Incidence, severity, mortality, and confounding factors for dissecting AAA detection in angiotensin II-infused mice: a meta-analysis.

AIMS While angiotensin II-infused mice are the most popular model for preclinical aneurysm research, representative data on incidence, severity, and mortality of dissecting abdominal aortic aneurysms (AAAs) have never been established, and the influence of confounding factors is unknown. METHODS AND RESULTS We performed a meta-analysis including 194 manuscripts representing 1679 saline-infused, 4729 non-treated angiotensin II-infused, and 4057 treated angiotensin II-infused mice. Incidence (60%) and mortality (20%) rates are reported overall as well as for grade I (22%), grade II (26%), grade III (29%), and grade IV (24%) aneurysms. Dissecting AAA incidence was significantly (P < 0.05) influenced by sex, age, genetic background, infusion time, and dose of angiotensin II. Mortality was influenced by sex, genetic background, and dose, but not by age or infusion time. Surprisingly, both incidence and mortality were significantly different (P < 0.05) when comparing angiotensin II-infused mice in descriptive studies (56% incidence and 19% mortality) with angiotensin II-infused mice that served as control animals in treatment studies designed to either enhance (35% incidence and 13% mortality) or reduce (73% incidence and 25% mortality) dissecting AAA formation. After stratification to account for confounding factors (selection bias), the observed effect was still present for incidence, but not for mortality. Possible underlying causes are detection bias (non-uniform definition for detection and quantification of dissecting AAA in mice) or publication bias (studies with negative results, related to incidence in the control group, not being published). CONCLUSIONS Our data provide a new meta-analysis-based reference for incidence and mortality of dissecting AAA in angiotensin II-infused mice, and indicate that treatment studies using this mouse model should be interpreted with caution.