Phase 1 multicenter study of the NKG2A targeting antibody S095029 as a single agent and in combination with anti-PD-1 in patients with advanced malignancies Authors: Aung Naing, Anna Spreafico, Minal Barve, Amita Patnaik, Xenophon Ianopoulos, Pauline Drean, Niels Jorgen Skartved, Agnes Hemon, Chahrazade Kantari-Mimoun, Peng He, Vasileios Askoxylakis, Nehal Lakhani Published: 2026 ABSTRACT AbstractBackgroundNKG2A is a C-type lectin that heterodimerizes with CD94 creating an inhibitory immunoreceptor. S095029 is a fully human monoclonal IgG1 anti-NKG2A antibody with attenuated Fc-effector functions that specifically binds the NKG2A/CD94 heterodimer, blocking the interaction with its ligand HLA-E. Sym021 is an anti-programmed cell death protein 1 (PD-1) IgG1 antibody. Preclinical data demonstrated that S095029 combined with Sym021 increases antitumor activity.MethodsThis first-in-human, open-label, multicenter study (NCT05162755) evaluated safety, tolerability, preliminary efficacy, pharmacokinetics and pharmacodynamics (PD) of S095029 as monotherapy or in combination with Sym021 in patients with advanced solid tumors. Monotherapy regimen included two doses of S095029 every 2 weeks (Q2W) followed by single-agent Sym021 (Q2W). Adverse events were defined per Common Terminology Criteria for Adverse Events V.5.0. Antitumor activity was evaluated per Response Evaluation Criteria In Solid Tumors (RECIST) V.1.1. Cytokine secretion as PD markers and target engagement/receptor occupancy were assessed in peripheral blood. Selected immune markers were assessed by multiplex immunofluorescence on tumor biopsies.ResultsAs of July 25, 2025, 21 patients with heavily pretreated solid tumors were treated with S095029 as a single agent (10, 30, 100, 300, 750 and 1,500 mg) and 20 patients were treated with S095029 (30, 100, 300, 750 and 1,500 mg) in combination with Sym021 (200 mg). S095029 was well tolerated without dose-limiting toxicities for both monotherapy and combination with anti-PD-1. A maximum tolerated dose was not reached. Two patients (9.5%; uterine sarcoma, porocarcinoma) who received one cycle of S095029 monotherapy followed by Sym021 and two patients (10%; leiomyosarcoma, squamous cell carcinoma of unknown primary) who received S095029 in combination with Sym021 showed confirmed partial responses. None of these patients had received prior anti-PD-1 treatment. The clinical benefit rate (complete response+partial response+stable disease ≥6 months) by RECIST V.1.1 in all enrolled patients was 19.5%. S095029 serum concentrations increased with dose for both monotherapy and combination with Sym021. Full receptor occupancy was achieved starting at 30 mg Q2W. Increases in levels of monokine induced by gamma interferon, macrophage inflammatory protein 1-beta and tumor necrosis factor-α in peripheral blood were detected following treatment with S095029+Sym021 suggesting immune cell activation on treatment.ConclusionsThe anti-NKG2A antibody S095029 in combination with anti-PD-1 Sym021 demonstrated a favorable safety profile and exhibited signal of antitumor activity in unselected, heavily pretreated patients with advanced malignancies.Trial registration numberNCT05162755.Keywords: Immunotherapy, Immune Checkpoint Inhibitor WHAT IS ALREADY KNOWN ON THIS TOPIC WHAT THIS STUDY ADDS NKG2A targeting may contribute to anti-programmed cell death protein 1 immune checkpoint activity. Our first-in-human, phase 1 study of an anti-NKG2A antibody demonstrated a favorable safety profile and preliminary antitumor activity in unselected, heavily pretreated patients with advanced malignancies. NKG2A targeting may contribute to anti-programmed cell death protein 1 immune checkpoint activity. Our first-in-human, phase 1 study of an anti-NKG2A antibody demonstrated a favorable safety profile and preliminary antitumor activity in unselected, heavily pretreated patients with advanced malignancies. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY The findings of this study may support the development of novel immunotherapy combinations for patients with cancer. The findings of this study may support the development of novel immunotherapy combinations for patients with cancer. Background NKG2A (CD159a) is a C-type lectin that heterodimerizes with CD94 creating an immune inhibitory receptor expressed on natural killer (NK) cells, NKT cells, gamma-delta (γδ) T cells and a subset of cytotoxic T cells.1 On binding to its ligand, human leukocyte antigen (HLA)-E, NKG2A/CD94 transmits an inhibitory signal via the 2 immune tyrosine-based inhibition motifs in its cytoplasmic tail and recruitment of SHP-1 tyrosine phosphatase.2 While this mechanism is part of the natural self-recognition/tolerance by NK cells, cancer cells exploit the NKG2A inhibitory pathway by overexpressing HLA-E, thereby protecting themselves against NK and T-cell mediated killing. NKG2A expression is increased on tumor infiltrating NK and T cells and can be induced by immunosuppressive factors.3 High intratumoral expression of NKG2A has been associated with unfavorable prognosis for some malignancies,4 while HLA-E expression may inhibit intratumoral cytotoxic T-cell lymphocytes.5 Inhibition of NKG2A restores the effector function of immune cells and early clinical studies indicate encouraging activity when NKG2A blockade is combined with other immune checkpoint inhibitors.6 S095029 is a fully human monoclonal IgG1 anti-NKG2A antibody with attenuated Fc effector functions (IgG1-LALA) that targets NKG2A. S095029 specifically binds the NKG2A/CD94 heterodimer, blocking the interaction with its ligand HLA-E. The blocking effect of S095029 increases the activation of primary NK cells and induces secretion of pro-inflammatory cytokines. Preclinical studies demonstrate that S095029 affects the cytotoxic activity of both primary human NK cells and γδ T cells in co-culture with cancer cell lines from various indications, leading to increased killing of cancer cells that express HLA-E on their surface.7 Sym021 is a humanized IgG1 antibody that inhibits binding of PD-1 to its ligands programmed death-ligand 1 (PD-L1) and PD-L2.8 Early-stage clinical trials of Sym021 showed an acceptable safety profile and single-agent antitumor activity in patients with advanced cancers.9 This is the first-in-human, phase 1, open-label, multicenter clinical trial evaluating the safety, tolerability, preliminary efficacy, pharmacokinetics (PK) and pharmacodynamics (PD) of S095029 as a single agent and in combination with the anti-PD-1 antibody Sym021 in patients with advanced solid tumors (ClinicalTrials.gov ID: NCT05162755). Methods Clinical study design This is a first-in-human, phase 1, multicenter, open-label dose escalation study designed to determine the safety and antitumor activity of S095029 (anti-NKG2A) as a single agent and in combination with Sym021 (anti-PD-1) in patients with advanced solid tumors. The study consisted of two dose finding parts: single-agent S095029 (Part 1a) followed in a staggered way by S095029 in combination with Sym021 (Part 1b). Data cut-off for this report: July 25, 2025. The clinical study design is presented in online supplemental figure 1. The full protocol of the study could be found in online supplemental file 1. Study objectives The primary objective of the phase 1 clinical trial was to assess the safety and tolerability and determine the maximum tolerated dose (MTD) and the recommended phase 2 dose (RP2D) of S095029 as a single agent and in combination with Sym021. Key secondary objectives included the evaluation of antitumor activity and characterization of PK and immunogenicity of S095029. PD biomarker analysis from paired tumor biopsies and/or peripheral blood was a key exploratory objective. Patient eligibility Eligible patients for this study were adult patients (18 years or older) with histologically confirmed unresectable, locally advanced or metastatic solid tumor who had disease progression after treatment with available standard of care therapies that are known to confer clinical benefit or were intolerant to standard treatment including prior immunotherapies. Key inclusion criteria were adequate hematological, renal and hepatic function. Key exclusion criteria were active autoimmune disease, clinically significant cardiovascular disease, history of uncontrolled chronic pulmonary condition or treatment with systemic immunosuppressive therapy other than steroids given in a prophylactic setting or at a chronic low dose. Patients with uncontrolled or symptomatic central nervous system or leptomeningeal metastases were excluded. Treatment The starting dose for S095029 was 10 mg administered via intravenous infusion once every 2 weeks (Q2W). The starting dose was determined based on non-clinical toxicology and pharmacology assessment.7 Patients in the monotherapy Part 1a received one 28-day cycle of treatment with single agent S095029 on day 1 and day 15 (10 mg, 30 mg, 100 mg, 300 mg, 750 mg, 1,500 mg). Following completion of the one-cycle dose limiting toxicity (DLT) evaluation period and review of monotherapy safety data, patients could start monotherapy treatment with Sym021 at the dose of 200 mg via intravenous infusion once Q2W. Patients in Part 1b received treatment with S095029 (30 mg, 100 mg, 300 mg, 750 mg, 1,500 mg) in combination with Sym021 (200 mg) Q2W. S095029 and Sym021 were infused over approximately 30 min (+10 min) with a 30 min separation between the infusions of S095029 and Sym021. Treatment for all patients continued until unacceptable toxicity, confirmed radiographic disease progression or patient/physician decision. For patients who achieved a partial response (PR) treatment continued until any of the prior criteria were met or until 24 months following the confirmed PR. Study assessments Safety assessments included incidence and severity of adverse events (AEs), incidence of DLTs, changes from baseline in laboratory values, vital signs, Eastern Cooperative Oncology Group (ECOG) status, physical examination and cardiac assessments and AEs leading to dose interruption, modification, treatment delays and treatment discontinuation. AEs were defined by the National Cancer Institute Common Terminology Criteria for Adverse Events V.5.0 and assessed at every visit. The DLT observation period was one cycle (eg, 28 days for Q2W treatment). The definition of DLTs is provided in online supplemental methods. Efficacy was evaluated by local investigator assessment per Response Evaluation Criteria In Solid Tumors (RECIST) V.1.1. Tumor assessments were performed at baseline and every 8 weeks (±7 days) until confirmed disease progression or withdrawal from the study. Per RECIST V.1.1, response needed to be confirmed by repeat imaging assessment not less than 4 weeks from the date the response was first documented. Pharmacokinetics and immunogenicity S095029 and Sym021 serum concentrations were measured using two target-specific validated assays, and the anti-drug antibodies (ADA) were measured using two drug specific bridging electrochemiluminescence (ECL) validated assays. PK parameters were derived using a non-compartmental analysis performed on WinNonLin V.8.3 (Certara). Blood samples in Part 1a were collected in cycle 1 at day 1 pre-dose, end of infusion (EOI), 2, 4, 8, 24, 48 hours after infusion and then at day 8, day 15 (pre-dose and EOI), day 22 and at day 1 of cycle 2 and 3 (pre-dose and EOI), day 1 of cycle 4 (pre-dose) and at the end of treatment visit (EOT). Blood samples in Part 1b were collected in cycle 1 at day 1 pre-dose, EOI of S095029, EOI of Sym021, 4 hours, 48 hours after EOI of Sym021 and then at day 8, day 15 (pre-dose and EOI of both S095029 and Sym021), and day 22. From cycle 2 to the EOT, blood samples were collected during each cycle at day 1 and day 15, at pre-dose and at both EOI of S095029 and Sym021, and finally at EOT. Receptor occupancy and biomarker assessment in blood Receptor occupancy (RO) was measured by flow cytometry in fresh whole blood, in CD3-CD56+NK cells, by measuring the binding of a competitive (Sym025-PE, Servier) and non-competitive anti-NKG2A antibody (CD159a_REA110, Miltenyi). The calculation of the percentage of RO was done as follows: %RO = ((CD159a + –Sym25-PE+) / CD159a+) × 100. It was assessed at different time points of cycle 1: Pre-dose, EOI, C1D1, C1D8, C1D15, C1D22, and C2D1. Soluble factors levels were measured in human plasma by Meso Scale Discovery (MSD) technology with a V-PLEX kit for the detection of interferon (IFN) gamma, tumor necrosis factor (TNF)-α, interleukin (IL)-1b, IL-6, IL-8, and IL-10, a V-PLEX kit for the detection of macrophage inflammatory protein (MIP)-1 alpha and MIP-1 beta, and a U-PLEX kit for the detection of monokine induced by gamma IFN (MIG or CXCL9). Analysis was performed on frozen plasma of n=31 patients between C1D1 and C2D1. Biomarker assessments in tumor Core needle formalin-fixed paraffin-embedded biopsies were collected at baseline and at the end of cycle 1. The samples were sent to NeoGenomics for multiplex immunofluorescence (mIF). Prior to analysis, an H&E-stained slide from each tissue block was examined by a pathologist to assess tumor content and tissue quality. For IF analysis, a multiplex staining with the following 12 markers was performed to assess target expression and immune populations infiltration and activation; CD3 (clone F7.2.38), CD8 (clone C8/144B), CD56 (clone MRQ-42), NKP46 (clone EPR22403-57), NKG2A (clone EPR23737), PanCK (clone AE-1/PCK26), HLA-E (clone MEM-E-02), granzyme-B (clone GrB-7), KI67, PD-L1 (clone SP6), 41BB (clone BLR051F) based on the following formula: (H-score=1 × (% of 1 + intensity positive cells) + 2 × (% of 2 + intensity positive cells) + 3 × (% of 3 + intensity positive cells)) and PD-L1 CPS (combined positive score). Statistical analysis Dose escalation Part 1a started with an accelerated titration period then a Bayesian optimal interval (BOIN) design.10 Dose escalation Part 1b started directly with the BOIN design. The decision to dose escalate, continue at the same dose, or dose de-escalate was guided by the BOIN design principles. Per the BOIN design, a target DLT rate of 0.30 was chosen and the rules for dose escalation or de-escalation were as follows: Dose escalation was allowed if the observed DLT rate at the tested dose level was <0.236, while de-escalation was required if the observed DLT rate at the tested dose level was >0.359. If the observed DLT rate at the tested dose level was between 0.236 and 0.359, a new cohort of up to three patients would be enrolled at the same tested dose level. The MTD is defined in accordance with dose escalation rules. Based on the number of provisional dose levels, it was determined that approximately 18 and 21 DLT-evaluable patients would be enrolled in Parts 1a and 1b, respectively. The sample size was proposed to ensure acceptable operating characteristics, such as the correct MTD selection percentage and number of patients exposed to toxic dose levels, under various hypothetical scenarios. In addition, a maximum of six patients could be enrolled as a backfill cohort in each Part, making the overall sample size for the dose escalation period up to approximately 51 patients. The primary analyses of safety and efficacy were based on the treated set, which includes all patients who received at least one dose of IMP (S095029 or Sym021) during Part 1. Tumor response is evaluated as per RECIST V.1.1 assessed by the investigator. Confirmation of complete response (CR) and PR is required, and a response is considered as confirmed if the same response criteria were met at a subsequent time point that is at least 4 weeks apart. Clinical efficacy endpoints were analyzed using point estimates and the Wilson score intervals. Statistical tests for flow cytometry measurements, mIF analysis and MSD soluble factors measurements were performed on changes between end of cycle 1 (between C1D15-C1D22) and screening for patients treated with S95029 alone or in combination with Sym021. Wilcoxon tests and Kruskal-Wallis test were performed and statistical significance was set at a p value threshold of 0.05. Results Patient demographics and characteristics As of the data cut-off of July 25, 2025, 41 patients with advanced/metastatic malignancies were treated in the phase 1 study: 21 patients received S095029 as a single agent (10 mg, 30 mg, 100 mg, 300 mg, 750 mg and 1,500 mg Q2W for one 28-day cycle) in Part 1a, and 20 patients received S095029 (30 mg, 100 mg, 300 mg, 750 mg and 1,500 mg Q2W) in combination with Sym021 (200 mg Q2W) in Part 1b. All patients in Part 1a continued treatment with single agent Sym021 at a fixed dose of 200 mg Q2W following completion of one cycle of treatment with single agent S095029. Key patient demographics and characteristics are presented in table 1 and tumor types are listed in online supplemental table 1. The median age was 55 years (range 29–82) for the monotherapy arm and 63 years (range 35–77) for the combination arm. The most frequent tumor type was colon adenocarcinoma (online supplemental table 1). 40 patients had received prior systemic antineoplastic therapies; seven patients in Part 1a (33%) and 8 patients in Part 1b (40%) had received prior anti-PD-1 treatment. Overall, 40 patients (%) had discontinued treatment as of the data cut-off date; 34 patients (%) due to disease progression, 3 due to AEs, 2 due to physician decision and 1 due to withdrawal of consent. Table 1Demographics and characteristics of patients treated with S095029 as a single agent or in combination with Sym021Demographics and baseline characteristicsPart 1a (N=21)Part 1b (N=20)Age, median (range), years55.0 (29–82)63.0 (35–77)Male, n (%)8 (38.1)13 (65.0)Female, n (%)13 (61.9)7 (35.0)Ethnic origin, n (%) Asian02 (10.0) Black or African American3 (14.3)2 (10.0) White18 (85.7)15 (75.0) Other01 (5.0)ECOG, n (%) 07 (33.3)5 (25.0) 114 (66.7)15 (75.0) Prior anti-PD-1 therapy, n (%)7 (33.3)8 (40.0)Prior treatment lines, n (%) 01 (4.8)0 11 (4.8)2 (10.0) 23 (14.2)10 (50.0) ≥316 (76.2)8 (40.0)ECOG, Eastern Cooperative Oncology Group; PD-1, programmed cell death protein 1. Demographics and characteristics of patients treated with S095029 as a single agent or in combination with Sym021 ECOG, Eastern Cooperative Oncology Group; PD-1, programmed cell death protein 1. Safety Overall, S095029 was well tolerated without any DLTs for both monotherapy (Part 1a) and the combination with anti-PD-1 Sym021 (Part 1b). Since no DLTs were observed across any dose level, and since after completion of the DLT-assessment period for the highest dose level in both the monotherapy and the combination arm the BOIN decision allowed further escalation, an MTD was not reached. Treatment-related AEs (TRAEs) of any grade were reported in 25 patients (61.0%): 13 patients in the monotherapy arm (61.9%) and 12 patients (60%) in the combination arm. The most common (≥2 patients) TRAEs of any grade for patients treated in Part 1a were diarrhea (33.3%), nausea, fatigue and immune-related hypothyroidism (14.3% each). The most common TRAEs (≥2 patients) of any grade in patients treated with S095029 in combination with Sym021 included maculopapular rash (15%), pruritus (15%) and infusion-related reactions (15%). One patient in the combination arm experienced a TRAE ≥Grade 3 (Grade 3 immune-mediated enterocolitis). A summary of all TRAEs is presented in table 2. Table 2Treatment-related adverse eventsAE related to S095029 or the combinationPart 1a monotherapy (N=21)Part 1b combination (N=20)All gradesn (%)Grade ≥3n (%)All gradesn (%)Grade ≥3n (%)All13 (61.9)2 (9.5)12 (60.0)1 (5.0)Nausea3 (14.3)–1 (5.0)–Diarrhea7 (33.3)–1 (5.0)–Abdominal discomfort1 (4.8)–––Gastroesophageal reflux disease1 (4.8)–––Immune-mediated enterocolitis––1 (5.0)1 (5.0)Chills2 (9.5)–1 (5.0)–Fatigue3 (14.3)–1 (5.0)–Fever––1 (5.0)–Pruritus––3 (15.0)–Rash maculopapular––3 (15.0)–Alopecia1 (4.8)–––Rash––1 (5.0)–Infusion-related reaction1 (4.8)–3 (15.0)–Headache1 (4.8)–2 (10.0)–Dizziness1 (4.8)–––Immune-mediated hypothyroidism3 (14.3)–––Immune-mediated hepatitis1 (4.8)–––Tumor pain––1 (5.0)–Hot flush––1 (5.0)–Immune-mediated adrenal insufficiency––1 (5.0)–Immune-mediated hyperthyroidism––2 (10.0)–Blood creatine phosphokinase increased1 (4.8)1 (4.8)––Decreased appetite––1 (5.0)–Hypertension1 (4.8)–––Cough1 (4.8)–––Hyperglycemia1 (4.8)1 (4.8)––Edema peripheral1 (4.8)– –Back pain1 (4.8)–––Pain in jaw1 (4.8)–––Insomnia––1 (5.0)–Pleuritic pain––1 (5.0)–AE, adverse event. Treatment-related adverse events AE, adverse event. AEs regardless of relationship to study drug were observed in 19 patients (90.5%) in the monotherapy arm and 19 patients (95%) in the combination arm. The most common AEs (≥2 patients) in the monotherapy arm were diarrhea (33.3%), nausea, fatigue and urinary tract infection (23.8% each), anemia, and vomiting (19% each) and constipation, cough, dyspnea, back pain and immune-mediated hypothyroidism (14.3% each). The most common AEs (≥2 patients) in the combination arm were headache and decreased appetite (20% each) and infusion-related reaction, dyspnea, fatigue, pruritus and maculopapular rash (15% each). AEs≥Grade 3 regardless of relationship to study drug were reported in 12 patients (57.1%) in the single-agent arm and in 7 patients (35%) in the combination arm. An overview of AEs occurring in ≥10% of treated patients regardless of relationship to study drug is provided in online supplemental table 2. Serious adverse events (SAEs) regardless of relationship to study drug were reported in eight patients (38.1%) in Part 1a and seven patients (35%) in the combination Part 1b. Treatment withdrawal due to TRAEs was reported in two patients: In the monotherapy arm, one TRAE of Grade 3 immune-mediated blood creatine phosphokinase increase leading to study withdrawal per-protocol was observed at the 30 mg dose level in cycle 2. In the combination arm, one participant experienced an SAE of Grade 3 immune-mediated colitis with diarrhea, abdominal pain, nausea and vomiting after five treatment cycles at the 100 mg S095029 dose level that were considered related to S095029 and Sym021. This event led to treatment withdrawal per-protocol, and the participant recovered 44 days thereafter. Efficacy Median exposure to S095029 was 4 weeks (range 3.9–4.8) in the monotherapy arm and 10.1 weeks (range 3.9–80.0) in the combination arm. In the monotherapy arm, 2 patients (9.5%) had confirmed PR, 9 patients (42.9%) had stable disease (SD) and 10 patients (47.6%) had progressive disease (PD) as investigator-assessed best overall response (BOR) per RECIST V.1.1. The disease control rate (CR, PR and SD) was 52.4% (95% CI 32.4% to 71.7%) and the clinical benefit rate (CR, PR, and SD ≥6 months) was 23.8% (95% CI 10.6% to 45.1%) (table 3). The two patients who had confirmed PRs had uterine sarcoma and porocarcinoma. Both patients were naïve to prior anti-PD-(L)1 treatment and received S095029 at the dose level of 300 mg Q2W. Table 3Investigator-assessed confirmed best overall response by Response Evaluation Criteria In Solid Tumors V.1.1Part 1a (N=21)Part 1b (N=20)Best overall response, n (%) PR2 (9.5)2 (10.0) SD9 (42.9)5 (25.0) PD10 (47.6)13 (65.0)Disease control rate (CR+PR+SD), n (%)11 (52.4)7 (35.0) 95% CI(32.4 to 71.7)(18.1 to 56.7)Clinical benefit rate (CR+PR+SD ≥6 months), n (%)5 (23.8)3 (15.0) 95% CI(10.6 to 45.1)(5.2 to 36.0)CR, complete response; PD, progressive disease; PR, partial response; SD, stable disease. Investigator-assessed confirmed best overall response by Response Evaluation Criteria In Solid Tumors V.1.1 CR, complete response; PD, progressive disease; PR, partial response; SD, stable disease. In the combination arm, 2 patients (10%) had confirmed PR, 5 patients (25%) had SD and 13 patients (65%) had PD as investigator-assessed BOR per RECIST V.1.1. The disease control rate (CR, PR and SD) was 35% (95% CI 18.1% to 56.7%) and the clinical benefit rate (CR, PR, and SD ≥6 months) was 15% (95% CI 5.2% to 36.0%) (table 3). The two patients who had confirmed PRs had leiomyosarcoma and squamous cell carcinoma of unknown primary. Both patients had PD-L1 CPS >1, were naïve to prior anti-PD-(L)1 treatment and received S095029 at the dose level of 100 mg Q2W in combination with Sym021 at 200 mg Q2W. The percentage change from baseline in tumor size and the duration of response for all patients are presented in figure 1. Clinical efficacy of S095029 as monotherapy (two doses of S095029 every 2 weeks followed by single-agent Sym021 in Part 1a) and in combination with Sym021 in patients with advanced malignancies. (A) Best percentage change from baseline and best overall responses. (B) Duration of responses. AE, adverse event; CR, complete response; esoph, esophageal; HNSCC, head and neck squamous cell carcinoma; NE, non evaluable; NSCLC, non-small cell lung cancer; PD, progressive disease; perito, peritoneal; PR, partial response; RCC, renal cell carcinoma; RECIST, Response Evaluation Criteria in Solid Tumors; SD, stable disease; TPBC, triple-positive breast cancer. Pharmacokinetics and immunogenicity The PK exposure of S095029 increased with dose between doses of 10 and 1,500 mg (figure 2A). No PK interaction between Sym021 and S095029 was observed, and more specifically the PK exposure of S095029 was not affected by the co-administration of Sym021. Pharmacokinetics and NKG2A RO in peripheral blood. (A) S095029 mean concentration after the two first Q2W administrations. (B) NKG2A RO was measured in CD3-CD56+NK cells in peripheral blood by flow cytometry throughout the cycle 1 (n=41 patients). NK, natural killer; Q2W, every 2 weeks; RO, receptor occupancy. The mean S095029 exposure (area under the plasma concentration time-curve over the dose interval (AUCtau) was between 536 and 62,222 hours×µg×mL−1 (coefficient of variation (CV%) between 23% and 42%) and the mean maximum serum concentration (Cmax) was between 3.96 and 401 µg×mL−1 from 10 to 1,500 mg (CV% between 22% and 53%), respectively (online supplemental table 3A). The median S095029 half-life (t1/2), estimated after the first administration, was 8 days. Trough concentration assessment (Ctrough) indicated moderate accumulation of S095029 at least up to cycle 5. At the dose of 200 mg, the mean Sym021 AUCtau was 12 315 h×µg×mL−1 (CV% of 55%) and the mean Cmax was 73.1 µg×mL−1 (CV% of 54%). The median Sym021 half-life, estimated after the first administration, was 10 days. PK parameters for Sym021 are presented in online supplemental table 3B. S095029 and Sym021 ADA analysis were performed on 252 available samples collected prior to the start of study treatment, at Ctrough time points during the study, and at EOT or follow-up visits. No immunogenicity alerts for S095029 or Sym021 were encountered in any of the 41 patients; treatment-induced ADA against S095029 were detected in only 1 patient and for Sym021 in 2 patients. Full NKG2A RO in peripheral blood was achieved starting at low doses of 30 mg Q2W both in the monotherapy and the combination arm. Full NKG2A RO in whole blood was retained throughout the entire cycle 1 (figure 2B). Based on all available clinical safety and efficacy, PK, RO and PD data and using modeling and simulation analysis for target saturation at the tumor level,11 the recommended phase 2 dose in combination with PD-1 blockade was determined at 1,140 mg once every 3 weeks. The specific criteria for RP2D determination are provided in online supplemental methods. Biomarkers To study the PD effects of S095029, longitudinal analysis of soluble cytokines/chemokines was performed in frozen plasma at multiple time points including baseline, C1D1, C1D8, C1D15, and C1D22. While no significant changes were detected after treatment with S095029 as a single agent, significant increases in levels of CXCL9, TNF-α and MIP1b in peripheral blood were detected following treatment with S095029 in combination with Sym021, suggesting immune cell activation on combination treatment (figure 3). In addition to cytokine analysis, immunofluorescence staining was performed in paired tumor biopsies collected at baseline and at the end of cycle 1. Assessment of HLA-E and NKG2A indicated numerically higher HLA-E and NKG2A levels in patients with PR who had received combination treatment, but the differences did not reach statistical significance (online supplemental figure 2). Notably, paired analysis of tumor tissue from a patient with metastatic squamous cell carcinoma of unknown primary site who showed a PR to treatment with S095029 in combination with Sym021 revealed an increase in both cytotoxic granzyme B+CD8+ T cell and NKG2A+cells infiltration at the end of cycle 1 (online supplemental figure 3). Regarding PD-L1 expression, among 18 evaluable patients, 12 had positive PD-L1 status (CPS ≥1), and 6 were negative for PD-L1 expression (CPS <1) (online supplemental table 4). No significant correlation was detected between PD-L1 expression and BOR. Pharmacodynamic effects and immune cell activation in peripheral blood on treatment with S095029 as monotherapy and in combination with Sym021. CXCL9, MIP1b and TNF-α level changes on treatment were measured in frozen plasma from patients between C1D1 and C2D1 (n=31 patients). CXCL9 (MIG), monokine induced by gamma interferon; MIP1b, macrophage inflammatory protein 1-beta; TNF, tumor necrosis factor. Discussion Immune checkpoint blockade (ICB) has a transformational impact on the treatment of cancer; however, many patients do not respond to ICB or develop therapeutic resistance. NK cells play a critical role in the antitumor immune response,12 and their inhibition or impairment is associated with cancer initiation and progression.13,15 Malignant cells use various strategies to evade the antitumor activity of NK cells, including the recruitment of suppressor cells such as regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs) that cause NK cell exhaustion,16 and the release of cytokines such as transforming growth factor-beta (TGF-ß) that suppress NK cell cytotoxicity.17 NKG2A is an inhibitory immune receptor expressed on NK cells and subsets of T cells.18 Cancer cells take advantage of the inhibitory activity of NKG2A by overexpressing its ligand HLA-E, thus inducing an immunosuppressive tumor microenvironment that protects tumors. Overexpression of HLA-E has been associated with worse prognosis in different malignancies and using monoclonal antibodies to block NKG2A has been shown to unleash T and NK cell effector function and promote antitumor immunity.19 20 The HLA-E/NKG2A pathway can promote tumor growth, and research data support the potential antitumoral activity of combining antibodies against NKG2A and PD-1.21 Here, we report the results from the first-in-human clinical trial of the anti-NKG2A antibody S095029. Based on preclinical data demonstrating that S095029 induces the secretion of pro-inflammatory cytokines and increases the cytotoxic activity of both primary human NK cells and γδ T cells, we tested S095029 as a single agent and in combination with anti-PD-1 in heavily pretreated patients with advanced/metastatic cancers. This phase 1 dose-finding trial demonstrated that S095029 is well tolerated both as monotherapy and in combination with anti-PD-1 without an increase in the incidence or severity of immune-mediated toxicities or new and unexpected safety signals compared to established ICI therapies. The most common single-agent toxicities were mild, and largely comparable to AEs reported for the anti-NKG2A antibody monalizumab.22 23 The safety profile of S095029 in combination with Sym021 was comparable to that seen with Sym021 alone.9 No DLTs were observed for S095029 in combination with Sym021 and no MTD was reached, consistent with safety data from the phase 1/2 study of monalizumab in combination with durvalumab.24 The overall incidence of treatment-related AEs in the dose-finding parts of both studies was approximately 60%. Grade 3 immune-mediated TRAEs were reported for only one patient in our study (immune-mediated enterocolitis); in comparison 14.3% (N=20) of participants in the dose expansion part of the monalizumab plus durvalumab study experienced grade 3/4 TRAEs, the most common being colitis, anemia, alanine aminotransferase increase and hypokalemia.24 Preliminary signal of clinical efficacy was seen in both monotherapy and the combination arms with a few patients having confirmed PRs. Responses occurred in heavily pretreated patients with tumor types for which the effectiveness of immunotherapies has not been previously established, such as uterine sarcoma, leiomyosarcoma and porocarcinoma. Moreover, responses were durable lasting in three cases more than 16 months. However, none of the patients who achieved PR had received prior treatment with immune checkpoint inhibitors, while patients in the S095029 monotherapy arm received only one cycle of S095029 and then switched to single-agent anti-PD-1 Sym021. Thus, while the responses to anti-NKG2A-based treatment are encouraging, the current data do not provide safe conclusions on the contribution of S095029 to the response effect. In addition to PRs, 42.9% of patients in the monotherapy arm and 25% of patients in the combination arm had SD as BOR; the disease control rate for all patients in the study was 43.9%. In a phase 1 study of single-agent monalizumab in patients with advanced gynecologic malignancies, no responses were observed for monotherapy while approximately 25% of all evaluable patients (41% of patients in dose escalation and 18% of patients in dose expansion) had SD as the best response.22 Monalizumab in combination with durvalumab did not show any responses in the mixed population of advanced solid tumors in the dose escalation part of the phase 1/2 study, and approximately 44% of patients had stable disease as BOR.24 In the same study, clinical activity in expansion cohorts of homogeneous patient populations with microsatellite-stable colorectal cancer (MSS-CRC), ovarian cancer, endometrial cancer and non-small cell lung cancer (NSCLC) that had not received prior treatment with anti-PD-1, anti-PD-L1 or anti-cytotoxic T-lymphocyte associated protein 4 immunotherapy was relatively modest with objective response rate (ORR) not exceeding 10% in any population.24 In addition to the clinical data, our PD findings were largely consistent with prior reports. Our biomarker longitudinal analysis in frozen plasma demonstrated increases in levels of soluble cytokines, such as CXCL9, MIP-1b and TNF-α, suggesting immune cell activation on treatment with S095029 in combination with anti-PD-1, consistent with data of monalizumab in combination with durvalumab showing increases of immune-activating blood cytokines.24 25 Our mIF analysis of the paired tumor tissue from a patient with metastatic squamous cell carcinoma who showed a PR to treatment with S095029 in combination with Sym021 are also in line with these two studies, highlighting an increase in both CD8+T cell and NKG2A+cells infiltration post-treatment. Notably, HLA-E and NKG2A levels were numerically higher in patients with PR who had received combination treatment, but the differences did not reach statistical significance. Similar observations were reported in the COAST study with no significant differences in the levels of NKG2A and HLA-E between the groups. In a recent exploratory biomarker analysis, progression-free survival (PFS) benefit was observed among patients with median or higher NKG2A or HLA-E expression.26 Based on the data from our phase 1 study, S095029 is currently being investigated in combination with anti-PD-1 in patients with previously untreated advanced NSCLC with high PD-L1 expression (tumor proportion score ≥50%)27 and in patients with microsatellite instability-high/deficient mismatch repair (MSI-H/dMMR) advanced gastric/GEJ cancer (gastroesophageal junction).28 The choice of NSCLC and gastric/GEJ is based on data indicating a relevance of NK cells and the HLA-E/NKG2A axis in these malignancies.29 30 This is further supported by internal data evaluating the expression of the NKG2A receptor, HLA-E and other NK-related genes in human samples, demonstrating a high presence within the tumor microenvironment in different solid cancers, including MSI-H tumors (data not shown). HLA-E and NKG2A expression have been reported to be independent factors for poor prognosis of patients with advanced gastric cancer.29 The hypothesis of a potential role for NKG2A in NSCLC is strengthened by findings from two independent clinical trials: A randomized phase 2 study (COAST) in patients with unresectable stage III NSCLC and no progression following concurrent chemoradiotherapy showed that adding the anti-NKG2A monoclonal antibody monalizumab to durvalumab resulted in numerically higher confirmed ORR and prolonged PFS relative to durvalumab alone.31 Moreover, a second randomized phase 2 study (NeoCOAST) assessing the effectiveness and safety of neoadjuvant durvalumab alone or in combination with monalizumab in patients with resectable, early stage (stage I-IIIA) NSCLC showed numerically improved major pathological response rates for the combination treatment.25 Building on the findings from the NeoCOAST trial and other perioperative studies, the phase 2 NeoCOAST-2 platform study tested neoadjuvant monalizumab in combination with durvalumab and platinum-doublet chemotherapy in 70 patients with untreated, resectable stage IIA-IIIB NSCLC. The combination of anti-NKG2A with anti-PD-L1 and chemotherapy showed a pathological complete response (pCR) rate of approximately 26% and an major pathological response (mPR) rate of 50% in the modified intention-to-treat population further highlighting the potential role of targeting NKG2A in NSCLC.32 In conclusion, our data suggest that NKG2A targeting may contribute to anti-PD-1 immune checkpoint activity. Our study demonstrated a favorable safety profile and preliminary antitumor activity in unselected, heavily pretreated patients with advanced malignancies. The clinical impact of S095029 in combination with established anti-PD-(L)1 therapies warrants further investigation. Supplementary material Acknowledgements This work was supported by Symphogen A/S and Servier. The authors thank Rikke Hald, MSc, PhD, for work on the clinical pharmacokinetic assay development and validation as lead bioanalytical senior scientist on the project. The authors thank the patients who participated in the clinical trial and their families, as well as the staff at the sites who assisted with the study. The authors acknowledge Fanny Cazade, Sofie Ellebaek Pollmann, Matthieu Bouaziz and Johanna Verneau. Footnotes Funding: This study was sponsored by Symphogen A/S and Servier. Preliminary results were previously presented at the 2024 Annual Meeting of the Society for Immunotherapy of Cancer. The sponsor was involved in the study design; in the collection, analysis and interpretation of the data; in the writing of the report; and in the decision to submit the paper for publication. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: This study was performed in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice. The study was approved by an independent ethics committee or institutional review board at each study center (Princess Margaret Cancer Centre (UHN): 21-5821 on 24 January 2022; START Midwest: STMW2021.35 on 28 September 2021; MD Anderson Cancer Center: 2021-0757 on 4 November 2021; Mary Crowley Cancer Research: #21‐43 on 9 November 2021; START San Antonio: START2021.43 on 16 September 2021). All patients provided written informed consent before taking part and before any study procedures. Data availability statement Data are available upon reasonable request. References Associated Data Supplementary Materials Data Availability Statement Data are available upon reasonable request.