Tumor angiogenesis is one of the hallmarks of cancer (Cook and Figg, 2010; El-Kenawi and El-Remessy, 2013). Angiogenic factors such as VEGF (Hicklin and Ellis, 2005), PDGF (Gialeli et al., 2014; Farooqi and Siddik, 2015), FGF (Korc and Friesel, 2009), c-KIT (Foster et al., 2018), RET (Jhiang, 2000), TIE (Partanen and Dumont, 1999) and their receptors (tyrosine kinases) such as VEGFR, PDGFR, FGFR mediate inter- and intra-cellular signalling cascades that activate cellular pathways culminating in the formation and branching of blood vessels which promotes rapid tumor growth. These angiogenesis pathways are known to cross-talk (Lai et al., 2018) with other key tumor signalling pathways such as the PI3K/Akt/mTOR (Conciatori et al., 2018), EGFR/ERBB2 (van Cruijsen, Giaccone, and Hoekman, 2005) and Raf-MEK-ERK-MAPK (Song and Finley, 2018) in a complex web of redundant and escape mechanisms which are favourable for tumor growth and proliferation; blockade of these signalling pathways is a viable strategy to control tumor growth and spread (Pietras et al., 2003; Press and Lenz, 2007; Shibuya, 2011; Sitohy, Nagy, and Dvorak, 2012; Dieci et al., 2013; Heldin, 2013; Stankov, Popovic, and Mikov, 2014; Abbaspour Babaei et al., 2016; Navid et al., 2020; Subbiah and Cote, 2020). Several anti-angiogenic agents (angiogenesis inhibitors, AGIs) including monoclonal antibodies (mABs) such as Bevacizumab (Shih and Lindley, 2006) (anti-VEGF) or Ramucirumab (Fala, 2015) (anti-VEGFR2) and small molecule Tyrosine Kinase Inhibitors (TKIs) such as Axitinib (Rini et al., 2011), Pazopanib (Sternberg et al., 2010; van der Graaf et al., 2012), Regorafenib (Demetri et al., 2013; Grothey et al., 2013; Bruix et al., 2017), Sorafenib (Wilhelm et al., 2004; Llovet et al., 2008; Brose et al., 2014) and Sunitinib (G D Demetri et al., 2005; Ravaud et al., 2016; Raymond et al., 2011) have been developed and approved for use in various cancers; though Imatinib (Dagher et al., 2002) is not a direct inhibitor of angiokinases, its activity has been linked to downstream suppression of VEGF expression and thus angiogenesis. Unlike the mABs which are specific to a target, anti-angiogenic TKIs are known to have intracellular activity and downstream signalling a broad range of targets such as VEGFR, PDGFR, FGFR, TIE, c-KIT, RET and RAF. An overview of AGIs used in several solid organ cancers is presented in Table 1 along with molecular targets and indicated usage.

It has been shown that blockade of VEGF/VEGFR signalling axis by monotherapies often leads to transient responses followed by eventual resistance and progression (Zhao and Adjei, 2015; Haibe et al., 2020). Similarly, combination strategies that achieve tandem blockade of multiple signalling pathways have been proposed as a viable anticancer treatment strategy (Gotink and Verheul, 2010; Qin et al., 2019; Yi et al., 2019; Deng et al., 2020). Since the role of the Notch signaling pathway in Angiogenesis is well recognized (Akil et al., 2021; Majumder et al., 2021), targeting angiogenesis via inhibition of the Notch pathway is also considered as a viable anti-cancer treatment approach. However, apart from Bevacizumab (Axitinib to a limited extent due to its use in combinations with Immune Checkpoint Inhibitors) most AGIs have been used as monotherapy and as such they offer limited therapeutic benefit. Secondly, though the molecular targets of AGIs are known, the selection of AGIs is not on the basis of tumor molecular profiling; currently there is no reliable molecular biomarker for selection of these agents. It is pertinent to add that prior efforts at informing use of targeted agents based on limited (single gene) tumor molecular profiling have yielded discouraging outcomes. Owing to these challenges, the aim of precision oncology to offer personalized (combination) regimens remains unachieved.

We have previously reported (Nagarkar et al., 2019; Ranade et al., 2019) the clinical utility of multi-analyte tumor profiling (ETA: Encyclopedic Tumor Analysis) to identify latent actionable vulnerabilities in advanced refractory cancers and guide selection of safe and efficacious patient-specific combination regimens. We hypothesized that ETA can be used to inform personalized combination regimens of AGIs with other anti-neoplastic agents, which can yield superior or improved clinical outcomes in patients with advanced refractory solid organ cancers.

The findings of the present study show that co-administration of AGI and other anticancer agents are well tolerated and achieve significant response even in a heavily pre-treated cohort with advanced solid organ cancers. It is generally accepted that the probability of success of anticancer treatments decrease with successive lines of treatment. However, our study shows a significant increase in median PFS as compared to the patient’s last failed line of treatment. Since several patients are continuing to receive treatments at the time of submission, we anticipate further improvements to the treatment benefit metrics reported in this manuscript.

Presently, the selection of AGI for use in anticancer treatment regimens is not informed by molecular genetic indication. This is the case for most, if not all, targeted anticancer agents such as tyrosine kinase inhibitors (TKI) as well as mABs. Though the paradigm of personalized label agnostic treatment selection based on tumor molecular profiling has been explored for several targeted anticancer agents, this appears to be less so in case of AGIs where there are fewer studies on organ agnostic treatment selection. Among the NCI-MATCH (NCI-MATCH, 2020)basket trials, there are 3 trials for FGFR aberrations and 1 trial on cKIT aberrations for selection of AGI; interim results for one of these studies indicated absence of significant treatment benefit with 9% ORR (Chae et al., 2020). Most AGIs are administered as monotherapy which yield modest treatment benefits. Prior studies have established the benefits of combination regimens of various TKIs in comparison to monotherapy with the same agents. In several cancers, targeted agents are administered with other targeted or endocrine or cytotoxic agents; Axitinib and Bevacizumab appear to be researched more extensively in comparison to other AGIs reported in this manuscript. Several combinations of Axitinib and Bevacizumab with other targeted agents, checkpoint inhibitors or cytotoxic agents are presently approved for treatment of various cancers since they offer varying levels of improvement over existing monotherapy options. In a recent study (Donato et al., 2020) it was observed that VEGF blockade, though effective in suppressing the primary tumour, could present hypoxic conditions conducive to release and survival of CTC clusters with metastatic potential. It follows that a combination regimen that targets angiogenesis as well as other activated pathways or cellular mechanisms (such as e.g., DNA replication, DNA synthesis, or mitosis) in tandem could avoid such potential pitfalls by acting on CTCs or clusters released from the tumor which are no longer susceptible to inhibition of angiogenesis.

The present study was designed to select AGIs based on molecular indications for use in patient specific combination regimens with other anticancer drugs which were selected based on de novo functional and molecular profiling of the tumors. The scope of AGIs in the present study was limited by the availability of approved drugs at the trial location and included Axitinib, Bevacizumab, Pazopanib, Regorafenib, Sorafenib and Sunitinib; Imatinib, though not a classical AGI is known to inhibit angiogenesis as a downstream function and was also considered. Except for Bevacizumab, a mAb specific for VEGF, all other anticancer agents are small molecule TKIs with varying activities against a repertoire of angiogenesis-related factors including VEGFR, PDGFR, FGFR, c-KIT, RET, and RET families. Accordingly, ETA evaluated gene alterations including single nucleotide variations (SNV) and copy number alterations (CNA, specifically gain of copy number) as well as overexpression of all genes where the polypeptide product can be targeted by AGIs. While SNV and CNA were determined by NGS, gene expression was evaluated by NGS of tumor tissue RNA or exosomal RNA, as well as by ICC of peripheral blood C-TACs. The study cohort included 18 patients who were unable to undergo a biopsy for tumor profiling–in these patients, peripheral blood tumor analytes including circulating tumor DNA (ctDNA), exosomal RNA and C-TACs were evaluated.

The molecular profiles of angiogenesis related genes in each patient are illustrated in Figure 1. In 21 patients, molecular features associated with multiple angiogenesis related genes was observed while in 3 patients, multiple alterations in same gene (copy number variation as well as point mutation) were observed. This aided the selection of AGIs that can efficiently target multiple vulnerabilities. In 17 patients, ETA also identified other tumor features such as (but not limited to) activation of mTOR or hormone receptor signalling pathways. In these patients, the corresponding targeted agent or endocrine antagonist was evaluated for incorporation in the treatment regimen.

The study findings suggest that ETA based comprehensive multi-analyte cancer profiling provides relevant molecular and functional evidence which in turn can inform selection of patient-specific combination regimens of AGIs with other anticancer agents. The benefits of such combination regimens may exceed those obtained with arbitrary selection of monotherapy AGIs or their combination regimens, respectively. While the merits of combination treatment strategies have been expounded previously (Liu, Nikanjam, and Kurzrock, 2016; Nikanjam, Liu, and Kurzrock, 2016; Nikanjam et al., 2017), the safety of such regimens are an important consideration. Meta-analyses of clinical trials have observed that de novo combinations of targeted and cytotoxic agents can be safely administered to patients without any increased risks of toxicity. The present study comprised exclusively of heavily pre-treated cases of advanced cancers with an inherently higher risk of AEs due to cumulative toxicities from prior treatments; ETA guided combination regimens of AGI and other anticancer agents were generally well tolerated with manageable profile of treatment related AEs. Though the present study did not include any therapy naïve patients, we speculate that ETA-guided approach can be beneficial as first line for advanced metastatic cases or in the neoadjuvant setting for other (operable) cases. In conclusion, our study demonstrates 1) the clinical efficacy of combination regimens with AGI and other anticancer agents in yielding clinically meaningful response rates and survival benefits, and 2) the utility of multi-analyte tumor profiling as a decision support system for clinicians to design personalized treatment strategies for patients with advanced or refractory cancers.