Background

Increased numbers of tumor-infiltrating CD3 T cells and a high M1/M2 ratio of tumor-associated macrophages in newly diagnosed advanced ovarian cancers have independent prognostic value for improved survival.1–3 Tumor-infiltrating lymphocytes are suppressed and functionally exhausted by a variety of mechanisms, including abundant immunosuppressive myeloid cells.4 5 In the context of immunosuppression, PD-1/L1 immune checkpoint inhibitors have shown limited benefit in patients with ovarian cancer either as monotherapy or in combination, despite the expression of PD-L1 in ovarian cancer.6 7 With the exception of the recent approval of the folate receptor alpha-targeted antibody drug conjugate, mirvetuximab soravtansine, the mainstays of therapy for ovarian cancer patients include platinum, taxanes, anthracyclines, and poly-ADP ribose polymerase inhibitors, and improvements in overall survival over the past several decades have been incremental.

CD47 and CD40 are two immune receptors that are frequently expressed by myeloid cells and are relevant in ovarian cancer. Among all solid tumors, expression of the macrophage checkpoint receptor CD47 is highest in ovarian cancer,8 9 suggesting that the fitness of ovarian cancer cells may be improved if they can defend themselves from macrophage-mediated phagocytosis. In phase I clinical trial in advanced solid tumors, two monotherapy responses were observed in ovarian cancer patients treated with a CD47-blocking antibody, magrolimab.10 SL-172154 was specifically designed with an inactive Fc-domain to prevent hemolytic anemia that has been reported with anti-CD47 antibodies including magrolimab.11 CD47 blockade enables macrophage-mediated phagocytosis of tumor cells if sufficient prophagocytic signals are present, but tumor shrinkage requires antigen presentation to T cells.12–16 CD40 is an immune-activating TNF superfamily receptor that is most abundantly expressed by antigen-presenting cells (including myeloid cells) and endothelial cells, and stimulation of CD40 provides innate immune activation that is supportive of antigen presentation and activation of an adaptive T-cell response.12 17–19 In preclinical studies, a murine SIRPα-Fc-CD40L fusion protein was shown to enhance antigen presentation, and CD8 T-cell-mediated antitumor immunity in comparison to CD47-blocking or CD40 agonist antibodies alone or in combination.11

CD40 agonist therapeutics first entered clinical development over 20 years ago, and most were discontinued in phase I clinical trials due to several reasons: toxicity at low doses, insufficient evidence of immune activation, and atypical “bell-shaped” dose-response curves even at μg/kg dose ranges. These agents and studies have been recently reviewed20; most agents used monovalent or bivalent antibody-derived CD40-binding domains and required cross-linking either through Fcγ receptors or tumor-associated antigens to facilitate trimerization of CD40.21 Like other TNF receptors, CD40 requires trimerization to signal, but hexamerization and higher-order network formation facilitate more potent activation. SL-172154 is structurally distinct from prior CD40 agonists in that it contains two preformed CD40L trimer domains that facilitate ligand-induced CD40 receptor trimerization without a requirement for secondary cross-linking. This structural distinction was predicted to provide dose-dependent CD40 activation without observing bell-shaped pharmacodynamic (PD) effects as the dose approached and exceeded receptor saturation.

Consistent with the mechanism of action, both CD47- and CD40-targeted agents have been developed in combination with agents that either induce a prophagocytic signal on tumor cells (eg, chemotherapy or antibody drug conjugates) or induce antibody-dependent phagocytosis (monoclonal antibodies or antibody drug conjugates). Here, we report the results from this phase I, open-label, multipart study to evaluate the safety, tolerability, pharmacokinetics (PKs), PDs, and initial antitumor activity of SL-172154 as monotherapy in patients with platinum-resistant ovarian cancer. This study enabled the selection of a recommended dose of SL-172154 monotherapy for further development in combination with other agents. The selection of dose was based on safety and PK and PD activity, which maximized the biology of CD47 inhibition and CD40 agonism.

MethodsStudy design

Study SL03-OHD-101 (NCT04406623) was an open-label, multicenter, first-in-human, phase I study of SL-172154 (SIRPα-Fc-CD40L) in patients with platinum-resistant ovarian cancer. This dose-escalation study had a modified toxicity probability interval-2 design to evaluate five dose levels (0.1, 0.3, 1.0, 3.0, 10.0 mg/kg) and two dosing schedules. On schedule 1, SL-172154 was administered by intravenous infusion on D1, D8, and D15 of a 28-day cycle and then once every 2 weeks (D1, D15) in 28-day cycles starting at cycle 2, while administration of study drug on schedule 2 was once weekly (D1, D8, D15, and D22) in 28-day cycles. Dose levels and schedules are described in online supplemental figure 1.

The primary objective of this study was to evaluate the safety and tolerability of SL-172154 and to identify the maximum tolerated dose (MTD) of SL-172154 as measured by adverse events (AEs), serious adverse events (SAEs), fatal SAEs, dose-limiting toxicities (DLTs), AEs leading to discontinuation, and changes in safety assessments (eg, laboratory parameters, vital signs). Secondary objectives included selection of the dose for SL-172154 in future studies, and assessment of preliminary evidence of antitumor activity as measured by objective response rate, time to response, and duration of response. An evaluation of immunogenicity to SL-172154, measured by antidrug antibody (ADA) titers, and characterization of PK, were also objectives of this study. Exploratory objectives included assessment of PD biomarkers such as CD47 receptor saturation on CD4 T cells, serum cytokines, flow cytometry of blood cells, and immunohistochemistry (IHC) of tumors.

Study population

Patients with histologically confirmed unresectable, locally advanced or metastatic ovarian cancer, primary peritoneal cancer or fallopian tube cancer, deemed platinum resistant or ineligible for further platinum therapy, and who had measurable disease per Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 were eligible for the study. Additional eligibility criteria included: age of 18 years or older with an Eastern Cooperative Oncology Group performance status of 0 or 1, and adequate hematological, hepatic, renal, and cardiac function. Key exclusion criteria were primary platinum refractoriness as defined by progression during or within 1 month of upfront platinum therapy, prior treatment with an anti-CD47- or anti-SIRPα-targeting agent or a CD40 agonist, documented history of autoimmune disease or active pneumonitis, and concurrent use of systemic corticosteroids or other immunosuppressive medication.

Safety and tolerability

Toxicities were graded as per the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) v5. The determinate period for a DLT was the first 28 days of SL-172154 dosing on schedule 1 or schedule 2. Treatment-emergent AEs (TEAEs) clearly related to disease progression or intercurrent illness were not considered DLTs. For grade 2 or 3 infusion-related reaction (IRR) or cytokine release syndrome (CRS), the SL-172154 infusion must be interrupted, and the patient must be treated per guidance. The infusion may be restarted on resolution at 50% of the rate with close monitoring. For grade 2 or 3 liver enzyme elevations, SL-172154 treatment must be withheld until resolution to grade ≤1. For grade 4 IRR, CRS, or liver enzyme elevations SL-172154 must be permanently discontinued.

PK assessments

Serial PK samples were collected for all patients. The non-compartmental PK analysis for SL-172154 serum concentrations was estimated using Phoenix WinNonlin (Certara; Philadelphia, PA). Accumulation ratio was defined as the ratio of maximum observed concentration over a dosing interval (Cmax) or area under the curve (AUC) at cycle 1 day 15 versus cycle 1 day 1, or at cycle 2 day 1 versus cycle 1 day 1. As the accumulation of SL-172154 after multiple doses was not observed (geometric mean of accumulation ratio <1.6 for both Cmax and AUC), schedules 1 and 2 were pooled for data presentations. Additional PK methodology is available in online supplemental methods.

Immunogenicity assessment

Predose blood samples were collected for determination of ADAs starting on day 1 (predose) and then on cycle 1 day 8, day 15, cycle 3 day 1, and day 1 of cycles 4, 7, 10, 13, and 24 and at the post-treatment visit. An unscheduled ADA sample was collected if a patient experienced an IRR. ADA response was defined as treatment-induced ADA in ≥2 consecutive samples without a subsequent negative sample or only in the last sample. Additional methodology is available in online supplemental methods.

PD assessmentsCD47 receptor occupancy

CD47 receptor occupancy by SL-172154 on CD4+ T Cells and CD40 receptor engagement by SL-172154 on B lymphocytes and CD40 expressing monocytes was measured by quantitative multiparameter flow cytometry assay. Additional methodology is available in online supplemental methods.

Serum cytokines

The levels of cytokines were measured in serum samples collected at predose, 2, and 24 hours post-end of infusion (post-EOI) on cycle 1 day 1, day 15, and cycle 2 day 1, day 15. Ad hoc samples were collected in patients who experienced an IRR. The multiplex kits from Meso Scale Diagnostics (MSD) (Rockville, MD), included the V-PLEX Plus Chemokine Panel 1, V-PLEX Plus Proinflammatory Panel 1 Human Kit, V-PLEX Plus Cytokine Panel 1 Human Kit, and V-PLEX Th17 Panel 1 Human Kit for a total of 37 chemokines and cytokines. The analysis was performed according to the manufacturer’s procedure.

Immunophenotyping

Expression of phenotypic and activation markers and the absolute number of cells were assessed by multiparameter flow cytometry. The frequency of CD95+ and CD86+ B cells, activated CD40+ monocytes (Lin-HLA-DR+CD14+CD16−) and CD4+ T cells were examined. See online supplemental methods.

Immunohistochemistry

To assess the pharmacological effect of SL-172154 in the tumor microenvironment, tumor biopsies were obtained at baseline and cycle 1 day 16–22 and analyzed by IHC using two multiplex immunofluorescence assays for the detection of cytotoxic T cells (Granzyme B, CD8, Ki-67, CD3, pan-cytokeratin) and detection of macrophage subsets (CD80, CD206, MHCII, CD68, and PanCK). Type 1 or M1 macrophages or inflammatory macrophages were defined as CD68+MHCII+ and CD68+CD80+, and type 2 or M2 macrophages or alternatively activated macrophages were defined as CD68+ and CD206+.

Efficacy assessments

All patients treated with at least one dose of SL-172154 (“All-Treated Population”) were evaluated for efficacy according to RECIST v1.1. Disease assessment was performed at baseline and every 8 weeks through week 24, and every 12 weeks thereafter until year 2, and every 24 weeks or until disease progression.

Statistical analyses

All data summaries were presented by dose schedule/level based on the all-treated population. Summaries were also provided for pooled data on patients at the same dose level in either schedule 1 or schedule 2. A DLT summary was provided for the DLT-evaluable population. All AEs with onset dates on or after the first dose of study treatment were defined as TEAEs. Drug-related AEs were defined as TEAEs with relationship to study treatment being related or possibly related.

PK/PD modeling methods

PK/PD modeling of SL-172154 was performed across a range of SL-172154 doses and included B-cell margination and interleukin 12 (IL-12) production in the blood after each dose. Additional methodology available in online supplemental methods.

ResultsPatient characteristics, disposition, and study drug exposure

A total of 27 patients were enrolled and received SL-172154, and an additional 7 patients were screen failures. All 27 patients were included in the PK population, 20 patients were included in the DLT-evaluable population, and 22 patients were in the PK/PD modeling population. Nine patients were treated on schedule 1 (0.1 mg/kg, n=3; 0.3 mg/kg, n=3; 3.0 mg/kg, n=3) and 18 patients were treated on schedule 2 (0.3 mg/kg, n=3; 1.0 mg/kg, n=4; 3.0 mg/kg, n=6; 10.0 mg/kg, n=5). 25 patients ended treatment due to disease progression based on radiological or clinical assessment, and 2 patients withdrew from the study (online supplemental figure 1). The median age of patients in the study was 66 years (range, 33–85 years) and of the 27 patients treated, 19 patients (70.4%) had ovarian cancer, 4 patients (14.8%) had primary peritoneal cancer, and 4 patients (14.8%) had fallopian tube cancer (table 1). The breakdown by race and ethnicity is limited by the small sample size: 21 (78%) of patients were white, 3 (11.1%) were black, 1 (3.7%) was Asian, and no patients reported Hispanic ethnicity (table 1).

Table 1

Patient demographics and disease characteristics

The median number of prior systemic regimens was 4 (range, 2–9). Most patients had prior surgical treatment (92.6%). The median number of intravenous doses of SL-172154 received during the study was 7 (range: 2–33), and the median treatment duration was 8 weeks (range, 2–36 weeks). For 24 patients (88.9%), the relative dose intensity of SL-172154 was >80% to ≤110%, while 3 patients (11.1%) had relative dose intensity of ≤80% due to missed doses as a result of TEAEs (intestinal obstruction, sepsis, and alanine aminotransferase (ALT)/aspartate aminotransferase (AST) increase).

Safety

Similar incidence and severity of TEAEs were observed between schedule 1 and schedule 2 at the two dose levels (0.3 and 3.0 mg/kg) evaluated in both schedules; thus, the results of patients are described by dose, regardless of schedule, throughout this paper. All 27 patients (100%) experienced a TEAE, with 24 patients (88.9%) having a TEAE considered by the investigator to be drug related (table 2, online supplemental table 1).

Table 2

Drug-related TEAEs (any grade in >10% patients and all grade 3 or 4)

12 patients (44.4%) had grade 3/4 TEAEs, and half of these patients (6 patients, 22.2%) had a drug-related grade 3/4 TEAE. No fatal TEAEs were reported. The drug-related grade 3 or 4 TEAEs were AST increased (n=2), lymphopenia (n=2), neutropenia, thrombocytopenia, anemia, ALT increased, IRR, back pain, and muscular weakness, each in one patient. These events occurred at doses of 3.0 or 10.0 mg/kg. The liver enzyme elevations and cytopenias fully resolved with no dose modifications. The event had not resolved for the patient with anemia who was found to have occult bleeding from a metastatic lesion in the gastrointestinal tract. There were seven serious TEAEs (all grade 3, not considered treatment related) reported in six patients (22.2%): embolism (2), sepsis (2), large intestine infection, lower gastrointestinal (GI) hemorrhage, and small intestinal obstruction. No TEAEs led to dose reduction or permanent discontinuation of SL-172154. 10 patients had a TEAE that led to a dose not being given or being delayed, with the most common reason being fatigue (n=4). Among the 20 patients in the DLT-evaluable population, one patient at 10.0 mg/kg experienced a DLT of grade 3 ALT increase. The event recurred with repeat dosing but resolved to grade ≤1 within 7 days on each occasion without intervention.

The most common drug-related TEAE was IRR, in which a total of 45 events were reported in 18 patients (66.7%). All drug-related IRRs were grade 1 or 2 events except for one grade 3 IRR. Infusion was interrupted for the grade 3 IRR, the patient was treated with steroids, and the event resolved the same day. Severity and frequency of IRRs events were dose dependent and correlated with rate of infusion. At ≤3.0 mg/kg, IRRs were typically limited to cycle 1 day 1 (C1D1) and C1D8. At 10.0 mg/kg IRRs were recurrent, occurring beyond C1D1 and C1D8. All IRRs occurred either during the infusion or within 2 hours after the EOI, except one event which occurred between 2 and 24 hours after the EOI. Common symptoms of the IRR events variably included fever, chills, rigors, back pain, rash, hyper/hypotension, tachycardia, nausea, and visual symptoms. No IRRs led to discontinuation of SL-172154, and all infusions associated with an IRR were completed, with the exception of two infusions.

Pharmacokinetics

A summary of the SL-172154 PK parameters at C1D1 for the PK population is provided in online supplemental table 2. Cmax ranged from 117 to 87 509 ng/hour over the dose range of 0.1–10.0 mg/kg. Following the first (C1D1) intravenous infusion of 0.1–10.0 mg/kg, serum SL-172154 concentrations showed multiphasic decline and quickly returned to below the limit of quantitation 8–24 hours post-EOI (figure 1A). The geometric mean derived from non-compartmental analysis Cmax and AUC0-last increased, with increasing doses following single intravenous infusions. Descriptive analysis of dose-normalized PK parameters illustrated that SL-172154 exposure increased in a greater than proportional manner between the 0.3 and 10.0 mg/kg dose ranges, and was most apparent between the 3.0 and 10.0 mg/kg dose levels (figure 1B,C). Similar trends were observed following multiple intravenous infusions.

Figure 1

Pharmacokinetics of SL-172154. (A) Cycle 1 day 1 (C1D1) concentration-time profile, (B) C1D1 dose-normalized Cmax, (C) C1D1 dose normalized AUC0-last. LLOQ, lower limit of quantification.

Immunogenicity

Out of the 25 patients who were ADA negative or missing at baseline, 8 patients (32.0%) had an ADA response anytime post cycle 1 day 1. Two patients were ADA positive at baseline. The median onset of treatment-induced ADA was 50 days (range, 15–169 days), and the median duration of treatment-induced ADA was 43 days (range, 8–239 days). There was no dose-response in the occurrence of ADA, and the titer of ADA was low (ranging from <10 to 1280); 7 of 8 patients had values being ≤320. PK profiles were available through cycle 2 day 1, while ADA occurred at later timepoints. In patients who became ADA positive on cycle 1 day 15 (n=2) and on cycle 2 day 1 (n=2), there was no evidence of increased study drug clearance as observed in the Cmax and AUC.

PD assessmentsCD47 receptor occupancy

There was a dose-dependent increase in CD47 receptor saturation on CD4+ T cells in the peripheral blood that approached 100% by ≥3.0 mg/kg dose. At 10.0 mg/kg, the saturation levels trended lower to what is observed at 3.0 mg/kg and this low saturation could be due to several factors, including dose interruption and CD4 T-cell margination from the circulation (figure 2).

Figure 2

CD47 receptor saturation on CD4+ T cells. + is median, o is individual patient value. (A) B-cells, (B) CD40 monocytes, (C) CD4 T cells.

B-cell and monocyte cell margination

CD40 is expressed on the majority of B cells and a proportion of monocytes. B cells, and CD40-expressing monocytes rapidly marginated from the circulation postinfusion in a dose-dependent manner. Figure 3 shows the percentage change from predose at 1-hour and 24-hour postinfusion on C1D1. Maximal margination of B cells and monocytes was observed at 3.0 and 10.0 mg/kg doses.

Figure 3

SL-172154 stimulates dose-dependent margination of B cells (A), CD40+ monocytes (B), and CD4+ T cells (C). Individual patient level data.

Serum cytokines

A dose-dependent and cyclical increase in IL-12, a mediator of TH1 proinflammatory responses, was observed in patients (figure 4A,B). The IL-12 concentrations achieved at 3 mg/kg on cycle 1 day 1 were in the range of 295–4450 pg/mL. Dose-dependent increases of other on-target cytokines, including CCL-22 (MDC), CXCL-8 (IL-8), CCL-20 (MIP-3a), IL-10, IL-15, and IL-27 (online supplemental figure 2), were also observed following SL-172154 administration and, similar to IL-12, these cytokine responses were cyclical with each infusion. Moreover, CXCL-10, a chemokine produced in response to IFN-γ, was also produced in a dose-dependent manner, underpinning the involvement of IFN-γ-producing cells such as NK and T cells. The CXCL-10 concentrations achieved at 3.0 mg/kg were in the range of 139–4410 pg/mL. All cytokines peaked 2 hours post-EOI and returned to baseline 24 hours post-EOI except for MDC, which peaked at 24 hours post-EOI. The difference of the fold change in cytokine production was higher between 1.0 mg/kg and 3.0 mg/kg than between 3.0 mg/kg and 10.0 mg/kg, indicative of a plateauing at 3.0 mg/kg. No bell-shaped dose response was observed. Other cytokines such as IL-6, IL-10, TNF-α, and IL-1β that have been shown to be involved in CRS were induced to levels slightly above the lower limit of quantification.

Figure 4

(A) Dose-dependent and (B) cyclical IL-12 cytokine responses to SL-172154. IL-12, interleukin 12.

Polarization of M1 macrophages and infiltration of T cells in paired tumor biopsies

Paired biopsies were available for 12 patients (0.1 mg/kg (n=1), 0.3 and 1.0 mg/kg (n=3 each), 3.0 mg/kg (n=4), and 10.0 mg/kg (n=1)). Due to the limited number of paired biopsy samples, the samples were grouped for dose levels 0.1 mg/kg, 0.3 mg/kg and 1.0 mg/kg (lower-dose levels) versus 3.0 mg/kg and 10.0 mg/kg (higher-dose levels) based on the observed plateau in serum IL-12 and B cell margination at ≥3.0 mg/kg. At the higher-dose levels, the on-treatment biopsies relative to baseline demonstrated enrichment for classical M1 macrophages (CD68+CD80+ cells/mm2, CD68+MHCII+ cells/mm2) compared with M2 macrophages (CD68+CD206+ cells/mm2), in the stroma and tumor nest (figure 5 top panels, online supplemental figure 5). In contrast, at lower-dose levels, the M2-dominant phenotype persisted before and after treatment. This observation suggests that higher doses of SL-172154 induced a shift in macrophages differentiation from a suppressive M2- to an inflammatory M1-dominant phenotype. The frequencies of granzyme B-producing, cytotoxic T cells similarly increased in the tumor nest and stroma in the higher-dose group compared with the lower-dose group, with the magnitude of change being higher in the tumor nest compared with the stroma (figure 5 bottom panels). All patients with a M1-dominant phenotype had an increase in infiltration of CD8 T cells in the tumor and stroma. The aggregate data and individual patient level data for the four patients in the 3 mg/kg group is provided in online supplemental figure 4.

Figure 5

SL-172154 induces polarization to M1 macrophages and infiltration of cytotoxic T cells in the stroma and tumor nest (the bar graphs represent the mean of the fold change and the SE).

Antitumor activity

The best response among the 27 treated patients with postbaseline scans was stable disease in 6 (22%) patients. The median duration of stable disease was 138 days (range, 50–252 days).

PK/PD modeling

PK/PD modeling was performed using both the B cell margination and IL-12 production described above. The models demonstrated that near maximal margination of B cells was reached at 3.0 mg/kg and did not change at 10.0 mg/kg (online supplemental figure 6A). Similarly, the production of IL-12 was dose dependent, with a trend towards a plateau at >3.0 mg/kg (online supplemental figure 6B).

Discussion

SL-172154 was overall well-tolerated in patients with platinum-resistant ovarian cancer. An MTD was not defined at doses up to 10.0 mg/kg, the maximum administered dose. Infusion related reaction was the most common TEAE. In subsequent studies with SL-172154, premedication with dexamethasone led to a reduction in incidence and severity of IRR (unpublished data). One patient at 10.0 mg/kg experienced DLT due to grade 3 ALT and AST elevations. The recommended SL-172154 dose of 3 mg/kg saturated both CD40 and CD47, and induced dose-dependent immune cell activation in peripheral blood and tumor microenvironment, rapid increases in multiple serum cytokines, and trafficking of CD40-positive B cells and monocytes following each infusion, consistent with the mechanism of action.

The CD40 agonist antibody, mitazalimab, was not dose escalated beyond 2 mg/kg, despite premedication with corticosteroids, due to dose-limiting toxicities. Another CD40 agonist antibody, selicrelumab (CP-870893), defined an MTD of 0.2 mg/kg as a result of dose-limiting toxicities of CRS and urticaria, was observed at 0.25 mg/kg.22 Similarly, sotigalimab reported a high frequency of grade 3 or 4 IRR/CRS when dose escalated to 0.6 and 1.0 mg/kg. Mitazalimab had dose-limiting toxicities of transaminase elevations at 1200 µg/kg.23 Finally, intravenous infusion of ChiLob7/4 also encountered dose-limiting toxicities of transaminase elevations at ~2–3 mg/kg.24 Preclinical studies have suggested that the toxicity observed with CD40 agonist antibodies was in part due to the requirement of those antibodies to engage Fc-gamma receptors, and to immune complex formation of CD40 antibodies within the liver.25–27 The relative tolerability of SL-172154 in comparison to prior CD40 agonist antibodies may be related to the hexameric structure of the CD40L domains in SL-172154 and the absence of Fc-gamma receptor binding function within the Fc domain.

Fc-gamma receptor binding has also contributed to toxicities in patients treated with CD47 inhibitory antibodies and fusion proteins. A high rate of hemolytic anemia was observed in acute myeloid leukemia and higher-risk myelodysplastic syndrome patients treated with magrolimab, a CD47-targeted antibody that includes an IgG4-derived Fc domain with Fc-gamma receptor binding function.28 29 Anemia and other cytopenias have been reported for a number of other CD47 inhibitors, including TTI-621, CC-90002 and AO-176, all of which have since been discontinued from clinical development.30–33 These agents were designed with Fc-gamma receptor binding function in order to mediate antitumor activity of a CD47 inhibitor as a monotherapy, with localization of both CD47 inhibition (blocking the “don’t eat me”) and Fc-gamma receptor-mediated prophagocytic (“eat me”) signals within a single compound.34 The tradeoff with this approach is secondary to binding and destruction of non-cancer cells expressing CD47. A growing body of clinical evidence now suggests that a sufficient therapeutic window may not exist for CD47 inhibitors with Fc-gamma receptor binding, particularly for hematological malignancies wherein anemia and other cytopenias may be present at baseline. As a result of these considerations, SL-172154 was designed with an Fc domain which lacks Fc-gamma receptor binding to avoid anemia and other cytopenias, and no evidence of hemolytic anemia or other destructive cytopenias was observed in patients treated with SL-172154. Changes relative to baseline in peripheral blood counts are transient and considered to reflect PD effects of CD40 agonism.35

Infusion-related reactions were the most common treatment-related adverse event observed in patients treated with SL-172154. Patients were not premedicated with corticosteroids prior to the first infusion, and IRRs were commonly observed during infusion of SL-172154. Premedication with corticosteroids mitigated IRRs on subsequent infusions for patients where IRRs were previously observed in the absence of corticosteroid premedication. No correlations between IRRs and cytokine concentrations could be found, despite the high absolute concentration of many cytokines. The magnitude of cytokine elevations was similar regardless of IRR. Specific attention was given to potential elevations of serum IL-6 and TNFα, because these cytokines have been associated with CRS for a number of immune agonists, including selicrelumab; however, neither consistent nor highly elevated IL-6 or TNFα concentrations were detected in the serum of patients treated with SL-172154. The absolute concentration of IL-12, CCL2, CCL4, CCL20, CXCL8, and CXCL10 consistently rose into the μg/mL range within hours of infusing SL-172154. The rapid elevation in these serum cytokines suggests that the mechanism was via CD40-mediated release of preformed intracellular cytokines rather than by de novo synthesis. Both the cytokine profile and magnitude of cytokine elevations for patients treated with SL-172154 were distinct from observations reported for patients treated with mitazalimab, selicrelumab, sotigalimab, or ChiLob7/4. Further, no evidence of a bell-shaped dose–response relationship was observed in patients treated with SL-172154, suggesting that hexameric display of CD40L may enable full activation of CD40 signaling at receptor-saturating concentrations. Finally, elevations in serum cytokines were observed following each weekly infusion of SL-172154, suggesting that desensitization of CD40 did not occur within a weekly interval.

The increase in SL-172154 exposure (Cmax and AUC) was nonlinear, and the pattern of non-linearity in the concentration-time profiles is consistent with that of target-mediated drug disposition. SL-172154 binds to both CD47 and CD40 receptors, and the observed PK pattern may be a consequence of CD40-mediated PD effects, including rapid cell egress and receptor-mediated endocytosis. Of note, CD40 agonist antibodies have a short serum half-life, which is most likely related to binding and internalization of the drug on widely distributed CD40 receptor-expressing cells.23 24 36 Despite the short half-life, potent PD activity was observed. The ADA titer was low in the majority of ADA-positive patients. There was no clear association between occurrence of IRRs and presence of ADA and no evidence of increased drug clearance.

It has been shown that the engagement of CD40 leads to migration of CD40-expressing cells from the peripheral circulation to secondary lymphoid organs.37 38 In the current study, we assessed the effect of CD40 target engagement on B cells (majority of B cells express CD40), and CD40-expressing monocyte margination. Binding of SL-172154 to CD40+ B cells and monocytes led to rapid activation (online supplemental figure 3) and dose-dependent margination of B cells and CD40-expressing monocytes from the circulation postinfusion. Margination was >90% at doses of 3.0 and 10.0 mg/kg. This phenomenon of cells marginating from the circulation may explain the variability in the estimated CD40 target engagement observed across the dose range.

Additionally, high CD47 receptor occupancy by SL-172154 on CD4+ T cells was maximal at 3 mg/kg. The dose-dependent margination of CD4 T cells due to a bystander effect of cytokine production may explain the variability in CD47 receptor occupancy. Engagement of CD40 on APC promotes antigen processing and presentation, activation of T cells, and production of proinflammatory cytokines among which IL-12, a key initiator of Th1 differentiation, is an inducer of IFN-γ release by NK and T cells. We observed a dose-dependent increase in CXCL-10, which is induced by IFN-γ, demonstrating the pharmacological effect of SL-172154 on APC and T cells. CXCL-10 binds CXCR3 expressed by Th1 cells and attracts them to the site of production through a chemokine gradient.

The safety, PD activity, and PK/PD modeling support 3.0 mg/kg as a safe and pharmacologically active dose. Fc silencing of SL-172154 likely explains the tolerability relative to anti-CD47 antibodies and CD40 agonist antibodies. Further, the hexameric structure of the CD40L domain provides potent activation of CD40 at receptor-saturating doses. The CD47 “don’t eat me” mechanism requires a prophagocytic “eat me” signal to facilitate macrophage-mediated phagocytosis of tumor cells. While CD40 agonism provides an activation signal to APCs, CD40 activation does not cause phagocytosis. Indeed, SL-172154 was deliberately designed to lack prophagocytic signals in an effort to avoid the cytopenias that have limited the development of other CD47-targeted agents. As a result, antitumor activity was predicted to require coadministration of SL-172154 with another agent that directly or indirectly increases the density of prophagocytic signals on tumor cells. This prophagocytic signal can be provided by chemotherapeutic agents, antibody drug conjugates, or monoclonal antibodies with antibody-dependent cellular phagocytosis activity. In a phase Ib study conducted in patients with platinum resistant ovarian cancer, SL-172154 is combined with pegylated liposomal doxorubicin or mirvetuximab soravtansine, as two distinct mechanisms to provide such a prophagocytic signal.