Combining Immunocytokine and Ex Vivo Activated NK Cells as a Platform for Enhancing Graft-Versus-Tumor Effects Against GD2+ Murine Neuroblastoma

Abstract

Management for high-risk neuroblastoma (NBL) has included autologous hematopoietic stem cell transplant (HSCT) and anti-GD2 immunotherapy, but survival remains around 50%. The aim of this study was to determine if allogeneic HSCT could serve as a platform for inducing a graft-versus-tumor (GVT) effect against NBL with combination immunocytokine and NK cells in a murine model. Lethally irradiated C57BL/6 (B6) x A/J recipients were transplanted with B6 bone marrow on Day +0. On day +10, allogeneic HSCT recipients were challenged with NXS2, a GD2⁺ NBL. On days +14-16, mice were treated with the anti-GD2 immunocytokine hu14.18-IL2. In select groups, hu14.18-IL2 was combined with infusions of B6 NK cells activated with IL-15/IL-15Rα and CD137L ex vivo. Allogeneic HSCT alone was insufficient to control NXS2 tumor growth, but the addition of hu14.18-IL2 controlled tumor growth and improved survival. Adoptive transfer of ex vivo CD137L/IL-15/IL-15Rα activated NK cells with or without hu14.18-IL2 exacerbated lethality. CD137L/IL-15/IL-15Rα activated NK cells showed enhanced cytotoxicity and produced high levels of TNF-α in vitro, but induced cytokine release syndrome (CRS) in vivo. Infusing Perforin^-/- CD137L/IL-15/IL-15Rα activated NK cells had no impact on GVT, whereas TNF-α^-/- CD137L/IL-15/IL-15Rα activated NK cells improved GVT by decreasing peripheral effector cell subsets while preserving tumor-infiltrating lymphocytes. Depletion of Ly49H⁺ NK cells also improved GVT. Using allogeneic HSCT for NBL is a viable platform for immunocytokines and ex vivo activated NK cell infusions, but must be balanced with induction of CRS. Regulation of TNFα or activating NK subsets may be needed to improve GVT effects.

Keywords: NK cells; cytokine release syndrome; graft-versus-tumor effect; immunocytokine; neuroblastoma.

Figure 1

Effect of hu14.18-IL2 after T cell deplete and replete allogeneic HSCT. (A) Lethally irradiated CD45.2⁺ B6AJF1 mice (H-2^{b x a}) were transplanted with: either CD3e cell depleted (No T cell) or CD3e depleted BM replenished with 2.5 x 10⁶ T cells (T cell) from congenic CD45.1⁺ B6 mice (H-2^b) on Day +0. On days +14-16, PBS or hu14.18-IL-2 (IC) was administered and allogeneic HSCT recipients were followed for (B) GVHD-associated weight loss and (C) survival. N=5 mice/group. The no T cell group was compared to the corresponding T cell group. (D) Lethally irradiated CD45.2⁺ B6AJF1 mice were transplanted with either no T cells or CD3e depleted BM from congenic CD45.1⁺ B6 mice replenished with logarithmically increasing doses of T cells (2.5 x 10²-10⁶) on Day +0. On days +14-16, IC was administered and allogeneic HSCT recipients were followed for survival. Each group was compared to the 2.5 x 10⁶ T cell group. Results pooled from 2 similar experiments, 5-10 mice/group. (E) Lethally irradiated CD45.2⁺ B6AJF1 mice were transplanted with either no T cells or CD3e depleted BM from congenic CD45.1⁺ B6 mice replenished with no T cells, 2.5 x 10³ or 10⁶ T cells on Day +0. On days +14-16, IC was administered and allogeneic HSCT recipients were sacrificed at Day +21 for flow cytometric analysis of B cells, (F) NK cells, (G) CD8⁺ T cells and (H) CD4⁺ regulatory T cells in the spleen. Results pooled from 2 similar experiments, 5-10 mice/group. NS, not significant *p < 0.05, ***p < 0.001.

Figure 2

GD2⁺ NXS2 neuroblastoma growth after allogeneic HSCT and hu14.18-IL2. (A) Lethally irradiated CD45.2⁺ B6AJF1 mice (H-2^{b x a}) were transplanted with either CD3e cell depleted (T cell depleted) or CD3e depleted BM replenished with 2.5 x 10³ T cells (T cell replete) from congenic CD45.1⁺ B6 mice (H-2^b) on Day +0. On Day +10, 2 x 10⁶ NXS2 tumor cells were inoculated into the right flank. On days +14-16, PBS (No IC) or hu14.18-IL-2 (IC) was administered and (B) tumor growth was monitored by using a digital caliper as well as (C) clinical GVHD scores. N=5 mice/group. ** p = 0.01.

Figure 3

In vitro characterization of ex vivo CD137L/IL-15/IL-15Rα expanded NK cells. Donor B6 NK cells were expanded in vitro with IL-15/IL-15Rα alone (0:1) or with IL-15/IL-15Rα plus an aAPC expressing 4-1BBL (CD137L) at a logarithmically increasing dose of aAPC : NK cell ratios (0.1:1-10:1). (A) Cell counts were enumerated twice per week. Results pooled from 4 separate cultures. (B) After 1 week, NK cells expanded without aAPC or IL-15/IL-15Rα (unexpanded) were compared to NK cells expanded with the aAPC at a 1:1 ratio (CD137L/IL-15/IL-15Rα expanded) for (C) NK1.1 purity, (D) inhibitory Ly49C/I expression and (E) activating Ly49H expression. Results pooled from 2 separate experiments, 2-7 mice/group. (F) Balb/c NK cells (H-2^d) expanded with IL-15/IL-15Rα or with IL-15/IL-15Rα and CD137L aAPCs at a 1:1 NK:aAPC ratio for 1 week were compared for their ability to lyse syngeneic A20 (H-2^d) or allogeneic Yac-1 (H-2^a) lymphoma cells at various E:T ratios using a 4 hour LDH release cytotoxicity assay, performed in triplicate. (G) NK cells expanded without aAPC (IL-15) were compared to NK cells expanded with the aAPC at a 1:1 ratio (IL-15/CD137L) and examined for TNF-α production by ELISA, performed in triplicate. (H) B6 (H-2^b) NK cells expanded with IL-15/IL-15Rα and CD137L aAPCs at a 1:1 NK:aAPC ratio for 1 week were tested for their ability to lyse murine neuroblastoma cell lines Neuro2a, N18TG2, NXS2, and 9464D (H-2^a) using a 4 hour calcein-AM release cytotoxicity assay, performed in triplicate. *p < 0.05, **p < 0.01, ***p < 0.001, NS, not significant.

Figure 4

Effects of ex vivo expanded NK cells after syngeneic and allogeneic MHC-matched and mismatched HSCT. (A) Child into parent HSCT schema showing how “missing self” in host prevents engagement of Ly49 inhibitory receptors on donor NK cells. (B) Lethally irradiated B6 mice (H-2^b) were transplanted with CD3e cell depleted BM from B6 (H-2^b) or CB6F1 mice (H-2^bxd) on Day +0. On Days +3, +10, and +17 HSCT recipients received donor-derived, CD137L/IL-15/IL-15Rα expanded NK cells in increasing dose increments of 1, 2 or 3 x 10⁶ cells, and were followed for weight loss and survival. N=5 mice/group. (C) Lethally irradiated Balb/c mice (H-2^d) were transplanted with CD3e cell depleted BM from CB6F1 mice (H-2^{b x d}) on Day +0. On Day +1 HSCT recipients received 5 x 10⁶ CB6F1 NK cells (H-2bxd) cultured in IL-15 alone, or activated with CD137L/IL-15/IL-15Rα, and were followed for survival and (D) weight loss. Lethally irradiated Balb/c mice (H-2^d) were transplanted with CD3e cell depleted BM from CB6F1 mice (H-2^{b x d}) on Day +0. On Day +1 HSCT recipients received 5 x 10⁶ CB6F1 NK cells (H-2^bxd) cultured in IL-15 alone, or activated with CD137L/IL-15/IL-15Rα, and were followed for survival. Results pooled from 2 separate experiments, 9 mice/group. (E) Lethally irradiated A/J mice (H-2^a) were transplanted with CD3e cell depleted BM from B6 mice (H-2^b) on Day +0. On Days +14-16, PBS (No IC) or hu14.18-IL-2 (IC) was administered. On Day +16, select groups were infused with 2.5 x 10⁶ CD137L/IL-15/IL-15Rα expanded B6 NK cells and followed for survival. N= 5 mice/group. (F) Lethally irradiated Balb/c mice (H-2^d) were transplanted with CD3e cell depleted BM from DBA mice (H-2^d) on Day +0. On Day +1 HSCT recipients received 5 x 10⁶ CB6F1 NK cells (H-2^{b x d}) cultured in IL-15/IL-15Rα alone, or activated with CD137L/IL-15/IL-15Rα, and were followed for weight loss and survival. N=5 mice/group. *p < 0.05.

Figure 5

Ex vivo CD137L/IL-15/IL-15Rα expanded NK cells mediate cytokine release syndrome in vivo. Lethally irradiated B6AJF1 mice (H-2^{b x a}) were transplanted with CD3e cell depleted BM from B6 mice (H-2^b) on Day +0. On Day +10, 2 x 10⁶ NXS2 tumor cells were inoculated into the right flank. On days +14-16, hu14.18-IL-2 (IC) was administered. On Day +16, groups were infused with PBS or 2.5 x 10⁶ CD137L/IL-15/IL-15Rα expanded B6 NK cells. Peripheral blood was collected pre-alloHSCT, and then 3 and 4 weeks post-alloHSCT. Serum was isolated from each blood sample and all timepoints were frozen, batched and analyzed in duplicate by multiplex cytokine array. N= 5 mice/group. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6

Impact of infusing ex vivo CD137L/IL-15/IL-15Rα expanded NK cells on GD2⁺ NXS2 neuroblastoma growth after allogeneic HSCT. Lethally irradiated A/J mice (H-2^a) were transplanted with CD3e cell depleted B6 BM (H-2^b) with 2.5 x 10² B6 T cells on Day +0. On Day +10, 2 x 10⁶ NXS2 tumor cells were inoculated into the right flank. On Days +14-16, hu14.18-IL-2 (IC) was administered. (A) Flow cytometric analysis of the spleen was performed on Day +7 (pre-NXS2), +11 (post-NXS2, pre-IC or NK cells) or +18 (post-IC and/or NK cells). (B) On Day +16, select groups were infused with 2.5 x 10⁶ CD137L/IL-15/IL-15Rα expanded NK cells from B6 wild type (Perf^+/+) or B6 Perforin (Perf^-/-) donors, or (C) B6 wild type (TNFα^+/+) or B6 TNFα^-/- donors. (D) Both spleens and tumors were harvested from recipients of B6 wild type (TNFα^+/+) or B6 TNFα^-/- donors at Day +30 and analyzed for T cells and NK cell subsets. (E) Mice were transplanted and challenged with NXS2 as above but on Day +16 were infused with 2.5 x 10⁶ CD137L/IL-15/IL-15Rα expanded NK cells from B6 wild type and then treated with anti-Ly49H depletion or an isotype control. All mice were followed for NXS2 tumor growth. N=3-5 mice/group. *p < 0.05. ****p < 0.001. NS, not significant.