Rationale for IL-15 superagonists in cancer immunotherapy

KEYWORDS:

• IL-15 superagonist
• NK cells
• CD8 T cells
• immune activation

View correction statement:

Correction

1. Introduction

Interleukin (IL)-15 (IL-15) is a promising cytokine for the treatment of hematological and solid malignancies. It is a member of the common receptor gamma chain (γ_c) family, which also includes IL-2, IL-4, IL-7, IL-9, and IL-21. This group of cytokines has broad pleiotropic activity on both the innate and adaptive immune systems, with important therapeutic ramifications, reviewed elsewhere [1,2]. IL-15 signals through a receptor complex composed of the IL-2/IL-15 receptor β (CD122) and the common γ_c receptor subunit (CD132), also shared by all other family members [3]. IL-15Rα is not a signaling receptor for IL-15 per se, but it is critical for IL-15 biological activity [4]. Due to its high affinity for IL-15, the IL-15Rα transmembrane receptor binds intracellularly to IL-15, facilitating its trafficking and cell surface expression as IL-15/IL-15Rα heterodimer complexes, which can also be cleaved from the cell surface [4]. Primarily expressed by activated monocytes, macrophages, and dendritic cells, membrane-bound IL-15/IL-15Rα complexes are presented in trans to cells that express the receptor complex [3]. As such, IL-15 preferentially stimulates activation, proliferation, survival, and cytotoxicity of NK and CD8⁺ T cell populations, including regulation of memory CD8⁺ T cell homeostatic proliferation and survival [3]. IL-15 has similar receptor binding, biologic activity, and signaling to IL-2, a cytokine with a long history of application as a cancer immunotherapeutic agent. However, in contrast to IL-2, IL-15 does not induce activation-induced cell death (AICD) of T cells nor enhance the proliferation, function, or differentiation of immunosuppressive CD4⁺ T regulatory cells (T_reg) [3]. In addition, IL-15 does not induce severe capillary leak syndrome in nonhuman primates (NHP) or humans, a toxicity associated with IL-2 therapy [3]. Thus, IL-15 has the potential to be a superior cancer therapeutic to IL-2 given its ability to activate effector immune cells necessary for tumor control with fewer associated toxicities.

2. Recombinant IL-15 as a cancer therapy

Although unassociated IL-15 is not found naturally in vivo, soluble recombinant (r) IL-15 was initially examined for therapeutic potential. In multiple murine carcinoma models, rIL-15 promoted significant anti-tumor efficacy associated with enhanced CD8⁺ T and NK cell function [5]. In murine models of lymphoma and adult T cell leukemia, rIL-15 enhanced the antitumor efficacy of both rituximab (anti-CD20) and alemtuzumab (anti-CD52), respectively. The higher antitumor effects observed with these combinations were associated with augmented antibody-dependent cell-mediated cytotoxicity (ADCC) mediated by NK cells and macrophages [6].

The first-in-human trial of bolus intravenous (i.v.) rIL-15 in metastatic malignant melanoma (MM) and metastatic renal cell cancer (RCC) also reported hyperproliferation and activation of these effector cells [7]. In addition, while there were no objective remissions, two patients with MM experienced clearance of lung lesions [7]. However, dose-limiting toxicities (DLTs), including fever, chills, and blood pressure changes, were observed at the doses required to induce immune and anti-tumor responses. When rIL-15 was administered subcutaneously (s.c.), 6.67-fold more rIL-15 was administered before reaching the maximum tolerated dose (MTD) [8]. NK cell numbers were increased at all administered dose levels [8]. Disease stabilization was observed in non-small cell lung cancer (NSCLC) and RCC patients, but no overall objective clinical response was reached. Again, DLTs were observed at the higher doses of rIL-15, including grade 2 or 3 fevers and grade 3 chest pain in one patient [8]. Alternative dosing strategies, including continuous intravenous infusion, are being examined for feasibility, safety, and efficacy of rIL-15 treatment. rIL-15 is currently in four active Phase I clinical trials in combination with alemtuzumab (anti-CD52) (NCT02689453), ipilimumab (anti-CTLA4 and nivolumab (anti-PD-1) (NCT03388632), mogamulizumab (anti-CCR4) (NCT04185220), and obinutuzumab (anti-CD20) (NCT03759184)).

3. IL-15 superagonists

The efficacy of rIL-15 is limited by its short half-life due to instability, ranging from <1 hour in mice and 2.5 to 12 hours in patients depending on the route of administration [7–9], and IL-15Rα availability in vivo. To improve the therapeutic potential of IL-15 therapy, several types of IL-15/IL-15Rα complexes were developed. These complexes, or IL-15 superagonists, are overall described to have greater potency, bioavailability, and stability than soluble rIL-15 due to their structure. In addition, IL-15 superagonists do not require transpresentation or cell-to-cell contact to induce IL-15-mediated responses. Of note, while cellular and viral IL-15 superagonist expression systems have been generated (see Guo et al. review [10]) and additional IL-15 superagonists are being examined preclinically, this discussion includes only soluble IL-15/IL-15Rα superagonists that are currently in clinical development.

There are four main soluble structurally distinct [4] IL-15 superagonists in clinical development: heterodimeric (het) IL-15 [11], receptor-linker-IL-15 (RLI) [12], IL-15/IL-15Rα-Fc [9,13], and N-803 (formerly ALT-803) [14] (Table 1). N-803 is produced in Chinese hamster ovary cells [14], (het) IL-15 is produced in human 293 cells [11], and RLI is produced in insect SF9 cells [12]. The fusion of hIgG1-Fc domain in IL-15/IL-15Rα-Fc and N-803 enhances the plasma half-life and mediates ADCC [3,15]. In general, these IL-15 superagonists have greatly increased IL-15 potency and serum stability compared to rIL-15, with half-lives of 7–20 hours in mice [9,16,17]. N-803 was also detectable in patient serum up to 4–7 days after s.c. injection [18]. In vitro, N-803 increased human NK-mediated cytotoxicity against B cell lymphoma, head and neck, lung, and breast carcinoma cells, including in the presence of cetuximab (anti-EGFR) or avelumab (anti-PD-L1) [19–2021]. hetIL-15 significantly expanded CD8⁺ T cells in the lymph nodes and overall lymphocyte numbers in the blood and spleen of rhesus macaques [22,23]. In preclinical murine tumor models, the expansion of NK and CD8⁺ T cell populations [9,16,17,24,25] was associated with a reduction in B16 melanoma and CT26 colon tumor burden by RLI [3,16] or B16 melanoma and pancreatic tumor growth by IL-15/IL-15Rα-Fc [9,26]. The combination of RLI with anti-PD-1 therapy further improved responses against CT26 murine colon tumors versus either monotherapy [27]. N-803 restricted tumor growth and improved the median overall survival (mOS) of mice with colon and breast tumors, among others [24,25]. This effect was further improved upon combination with anti-CTLA4 and/or anti-PD-L1 [24,25]. Additionally, the combination of N-803 with anti-CD20 antibody triggered NK cell responses and in vivo clearance of murine B cell lymphomas [21].

Table 1. IL-15/IL-15Rα superagonists in clinical development.

Agent	Description^§	Reference	Clinical Designation	Company	Clinical Development Status
Heterodimeric IL-15 (hetIL-15)	Secreted IL-15/IL-15Rα complex	[7]	NIZ985	Novartis	Phase I (1 active trial)*
Receptor-linked IL-15 (RLI)	Fusion protein of rIL-15 linked to IL-15Rα sushi (cytokine-binding) domain via a 20-amino acid flexible linker	[8]	SO-C101	SOTIO/Cytune Pharma	Phase I (1 active trial)**
NKTR-255	IL-15Rα-dependent polymer-conjugated recombinant human IL-15 (rhIL-15) agonist	[9]	NKTR-255	Nektar Therapeutics	Phase I (1 active trial)*
N-803 (formerly ALT-803)	Fusion protein consisting of a mutated superagonist IL-15 (N72D) bound to IL-15Rα sushi domain linked to IgG1-Fc domain	[10]	N-803 (ALT-803)	ImmunityBio	Phase I/II (23 active trials)*

^§A schematic of all structural variants can be found in Ref. [4]. fc, fragment crystallizable region; h, human; IgG1, immunoglobulin G1; r, recombinant. * www.clinicaltrials.gov; ** www.SOTIO.com.

4. Clinical development of IL-15 superagonists

Given the therapeutic potential seen in preclinical models, all aforementioned IL-15 superagonists are being examined clinically for the treatment of both solid and hematological malignancies (Table 1). RLI, hetIL-15, and NKTR-255 are currently in their first-in-human Phase I clinical trials, whereas N-803 is in Phase I and Phase II clinical trials (Table 2). To our knowledge, N-803 is the only IL-15 superagonist with published clinical results at this time.

Table 2. Selected IL-15 superagonist clinical trials.

Agent	Combination Agent	Cancer Type	Phase	Trial Identifier*
NIZ985	PDR001	Metastatic and advanced solid tumors	Phase I/Ib	NCT02452268
SO-C101	-	Metastatic and advanced solid tumors	Phase I/Ib	SC103
NKTR-255	Daratumumab	Multiple myeloma, NHL	Phase I	NCT04136756
N-803	-	Relapsed or refractory multiple myeloma	Phase I/II	NCT02099539
	-	AML, MDS (post-allogenic HCT)	Phase II	NCT02989844
	Intravesical BCG	High grade NMIBC (BCG unresponsive)	Phase II	NCT03022825
	Intravesical BCG	NMIBC (BCG naïve)	Phase Ib/IIb	NCT02138734
	Nivolumab	Advanced or metastatic NSCLC (PD-1/PD-L1 checkpoint inhibitor refractory or naïve)	Phase Ib/II	NCT02523469
	Pembrolizumab, nivolumab, atezolizumab, avelumab	Advanced solid carcinomas (progression following initial response to PD-1/PD-L1 checkpoint inhibitors)	Phase IIb	NCT03228667
	M7824, MVA-BN-Brachyury, epacadostat	Prostate, advanced solid tumors	Phase I/II	NCT03493945
	Rituximab	Relapsed or refractory iNHL	Phase I/II	NCT02384954
	aNK (NK-92)	Stage III/IV MCC	Phase II	NCT02465957
	Avelumab, haNK	MCC (PD-1/PD-L1 checkpoint inhibitor refractory)	Phase II	NCT03853317
	Pembrolizumab	Stage III/IV NSCLC	Phase II	NCT03520686

AML: acute myeloid leukemia; aNK: activated natural killer cells; BCG: Bacillus Calmette-Guerin; haNK: high affinity natural killer cells; HCT: hematopoietic cell transplant; iNHL: indolent B cell non-Hodgkin's lymphoma; MCC: Merkel cell carcinoma; MDS: myeloid dysplastic syndromes; NHL: non-Hodgkin’s lymphoma; NMIBC: non-muscle invasive bladder cancer; NSCLC: non-small cell lung cancer; SCT: stem cell transplant; SoC: standard of care; TNBC: triple negative breast cancer. *NCT identifiers refer to trials registered at www.clinicaltrials.gov; SC103 trial information can be found at www.SOTIO.com.

N-803 is well-tolerated at doses up to 20 μg/kg (MTD not reached), with most toxicities being mild and non-dose limiting [18,28,29]. Based on preclinical [30] and initial Phase I data [28], N-803 is now administered subcutaneously, as this reduced the severity and number of adverse events, prolonged N-803 serum concentrations, and increased NK cell numbers compared to intravenous delivery [18,28,29]. Numbers of NK cells, and to a lesser extent CD8⁺ T cells, were expanded upon N-803 treatment, correlating with significant increases to serum IFNγ and TNFα [18,28,29].

A recent Phase 1b clinical trial in NSCLC patients (NCT02523469) demonstrated that N-803 in combination with the anti-PD-1 antibody nivolumab produced a 29% objective response rate with nine of 21 (43%) patients experiencing a decrease in target lesion size and 16 of 21 (76%) patients achieving disease control [29]. Of particular interest, 11 patients in the trial had progressed on previous anti-PD-1 treatment. Ten of 11 (91%) of these treatment-resistant patients achieved disease control with 3 (27%) having a partial response and 7 (64%) achieving stable disease [29]. A Phase II trial (NCT03022825) of intravesical N-803 and Bacillus Calmette-Guerin (BCG) in BCG-unresponsive non-muscle invasive bladder cancer (NMIBC) patients has reported preliminary results, including 18 of 20 (90%) subjects with carcinoma in situ (CIS) having a complete response [31]. In the high-grade Ta/T1 papillary disease cohort, 12 of 16 (75%) at 6 months and 7 of 13 (54%) at 9 months remain disease-free [31]. In addition, preliminary results from a Phase Ib trial of N-803 + BCG in BCG-naïve NMIBC patients (NCT02138734) demonstrated that 9/9 (100%) of patients were disease-free at 24 months with no recurrence or progression [32]. A randomized Phase II trial in BCG-naïve NMIBC is now underway (NCT02138734). N-803 is currently in other clinical trials as both a monotherapy and in combination (Table 2).

5. Generation of novel multi-functional IL-15 superagonists

The use of IL-15 superagonists, especially N-803, with other immunotherapies has driven the generation of novel IL-15 superagonists with bi- or multi-specific functions. These agents fuse IL-15/IL-15Rα complexes with oncolytic virus, proteins, or antibodies, thus targeting IL-15 to the tumor site or specific cell type, providing additional costimulatory or activating signals, and/or inhibiting immunosuppression. For example, in preclinical studies, RLI has been fused to the heavy chain of an anti-GD2 ganglioside antibody (GD2-RLI) to target IL-15 to neuroectodermal tumors [33]. GD2-RLI maintains its IL-15 activity and significantly reduces liver metastasis of murine NXS2 neuroblastomas [33]. A novel oncolytic vaccinia virus expressing N-803 demonstrated significant antitumor activity in murine models of colon and ovarian cancers, which was further increased in combination with an anti-PD1 antibody [34]. The N-803-anti-CD20 fusion protein (2B8T2 M, ImmunityBio) successfully mediated ADCC against Daudi B cell lymphoma cells and reduced tumor burden, even compared to rituximab (anti-CD20) monotherapy [35]. N-803 has also been fused to an anti-PD-L1 antibody (designated N-809), allowing it to block immunosuppression and potentially deliver IL-15 to the tumor microenvironment via PD-L1 binding [36]. Whether these agents will provide superior anti-tumor responses in patients compared to IL-15 superagonist monotherapy or combinations remains to be determined.

6. Expert opinion

Given the current clinical results, IL-15 superagonists appear to have an advantage over rIL-15 in mediating immune effects, specifically, greater expansion of NK and CD8⁺ T cells, without inducing significant toxicity. This is most likely due to increased stability/half-life and bioactivity of IL-15 superagonists versus rIL-15. Whether these differences between rIL-15 and IL-15 superagonists translate to better patient outcome remains to be seen. It is possible, however, that the greater activation and expansion of effector immune cells by IL-15 superagonists will lead to negative immune effects, such as NK and T cell exhaustion through overstimulation, overproduction of proinflammatory cytokines by either endogenous immune cells or adoptively transferred cell types (unmanipulated T or NK cells, chimeric antigen receptor (CAR) T cells, high-avidity NK (haNK) cells), induction of autoimmunity, or development of leukemia. The potential of these negative effects is of particular concern given the similarity of IL-15 receptor binding, biologic activity, and signaling to IL-2. High dose IL-2 treatment can induce immune-related adverse events such as vascular leak syndrome, cardiac toxicity, and development of autoimmune syndromes.

It will be important in future clinical trials to optimize the treatment schedule and dose of IL-15 superagonists, especially when combining with other agents that provide costimulatory and activating signals or checkpoint blockade to override immunosuppression. In addition, it will be necessary to examine the differences between the various IL-15 superagonists, since they have different potency and half-life. This differential may result in preferential stimulation of specific immune subsets or broaden the IL-15 activity to populations usually minimally stimulated by IL-15, such as naïve T cells versus effector and memory T cells or NK cells, or different levels and signatures of lymphocyte exhaustion or toxicity profiles.

IL-15 superagonists have a distinct advantage over IL-2 therapy in that they do not promote the expansion of immunosuppressive T_reg and have a better safety profile. However, it is unknown whether new IL-2 therapies like NKTR-214, which does not expand T_reg populations but significantly increases intratumoral CD8⁺ T cell frequency [37], or tumor-targeted IL-2 molecules, which have the potential to reduce systemic toxicity, compare to IL-15 superagonists.

The current positive clinical results demonstrate that IL-15 superagonists are promising agents for the treatment of hematological and solid malignancies. Future studies will unravel if IL-15 superagonists can modulate the immune system to convert treatment-refractory patients into responders. The identification of biomarkers that identify patients or tumor types that are particularly responsive to IL-15 immunotherapy will aid in this clinical development. In addition, the study of new IL-15 superagonist combinations with immune checkpoint inhibitors, chemotherapy, radiation, or cancer vaccines may open the possibility of successful cancer treatment to a broader patient population.

Declaration of Interest

The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial relationships or otherwise to disclose.

Additional information

Funding

This work was supported by the Intramural Research Program of the Center for Cancer Research, National Cancer Institute, National Institutes of Health.