Risk of Hypertension Associated with Antivascular Endothelial Growth Factor Monoclonal Antibodies: A Meta-Analysis From 51088 Patients with Cancer

Angiogenesis has an important role in promoting tumor growth, invasion, and metastasis. VEGF-A, one of the members of the vascular endothelial growth factor family, plays a vital role in angiogenesis and tissue neovascularization. Among all receptors, VEGFR-2 is widely considered as the most critical driver of tumor angiogenesis. Therefore, inhibition of angiogenesis via blocking VEGF-A or VEGFR-2 receptor signaling pathway is the key approach in current tumor therapeutics (1).

As a VEGF-A-targeted monoclonal antibody, bevacizumab widely administers in the treatment of various cancers. Ramucirumab is a VEGFR2- targeted monoclonal antibody and inhibits the signaling pathways in endothelial cells that mediate angiogenesis. Ramucirumab is approved for the treatment of advanced gastric, lung, and colorectal cancers.

Hypertension events are the most common adverse events of patients who receive the antivascular endothelial growth factor (anti-VEGF) monoclonal antibodies bevacizumab and ramucirumab treatment, which severe hypertension, particularly hypertensive emergencies may cause acute target organ injury and major cardiovascular events, that has limited the administration of anti-VEGF monoclonal antibodies (2). Previously conducted meta-analyses have investigated the risk of hypertension for bevacizumab or ramucirumab. However, the risk of hypertension caused by the anti-VEGF monoclonal antibodies is not yet evaluated systematically. Therefore, we performed the first meta-analysis to examine the risk of hypertension associated with anti-VEGF monoclonal antibodies.

The current meta-analysis aimed to examine the risk of hypertension associated with anti-VEGF monoclonal antibodies.

2.1. Search Strategy

To conduct the current meta-analysis, using the “bevacizumab”, “Avastin”, “ramucirumab”, “IMC1121B”, “LY3009806”, and “cancer” keywords, the following databases were searched: PubMed, ASCO abstracts, ESMO abstracts, and the clinical trial registration website (https://www.ClinicalTrials.gov) for relevant trials till July 2019. To ensure that all relevant clinical trials are incorporated into the meta-analysis, an independent search was conducted using the Web of Science databases.

2.2. Selection of Trials

The publications and data were reviewed and extracted by two independent investigators. Discrepancies were resolved by consensus with a third researcher. The randomized controlled trials that met the following criteria were included (1) phase II and III randomized controlled trials on cancer patients; (2) having a case group with anti-VEGF monoclonal antibodies treatment alone/concurrent chemotherapy or a control group with placebo/chemotherapy alone; and (3) Events or incidence and sample size available for hypertension.

2.3. Data Extraction

Data on study characteristics, therapeutic strategy, and results and reports of hypertension of all eligible studies were collected. The primary endpoint was the relative risk of hypertension with anti-VEGF monoclonal antibodies. Grading of hypertension events was based on versions 3.0 or 4.0 of Common Terminology Criteria.

2.4. Statistical Analysis

Analyses were performed using the RevMan 5.2. The heterogeneity of eligible studies was assessed by the I2 statistic. If P ≥ 0.1 and I2 ≤ 50%, data were analyzed by the fixed-effects model, the analysis was conducted using the random-effects model. Subgroup analyses were performed separated by the drug type, dosage, and cancer type to explore possible reasons for heterogeneity. We performed a meta-regression analysis to investigate various variables on hypertension events by Stata version 12.0.

3.1. Search Results

The initial literature review resulted in 2723 potentially relevant studies, of which 2491 were excluded because of the following reasons: reviews, commentaries, letters, basic studies, case reports, non-randomized controlled trials, irrelevant topics, and duplications. The 232 remaining studies were carefully screened and 138 studies were removed because both control and treatment groups were receiving anti-VEGF monoclonal antibodies or data required for assessment of hypertension were not available. The remaining 94 randomized controlled trials were judged as eligible for the purpose of the analysis (Figure 1).

 

 

Figure 1. Outline of the search-flow diagram

3.2. Characteristics of Studies

94 RCTs and 51088 patients were selected for this meta-analysis, which were as follows 80 trials of bevacizumab (3 - 82) and 14 trials of ramucirumab (83 - 96) were investigated (3 - 82). All malignancies, including lung cancer (19 trials), colorectal cancer (22 trials), breast cancer (19 trials), ovarian cancer (4 trials), pancreatic cancer (2 trials), renal cell cancer (4 trials), gastric or gastro-oesophageal junction adenocarcinoma (7 trials), glioblastoma (3 trials), lymphoma (2 trials), melanoma (2 trials), lymphocytic leukemia (one trial), prostate cancer (one trial), two malignant mesothelioma (one trial), leiomyosarcoma (one trial), urothelial carcinoma (2 trials), hepatocellular carcinoma (one trial), multiple myeloma (one trial), and soft tissue sarcoma (one trial). The quality of all the trials included in the meta-analysis was acceptable. The characteristics of 94 trials are listed in Table 1.

Table 1. Characteristics of Studies Included in the Meta-Analysis

3.3. RR of All-Grade Hypertension Events

To assess the relative risk of all-grade hypertension 58 RCTs were reviewed. The random-effects model (I2 = 72%) revealed that anti-VEGF monoclonal antibodies significantly increased the relative risk of all-grade (3.45, 95% CI: 2.98 - 4.00, P < 0.00001) hypertension compared to control arms (Figure 2).

 

 

3.4. RR of High-Grade Hypertension Events

The relative risk of high-grade hypertension was assessed by 91 RCTs. Using a fixed-effects model (I2 = 46%), anti-VEGF monoclonal antibodies significantly increased the relative risk of high-grade hypertension (5.63, 95% CI: 5.05 - 6.26, P < 0.00001) (I2 = 46%). (Figure 3).

 

 

3.5. Subgroup Analysis According to Drug Type

To explore the association between risk of hypertension events and drug type, a subgroup analysis was performed according to the type of the administered drug, which indicated the relative risk of all-grade hypertension events for bevacizumab (3.57: 95% CI: 2.98 - 4.28, P < 0.00001) and for ramucirumab (2.92: 95% CI: 2.42 - 3.52, P < 0.00001). No significant difference was observed between the subgroups (P = 0.571, Table 2).

 

Table 2. Relative Risk of All-Grade Hypertension Associated with Angiogenesis Inhibitors in the Subgroup Analysis

 

For high-grade hypertension events, the relative risk (RR) was 6.04 (95% CI: 5.37 - 6.80, P < 0.00001) for bevacizumab and 3.83 (95% CI: 2.96 - 4.96, P < 0.00001) for ramucirumab. There was no significant difference (P = 0.224) between bevacizumab and ramucirumab (Table 3).

 

Table 3. Relative Risk of High-Grade Hypertension Associated with Angiogenesis Inhibitors in the Subgroup Analysis

3.6. Subgroup Analysis According to Drug Dose

In the stratified analysis, which was performed according to the dose of anti-VEGF monoclonal antibodies, RR of all-grade hypertension for bevacizumab at 2.5 mg/kg/week was 3.13 (95% CI: 2.51 - 3.91), for bevacizumab at 5 mg/kg/week was 4.06 (95% CI: 3.09 - 5.34), for ramucirumab at 3.3 mg/kg/week was 2.53 (95% CI: 2.02 - 3.17), and for ramucirumab at 4 mg/kg/week was 3.21 (95% CI: 2.46 - 4.18). No significant difference was found by various drug doses (P = 0.408, Table 2).

For high-grade hypertension events, RR for bevacizumab at 2.5 mg/kg/week was 5.70 (95% CI: 4.66 - 6.97), for bevacizumab at 5mg/kg/week was 6.16 (95% CI: 5.34 - 7.11), for ramucirumab at 3.3mg/kg/week was 3.39 (95% CI: 2.26 - 5.08), and for ramucirumab at 4 mg/kg/week was 4.17 (95% CI: 2.98 - 5.83). No significant difference was observed by various drug dose (P = 0.369, Table 3).

3.7. Subgroup Analysis According to Cancer Type

In the subgroup analysis, which was performed according to the cancer type (i.e., ovarian cancer, renal cell carcinoma, and other cancer types), the risk of all-grade hypertension events was higher for these cancers. RR for ovarian cancer patients was 6.15 (95% CI: 3.34 - 11.35), for renal cell carcinoma patients was 4.02 (95% CI: 2.35 - 6.87), and for other tumor patients was 3.23 (95% CI: 2.76 - 3.76). However, RR of all-grade hypertension events did not vary significantly according to the cancer type (P = 0.074, Table 2).

For high-grade hypertension events, RR varied significantly (P = 0.003), the highest and lowest RR was for ovarian cancer patients (17.27, 95% CI: 8.50 - 35.08) and for other cancer patients (5.27, 95% CI: 4.72 - 5.88, Table 3), respectively.

3.8. Publication Bias

Funnel plots were performed to assess publication bias. No apparent publication bias was detected for all-grade and high-grade hypertension by the funnel plots.

To the best of our knowledge, this is the first meta-analysis that examined the risk of hypertension events associated with anti-VEGF monoclonal antibodies.

Analysis of the data from RCTs demonstrated that anti-VEGF monoclonal antibodies increase the risk of all-grade (RR: 3.45, 95% CI: 2.98 - 4.00) and high-grade (RR: 5.63, 95% CI: 5.05 - 6.26) hypertension compared to control arms. The mechanisms of hypertension induced by angiogenesis inhibitor included increasing cell apoptosis, decreasing endothelial renewal capacity, suppressing the production of nitric oxide in vessels, and decreasing the number of capillaries and arterioles (2).

Because severe hypertension, particularly hypertensive emergencies, may cause acute target organ injury and major cardiovascular events, which in turn leads to limited administration of the anti-VEGF monoclonal antibodies. So, clinical monitoring and effective management might be important ways for the safe application of these agents.

To explore possible risk factors, subgroup analysis was performed according to the types of administered drugs. In the current meta-analysis, there was a higher risk of high-grade hypertension in patients using bevacizumab compared with ramucirumab (RR: 6.04 VS 3.83). However, no significant difference (P = 0.224) was discovered between bevacizumab and ramucirumab.

As a VEGF-A-targeted monoclonal antibody, bevacizumab prevents the activation of VEGFR-1 and VEGFR-2, whereas ramucirumab only inhibits the VEGFR-2 receptor. VEGFR-2 is a critical receptor for angiogenesis, and blockade of the VEGF-A/VEGFR-2 signaling may result in endothelial dysfunction and hypertension (2). The precise mechanism of VEGFR-1 is not entirely understood, and the study showed that VEGFR-1 tyrosine kinase signaling also had an effect on angiogenesis (97). Blockade of the VEGF-A/VEGFR-1 signaling resulted in the endothelial dysfunction and hypertension, but simultaneously played only a minor role compared with VEGF-A/VEGFR-2 signaling (98). This may be the reason for the higher risk of high-grade hypertension in patients who were receiving bevacizumab compared to those who were receiving ramucirumab; however, the difference was not statistically significant. The difference between bevacizumab and ramucirumab was more obvious for the risk of ATE, VTE, and high-grade bleeding (99).

Based on the results, all doses of angiogenesis inhibitors increase the risk of all-grade and high-grade hypertension events, but no significant difference was found between various doses of antiangiogenic monoclonal antibodies bevacizumab (2.5 mg/kg/week and 5 mg/kg/week) and ramucirumab (3.3 mg/kg/week and 4 mg/kg/week), no matter for the risk of all-grade hypertension (P = 0.408) or the risk of high-grade hypertension (P = 0.369), suggesting that the risk of hypertension events may not be dose-dependent.

In a meta-analysis on angiogenesis inhibitors, patients with renal cell carcinoma or ovarian cancer had a high risk of hypertension. Therefore, we performed subgroup analysis according to renal cell carcinoma, ovarian cancer, and other cancer types to identify potential risk factors. The risk of high-grade hypertension varied significantly according to the cancer type, with the highest and lowest RR was for ovarian cancer and other types of cancer. The underlying mechanisms of these differences are still unclear. A possible explanation is that patients with hypertension were not excluded from ovarian trials, despite antihypertensive treatment. Also, hypertension events are relatively common in women with ovarian cancer. The high incidence rate of hypertension may be related to cancer or its ovariectomy treatment. The depletion of endogenous estrogen by ovariectomy, at least in part, induces hypertension. For example, at least 40% of the participants of the OCEANS trial had baseline hypertension, and according to the literature, pre-existing hypertension predicts the increased risk for anti-VEGF therapy-induced hypertension (63). Besides, the RR of patients with renal cell carcinoma was times higher than those with other cancers. Although nephrectomy performed among renal cell carcinoma patients can decrease glomerular filtration, the concentration of antiangiogenic monoclonal antibodies is not be influenced by a decreased GFR likely, because the metabolism and elimination of these agents primarily rely on proteolytic catabolism throughout the body, and does not depend primarily on elimination through the kidneys and livers (2, 100). Thus, the possible explanation for this phenomenon is that post nephrectomy glomerular hypertrophy may be more dependent on VEGF to keep structural completeness than a normal kidney, leading to increased sensitiveness to angiogenesis inhibitors (101). So, patients with ovarian cancer or renal cell carcinoma should pay more attention to hypertension when receiving antiangiogenic monoclonal antibodies.

4.1. Limitations

One limitation of this meta-analysis is that we conducted subgroup analysis only for ovarian cancer, renal cell carcinoma, and other cancer types, mainly because it was difficult to assess so many cancer types included in the current meta-analysis. Besides, except for drug type, drug dose, and cancer type, other potential risk factors, such as age, race, sex, and treatment duration, could increase clinical heterogeneity and, therefore, were not evaluated in the study. Finally, the literature search was limited to articles published in English, which may have led to selection bias.

In conclusion, the results showed that anti-VEGF monoclonal antibodies significantly increased the risk of hypertension. The risk may vary with cancer type, in which the highest RR was for patients with ovarian cancer (17.27, 95% CI: 8.50 - 35.08). When patients were stratified based on the type of administered drugs and dosage, no significant difference was observed. Clinicians should be aware of the adverse reaction and clinical monitoring as well as effective management of such situations, particularly for high-risk patients.

Thanks to the Gastroenterology Clinic of Taleghani Hospital in Tehran for identifying and referring patients with reflux and all the participants who participated in this study.

• Authors’ Contribution:  Bingkun Xiao was responsi-  ble for the idea and design for the meta-analysis. Weilan Wang and Le Cai contributed to data search and selection. Rongqing Huang carried out the statistical analyses.
• Conflict of Interests: The authors declare no conflicts of interest.
• Ethical Approval: This article does not contain any stud- ies with human participants or animals performed by any of the authors.
• Funding/Support: None.