Pharmacologic Properties and Preclinical Activity of Sasanlimab, A High-affinity Engineered Anti-Human PD-1 Antibody

The ability to escape immune recognition is a hallmark of cancer progression. Indeed, even in the presence of tumor antigens, T cells often become dysfunctional and are unable to respond to repeated antigen encounters (1–3). The response of tumor-infiltrating T lymphocytes (TIL) is regulated by costimulatory and coinhibitory immune checkpoint receptors (4). TILs express high levels of programmed cell death protein-1 (PD-1), a checkpoint receptor that dampens the T-cell response once it binds to its ligands, PD-L1 (also known as CD274 or B7-H1) and/or PD-L2 (also known as CD273 or B7-DC; refs. 5–7). Many tumor cells express PD-L1 (4, 6–8) and exploit the PD-1/PD-L1 pathway, thus maintaining an immunosuppressive environment (2, 4).

PD-1 is an immunoglobulin (Ig) superfamily member, belonging to the CD28 family, together with BTLA and CTLA-4 (1, 9). PD-1 functions as a coinhibitory regulator of T cells to maintain equilibrium between activation, tolerance, and immune-mediated tissue damage in the periphery (1, 9). Moreover, broad expression of the PD-1 receptor has been reported on regulatory T cells, B cells, natural killer cells, dendritic cells (DC), and macrophages and is evident following immune activation (1, 10, 11). The PD-1 cytoplasmic region constitutes an immunoreceptor tyrosine-based inhibitory motif and an immunoreceptor tyrosine-based switch motif, both of which are essential for the inhibitory function of PD-1 (2, 9, 11). Following PD-1/PD-L1 or PD-L2 ligation, the induction of PI3K activity is blocked, suppressing protein kinase B (AKT) phosphorylation. This ultimately leads to downregulation of effector cytokine production (including IFNγ, TNFα, and IL2), cell survival proteins (including Bcl-xL), and T-cell proliferation, thereby dampening the immune response and preventing tissue damage and autoimmunity (10, 11).

mAb-based therapies targeting the PD-1 receptor have yielded clinical benefits and durable responses in some patients with multiple tumor indications (12–14). Nivolumab (fully human) and pembrolizumab (humanized) anti-PD-1 antibodies of IgG4 isotype have been approved for use as monotherapy or in combination with other agents in melanoma, non–small cell lung cancer, head and neck squamous cell cancer, gastric, esophageal, cervical, and microsatellite instability-high cancers; urothelial, hepatocellular, Merkel cell, renal cell, and endometrial carcinomas; and classical Hodgkin and primary mediastinal large B-cell lymphomas (15, 16). Ongoing trials with antibodies generated against PD-L1 have also yielded encouraging results (17, 18), further highlighting the importance of blocking the PD-1/PD-L1 axis in immuno-oncology as a single agent or in combination with other agents (19, 20).

Sasanlimab (PF-06801591) is a humanized, hinge region–stabilized (21) IgG4 mAb directed against human PD-1 (hu-PD-1). It selectively and reversibly binds to human and cynomolgus monkey PD-1 (cy-PD-1) with similar high affinities, thereby antagonizing the inhibitory interaction between PD-1 and PD-L1/PD-L2. Sasanlimab has been tested in a phase I clinical study as a monotherapy administered to patients with solid tumors via intravenous infusion or subcutaneous injection (22). Sasanlimab treatment is well-tolerated and associated with antitumor activity in a variety of tumor types. Here, we present the biophysical characteristics and preclinical antagonistic activities of sasanlimab that supported the initiation of clinical trials in patients with cancer.

Detailed methodologies relating to antibody generation, drug manufacturing, and basic molecular and biophysical characterization (Supplementary Figs. S1–S7), together with a detailed description of performed studies, are provided in the Supplementary Materials and Methods section.

All animal procedures were performed following the guidelines established by the Institutional Animal Care and Use Committee.

Kinetics and affinity determinations for hu-PD-1, cy-PD-1, and mouse PD-1 (mu-PD-1) interactions with sasanlimab were conducted at 25°C in a running buffer (PBS, 0.01% Tween 20, pH 7.4). Hu-PD-1 and cy-PD-1 were additionally analyzed at 37°C. Experiments were performed on a Bio-Rad ProteOn XPR36 Surface Plasmon Resonance (SPR) Biosensor equipped with NLC (neutravidin coated) Chips (Bio-Rad). Biotinylated sasanlimab was captured onto the ligand channels of the NLC chip, and monomeric PD-1 analytes were injected in the analyte channels, using a “one-shot” kinetics methodology as described previously (23). A similar experiment was performed with the dimeric rat PD-1 Fc-fusion protein to evaluate cross-reactivity with rat PD-1. Kinetics experiments with neonatal Fc receptors (FcRn) as analytes were performed using the “one-shot” kinetics methodology at 25°C. For this purpose, 10 mmol/L sodium phosphate, pH 5.9, 150 mmol/L NaCl, and 0.01% Tween 20 were used as a running buffer. Active concentrations determined by titration against sasanlimab were used for hu-PD-1 and cy-PD-1 (24). Active concentrations determined by calibration-free concentration analysis (25) were used for human FcγRI (using human IgG1 isotype control as a detection molecule), and human and cynomolgus monkey FcRn (using trastuzumab human IgG1 N434W mutant as the detection molecule). Nominal concentrations (as determined by A₂₈₀) were used for all other analytes.

Hu-PD-1 (ReqSeq NP_005009.2) residues 32–160 with a C-terminal GS linker and 8xHis tag were expressed in HighFive Insect Cells (Gibco) using the Bac-to-Bac Baculovirus Expression System (Gibco). The sasanlimab kappa light chain and sasanlimab human IgG1 heavy chain Fab region with a C-terminal 10xHis tag were expressed in HEK Cells (Gibco). Harvested PD-1 and sasanlimab Fab were purified separately over a Ni Sepharose Excel Column (GE Healthcare) before gel filtration on a Superdex 75 Column (GE Healthcare). PD-1/sasanlimab Fab complex was prepared by mixing purified Fab and PD-1 at a 1.5:1 molar ratio followed by a 1-hour incubation on ice and further purification by gel filtration on a Superdex 200 Column (GE Healthcare). The complex was concentrated to 10 mg/mL for crystallization. PD-1/sasanlimab complex was crystallized in 0.5 mol/L LiCl, 0.1 mol/L citric acid, pH 4, 22% w/v PEG 6000, and 0.5% w/v n-Octyl-b-d-glucoside, and flash cooled before data collection at the ALS beamline 5.0.2. Data were processed using XDS (26) and phased by molecular replacement. Refinement was carried out using PHENIX (27) and manual rebuilding was performed in COOT (28). Figures were generated using PyMOL.

The atomic coordinates and structure factors are deposited in the Protein Data Bank (PDB), www.pdb.org (PDB codes 6XKR).

HEK-293T cells transfected with hu-, cy-, mo-, or rat-PD-1 vectors, or empty vector were harvested at 48 hours after transfection. The Jurkat cell line (clone E6-1-TIB-152TM; ATCC) was used to express hu-PD-1 and cy-PD-1 stably. Cell lines were generated using electroporation via Amaxa Nucleofector System (Lonza). Transfections were performed using a Kit V as instructed by the manufacturer's protocol (Lonza).

Primary T cells were purified from human and cynomolgus monkey peripheral blood mononuclear cells (PBMC) or from mouse and rat spleens. Human, cynomolgus monkey, and mouse T cells were activated using magnetic beads coated with anti-CD3/CD28 antibodies (Dynabeads, human or mouse T Cell-Activator Kit, Life Technologies; and Primate T Cell-Activator, Miltenyi Biotec). Rat T cells were stimulated using 0.1 μg/mL Phytohemagglutinin (Sigma). Cultures were incubated at 37°C in 5% CO₂ for up to 72 hours.

After appropriate PD-1 expression was confirmed using different anti-PD1 antibodies, cells were treated with serial dilutions of sasanlimab or negative control antibody (1–0.00001 μg/mL) on ice for 1 hour. Cells were then washed and stained with APC-labeled donkey anti-human-affiniPure F(ab')₂ Fragment IgG, Fcγ-specific (1:100 dilution, Jackson ImmunoResearch Laboratories) on ice for 30 minutes. Data were acquired on LSRFortessa Cell Analyzer (BD Biosciences).

Hu-PD-1–expressing Jurkat cells were incubated with biotinylated human PD-L1 or PD-L2 at a saturating concentration of 20 μg/mL, followed by a 1-hour incubation on ice in the presence of serial dilutions of sasanlimab or a negative control IgG4 antibody. Cells were washed and incubated for 30 minutes on ice with PE streptavidin (1:100 dilution). Cy-PD-1–expressing Jurkat cells were incubated with 20 μg/mL Alexa Fluor 647 cynomolgus monkey PD-L1 or PD-L2 and treated with serial dilutions of sasanlimab or a negative control antibody. Data were acquired using LSRFortessa (BD Biosciences) Cell Analyzer (BD Biosciences).

For this purpose, we adapted a previously published protocol (29). Primary human DCs differentiated in vitro were harvested 7 days after initiation of the culture. Freshly isolated human T cells from allogeneic donors were plated with irradiated DCs at a ratio of 10:1 in the presence of different concentrations of sasanlimab or a negative control antibody. Cultures were incubated at 37°C with 5% CO₂ for 5 days. Supernatants were collected, and cytokine concentrations measured using Cytometric Bead Array (CBA, BD Biosciences). Data were acquired using LSRFortessa (BD Biosciences). Proliferation was measured in parallel cultures by adding 1 μCi of 3H methyl-titrated thymidine (PerkinElmer) to each well followed by an additional 18-hour incubation. Tritiated thymidine incorporation provided an index of cell proliferation measured as counts per minutes using the MicroBeta2 Machine (PerkinElmer). Where indicated, T cells were cultured with JeKo-1-PD-L1 cell line at a 5:1 ratio, and supernatants were collected 5 days later.

Whole blood from cynomolgus monkey was incubated for 1 hour at 37°C in 5% CO₂ in the presence of sasanlimab or a negative control antibody at 0.1–100 μg/mL. Initial incubation was followed by stimulation with 0.1 μg/mL Staphylococcal enterotoxin B (SEB, Toxic Technologies) and an additional 3-day incubation. Plasma was harvested, and the concentrations of IFNγ, TNFα, and IL2 were measured using MSD Platform according to the manufacturer's protocol (Meso Scale Diagnostics).

Immunocompromised NSG mice were injected with 1 × 10⁷ human PBMCs. Mice received injections of sasanlimab (0.1, 1, or 10 mg/kg) at 2 and 8 days after transfer. An isotype-matched control antibody was dosed at 10 mg/kg. Body weight change and survival were monitored daily. Where indicated, selected groups were sacrificed at day 19 to determine the effect of sasanlimab treatment on the number of CD3⁺ T cells in blood, spleen, and liver. In addition, human CD45⁺ cells were isolated from spleen and liver tissue, and 1 × 10⁶ cells were incubated for 8 hours with phorbol myristate acetate (PMA; 10 ng/mL) and ionomycin (125 ng/mL) to assess the production of human IFNγ in culture supernatants.

Hu-PD-1 knock-in mice (GenOway) were subcutaneously inoculated with 5 × 10⁵ MC-38 cells/mouse. Mice were randomized into treatment groups 9 days after engraftment, when tumor volumes were approximately 60 mm³. Sasanlimab or isotype control antibody was intraperitoneally injected at days 10, 13, 16, and 23 after tumor inoculation at 10 mg/kg. Body weight change and tumor size were recorded throughout the study.

To assess toxicity, sasanlimab was administered by intravenous bolus injection once per week (five total doses) to male and female cynomolgus monkeys (3/sex/group) at doses of 20, 60, or 200 mg/kg/week. For the control and 200 mg/kg/week dose levels, an additional 2 monkeys/sex/group were added to assess the potential reversal of toxicity following a 2-month recovery phase. Monkeys were euthanized at days 30 (main study) or 90 (recovery). Blood was collected for clinical pathology, immunophenotyping SEB assay, and pharmacokinetics. Tissues were collected for histologic analysis.

Pharmacokinetic studies were performed following a single intravenous dose administration at 0.1 and 1 mg/kg or as part of the 1-month toxicity study. Serum concentrations of sasanlimab were determined using MSD platform with monoclonal neutralizing antibody, in-house generated mouse-anti-sasanlimab clone F10 as a capture reagent, and goat anti-human IgG, Sulfo-Tag labeled (MSD, R32AJ-1) as a detection reagent. The range of quantitation in 100% matrix was 3.3–5,000 ng/mL. The pharmacokinetic parameters were determined using noncompartmental analysis in Watson LIMS (version 7.4, Thermo Fisher Scientific).

The availability of the free PD-1 receptor was assessed in whole-blood samples after a single intravenous injection of sasanlimab. The free PD-1 receptor was detected using anti-PD-1 (clone EH12.1, BD Biosciences), which competes with sasanlimab for the epitope binding. To further assess the in vivo blockade of PD-1 with sasanlimab, whole-blood samples were collected twice prior to the initiation of dosing, at days 14 and 28, and during recovery at days 43/42 (males/females) and 53/52 (males/females). Blood samples were cultured for 3 days in the presence or absence of 0.1 μg/mL SEB, or with 0.1 μg/mL SEB in the presence or absence of 10 μg/mL exogenously added sasanlimab. Plasma was collected for assessment of IFNγ concentrations using ELISA (Cell Sciences). The functional blockade of PD-1 was determined on the basis of the ratio of IFNγ concentrations in samples coincubated with SEB and sasanlimab and those incubated with SEB alone.

Statistical differences between two study groups were evaluated using an unpaired, two-tailed t test. Statistical differences between more than two groups were evaluated using a one-way ANOVA with Tukey multiple comparison post hoc test. Two-way ANOVA with Sidak multiple comparison post hoc test was used to assess comparison between more than two groups based on more than one parameter (multiple factors). Statistical significance was assigned as *, P < 0.05; **, P < 0.01; ***, P < 0.001; and ****, P < 0.0001. Statistical analyses were performed using GraphPad Prism 8 software.

Sasanlimab's (PF-06801591) binding and specificity for PD-1 proteins from different species were measured using the kinetic exclusion assay (KinExA) and SPR analyses. The KinExA method provided high sensitivity measurements of the affinity (K_D) in solution and showed a similar affinity of sasanlimab for hu-PD-1 and cy-PD-1 proteins. The K_D values at 23°C were 17 and 23 pmol/L for hu-PD-1 (when studied in two opposing assay orientations) and 28 pmol/L for cy-PD-1 (Table 1; Supplementary Fig. S8). Using SPR analysis, the K_D values of sasanlimab for hu-, cy-, and mu-PD-1 at 25°C were 42 pmol/L, 69 pmol/L, and 0.9 μmol/L, respectively. No binding to rat PD-1 was detected when tested at 0.5 μmol/L by SPR analysis (Table 1; Supplementary Fig. S9). K_D values determined by SPR analysis were up to two times higher at 37°C compared with 25°C for both hu-PD-1 and cy-PD-1. Low K_D values observed with sasanlimab were due to the fast on-rate (k_a) and slow off-rate (k_d; Table 1).

In different cell-based assays, sasanlimab bound to hu- and cy-PD-1 with high affinity and specificity. In HEK-293T transiently transfected cell line expressing hu-, cy-, mu-, or rat-PD-1, sasanlimab showed similar binding patterns, indicated by mean fluorescence intensity (MFI) values, to hu-PD-1 and cy-PD-1 (Supplementary Fig. S10A). Binding to mu-PD-1 was only achieved at a high, biologically irrelevant concentration of sasanlimab. No binding of sasanlimab was detected for rat PD-1, and minimal to no binding was observed in the parental cell line transfected with an empty vector (vehicle; Supplementary Fig. S10A). Notably, sasanlimab selectively bound to PD-1, but not to other proteins associated with T-cell activation and exhaustion, such as PD-L1, PD-L2, CTLA-4, LAG-3, TIGIT, TIM-3, CD28, and ICOS (Supplementary Fig. S10B).

Next, activated primary T cells were used to accurately estimate half maximal effective binding concentration (EC₅₀) values of sasanlimab binding to hu-PD-1 and cy-PD-1. The binding was determined 72 hours postactivation when PD-1 expression on the cell surface and cell viability were optimal (Fig. 1A). The EC₅₀ values of sasanlimab binding to both human and cynomolgus monkey activated T cells were low and similar (51.04 ± 5.07 and 85.34 ± 11.73 pmol/L, respectively; Fig. 1B). No binding was observed with the negative control antibody. Because different donors can express variable levels of PD-1 receptors upon activation, stably transfected cell clones expressing high levels of hu-PD-1 and cy-PD-1 were generated to verify EC₅₀ values.

Using Jurkat T cells stably transfected with hu-PD-1 or cy-PD-1, sasanlimab showed a high affinity for both human and cynomolgus monkey receptors (Fig. 1C). EC₅₀ values for the two species were similar to those obtained using primary activated T cells. EC₅₀ values for cy-PD-1–expressing cells were more variable than those for hu-PD-1 molecules in two experimental runs. This was due to high EC₅₀ values obtained in one of the experimental repeats; therefore, EC₅₀ values were 2- to 4-fold lower for hu-PD-1- compared with cy-PD-1–transfected cells. As the negative control antibody did not exceed baseline biding values, no EC₅₀ values were calculated. As molecular interaction of full IgG molecule with PD-1 is more applicable to physiologic setting, the affinity of sasanlimab Fab to hu-PD-1/cy-PD-1 was not determined.

The IgG4 antibody subclass was selected to ensure that the binding of sasanlimab would not mediate the depletion of tumor antigen–reactive T cells. Our data showed that binding kinetics and specificity values for sasanlimab over a range of Fc receptors, including FcγRs and FcRn, from both human (Supplementary Fig. S11A) and cynomolgus monkey (Supplementary Fig. S11B), recapitulated those values expected for an isotype-matched human IgG4 control. All other interactions were of weak affinities (K_D values ∼1 μmol/L or higher).

Binding to C1q is considered a surrogate for potential complement-dependent cytotoxicity (CDC; ref. 30). Sasanlimab binding to C1q was barely detectable, similar to the binding of the human IgG4 isotype-matched control. Sasanlimab binding was even lower than that of human IgG2, which is known to have a weak effector function (31). In contrast, positive control antibodies, human IgG1 and human IgG3, displayed high binding signals (Supplementary Fig. S12A).

To further interrogate the lack of sasanlimab ability to induce killing of T cells in vitro, an antibody-dependent cell-mediated cytotoxicity (ADCC) assay was performed with activated T cells (target cells) expressing high levels of PD-1 receptors and PBMCs (effector cells) expressing high levels of FcγRs. Sasanlimab showed minimal ADCC activity in multiple donors, in contrast to the high ADCC-specific lysis of the IgG1-positive control (Supplementary Fig. S12B). Thus, sasanlimab is unlikely to elicit ADCC of T cells expressing PD-1 in the tumor microenvironment.

To better understand the binding mode of sasanlimab to hu-PD-1, we crystallized the complex of PD-1 with the Fab fragment of sasanlimab (Fig. 2A; Supplementary Fig. S13A). The structure contains a single copy of the hu-PD-1/sasanlimab Fab complex in the asymmetric unit with the interactions between the Fab variable domain and hu-PD-1 clearly visible in the electron density (Supplementary Fig. S13B). The structure hu-PD-1 adopts is a canonical IgV-type, B-sandwich fold seen in previous structures (32, 33). The interface between sasanlimab and hu-PD-1 was dominated by hydrogen bonds and Van der Waals interactions with only two salt bridges between the kappa chain and PD-1 (Fig. 2B). Analysis of the complex with PISA showed that hu-PD-1 binding to sasanlimab heavy chain and kappa chains bury 443Å2 and 544Å2, respectively (Fig. 2C; ref. 34). Importantly, the binding site of sasanlimab on the surface of hu-PD-1 involves residues that have been previously identified as being critical to binding hu-PD-L1 (Fig. 2D; ref. 33). Thus, the binding of sasanlimab to hu-PD-1 would preclude interactions with PD-L1.

Sasanlimab blocked PD-1 interaction with PD-L1 and PD-L2 in human- and cynomolgus monkey–transfected cell lines (Fig. 3A). The pharmacologic potency of sasanlimab, measured as IC₅₀, was 880 and 1,058 pmol/L for human PD-L1 and PD-L2, respectively, with no statistically significant difference (Fig. 3A). Similarly, in the cynomolgus monkey system, IC₅₀ values were not statistically different and measured 943 and 839 pmol/L for PD-L1 and PD-L2, respectively (Fig. 3B). A negative control antibody did not change PD-1/PD-L1 or PD-1/PD-L2 binding in either human or cynomolgus monkey systems. Consistent with these cell-based assays, SPR analysis demonstrated that sasanlimab blocks the binding of hu-PD-1 to its ligands, PD-L1 and PD-L2 (Supplementary Fig. S14).

The ability of sasanlimab to enhance T-cell function was investigated in different in vitro cell culture systems. More specifically, sasanlimab reversed the inhibition of NFAT-responsive luciferase activity in hu-PD-1–expressing Jurkat T cells when cocultured with CHO cell line expressing hu-PD-L1. Relative to a control antibody, sasanlimab restored NFAT signaling in a dose-dependent manner (Fig. 4A).

Furthermore, in the human mixed lymphocyte reaction (MLR) system, primary human T cells isolated from healthy donors were mixed with allogeneic human DCs expressing high levels of hu-PD-L1. In the presence of sasanlimab, a dose-dependent increase of T-cell proliferation was observed when compared with the control antibody (Fig. 4B). Sasanlimab, at the 10 μg/mL concentration, increased the proliferation up to 2.5-times compared with the control antibody. T-cell activation was depicted by a dose-dependent increase in IFNγ and TNFα production (Fig. 4C). IFNγ and TNFα levels were increased up to 8-fold and 5-fold, respectively, compared with the control antibody treatment. IL2 expression was not detected in these cultures. To evaluate the potential increase of IL2 following sasanlimab treatment, an in vitro MLR assay was performed using human T cells mixed with a JeKo-1 cell line expressing high levels of human PD-L1. Sasanlimab treatment induced a dose-dependent increase of IL2 (up to 5-fold) compared with the control antibody (Fig. 4C). Because the binding affinity of sasanlimab to hu- and cy-PD-1 was similar, the effect of sasanlimab on cynomolgus monkey T-cell activation was also examined. Notably, a similar increase in the production of IFNγ and IL2 was observed after stimulation of cynomolgus monkey whole blood with SEB superantigen (Fig. 4D).

Multiple reports have indicated that the blockade of the PD-1 pathway augments GvHD in mouse models (35, 36). Therefore, a xenogeneic model of acute GvHD, in which human PBMCs were transferred to NSG immunocompromised mice, was utilized to test the effects of sasanlimab on human T cells in vivo. In this model, body weight loss was expected to accelerate between day 20 and day 30 after transfer (37–39). Indeed, sasanlimab accelerated body weight loss (Fig. 5A) and induced additional disease signs, such as hunched posture, ruffled fur, and decreased mobility, and in some animals, diarrhea, which were commonly seen in this model. Accordingly, sasanlimab-treated mice exhibited a dose-dependent decrease in survival compared with mice treated with the control antibody (Fig. 5B).

To evaluate the effects of sasanlimab treatment on T-cell frequency and function, two groups of mice received transfer of human PBMCs and were treated with either 10 mg/kg sasanlimab or control antibody. On day 19, at the onset of acute GvHD, human leukocytes were isolated from blood, spleen, and liver as representative target tissues of GvHD. Results showed that the relative abundance of human CD3⁺ T cells was elevated in blood, spleen, and liver of sasanlimab-treated mice (Fig. 5C). In addition, no free PD-1 was detected on the surface of human T cells upon sasanlimab treatment (Fig. 5D). To determine the effect of sasanlimab on T-cell function, human lymphocytes isolated from spleens and livers were stimulated ex vivo using a mixture of PMA and ionomycin for 8 hours. Our results showed significantly higher levels of IFNγ produced by lymphocytes isolated from sasanlimab-treated mice relative to the control group (Fig. 5E). Collectively, our results demonstrate the antagonistic activity of sasanlimab in human T cells.

To further investigate the antitumor effects of sasanlimab in an immunocompetent mouse system, we used hu-PD-1 knock-in mice (GenOway) expressing the hu-PD-1 protein. Dimeric hu-PD-1 was previously shown to bind human and mouse PD-L1 with identical K_D values (40). Our data confirmed the induction of hu-PD-1 and the absence of mu-PD-1 on the surface of tumor-infiltrating T cells 3 weeks after MC-38 inoculation (Fig. 5F). These T cells exhibited a heightened expression of additional coinhibitory proteins, such as TIM-3 (Fig. 5F). Treatment of MC-38–bearing mice with 10 mg/kg of sasanlimab or corresponding isotype control antibody at days 10, 13, 16, and 23 after tumor engraftments was not associated with any aberrant body weight changes or signs of toxicity during the study (Supplementary Fig. S15). Importantly, we observed a substantial delay in the growth of MC-38 tumors in sasanlimab-treated mice (Fig. 5G).

On the basis of the C_max and AUC₁₆₈ values, pharmacokinetics of sasanlimab was linear at doses of 1 mg/kg and higher. At 0.1 mg/kg dose, an accelerated elimination was observed, possibly due to the PD-1 target-mediated drug disposition. Because of the short duration of sample collection from the 1-month toxicity study, those data could not be used for the calculation of pharmacokinetic parameters (Fig. 6A). Therefore, the linear pharmacokinetic parameters in monkeys were calculated from 1 mg/kg dose. The clearance, volume of distribution, and half-life of sasanlimab in monkeys were similar to the typical human IgG and were estimated to be 0.26 mL/hour/kg, 45 mL/kg, and 4.4 days, respectively.

No free PD-1 receptor was detectable on the cell surface at 0.25 hours after dosing in sasanlimab-treated groups (Fig. 6B). In general, a higher sasanlimab dose was associated with a prolonged reduction in free PD-1 receptor levels. Free PD-1 receptor levels started returning to baseline after 4 and 11 days of treatment with sasanlimab at 0.1 and 1 mg/kg dose, respectively, with no difference observed between CD4⁺ and CD8⁺ T cells (Fig. 6B). In addition, to measure the functional blockade of PD-1 receptors, blood samples collected before and after dosing with sasanlimab were stimulated with SEB with or without the addition of 10 μg/mL of sasanlimab to the cultures. Under these conditions, if cell surface PD-1 receptors were already blocked by sasanlimab in the circulation from intravenous dosing, then adding sasanlimab to the culture would not increase the (SEB + sasanlimab)/(SEB) ratio. Results showed that adding sasanlimab into blood samples potentiated SEB-induced IFNγ production by blood lymphocytes that were not previously exposed to sasanlimab. However, spiking sasanlimab into blood samples from sasanlimab-dosed monkeys did not show any potentiation effect (Supplementary Fig. S16), suggesting that PD-1 was already blocked by sasanlimab dosed in vivo. This functional blockade occurred at all doses (20, 60, and 200 mg/kg) and at all timepoints during the dosing phase and was still observed 24 days following cessation of dosing at 200 mg/kg (Supplementary Fig. S16).

We also assessed downstream pharmacologic effects of sasanlimab in monkeys by evaluating changes in the expression of T-cell activation and proliferation markers. At all doses tested (20, 60, and 200 mg/kg), we observed an increase in the relative abundance of activated (HLA-DR⁺) and proliferating (Ki-67⁺) CD4⁺ and CD8⁺ T cells (Fig. 6C). Changes in T-cell activation and proliferation were not dose dependent, consistent with a complete blockade of the PD-1 pathway at all doses tested (Fig. 6B; Supplementary Fig. S16).

The combined toxicity study results are summarized in Supplementary Table S1. Overall, a diffuse pattern of mild immune cell infiltration was observed in all sasanlimab-treated monkeys following repeat dosing, consistent with an elevated pattern of immune surveillance, an expected mechanism of action of PD-1 inhibition. Immune cell infiltrates were detected in the choroid plexus and/or meninges at the end of the dosing phase. Because of the high dose and relatively long half-life, the exposure in the 200 mg/kg/week dose group at the end of the recovery period was similar to the exposure in the 20 mg/kg/week group during treatment. Because this exposure causes maximal pharmacologic activity, reversal was not evaluable.

Consistent with a prolonged exposure at pharmacologically active doses, at the end of the 2-month recovery phase, the increased immune cell infiltration was observed in the spinal cord, brain, peripheral nerve, pituitary gland, kidney, or thyroid tissue of individual monkeys. There was no associated tissue inflammation and/or damage altering the organ function. An additional finding in a single male monkey in the 200 mg/kg/week recovery group was mild perivascular mononuclear cell infiltration in the neuropil (brain), which was rarely associated with astrocytes or glial cells. However, this finding was not associated with any observable behavior changes or general functional deficits throughout the study.

Mechanistically, inhibition of PD-1 signaling induces enhanced tumor immune surveillance and antitumor immune responses (1, 3, 9). Agents targeting PD-1 and PD-L1 show efficacy in various tumor types and have been approved for multiple indications (15, 16). In this article, we presented the biophysical and functional properties of sasanlimab (PF-06801591), a humanized, hinge region–stabilized IgG4 anti-PD-1 mAb.

Sasanlimab binds with similar high affinities to both human and cynomolgus monkey PD-1 proteins and blocks the interaction between PD-1 and its ligands, PD-L1 and PD-L2. Of note, the binding of sasanlimab to hu- and cy-PD-1 is associated with the formation of a very stable complex, which is likely to be the main driver of this high-affinity interaction. We used MLR studies to create an in vitro setting resembling the tumor microenvironment where T cells express high levels of PD-1 in the presence of allogeneic DCs or tumor cells expressing PD-L1. PD-1/PD-L1 interaction inhibited T-cell proliferation and cytokine release. Consistent with previously published reports, treatment with sasanlimab markedly restored T-cell proliferation and effector function (40–42). The overall risk of T-cell receptor–independent systemic cytokine release in humans seems to be low as soluble or solid phase assays with sasanlimab cultured with human whole blood or PBMCs, respectively, did not show the release of IL6, TNFα, or IFNγ. Moreover, the IgG4 subclass of sasanlimab did not mediate ADCC or CDC activity and exhibited reduced binding to low affinity FcγRs, similar to published data (43, 44), suggesting a minimal risk for the depletion of antitumor effector T cells.

Sasanlimab displayed weak binding to mouse PD-1 protein only at high, biologically irrelevant concentrations. Therefore, we initially used a xenogeneic model for acute GvHD in NSG immunocompromised mice, previously engrafted with human PBMCs, as a surrogate to explore the biological functions of sasanlimab in vivo. In these studies, we confirmed the in vivo binding of sasanlimab to hu-PD-1 on T cells. Of interest, sasanlimab accelerated the incidence of GvHD by enhancing T-cell proliferation and proinflammatory cytokine secretion, supporting the antagonistic effect of sasanlimab on PD-1 receptor. We further demonstrated a significant delay in growth of mouse MC-38 colorectal tumor implanted in hu-PD-1 knock-in mice (40) upon treatment with sasanlimab. Collectively, our results further support the sasanlimab mechanism of action as an antagonist of the PD-1 pathway.

In-line with the published PD-1 expression pattern (45, 46), tissue cross-reactivity studies revealed staining in mononuclear cell types, including T and B lymphocytes, and myeloid cell subtypes. The similarity in sasanlimab staining patterns across human and monkey tissues, together with similarity in biochemical and cell-based interactions of the molecule with hu-PD-1 and cy-PD-1 justify the usage of monkeys as an appropriate model for evaluating potential human toxicities. The mean systemic exposure of sasanlimab increased with increasing dose and was higher after repeated dosing in cynomolgus monkeys. The incidence of antidrug–antibody was estimated to be approximately 4.5% across all treated groups after repeated dosing. However, the presence of such antibodies was not associated with any adverse events. The actions of marketed PD-1 inhibitors have been previously described, suggesting the increased immune cell infiltration likely to be a secondary effect of enhanced immune response related to the primary pharmacology of PD-1 inhibition. The mononuclear infiltration observed after sasanlimab administration in monkeys was similar to the findings previously shown for the marketed PD-1 antagonist drugs, nivolumab (47) and pembrolizumab (48), where inflammatory cell infiltrates were observed in several tissues, including the brain (choroid plexus and meninges).

Altogether, the nonclinical profile of sasanlimab has been well-characterized, therefore, strongly supporting its clinical development for the treatment of patients with advanced cancers. Indeed, clinical testing of sasanlimab has begun in a phase I open-label, multi-center, multiple-dose, dose escalation, safety, pharmacokinetic, and pharmacodynamic study in patients with locally advanced or metastatic solid tumors (22). This is the first study that demonstrates the feasibility of subcutaneous administration of an anti-PD-1 antibody. Of note, the safety and tolerability of intravenous and subcutaneous administration of sasanlimab were comparable, and objective responses were seen with both intravenous and subcutaneous dosing. Clinical safety and efficacy of sasanlimab are similar to published data of other PD-1 checkpoint inhibitors (49, 50). It will also be of interest to interrogate how sasanlimab distinguishes in comparison with other relevant PD-1 checkpoint inhibitors in the antibody epitope, biophysical properties, and biological activity. Taken together, the preclinical findings and the recent clinical results support the further clinical characterization of sasanlimab as a promising immunotherapeutic agent.

S. Youssef reports other from Pfizer (hold stock/stock options in the company) outside the submitted work, as well as has a patent for US10155037B2 issued (this is the paten that has the sequences of this article). Y. Abdiche reports other from Pfizer (hold stock/stock options in the company) outside the submitted work, as well as has a patent for US10155037B2 issued (anti-PD-1 antibodies and methods of use thereof). C.R. Kimberlin reports other from Pfizer (hold stock/stock options in the company) outside the submitted work. S.M. Chin reports former employment with Pfizer Inc. C. Kamperschroer reports other from Pfizer (hold stock/stock options in the company) outside the submitted work. B. Kern reports other from Pfizer (hold stock/stock options in the company) outside the submitted work. N. Budimir reports other from Pfizer (hold stock/stock options in the company) outside the submitted work. C.P. Dillon reports employment with Pfizer. J. Chou reports other from Pfizer (hold stock/stock options in the company) during the conduct of the study and outside the submitted work. E. Kraynov reports other from Pfizer (hold stock/stock options in the company) outside the submitted work. A. Rajpal reports a patent for US20160159905A1 pending to Pfizer (application). S. Salek-Ardakani reports other from Pfizer (hold stock/stock options in the company) outside the submitted work. No potential conflicts of interest were disclosed by the other authors.

A.A. Al-Khami: Resources, data curation, formal analysis, validation, investigation, visualization, methodology, writing-original draft, writing-review and editing. S. Youssef: Conceptualization, data curation, formal analysis, validation, investigation, visualization, methodology. Y. Abdiche: Conceptualization, resources, data curation, software, formal analysis, supervision, validation, investigation, visualization, methodology, writing-review and editing. H. Nguyen: Data curation, formal analysis, investigation, methodology, writing-review and editing. J. Chou: Resources, data curation, formal analysis, investigation, visualization, methodology, writing-review and editing. C.R. Kimberlin: Resources, data curation, software, formal analysis, validation, investigation, visualization, methodology, writing-review and editing. S.M. Chin: Resources, formal analysis, supervision, validation, investigation, visualization, methodology, writing-review and editing. C. Kamperschroer: Resources, data curation, formal analysis, supervision, validation, investigation, visualization, methodology, writing-original draft, writing-review and editing. B. Jessen: Resources, data curation, formal analysis, supervision, validation, investigation, visualization, methodology, writing-review and editing. B. Kern: Resources, data curation, formal analysis, validation, investigation, visualization, methodology, writing-review and editing. N. Budimir: Resources, data curation, formal analysis, validation, investigation, visualization, methodology, writing-original draft, writing-review and editing. C.P. Dillon: Resources, formal analysis, investigation, visualization, methodology, writing-review and editing. A. Xu: Resources, formal analysis, validation, investigation, visualization, methodology, writing-review and editing. J.D. Clark: Resources, data curation, formal analysis, supervision, investigation, visualization, methodology, writing-review and editing. J. Chou: Formal analysis, writing-review and editing. E. Kraynov: Formal analysis, supervision, investigation, visualization, methodology, writing-original draft, writing-review and editing. A. Rajpal: Conceptualization, resources, data curation, formal analysis, validation, investigation, visualization, methodology, project administration, writing-review and editing. J.C. Lin: Conceptualization, formal analysis, supervision, project administration, writing-review and editing. S. Salek-Ardakani: Conceptualization, data curation, formal analysis, supervision, writing-original draft, project administration, writing-review and editing.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.