After improving the core structure of ARS-853, Janes et al [15] synthesized a quinazoline core as a versatile lead scaffold. Among the drugs using this scaffold, the representative drug is ARS-1620 (See Figure 5). ARS-1620 covalently modified KRAS G12C at a rate of 1,100 ± 200 M⁻¹s⁻¹ (kobs/[I]), 10 times higher than the rate of ARS-853. In H358 cell lines, ARS-1620 inhibited RAS-GTP binding and the phosphorylation of MEK, ERK, RSK, S6, and AKT in a dose-dependent and selective manner but had no effect on the negative control group (A549, H460, and H44 cells), indicating that it selectively inhibited G12C mutant cells but not wild-type cells. Moreover, ARS-1620 showed excellent oral bioavailability (F >60%) and had sufficient plasma stability in mice. To further determine whether ARS-1620 has sufficient covalent and pharmacokinetic properties to enable it to function in vivo, the researchers administered ARS-1620 to a subcutaneous xenograft model with KRAS G12C-bearing tumors at a dose of 200 mg/kg/day and found that the tumor obviously receded.

Using ARS-1620 alone confers drug resistance to tumor cells and leads to tumor recurrence after regression. Miriam et al [16] found that deletion of MTOR genes can make cells significantly sensitive to KRAS and IGF1R inhibitors. The researchers also found that blocking the IGF1R, MAPK, and PI3K/AKT/mTOR signaling pathways simultaneously led to the death of cancer cells carrying KRAS mutations. On the basis of previous findings, they screened the entire genome (16,019 genes) in shRNA-infected KRAS-mutant non-small-cell lung cancer (NSCLC) H23 cells, trying to identify triple combination therapies using ARS-1620, mTOR inhibitors and the IGF1R inhibitor linsitinib. They found that triple therapy could significantly reduce tumor size in mice and humans, and cell growth was not detected for 17 days. Signaling pathway analysis showed not only that ARS-1620 inhibited ERK and S6 phosphorylation but also that the triple therapy persistently inhibited AKT and S6 phosphorylation. In addition, the triple therapy reduced the reactivation of RAS and ERK activation for 48 h.

Due to the small size of the switch II pocket, additional protein-ligand interactions are limited; this limitation makes further improvements in ARS-1620 difficult. However, with the discovery of the obvious hydrogen bonding between ARS-1620 and His95, new progress has been achieved by researchers. A surface groove formed by the alternative orientation of His95 can be occupied by aromatic rings, and this discovery might enhance the interaction between drugs and KRAS G12C [17]. Therefore, a new drug called AMG-510 (See Figure 6), which has similar ligand structure to ARS-1620, was developed by Cannon et al. [18] However, His95 is a newly identified site for AMG-510 binding. AMG-510 almost completely inhibited pERK (IC50 ≈ 0.03 μM) in NCI-H358 and MIA PaCa-2 cells after 2 h of treatment, and its IC50 value was 20 times that of ARS-1620. In terms of cell viability, AMG-510 decreased the viability of NCI-H358 and MIA PaCa-2 cells (IC50 ≈ 0.006 μm and 0.009 μm, respectively), with IC50 values approximately 40 times those of ARS-1620. However, the strong binding affinity of AMG-510 may covalently link it to other cysteine-containing proteins, which is a problem that needs to be addressed in the clinical trial stage [19].

In a patient trial (four patients with NSCLC), the researchers found that treatment with AMG-510 resulted in objective partial responses (PRs) in two patients and stable disease (SD) in two patients. The two patients with a PR had documented disease progression on treatments including carboplatin, pemetrexed, and nivolumab. After 6 weeks of AMG-510 treatment, the first group (180 mg) exhibited tumor shrinkage of 34%, and the second (360 mg) exhibited tumor shrinkage of 67%. A computed tomography (CT) scan 18 weeks later showed that the second group of tumors had shrunk to the point at which they were no longer visible. Amgen [20] presented preclinical data of AMG-510 at the American Association for Cancer Research (AACR) meeting in 2019. They used homologous tumor cell lines in vitro that were suitable for testing AMG-510, checkpoint inhibitors, and AMG-510 combined with checkpoint inhibitors and ultimately found that AMG-510 could clear colon cancer cells from mice. In the phase I/II study [21, 22], a total of 35 patients were recruited into four low-dose groups (180, 360, 720, and 960 mg). The outcome was evaluated in 29 patients, 10 of whom had NSCLC (most of them had previously received ≥3 kinds of treatments). The results showed that half of the NSCLC patients had objective PRs. The most common treatment-related adverse events (AEs) were loss of appetite, diarrhea, fatigue, headache, cough, hot flashes, and nausea. The severe adverse events (AEs) in 6 patients, including grade III pneumonia, malignant biliary obstruction, and grade IV pericardial effusion, were all considered to be AMG-510 related. A recent study, presented in September at the 2019 World Lung Cancer Congress meeting, covered 34 NSCLC patients who received AMG-510 treatment; 19 of them were treated at a dose of 180 mg, and 15 of them had stage II disease and received a dose of 960 mg; no dose-limiting toxicity was observed. Of the 23 patients who received a CT scan for 6 weeks or had early progressive disease, the objective response rate was 48%, and the disease control rate was 96%. Among the evaluable patients, 13 received 960 mg of phase II treatment, 7 people (54%) achieved PR, and 6 people (46%) achieved SD.

The clinical results of strategies combining BRAF and MEK inhibitors [23] suggest that combinations with AMG-510 and other inhibitors may show enhanced tumor cell killing effects and overcome drug resistance. One team conducted experiments in vitro in KRAS G12C cell lines using AMG-510 and HER kinase, EGFR, SHP2, PI3K, AKT, and MEK inhibitors and found that the combination of AMG-510 and MEK inhibitors significantly enhanced antitumor activity. In an AMG-510 monotherapy mouse experiment, the tumors of mice lacking T cells were likely to recur. The researchers used a model called CT-26KRASG12C [24, 25], which is widely used to assess the effectiveness of immunotherapy and combinations of immunotherapy and targeted drugs, to assess AMG-510 combined with a PD-1 inhibitor. As a result, 1/10 of the tumors completely disappeared when AMG-510 or the PD-1 inhibitor was used alone. However, 9/10 of the tumors disappeared completely with the combination therapy. Treatment was stopped after the 43rd day, and no recurrence was seen in any complete responder after 112 days.

Hallin et al [26] discovered MRTX849 (See Figure 7) while performing favorable drug optimization experiments. Through LC/MS-based tests of KRAS G12C protein modification, they found that MRTX849 also combined with GDP-bound KRAS G12C and locked it in the inactive state. In addition, MRTX849 also inhibited KRAS-dependent signal transduction effects, including ERK1/2 phosphorylation, S6 phosphorylation, and the expression of ERK-regulated DUSP6, and the greatest effect was observed at 24 h. The half-life was approximately 25 h after a single dose. To assess the pharmacodynamics of MRTX849, the researchers fed mice H358 xenotransplantation by gavage. The dose of MRTX849 found to covalently modify KRAS G12C was 30 mg/kg, and the effects increased in a dose-dependent manner; the maximum effective dose of MRTX849 was between 30 and 100 mg/kg/day. The researchers performed experiments with a dose of 100 mg/kg/day in cell line-derived xenograft (CDX) and patient-derived xenograft (PDX) models. After approximately 3 weeks of treatment, 17 of the 26 models (65%) showed tumor regression over 30%.

The preliminary results of a phase I/II trial, which began in January 2019, were first presented at the AACR-NCI-EORTC Conference in 2019 (International Conference on Molecular Targets and Cancer Therapy) [27]. Of the 12 assessable patients (6 with NSCLC, 4 with colorectal cancer (CRC) and 2 with appendix cancer), 3 (50%) NSCLC patients and 1 (25%) CRC patient achieved a PR. The most common adverse events from MRTX849 were grade 1/2 diarrhea or nausea, and no other serious adverse reactions have yet been found.

MRTX849 combined with HER family inhibitors synergistically inhibited tumor cell viability in most cell lines and was first tested in combined screening in vitro. The researchers selected afatinib as a representative HER family inhibitor and chose five tumor models (KYSE410, SW1573, H2122, H2030, and LU6405 cells) with partial sensitivity to single-agent MRTX849 or refractoriness to the treatment, and all the cell lines showed obvious antitumor activity. They found that HER family activation in vivo may limit MRTX849-mediated inhibition of the ERK and mTOR-S6 signaling pathways, while administration of afatinib combined with MRTX849 could limit feedback reactivation of ERK signaling and showed complementary inhibition of AKT-mTOR-S6 signaling, significantly enhancing antitumor activity. Moreover, the combination of SHP2 inhibitors and MRTX849 also achieved inhibition of tumor activity. Administration of the mTOR inhibitor vistusertib with MRTX849 induced complementary inhibition of the ERK and mTOR-S6 signaling pathways, showing extensive antitumor activity.