Genetic and ethnic modulation of cardiovascular toxicity of vascular endothelial growth factor inhibitors

Abstract

Vascular endothelial growth factor (VEGF) inhibitors, including monoclonal antibodies and tyrosine kinase inhibitors (TKIs), are important as anticancer treatments through curbing tumour angiogenesis and growth. VEGF inhibitors have significant cardiovascular effects. By blocking VEGF receptors, ligands, or signal pathways, VEGF inhibitors disturb the balance between vasodilation and vasoconstriction, undermine endothelial cell integrity, and activate cardiomyocyte apoptosis. VEGF inhibitors increase risks of hypertension, heart failure, thromboembolism and arrhythmia. Genetic and geographic studies showed that genetic polymorphisms likely play significant predictive or prognostic roles in cardiovascular toxicity associated with VEGF inhibitors. This review updates current understandings of VEGF inhibitors on cardiovascular toxicity, explores potential mechanisms, and clarifies whether genetic or ethnic factors contribute to their adverse effects.

Key Messages

• VEGF inhibitors disturb the balance between vasodilation and vasoconstriction, undermine endothelial cell integrity and activate cardiomyocyte apoptosis.

• VEGF inhibitors increase risks of hypertension, heart failure, thromboembolism and arrhythmia.

• Genetic and geographic studies showed that genetic polymorphisms likely play significant predictive or prognostic roles in cardiovascular toxicity associated with VEGF inhibitors.

Keywords:

• Angiogenesis inhibitors
• antibodies
• cardiovascular system
• cardiotoxicity
• ethnic groups
• tyrosine kinase inhibitors
• vascular endothelial growth factors

Introduction

Onco-cardiology is critical for optimal cancer care, since cancer and cardiovascular diseases (CVDs) often coexist in middle-aged and senior patients. Anticancer therapies may aggravate CVD, and their side effects may cause cancer treatments to fail. Angiogenesis inhibitors are novel agents that inhibit tumour growth and angiogenesis by blocking vascular endothelial growth factor (VEGF) receptors or interfering in their signal pathways [1]. VEGF monoclonal antibodies, such as bevacizumab [2], and tyrosine kinase inhibitors (TKIs) with anti-VEGF activities, such as sunitinib [3], sorafenib [4], pazopanib [5] and axitinib [6], were proven to prolong the survival of cancer patients with various advanced or metastatic malignant solid cancers (Table 1). However, VEGF can promote endothelial proliferation, angiogenesis, and the production of nitric oxide (NO) and prostacyclin, and reduce endothelin-1 [7,8]. Theoretically, anti-VEGF agents can also undermine normal angiogenesis and cardiovascular homeostasis.

Table 1. Vascular endothelial growth factor inhibitors.

Name	Mechanism	Route	Indication	Cardiovascular toxicity (reported)
Monoclonal Antibody
Bevacizumab	VEGF-A monoclonal antibody	IV	Colorectal cancer, non-squamous non-small cell lung cancer, glioblastoma, renal cell carcinoma, carcinoma of the cervix, epithelial ovarian, fallopian tube	Hypertension, heart failure, thromboembolic events, proteinuria, thrombotic microangiopathy, conduction disorder
Aflibercept	Monoclonal antibody to VEGF-A, VEGF-B, PGF	IV	Colorectal cancer	Hypertension, heart failure, thromboembolic events, proteinuria, thrombotic microangiopathy
Ramucirumab	VEGFR2 monoclonal antibody	IV	Colorectal cancer, non-small cell lung cancer, gastric cancer	Hypertension, thromboembolic events, proteinuria
Multi-Targeting Tyrosine Kinase Inhibitors with an Anti-VEGF Effect
Sunitinib	VEGFR1-3, PDGFRα/β, KIT, FLT3, CSF1R, RET	PO	Gastrointestinal stromal tumour, renal cell carcinoma, pancreatic neuroendocrine tumour	Hypertension, heart failure, myocardial ischemia/infarction, proteinuria, thrombotic microangiopathy, QTc prolongation, torsades de pointes
Sorafenib	VEGFR2, PDGFRβ, RAFs, KIT, FLT3, RET	PO	Hepatocellular carcinoma, renal cell carcinoma, differentiated thyroid carcinoma	Hypertension, heart failure, myocardial ischemia/infarction, proteinuria, QTc prolongation
Pazopanib	VEGFR1-3, PDGFRα/β, FGFR1/3, KIT, ITK, LCK, c-FMS	PO	Renal cell carcinoma, soft-tissue sarcoma	Hypertension, heart failure, myocardial ischemia/infarction, proteinuria, thrombotic microangiopathy, QTc prolongation, torsades de pointes
Axitinib	VEGFR1-3, PDGFRα/β, KIT	PO	Renal cell carcinoma	Hypertension, heart failure, thromboembolic events, proteinuria
Vandetanib	VEGFR1-3, EGFRs, RET, BRK, TIE2, EPHRs, Src kinases	PO	Medullary thyroid cancer	Hypertension, heart failure, myocardial ischemia/infarction, proteinuria, QTc prolongation, torsades de pointes
Regorafenib	VEGFR1-3, PDGFRα/β, FGFRs, RAFs, RET, KIT, TIE2, DDR2, TRKA, EPH2A, SAPK2, PTK5, Abl	PO	Gastrointestinal stromal tumour, colorectal cancer	Hypertension, myocardial ischemia/infarction, proteinuria
Cabozantinib	VEGFR1-3, MET, AXL, RET, ROS1, TYRO3, MER, KIT, TRKB, FLT3, TIE2	PO	Renal cell carcinoma	Hypertension, thrombotic events, proteinuria
Lenvatinib	VEGFR1-3, PDGFRα, FGFRs, KIT, RET	PO	Differentiated thyroid cancer, renal cell carcinoma	Hypertension, heart failure, thromboembolic events, proteinuria, QTc prolongation
Under Investigation
Tivozanib	Tyrosine kinase inhibitor, under phase III trial for renal cell carcinoma
Lucitanib	Tyrosine kinase inhibitor, under phase II trial for breast cancer

VEGF: vascular endothelial growth factor; VEGFR: vascular endothelial growth factor receptor; PGF: placental growth factor; PDGFR: platelet-derived growth factor receptor; FLT: feline McDonough sarcoma (FMS)-like tyrosine kinase; CSF1R: colony stimulating factor 1 receptor; RET: rearranged during transfection; RAF: rapidly accelerated fibrosarcoma; FGFR: fibroblast growth factor receptor; ITK: interleukin-2-inducible T-cell kinase; LCK: lymphocyte-specific protein tyrosine kinase; FMS: feline McDonough sarcoma; EGFR: epidermal growth factor receptor; BRK: breast tumour kinase; TIE: tyrosine kinase with immunoglobulin-like and epidermal growth factor-like domains; EPH: ephrins; Scr: sarcoma; DDR: discoidin domain receptor; TRK: tropomyosin receptor kinase; SAPK: stress-activated protein kinase; PTK: protein tyrosine kinase; ABL: Abelson murine leukaemia; MET: mesenchymal epithelial transition; AXL: anexelekto; ROS1: proto-oncogene tyrosine-protein kinase ROS; TYRO3: TYRO3 protein tyrosine kinase; MER: membrane oestrogen receptor; IV: intravenous; PO: oral (Reference: DailyMed; ClinicalTrials.gov).

Genetic and geographic studies showed that genetic polymorphisms or ethnic differences probably play significant predictive or prognostic roles in cardiovascular toxicity associated with VEGF inhibitors. Genetic variants may increase or reduce the potential toxicity of VEGF inhibitors, influencing their short- or long-term treatment outcomes. Accordingly, genetic monitoring may optimize the application of VEGF inhibitors in cancer treatment. This review updates current understandings of VEGF inhibitors on cardiovascular toxicity, explores potential mechanisms and clarifies whether genetic or ethnic factors contribute to their adverse effects.

Methodology

We examined the publications in PubMed from 2004 to 2017 by combining the keywords “vascular endothelial growth factor” and “cancer”, and identified associated publications with the focus on VEGF inhibitors by the keywords of “angiogenesis inhibitor”, “monoclonal antibody”, “tyrosine kinase inhibitor”, “protein kinase inhibitor”, “bevacizumab”, “aflibercept”, “ramucirumab”, “sunitinib,” “sorafenib, “pazopanib”, “axitinib”, “vandetanib”, “regorafenib”, “cabozantinib”, or “lenvatinib”, respectively. We also searched literatures in PubMed for VEGF inhibitor-induced cardiotoxicities by the keywords of “hypertension”, “heart failure”, “thromboembolism”, “cardiac ischemia”, “arrhythmia” or “cardiovascular toxicity” with or without related to “genetic polymorphism” and “racial”. In the 925 publications screened, we excluded case reports, disease reviews, editorials, duplicates, non-English articles and the studies not related to adult cancer therapy or VEGF-inhibitor cardiovascular toxicity, but additionally included some references from the published clinical trials. Under this setting, a total of 398 articles were included and reviewed.

The grading system of cardiovascular toxicity of VEGF inhibitor in this review and associated clinical trials is based on Common Terminology Criteria for Adverse Events version 1, 2, 3 or 4 [9]. Grade 3 or higher represents high grade of adverse events. The grading criteria of hypertension or raised blood pressure is as followed: grade 1, systolic blood pressure (BP) 120-139 mmHg or diastolic BP 80-89 mmHg; grade 2, recurrent or persistent (≥24 h) systolic BP 140-159 mmHg or diastolic BP 90-99 mmHg, or symptomatic diastolic BP increases by >20 mmHg, and monotherapy might be indicated; grade 3, systolic BP ≥ 160 mmHg or diastolic BP ≥ 100 mmHg and more than one drug or more intensive therapy than previously might be indicated; grade 4, life-threatening consequences, such as malignant hypertension, transient or permanent neurologic deficit and hypertensive crisis, and urgent intervention might be indicated [9]. The grading criteria of QTc prolongation is as followed: grade 1, QTc 450-480 ms; grade 2, QTc 481-500 ms; grade 3, QTc ≥ 501 ms on at least two separate electrical leads; grade 4, QTc ≥ 501 ms or > 60 ms change from baseline and torsade de pointes or polymorphic ventricular tachycardia or signs/symptoms of serious arrhythmia [9].

Cellular and molecular mechanisms of cardiovascular toxicity of VEGF inhibitors

VEGFs promote endothelial cell proliferation, induce two vasodilators, NO, and prostacyclin, and reduce the vasoconstrictor, endothelin-1 (Figure 1) [7,8]. By blocking VEGFs, their receptors, or signal pathways, VEGF inhibitors undermine endothelial cell integrity and the balance between vasodilation and vasoconstriction (Figure 2) [7,8]. Endothelial cells are important components for maintaining vascular homeostasis and avoiding haemorrhaging and thrombosis. Thus, VEGF inhibitors interfere with normal vascular homeostasis, leading to higher tendencies for hypertension and thrombosis [7,8]. Previous studies reported similar manifestations between cardiovascular toxicity associated with VEGF inhibitors and pregnancy-associated cardiovascular complications, such as preeclampsia and peripartum cardiomyopathies [10,11]. VEGF inhibitors induce a syndrome-like preeclampsia, including hypertension, proteinuria and deteriorating renal function. Renal biopsies of such patients revealed thrombotic microangiopathy, a feature similar to preeclampsia [12–14]. In addition, soluble feline McDonough sarcoma (FMS)-related tyrosine kinase 1 (FLT1), also called soluble VEGF receptor 1 (VEGFR1), can cause preeclampsia and peripartum cardiomyopathies [15,16]. Like VEGF inhibitors, FLT1, secreted by the placenta in late gestation, blocks the VEGF signal pathway [16], indicating a potential causal relation between VEGF inhibitors and their cardiovascular toxicity.

Figure 1. Vascular endothelial growth factor (VEGF) and its associated signal pathways maintain vascular homeostasis through promoting endothelial cell proliferation, inducing two vasodilators, nitric oxide and prostacyclin, and reducing endothelin-1, a vasoconstrictor.

Figure 2. By blocking vascular endothelial growth factors (VEGFs), their receptors, or signal pathways, VEGF inhibitors undermine endothelial cell integrity and the balance between vasodilation and vasoconstriction, leading to hypertension, cardiac ischemia, thromboembolism, heart failure and microangiopathy.

VEGFs also play pivotal roles in cardiomyocyte protection. VEGFs protect against doxorubicin-induced apoptosis of cardiomyocytes through upregulating the pro-survival Akt/nuclear factor (NF)-κB/bcl-2 signalling pathway [17]. Overexpression of VEGFs in doxorubicin-induced heart failure (HF) increases cardiac mitochondrial respiration, thereby preserving the left ventricular volume [18]. In pacing-induced dilated cardiomyopathy, VEGFs activate the antiapoptotic regulator, Akt, thereby enhancing cardiomyocyte survival [19]. VEGFs also abolish hypoxia-induced apoptotic factors, such as cleaved caspases-3, -8 and -9, as well as Bax in neonatal cardiomyocytes [20]. Therefore, it is possible that VEGF inhibitors undermine these protective effects of VEGFs and activate cardiomyocyte apoptosis, leading to cardiac systolic dysfunction and HF.

While VEGF monoclonal antibodies mainly focus on VEGF receptors or ligands, TKIs not only block the VEGF signal pathway through targeting VEGF receptors (VEGFRs) and platelet-derived growth factor receptor, but also block multiple other kinases or signal pathways associated with tumour growth and inflammation, such as KIT kinases, colony stimulating factor 1 receptor (CSF1R), rearranged during transfection (RET), rapidly accelerated fibrosarcoma (RAF), fibroblast growth factor receptor (FGFR) and so on (Table 1). This multi-targeted property of TKIs may expand or strengthen their anticancer efficacy, but may also increase potential cardiovascular toxicity, such as HF [21,22].

Although the mechanisms of reduced left ventricular systolic function are not clear, it is hypothesized that TKIs inhibit multiple off-target kinase activities and the VEGF signalling pathway, leading to cardiac dysfunction [23,24]. Sunitinib was shown to inhibit ribosomal S6 kinase and AMP-activated protein kinase (AMPK), promoting ATP depletion and potential cardiomyocyte apoptosis [23,24]. Sorafenib, through blocking RAF1, inhibits the extracellular signal-regulated kinase kinase cascade, undermining regulation of cardiomyocyte survival, especially under stress [23]. Currently, it remains unclear whether inhibition of RAF1 by sorafenib also activates two stress-induced, proapoptotic pathways associated with apoptosis, viz., signal-regulating kinase 1 and mammalian sterile 20 kinase 2 (MST2), which could lead to even more profound effects of sorafenib on cardiomyocyte apoptosis and potential HF [23]. These potential detrimental effects of TKIs on regulating cardiomyocyte apoptosis and survival as well as energy utilization were hypothesized to cause cardiac dysfunction and HF.

Furthermore, VEGF inhibitors may also induce arrhythmia. The mechanism of drug-induced QTc prolongation was speculated to be a direct interaction between the studied drugs and the human ether-a-go-go-related gene (hERG) potassium channels, interfering with delayed rectifier potassium currents by delaying cardiac repolarization [25–28]. In preclinical studies, sunitinib and vandetanib were found to have a significant interaction with the hERG channel [29]. In addition, cardiac ischemia and HF induced by TKIs can cause QTc prolongation [30–32], thus the electrophysiological effects of sunitinib and vandetanib on the hERG channel may aggravate the risk of ventricular arrhythmia related to QTc prolongation in cancer patients.

Although there are limited clinical reports about atrial fibrillation (AF) associated with TKIs, sunitinib was reported to inhibit AMPK [23], a compensatory pathway activated by AF [33]. Activation of AMPK by AF helps maintain atrial calcium handling and atrial contractions [33]. AMPK was hypothesized to protect paroxysmal AF from progressing into chronic AF, while fractional AMPK phosphorylation increased with paroxysmal AF and decreased with persistent AF [33]. Therefore, further studies are needed to clarify this potential interaction and relationship between sunitinib and AF through blocking AMPK.

VEGF inhibitor-induced hypertension

Hypertension is the most common cardiovascular toxicity of VEGF inhibitors. One meta-analysis of 12,949 patients treated with or without bevacizumab for various solid tumours from 19 randomized controlled trials (RCTs) revealed that the incidence of all cases of significantly elevated blood pressure was 24%, while high-grade elevated blood pressure was 8% [34]. Bevacizumab increased the risk of significantly elevated blood pressure (relative risk [RR]: 5.38, 95% confidence interval [CI]: 3.6 3 ∼ 7.97) [34]. A dose-dependent association was also identified. The RRs for significantly elevated blood pressure associated with low-dose (2.5 mg/kg per week) and high-dose bevacizumab (5 mg/kg per week) were 4.11 (95% CI: 2.4 9 ∼ 6.78) and 7.17 (95% CI: 3.9 1 ∼ 13.13), respectively [34]. VEGF TKIs, such as sunitinib and sorafenib, also have similar incidences and risks of significantly elevated blood pressure. One meta-analysis of 4999 patients treated with or without sunitinib from 13 prospective trials revealed that the incidence of all cases of elevated blood pressure was 21.6%, while high-grade elevated blood pressure was 6.8% [35]. Sunitinib significantly increased the risk of high-grade elevated blood pressure (RR: 22.72, 95% CI: 4.4 8 ∼ 115.29) [35]. Another meta-analysis of 4599 patients using sorafenib showed that the incidence of all cases of elevated blood pressure was 23.4%, while high-grade elevated blood pressure was 5.7% [36]. Other TKIs and monoclonal antibodies, including pazopanib [37], axitinib [6], lenvatinib [38], ramucirumab [39] and aflibercept [40], all showed similar or even higher incidences of elevated blood pressure in phase III trials or meta-analyses.

Currently, it is still controversial as to whether a higher risk of hypertension induced by VEGF inhibitors predicts better efficacy and outcomes of cancer treatment. Previous studies showed that hypertension induced by VEGF inhibitors was associated with better treatment outcomes with sunitinib for renal cell carcinoma [41,42], sorafenib for hepatocellular carcinoma [43], and bevacizumab for non-small cell lung cancer, colorectal cancer, ovarian cancer, and breast cancer [44–48]. However, one systemic review of 6487 patients from seven phase III trials of bevacizumab for various cancers found no significant association between survival outcomes and early development of hypertension (blood pressure increased by >20 mmHg systolic or >10 mmHg diastolic within the first 60 days of treatment) [49]. One retrospective study of 337 patients treated with pazopanib for non-adipocytic soft-tissue sarcomas showed no correlation between hypertension and treatment efficacy within 5 weeks of treatment [50]. Thus, further investigations are needed to clarify these clinical implications between the risk of hypertension and treatment efficacy.

VEGF inhibitor-induced thromboembolisms

Thromboembolisms are another potential cardiovascular toxicity of VEGF inhibitors. VEGF inhibitors, including monoclonal antibodies and TKIs, have a significant association with arterial thromboembolisms, including cerebrovascular and cardiovascular ischemia or infarction [51–54]. In one meta-analysis of 1745 patients from five RCTs for various cancers, compared to chemotherapy alone, chemotherapy with bevacizumab was associated with a higher risk of arterial thromboembolic events (ATEs) (hazard ratio [HR]: 2.0, 95% CI: 1.0 5 ∼ 3.75), but not with venous thromboembolic events (VTEs) [51]. Two other larger meta-analyses also revealed that bevacizumab increased the risk of ATEs with no dose-dependent relation between low- and high-dose bevacizumab [52,53]. Of these two meta-analyses, one study involving 12,617 patients from 20 RCTs revealed that bevacizumab increased the risk of ATEs (RR: 1.44, 95% CI: 1.0 8 ∼ 1.91), with a higher risk of cardiac ischemia (RR: 2.14, 95% CI: 1.1 2 ∼ 4.08) and no significant risk of cerebral ischemia (RR: 1.37, 95% CI: 0.6 7 ∼ 2.79) [53]. The other study with 13,026 patients from 20 RCTs also showed that bevacizumab increased the risk of ATEs (RR: 1.46, 95% CI: 1.1 1 ∼ 1.93) but provided no further information about different types of ATE [52]. Recently, two newly introduced VEGF monoclonal antibodies, ramucirumab and aflibercept, were also reported to induce higher incidences of arterial embolism, compared to chemotherapy alone in their phase III trials [39,40]. VEGF-targeted TKIs are also associated with higher risks or incidences of ATEs in clinical trials or meta-analyses. One meta-analysis of 10,225 subjects from 10 studies revealed that sunitinib and sorafenib brought a higher risk of ATEs (RR: 3.03, 95% CI: 1.2 5 ∼ 7.37) [54]. One phase III trial of 392 patients treated with or without lenvatinib for differentiated thyroid cancer refractory to radioiodine also revealed a higher incidence of ATEs (5% vs. 2%) [38].

The risk of VTEs with VEGF inhibitors remains controversial. Although one meta-analysis of 7956 patients from 15 RCTs in 2008 revealed a higher risk of VTEs with bevacizumab [55], another meta-analysis of 1745 patients from five RCTs in 2007 showed no significant association between VTEs and bevacizumab [51]. One larger recent meta-analysis of 6055 patients from 10 RCTs in 2011 also showed no significant association between VTEs and bevacizumab [56]. The author of the 2011 study speculated that the difference in results among these trials probably came from different study designs and methodologies [56]. While the 2011 study pooled individual patient data to reduce the limitation of a meta-analysis of aggregated data, the 2008 study developed its meta-analysis by summarizing rates of VTEs and included trials which did not report all-grade VTEs and did not distinguish between venous and arterial events [56]. As for TKIs, one recent meta-analysis of 7441 patients from 17 RCTs in 2013 showed no significant association between VTEs and TKIs [57]. However, further clarifying investigations are still needed to resolve these conflicting reports.

VEGF inhibitor-induced HF

The use of VEGF TKIs is commonly associated with asymptomatic reduction in left ventricular systolic function and symptomatic HF [21,22]. In contrast, bevacizumab-associated HF was infrequently reported in some trials of advanced breast cancer with paclitaxel, capecitabine, or anthracycline, an indication for which it is not now approved [58].

One meta-analysis involving 10,647 patients treated with different TKIs from 16 phase III trials and 5 phase II trials revealed that the incidence of all-grade HF was 2.39%, while that of high-grade HF was 1.19% [22]. TKIs increased risks of both all-grade (RR: 2.69, 95% CI: 1.8 6 ∼ 3.87) and high-grade HF (CHF; RR: 1.65, 95% CI: 0.7 3 ∼ 3.70) [22]. Although it is hypothesized that more specific TKIs possibly had lower cardiac toxicity [24], this analysis found that more specific TKIs, such as axitinib, and non-specific TKIs, including sunitinib, sorafenib, vandetanib and pazopanib, seemed to have similar risks of HF [22]. One retrospective study of 75 patients treated with sunitinib suggested that the incidences and impacts of TKI-associated HF were probably higher in the real world (28% for a ≥10% reduction in the left ventricular ejection fraction (LVEF); 19% for a ≥15% reduction in LVEF; and 8% for clinical HF), which was significantly higher than the risk reported in clinical trials [59].

VEGF inhibitor-induced cardiac arrhythmogenesis

VEGF inhibitors were reported to induce arrhythmia in clinical trials and post-market reports. Bevacizumab was found to induce supraventricular arrhythmia and conduction disorders in one retrospective analysis, whereas bevacizumab-induced cardiac ischemia was hypothesized to induce these arrhythmic events [60]. Moreover, interactions of TKIs with the hERG channel [29] and AMPK pathway [23] or genesis of cardiac ischemia [54] and HF (22) by TKIs also increase potential arrhythmic risks.

Vandetanib and sunitinib were reported to possess greater arrhythmic risks. Moreover, it was reported that there was a dose-dependent relation about QTc prolongation induced by vandetanib [61] and sunitinib [62]. One meta-analysis of a total of 6548 patients from 18 clinical trials for arrhythmic events of TKIs, including sunitinib, pazopanib, axitinib and vandetanib, revealed that patients receiving TKIs had a significantly higher risk of all-grade (RR: 8.66, 95% CI: 4.9 2 ∼ 15.2) and high-grade (RR: 2.69, 95% CI: 1.3 3 ∼ 5.44) QTc prolongation than those receiving no TKIs [61]. However, in that study, sunitinib and vandetanib were two major contributors to the significantly increased risk of QTc prolongation. That study included only two trials of pazopanib with a total of 806 patents and only one trial of axitinib with 630 patients, leading to limited study power for pazopanib and axitinib [61]. Therefore, the arrhythmic risks of other TKIs, including sorafenib, pazopanib, axitinib, cabozantinib, ponatinib and regorafenib, remain unclear [29,61].

Ethnic and genetic modulation of VEGF inhibitors-induced hypertension

Ethnic differences are common in pharmacological treatment. Previous studies revealed that there was an ethnic difference in VEGF inhibitors-induced hypertension. One prospective observational cohort study including 227 African American patients and 1589 Caucasian patients with metastatic colorectal cancer receiving first-line bevacizumab therapy showed African Americans (13.7%) had a higher rate of worsening hypertension than Caucasian patients (8.9%), but had similar rates of arterial (1.8% vs. 2.3%) and venous (5.3% vs. 4.8%) thromboembolism [63]. African Americans are known to have the highest incidence, prevalence, and severity of hypertension among different ethnic groups [64–66], the cardiovascular toxicity of VEGF inhibitor may potentiate cardiovascular risks for this group, although clinical reports are limited.

One meta-analysis of 4308 patients with lung cancer showed that non-Asians had a higher risk of severe bleeding (RR: 2.17, 95% CI: 1.1 4 ∼ 4.15) and thromboembolisms (RR: 3.65, 95% CI: 1.8 9 ∼ 7.04) and a lower risk of severe proteinuria (RR: 0.43, 95% CI: 0.2 8 ∼ 0.66) than Asian populations, while severe hypertension and haemoptysis were similar in the two groups [67]. In one phase III trial of 723 patients treated with axitinib or sorafenib for renal cell carcinoma, a higher incidence of all-grade or high-grade hypertension occurred in a Japanese population (axitinib: all-grade 64%, high-grade 44%; sorafenib: all-grade 62%, high-grade 45%) than in the overall population (axitinib: all-grade 40%, high-grade 16%; sorafenib: all-grade 29%, high-grade 11%) [68].

One multinational, prospective study of 325 Asians and 4046 non-Asians treated with sunitinib for renal cell carcinoma revealed that Asians (23%; 4%) had a similar incidence of all-grade and high-grade hypertension to non-Asians (21%; 5%), but Asians treated in Asia had a higher incidence (30%) of all-grade hypertension than Asians treated outside Asia (8%) [69]. The conclusion of that study suggested that even environmental factors should be considered with this difference, since these factors are difficult to control and evaluate [69]. One pharmacokinetics study of axitinib for Japanese, Chinese and Caucasians suggested that there was no significant pharmacokinetic difference between Asians and Caucasians [70], possibly indicating that clinical differences in VEGF therapy were based on other types of genetic polymorphisms rather than metabolic differences.

Genetic polymorphisms are also important factors regulating pharmacological responses and treatment outcomes. Some genetic studies showed that genetic polymorphisms or markers were able to predict bevacizumab- or sunitinib-induced hypertension (Table 2) [46, 71–78]. Genetic polymorphisms, such as SV2C(71) and VEGF-A c.*237C > T [72], were more associated with bevacizumab-induced hypertension. VEGF-634 CC polymorphism was associated with higher risks of both bevacizumab-induced hypertension and thromboembolism [73]. On the other hand, VEGF-1498 TT [46], VEGF-2578 C/C [74] and VEGF-936 C/C [74] were associated with less high-grade bevacizumab-induced hypertension. In addition, Schneider et al. examined VEGF genotypes, hypertension and overall survival in metastatic breast cancer patients treated with bevacizumab [46]. They found that the VEGF-2578AA genotype was associated with superior survival and a higher incidence of hypertension [46]. Sibertin-Blanc et al. also found VEGF-A c.*237C > T was associated with better survival and higher risk of bevacizumab-induced hypertension in metastatic colorectal cancer [72]. These findings indicated that higher risk of hypertension could be a surrogate marker for treatment outcome.

Table 2. Potential effects of genetic polymorphism on hypertension induced by vascular endothelial growth factor inhibitors.

Cancer	Genetic Polymorphism/Markers	Toxic Effect
Bevacizumab
Breast cancer	VEGF-634 CC	Controversial results for hypertension [46,73]
	VEGF-1498 TT	Less hypertension [46]
	SV2C variant	More hypertension [71]
Colorectal cancer	VEGF-A c.*237C > T	More hypertension [72]
	VEGF-2578 C/C	Less hypertension [74]
	VEGF-936 C/C	Less hypertension [74]
Various solid cancers	EGLN3	More hypertension [75]
	EGF	More hypertension [75]
	KDR	Less hypertension [75]
	WNK1	More hypertension [75]
Sunitinib
Renal cell carcinoma	VEGF SNP-634 G/G	More hypertension [77]
	VEGF ACG haplotype-2578A > C	More hypertension [78]
	VEGF ACG haplotype-460C > T	More hypertension [78]
	VEGF ACG haplotype-405C > G	More hypertension [78]

Genetic markers, such as EGLN3, EGF and WNK1, are more closely associated with bevacizumab-induced hypertension, while KDR was associated with fewer hypertensive events [75]. WNK kinases regulate sodium-chloride co-transporters in distal renal tubules [79]. Mutation of the WNK1 gene causes overexpression of co-transporters, leading to Gordon’s syndrome, a Mendelian disease with hypertension and hyperkalaemia [80]. Polymorphisms of the WNK1 gene are also associated with hypertension [81] and the treatment response to thiazide diuretics [82]. In addition, the ECF gene encodes epidermal growth factors (EGFs). It is known that the EGF receptor may increase blood pressure through influencing glomerular arterioles and sodium reabsorption in the kidneys [83]. Furthermore, both the EGLN3 and KDR genes are involved in the VEGF-A pathway. The KDR protein is a VEGF receptor. It was reported that polymorphisms of the KDR gene decrease VEGF-A’s binding ability to the KDR protein, increasing the risk of coronary artery disease [84]. EGLN3 is a member of the Egg-laying Nine (EGLN) gene family, which regulates hypoxia-inducible transcription factors (HIFs) and cellular oxygen homeostasis [85]. HIFs also increase several gene expressions related to angiogenesis, including VEGF [85]. Thus, it is reasonable that these gene markers would play certain roles in the development or prediction of bevacizumab-induced hypertension.

Polymorphisms in VEGF ACG haplotypes −2578A > C, −460C > T and −405C > G are more closely associated with sunitinib-induced hypertension [78]. Clinical genetic studies reported that ACG haplotypes are probably associated with ischemic heart disease and coronary artery disease [86–88]. Therefore, ACG haplotypes may also contribute to sunitinib-induced hypertension through influencing the VEGF signal pathway. However, the biologic roles of these ACG haplotypes remain controversial. Some reports suggested that ACG haplotypes −2578A, −460C and 405G were associated with reduced levels of serum VEGF [89,90], while other reports found no association between serum VEGF levels and haplotype −2578A [91]. Additionally, these genetic polymorphisms may also be differently distributed between different ethnic groups, leading to different risks of VEGF-associated cardiovascular toxicity. One Korean study showed that VEGF SNP-634 G/G was associated with a higher risk of sunitinib-induced hypertension [77], while it was reported that polymorphisms of VEGF-634 were associated with serum VEGF levels and its production [92,93]. Another Korean study found that the genetic polymorphism, ABCG2 421C > A, increased the risk of sunitinib-related toxicity, such as thrombocytopenia, neutropenia and hand-foot syndrome [76]. It was reported that the ABCG2 transporter, a multidrug transporter, can be inhibited by binding to sunitinib, and also decreases the extrusion of sunitinib [94–96]. Polymorphisms of the ABCG2 gene are also associated with the function and amount of ABCG2 [96,97], leading to potential differences in inhibitory effects of sunitinib. In addition, a higher frequency of 421C > A was found in Asians [98,99]. Thus, ethnic differences in genetic polymorphisms might also contribute to different presentations of VEGF toxicity. However, some of these genetic findings provide conflicting evidence. VEGF-634 CC was associated with higher or lower bevacizumab-induced hypertension in different studies [46,73]. More studies are still needed to clarify the underlying mechanisms and correlations before these clinical and genetic findings can be applied in clinical practice.

In conclusion, VEGF inhibitors disturb the balance between vasodilation and vasoconstriction, undermine endothelial cell integrity, and interact with off-target pathways, increasing potential risks of hypertension, thromboembolisms, HF and arrhythmias. Current VEGF research has not completely elucidated the underlying mechanisms of long-term impacts on and genetic or ethnic predictors of cardiovascular toxicity. Future clinical and genetic studies are needed to resolve these issues.

Disclosure statement

The authors report no conflicts of interest.