Bempegaldesleukin (BEMPEG; NKTR‐214) efficacy as a single agent and in combination with checkpoint‐inhibitor therapy in mouse models of osteosarcoma

Abstract Survival of patients with relapsed/refractory osteosarcoma has not improved in the last 30 years. Several immunotherapeutic approaches have shown benefit in murine osteosarcoma models, including the anti‐programmed death‐1 (anti‐PD‐1) and anti‐cytotoxic T‐lymphocyte antigen‐4 (anti‐CTLA‐4) immune checkpoint inhibitors. Treatment with the T‐cell growth factor interleukin‐2 (IL‐2) has shown some clinical benefit but has limitations due to poor tolerability. Therefore, we evaluated the efficacy of bempegaldesleukin (BEMPEG; NKTR‐214), a first‐in‐class CD122‐preferential IL‐2 pathway agonist, alone and in combination with anti‐PD‐1 or anti‐CTLA‐4 immune checkpoint inhibitors in metastatic and orthotopic murine models of osteosarcoma. Treatment with BEMPEG delayed tumor growth and increased overall survival of mice with K7M2‐WT osteosarcoma pulmonary metastases. BEMPEG also inhibited primary tumor growth and metastatic relapse in lungs and bone in the K7M3 orthotopic osteosarcoma mouse model. In addition, it enhanced therapeutic activity of anti‐CTLA‐4 and anti‐PD‐1 checkpoint blockade in the DLM8 subcutaneous murine osteosarcoma model. Finally, BEMPEG strongly increased accumulation of intratumoral effector T cells and natural killer cells, but not T‐regulatory cells, resulting in improved effector:inhibitory cell ratios. Collectively, these data in multiple murine models of osteosarcoma provide a path toward clinical evaluation of BEMPEG‐based regimens in human osteosarcoma.


| INTRODUCTION
Osteosarcoma is the most common primary malignant bone tumor. 1 Primary osteosarcoma occurs predominantly in adolescents and young adults under the age of 25 years, while it generally presents as a secondary malignancy in adults over 65 years of age. 2,3 Along with surgical resection, combination chemotherapy plays a central role in the management of osteosarcoma; however, a lack of new therapies and the inability to optimize existing ones has prevented significant improvement in overall survival for the last 30 years. Moreover, treated patients remain at high risk for developing pulmonary metastases, the leading cause of death from osteosarcoma. 4 Thus, relapsed/ refractory osteosarcoma remains challenging to treat and constitutes an unmet clinical need. [5][6][7][8][9][10] Several novel immunotherapeutic approaches have been evaluated for the treatment of sarcomas, including osteosarcoma, such as monoclonal antibodies directed against the immune checkpoint protein targets cytotoxic T-lymphocyte antigen-4 (CTLA-4) and programmed death-1 (PD-1). 1,[11][12][13][14] These immune checkpoint inhibitors (ICI) modulate the adaptive immune response by preventing the downregulation of T-cell function induced by immune checkpoints. 15 In recent years, ICIs have induced high response rates, particularly in "hot tumors" that have high immune infiltrates, including melanoma and lung cancer. 16 In "cold tumors" with low infiltrating T cells, such as osteosarcoma, the benefit of ICIs is less clear. 16 In preclinical murine models of osteosarcoma, when used alone or in combination with other therapies, antibodies against CTLA-4 or PD-1 have shown benefit by upregulating the antitumor activity of cytotoxic T cells. [11][12][13] In a xenograft osteosarcoma mouse model, anti-PD-1 therapy resulted in regression of pulmonary metastases; such efficacy was mediated by local M1-polarized macrophages. 4 However, in clinical studies of osteosarcoma, ICIs have demonstrated limited clinical activity. [17][18][19] Nevertheless, a recent immunogenomic study using multiplatform profiling of osteosarcoma specimens from 48 pediatric and adult patients with primary, relapsed and metastatic osteosarcoma confirmed multiple immunosuppressive features of osteosarcoma, suggesting potential immunotherapeutic opportunities. 19 Thus, it is thought that novel therapeutic approaches that stimulate T cells may synergize with ICIs to induce antitumor responses.
High-dose systemic recombinant human interleukin-2 (rhIL-2) is an immunotherapeutic approach with efficacy in melanoma and renal cell carcinoma (RCC), likely through stimulating the proliferation and antitumor activity of T cells and natural killer (NK) cells. Although some efficacy has also been observed in a limited group of pediatric patients with sarcoma, high toxicity rates, poor pharmacokinetics and the pleiotropic effects of systemic rhIL-2 have precluded its further development as a viable therapeutic option. [20][21][22] To improve efficacy while minimizing toxicity, alternate formulations of IL-2 have been developed, such as aerosols that can directly target immune cells in the lungs. Preclinical studies of aerosolized rhIL-2, both as a single agent and in combination with adoptive transfer of NK and/or T cells, have demonstrated efficacy in osteosarcoma. [23][24][25] Recently, novel forms of IL-2 have been engineered to improve efficacy and decrease toxicity relative to rhIL-2. Bempegaldesleukin (also known as BEMPEG or NKTR-214) is a CD122-preferential IL-2-pathway agonist that is being investigated for its ability to stimulate an antitumor response via induction of T-effector cells. 26 It consists of an rhIL-2 moiety conjugated to an average of six releasable polyethylene glycol (PEG) chains. 27 When BEMPEG is administered systemically, the PEG chains are progressively released from rhIL-2, unmasking its activity as an IL-2 receptor agonist with preferential IL-2 receptor beta (CD122) binding in a pharmacokinetically predictable, controlled and sustained manner. [27][28][29] Compared to native IL-2, the placement of the PEG chains directs BEMPEG to preferentially bind the heterodimeric IL-2 receptor beta gamma complex (IL-2Rβγ; CD122/CD132), most frequently expressed on T-effector cells, over the heterotrimeric IL-2Rαβγ complex, typically expressed on Tregs. [27][28][29] BEMPEG favors activation and expansion of T-effector cells and NK cells, without expansion of unwanted regulatory CD4 + T cells (Treg) in the tumor tissue. [27][28][29] Mechanistically, the lack of Treg expansion observed in the tumor microenvironment is mediated by CD8 + T-cell-associated cytokines, including IFNγ and TNFα, which are released in response to cognate tumor antigen recognition in the tumor tissue. 28 BEMPEG treatment has resulted in stronger antitumor activity with better tolerability than that with rhIL-2 in a variety of mouse models, when given either alone or in combination with anti-CTLA-4, anti-PD-1, adoptive T-cell therapy or cancer vaccines. [27][28][29][30] Early clinical data revealed that BEMPEG was well tolerated and strongly increased numbers of peripheral and intratumoral T cells and NK cells, and promoted a favorable intratumoral ratio of effector T cells to Tregs. 26

| K7M3 primary tibial osteosarcoma mouse model
BALB/c mice underwent two intratibial inoculations of 1 × 10 5 cultured K7M3 cells passaged no more than four times. Prior to inoculation, the left rear leg was cleaned with iodine and 70% ethanol.
Tumor cell suspension was aspirated into a 1 mL disposable syringe fitted with a 27-gauge needle. Ten microliters of the single-cell suspension containing 1 × 10 5 cells was injected twice. Injections were given 24 hours apart and to the same leg following the same procedure. Inoculated mice were assigned to treatment groups of 10 animals per group and treated with vehicle or BEMPEG (0.8 mg/kg intravenously on Days 11, 20, 29, 38 and 47 after inoculation). Treatment efficacy against the primary tumor was evaluated by weekly X-rays.
All X-ray image analyses were performed using National Institutes of Health's ImageJ software (available at http://www.rsb.info.nih.gov/ij).
For assessing tumor volume, the diameter-based calculation was computed by measuring the greatest longitudinal diameter (length) and the greatest transverse diameter (width) of the tumors. Diameterbased measurements were determined using the modified ellipsoidal formula 34,35 to calculate diameter-based volume (V) in mm 3 Animals underwent amputation 33 days after inoculation when primary tumor reached 1.5 cm 3 . Tibial tumors were evaluated by histology. All mice were sacrificed 45 to 50 days (7 weeks) after initial K7M3 inoculation, at which time relapse and metastasis were assessed by histology of the femur and lungs. Histological response was determined as follows: measurements were carried out using a microscope equipped with a stage clip and an eyepiece graticule.
Areas of tumor necrosis were measured at ×100 magnification, whereas areas of viable tumor were measured at ×40 magnification and converted to ×100. One unit area at ×40 is equivalent to 0.087 mm 2 . Measurements were performed twice and averaged.

| DLM8 subcutaneous osteosarcoma mouse model
DLM8 tumors were implanted subcutaneously in the flank of C3H mice. Seven days after implantation, animals were randomized; treatment was initiated at a tumor volume of 106 ± 3 mm 3 (mean ± SE of the mean) and designated as Day 0. Specific treatments and dosing schedules for each group are summarized in Table 1.

| Immunohistochemistry
The resected lungs were washed in saline, fixed in 10% formalin buffer and embedded in paraffin. Five-micrometer-thick tissue sections were deparaffinized in xylene and rehydrated. For the primary bone tumors, tissues were fixed in 10% formalin and then decalcified using ethylenediaminetetraacetic acid for 24 hours before embedding in paraffin. Bone tumor tissue sections were heated at 60 C for 24 hours and then subjected to the same process of deparaffinization and rehydration as the lungs. Antigen retrieval was performed for tissues using vector unmasking solution pH 6 diluted in ddH 2 O (1:10).
Sections were blocked using normal blocking serum (Vector Laboratories, Burlingame, CA) diluted in phosphate-buffered saline (PBS) followed by overnight incubation with primary antibodies against CD4 Sections not exposed to the primary antibody served as negative controls, and normal mouse spleen was used as a positive control. Three slides per group were selected, then three different tumor areas per slide were quantified and averaged, and the mean ratio of positive pixel area to total positive counted nuclei was plotted. Measurements from each field and from each group (control or treatment) were analyzed for statistical significance using the Mann-Whitney U test.
Systemic BEMPEG treatment-driven increase in the ratio of CD4 + effector to CD4 + regulatory T cells was most prominent in the tumorbearing lungs. Significantly lower increase in CD4 + effector to Treg ratio was observed in the spleen and blood. Notably, CD4 + Treg fraction did not change in lungs of treated mice, suggesting that BEMPEG selectively increased the effector CD4 + T-cell fraction at the primary site of metastatic tumor lesion development.

| BEMPEG monotherapy decreases K7M3 primary tumor growth in tibia and increases T-cell infiltration into the tumor
To examine the effect of BEMPEG on the growth of primary orthotopic osteosarcoma, we performed intratibial injection of K7M3 cells.

| BEMPEG monotherapy decreases osteosarcoma tumor relapse in femurs after amputation of primary tibial tumors
To evaluate local recurrence after surgical resection, the femurs of mice were analyzed 13 days after amputation of the primary tibial

| BEMPEG enhances CTLA-4 monotherapy efficacy on osteosarcoma primary tumor growth inhibition and survival
Anti-PD-1 and anti-CTLA-4 each suppressed the growth of DLM8 osteosarcoma compared to vehicle control, leading to respective TGI of 57% (P < .01) and 79% (P < .0001) and 86% (P < .0001) for the combination. Addition of BEMPEG to anti-PD-1 and anti-CTLA-4 resulted in antitumor efficacy of 60% TGI (P < .01) and 96% TGI (P < .0001), respectively (Table 1) were 45% and 34%, respectively, suggesting a potential role for IL-2 in the treatment of osteosarcomas. Another study in relapsed pediatric sarcoma patients found that high-dose IL-2 therapy elicited complete responses in two of four osteosarcoma cases. 22 However, systemic IL-2 therapy can be difficult to tolerate, primarily due to serious side effects associated with capillary leak syndrome. 23 Furthermore, its short in vivo half-life necessitates frequent, repeated dosing.
BEMPEG is designed to improve these characteristics through modification of the IL-2 molecule by conjugation of an average of six PEG chains, which progressively release after administration, creating a long-acting IL-2 receptor agonist with preferential IL-2 receptor beta (CD122) binding. 27 In this study, BEMPEG also inhibited primary tumor growth and metastatic relapse in the lungs and bone tissue after K7M3 primary tumor resection, with a concurrent increase in CD8 + and CD4 + T-cell infiltration in the bone and lungs. Finally, although BEMPEG monotherapy showed modest efficacy against the DLM8 subcutaneous osteosarcoma model, combination with anti-CTLA-4 or anti-PD-1 checkpoint inhibitor antibodies led to durable tumor growth control with long-term survival, including complete cures in a significant fraction of treated animals. Collectively, these data provide a path toward clinical evaluation of BEMPEG-based regimens in human osteosarcoma.

ACKNOWLEDGEMENTS
The study was sponsored by Nektar Therapeutics. The authors thank Phillips Gilmore Oncology Communications for assistance with manuscript preparation, and BOLDSCIENCE Inc. for editorial assistance, both funded by Nektar Therapeutics.

CONFLICT OF INTEREST
MH, RP, PQ, GK, SK, LM, WWO and JZ are/were Nektar Therapeutics employees and have/had Nektar Therapeutics stock ownership interests to disclose. All other authors declare that they have no conflicts of interest.

ETHICS STATEMENT
All studies met the ethical, humane and current regulatory standards for transportation, housing and care established by the American Association of Laboratory Animal Care, the U.S. Department of Agriculture, Department of Health and Human Services and the National Institutes of Health.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.