First-in-human phase I/Ib study of QL1706 (PSB205), a bifunctional PD1/CTLA4 dual blocker, in patients with advanced solid tumors

Abstract

Background: QL1706 (PSB205) is a single bifunctional MabPair (a novel technical platform) product consisting of two engineered monoclonal antibodies (anti-PD-1 IgG4 and anti-CTLA-4 IgG1), with a shorter elimination half-life (t_1/2) for CTLA-4. We report results from a phase I/Ib study of QL1706 in patients with advanced solid tumors who failed standard therapies.

Methods: In the phase I study, QL1706 was administered intravenously once every 3 weeks at one of five doses ranging from 0.3 to 10 mg/kg, and the maximum tolerated dose, recommended phase 2 dose (RP2D), safety, pharmacokinetics (PK), and pharmacodynamics (PD) of QL1706 were investigated. In the phase Ib study, QL1706 was administered at the RP2D intravenously every 3 weeks, and the preliminary efficacies in non-small cell lung cancer (NSCLC), nasopharyngeal carcinoma (NPC), cervical cancer (CC), and other solid tumors were evaluated.

Results: Between March 2020 and July 2021, 518 patients with advanced solid tumors were enrolled (phase I, n = 99; phase Ib, n = 419). For all patients, the three most common treatment-related adverse events (TRAEs) were rash (19.7%), hypothyroidism (13.5%), and pruritus (13.3%). The TRAEs and immune-related adverse events (irAEs) of grade ≥ 3 occurred in 16.0% and 8.1% of patients, respectively. In phase I, 2 of 6 patients in the 10mg/kg group experienced dose-limiting toxicities (DLTs) (grade 3 thrombocytopenia and grade 4 immune-mediated nephritis), so the maximum tolerated dose (MTD) was reached at 10 mg/kg. The RP2D was determined to be 5 mg/kg based on comprehensive analysis of tolerability, PK/PD, and efficacy. For all patients who received QL1706 at the RP2D, the objective response rate (ORR) and median duration of response were 16.9% (79/468) and 11.7 months (8.3-not reached [NR]), respectively; and the ORRs were 14.0% (17/121) in NSCLC, 24.5% (27/110) in NPC, 27.3% (15/55) in CC, 7.4% (2/27) in colorectal cancer, 23.1% (6/26) in small cell lung cancer. For immunotherapy-naive patients, QL1706 exhibited promising antitumor activities, especially in NSCLC, NPC, and CC, with ORRs of 24.2%, 38.7%, and 28.3%, respectively.

Conclusions: QL1706 was well tolerated and demonstrated promising antitumor activity in solid tumors, especially in NSCLC, NPC, and CC patients. It is currently being evaluated in randomized phase II (NCT05576272, NCT05179317) and phase III (NCT05446883, NCT05487391) trials. Trial Registration ClinicalTrials.gov Identifier: NCT04296994 and NCT05171790.

Keywords: Bifunctional PD-1; CTLA4 antibody; Cervical cancer; MabPair; Nasopharyngeal carcinoma; Phase I trial.

Fig. 1

Generation and characterization of PSB205. A The principle of MabPair technology for producing two correctly assembled antibodies from a single mammalian cell line. Top panel, co-expression of two different antibodies in a single production cell line requires the simultaneous introduction of DNAs encoding two heavy chains (HCs) and two light chains (LCs) into the same cell. Under normal conditions, the two HCs can randomly dimerize to form two separate homodimers and one heterodimer species; the two LCs can also pair with either of the two HCs. The random combinations result in a total of 10 possible products generated, but only two of them are the desirable antibody products that contain the cognate HC/HC and LC/HC pairings (yellow-circled ones). Middle panel, using a charge-pair approach (referred as “HC pairing keys”) to correctly control the homodimeric pairing of the HCs of two different antibodies, four undesirable side products containing heterodimeric HCs are eliminated, so 10 combinations are reduced to 6. Bottom panel, applying a combined charge-pair and cysteine-pair approach (referred as “HC/LC pairing keys”) to control the cognate LC/HC pairings, only the two correctly paired and structurally stable products can pass the endogenous quality control system inside cells before they are secreted out. Other byproducts are fully eliminated due to their instability. B Fluorescence-assisted cell sorting plots showing the coexpression of PSB103 and PSB105 in the production cell line after intracellular staining of conjugated anti-hu IgG4- and anti-hu IgG1-specific antibodies, respectively. A panel of orthogonal analytical methods was used to characterize the PSB205 MabPair. The results confirmed the molecular integrity, and no mispaired antibodies were detected in the sequential characterization. C PSB205 size variants were analyzed by size-exclusion high-performance liquid chromatography (HPLC). The chromatogram shows the main peak for the monomers of the two mAbs overlaid, frontal minor peak(s) for high-molecular weight species, and post minor peak(s) for low-molecular weight species (not detected) in PSB205. As a result, the PSB205 purity as defined by the monomers (the main peak) was typically measured as 97–99% for different batches. D Baseline separation of the two mAbs in PSB205 was achieved by the hydrophobic interaction HPLC method. Thus, it served as a tool to determine the concentration ratio of the two mAbs, [anti-PD-1]:[anti-CTLA-4] (w/w). E The intact glycoform mass profile was obtained by liquid chromatography–mass spectrometry (LC–MS) analysis. As a result, the two main peaks at 149,320 Da and 147,610 Da in the deconvoluted mass spectra closely match the G0F/G0F glycoforms of anti-PD-1 and anti-CTLA-4, respectively, with their HC N-terminal Gln converted to pyroglutamic acid and the C-terminal Lys removed

Fig. 2

Flowchart of the study

Fig. 3

Tumor response. The waterfall plot (A, C, E) of the best overall responses with respect to the tumor size and the swimmer plot (B, D, F) of time to tumor response in non-small-cell lung cancer (NSCLC) patients (A, B), nasopharyngeal carcinoma (NPC) patients (C, D), and cervical cancer (CC) patients (E, F)

Fig. 4

Mean (± standard deviation) plasma concentrations of anti-CTLA-4 (A) and anti-PD-1 (B) as a function of time following dosing in cycle 1 and at steady state (cycle 6) shown on a log10 scale in μg/mL across dose levels from 0.3 mg/kg to 10.0 mg/kg Q3W. C Mean % PD-1 receptor occupancy. Expression changes of Ki67 (D) and ICOS (E) on T cells in each dose group (Note, the values out of the visit window range or deviated from the protocol were not included in the summary analysis. When more than half (> 50%) of the values at a single time point are below the quantization limit (BQL), the mean values are reported as 0. For those BQL values, they are omitted on the semi-log scale plot. When there are only two samples at a single time point, the error bars are not presented