Association Between Aortic Dissection and Systemic Exposure of Vascular Endothelial Growth Factor Pathway Inhibitors in the Japanese Adverse Drug Event Report Database
- aortic dissection
- vascular endothelial growth factor
Angiogenesis inhibition therapy using vascular endothelial growth factor pathway inhibitors (VPIs), such as bevacizumab, sunitinib, and sorafenib, is used to treat various solid tumors. Patients treated with VPIs frequently develop hypertension. Aortic dissection involves splitting of the layers of the aortic wall, and systemic hypertension is a major predisposing risk factor. Aortic dissection is often fatal unless it is diagnosed early and aggressive treatment is initiated in a timely manner. Although VPIs have been suspected in the onset of aortic dissection in anecdotal case reports, no comparative study has assessed the link. Using disproportionality analysis1 of a large Japanese postmarketing database, we detected a signal of a potential association between aortic dissection and VPIs. To examine this signal further, we performed a comparative study within the database.
The Japanese Adverse Drug Event Report (JADER) database was obtained from the webpage of the Pharmaceuticals and Medical Devices Agency (PMDA) (https://www.pmda.go.jp/). Institutional review board approval was not required because JADER is an unlinkable anonymized database open to the public. We extracted spontaneous adverse drug reaction (ADR) reports published between April 2004 and October 2015 and associated with systemically administered drugs. As of October 2015, 8 VPIs had been approved in Japan for the treatment of various malignant tumors, including renal cell carcinoma, hepatocellular carcinoma, and soft tissue sarcomas. The 8 approved VPIs are sunitinib, sorafenib, bevacizumab, ramucirumab, lenvatinib, vandetanib, axitinib, and pazopanib. Subjects treated with at least 1 of these inhibitors were defined as the VPI group, whereas the reference group comprised subjects with cancer(s) as an underlying disease, as well as those treated with at least 1 antimalignant neoplasm agent, including tyrosine kinase inhibitors, monoclonal antibodies, and other drugs with an approved indication for any kind of cancer. The data were analyzed using R software (version 3.1.2; R Development Core Team, Vienna, Austria).
Of the 91 055 subjects with underlying malignant neoplasms, 16 441 (18.1%) were treated with a systemic VPI, whereas 59 (0.065%) had aortic dissection. Of the 59 cases, 49 [0.30% (95% confidence interval [CI], 0.22–0.39)] were associated with the use of VPIs (sunitinib, n = 14; sorafenib, n = 11; bevacizumab, n = 21; lenvatinib, n = 1; axitinib n = 1; and pazopanib n = 3; 2 were treated with both sunitinib and pazopanib). Of the 74 614 subjects with an underlying malignant neoplasm not treated with VPIs, 10 [0.01% (95% CI, 0.01–0.02)] developed aortic dissection. The crude odds ratio for aortic dissection in the group treated with VPIs relative to the reference group was 22.3 (95% CI, 11.2–49.4) (Table 1). The median time to onset was 105 days (range: 4–1363 days).
Hypertension appeared to be a confounding factor because 25 of the 49 aortic dissection cases in the VPI group had baseline hypertension, and the percentage of hypertension cases was higher in the VPI group than in the reference group. When we performed a multivariate logistic regression analysis, although hypertension had a high adjusted odds ratio of 3.8 (95% CI, 2.3–6.4), use of a VPI still had a high adjusted odds ratio of 19.4 (95% CI, 10.2–40.8).
In this study, we found a signal for an increased risk of aortic dissection after VPI treatment. Because the signal was seen with several VPIs, including sunitinib, sorafenib, and bevacizumab, this suggests that the risk is a class effect. The mechanisms could include the known potential to induce increased blood pressure and suppress the healing of damaged endothelium.
Of the 4 published cases of aortic dissection during VPI treatment, 2 patients had preexisting hypertension2,3 and 2 did not.4,5 These reports implied an association between VPIs and aortic dissection, and the current report reinforces this association based on a large database of ADR reports.
There are limitations to using the ADR database. Because the database registers only cases with an ADR, there is no information available on those who did not develop an ADR. Consequently, there was no information found on the drug-exposed population. There could also be reporting bias and noise in the data. Because of the lack of a research protocol, ADR reports are not collected systematically. In addition, data on possible confounding factors are not collected systematically, including patient background information, underlying conditions, and concomitant drugs. Therefore, our results show only a signal for an increased risk, which must be confirmed in further evaluations. Consequently, we do not recommend that clinicians alter the prescription of VPIs based on these results, although appropriate monitoring should be considered in patients with an increased risk of aortic dissection, including those with hypertension.
Our analysis found a signal for an increased risk of aortic dissection during systemic exposure to VPIs. This finding appears to be a class effect.
Sources of Funding
Dr Oshima reports personal fees from Novartis Pharma K.K. and Sanofi K.K., outside the submitted work. Dr Tojo reports grants from Bristol-Myers, Pfizer, Chugai Pharmaceutical, Daiichi Sankyo, and Sumitomo Dainippon-pharma; personal fees from Taiho Pharmaceutical, Rohto Pharmaceutical, Celgene, Novartis, Ohtsuka Pharmaceutical, and Sysmex, outside the submitted work. The other authors report no conflicts.
Circulation is available at http://circ.ahajournals.org.
- © 2017 American Heart Association, Inc.
- Edeline J,
- Laguerre B,
- Rolland Y,
- Patard JJ
- Serrano C,
- Suárez C,
- Andreu J,
- Carles J
- Formiga MN,
- Fanelli MF