Among the antibody drugs targeting the VEGF/VEGFR axis ( Table 1 ), Bevacizumab was approved primarily as an anti-angiogenic drug in various malignancies, and more clinical trials were taken to assess its precise anti-lymphangiogenic effects in combination therapy. A current study showed that, by inhibiting VEGF-A induced inflammatory (lymph) angiogenesis, Bevacizumab could benefit corneal pathologies and was studied as a novel strategy in corneal and ocular surface diseases (41). In 2011, a human IgG subclass 1 mAb (monoclonal antibody) specific to VEGFR-3, IMC-3C5, entered the clinical trial and has recently completed the phase 1 study. It was shown to be well tolerated up to a dosage of 30 mg/kg weekly (qwk) in advanced solid tumors, though the following expanded evaluation of its activity on colorectal cancer (CRC) was minimal (22). Similarly, VGX-100 was another highly specific monoclonal antibody that explicitly binds to VEGF-C and dampens VEGFR-2 and VEGFR-3 activation. It entered the phase 1 trial of late-stage cancers, among which the primary indications were glioblastoma and metastatic CRC (42). VGX-100 was well tolerated when it was used alone or combined with Bevacizumab, and 12% of the evaluated patients obtained a best response of durable stable disease for more than four months (23). Another preclinical stage study revealed a phage-derived single-chain fragment of anti-VEGF-C mAb (anti-VEGF-C scFv). By interacting with the epitope on VEGF-C, scFv showed distinct specificity and affinity. Significant advantages of such drugs can be shown in cancers depending on the direct VEGF-C pathway for growth, like Kaposi sarcoma and acute myeloid or lymphocytic leukemia. Researchers also speculated additional benefits of these inhibitors in VEGF-C induced bone and macular degeneration (28). An IgG-like fusion protein molecule, VEGFR-31-Ig, which could simultaneously bind the angiogenic VEGF-A and the lymphangiogenic VEGF-C, has been reported. It was remarkable for its outstanding stability and comparable effect with the complex activity of VEGF-Trap (25 mg/kg) and sVEGFR-3 (25 mg/kg), whereas excluding their conventional drawbacks like impractical clinical use and inadequate preclinical safety (27). Also, based on a current database analysis of the VEGF-C/VEGFR-3 triggered KRAS/MAPK-YAP1/Slug signaling in skin cancer progression, a unique anti-VEGFR-3 peptide was discovered able to abrogate the process (25).

Another axis targeted by antibody drugs was the HGF/HGFR signaling. A phase 3 study of Rilotumumab (AMG 102) with Epirubicin, Cisplatin and Capecitabine (CX) as first-line therapy in advanced MET-positive gastric or gastroesophageal junction adenocarcinoma was taken. However, the trial was terminated with a disappointing result as the Rilotumumab group presented a lower median overall survival (OS) and 12-month survival rate compared to the placebo arm, and it was associated with frequent fetal events such as neutropenia (29%) and anemia (12%) (34). Filcatuzumab was assessed in combination with Cetuximab for refractory head and neck squamous cell carcinoma due to minor benefits shown by Cetuximab alone. The combination had an acceptable safety profile and a promising anti-tumor activity by dual inactivation of HGF and EGFR pathways, indicating an active mitigation of cancer progression (33). Nonetheless, results from a phase 3 study of Onartuzumab with Erlotinib in non-small cell lung cancer (NSCLC) were in stark contrast to the previous phase consequence, partially due to the small size of the phase 2 study (29). Although it is dubious that the selection of patients group would affect, future investigations were expected to pick splice-site mutations harbored tumor samples (43). Taken together with the failure to meet the endpoint by many other MET-signaling inhibitors, researchers implied an unsatisfactory perspective of Ab therapy targeting this axis.

Fortunately, several agents inhibiting the interaction between Angiopoietin-1/-2 and the Tie2 receptor were demonstrated with promising efficacy. Trebananib, an angiopoietin neutralizing peptibody, prolonged the progression-free survival in a phase 3 trial of patients with recurrent epithelial ovarian cancer, proving the effectiveness of targeting this non-VEGF pathway (35). But some recent trials of Trebananib in combination with standard ovarian cancer chemotherapy didn’t succeed in the overall survival improvement (44). Later designed AMG 780 had a relatively longer half-life duration (8 to 13 days) than that of Trebananib (3.1 to 6.3 days), and the phase 1 study of AMG 780 suggested a maximum dose of up to 30 mg/kg every 2 weeks in patients with advanced solid tumors (37). Selective anti-Ang2 mAb MEDI3617 was studied for its recommended monotherapy dosage and effects when combined with Bevacizumab and several other chemotherapies. Within the recommended single-agent dosage (1000 mg Q2W; 1500 mg Q3W), there was no worsening of adverse effects like edema in combined treatments, yet the study of MEDI 3617 monotherapy in advanced ovarian cancers was discontinued due to the observation of peripheral edema (38). More recently, Nesvacumab showed preliminary anti-tumor activity in a phase 1 trial and the safety profile was acceptable at all dose levels tested (39). Combination therapy of Nesvacumab with Aflibercept was also well tolerated (45). Some additional investigations were taken in retinal diseases like macular degeneration though they failed to meet the phase 2 endpoint (46). Some chemically programmed antibodies behaving as potent endothelin receptor antagonists, like CVX-060 and CVX-241, were designed to target Ang2. However, CVX-241 was prematurely terminated because of the absence of ideal pharmacological effects through the 25 mg/kg cohort and shorter than expected half-life duration (40). CVX-060 plus Axitinib were tested in patients with previously treated metastatic renal cell cancer, and as a result of the higher than anticipated treatment-related thromboembolic events, enrollment to the next part was discontinued, but alternate doses and/or disease settings were considered (36). Although peripheral edema occurred as common toxicity that appeared in anti-angiopoietin treatments, all the known cases were limited to grade 1 or 2, and no anti-VEGF therapy associated adverse effects were observed. Therefore, it might be noteworthy to combine anti-angiopoietin drugs with conventional inhibitors of VEGF pathways to further augment the overall survival in some refractory cases.

So far, FDA has approved several tyrosine kinase inhibitors (TKI) targeting VEGFRs for their significant improvement of progression-free survival in patients with renal cell carcinoma, thyroid cancer, hepatocellular carcinoma, and multiple other tumor types ( Table 2 ). For example, in addition to the approved anti-angiogenic activity in malignancies, Pazopanib demonstrated favorable inhibitory effects on tumor growth, lymphatic metastasis, and tumor lymphangiogenesis in an orthotopic model of mice with colorectal cancer (56). Lenvatinib had versatile applications and predominantly suppressed lymph node metastasis in a breast cancer model, and for further studies of Lenvatinib-therapies, guiding biomarkers and sufficient selection of patients are needed (53). Triple angiokinase inhibitor, Nintedanib, adequately dampened lymphangiogenesis stimulated by VEGF-C, bFGF, and PDGF-BB in a suture-induced corneal neovascularization assay. Both topical and systemic applications showed positive results, though the in vivo safety of topical use wasn’t tested yet (52). Interestingly, other than the clinical application in renal cell carcinoma, Axitinib significantly inhibited LEC proliferation and lymphangiogenesis in a study of allergic asthma (68). An earlier establishment of Sunitinib in the breast cancer model presented marked reduction of tumor lymphatic and blood vessels density, as well as decreased axillary lymph nodes invasion, suggesting a prosperous blockade of LEC cellular function by dampening VEGFR-2/-3 activation from VEGF-C/-D (54). However, a controversy arose based on the observation of promoted lymphangiogenesis caused by the increase of VEGF-C transcription and mRNA stabilization in clear cell renal cell carcinomas (55). The paradoxical effects of Sunitinib treatment in different cancer models suggested a variation of LEC response in different tissues, implying a more appropriate patient sample will be needed for further confirmation of many other drugs also. It is noticeable that, as for future studies of these multi-kinase inhibitors, management of some common adverse effects, including drug-induced hypertension and dermatotoxicity like the hand-foot skin reaction, should be considered while estimating the clinical parameters.

Concurrently, there are extensive ongoing trials and preclinical studies of novel therapeutics. Investigations have been taken to test the synergistic effect of combining Vatalanib (750 mg BID) with the mTOR inhibitor, Everolimus (10 mg QD), and a 20% increase in the response rate was detected through this combination (compared to Everolimus therapy alone) in neuroendocrine tumors (69). Selective dual inhibitor Brivanib was studied in patients with hepatocellular carcinoma who were intolerant to Sorafenib or for whom Sorafenib failed. Still, the trial didn’t meet the primary endpoint of overall survival, and similar negative results were taken by several other phase 3 trials of hepatocellular carcinoma (HCC) (58). Favorable results were shown in combination therapy of Cediranib and Olaparib (PARP inhibitor) with significantly extended effectiveness than Olaparib alone in patients having relapsed, platinum-sensitive ovarian cancer (57), which was consistent with previous studies. Moreover, preclinical studies of MAZ51 (a selective VEGFR-3 inhibitor) established a valid blockade of prostate tumor growth (70) and melanoma metastasis (62), as well as a 30% reduction in gastric cancer cell migration (71). Another highly selective strategy using SAR131675 has presented diminished lipotoxicity-induced lymphangiogenesis in a type 2 diabetic nephropathy model (66). AD0157 (a marine fungal metabolite), as a natural anti-angiogenic compound, was shown to possess a pro-apoptotic property in human myeloid leukemia cells by inducing the caspase-dependent cascades (64). Concomitantly, other than the caspase-dependent apoptosis, it could also attenuate tumor-associated lymphangiogenesis and metastatic dissemination to both regional lymph nodes and distant organs by decreasing VEGFR-3/-2, ERK1/2, and Akt phosphorylation in human myeloid leukemia cells (65).

The other potent group of TKI is the MET kinase inhibitor ( Table 3 ). In the recent two years, FDA granted approval to selective MET inhibitor Capmatinib and Tepotinib for adult patients with metastatic NSCLC whose tumors were MET exon 14 mutated. However, adverse effects like phototoxicity and interstitial lung diseases were reported (74). In the study of EMD 1214063 and EMD 1204831, both candidates triggered tumor regression in a dose-related fashion by inhibiting c-MET phosphorylation, regardless of HGF-dependent or HGF-independent activation. Comparatively, the former was more sustained. A single dose of EMD 1214063 at 10 or 30mg/kg could achieve a complete and persistent target inhibition, making it better than many other MET inhibitors (81). PF-04217903 is one of the most selective c-MET inhibitors. It provided a robust TGI (tumor growth inhibition) in the MET-amplified GTL-16 model and a U87MG model exhibiting the HGF/c-MET autocrine loop (80). In clinical trials, multi-kinase inhibitor Crizotinib was approved for the treatment of ALK-positive NSCLC. It was relatively non-toxic though resistance may occur within one year (72). Several other drugs were discontinued in the study, like AMG 337, which was terminated early because of low response (13% of evaluable patients) observed during the phase 2 trial partially due to the larger sample and narrower range of tumor types (77).

A previous study utilizing the CI66 tumor model suggested that the siRNA-mediated VEGF-C gene silencing could efficiently suppress tumor lymphangiogenesis as well as recruitment of inflammatory cells in TME (tumor microenvironment) and calcium carbonate nanoparticle was believed to be a prosperous vector for it ( Table 4 ). Furthermore, a higher level of dendritic cells and concanavalin A-induced proliferation in tumor-associated leukocytes was associated with the siVEGF-C application, indicating an additional positive modulation of the immune response (95). In vitro study of Fucoxanthin showed decreased lymph node metastases and significant reduction of in vivo LVD (lymphatic vessel density) in an MDA-MB-231 breast cancer model. Enhanced blocking of matrix metalloproteinases (MMP) secretion and increased TIMP-1 protein expression were also demonstrated in the experimental data of Fucoxanthin (84). Shikonin, according to the report, possessing a favorable inhibitory effect of NF-kB nuclear translocation and activation, causing the decrement of LEC cord formation (87). As an endogenous angiogenic inhibitor, Kallistatin was shown to be a potential dual-effect substance disrupting both VEGFR-2 & VEGFR-3. Because of severe adverse effects appearing in VEGFR TKIs treatment, combination therapy of Kallistatin with the kinase inhibitors could thus be considered a new strategy (85). Meanwhile, a highly metastatic human lung adenocarcinoma cell line Anip973 revealed an increase in both VEGF-C & COX-2 immunoreactivity, indicating a possible correlation in between. Therefore COX-2 (largely from PGE2-stimulated EP4 receptors) and EP4 may serve as novel targets for treatment in light of its potential regulatory activity on VEGF-C expression (83). Many other agents like Fucoidan (89), Aspirin (90), Hedyotis diffusa Willd (EEHDW) (91), and Curcumin (93) were able to block the important downstream signalings of VEGFR-3 in various cancer models. Interestingly, an earlier study explicated an advantageous inhibitory effect of VEGF-A induced lymphangiogenesis and sentinel lymph node metastasis in oral cancer by 3AOA. Besides the VEGF-A inhibition, 3-O-Acetyloleanolic acid (3AOA) also interfered with an alternative lymphangiogenic pathway mediated by Ang-1/Tie-2 in the CT-26 colon carcinoma model. 5 μM 3AOA could significantly dampen 87% of human umbilical vein endothelial cells (HUVEC) proliferation, migration, and tube formation stimulated by Ang-1 (96, 97). Similarly, 6,8-Diprenylgenistein (6,8-DG), an isoflavonoid isolated from Cudrania tricuspidata has been reported with its inhibitory effect on VEGF-A induced lymphangiogenesis as well. In the oral cancer model, 6,8-DG inhibited VEGF-A expression and blocked the VEGFR-2 interaction with VEGF-A, decreasing cervical lymph nodes metastasis of the oral squamous cell carcinoma (101). Newly examined histone deacetylase inhibitors Trichostatin A (TSA) (100) and Dimethyl fumarate (DMF) (94) both caused G1 cell cycle arrest in a p21-dependent manner. DMF had a pronounced anti-tumorigenic effect for melanoma and a cell type-specific cell cycle arrest. No apoptotic influence was observed in DMF-treated human DLEC (dermal lymphatic endothelial cells), while TSA, on the other hand, downregulated the anti-apoptotic proteins cIAP-1/2, causing cell death.