Phase I Trial Characterizing the Pharmacokinetic Profile of N-803, a Chimeric IL-15 Superagonist, in Healthy Volunteers

In less than a decade, immunotherapies have transformed cancer treatment, while highlighting the fundamental role of the immune system in suppressing tumor initiation and growth. As a central regulator of NK cell and T cell homeostasis and function, the cytokine IL-15 shows particular promise as a target for development of novel immunotherapies (1–3).

IL-15 is a member of a family of six known cytokines that signal through an IL-2 common γ chain receptor subunit (CD132) (1–4). IL-2 is one of the first cytokines that proved effective in inducing clinical responses in cancer patients (5–7). In addition to signaling through the common γ chain receptor, IL-2 and IL-15 share a common β receptor (CD122). IL-2 and IL-15 are both effective at inducing the expansion and functional activation of lymphocytes, including NK cells and CD8⁺ T cells. However, a critical distinction is that cell surface expression of the IL-2Rα subunit selectively enhances the ability of IL-2 to signal through the IL-2/IL-15Rβγ receptor. As FOXP3⁺ T regulatory cells constitutively express high levels of IL-2Rα, administration of IL-2 but not IL-15 preferentially induces the expansion and activation of regulatory T cells.

A first-in-human study assessing rhIL-15 as treatment for patients with metastatic melanoma or renal cell cancer showed that rhIL-15 administration induced proliferation of NK cells and CD8⁺ T cells (8). Although treatment did not result in objective responses, two patients experienced clearing of lung lesions. Though rhIL-15–stimulated lymphocyte proliferation, the protein’s serum t_1/2 of 2–3 h made daily administration of the drug necessary.

N-803 (also known as ALT-803) is an engineered chimeric IL-15 superagonist developed to have increased signaling strength and serum half-life relative to rhIL-15 (9). N-803 consists of an IL-15 mutein (N72D) associated with high affinity to a dimeric IL-15Rα sushi domain/IgG1 Fc fusion protein. The N72D mutation, as well as association of the IL-15Rα with the IL-15 mutein, increases the biological activity of the molecule relative to IL-15 (10–15). Preclinical studies have shown that N-803 administration markedly increases expansion of NK cells and CD8⁺ T cells, as well as increasing the functionality and cytotoxicity of these cells through a number of mechanisms including upregulated expression of the NKG2D receptor and granzyme B (GzB) (10–18).

In clinical settings, N-803 is well tolerated and has promising anticancer activity. When administered s.c. to patients with advanced solid tumors, the most common adverse events (AEs) were grade 1–2 injection site reactions, fatigue, and hypoalbuminemia (19). When administered to patients with hematologic malignancy, similar types of AEs were observed, and clinical benefit was observed in 19% of evaluable patients (20). In a phase 1b study of N-803 administered in combination with nivolumab to patients with metastatic non-small cell lung cancer (NSCLC), the most common AEs were grade 1–3 injection site reactions, flu-like symptoms, fever, and fatigue; no grade ≥4 AEs occurred (21). Although this phase 1b study was also not designed to assess efficacy, 6 of 21 patients (29%) experienced an objective response, including patients that had relapsed on prior anti–PD-1 therapy.

Given the importance of the N-803 complex in modulating the immune system and antitumor immunity, we conducted a single-center, open-label phase 1 clinical trial with the primary objective of establishing the pharmacokinetic (PK) profile of N-803 after s.c. administration to healthy volunteers. The secondary objective was to evaluate the safety of N-803. In addition, N-803 effects on lymphocyte proliferation, inflammatory cytokine levels, and immunogenicity were measured.

Twenty healthy adult subjects (11 males, 9 females) 23–57 y of age with no recent cancer history were enrolled in the study. Fourteen subjects completed the trial. Subjects were randomized 1:1 to receive N-803 formulated at either 1.0 mg/ml (Fig. 1; group A, 10 subjects) or 2.0 mg/ml N-803 (group B, 10 subjects). In study period 1, each subject received a single s.c. dose of 10 µg/kg N-803 on day 1, and assessments were made daily through day 9. A rest period of ≥6 d followed the conclusion of study period 1. In study period 2, each subject received a single s.c. dose of 20 µg/kg N-803 on day 1, and assessments were again made through day 9. Subjects returned on day 15 of study period 2 for an end-of-study (EOS) visit.

Blood samples for PK analysis of N-803 serum levels were collected prior to each drug administration and at 1, 4, 24, 48, 72, 96, 120, 144, 168, and 192 h after dosing. Blood samples for hematologic assessments, analysis of immune cell levels, cytokine levels, and immunogenicity testing were collected prior to each N-803 administration and at 24, 48, 72, 96, 120, 144, 168, and 192 h after dosing. In addition, an EOS blood sample was drawn on day 15 of study period 2.

All 20 subjects received one or more doses of N-803 and comprised the evaluable population for PK, safety, immune cells, cytokine levels, and immunogenicity. Serum was frozen and N-803 concentrations were determined using an assay developed at Altor BioScience (Miramar, FL) using a human IL-15–specific ELISA kit (R&D Systems, Minneapolis, MN). N-803 serum concentrations versus time data were analyzed by noncompartmental analysis with Phoenix WinNonlin version 8.0 (Certara, Princeton, NJ). The PK parameters calculated for serum N-803 levels were the maximum observed concentration (C_max), the time to the observed maximum concentration (T_max), t_1/2, the area under the plasma concentration curve (AUC_0-t), the extravascular volume of distribution (Vz/F), and the extravascular clearance (CL/F).

Safety was assessed by AEs, clinical laboratory tests, and subject diaries of N-803 injection site reactions. AEs were assessed for severity using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE version 4.03), and attribution to study drug was determined by the principal investigator.

Blood samples for PBMC analysis were collected in 6-ml purple top tubes (at the Quotient Sciences trial site, Miami, FL). To isolate PBMCs, whole blood was diluted, layered over sterile density gradient media, and centrifuged. The plasma layer was discarded, and the opaque interface was collected and aliquots of PBMCs were frozen for batch analysis. For flow cytometric analysis, after thawing, viable cells were identified (Live/Dead fixable blue dead cell stain; Invitrogen, Carlsbad, CA) and ranged from ∼5 to 30% per sample. Cells were simultaneously labeled with fluorescently conjugated Abs to the cell-surface molecules CD8 (SK1) and CD4 (RPTA-4; both from eBioscience); CD45RO (UCHL1) and CD56 (HCD56; both from BioLegend); and CD3 (SK7) and CD16 (3G8; both from BD Biosciences). Cells were permeabilized and labeled with fluorescently conjugated Abs to the intracellular molecules Ki67 (20Raj1; eBioscience) and GzB (GB11; BD Biosciences). Cell acquisition was performed using an LSR II flow cytometer (BD Biosciences, San Jose, CA), and data were analyzed with FlowJo software (BD Biosciences). We did not observe evidence of sampling bias as a result of reduced viability, as a similar phenotype and frequency of immune cell populations were observed independent of whether the analysis was performed on viability dye positive or negative cells (Supplemental Fig. 1). Except for Supplemental Fig. 1, all data are gated on viable cells.

Cytokine levels were analyzed from serum stored at −80°C. The cytokines IL-2, IL-4, IL-6, IL-10, TNF-α, and IFN-γ concentrations were determined using a commercial cytometric bead array kit (Th1/Th2 cytokine kit II; BD Biosciences).

Mass cytometry was performed with protocols and reagents as previously described (21, 22). Briefly, Live/Dead staining was performed with cisplatin and samples were combined using barcoding. Surface and intracellular proteins were stained with Abs, and cells were incubated with an iridium intercalator to label cellular DNA. Mass cytometry acquisition was performed on a CyTOF2.1 (Helios) mass cytometer. After acquisition, data were normalized by using the bead standard and the executable MATLAB normalizer application (23), and dead cells and beads were removed and debarcoded using Boolean gating using FlowJo software. All analyses on CyTOF data were performed after arcsinh transformation (with cofactor equal to 5) of marker expression. The bioinformatics workflow was based on FlowSOM (24) clustering and t-SNE (25) visualization with manual annotation using heat maps of normalized marker expression and has been published as a workflow article (26).

An ELISA developed by Altor BioScience using N-803 as the capture reagent and HRP-conjugated N-803 was used to detect serum anti-N-803 Abs. Samples were assayed in triplicate in a dilution series (1:100–1:6400) and considered positive for anti–N-803 Abs when the OD of the postdose sample was >2-fold the OD₄₅₀ of the predose sample (average uncorrected measures).

Statistical analyses were conducted using R version 3.6.0 (R Core Team 2019), SAS 9.4 (SAS Institute, Cary, NC), and GraphPad Prism 8 (GraphPad Software, San Diego CA). PK analysis excluded data collected from a single subject in study period 2 with AUC_0-t >10-fold higher than the mean value, possibly due to an experimental error or inadvertent injection into a superficial vein or lymphatic vessel. Exclusion of these data did not alter conclusions drawn from any statistical analyses. N-803 serum concentrations and PK parameters with all data included are shown in Supplemental Fig. 2.

To determine whether average patient N-803 serum concentration levels differed significantly by N-803 formulation (1.0 versus 2.0 mg/ml) or by dose (10 versus 20 µg/kg), data were analyzed using longitudinal mixed effects regression models, with N-803 serum concentration levels below the limit of quantification (23.4 pg/ml) treated as left-censored (27). Serum N-803 concentrations were log-transformed to induce approximate normality and stabilize variance.

Models investigating differences in serum N-803 concentrations for the two drug formulations (1.0 or 2.0 mg/ml) were constructed separately for study periods 1 and 2. Time (measured as a continuous variable), drug formulation, and their interaction were included in the model as independent variables.

Models investigating serum N-803 concentrations and cytokine levels versus the limit of quantification as a function of N-803 dose were constructed using a similar approach. We used the limit of quantification (as opposed to baseline values) as a comparator for each outcome, as most N-803 concentrations and cytokine levels at baseline were below the limit of quantification. Comparisons between drug doses at specific time points were performed using model-based linear contrasts. A similar model with only main effects for dose and time was used to estimate an average dose effect. Cytokine data were log-transformed to induce approximate normality and stabilize variance.

Immune cell proliferation data were analyzed using linear mixed effects regression models, with the determination of a significant dose effect investigated using the same approach as described for serum N-803 levels. Data were transformed via the logit transformation to satisfy normality assumptions. For all analyses, time was transformed using fractional polynomials to appropriately model significant non-linearity (28).

PK parameters were analyzed using two-way ANOVA (factors of N-803 concentration and dose). Injection site reaction symptoms were analyzed using paired t tests.

This study was conducted in accordance with the Declaration of Helsinki ethical principles and the International Conference on Harmonization Good Clinical Practice Guidelines. A local Institutional Review Board approved the study protocol, and all subjects provided written informed consent prior to treatment.

Subject demographics are shown in Table I. A total of 20 healthy volunteers were enrolled in the study and received s.c. injection of 10 µg/kg N-803 in study period 1; 14 subjects went on to receive s.c. injection of 20 µg/kg N-803 in study period 2 and complete the study (Fig. 1). Subjects were randomized 1:1 to receive N-803 formulated at 1.0 mg/ml (group A; 10 subjects) or 2.0 mg/ml (group B; 10 subjects). Six subjects (three subjects each in groups A and B) discontinued participation in the study after receiving N-803 in study period 1. For subjects in group A, reasons for discontinuation included an AE (peripheral neuropathy), protocol deviation (positive drug test), and withdrawal by subject. For subjects in group B, reasons for discontinuation included an AE (urinary tract infection) and withdrawal by subject (two subjects).

All subjects enrolled experienced at least one AE, all of which were mild or moderate (CTCAE grade 1 or 2); none of the subjects experienced treatment-emergent grade ≥3 AEs or serious AEs. The most common (≥50% of subjects) treatment-emergent AEs included injection site reactions, chills, pyrexia, abdominal pain, headache, and feelings of body temperature change (Table II). The incidence of AEs was similar after administration of 1.0 versus 2.0 mg/ml formulations of N-803, and all treatment-related AEs resolved with supportive care.

Two subjects (one in group A and one in group B) experienced AEs that resulted in study discontinuation. The subject in group A experienced grade 1 peripheral neuropathy considered related to the study drug that persisted for 18 d, but ultimately resolved without treatment. The subject in group B experienced grade 2 urinary tract infection (considered unrelated to the study drug) that resolved after treatment.

All subjects experienced injection site reactions consisting of circular raised welts characterized by redness, hardness, and pain or itching. Symptoms usually resolved during the course of ∼1 wk with topical steroid application, and in the absence of any systemic treatment.

Injection site reaction symptoms were generally similar after administration of 1.0 versus 2.0 mg/ml formulations. However, reaction symptoms typically occurred with significantly shorter onset latency and persisted longer after administration of 20 versus 10 µg/kg (Supplemental Fig. 3).

The PK of N-803 upon s.c. administration were represented by a typical absorption profile followed by biphasic elimination, characterized by an initial rapid tissue distribution phase accompanied by a slower elimination phase (Fig. 2A, 2B). N-803 serum levels did not differ significantly between subjects who received the 1.0 versus 2.0 mg/ml drug formulation in either study period. Thus, means reported below report the average across both formulations of N-803.

After administration of 10 µg/kg in study period 1, serum N-803 concentrations peaked with a median T_max of 4 h (range of 4–48 h), reaching a mean peak concentration (C_max) of 1110 ± 900 pg/ml (Fig. 2A). Mean N-803 levels remained elevated above the limit of quantification for 72 h after injection (p < 0.001 at each time point). After administration of 20 µg/kg N-803 in study period 2 (Fig. 2B), N-803 serum concentrations were significantly increased relative to study period 1 (p = 0.0016). Serum N-803 concentrations peaked with a median T_max of 4 h (range of 4–24 h), reaching a C_max of 1680 ± 1420 pg/ml. Mean serum levels of N-803 remained significantly elevated for 72 h after drug injection (p < 0.01).

There was an ∼1.5-fold increase in C_max levels after administration of 20 µg/kg in study period 2 when compared with 10 µg/kg in study period 1 (Fig. 2C, Table III; p = 0.098), supporting a dose-dependent increase in N-803 exposure. Similarly, mean AUC_0-t values showed a significant 1.6-fold increase after 20 versus 10 µg/kg (Fig. 2D; p = 0.005).

Mean N-803 half-life was similar after administration of 10 and 20 µg/kg, ranging from 20.0 to 20.7 h, respectively (Fig. 2E). Neither CL/F (Fig. 2F) nor V_z/F (Fig. 2G) showed concentration- or dose-dependent effects. Notably, mean V_z/F estimates were 10,000 and 13800 ml/kg in study periods 1 and 2m respectively. These values are notably larger than human serum volume (43 ml/kg), suggesting some distribution of N-803 into tissues.

N-803 administration caused a transient decline in subject absolute lymphocyte count (ALC; (Fig. 3A). In study period 1, this decline was followed by recovery and then overshoot of the baseline value, so that ALC levels were significantly elevated on days 7 and 8 relative to baseline levels on day 1 (p = 0.013 and p = 0.0002, respectively).

This elevation in ALC persisted through day 1 of study period 2 (Fig. 3A, red; p = 0.0011 versus day 1 of study period 1) despite the ≥6-d rest period between study periods 1 and 2. Mean ALC levels were higher throughout study period 2 versus study period 1 (p = 0.0019), and the EOS ALC (black symbol) was significantly higher than baseline values measured in study period 1 (p < 0.0001) and study period 2 (p = 0.0026).

Flow cytometry was used to detect immune cell frequency and proliferation (Fig. 3B). NK cell, CD8⁺ T cell, and CD4⁺ T cell frequency all showed a transient decline after N-803 administration, paralleling the dip seen in the ALC (Fig. 3C, 3D). Notably, whereas all three lymphocyte subtypes showed a transient decline in cell number, only NK cell number recovered to and then overshot baseline levels.

In study period 1, NK cell frequency was significantly increased on days 4–8 relative to baseline (all p < 0.05). This elevation persisted throughout the ≥6-d rest period after study period 1, and it was evident on day 1 of study period 2 prior to the second N-803 injection (Fig. 3D; p = 0.0036 versus day 1 of study period 1). Mean NK cells numbers were significantly elevated in study period 2 relative to study period 1 (p = 0.0076). NK cell numbers in study period 2 again declined relative to baseline value on days 1 and 2 before significantly increasing on days 5–8 (Fig. 3D; p < 0.05). NK cell levels measured at the EOS visit were significantly increased relative to the baseline value measured in study period 1 (p = 0.008). Neither the CD8⁺ or CD4⁺ T cell population increased above baseline levels (Fig. 3C–E). (Fig. 3E indicates the fold change in each cell type across each study period, and it shows that the increase in cell number was confined to NK cells.

Administration of N-803 in both study periods induced robust proliferation (as indicated by Ki67 staining) of NK cells and CD8⁺ T cells, with modest increases in proliferation of CD4⁺ T cells (Fig. 3F–H). For all cell types, immune cell proliferation peaked 4–6 d after drug administration. The magnitude of this effect (measured as percentage of Ki67⁺ cells) was most robust for NK cells (peak of 73% of cells Ki67⁺ in study period 1), followed by CD8⁺ T cells (32%) and CD4⁺ T cells (14%). For all three immune cell types, the magnitude of the proliferation induced in study period 2 was modestly but significantly decreased relative to that occurring in study period 1 (Fig. 3F, 3G, all p < 0.001).

GzB importantly mediates cytotoxicity of NK cells and CD8⁺ T cells. N-803 administration resulted in expansion of Ki67⁺GzB⁺ cells for both of these immune cell subtypes. This subset of cells was modestly increased in CD4⁺ T cells after N-803 administration (Supplemental Fig. 4A–C). We did not observe notable differences in Ki67 upregulation between CD8⁺ or CD4⁺ T cell subsets gated based on CD45RO expression (Supplemental Fig. 4D).

We also characterized immune cell changes in three volunteers during study period 1 using mass cytometry (Fig. 4). These results largely paralleled findings from flow cytometry, but also showed that Ki67 was induced in γδ T cells with roughly similar kinetics to those seen in NK and CD8⁺ T cells, consistent with the known ability of IL-15 to induce expansion of γδ T cells (8).

N-803 administration significantly increased levels of IFN-γ, IL-6, and IL-10 cytokines after administration of both 10 and 20 µg/kg (Fig. 5; all p < 0.05). These cytokines peaked 3–4 d after N-803 administration. For both IFN-γ and IL-10, cytokine levels in study period 2 were significantly higher than levels in study period 1 (both p < 0.05). N-803 administration did not induce statistically significant increases in serum concentrations of TNF-α, IL-4, or IL-2 (data not shown).

Of the 10 subjects in group A who received N-803 formulated at 1.0 mg/ml, 7 met criteria for evaluation of anti–N-803 Abs (samples drawn in both study periods). None of these subjects developed detectable levels of anti–N-803 Abs. Of the 10 subjects in group B who received N-803 formulated at 2.0 mg/ml, 8 met criteria for evaluation. A single subject had measurable levels of anti–N-803 Abs, with a titer of 104 detected in blood samples collected at the EOS visit (data not shown).

The primary objective of this study was to establish the PK profile and safety of serum N-803 levels after s.c. administration to healthy volunteers at dose levels of 10 and 20 µg/kg and drug formulations of 1.0 and 2.0 mg/ml. N-803 was well tolerated and AEs were exclusively mild to moderate (grade 1–2). None of the subjects enrolled in the current study experienced grade ≥3 AEs or serious AEs. Our data strengthen results from previous clinical studies, which provide evidence that s.c. administration of N-803 is well tolerated and that AEs can typically be managed with supportive care (19–21, 29).

The most common AE experienced was injection site reaction. These reactions were characterized by reddening and swelling, symptoms previously associated with injection site infiltration with CD68⁺ macrophages and CD3⁺ lymphocytes (19). Injection site reactions were managed by topical steroid application and resolved over the course of ∼1 wk following N-803 injection, and they did not require systemic treatment.

The PK profile of serum N-803 was characterized by an initial absorption phase, in which drug concentrations peaked 4 h (range of 4–48 h) after administration. A more gradual decline in N-803 levels followed, with a half-life of 20.0–20.7 h. Thus, drug levels remained elevated above baseline levels for ∼3 d after a single s.c. injection. Serum N-803 levels increased dose-dependently. C_max and AUC_0-t were increased 1.5- and 1.6-fold, respectively, after administration of 10 versus 20 µg/kg. No significant differences in PK profiles were observed after administration of 1.0 versus 2.0 mg/ml formulations.

The PK profile of N-803 observed in the current study is consistent with results from clinical studies that have investigated the PK of N-803 after s.c. administration of N-803 to cancer patients (Table IV) (19–21, 29). Findings from these studies converge with the current results in showing that after s.c. administration of N-803 doses of 10–20 µg/kg, serum levels generally peak within 24 h, reach C_max concentrations of ∼1000–1700 pg/ml, and persist with a t_1/2 of ∼24 h. Notably, injection site reaction and flu-like symptoms are consistently reported as the most common AEs.

The PK profile of N-803 after s.c. administration contrasts with that observed after i.v. injection. The i.v. administration transiently produces a peak in N-803 serum concentration that is higher (with a C_max ∼100-fold higher) and more fleeting (t_1/2 ∼10-fold lower) than that produced by s.c. injection (19, 20). Because high N-803 levels after i.v. administration have been associated with fever and tachycardia, s.c. delivery is the preferred route of systemic delivery in ongoing and planned clinical studies.

The PK profile of N-803 notably contrasts with that of rhIL-15, which has a short mean t_1/2 of <3 h after i.v. administration (8, 30). This may be in part due to increased partitioning of N-803 into tissue versus blood relative to rhIL-15, which is supported by volume of distribution values ∼2.3-fold higher than those reported for rhIL-15 (8). Regardless of the specific mechanism, our data suggest that s.c. administered N-803 persists in circulation ∼10-fold longer than i.v. rhIL-15, allowing for increased activation and proliferation of target NK and CD8⁺ T cells.

N-803 administration produced a transient decline in lymphocyte levels. Decreased lymphocyte levels persisting 24–48 h have consistently been observed after IL-2, rhIL-15, and N-803 administration (8, 19–21, 29–31). Because the onset of recovery of lymphocyte levels precedes the induction of proliferation, these changes may reflect redistribution of lymphocyte levels between blood and tissue, rather than acute changes in total cell levels (8, 21).

The current study shows that the transient decrease in lymphocyte levels was followed by a selective and persistent increase in NK cell number. A single s.c. administration of N-803 caused increases in NK cell number that persisted for at least 15 d (i.e., the 9-d interval of study period 1, followed by the ≥6-d rest period).

N-803 administration induced proliferation of NK cells and CD8⁺ T cells, and, more modestly, CD4⁺ T cells, consistent with previous clinical results (8, 19–21, 29, 30). Also consistent with previous studies of N-803, we observed increased expression of GzB in CD8⁺ T cells and NK cells, suggesting increased cytotoxic potential (32, 33). N-803–induced upregulation of the NK cell–expressed activating receptors NKG2D and NKp30, without changes in expression of the inhibitor receptors KIR or NKG2A, have also been reported in previous studies (20). Increased NK cell levels may be of particular value for cancer patients with loss of tumoral MHC class I expression (34, 35). Our findings complement a substantial literature demonstrating that N-803 administration can increase the number, cytotoxicity, and activation of immune cells contributing to patient immune antitumor responses.

N-803 administration increased circulating levels of the inflammatory cytokines IFN-γ, IL-6, and IL-10. This increase appeared to be dose-dependent for IFN-γ and IL-10, as the magnitude of cytokine induction was significantly higher in study period 2 versus study period 1. These findings of elevated inflammatory cytokines are consistent with other studies using rIL-2 and rIL-15 (8, 36). In parallel with cytokine changes, the magnitude of injection site reaction symptoms was increased and the latency of symptom onset was reduced in study period 2 versus study period 1.

A single subject (of 20 treated with N-803; 5%) developed detectable anti–N-803 Abs at the EOS visit. Previous studies have reported similar findings, with detectable immunogenicity in 0 of 24 patients (19) and 1 of 33 patients (20), consistent with low overall levels of immunogenicity for the protein. Anti–N-803 Abs occurred more frequently in the phase 1b study of N-803 administered in combination with nivolumab in NSCLSC patients (21). In that study, 7 of 21 patients (33%) developed Abs against N-803. However, immunogenicity had no apparent effects on N-803 safety or efficacy. Four of the seven patients with anti–N-803 Abs responded to treatment (all partial responses), and no changes in the magnitude or frequency of AEs were noted in this subset of patients.

The current study enrolled healthy volunteers, and the efficacy of N-803 was not assessed. However, previous studies suggest that N-803 harbors significant promise in treating patients with a variety of cancer indications. In a phase 1 study of N-803 treatment for patients with hematologic malignancies who had previously relapsed after allogeneic hematopoietic cell transplantation, clinical responses were observed in 19% of evaluable patients (20). This included one patient with a complete response lasting 7 mo. Wrangle et al. (21) showed that 29% of patients with NSCLC treated with a combination therapy of N-803 and nivolumab experienced a clinical response (all partial responses). Notably, this group included patients that had relapsed after previous treatment with a checkpoint inhibitor. These data suggest that N-803 holds promise not only in stimulating patients’ endogenous immune response, but also in complementing, and possibly reinvigorating, the antitumor effects of other immunomodulatory therapies.

In conclusion, we demonstrate that s.c. administration of N-803 results in prolonged elevation of drug serum concentrations, accompanied by robust proliferation of cytotoxic lymphocytes. Ongoing studies continue to investigate the oncotherapeutic promise of N-803 in combination with other immunotherapeutics in a variety of settings, including intravesicular administration in combination with bacillus Calmette-Guérin for non-muscle invasive bladder cancer (NCT03022825 and NCT02138734) and in combination with a checkpoint inhibitor in cancer patients (NCT03520686 and NCT03228667). Our PK and immune correlative data will help provide a foundation for the rational design of new clinical trials.

We thank Laura Biggs for critical review of the manuscript.

M.P.R. is a co-inventor on a patent application related to IL-2/anti–IL-2 monoclonal Ab complexes. M.D.R. has received research support from Hoffman La Rouche, Ltd. E.G.H. has equity in Eli Lilly. S.A.T., A.D.R., and J.H.L. were employees of ImmunityBio. P.S.-S. is chief executive officer of NantHealth, NantKwest, and ImmunityBio and has equity in each of the companies. J.W. has equity in Precision Genetics. The other authors have no financial conflicts of interest.