Discovery of a novel ALK/ROS1/FAK inhibitor, APG-2449, in preclinical non-small cell lung cancer and ovarian cancer models

Tyrosine kinase inhibitors (TKIs) are mainstays of cancer treatment. However, their clinical benefits are often constrained by acquired resistance. To overcome such outcomes, we have rationally engineered APG-2449 as a novel multikinase inhibitor that is highly potent against oncogenic alterations of anaplastic lymphoma kinase (ALK), ROS proto-oncogene 1 receptor tyrosine kinase (ROS1), and focal adhesion kinase (FAK). Here we present the preclinical evaluation of APG-2449, which exhibits antiproliferative activity in cells carrying ALK fusion or secondary mutations. KINOMEscan® and LANCE TR-FRET were used to characterize targets and selectivity of APG-2449. Water-soluble tetrazolium salt (WST-8) viability assay and xenograft tumorigenicity were employed to evaluate therapeutic efficacy of monotherapy or drug combination in preclinical models of solid tumors. Western blot, pharmacokinetic, and flow cytometry analyses, as well as RNA sequencing were used to explore pharmacokinetic–pharmacodynamic correlations and the mechanism of actions driving drug combination synergy. In mice bearing wild-type or ALK/ROS1-mutant non-small-cell lung cancer (NSCLC), APG-2449 demonstrates potent antitumor activity, with correlations between pharmacokinetics and pharmacodynamics in vivo. Through FAK inhibition, APG-2449 sensitizes ovarian xenograft tumors to paclitaxel by reducing CD44+ and aldehyde dehydrogenase 1-positive (ALDH1+) cancer stem cell populations, including ovarian tumors insensitive to carboplatin. In epidermal growth factor receptor (EGFR)-mutated NSCLC xenograft models, APG-2449 enhances EGFR TKI-induced tumor growth inhibition, while the ternary combination of APG-2449 with EGFR (osimertinib) and mitogen-activated extracellular signal-regulated kinase (MEK; trametinib) inhibitors overcomes osimertinib resistance. Mechanistically, phosphorylation of ALK, ROS1, and FAK, as well as their downstream components, is effectively inhibited by APG-2449. Taken together, our studies demonstrate that APG-2449 exerts potent and durable antitumor activity in human NSCLC and ovarian tumor models when administered alone or in combination with other therapies. A phase 1 clinical trial has been initiated to evaluate the safety and preliminary efficacy of APG-2449 in patients with advanced solid tumors, including ALK+ NSCLC refractory to earlier-generation ALK inhibitors. Clinicaltrial.gov registration: NCT03917043 (date of first registration, 16/04/2019) and Chinese clinical trial registration: CTR20190468 (date of first registration, 09/04/2019).


Significance
In preclinical experiments, APG-2449 is active against treatment-resistant (e.g., lung, ovarian) tumors. If confirmed in clinical trials, these findings may support the role of APG-2449 as a novel therapy for these conditions.

Background
Molecular targeted therapies (e.g., tyrosine kinase inhibitors [TKIs]), have achieved clinical successes when directed against strong oncogenic drivers, such as anaplastic lymphoma kinase (ALK) rearrangement and epidermal growth factor receptor (EGFR) mutation. Despite initial responses, patients almost inevitably develop TKI resistance by acquiring or expanding resistance mutations [1], within 12-24 months for crizotinib [2]. Treatment with crizotinib, an ALK inhibitor and one of the earliest validated TKIs, results in an overall response rate of ~ 60% in patients with ALK-positive (ALK + ) nonsmall-cell lung cancer (NSCLC) [2,3]. Extensive efforts have been undertaken to develop novel TKIs that deepen treatment responses and overcome drug resistance, culminating in discovery of second-generation (2G) TKIs ceritinib, alectinib, and brigatinib, as well as third-generation (3G) agent lorlatinib.
Our efforts to develop a novel TKI resulted in APG-2449, which demonstrated activity against ALK, ROS proto-oncogene 1 receptor tyrosine kinase (ROS1), and focal adhesion kinase (FAK) [4]. ROS1 gene fusions account for ~ 1 to 2% of all cases of NSCLC [5]. FAK is a ubiquitous intracellular nonreceptor tyrosine kinase localized at focal adhesions and is a key regulator of cellular adhesion, migration, and proliferation [6,7]. FAK gene overexpression or amplification occurs in several malignancies [8] and is linked to tumor progression, metastasis, drug resistance, and a poor prognosis [9][10][11]. To date, several FAK inhibitors such as defactinib (VS-6063) are under clinical development. Despite preclinical antitumor effects [12][13][14][15][16], objective clinical responses have not been achieved [17][18][19], apart from extending progression-free survival and stabilizing disease. Pharmaceutical development strategies have largely shifted to combinations of TKIs with chemotherapeutics, targeted therapies, and/or immune modulators for drug-resistant cancers, especially chemotherapy-resistant ovarian cancer [10] and EGFR TKI-resistant NSCLC [20,21]. To advance these initiatives, we conducted in vitro and in vivo studies on the effects of multikinase inhibitor APG-2449, alone or with available therapies, in overcoming TKI resistance observed in ALK-, ROS1-, and FAKmutated tumors.
an N-terminal glutathione S-transferase fusion protein was purchased from Carna Biosciences (Kobe, Japan). Mutated proteins from mutant ALK were expressed in SF9 insect cells with N-terminal tags cleaved after purification. Kinase activities of all enzymes were assessed using a LANCE TR-FRET assay kit (PerkinElmer, Waltham, MA), as previously reported [24].

Cell proliferation assay
Cell viability was determined using the water-soluble tetrazolium salt (WST-8) assay (Cell counting Kit-8, Shanghai Life iLab). Cell viability was calculated as follows, where "RLU" signifies relative light units: Half-maximal inhibitory concentration (IC 50 ) values were calculated using Prism (GraphPad, San Diego, CA).
The following were computed as previously reported (23): mean ± SEM tumor volumes, percent changes in body weight, tumor growth inhibition in treatment/control (T/C; %), modified Response Evaluation Criteria in Solid Tumors (mRECIST) [25] objective response rate (ORR %), and disease control rate (DCR, %).

Pharmacokinetic (PK) analysis
Systemic and tumor tissue APG-2449 exposures were measured by liquid chromatography tandem mass spectrometry (LC/MS/MS) using an Exion high-performance LC system (AB Sciex, Ontario, Canada) coupled to an Atmospheric Pressure Ionization (API 5500) MS (AB Sciex) equipped with an API electrospray ionization source.

Western blot analysis
Western blot analysis was performed as described previously [23]. Blots were cut prior to hybridization with the antibodies. Images of original cut blots with membrane edge visible are included (where applicable) in the Supplementary Material. Protein expression was quantitated using ImageJ (Image Processing and Analysis in Java) software (GitHub; Wayne Rasband, US National Institute of Mental Health) and normalized with β-actin to compute relative fold changes of proteins (vs. vehicle control). Antibody information is enumerated in Supplementary Table S1.
IHC staining was performed on the BOND RX automated IHC & In Situ Hybridization system with the Detection System Horseradish Peroxidase (HRP) Polymer Kit (each from Leica Biosystems, Danvers, MA). Stained sections were scanned using a Hamamatsu Photonics NanoZoomer 2.0-HT Image system (Hamamatsu, Japan) at 40× magnification. Images were analyzed on a HALO ® image analysis platform (v3.0.311.363; Indica Labs, Albuquerque, NM). IHC staining intensity was scored as 0 (negative), 1 + (weak), 2 + (medium), or 3 + (strong). The percentages of tumor cells at different intensity levels were evaluated as:

RNA-sequencing (RNA-seq)
RNA-seq of ovarian PDX tumors was performed using an Illumina NovaSeq 6000 system (Crown Bioscience Inc., San Diego, CA). Differential expression analysis was carried out using the Empirical Analysis of Digital Gene Expression Data in R (edgeR) package (Bioconductor Open-Source Software for Bioinformatics; Stanford, CA). Significantly differentially expressed genes were defined as a false-discovery rate of < 0.05 and a fold change of ≥ |2|.

Statistical analysis
To assess statistical significance of differences between treatment groups, we conducted two-tailed Student t-tests and one-way analyses of variance followed by Games-Howell's post-test (for multiple comparisons). The sample size calculation was based on a resource equation [26]. All data were analyzed using Statistical Product and Service Solutions (SPSS) version 18.0 (IBM, Armonk, NY), with GraphPad Prism for graphic presentation.
Cells harboring ALK or ROS1 rearrangements and other clinically relevant resistant mutations were among the most sensitive to APG-2449 (Fig. 1E). In cells without ALK or ROS1 rearrangements but with detectable expression of FAK protein, APG-2449 exhibited moderate antiproliferative activity (e.g., IC 50 = 3.55 μM in NCI-H1975 and 2.71 μM in PA-1). In conclusion, APG-2449 is a novel and potent multikinase inhibitor with selective activity against ALK, ROS1, and FAK.

Pharmacologic profiles of APG-2449
To explore pharmacologic characteristics of APG-2449, we treated SCID-beige mice bearing KARPAS-299 CDX tumors with various single oral doses of APG-2449, which resulted in dose-proportional increases in APG-2449 plasma and tumor exposures (Fig. 2D).
In conclusion, our in vivo studies demonstrated potent and durable activity of APG-2449 against tumors with ALK or ROS1 secondary mutations, including the critical ALK G1202R . Compared to ceritinib or ensartinib, APG-2449 exhibits greater antitumor activity against crizotinib-resistant tumors lacking secondary ALK mutations, indicating that APG-2449 may be effective in a large proportion of patients with ALK inhibitor-resistant NSCLC irrespective of ALK genotype.

APG-2449 sensitizes ovarian cancer to chemotherapy by inhibiting the FAK signaling pathway
Overexpression and amplification of FAK occurs in 68 and 26.7% of patients with ovarian cancer, respectively [29]. These genetic alterations are significantly associated with metastasis and a poor prognosis [30]. Therefore, the FAK signaling pathway represents a novel avenue for pharmacologic intervention against ovarian cancer. We thus evaluated on-target activity of APG-2449 against FAK in preclinical models of ovarian cancer. Immunoblotting analysis revealed that APG-2449 downregulated Y397-FAK autophosphorylation (p-FAK) and downstream signaling factors p-AKT, p-ERK1/2, and p-STAT3 in PA-1 ovarian cancer cells (Fig. 4A). Interestingly, the inhibitory effect of APG-2449 on p-AKT was more potent than that of FAK-selective inhibitor defactinib. However, defactinib inhibited p-FAK to a greater extent than APG-2449, suggesting that the 2 agents have distinct pharmacologic properties.
We next evaluated the antitumor activity of APG-2449, alone or combined with standard-of-care (SOC) chemotherapeutics, in ovarian cancer xenograft models. In SCID mice bearing PA-1 CDX tumors, APG-2449 50, 100, or 150 mg/kg showed dose-dependent antitumor activity (Fig. 4B) and downregulation of p-FAK (Suppl. Fig. 3A). Conversely, APG-2449, carboplatin, or paclitaxel monotherapy showed weak or marginal antitumor activity in BALB/c nude mice bearing ovarian cancer OVCAR-3 CDX tumors, (Fig. 4C). The SOC combination of carboplatin plus paclitaxel suppressed tumor growth, with a T/C value of 33.9%. Of potential interest, APG-2449 combined with paclitaxel had potent antitumor activity, with a T/C value of 8.2% and DCR of 80% (i.e., 1/5 mPR, 3/5 mSD). Ternary combinations of APG-2449, paclitaxel, and carboplatin did not further improve antitumor activity (data not shown). All animals were tolerant of the combinations, with limited body weight loss and no interruptions in dosing (Suppl. Fig. 3B).
To confirm the therapeutic potential of APG-2449 in combination with paclitaxel, we conducted an experiment in murine ovarian cancer PDX models. From a group of 78 ovarian cancer PDXs, we selected 6 exhibiting the highest expression levels of FAK mRNA or FAK amplification according to RNA-seq (Suppl . Table S2). Compared to each single agent, the combination of APG-2449 and paclitaxel enhanced antitumor activity in all 6 PDX models (Fig. 4D). In addition, the combination significantly augmented antitumor activity in 3 of the 6 PDX models, with a synergy ratio greater than 2 (Fig. 4D). Interestingly, these 3 responsive models were not sensitive to carboplatin treatment (Suppl . Table S2). In the aggregate, our studies demonstrate potent activity of APG-2449 plus paclitaxel in preclinical models of ovarian cancer exhibiting FAK overexpression or amplification, including those insensitive to carboplatin.

Reduction of ovarian Cancer Stem Cell (CSCs) by APG-2449 and paclitaxel
To explore the mechanism of action underlying enhanced and/or synergistic antitumor effects of  that is known to support proliferation, self-renewal, and differentiation of CSCs [31]. Upregulation of the glycine, serine, threonine pathway supports tumor homeostasis and promotes cancer cell survival [32]. Thought-provokingly, the most markedly upregulated gene in APG-2449 responders was CD44, which is a marker for CSCs in ovarian cancer [33]. IHC staining further confirmed that CD44 expression was significantly higher in responders (vs. nonresponders), whereas no such differences were observed for FAK, P-FAK or E-cadherin (Suppl. Fig. 3C and D). These results suggest that high-"stemness" ovarian cancer models may be more susceptible to APG-2449/paclitaxel treatment.
In ovarian cancer models, it has been reported that FAK inhibition reduces CD44 protein expression levels [34] and stem phenotype [35]. In our CD44-expressing OV2018 PDX model that responded to APG-2449/paclitaxel (Fig. 4D), this binary combination downregulated CD44 protein expression (Fig. 4F). Downregulation of FAK, p-FAK, YB1, and p-YB1 was also noted, suggesting that downregulation of CD44 may be mediated by its upstream transcription factor YB1. In contrast, CD44 expression levels were barely detectable in OV1396 tumors insensitive to the combination. IHC analysis of OVCAR-3 tumors consistently revealed that expression of CD44 and another CSC marker (ALDH1A1) decreased after APG-2449/paclitaxel treatment (Fig. 4G). To assess effects on numbers of CSCs, we exposed SKOV-3 cells to APG-2449, paclitaxel, and the combination, and then analyzed by flow cytometry, which showed that APG-2449 dose dependently decreased numbers of CD44 + and ALDH1 + CSCs (Fig. 4H). A further decrease in these CSC populations was observed when APG-2449 was combined with paclitaxel (Fig. 4I). These results demonstrate that the combination of APG-2449 and paclitaxel effectively inhibits tumor growth in ovarian cancers with FAK overexpression or amplification by downregulating CD44 + and ALDH + CSC populations.

APG-2449 enhances EGFR TKI-mediated tumor suppression in NSCLC
FAK and SRC family kinases (SFK) support downstream AKT and mitogen-activated protein kinase (MAPK) signaling when continuous EGFR inhibition is driven by EGFR TKIs [20,36]. Incomplete inhibition of these survival signals can gradually contribute to EGFR TKIacquired resistance. To enhance antitumor activity and delay relapse, combination therapy with FAK/SRC inhibitors may be a viable approach. We investigated the effect of combined APG-2449 and EGFR TKIs in mice bearing xenografts derived from (1) NSCLC HCC827 with EGFR Ex19del (SCID-beige mice); (2) NCI-H1975 with EGFR L858R_T790M mutation (BALB/c nude mice); and (3) a PDX model with EGFR L858R_T790M and ROS1 fusion (LD1-0006-215,676; NU/NU mice). A strong synergistic antitumor effect was observed in the HCC827 CDX model when APG-2449 was combined with first-generation EGFR inhibitor erlotinib (Fig. 5A). A similar effect was observed when APG-2449 was combined with 3G EGFR inhibitor osimertinib (2 mg/kg) in both a CDX (NCI-H1975; Fig. 5B) and a PDX (LD1-0006-215,676; Fig. 5C) model. The combination treatment resulted in 100% ORRs across all models tested. Administration of APG-2449 in concert with a higher dose of osimertinib (15 mg/ kg) extended the duration of antitumor response, with a temporary loss of body weight (Suppl. Fig. 4A and B).
We next evaluated antitumor activity of APG-2449 plus osimertinib in an osimertinib-relapsed setting using the NCI-H1975 CDX model in BALB/c nude mice. Tumorbearing mice were initially treated with a high dose of osimertinib (15 mg/kg) for 21 days, after which all treated animals achieved complete response. Treatment was then suspended from Day 22 until disease progression. Mice carrying relapsed tumors were randomly allocated to treatment with APG-2449, alone or combined with osimertinib, for 21 days. Rechallenge with APG-2449 resensitized relapsed tumors to osimertinib (1/5 mCR), with gradual disease progression after dose suspension. Intriguingly, treatment with APG-2449 plus osimertinib achieved deeper and more durable response (5/5 mCR) compared to osimertinib alone (Fig. 5D).
As a putative mechanism, we postulate that combined treatment with APG-2449 and osimertinib suppressed phosphorylation of EGFR (p-EGFR), FAK (p-FAK), SRC (p-SRC), and ERK (p-ERK) compared to each single agent, in NCI-H1975 xenografts after 2 weeks (Fig. 5E, Suppl. Fig. 4C). In summary, the combination of APG-2449 and EGFR TKIs synergistically enhances antitumor activity in EGFR-mutant NSCLC models, with synergy driven by downregulation of FAK, EGFR, SRC, and ERK phosphorylation and APG-2449 extending the duration of response to osimertinib.

Discussion
Targeting oncogenic-driving genetic aberrations using small-molecule drugs has met with considerable clinical successes and is a viable strategy of pharmacologic innovation against a range of malignancies. Among these druggable targets, ALK and ROS1 rearrangements, as well as EGFR mutations in NSCLC, are well-known examples. Despite high initial response rates, patients almost inevitably develop TKI resistance via multiple disease mechanisms, including: (1) acquiring de novo resistance mutations; (2) expanding pre-existing resistance mutations; (3) triggering gene amplification; and/or (4) activating alternative survival signaling pathways [37]. FAK overexpression or amplification has been linked to tumor progression, metastasis, drug resistance, and a poor prognosis in patients with various solid tumors [38]. Consequently, extensive efforts have been undertaken to: (1) develop next-generation TKIs that overcome clinically relevant resistant mutations and (2) discover novel combination therapies that suppress resistance mechanisms.
Our report suggests that APG-2449 may help to meet these needs by serving as (1) a novel ALK/ROS1 inhibitor that can overcome acquired drug resistance conferred by secondary mutations and other resistance mechanisms in ALK + /ROS1 + NSCLC; (2) a unique FAK inhibitor with a potentially different pharmacologic profile; and (3) an important constituent of combinations with SOC chemotherapeutics or targeted agents that can reverse primary or acquired drug resistance in FAK-overexpressing or amplified ovarian cancer and EGFR-mutant NSCLC models. As a putative mechanism of action, we propose that APG-2449 induces downregulation of p-ALK, p-ROS1, p-FAK, p-AKT, p-ERK1/2, and p-STAT3, hence suggesting that its antitumor activity is mediated by inhibiting multiple salient oncogenic pathways (Fig. 6).
Secondary mutations confer acquired resistance to first-generation ALK inhibitors in 20 to 30%, and 2G ALK inhibitors in 50 to 70%, of patients with ALK + NSCLC [28]. Another 20 to 70% of patients develop acquired resistance without secondary mutations [28]. In our preclinical NSCLC models, APG-2449 proved to be active against critical secondary mutations ALK L1196M and ALK G1202R appearing independently or concurrently and exerts antitumor activity in a crizotinib-resistant NSCLC PDX lacking secondary mutations.
If confirmed in clinical trials, these findings may support APG-2449 as a member of a new generation of ALK multikinase inhibitors that can offset resistance in certain subpopulations of patients with NSCLC that relapses on ALK inhibitors. Based on our preclinical data in this setting, APG-2449 demonstrated enhanced activity compared to current ALK inhibitors, including ceritinib, ensartinib, and alectinib.
Ovarian epithelial cancer is one of the deadliest malignancies in women, up to 70% of whom have FAK overexpression, amplification, or activation [29], which is in turn significantly associated with higher tumor stage, metastasis, and shorter overall survival. These effects may be ascribed to the pivotal role of FAK in regulating cellular migration and survival, growth factor signaling, cell cycle progression, and chemoresistance [10]. Inhibition of FAK is therefore a promising pharmaceutical approach to treat ovarian cancer and/or advance its SOC (including chemotherapy). In this context, FAK inhibition by defactinib is known to resensitize ovarian tumors to paclitaxel in both preclinical and clinical settings [14,34]. FAK inhibition by defactinib also targets the CD44 + CSC population and restores sensitivity in chemotherapyresistant cells [34]. Although APG-2449 alone effectively suppresses phosphorylation of FAK in ovarian PA-1 cells, it fails to exert meaningful antitumor activity in PA-1 and OVCAR-3 ovarian cancer xenograft tumor models. This observation comports with the prevailing understanding that monotherapy with a FAK inhibitor may be insufficient in certain clinical settings [11].
On the other hand, our studies have demonstrated that paclitaxel combined with APG-2449 is more effective in inhibiting tumor growth compared to paclitaxel plus carboplatin in a carboplatin-insensitive OVCAR-3 ovarian cancer xenograft model. Platinum/taxane-based chemotherapy remains the backbone of ovarian cancer treatment, and previous studies have shown that elevated FAK activation underlies intrinsic and acquired resistance to chemotherapy in residual tumors of patient-derived cells or mouse xenograft tumors treated with platinum or taxanes [35]. FAK inhibition by defactinib is known to enhance chemosensitivity in taxane-resistant cells [34]. Confirmation of the effectiveness of APG-2449/paclitaxel in 3 of 6 ovarian cancer PDX models supports the potential advantage of this combination over paclitaxel alone. These 6 PDX models expressed high levels of FAK mRNA and FAK amplification. Among these 6 PDX models, 3 that were sensitive to combined therapy with APG-2449 and paclitaxel had high levels of CD44 expression, suggesting that FAK alterations and CD44 expression warrant further clinical investigation as predictive biomarkers.
Also worthy of further clinical exploration are expression levels of CD44, ALDH1, p-FAK/FAK, and p-YB1/ YB1, as well as CSCs and other potential predictors and/ or pharmacodynamic biomarkers, in ovarian cancer. CSCs constitute a subgroup of cancer cells responsible for primary tumor growth, metastasis, chemoresistance, and cancer relapse [39][40][41]. In addition to driving tumor progression and metastasis in ovarian cancer, CSCs play a critical role in conferring chemoresistance, which in turn leads to relapsed or refractory tumors [40,42,43]. FAK is required to maintain ovarian cancer CSCs, potentially by activating the β-catenin pathway [35,44]. FAK overexpression upregulates ALDH1 activity in platinumresistant (and CD44 activity in taxane-resistant) ovarian tumors [10,34]. Expression of both ALDH1 and CD44 is associated with CSC populations in ovarian cancer [45]. In our preclinical models of ovarian cancer, tumors with high CD44 expression were more likely to respond to APG-2449/paclitaxel, and the combination also significantly reduced CSC populations expressing either ALDH1 or CD44. Collectively, synergistic antitumor activity of APG-2449/paclitaxel is likely mediated by downregulation of ALDH1 + , CD44 + , and CSC populations.
In lung cancer, SFK and FAK sustain AKT and MAPK pathway signaling under continuous EGFR inhibition [20]. Inhibiting either the AKT or MAPK pathway enhances the efficacy of osimertinib [20]. Therefore, the combination of EGFR or SFK/FAK inhibitors constitutes a promising therapeutic strategy for EGFR-mutant lung cancer. Accordingly, increased phosphorylation of FAK has been observed in erlotinib-resistant NSCLC cells [46]. Defactinib combined with osimertinib enhanced antitumor activity in PC-9-and PC-9-pemetrexed-resistant xenografts [46]. In vitro, defactinib and osimertinib also overcome erlotinib resistance [46]. However, the effects of concomitant defactinib and osimertinib in overcoming osimertinib resistance has not been evaluated. Our preclinical studies demonstrate that combining APG-2449 with osimertinib significantly enhanced antitumor activity in EGFR Ex19del HCC827, EGFR L858R_T790M NCI-H1975, and EGFR L858R_T790M plus ROS1 fusion in xenograft and PDX models. Compared to osimertinib alone, the combination with APG-2449 further suppressed phosphorylation of EGFR, SRC, AKT, and ERK in H1975 tumors.
In addition to SFK and FAK signaling cascades, cancer cells may acquire resistance to osimertinib by activating the ERK1/2 pathway [47]. Osimertinib combined with a MEK or ERK inhibitor synergistically decreased survival in EGFR-mutant, but not wt-EGFR NSCLC cells [48]. In osimertinib-resistant PC-9/OR cells and xenografts, APG-2449 plus osimertinib failed to achieve tumor regression, indicating that alternative signaling pathways may be activated to sustain tumor growth. However, the ternary combination of APG-2449, osimertinib, and MEK inhibitor trametinib achieved 100% ORR, resulting in a CR or PR in all animals.

Conclusions
In conclusion, investigational agent APG-2449 is a novel small-molecule inhibitor targeting ALK/ROS1/ FAK. In our preclinical models including APG-2449 alone or in concert with other agents, this multikinase inhibitor demonstrated anticancer activity in various types of solid tumors, including NSCLC carrying ALK, ROS1, or EGFR mutations and ovarian cancer expressing FAK alterations and elevated CD44 protein expression. Administered alone or in combination, APG-2449 can overcome primary and acquired TKI resistance in preclinical models of NSCLC and ovarian cancer. If confirmed in clinical trials, these data may pave the way for APG-2449 to become an innovative therapy for patients with certain solid tumors. Consistent with the encouraging preclinical findings, a phase 1 clinical trial has been initiated to evaluate the safety and preliminary efficacy of APG-2449 in patients with ALK + NSCLC and other solid tumors (NCT03917043/CTR20190468).