Preclinical pharmacological and toxicological studies were performed both in vivo and in vitro to evaluate safety and efficacy of IBI110 (Supplementary Results). In brief, IBI110 showed high affinity to human LAG-3 and cynomolgus monkey LAG-3. It disrupts MHCII/LAG-3 binding and triggers downstream signaling. IBI110 plus sintilimab showed enhanced IL-2 secretion in vitro and more favorable antitumor effect in vivo.

From December 4, 2019 to January 20, 2022, the phase Ia dose escalation of IBI110 monotherapy enrolled 28 patients and the phase Ib combination dose escalation of IBI110 plus sintilimab enrolled 45 patients (Table 1). In phase Ia escalation, patients were allocated to receive IBI110 every three weeks (Q3W) from 0.01 mg/kg (n = 1), 0.1 mg/kg (n = 1), 0.3 mg/kg (n = 3), 1 mg/kg (n = 3), 3 mg/kg (n = 6), 10 mg/kg (n = 7), to 20 mg/kg (n = 7). In phase Ib escalation part, patients were allocated to receive IBI110 Q3W from 0.3 mg/kg (n = 3), 0.7 mg/kg (n = 3), 1.5 mg/kg (n = 3), 3 mg/kg (n = 11), 5 mg/kg (n = 15), 8 mg/kg (n = 6), to 10 mg/kg (n = 4) plus sintilimab 200 mg Q3W. The CONSORT diagram is shown in Supplementary Figure S2.

The phase Ib expansion of IBI110 plus sintilimab and chemotherapy was conducted in previously untreated patients with advanced sqNSCLC or HER-2 negative GC. As of October 25, 2022, the sqNSCLC cohort enrolled 20 patients, including 5 patients with TNM stage III (25.0%) and 15 patients with TNM stage IV (75.0%). The median treatment duration of IBI110 plus sintilimab was 53.1 weeks (range: 6.1–70.3) with 10 patients remained on treatment (50.0%). As of March 22, 2023, the GC cohort enrolled 17 patients including 2 patients with TNM stage III (11.8%) and 15 patients with TNM stage IV (88.2%). There were 7 patients with liver metastasis (41.2%). The median treatment duration of IBI110 plus sintilimab was 16.6 weeks (range: 3.1–72.1) with 3 patients remained on treatment (17.6%). The baseline characteristics of each cohorts were listed in Table 1.

When IBI110 was administered as monotherapy or in combination with sintilimab at relatively low dose levels ranging from 0.01 to 1.5 mg/kg, decreased clearance rates were observed with escalating doses. These non-linear PK profiles were typical characteristics of target-mediated drug disposition (TMDD). When IBI110 dose increased to ≥ 3 mg/kg (200 mg), the clearance rates became stabilized and linear PK characteristics were observed (Supplementary Figure S3). Patients in phase Ia and Ib escalation were also tested for anti-drug antibody (ADA) to evaluate immunogenicity. No clinically significant changes in immunogenicity were observed across different dose levels indicating that immunogenicity was not dose-dependent. The PD analysis of IBI110 was described in Supplementary Results and presented in Supplement Figure S4.

During phase Ia escalation of IBI110 alone and phase Ib escalation of IBI110 plus sintilimab, no DLT was observed across all dose groups. The maximum tolerated dose was not reached. Safety profiles of phase Ia and Ib escalation as well as Ib expansion were summarized in Table 2.

In phase Ia escalation, TRAEs of any grade occurred in 19 patients (67.9%) while TRAEs of grade ≥ 3 occurred in 6 patients (21.4%). Common TRAEs were presented in Supplementary Table S1 with the most common being anemia (17.9%, including 3.6% ≥ G3). TRAEs leading to dose interruption and treatment discontinuation occurred in 2 (7.1%) and 1 (3.6%) patients, respectively. No TRAEs led to death. Immune-related adverse events (irAEs) occurred in 3 patients (10.7%) including 1 patient (3.6%) had grade ≥ 3 irAE (Supplementary Table S2). In phase Ib escalation, TRAEs of any grades occurred in 34 patients (75.6%) while TRAEs of grade ≥ 3 occurred in 10 patients (22.2%). Common TRAEs were presented in Supplementary Table S3 with the most common being aspartate aminotransferase increased (28.9%, all G1-G2). TRAEs leading to dose interruption in 6 patients (13.3%). No TRAEs led to treatment discontinuation or death. irAEs occurred in 14 patients (31.1%) including 3 patients (6.7%) had grade ≥ 3 irAE (Supplementary Table S4).

The phase Ib expansion of IBI110 plus sintilimab and chemotherapy was conducted in patients with advanced sqNSCLC or HER-2 negative GC. In sqNSCLC cohort (n = 20), TRAEs of any grade occurred in all patients while TRAEs of grade ≥ 3 occurred in 16 patients (80.0%). Common TRAEs in sqNSCLC cohort were presented in Supplementary Table S5 with the most common being anemia (70.0%, all G1-G2). TRAEs leading to dose interruption and treatment discontinuation occurred in 8 (40.0%) and 4 (20.0%) patients, respectively. No TRAEs lead to death. irAEs occurred in 14 patients (70.0%) including 4 patients (20.0%) had grade ≥ 3 irAE (Supplementary Table S6). In GC cohort (n = 17), TRAEs of any grade occurred in all patients (100%) while TRAEs of grade ≥ 3 occurred in 11 patients (64.7%). Common TRAEs in GC cohort were presented in Supplementary Table S7 with the most common being neutrophil count decreased (64.7%, including 17.6%≥ G3). TRAEs leading to dose interruption and treatment discontinuation occurred in 14 (82.4%) and 6 (35.3%) patients, respectively. No TRAEs lead to death. irAEs occurred in 10 patients (58.8%) including 3 patients (17.6%) had grade ≥ 3 irAE (Supplementary Table S8).

As of Jan 20, 2022, ORR and DCR of IBI110 monotherapy in phase Ia escalation (n = 28) were 3.6% and 25.0%, respectively (Fig. 1A). Only 1 patient with ovarian cancer at 3 mg/kg had PR. In phase Ib escalation, 43 of 45 patients had undergone at least 1 post-baseline tumor assessment. The ORR and DCR of IBI110 plus sintilimab were 14.0% and 67.4%, respectively (Fig. 1B). Based on the safety and efficacy profiles observed in phase Ia and Ib escalation, the RP2D was determined as IBI110 200 mg (3 mg/kg) Q3W plus sintilimab 200 mg Q3W. In phase Ib expansion, 20 patients with previously untreated, advanced sqNSCLC received RP2D of IBI110 in combination with sintilimab plus paclitaxel and carboplatin (TP) as first-line treatment; while 17 patients with previously untreated, advanced HER-2 negative GC received RP2D of IBI110 in combination with sintilimab plus capecitabine and oxaliplatin (XELOX) as first-line treatment (Table 3).

Tumor assessment of each patient in phase Ia (A) and Ib (B) dose escalation. (A) In phase Ia escalation of IBI110, the best response assessed by investigator was partial response (PR) in 1 patient and stable disease (SD) in 6 patients. After progressive disease (PD), patients may cross to combination treatment of IBI110 plus sintilimab while 8 patients had SD. (B) In phase Ib escalation of IBI110 plus sintilimab, 43 of 45 patients had undergone at least 1 post-baseline tumor assessment. The best response assessed by investigator was PR in 6 patients (4 non-small cell lung cancer, 1 small cell lung cancer and 1 endometrial cancer) and SD in 23 patients

As of October 25, 2022, the unconfirmed and confirmed ORRs in sqNSCLC cohort were 80.0% (95% CI, 56.3–94.3) and 75.0% (95% CI, 50.9–91.3), respectively (Fig. 2A and B). The DCR was 85.0% (95% CI, 62.1–96.8). The median DoR was not reached with events occurring in 4 of 15 patients with confirmed objective response. The 12-month DoR rate was 73.3% (95% CI, 43.6–89.1). The median PFS was not reached with events occurring in 8 (40.0%) patients. The 12-month PFS rate was 60.0% (95% CI, 35.7–77.6). The median OS was not reached with event occurred in 3 (15%) patients. The 12-month OS rate was 85.0% (95% CI, 60.4–94.9).

Efficacy of IBI110 in combination with sintilimab and chemotherapy in sqNSCLC and GC patients. (A) Confirmed best overall response in sqNSCLC: 1 patient had increase of 11.04% in size of the target lesion and 1 patient had decrease of 38.66% in size of the target lesion, but their overall responses were PD due to the presence of new lesions. (B) Tumor assessment in patients with sqNSCLC: 1 patient had the first and second tumor assessment of iUPD, and continued treatment until the third tumor assessment of iCPD. (C) Confirmed best overall response in GC: 1 patient had decrease of 0.83% in size of the target lesion, but the overall response was PD due to the presence of new lesion. (D) Tumor assessment in patients with GC

As of March 22, 2023, the unconfirmed and confirmed ORRs in GC cohort were 88.2% (95% CI, 63.6–98.5) and 70.6% (95% CI, 44.0-89.7), respectively (Fig. 2C and D). The DCR was 94.1% (95% CI, 71.3–99.9). The median DoR was 10.6 months (95% CI, 2.5–14.4). With a median follow up of 13.1 months (95% CI, 7.1-NR [not reported]), the progression-free survival (PFS) was 12.9 months (95% CI, 3.8–15.8). With a median follow up of 15.8 months (95% CI, 13.4–16.6), the median overall survival (OS) was 15.8 months (95% CI, 8.5-NR), and the 12-months OS rate was 70.6% (95% CI, 43.1–86.6).