Abstract
This pharmacovigilance study uses data from the World Health Organization database of adverse drug reactions to examine the association of antiangiogenic drugs with artery dissections or aneurysms among patients receiving antiangiogenic drugs for the treatment of cancer.
Anticancer drugs targeting angiogenesis act either directly on vascular endothelial growth factor (VEGF) receptors (VEGF inhibitors) or VEGF-mediated intracellular processes (tyrosine kinase inhibitors) or indirectly through downstream VEGF signaling within broader processes (mammalian target of rapamycin inhibitors and multiprotein kinase inhibitors). Although VEGF blockade may be implicated in arterial wall injuries, the literature on the association between antiangiogenic drugs and artery dissections or aneurysms is scant, comprising only a few case reports and 1 pharmacovigilance study conducted in Japan.1 The present study evaluated the association between all antiangiogenic drugs and dissections or aneurysms occurring in all arteries.
Methods
In this pharmacovigilance study, all artery dissection or aneurysm cases from July 18, 2005, to January 19, 2019, that were potentially associated with the receipt of 14 antiangiogenic drugs were identified using VigiBase, the World Health Organization’s centralized database of adverse drug reaction cases that are spontaneously reported to national pharmacovigilance systems by patients and clinicians (18 408 774 cases as of January 3, 2019). Per the Jardé law in France regarding research involving human participants, this study did not require ethical review or informed consent because an existing anonymized database was used.
Disproportionate reporting was assessed using 2 measures, the proportional reporting ratio (PRR) and the bayesian information component (IC), with all other anticancer drugs used as the reference group.2 A signal of disproportionate reporting was defined as either a PRR with a 95% CI lower limit that exceeded 1 or an IC with a 95% credibility interval (CrI) lower limit that exceeded 0. Analyses were performed using SAS software, version 9.4 (SAS Institute). Data were analyzed from March 8 to April 10, 2019, and from September 1 to 16, 2019.
Results
Among 1 521 231 patients who experienced adverse drug reactions after receiving anticancer drugs, 217 664 cases were potentially associated with the receipt of antiangiogenic drugs. Of those, 494 patients (median age, 67 years [interquartile range, 59-73 years]; 253 women [51.2%]) had artery dissections or aneurysms (Table). Most adverse drug reactions (436 cases [88.3%]) were serious; 120 cases (24.3%) were fatal, and 88 cases (17.8%) were life threatening. Drug reactions were most frequently associated with the receipt of bevacizumab (222 cases [44.9%]), sunitinib (71 cases [14.4%]), and everolimus (55 cases [11.1%]). Overall, the presence of hypertension was reported in 51 adverse drug reaction cases (10.3%); however, hypertension was reported in only 2 of 57 cases (3.5%) involving receipt of mammalian target of rapamycin inhibitors. A total of 138 patients (27.9%) were receiving drugs in addition to antiangiogenics that may have been associated with artery dissections or aneurysms.
Table. Characteristics of Patients With Artery Dissections or Aneurysms Associated With the Receipt of Antiangiogenic Drugsa .
| Characteristic | Patients, No. (%) |
|---|---|
| Total patients, No. | 494 |
| Age, y | |
| Median (IQR) | 67 (59-73) |
| Missing data | 132 (26.7) |
| Sex | |
| Male | 213 (43.1) |
| Female | 253 (51.2) |
| Missing data | 28 (5.7) |
| Seriousness of drug reaction | |
| Death | 120 (24.3) |
| Life threatening | 88 (17.8) |
| Caused or prolonged hospitalization | 122 (24.7) |
| Other | 106 (21.5) |
| Nonserious | 8 (1.6) |
| Missing data | 50 (10.1) |
| Time to onset of drug reaction, d | |
| Median (IQR) | 89 (27-212) |
| Missing data | 270 (54.7) |
| Artery dissections or aneurysmsb | |
| Aorta | 208 (42.1) |
| Cerebral arteries | 174 (35.2) |
| Other arteries or not specified | 133 (26.9) |
| Antiangiogenic drugsc | |
| Vascular endothelial growth factor inhibitors | |
| All | 254 (51.4) |
| Bevacizumab | 222 (44.9) |
| Aflibercept | 20 (4.0) |
| Ramucirumab | 12 (2.4) |
| Tyrosine kinase inhibitors | |
| All | 177 (35.8) |
| Sunitinib | 71 (14.4) |
| Sorafenib | 31 (6.3) |
| Pazopanib | 32 (6.5) |
| Axitinib | 15 (3.0) |
| Nintedanib | 16 (3.2) |
| Lenvatinib | 11 (2.2) |
| Regorafenib | 7 (1.4) |
| Mammalian target of rapamycin inhibitors | |
| All | 57 (11.5) |
| Everolimus | 55 (11.1) |
| Temsirolimus | 2 (0.4) |
| Multiprotein kinase inhibitor | |
| Cabozantinib | 11 (2.2) |
Abbreviation: IQR, interquartile range.
All cases of artery dissections or aneurysms associated with the receipt of antiangiogenic drugs were identified through VigiBase, the World Health Organization’s centralized database of adverse drug reaction cases.
Patients could report more than 1 artery dissection or aneurysm.
Patients could report more than 1 antiangiogenic drug.
Signals of disproportionate reporting of artery dissections or aneurysms were found among all antiangiogenic drugs (PRR, 2.76 [95% CI, 2.48-3.07]; IC, 1.14 [95% CrI, 0.99-1.25]) and among 9 specific antiangiogenic drugs: bevacizumab (PRR, 4.08 [95% CI, 3.54-4.70]; IC, 1.86 [95% CrI, 1.63-2.02]), ramucirumab (PRR, 3.34 [95% CI, 1.89-5.90]; IC, 1.60 [95% CrI, 0.63-2.27]), sunitinib (PRR, 2.38 [95% CI, 1.87-3.01]; IC, 1.20 [95% CrI, 0.80-1.48]), pazopanib (PRR, 1.57 [95% CI, 1.11-2.23]; IC, 0.63 [95% CrI, 0.04-1.05]), axitinib (PRR, 2.52 [95% CI, 1.51-4.19]; IC, 1.26 [95% CrI, 0.39-1.86]), nintedanib (PRR, 2.08 [95% CI, 1.27-3.40]; IC, 1.00 [95% CrI, 0.16-1.58]), lenvatinib (PRR, 4.17 [95% CI, 2.30-7.56]; IC, 1.87 [95% CrI, 0.84-2.56]), everolimus (PRR, 1.61 [95% CI, 1.23-2.11]; IC, 0.66 [95% CrI, 0.21-0.98]), and cabozantinib (PRR, 2.25 [95% CI, 1.24-4.07]; IC, 1.09 [95% CrI, 0.06-1.78]) (Figure).
Figure. Disproportionality Analyses for Artery Dissections or Aneurysms and Aortic or Cerebral Localizations.
Binary heat map representing disproportionality measures for adverse drug reactions. IC indicates information component; MKi, multiprotein kinase inhibitor; mTORi, mammalian target of rapamycin inhibitor; PRR, proportional reporting ratio; TKi, tyrosine kinase inhibitor; and VEGFi, vascular endothelial growth factor inhibitor. A, Analysis by antiangiogenic drug class. B, Analysis by antiangiogenic drug. The white squares indicate insufficient data: the PRR was not computable when there were 2 or fewer cases, and the IC was not calculable when there was no case.
When focusing on localized aortic and cerebral dissections or aneurysms, signals of disproportionate reporting were also found for VEGF inhibitors (for aortic, PRR, 4.16 [95% CI, 3.36-5.16]; IC, 1.83 [95% CrI, 1.50-2.06] and for cerebral, PRR, 3.63 [95% CI, 2.88-4.58]; IC, 1.66 [95% CrI, 1.30-1.92]) and tyrosine kinase inhibitors (for aortic, PRR, 2.67 [95% CI, 2.13-3.35]; IC, 1.26 [95% CrI, 0.91-1.51] and for cerebral, PRR, 1.51 [95% CI, 1.13-2.03]; IC, 0.54 [95% CrI, 0.07-0.88]) and for 3 specific antiangiogenic drugs: bevacizumab (for aortic, PRR, 5.05 [95% CI, 4.05-6.29]; IC, 2.10 [95% CrI, 1.76-2.34] and for cerebral, PRR, 3.76 [95% CI, 2.92-4.84]; IC, 1.74 [95% CrI, 1.34-2.03]), sunitinib (for aortic, PRR, 3.81 [95% CI, 2.76-5.25]; IC, 1.81 [95% CrI, 1.28-2.18] and for cerebral, PRR, 1.89 [95% CI, 1.19-2.99]; IC, 0.86 [95% CrI, 0.09-1.40]), and axitinib (for aortic, PRR, 3.27 [95% CI, 1.55-6.89]; IC, 1.49 [95% CrI, 0.19-2.35] and for cerebral, PRR, 2.52 [95% CI, 1.04-6.08]; IC, 1.14 [95% CrI, −0.42 to 2.12]). For other VEGF inhibitors and tyrosine kinase inhibitors, signals of disproportionate reporting were found only for localized aortic or cerebral dissections or aneurysms.
Discussion
The results of this pharmacovigilance study provide additional information through the detection of new signals for arterial wall injuries associated with the receipt of antiangiogenic drugs (eg, everolimus, cabozantinib, and cerebral localization). Vascular endothelial growth factor blockade can produce conditions favoring artery dissections or aneurysms, notably hypertension, through the inhibition of nitric oxide synthesis and the rarefaction of tissue capillaries.3 Although the association with hypertension was not quantifiable, the presence of hypertension was reported in 10.3% of adverse drug reaction cases, with the exception of cases treated with mammalian target of rapamycin inhibitors (3.5%); this finding is consistent with the antiangiogenic action of these inhibitors, which is indirectly induced by VEGF. Antiangiogenic drugs may also produce arterial wall injuries through alteration of the extracellular matrix by (1) rarefaction of vasa vasorum that alter vessel wall blood supply and may initiate acute rupture and (2) overexpression of degradation enzymes (ie, elastin and collagen) that may weaken the matrix structure.4,5 Furthermore, because VEGF is a growth factor, its blockade may induce vascular aging.
A strength of this study is the combined use of 2 complementary disproportionality methods with 89% consistency between PRR and IC. Underreporting is an intrinsic limitation to research performed using pharmacovigilance data; however, notoriety bias is unlikely for the adverse drug reactions examined because they are generally unrecognized in the literature. In addition, the use of all anticancer drugs as the reference group should have minimized any potential indication bias, even if differences in cancer types and patient characteristics partially accounted for discrepancies found between antiangiogenic drugs from the same class (eg, bevacizumab and aflibercept). These results are unlikely to be explained by competing drugs (eg, corticosteroids or fluoroquinolones)6; only 27.9% of patients concurrently received drugs that may have been associated with artery dissections or aneurysms, and those drugs were primarily anticancer medications. Nevertheless, the potential role of other risk factors, including smoking or uncontrolled hypertension, cannot be excluded.
Population-based studies are needed to confirm and quantify the potential risk of artery dissections or aneurysms associated with the receipt of antiangiogenic drugs. However, this study’s results warrant the cautious use of antiangiogenic drugs, whatever their mechanism, among individuals at risk of artery dissections or aneurysms, as already advised by Canadian and European agencies.7
References
- 1.Oshima Y, Tanimoto T, Yuji K, Tojo A. Association between aortic dissection and systemic exposure of vascular endothelial growth factor pathway inhibitors in the Japanese Adverse Drug Event Report database. Circulation. 2017;135(8):815-817. doi: 10.1161/CIRCULATIONAHA.116.025144 [DOI] [PubMed] [Google Scholar]
- 2.van Puijenbroek EP, Bate A, Leufkens HGM, Lindquist M, Orre R, Egberts ACG. A comparison of measures of disproportionality for signal detection in spontaneous reporting systems for adverse drug reactions. Pharmacoepidemiol Drug Saf. 2002;11(1):3-10. doi: 10.1002/pds.668 [DOI] [PubMed] [Google Scholar]
- 3.Mourad JJ, des Guetz G, Debbabi H, Levy BI. Blood pressure rise following angiogenesis inhibition by bevacizumab. a crucial role for microcirculation. Ann Oncol. 2008;19(5):927-934. doi: 10.1093/annonc/mdm550 [DOI] [PubMed] [Google Scholar]
- 4.Stefanadis C, Vlachopoulos C, Karayannacos P, et al. Effect of vasa vasorum flow on structure and function of the aorta in experimental animals. Circulation. 1995;91(10):2669-2678. doi: 10.1161/01.CIR.91.10.2669 [DOI] [PubMed] [Google Scholar]
- 5.Verheul HMW, Pinedo HM. Possible molecular mechanisms involved in the toxicity of angiogenesis inhibition. Nat Rev Cancer. 2007;7(6):475-485. doi: 10.1038/nrc2152 [DOI] [PubMed] [Google Scholar]
- 6.Maumus-Robert S, Berard X, Mansiaux Y, Tubert-Bitter P, Debette S, Pariente A. Short-term risk of aortoiliac aneurysm or dissection associated with fluoroquinolone use. J Am Coll Cardiol. 2019;73(7):875-877. doi: 10.1016/j.jacc.2018.12.012 [DOI] [PubMed] [Google Scholar]
- 7.European Medicines Agency. Pharmacovigilance risk assessment committee (PRAC): 8-11 July 2019. December 19, 2019. Accessed February 16, 2021. https://www.ema.europa.eu/en/events/pharmacovigilance-risk-assessment-committee-prac-8-11-july-2019
