Safety, antitumor activity and biomarkers of sugemalimab in Chinese patients with advanced solid tumors or lymphomas: results from the first-in-human phase 1 trial

Background This first-in-human phase 1 trial is to evaluate the safety, pharmacokinetics, preliminary efficacy, and biomarkers of sugemalimab, a full-length, fully human anti-PD-L1 monoclonal antibody, in Chinese patients with advanced malignancies. Methods Eligible patients with unresectable advanced or metastatic solid tumors or lymphomas were enrolled in phase 1a to receive sugemalimab following a modified 3 + 3 design. The primary endpoints included safety, tolerability, and the recommended Phase 2 dose (RP2D). In phase 1b, patients with 7 selected types of tumor received sugemalimab at the RP2D alone (monotherapy cohorts) or in combination with standard-of-care (SOC) chemotherapy (combination cohorts). The primary endpoint of phase 1b was investigator-assessed objective response rate (ORR). Results As of 19 February 2020, 29 and 178 patients were treated in phase 1a and 1b, respectively. No dose-limiting toxicities were observed in phase 1a, and the RP2D of sugemalimab was determined as 1200 mg fixed dose once every 3 weeks. Sugemalimab-related adverse events (AEs) were mostly (75.9%) grade 1–2 in phase 1a. Antitumor activity was observed across dose levels with an ORR of 24.1%. In phase 1b, 15.9% and 40.4% of patients in the monotherapy and combination cohorts, respectively, reported grade 3–5 sugemalimab-related AEs. Promising efficacy was observed in all combination cohorts, with ORRs ranging from 47.6 to 75.0%. Exploratory biomarker analysis did not indicate significant differences in responses at different PD-L1 expression/tumor mutation burden levels. Conclusions Sugemalimab was well-tolerated and showed promising antitumor activity as monotherapy or in combination with SOC chemotherapy in advanced malignancies. This trial was registered with ClinicalTrials.gov on Oct 18, 2017, number NCT03312842. Supplementary Information The online version contains supplementary material available at 10.1007/s00262-021-03102-3.


Introduction
Cancer has become the leading cause of death and a major barrier to extending life expectancy worldwide in the past decade [1]. As the most populous country, China carries the most significant burden, with one-fourth of the estimated global cancer cases and deaths in 2018 occurring in the country [1,2]. In China, cancers with poor prognosis, such as non-small cell lung cancer (NSCLC), esophageal squamous cell carcinoma (ESCC), hepatocellular carcinoma (HCC), and gastric cancer, account for more than half of all cancers diagnosed in China whereas they only comprise one-fifth of cancer incidence in developed countries [2]. Therefore, the demand for novel and effective anti-cancer therapies remains huge in China.
Sugemalimab is a full-length, high-affinity, fully human PD-L1 blocking, IgG4 monoclonal antibody developed using the OmniRat® transgenic rat platform. It mirrors natural IgG4 human antibody with expected pharmacokinetics (PK) profiles, which may potentially reduce the risk of immunogenicity and related toxicity in patients [17,18]. In vitro, sugemalimab specifically binds to PD-L1, competitively blocks the binding of human PD-L1 with PD-1 and CD80, and effectively induces CD4+ T lymphocyte proliferation and enhances the production of interferon-γ and interleukin-2, suggesting the potentials in enhancing antitumor immunity [19]. In vitro studies also suggest that unlike other Fc-null PD-L1 blocking antibodies such as durvalumab, sugemalimab retains the binding to FcγR I and therefore could efficiently induce antibody-dependent cellular phagocytosis through crosslinking of PD-L1 positive tumor cells with macrophages that are prevalently present in the tumor microenvironment [19]. This mechanism may further enhance tumor antigen presentation for long-term antitumor immunity. In in vivo efficacy studies, sugemalimab significantly inhibited tumor growth in a syngeneic tumor model in which the host mice were engineered to express humanized PD-1 and implanted with MC38 tumor cells expressing human PD-L1.
This first-in-human phase 1 trial aimed to evaluate the safety and tolerability, PK, preliminary antitumor activity, and potential predictive biomarkers of sugemalimab as monotherapy and in combination with standard-of-care (SOC) chemotherapy in Chinese patients with advanced solid tumors or lymphomas.

Patients
Patients were recruited from 2 study centers in phase 1a and 15 study centers in phase 1b in China. Eligible patients were 18-75 years old; had histologically or cytologically confirmed unresectable, locally advanced or metastatic solid tumors or lymphomas with at least one measurable/evaluable lesion according to Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST v1.1) (solid tumors) or Lugano Classification 2014 (lymphomas); progressed since previous standard anti-cancer therapy; and had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0-1. Key exclusion criteria included known primary central nervous system (CNS) tumors; prior malignancy other than those specified in phase 1b within the past 5 years; and major cardiovascular diseases. Full inclusion and exclusion criteria are listed in Supplementary materials. All patients provided written informed consent; study procedures were approved by an independent ethics committee at each study center.

Study design
This was a multi-center, open-label, phase 1 trial, consisting of a dose-escalation phase (1a) in solid tumors or lymphomas and an expansion phase (1b) in multiple disease-specific cohorts. Following a modified 3 + 3 dose-escalation design, patients in phase 1a were treated with sugemalimab once every 3 weeks (Q3W) intravenously at doses of 3 mg/kg, 10 mg/kg, 20 mg/kg, 1200 mg fixed dose, and 40 mg/kg, for up to 24 months. Dose-limiting toxicity (DLT) was evaluated within the first 21-day treatment cycle and defined as adverse events (AEs) fulfilling a set of prespecified criteria (see Supplementary materials) that were judged by the investigator to be probably or definitely related to sugemalimab, or whose causality was undeterminable.

Endpoints
The primary endpoints of phase 1a were safety and tolerability, maximum tolerated dose (MTD), and/or RP2D. For phase 1b, the primary endpoint was the investigator-assessed objective response rate (ORR) as a single agent or in combination with chemotherapy in specific types of tumors per RECIST v1.1 (solid tumors) or Lugano Classification 2014 (lymphomas). Secondary endpoints of phase 1a included preliminary antitumor activity, PK profile, and immunogenicity. For phase 1b, the secondary endpoints included safety and tolerability, PK profile, and immunogenicity. Exploratory endpoints of both phases 1a and 1b included exploration of the pharmacodynamic profile of sugemalimab and evaluation of PD-L1 expression level as a potential predictive biomarker.

Assessments
Safety assessments were performed throughout the study. The severity of AEs was graded according to National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03. Tumor assessment for patients with solid tumors was performed by computed tomography (CT) or magnetic resonance imaging at screening within 28 days before enrollment, every 9 weeks during the 1 st year on study and every 12 weeks thereafter according to RECIST v1.1. Patients with lymphomas accepted a CT examination for tumor assessment at screening, and while on study every 12 weeks according to Lugano Classification 2014.
Blood samples were collected at prespecified time points to determine the concentration of sugemalimab by ELISA. Sodium heparin anticoagulated whole blood samples were collected at pre-dose of C1D1, C2D1, and C4D1 and were sent to the central lab for flow cytometry analysis of receptor occupancy (RO) by sugemalimab. A bound strategy was used to perform RO bioanalysis of peripheral CD3+ T cells (see Supplementary materials).
PD-L1 expression in tumor tissues was retrospectively assessed by immunohistochemistry in a central lab using the VENTANA PD-L1 (SP263) assay per user manual on a BenchMark autostainer. For all treatment cohorts, PD-L1 expression was scored by the percentage of membranestained tumor cells (TC) and immune cells (IC) at any intensity. Tumor tissue slides with less than 100 viable tumor or immune cells were considered not evaluable. PD-L1 expression in the GAC/GEJAC and ESCC cohorts was also determined by combined positive score (CPS) [20] measuring the number of PD-L1-stained tumor cells, lymphocytes, and macrophages per 100 viable tumor cells.

Statistical analysis
The safety analysis set included patients who received at least one dose of sugemalimab. The efficacy analysis set included patients who received at least one dose of sugemalimab and had a measurable/evaluable lesion at baseline. Pharmacokinetics analysis set included patients who received at least one dose of sugemalimab and had available blood drug concentration data.
Systemic exposure of sugemalimab was dose-proportional from 3 mg/kg to 40 mg/kg including 1200 mg fixed dose (Supplementary Table S3 months). Notably, one patient with middle and low differentiated adenocarcinoma of ampulla initially administrated at the starting dose level of 3 mg/kg Q3W (switched to 1200 mg Q3W upon RP2D determination) experienced a durable response that lasted for almost 2 years and was still ongoing as of the cutoff date.
Based on the safety, tolerability, PK, pharmacodynamics, and preliminary antitumor efficacy data collected in phase 1a, 1200 mg fixed dose Q3W was determined as the RP2D by the safety monitoring committee.
In total, 12 of the 69 patients achieved a best overall response of PR, including 2 (1 unconfirmed) in the CC/ GBC cohort, 2 (1 unconfirmed) in the HCC cohort, and 8 (2 unconfirmed) in the MSI-H/dMMR cohort, leading to ORRs of 6.9%, 10.5%, and 38.1%, respectively ( Fig. 2 and Table 3). The DoRs of the 2 responders in the CC/ GBC cohort were 2.8 and 8.0 months, respectively, and the responses in both patients were still ongoing as of the cutoff date, whereas the 2 responders in the HCC cohort experienced disease progression after they have responded to the treatment for 2.2 and 6.9 months, respectively. For the 8 responders in the MSI-H/dMMR cohort, the median DoR was 13.8 months (95% CI: 2.1, -)
Robust and durable antitumor activities were observed among the 107 patients included in the efficacy analysis set in these 4 cohorts. Two patients in the ESCC cohort, who had not reached their first post-baseline tumor assessments and were still receiving study treatment as of the cutoff date, were excluded from the efficacy analysis. In the GAC/ GEJAC cohort, 18 of the 29 patients achieved PR (1 unconfirmed), resulting in an ORR of 62.1% (Fig. 2 and Table 3). The median DoR, PFS, and OS were 11.3 months (95% CI: 3.9, -), 8.3 months (95% CI: 4.8, 13.3), and 17.0 months (95% CI: 12.1, -), respectively. Among the 37 efficacy-evaluable patients in the ESCC cohort, 25 patients had PRs (5 unconfirmed). The ORR was 67.6%, with the median DoR not reached (range: 0.03 + to 13.3 + months). The median PFS was 9.0 months (95% CI: 4.4, -), and the OS ranged from 2.5 to 18.2 + months, with the median OS not reached. Ten of the 21 patients in the non-squamous NSCLC cohort experienced PRs (2 unconfirmed). The ORR was 47.6%. Among the responders, the median DoR was 8.7 months (95% CI:1.8, -). The median PFS and OS were 6.5 months (95% CI: 4.4, 11.7) and not reached (range: 2.7 to 16.4 + ), respectively. In

PD-L1 expression
PD-L1 expression was evaluable for 118 patients enrolled in phase 1a (1200 mg dose group only) and 1b, among which 88 (74.6%) had PD-L1 expression (TC/IC ≥ 1%). A total of 28 patients who received sugemalimab monotherapy at 1200 mg Q3W (RP2D) had TC/IC ≥ 1%, among which 8 (28.6%) patients achieved a best response of PR, while the response rate was 20.0% (n = 3) in the 15 patients without PD-L1 expression. For the 26 PD-L1-evaluable GAC/GEJAC patients, the ORRs were 63.2% and 57.1% for patients with CPS ≥ 5 (n = 19) and CPS < 5 (n = 7) (Table 4), with median DoR being not reached (95% CI: 9.7, -) and 5.0 months (95% CI: 3.2, -), and median PFS being 13.3 months (95% CI: 4.4, -) and 6.3 months (95% CI: 2.0, 13.3), respectively. In  The safety profiles of sugemalimab monotherapy and combined with chemotherapy observed in this study were considered generally consistent with those reported for other anti-PD-L1/anti-PD-1 monoclonal antibody therapeutics [21]. In both phase 1a and 1b, most sugemalimab-related AEs were of grade 1 or 2 and manageable. Only 2 (6.9%) and 18 (10.1%) patients in phase 1a and 1b discontinued sugemalimab due to AEs, respectively. Two fatal AEs (death and cerebral hemorrhage) considered related to sugemalimab and/or chemotherapy by the investigators were reported in one patient each from the ESCC and non-squamous NSCLC cohorts, respectively. These two events occurred when the patients were treated outside the research centers, thus complete exclusion of their relation to sugemalimab was restricted due to the lack of detailed information.
The responses observed in phase 1a were promising and durable, with an overall ORR of 24.1% in 29 heavily pre-treated patients of 22 different types of tumors and two patients continuing to benefit from the treatment for almost 2 years. In phase 1b, the regimen of sugemalimab combined with SOC chemotherapy also showed favorable efficacy in cohorts of major cancers in the 1L setting, consistent with the mechanism that the combination of immune checkpoint inhibitors (ICIs) and chemotherapy could yield synergistic effects by regulating tumor cell/ immune microenvironment interactions and improve the overall therapeutic outcome [22]. For non-squamous and squamous NSCLC, the addition of sugemalimab to carboplatin and pemetrexed or paclitaxel achieved ORRs of 47.6% and 75.0%, and PFS of 6.5 and 8.4 months, respectively, similar to those reported in phase 3 studies involving pembrolizumab plus chemotherapy [23,24]. Clinical benefits from similar synergistic treatment effects have also been shown for GAC/GEJAC and ESCC in phase 3 trials of nivolumab/pembrolizumab plus chemotherapy [25][26][27][28]. Sugemalimab in combination with SOC chemotherapy demonstrated comparable results in these two cohorts with overall ORRs of 62.1% and 67.6%, median PFS of 8.3 months and 9.0 months, and median OS of 17.0 months and not reached, respectively. These findings support additional studies for sugemalimab plus SOC chemotherapy as a novel and effective 1L treatment option for patients with NSCLC, GAC/GEJAC and ESCC, and three double-blind, randomized phase 3 trials of sugemalimab plus SOC have thus been initiated and are currently ongoing in the mentioned three indications (NCT03789604, NCT03802591, NCT04187352). While the conclusions in this study may be approached with caution due to the limited sample size in each indication cohort, they could be further validated in these phase 3 trials.
In the exploratory biomarker analysis, the limited number of biomarker-evaluable patients in this study precludes us to draw a statistically meaningful conclusion, but we observed a trend that a higher level of PD-L1 expression may potentially lead to a higher response rate in the GAC/ GEJAC cohort.
In summary, sugemalimab demonstrated a well-tolerated and manageable safety profile both as a single agent and combination therapy in Chinese patients with advanced solid tumors. The robust antitumor activity observed in the combination therapy cohorts provided solid evidence in supporting the ongoing phase 3 trials of sugemalimab plus chemotherapy as the 1L treatment for patients with advanced or metastatic NSCLC, GAC/GEJAC, and ESCC.