distinguish individual responses with optimal sensitivity. In contrast, 7 and 3 × 10 7 PBMC engraftment of these cells provides a sensitive test to screen for the occurrence of a cytokine storm in response to a drug in individual humans under these conditions. Example 11 Body temperature and clinical scores at different concentrations of PBMC engraftment

[0250] In this study, applicants measured different cell types, body temperature and clinical scores in humanized mice after engraftment with different PBMC concentrations. The PBMCs used in this study were obtained from donor 309. Figure 11A depicts the cell population 5 days after PBMC engraftment. The total human CD45+ cell percentage increased with increasing levels of PBMC engraftment in mice, while the percentages of different cell types within the CD45 population were similar. Figure 11B depicts the clinical scores in humanized mice after treatment with OKT3 mAb (0.5 mg / kg), anti-CD28 (1 mg / kg) and KEYTRUDA® (pembrolizumab; 10 mg / kg). Figure 11C, Figure 11D and Figure 11E depict the mouse body temperature changes. As shown in Figure 11C and Figure 11D, the 2×10 7 PBMCs / mouse compared to 3 × 10 7 A slight decrease in body temperature was observed in the PBMC / mouse group. Example 12 Cytokine levels increase with increasing drug dosage

[0251] Drug tolerance often varies between individuals. Humanized immunodeficient mouse models can be used to test the drug concentration dependence of cytokine release in individual patients.

[0252] In this study, we examined different concentrations of KEYTRUDA® (pembrolizumab) to determine if there is a dose-dependent effect on CRS in humanized mice. scid Il2rg tm1Wjl / SzJ (NSG, JAX line number 005557) mice were irradiated with 100 cGy of X-rays at least 4 hours prior to human PBMC engraftment. PBMCs were derived from donor 358 in these experiments. Purified / isolated human PBMCs were added at 3×10 7Mice were injected intravenously at 1000 cells / mouse. Six days after PBMC engraftment, mice were injected iv with PBS (negative control), 0.5 mg / kg OKT3 (positive control), 2.5 mg / kg, 5 mg / kg, or 10 mg / kg KEYTRUDA®. There were 5 mice per group. These KEYTRUDA® dosages are within the range of dosages tested in clinical trials described in the US prescribing information for KEYTRUDA®. Few reports of severe cytokine release syndrome adverse effects following KEYTRUDA® administration have been described. Mice were bled at 2 and 6 hours and circulating cytokine concentrations were measured using BD CBA Th1 / Th2 II kits. As shown in Figures 12A-F, levels measured following injection of the positive control OKT3 were very high for IFN-γ, IL-10, IL-6, IL-2, IL-4, and TNF, respectively. As shown in Figures 12A-12F, cytokine levels increased with increasing dosage of KEYTRUDA®. At 2.5 mg / kg KEYTRUDA®, cytokine levels were similar to the negative control group (i.e., the PBS control group) with little cytokine release. In contrast, cytokine levels increased with increasing KEYTRUDA® dosage, with a high cytokine response when 10 mg / kg KEYTRUDA® was used.

[0253] As can be seen from Figures 12A-12F, cytokine levels 6 hours after dosing showed a dose-dependent response in the amount of KEYTRUDA® (pembrolizumab) and increased with increasing dose. At the 2.5 mg / kg KEYTRUDA® concentration, the levels of each cytokine were within the error range of the levels measured for the PBS control group, indicating little cytokine release. At 10 mg / kg KEYTRUDA®, the levels of each cytokine were within the moderate range for predicting CRS. Thus, for donor 358, administration of 2.5 mg / kg KEYTRUDA® (pembrolizumab) would have a lower risk of developing CRS than administration of 10 mg / kg KEYTRUDA® (pembrolizumab).

[0254] The dose-dependence of cytokine release indicates that the in vivo humanized mouse model can be used to screen individual patients for the best drug concentration with respect to avoiding immunotoxicity. Such information can then be used in conjunction with other knowledge regarding the effective dosing range for treating the disorder afflicting the patient to determine an effective yet safe dose of the drug. (Example 13) Temperature and Clinical Scores for Different Doses of KEYTRUDA® (Pembrolizumab)

[0255] In this study, we measured various cell types, body temperature and clinical scores in humanized mice described in Example 12. Figure 13A depicts the cell population 5 days after cell engraftment. We observed that T cells and NK cells represent the predominant cell types in these mice. Figure 13B depicts the clinical scores after administration of immunotherapy (KEYTRUDA® (pembrolizumab)) with different doses of KEYTRUDA®. The clinical scores 6 hours after dosing were the same as the positive control (OKT administration) when dosed with 5 mg / kg and 10 mg / kg KEYTRUDA®. Figure 13C depicts the mouse body temperature changes with different doses of KEYTRUDA®. Example 14 Comparison – in vivo humanized mouse method vs. in vitro assay

[0256] In vitro whole blood or PBMC assays are currently the primary assays for testing cytokine release for drug screening. In this example, cytokine release in response to drug treatment was determined by two methods on the same PBMC donor: an in vitro PBMC assay and an in vivo humanized mouse method. in vivo method

[0257] Day 0: 6-week-old female NOD.Cg-Prkdc scid Il2rg tm1Wjl / SzJ (NSG, JAX lineage no. 005557) mice were irradiated with 100 cGy of X-rays at least 4 hours prior to human PBMC engraftment. Purified / isolated human PBMCs were engrafted at 3 × 10 7 Mice were intravenously injected with 1000 cells / mouse. PBMCs were derived from donors 213, 309, 345 or 366. On day 5, mice were bled for human cell engraftment studies. On day 6, mice were administered drugs; cytokine levels in mouse serum were determined 6 hours after drug administration using the BD CBA Th1 / Th2 II kit. Five mice per group were used in these experiments. in vitro method

[0258] Day -1: Coat plates with drug: Dilute drug (eg, antibody) in PBS; keep plates open in fume hood.

[0259] Day 0: Process PBMCs from each donor. PBMCs were from donors 213, 309, 345 or 366. Thaw PBMCs in supplemented RPMI and wash once. Count cells and culture at 1×10 6Resuspend in supplemented RPMI to 10 cells / ml. Wash coated wells twice with 200 μl PBS, then once with 200 μL supplemented RPMI. Plate 100 μl of cells per well (total of 1×10 cells per well). 5 for cells).

[0260] Day 2: Collect supernatant from each well and measure cytokine levels using the BD CBA Th1 / Th2 II kit. result

[0261] Two methods for determining cytokine release levels after drug treatment were compared for treatment of PBMCs from four different donors with anti-CD28 antibody (Ancell, Cat. No. 177-824). For in vitro assays, anti-CD28 was dosed at 10 μg / well in 96-well plates. For in vivo methods, anti-CD28 was dosed at 1 mg / kg. Supernatants of in vitro PBMC culture wells and serum from mice were harvested and cytokine levels were measured. Antibody OKT3 (0.5 mg / kg for in vivo assays or 1 mg / ml for in vitro assays) was used as a positive control, and either PBS or isotype antibodies were used as negative controls in the in vivo or in vitro assays, respectively. Determined cytokine levels are presented in Figure 14A (IFN-γ and IL-10) and Figure 14B (IL-6 and IL-4) as mean ± standard error of the mean (SEM). Figures 14A and 14B show that for each of the four PBMC donors, for any given cytokine, different levels were determined by the two assays.

[0262] Figure 15 re-plots the cytokine level data of Figures 14A and 14B for donor 213 to allow easier comparison of the differences in cytokine levels determined in the two studies. For each of the four cytokines, IFN-γ, IL-10, IL-6 and IL-4, the in vitro studies showed that relatively little cytokine release occurred in donor 213 cells after dosing with anti-CD28. In the in vitro assays, the cytokine levels measured after dosing with anti-CD28 were not significantly different from the levels determined after dosing with a control isotype antibody to anti-CD28 ("Vehicle 2"), while the cytokine levels after dosing with the control OKT3 antibody were much higher than the levels determined after dosing with a control isotype antibody to OKT3 ("Vehicle 1"). In contrast, in vivo testing methods demonstrated that for each of the four cytokines, IFN-γ, IL-10, IL-6 and IL-4, the cytokine levels produced in donor 213 cells after dosing with either anti-CD28 or OKT3 were much higher than those measured in the negative controls.

[0263] These results indicate that for some humans, in vivo testing indicates that the individual may react with a cytokine storm after treatment with an immunostimulatory drug, but in vitro testing may not be able to indicate that the individual will react with a cytokine storm after treatment with an immunostimulatory drug. The in vivo humanized mouse method is more sensitive than the in vitro assay and can predict the potential for cytokine storm in humans that in vitro testing may miss. Example 15 Cytokine release after treatment with drug combinations

[0264] The PBMC humanized mouse model can effectively test whether a single drug can induce cytokine storm in an individual. However, for drug screening and clinical therapy, it may be necessary to use drug combinations. This example shows that the PBMC humanized mouse model can be used to test drug combinations for cytokine release. The results obtained by the model are shown to be donor specific.

[0265] 6-week-old female NOD.Cg-Prkdc scid Il2rg tm1Wjl / SzJ (NSG, JAX lineage no. 005557) mice were irradiated with 100 cGy of X-rays at least 4 hours prior to engraftment with human PBMCs. Isolated human PBMCs from each donor were added at 3 × 10 7 Mice were intravenously injected with 1000 cells / mouse. PBMCs were derived from donor 213 or donor 364. Six days after PBMC engraftment, applicants administered the drug combinations to mice and determined cytokine levels in mouse serum 6 hours after drug administration using the BD CBA Th1 / Th2 II kit. Five mice were used in all experimental groups.

[0266] The drug combinations tested were KEYTRUDA® (pembrolizumab) and REVLIMID® (lenalidomide); KEYTRUDA® and anti-thymocyte globulin (ATG) (THYMOGLOBULIN® (rabbit)); and anti-CD28 and ATG.

[0267] Pembrolizumab (KEYTRUDA®) is a humanized antibody used in cancer immunotherapy. Lenalidomide (REVLIMID®) is a derivative of thalidomide, an oral immunomodulatory small molecule drug used in cancer treatment. Antithymocyte globulin (ATG), sold as THYMOGLOBULIN® (rabbit), is an immunosuppressant drug used to reduce the body's natural immunity in patients receiving transplants, such as kidney transplants.

[0268] In the KEYTRUDA® (pembrolizumab) and REVLIMID® (lenalidomide) experiments, humanized NSG mice were injected intravenously (iv) with 5 mg / kg KEYTRUDA® per 100 mg / kg lenalidomide administered orally, or received both drugs. Treatment with 5 ml / kg PBS iv was used as a control.

[0269] In the KEYTRUDA® (pembrolizumab) and ATG experiments, humanized NSG mice were injected intravenously with 5 mg / kg KEYTRUDA®, 1 mg / kg ATG, or received both drugs. Treatment with 5 ml / kg PBS iv was used as a control.

[0270] In anti-CD28 and ATG experiments, humanized NSG mice were injected intravenously with 1 mg / kg anti-CD28, 1 mg / kg ATG, or received both drugs. Treatment with 5 ml / kg PBS iv was used as a negative control.

[0271] Cytokine levels determined in the KEYTRUDA® (pembrolizumab) and REVLIMID® (lenalidomide) studies are shown for both donors in Figure 16. Cytokines whose levels were measured were IFN-γ (panel a); IL-10 (panel b); IL-6 (panel c); IL-2 (panel d); IL-4 (panel e); and TNF (panel f).

[0272] Based on data from two stopped clinical trials evaluating KEYTRUDA® (pembrolizumab) in combination with dexamethasone and the immunomodulatory agent REVLIMID® (lenalidomide) to treat multiple myeloma, the FDA issued a statement in 2017 that treatment with the combination of KEYTRUDA and lenalidomide posed an increased risk of severe toxicity and death.

[0273] As shown in FIG. 16 , IFN-γ, IL-6, IL-2, and TNF each demonstrated a significant increase in cytokine release when KEYTRUDA® (pembrolizumab) and REVLIMID® (lenalidomide) were used in combination above the levels produced by treatment with each drug alone, however this result was for donor 213 but not for donor 364.

[0274] Antithymocyte globulin (ATG) is used in the prevention and treatment of acute rejection in organ transplants and in the therapy of aplastic anemia. ATG has previously been demonstrated to stimulate clinical toxicity. Figure 17 shows cytokine results after treatment with KEYTRUDA® (pembrolizumab) alone, ATG alone, and the combination of KEYTRUDA® and ATG. Levels of IFN-γ, IL-6, IL-10, and TNF were higher after treatment with the combination of KEYTRUDA® and ATG than after treatment with either drug alone, but this result was for donor 213 but not for donor 364.

[0275] Figure 18 shows cytokine results after treatment with anti-CD28 alone, ATG alone, and the combination of anti-CD28 and ATG. For donor 213, levels of IFN(IIFN)-γ, IL-6, IL-2, and TNF were higher after treatment with the drug combination compared to treatment with either drug alone, while for donor 364, only IL-2 and TNF were increased after treatment with the drug combination compared to treatment with either drug alone.

[0276] These results indicate that the in vivo humanized mouse model can predict the likelihood of high cytokine release from single drugs and for drug combinations as well. In addition, different PBMC donors showed different cytokine release responses to combination therapy.

[0277] This study demonstrates the advantage of using a patient's own PBMCs in an in vivo method to screen whether a particular drug combination may be likely to induce severe cytokine release in a patient prior to clinical administration. Similarly, this method is useful for screening for possible drug-related toxicity associated with the administration of drug combinations. (Example 16) Human immune cell populations and cell type distribution in irradiated humanized mice

[0278] In this example, we evaluated human immune cell populations and cell type distribution in irradiated humanized mice. The effect of X-ray irradiation prior to and time after engraftment of human PBMCs into immunodeficient mice on human cell populations within the mice is demonstrated.

[0279] 6-week-old female NOD.Cg-Prkdc scid Il2rg tm1Wjl / SzJ (NSG, JAX line number 005557) mice were used for engraftment of human PBMCs. For mice subjected to irradiation treatment, the mice were irradiated with 100 cGy of X-rays at least 4 hours prior to engraftment of human PBMCs. Purified / isolated human PBMCs from donors were added at 2×10 7 Mice were injected intravenously with 1000 cells / mouse. PBMCs were derived from one of six different donors: 362, 345, 2785, 213, 364 or 3251. Mice were bled on days 5 and 10 post-engraftment and human cells in mouse blood were analyzed by flow cytometry to determine the human cell types present and the percentage of each type. Specifically, mouse PBMCs in a mixture with human PBMCs were stained with human antibodies: anti-CD45, anti-CD3, anti-CD14, anti-CD19 and anti-CD56 mAbs. The results are shown in Figure 19, which illustrates that human cell populations differ between irradiated and non-irradiated mice on days 5 and 10.

[0280] As shown in Figure 19, on day 5, the total human white blood cell population (CD45+) was approximately 8-15% in irradiated mice, but only 1-3% in non-irradiated mice, demonstrating diversity between donors with respect to the total population. Figure 19 further shows that on day 5, the distribution of white blood cell types within the total human cell (CD45+) population for irradiated or non-irradiated mice was nearly identical, with the predominant cell types in each being CD3+ cells (T cells) and CD56+ (NK cells), with smaller percentages of CD19+ cells (B cells) and CD14+ cells (monocytes).

[0281] Figure 19 shows that on day 10, the total human cell (CD45+) population grew to approximately 30-85% of viable cells in irradiated mice, but only 8-17% of viable cells in non-irradiated mice, with variability between individual donors.

[0282] Figure 19 further shows that on day 10, the distribution of leukocyte cell types within the total human cell (CD45+) population was again similar in irradiated or non-irradiated mice. In either irradiated or non-irradiated mice, the predominant cell type was CD3+ cells (T cells), which grew and expanded the population in the mice. Only a small or negligible percentage of viable CD56+ (NK cells), CD19+ cells (B cells) and CD14+ cells (monocytes) were present in irradiated or non-irradiated mice - i.e., these cells did not grow and expand the population in the mice.

[0283] The absolute numbers of human cells present on days 5 or 10 in irradiated mice were approximately 3-15 times higher than the absolute numbers of human cells present on the same days in non-irradiated mice, indicating that irradiated mice exhibited greater numbers of each of the human leukocyte types present. Importantly, on day 5, irradiated mice had greater absolute numbers of viable CD56+ (NK cells), CD19+ cells (B cells) and CD14+ cells (monocytes) within the human CD45+ cell population, providing an excellent model of the human immune system that enhances the sensitivity and accuracy of testing for cytokine release syndrome. (Example 17) Weight loss as an indicator of GVHD in irradiated and non-irradiated humanized mice

[0284] In this study, applicants used weight loss as an indicator of GVHD in humanized mice. Body weight measurements were taken daily in humanized mice engrafted with 20 million PBMCs from each of six different human donors (362, 345, 2785, 213, 364, or 3251). Mice were either irradiated or not prior to engraftment of donor PBMCs according to the procedure of Example 16. Body weight measurements for irradiated and non-irradiated groups per donor are shown in Figure 20, panel a (donors 362, 345, and 2785); panel b (donors 213, 364, and 3251). Figure 20 shows that for all donors, weight loss occurs earlier in irradiated mice than in non-irradiated mice, demonstrating that GVHD develops faster in irradiated humanized mice than in non-irradiated humanized mice. For all six human donors, the weight loss of irradiated humanized mice after day 8 was significant (at least about 10%), indicating significant GVHD after day 8. Non-irradiated mice did not show significant weight loss even 12-14 days after PBMC engraftment, indicating a much later onset of significant GVHD. (Example 18) GVHD alone (in the absence of drug treatment) triggers human cytokine release in humanized mice

[0285] This example demonstrates that mice experiencing significant GVHD secreted human cytokines as a result of GVHD. GVHD developed more quickly in irradiated humanized mice than in non-irradiated humanized mice.

[0286] Figure 21 shows human cytokine levels present in humanized mice 10 days after PBMC engraftment in the absence of any drug treatment. Cytokine measurements are shown for IFN-γ (panel a), IL-10 (panel b) and IL-6 (panel c) for humanized mice from six donors (362, 345, 2785, 213, 364 and 3251) with or without irradiation prior to PBMC engraftment. For each donor, more of each of the three cytokines was present in the blood of irradiated mice than non-irradiated mice 10 days after engraftment. At day 10, based on the body weight measurements disclosed in Example 17, irradiated mice had more severe GVHD than non-irradiated mice due to the resulting GVHD-associated secretion of human cytokines from present human leukocytes. (Example 19) Cytokine release in irradiated and non-irradiated humanized mice

[0287] This example evaluates the effect of X-irradiation of mice (prior to PBMC engraftment) on drug-induced human cytokine release.

[0288] Mice were engrafted with 20 million PBMCs from donor 362 (high responder) or donor 21 (intermediate responder). Mice were either irradiated or not prior to engraftment of donor hPBMCs. Six days after engraftment, mice were treated with immunomodulatory drugs (iv injection of 0.5 mg / kg OKT3, 10 mg / kg KEYTRUDA, 1 mg / kg ATG or 5 ml / kg PBS (negative control)). Cytokine levels were determined in mouse blood obtained 6 hours after drug treatment. Results comparing drug-induced cytokine levels in irradiated and non-irradiated humanized mice are shown in Figures 22A-B. Figure 22A presents data for IFN-γ (panel a), IL-10 (panel b) and IL-6 (panel c), while Figure 22B presents data for IL-2 (panel d), IL-4 (panel e) and TNF (panel f). Figures 22A-B show that for each cytokine, in either donor, cytokine levels in non-irradiated mice were consistently lower after a given drug treatment compared to those determined in irradiated mice after the same drug treatment.

[0289] The higher cytokine levels produced in irradiated mice may be due to the higher number of human immune cells (T cells and NK cells) present in the mice, as disclosed in Example 16. Thus, the irradiated humanized mouse model offers greater sensitivity for the detection of individual variability in immune activation in response to a given drug. Materials & Methods 1. PBMC Humanized Mouse Reconstitution

[0290] 6-week-old female NOD.Cg-Prkdc scid Il2rg tm1Wjl NSG / SzJ (NSG, JAX lineage no. 005557) mice and their derived lines, NSG-CSF-1, NSG-IL-6 mice were irradiated with 100 cGy of X-rays at least 4 h prior to human PBMC engraftment. Purified human PBMCs were purchased commercially (Astarte Biologics or Allcells technologies). PBMCs were washed twice with PBS after thawing, then 10–50 million cells / mouse were injected intravenously (iv) into NSG mice. Mice were observed daily after hPBMC injection for overall appearance of fur and motility. On another day after hPBMC engraftment, mice were bled to test human cell percentages by flow cytometry. Mouse PBMCs were stained with human anti-CD45, anti-CD3, anti-CD14, anti-CD19, and anti-CD56. 2. Induction and Measurement of Cytokine Release in Humanized Mice

[0291] On day 6 of hPBMC engraftment, mice were induced for human cytokine release by iv injection of antibodies OKT3 (anti-CD3 mAb) and ANC28 (anti-CD28 mAb), KEYTRUDA® (pembrolizumab) (anti-PD-1), ATG, REVLIMID® (lenalidomide) and PBS buffer as a baseline control. Mice were bled at 2 and / or 6 hours and serum was collected and analyzed for cytokine concentrations using the BD Cytometric Bead Array (CBA) Human Th1 / Th2 Cytokine Kit II (BD Cat. No. 551809). The detection limits of the assay for various cytokines are as follows: IFN-γ, 7 pg / ml; IL-2, 2.6 pg / ml; IL-4, 2.6 pg / ml; IL-6, 3.0 pg / ml; IL-10, 2.8 pg / ml; and TNF, 2.8 pg / ml. 3. Measuring body temperature

[0292] Rectal temperatures of mice were measured prior to treatment and immediately prior to bleeding at each time point. Temperatures were measured by inserting a rectal thermocouple probe and waiting until a stable reading was obtained. 4. Clinical Score Assessment

[0293] As referenced by Brady et al., Clinical & Translational Immunology, 2014 Similarly, we graded the signs and scores as follows: Score: 0 = normal activity; 1 = normal activity, piloerection, walking on tiptoes; 2 = hunched, less active but still mobile; 3 = less mobile but mobile when prompted; 4 = moribund (critically ill). Mice with a clinical score of 4 were euthanized. 5.Statistical analysis

[0294] Results were analyzed using GraphPad Prism 5.0.

[0295] The present disclosure further encompasses the following aspects.

[0296] Aspect 1. A method for determining whether an immunomodulatory agent is likely to induce severe cytokine release syndrome in humans following administration of the immunomodulatory agent, comprising the steps of: (a) providing an immunodeficient mouse, said mouse being irradiated with 75-125 cGy of X-rays; and (b) detecting 1.5-3.0 x 10 7 (c) administering an immunomodulatory agent to the mouse 5-7 days after engraftment of the PBMCs; (d) determining blood levels of a plurality of cytokines in the mouse, including IFN-γ and IL-10, wherein blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse; and (e) determining that the immunomodulatory agent is likely to induce severe cytokine release syndrome in the human, wherein the presence of severe cytokine release syndrome in the mouse indicates that administration of the immunomodulatory agent is likely to induce severe cytokine release syndrome in the human.

[0297] Aspect 2. A method for determining the likelihood that administration of an immunomodulatory agent to a human will induce severe cytokine release syndrome in the human, comprising: (a) detecting 1.5 to 3.0 x 10 human-derived cytokines. 7 (b) providing a blood sample from a humanized irradiated immunodeficient mouse administered an immunomodulatory agent 5-7 days after engraftment of isolated peripheral blood mononuclear cells (PBMCs); and (b) detecting in vitro the concentration of a plurality of cytokines, including IFN-γ and / or IL-10, present in the mouse blood sample, wherein a concentration of IFN-γ≧1,800 pg / ml or IL-10≧120 pg / ml in the mouse blood sample indicates that administration of the immunomodulatory agent to a human may induce severe cytokine release syndrome.

[0298] Aspect 3. A method of determining whether a combination of a first immunomodulatory agent and a second immunomodulatory agent is likely to induce severe cytokine release syndrome in a human following administration of said combination of immunomodulatory agents, the method comprising the steps of: (a) providing an immunodeficient mouse, said mouse being irradiated with 75-125 cGy of X-rays; and (b) detecting 1.5-3.0 x 10 7 (c) administering a first immunomodulatory agent and a second immunomodulatory agent to the mouse 5-7 days after engrafting the PBMCs; (d) determining blood levels of a plurality of cytokines in the mouse, including IFN-γ and IL-10, wherein blood levels of IFN-γ≧1,800 pg / ml and IL-10≧120 pg / ml indicate severe cytokine release syndrome in the mouse; and (e) determining that the combination of immunomodulatory agents is likely to induce severe cytokine release syndrome in the human, wherein the presence of severe cytokine release syndrome in the mouse indicates that administration of the combination of immunomodulatory agents is likely to induce severe cytokine release syndrome in the human.

[0299] Aspect 4. A method of determining the likelihood that administration of a combination of a first and a second immunomodulatory agent to a human will induce severe cytokine release syndrome in the human, comprising: (a) providing a blood sample from an irradiated humanized immunodeficient mouse administered the combination of a first and a second immunomodulatory agent 5-7 days after engraftment of 1.5-3.0×107 isolated peripheral blood mononuclear cells (PBMCs) from the human; and (b) detecting in vitro the concentration of a plurality of cytokines, including IFN-γ and / or IL-10, present in the mouse blood sample, wherein a concentration of IFN-γ≧1,800 pg / ml or IL-10≧120 pg / ml indicates that administration of the combination of a first and a second immunomodulatory agent to a human is likely to induce severe cytokine release syndrome.

[0300] Aspect 5. A method for determining a safe dosage of an immunomodulatory agent that does not induce cytokine release syndrome in humans following administration of the immunomodulatory agent, comprising the steps of: (a) providing an immunomodulatory agent having a first dosage, wherein the first dosage of the immunomodulatory agent is determined to induce mild or severe cytokine release syndrome in a first humanized irradiated immunodeficient mouse following administration of the immunomodulatory agent; (b) providing a second immunodeficient mouse, wherein the second mouse is irradiated with 75-125 cGy of X-rays; and (c) detecting 1.5-3.0 x 10 cytokines isolated from a human. 7 (d) administering an immunomodulatory agent to the second mouse 5-7 days after engrafting the PBMCs, wherein the immunomodulatory agent is administered at a second dosage less than the first dosage; (e) determining blood levels of a plurality of cytokines in the second mouse, including IFN-γ and IL-10; and (f) determining a safe dosage of the immunomodulatory agent for administration in the human. determining a safe dosage that results in blood levels of IFN-γ <300 pg / ml and IL-10 <25 pg / ml following administration of the immunomodulatory agent to the second mouse, wherein blood levels of IFN-γ<300 pg / ml and IL-10<25 pg / ml in the second mouse indicate that administration of the safe dosage of the immunomodulatory agent is likely not to induce cytokine release syndrome in the human.

[0301] Aspect 6. A method for determining the immunotoxicity of a drug candidate for use in humans, comprising the steps of: (a) providing an immunodeficient mouse, said mouse being irradiated with 75-125 cGy of X-rays; and (b) administering 4.5-5.5×10 7(c) administering a drug candidate to the mouse 4 to 7 days after engraftment; (d) determining a cytokine concentration in the blood of the mouse, wherein the cytokine is at least one cytokine selected from the group consisting of IFN-γ≧300pg / ml, IL-2≧15pg / ml, IL-4≧10pg / ml, IL-6≧10pg / ml, IL-10≧25pg / ml, or TNF≧5pg / ml; and (e) determining the immunotoxicity of the drug candidate, wherein a blood concentration in the mouse of at least one cytokine selected from the group consisting of IFN-γ≧300pg / ml, IL-2≧15pg / ml, IL-4≧10pg / ml, IL-6≧10pg / ml, IL-10≧25pg / ml, or TNF≧5pg / ml is indicative of the immunotoxicity of the drug candidate in humans.

[0302] A method for determining the immunotoxicity of a drug candidate in humans, comprising: (a) detecting 4.5 to 5.5 × 10 7 (b) providing a blood sample from a humanized irradiated immunodeficient mouse administered the drug candidate 4-7 days after engraftment with isolated human peripheral blood mononuclear cells (PBMCs); and (b) detecting in vitro the concentration of at least one human cytokine present in the mouse blood sample to determine human immunotoxicity of the drug candidate, wherein the at least one human cytokine is selected from the group consisting of IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNF, and wherein the drug candidate has low human immunotoxicity when a low human cytokine concentration is detected in the mouse blood sample.

[0303] Embodiment 8. The method according to any one of embodiments 1 to 7, wherein said mouse is an NSG, NSG-IL-6 or NSG-CSF-1 mouse.

[0304] Embodiment 9. The method of any one of embodiments 1 to 8, wherein said mouse is an NSG mouse.

[0305] Aspect 10. The method of any one of aspects 1 to 9, wherein said mouse is irradiated with 100 cGy of X-rays.

[0306] Aspect 11. The engraftment step comprises: 7 The method according to any one of embodiments 1 to 5 and 8 to 10, wherein the method is performed on individual PBMCs.

[0307] Aspect 12. The method of any one of aspects 1 to 11, wherein said administering step occurs 6 days after engraftment.

[0308] Embodiment 13. The method of any one of embodiments 1 to 12, wherein said plurality of cytokines further comprises IL-2, IL-4, IL-6, and TNF.

[0309] Aspect 14. The method of any one of aspects 1 to 13, wherein the blood levels of a plurality of cytokines are determined 2 to 6 hours after administration of said immunomodulatory agent or said combination of immunomodulatory agents.

[0310] Embodiment 15. The method of any one of embodiments 1 to 14, wherein the blood levels of a plurality of cytokines are determined 6 hours after administration of said immunomodulatory agent or said combination of immunomodulatory agents.

[0311] Aspect 16. The immunomodulatory agent is selected from the group consisting of anti-CD28 monoclonal antibody (mAb), anti-CD3 mAb, anti-CD20 mAb, anti-CD52 mAb; granulocyte colony-stimulating factor (G-CSF); interferon; imiquimod; thalidomide, lenalidomide, pomalidomide, apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; talimogene laherparepvec; adalimumab, catumaxomab, ibritumomab tiuxetan, tositumomab-I 13116. The method of any one of embodiments 1 to 15, wherein the IL-16 antibody is selected from the group consisting of: brentuximab vedotin, cetuximab, rituximab, alemtuzumab, bevacizumab, pertuzumab, trastuzumab, trastuzumab emtansine, denosumab, ofatumumab, panitumumab, pembrolizumab, nivolumab, ipilimumab, atezolizumab, avelumab, durvalumab, daratumumab, ceritinib, elotuzumab, and anti-thymocyte globulin.

[0312] Embodiment 17 The method of embodiment 16, wherein said anti-CD28 mAb is TGN1412.

[0313] The method of claim 16, wherein said anti-CD3 mAb is OKT3.

[0314] Aspect 19. The method of aspect 16, wherein said anti-C20 mAb is rituximab.

[0315] Aspect 20. The method of aspect 16, wherein said anti-CD52 mAb is alemtuzumab.

[0316] Aspect 21. The first immunomodulatory agent and the second immunomodulatory agent are selected from the group consisting of anti-CD28 monoclonal antibody (mAb), anti-CD3 mAb, anti-CD20 mAb, anti-CD52 mAb; granulocyte colony-stimulating factor (G-CSF); interferon; imiquimod; thalidomide, lenalidomide, pomalidomide), apremilast; azathioprine, cladribine, cyclophosphamide, intravenous immunoglobulin, methotrexate, mitoxantrone; talimogene laherparepvec; adalimumab, catumaxomab, ibritumomab tiuxetan, tositumomab-I 131, brentuximab vedotin, cetuximab, rituximab, alemtuzumab, bevacizumab, pertuzumab, trastuzumab, trastuzumab emtansine, denosumab, ofatumumab, panitumumab, pembrolizumab, nivolumab, ipilimumab, atezolizumab, avelumab), durvalumab, daratumumab, ceritinib, elotuzumab and anti-thymocyte globulin.

[0317] Aspect 22. The method of any one of aspects 3-4, 5-15, and 21, wherein said first immunomodulatory agent is pembrolizumab or nivolumab, and said second immunomodulatory agent is lenalidomide, pomalidomide, epacadostat, talimogene laherparepvec, ipilimumab, atezolizumab, avelumab, rituximab, alemtuzumab, ceritinib, daratumumab, or durvalumab.

[0318] Aspect 22. The method of any one of aspects 3-4, 5-15, and 21, wherein said first immunomodulatory agent is ipilimumab and said second immunomodulatory agent is lenalidomide, pomalidomide, pembrolizumab, atezolizumab, avelumab, rituximab, alemtuzumab, ceritinib, daratumumab, or durvalumab.

[0319] Aspect 23. The method of any one of aspects 3-4, 5-15, and 21, wherein said first immunomodulatory agent is atezolizumab, avelumab, or durvalumab, and said second immunomodulatory agent is lenalidomide, pomalidomide, pembrolizumab, ipilimumab, rituximab, ceritinib, daratumumab, or alemtuzumab.

[0320] Aspect 24. The method of any one of aspects 3 to 4, 5 to 15 and 21, wherein said anti-CD52 mAb is alemtuzumab, said anti-C20 mAb is rituximab, said anti-CD3 mAb is OKT3, or said anti-CD28 mAb is TGN1412.

[0321] Aspect 25. The engraftment step (b) is performed using 5×10 7 8. The method of embodiment 6 or 7, wherein the method is performed on PBMCs.

[0322] Any publication or reference mentioned in this specification is indicative of the level of skill of those skilled in the art to which this invention pertains. All patents, publications and / or references herein are incorporated by reference to the same extent as if each individual publication was specifically and individually indicated to be incorporated by reference in its entirety.

[0323] Although an example embodiment is described herein with respect to the use of a humanized immunosuppressed mouse to determine whether an immunomodulatory drug can be administered to a human without inducing an unacceptably high cytokine response in humans, it should be understood that the method can be used with a variety of mammals and / or drugs, and / or can be used to treat mammals other than humans. Thus, the present invention is not limited to the examples. In view of the teachings provided herein, one skilled in the art will recognize other applications in which the present invention can be used. Thus, one skilled in the art will be able to use the method of the present invention in other applications. Thus, such alternative uses are intended to be part of the present invention.

[0324] In the foregoing specification, the invention has been described with reference to specific embodiments thereof. It will be apparent, however, that various modifications and changes can be made therein without departing from the broader spirit and scope of the invention. It is therefore intended that such changes and modifications fall within the scope of the invention as defined by the claims appended hereto. The specification and drawings are therefore to be regarded in an illustrative rather than a restrictive sense.

Claims

1. 1. A method for determining whether an immunomodulatory agent is likely to induce severe cytokine release syndrome in a human following administration of said immunomodulatory agent, said method comprising: (a) administering human peripheral blood mononuclear cells (huPBMCs) to irradiated immunodeficient mice in an amount that induces measurable human cytokine release in said mice in response to a human T cell activator; (b) administering an immunomodulatory agent to the humanized mouse 4-7 days after administering the huPBMCs; (c) determining a circulating concentration of a cytokine selected from IFN-γ, IL-10, IL-6, IL-2, IL-4, and TNFα in a blood sample from the mouse before the mouse develops symptoms of graft-versus-host disease (GVHD); Including, The method, wherein a blood concentration of the cytokine equal to or exceeding a threshold concentration value indicates severe cytokine release syndrome in the mouse and indicates that the immunomodulatory agent may induce severe cytokine release syndrome in the human.

2. 2. The method of claim 1, wherein the symptoms of GVHD are selected from weight loss, hunched posture, decreased mobility, hair loss, or tachypnea.

3. 3. The method of claim 2, wherein the symptom of GVHD is weight loss.

4. The method of claim 1 , wherein the immunomodulatory agent is an antibody.

5. The method of claim 4 , wherein the antibody is a monoclonal antibody.

6. 10. The method of claim 1, wherein the immunomodulatory agent is a checkpoint inhibitor.

7. 7. The method of claim 6, wherein the checkpoint inhibitor is an antibody.

8. 2. The method of claim 1, wherein the threshold concentration values ​​for IFN-γ, IL-10, IL-6, IL-2, IL-4, and TNFα are 1,800 pg / ml, 120 pg / ml, 25 pg / ml, 80 pg / ml, 25 pg / ml, and 20 pg / ml, respectively.

Images