Pegylated Engineered IL2 plus Anti–PD-1 Monoclonal Antibody: The Nectar Comes from the Combination

Human IL2 cytokine was identified 40 years ago and described as a T-cell and natural killer (NK)–cell growth factor that engages the high-affinity IL2 receptor (IL2R) on T effector, NK, and T regulatory (Treg) cells (1). IL2 binds to T-cell IL2R α, β, and γ chains, which heterodimerize prior to signal transduction. Owing to its immune-stimulatory effect on conventional T and NK cells, human recombinant IL2 has been evaluated in early-phase clinical trials in patients with advanced refractory cancers. The administration of intravenous IL2 led to complete and durable responses in a subset of patients with metastatic cancers. IL2 was first approved by the FDA as aldesleukin (Proleukin) for treating metastatic renal cell carcinoma (RCC) and melanoma in 1992 and 1998, respectively. However, one of the major drawbacks that limited its broad use was its short life, which required frequent administration of high-dose levels of IL2 associated with dose-limiting toxicities, characterized by a capillary leak syndrome and multiorgan toxicity. At lower dosing, IL2 promotes the survival and function of Tregs through the engagement of the IL2R α-subunit (CD25), leading to significant amelioration of autoimmune diseases (2).

Hence, tilting the balance between effector cells and Tregs during systemic therapy with recombinant IL2 toward a better therapeutic index has been challenging in patients with cancer so far, as indicated through pioneering studies attempting to combine CTLA4 blockade and recombinant IL2 (3, 4).

Recently, the engineered IL2 bempegaldesleukin (NKTR-214; BEMPEG) has been developed with, on average, six releasable polyethylene glycol sites that preferentially lead to binding of the IL2R β-subunit (CD122) as opposed to the α-subunit (CD25). This pegylated IL2 preferentially promotes the activation of T effectors and NK cells as opposed to Tregs in preclinical tumor models (5, 6). In the first-in-human study, bempegaldesleukin monotherapy led to a significant expansion of CD4⁺ CD8⁺ and NK cells both in the tumor and in peripheral blood, along with a transient increase of Tregs in the peripheral blood, but not in the tumor (7). Although the agent alone showed modest antitumor efficacy, its potent immune activity together with a favorable safety profile supported further development of bempegaldesleukin combined with immune checkpoint blockade (ICB) therapy.

In this issue of Cancer Discovery, Diab and colleagues report on the phase I dose-escalation study evaluating the combination of bempegaldesleukin with the anti–PD-1 mAb nivolumab in patients with selected solid tumors (8). In all, 38 patients with immunotherapy-naïve advanced melanoma (n = 11), RCC (n = 22), and non–small cell lung carcinoma (NSCLC; n = 5) were enrolled in this phase I trial. The authors demonstrated that the CD122-preferential IL2 pathway agonist bempegaldesleukin and nivolumab can be safely combined, with a recommended phase II dose (RP2D) of bempegaldesleukin 0.006 mg/kg plus nivolumab 360 mg every 3 weeks. With a median duration of exposure of 13.3 months [interquartile range (IQR), 3.7–16.9] and a median of 15 cycles (IQR, 6.0–22.0) across all dose cohorts, bempegaldesleukin plus nivolumab was well tolerated, with grade ≥3 treatment-related adverse events (AE) occurring in 8 patients (21.1%). Specifically, the most common treatment-related AEs were flu-like symptoms (86.8%), rash (78.9%), fatigue (73.7%), and pruritus (52.6%). Of note, adequate hydration protocol mitigated the development of hypotension associated with bempegaldesleukin administration.

With regard to efficacy, the combination of bempegaldesleukin plus nivolumab demonstrated durable clinical activity across all tumor types and dose cohorts with a median duration follow-up of 18 months. In all, total objective response rate (ORR) was 59.5% (22/37), with 18.9% (7/37) complete responses (CR); tumor-specific ORRs were 63.6% in first-line (1L) melanoma (including four CRs, even in patients presenting elevated basal lactate dehydrogenase and liver metastases), 71.4% in 1L RCC, 28.6% in second-line (2L) immuno-oncology (IO)–naïve RCC, and 60.0% in IO-naïve NSCLC. Among the 27 patients eligible for the efficacy analysis who received the RP2D of bempegaldesleukin plus nivolumab, the ORR was 66.7% (95% confidence interval, 44.7%–84.4%; 16/24), and disease control rate was 83.3% (20/24).

Although head-to-head comparisons with nivolumab monotherapy cannot be formally made, especially with the low number of patients in this phase I trial, these findings suggest that IL2-induced immune activation with bempegaldesleukin could enhance the antitumor efficacy mediated by nivolumab monotherapy. However, we should note that the majority (70%) of patients included in this phase I trial received bempegaldesleukin plus nivolumab as a first-line therapy. This might partially explain the improved ORR by comparison with nivolumab monotherapy, as illustrated by the lower ORR in the 2L RCC cohort (28.6%) compared with the 1L RCC cohort (71.4%).

However, it is of utmost importance to note that the antitumor responses upon bempegaldesleukin plus nivolumab administration continued to deepen over time with additional cycles of combination treatment. Notably, 5 of the 7 patients who achieved a CR did so on or after the fourth post-baseline scan (≥8 treatment cycles). This deepening of response over time could possibly be due to the gradual release of polyethylene glycol sites, and bempegaldesleukin's ability to continuously mobilize or expand antigen-specific lymphocytes.

Mechanistically, the combination therapy induced powerful immunostimulatory effects in blood and tumor beds, as illustrated in Fig. 1. In peripheral blood, the evidence for tachyphylaxis reaction resulted in bempegaldesleukin-induced lymphocyte mobilization at each treatment cycle with similar magnitude of mobilization in the bempegaldesleukin 0.006 and 0.009 mg/kg dose cohorts regardless of dosing schedule. Flow cytometry analysis of matched blood samples identified a significant increase in proliferating (Ki-67⁺) CD4⁺, CD8⁺, and NK cells at day 8 of treatment compared with baseline. In addition, the activation markers CTLA4, TIGIT, and LAG3 were significantly upregulated in proliferating (Ki-67⁺) T cells compared with nonproliferating (Ki-67⁻) T cells upon therapy. Although the percentage of CD4⁺ CD25^hi FOXP3⁺ Tregs also increased in peripheral blood at day 8, likely limiting immune-related AEs, Tregs failed to accumulate in tumors, allowing a marked elevation of the CD8/Treg ratio in lesions.

To investigate the effects of bempegaldesleukin plus nivolumab on the tumor microenvironment (TME), longitudinal tumor biopsies were obtained at baseline and on treatment from a significant percentage of patients enrolled in this dose-escalation study. First, flow cytometry analysis on 12 matched pretreatment and on-treatment biopsies identified a significant increase in absolute number of intratumoral CD8⁺ T cells. Second, T-cell receptor sequencing analysis revealed increased T-cell infiltration and a trend toward increased T-cell clonality in responders (n = 5), but not in nonresponders (n = 7). Third, changes in gene expression were analyzed prior to (n = 19) and on therapy (n = 9) with the NanoString Human PanCancer Immune Profiling panelacross 770 genes. Strikingly, genes associated with T-cell signaling (CD3G, CD3E, CD3D, CD8A, and CD8B), T-cell activation(ICOS, PD-1, CTLA4, TIGIT, and LAG3), Th1 differentiation (CXCR3, INFg, EOMES, and TBX1), and cytotoxic function (PRF1, GZMA, GZMB, and GZMK) significantly increased on treatment. This, along with a significant decrease in TGFβ1 and TGFβ2 isoforms in the TME, may have improved ICB efficacy by reducing primary resistance to anti–PD-1 monotherapy. Although selective depletion of tumor-associated Tregs has been a proposed mechanism of action of bempegaldesleukin plus nivolumab in preclinical models (6), no data support this assumption in vivo because increased transcription of CTLA4 (8.8-fold change; P = 0.005) and FOXP3 (6-fold change; P = 0.069) Treg-associated genes has been observed upon treatment.

Overall, this dose-escalation phase I study demonstrated that bempegaldesleukin plus nivolumab rapidly induced NK- and cytotoxic T-lymphocyte–associated tumor regressions, independent of baseline PD-L1 expression or tumor-infiltrating lymphocyte densities. In addition, this combination did not appear to exacerbate the incidence of AEs typically associated with PD-1 blockade, which suggests a lack of overlapping toxicities. Building on these findings, additional cohorts in other tumor types including breast, urothelial, and colorectal carcinomas have opened in the PIVOT-02 trial. Of utmost interest, the bempegaldesleukin plus nivolumab combination is also being evaluated in separate phase II and III pivotal studies, particularly in the neoadjuvant setting in invasive bladder cancer before radical cystectomy (NCT04209114) and in the adjuvant setting after complete resection in high-risk melanoma (NCT04410445). Proper sequencing of the combination of immune checkpoint blockade and cytokine therapy has been discussed in retrospective studies (9).

Finally, we surmise that these findings support the exploration of other combinatorial approaches targeting alternative inhibitory receptors or local immunomodulators [such as oncolytic viruses or the TLR7/8 agonist NKT-262 (NCT03435640)], and may also synergize with radiotherapy. Likewise, the use of an anti-CD25 antibody with enhanced binding to activating FcγRs alone or combined with anti–PD-1 mAbs also led to effective depletion of tumor-infiltrating Tregs, increased effector to Treg cell ratios, and improved control of established tumors (10). Altogether, this report reinforces the relevance of repurposing cytokine-based therapies within the armamentarium of new-generation immune stimulatory drugs to achieve deep and durable responses sparing self-tissues.

L. Zitvogel reports personal fees from Transgene (board of administration) and Lytix Biopharma (research contract), grants from Kaleido (research contract), Innovate Pharma (research contract), GlaxoSmithKline (research contract), and MSD (research contract), and is a founder of the biotech EverImmune, all outside the submitted work. A. Marabelle reports grants, personal fees, and non-financial support from AstraZeneca and Roche, grants and non-financial support from BMS, grants fromBoehringer Ingelheim, Janssen Cilag, Novartis, Pfizer, Merus, Transgene, and Fondation MSD Avenir, grants and personal fees from Sanofi, personal fees and non-financial support from Merck (MSD), personal fees from Pierre Fabre, Onxeo, EISAI, Bayer, Genticel, Rigontec, Daiichi Sankyo, Imaxio, Sanofi/BioNTech, Molecular Partners, Pillar Partners, BPI, Faron, Merck Serono, Servier, and Amgen outside the submitted work; has a patent for 62/351,054 issued; is co-founder and shareholder of Pegascy SAS & PEGA-1 SAS; is a member of Clinical Trial Steering Committee: NCT02528357 (GSK) and NCT03334617 (AZ); is a member of Data Safety and Monitoring Board: NCT02423863 (sponsor: Oncovir) and NCT03818685 (sponsor: Centre Leéon Beérard); and has participated in Scientific Advisory Boards: Merck Serono, eTheRNA, Lytix Pharma, Kyowa Kirin Pharma, Novartis, BMS, Symphogen, Genmab, Amgen, Biothera, Nektar, Tesaro/GSK, Oncosec, Pfizer, Seattle Genetics, AstraZeneca/Medimmune, Servier, Gritstone, Molecular Partners, Bayer, Partner Therapeutics, Sanofi, Pierre Fabre, RedX pharma, OSE Immunotherapeutics, Medicxi, HiFiBio, IMCheck, and MSD. No potential conflicts of interest were disclosed by the other author.