Advertisement

Editor's Choice – Association Between Metformin Prescription and Abdominal Aortic Aneurysm Growth and Clinical Events: a Systematic Review and Meta-Analysis

  • Shivshankar Thanigaimani
    Affiliations
    The Queensland Research Centre for Peripheral Vascular Disease (QRC-PVD), College of Medicine and Dentistry, James Cook University, Queensland, Australia

    The Australian Institute of Tropical Health and Medicine, James Cook University, Townsville, Queensland, Australia
    Search for articles by this author
  • Tejas P. Singh
    Affiliations
    The Queensland Research Centre for Peripheral Vascular Disease (QRC-PVD), College of Medicine and Dentistry, James Cook University, Queensland, Australia

    The Department of Vascular and Endovascular Surgery, Townsville University Hospital, Townsville, Queensland, Australia
    Search for articles by this author
  • Jon Unosson
    Affiliations
    Department of Surgical Sciences, Vascular Surgery, Uppsala University, Uppsala, Sweden
    Search for articles by this author
  • James Phie
    Affiliations
    The Queensland Research Centre for Peripheral Vascular Disease (QRC-PVD), College of Medicine and Dentistry, James Cook University, Queensland, Australia

    The Australian Institute of Tropical Health and Medicine, James Cook University, Townsville, Queensland, Australia
    Search for articles by this author
  • Joseph Moxon
    Affiliations
    The Queensland Research Centre for Peripheral Vascular Disease (QRC-PVD), College of Medicine and Dentistry, James Cook University, Queensland, Australia

    The Australian Institute of Tropical Health and Medicine, James Cook University, Townsville, Queensland, Australia
    Search for articles by this author
  • Anders Wanhainen
    Affiliations
    Department of Surgical Sciences, Vascular Surgery, Uppsala University, Uppsala, Sweden
    Search for articles by this author
  • Jonathan Golledge
    Correspondence
    Corresponding author. Queensland Research Centre for Peripheral Vascular Disease, College of Medicine and Dentistry, James Cook University, Townsville, Queensland, 4811, Australia.
    Affiliations
    The Queensland Research Centre for Peripheral Vascular Disease (QRC-PVD), College of Medicine and Dentistry, James Cook University, Queensland, Australia

    The Australian Institute of Tropical Health and Medicine, James Cook University, Townsville, Queensland, Australia

    The Department of Vascular and Endovascular Surgery, Townsville University Hospital, Townsville, Queensland, Australia
    Search for articles by this author
Open ArchivePublished:September 20, 2021DOI:https://doi.org/10.1016/j.ejvs.2021.06.013

      Objective

      A meta-analysis of the association between metformin prescription and abdominal aortic aneurysm (AAA) growth and events (rupture or surgical repair) was performed.

      Methods

      Open source databases were searched for observational studies reporting the association between metformin prescription and AAA growth or events. Meta-analyses were performed using random effects models. The risk of bias of included studies was assessed using a quality assessment tool developed in a previous systematic review. Sensitivity analyses restricted to people with diabetes, leave one out analyses, and an individual patient risk factor adjusted sub-analysis were performed. Funnel plots assessed reporting bias.

      Results

      Eight studies comprising 153 553 patients were included, of whom 35 240 were and 118 313 were not prescribed metformin. Pooled weighted mean (± standard deviation) AAA growth was significantly reduced in patients prescribed metformin (0.9 ± 0.4 mm/year) compared with those not receiving the medication (1.8 ± 0.4 mm/year; weighted mean difference [WMD] 0.8 mm/year, 95% confidence interval [CI] 0.5 – 1.1; p < .001; I2 = 89%). Leave one out analysis suggested that the significance of findings did not change after removal of individual studies. A sub-analysis within people with diabetes suggested that metformin reduced AAA growth (WMD 0.7 mm/year, 95% CI 0.3 – 1.0). Metformin prescription was associated with a reduced risk of AAA events (risk ratio 0.6, 95% CI 0.4 – 0.9, p = .028). Three, four, and one studies had low, moderate, and high risk of bias, respectively. Individual patient data analysis suggested that metformin prescription slowed annual AAA growth by 0.5 mm/year (95% CI 0.2 – 0.7). The GRADE summary suggested that the certainty of evidence that metformin limited AAA growth and prevented AAA events was very low.

      Conclusion

      Observational studies suggest that metformin prescription is associated with a clinically important significant reduction in both growth and clinically relevant events in people with AAA. These findings support the need for randomised trials to examine the benefit of metformin.

      Keywords

      Most abdominal aortic aneurysms (AAA) are identified when small and at low risk of rupture but often continue to grow in size and eventually undergo surgical repair because of concerns over rupture risk. There are currently no medications that have been shown to slow the AAA growth rate. This systematic review and meta-analysis suggested that metformin prescription was associated with a significant reduction in both growth and clinically relevant events in people with AAA. The findings support the need for randomised controlled trials to examine the benefit of metformin in people with small AAAs.

      Introduction

      Abdominal aortic aneurysm (AAA) rupture is an important cause of death in older adults.
      • Sampson U.K.
      • Norman P.E.
      • Fowkes F.G.
      • Aboyans V.
      • Yanna S.
      • Harrell Jr., F.E.
      • et al.
      Global and regional burden of aortic dissection and aneurysms: mortality trends in 21 world regions, 1990 to 2010.
      Many AAAs are identified when small and at low risk of rupture, but most small AAAs continue to grow in size and eventually undergo surgical repair because of concerns over the risk of AAA rupture.
      • United Kingdom Small Aneurysm Trial P.
      • Powell J.T.
      • Brady A.R.
      • Brown L.C.
      • Fowkes F.G.
      • Greenhalgh R.M.
      • et al.
      Long-term outcomes of immediate repair compared with surveillance of small abdominal aortic aneurysms.
      Prior randomised trials have failed to identify an effective drug therapy for AAA and current guidelines recommend that small AAAs are simply managed by imaging surveillance.
      • Golledge J.
      • Moxon J.V.
      • Singh T.P.
      • Bown M.J.
      • Mani K.
      • Wanhainen A.
      Lack of an effective drug therapy for abdominal aortic aneurysm.
      • Wanhainen A.
      • Verzini F.
      • Van Herzeele I.
      • Allaire E.
      • Bown M.
      • Cohnert T.
      • et al.
      Editor's choice - european society for vascular surgery (ESVS) 2019 clinical practice guidelines on the management of abdominal aorto-iliac artery aneurysms.
      • Golledge J.
      • Singh T.P.
      Effect of blood pressure lowering drugs and antibiotics on abdominal aortic aneurysm growth: a systematic review and meta-analysis.
      There has been growing interest in the potential value of the diabetes drug metformin as a treatment for AAA, which may act through activation of 5' adenosine monophosphate-activated protein kinase (AMPK) to reduce matrix remodelling and aortic inflammation.
      • Golledge J.
      • Moxon J.V.
      • Singh T.P.
      • Bown M.J.
      • Mani K.
      • Wanhainen A.
      Lack of an effective drug therapy for abdominal aortic aneurysm.
      ,
      • Golledge J.
      Abdominal aortic aneurysm: update on pathogenesis and medical treatments.
      ,
      • He J.
      • Li N.
      • Fan Y.
      • Zhao X.
      • Liu C.
      • Hu X.
      Metformin inhibits abdominal aortic aneurysm formation through the activation of the AMPK/mTOR signaling pathway.
      A number of animal studies and human observational studies have suggested that metformin may be effective at limiting AAA progression via these mechanisms.
      • Yang L.
      • Shen L.
      • Gao P.
      • Li G.
      • He Y.
      • Wang M.
      • et al.
      Effect of AMPK signal pathway on pathogenesis of abdominal aortic aneurysms.
      • Wang Z.
      • Guo J.
      • Han X.
      • Xue M.
      • Wang W.
      • Mi L.
      • et al.
      Metformin represses the pathophysiology of AAA by suppressing the activation of PI3K/AKT/mTOR/autophagy pathway in ApoE(-/-) mice.
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      • Vasamsetti S.B.
      • Karnewar S.
      • Kanugula A.K.
      • Thatipalli A.R.
      • Kumar J.M.
      • Kotamraju S.
      Metformin inhibits monocyte-to-macrophage differentiation via AMPK-mediated inhibition of STAT3 activation: potential role in atherosclerosis.
      A previous meta-analysis examined the findings from human observational studies and reported that metformin prescription was associated with a significant reduction in AAA growth although no meta-analysis was possible for the clinically relevant events of AAA rupture and requirement for AAA repair.
      • Yu X.
      • Jiang D.
      • Wang J.
      • Wang R.
      • Chen T.
      • Wang K.
      • et al.
      Metformin prescription and aortic aneurysm: systematic review and meta-analysis.
      Since the publication of this meta-analysis, two further large cohort studies have been reported expanding the pool of evidence.
      • Sutton S.S.
      • Magagnoli J.
      • Cummings T.H.
      • Hardin J.W.
      Association between metformin and abdominal aortic aneurysm in diabetic and non-diabetic US veterans.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      The aim of this systematic review and meta-analysis was to provide an up to date synthesis of evidence regarding the association between metformin prescription and AAA growth and also to examine its association with clinically relevant events.

      Methods

      Search strategy

      This systematic review and meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.
      • Moher D.
      • Liberati A.
      • Tetzlaff J.
      • Altman D.G.
      Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement.
      The study protocol was registered with PROSPERO (Registration ID – CRD42021228160). The literature search strategy was developed and conducted by one author (ST) and peer reviewed by another (TS). The databases Medline and Scopus were searched on 6 August 2020. The search string used is shown in Appendix 1 of the Supplementary material. No date or language restrictions were applied. To be eligible for inclusion, studies needed to report the rate of AAA growth and/or clinically relevant events, defined to include AAA repair or rupture, in people who were and were not prescribed metformin and have an observational design. Included articles were identified by one author (ST) and reviewed by another (JG).

      Data extraction and outcomes

      Data were extracted independently on a customised spreadsheet by two authors (ST and TJ). The primary outcome of the study was annual AAA growth rate (mm/year) extracted as mean ± standard deviation (SD). The secondary outcome was the risk of AAA events defined as surgical AAA repair or rupture. For AAA events, reporting was not uniform, so the number of events and total number of “at risk” patients were collected to calculate an unadjusted risk ratio (RR). The following data were also extracted from the included studies: primary diagnosis of the population, study design, sample size for both control and intervention groups, age, sex, duration of follow up, study outcomes, diabetes, hypertension, ever smoked history, initial AAA diameter, and prescription of statins.

      Risk of bias

      Two authors (ST and TJ) independently assessed the quality of all included studies using a quality assessment tool developed in a previously published systematic review.
      • Singh T.P.
      • Wong S.A.
      • Moxon J.V.
      • Gasser T.C.
      • Golledge J.
      Systematic review and meta-analysis of the association between intraluminal thrombus volume and abdominal aortic aneurysm rupture.
      This assessed the reporting of study objective, study design, sample size estimation, patient characteristics, AAA imaging, method of AAA size assessment, adjustment for confounders, and definition of AAA repair or rupture.
      • Singh T.P.
      • Wong S.A.
      • Moxon J.V.
      • Gasser T.C.
      • Golledge J.
      Systematic review and meta-analysis of the association between intraluminal thrombus volume and abdominal aortic aneurysm rupture.
      Any inconsistencies were resolved through discussion between the authors until consensus was reached. The scores were specified based on the predetermined questionnaire, with responses scored as Yes = 2, Partial = 1, and No = 0. In questions that had multiple partial responses, the scores were incremental at the rate of 0.5 based on the amount of detail reported (Supplementary Table 1). The final quality assessment scores were expressed in terms of percentage (%). Percentage scores of < 50%, 50% – 70%, and > 70% were considered as high, moderate, and low risk of bias, respectively.

      Analysis of published data

      A minimum of three studies reporting the primary or secondary outcome in both the metformin and non-metformin groups were required for meta-analyses to be performed. For the main analysis, non-metformin users were pooled together regardless of their diabetes status and compared against patients with diabetes receiving metformin. A sub-analysis was performed to compare AAA growth in metformin and non-metformin users who all had diabetes. Pooled weighted means were calculated using the formula [Pooled weighted mean = (Sample size1 ∗ Mean1 + Sample sizen ∗ Meann)/ Total sample size]. Pooled weighted SD was calculated using the formula [Average of SD = √ ((N1–1) SD2 + (Nn-1) SD2) + ((N1∗Nn) / (N1+Nn)) ∗ (Mean2+Mean2 – n ∗ Mean1 ∗ Meann/ Total sample size-1], where n is the study number.
      • Higgins JPT T.J.
      • Chandler J.
      • Cumpston M.
      • Li T.
      • Page M.J.
      • Welch V.A.
      Continuous data (AAA growth per year) were expressed as weighted mean difference (WMD) and 95% confidence intervals (CI) and dichotomous data (AAA events) were expressed as RR and 95% CI. Because of the small number of eligible studies, meta-analyses were performed using the inverse variance method with random effects model and employed Sidik-Jonkman method to estimate the between study variance (tau2). This estimator uses the Q profile method to provide a conservative and broader CI to minimise the risk of false positive results. Forest plots were developed using the WMD showing the average AAA growth in mm per year. Statistical heterogeneity between studies was assessed using the I2 statistic and interpreted as low (0 – 49%), moderate (50% – 74%), or high (75% – 100%).
      • Higgins J.P.T.
      • Thompson S.G.
      • Deeks J.J.
      • Altman D.G.
      Measuring inconsistency in meta-analyses.
      Sensitivity analyses were performed to examine the heterogeneity of the included studies and the consistency of findings using the leave one out analysis approach. L’Abbé plots were used to visualise the event rate of metformin users and plotted against the event rate in non-metformin users.
      • L'Abbe K.A.
      • Detsky A.S.
      • O'Rourke K.
      Meta-analysis in clinical research.
      Meta-analyses were conducted using “meta” package and sensitivity analysis was performed using “dmetar” package of R software version 3.4.4. The certainty of evidence was assessed according to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) using the GRADEpro guideline development tool (GDT) (https://gradepro.org/).

      Individual patient data analysis

      Corresponding authors of all included studies were contacted to request de-identified individual patient data for analysis. Data from patients included in two published reports were obtained.
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      These data allowed more detailed analysis of the association between metformin and AAA growth using linear mixed effects. Demographic characteristics, risk factors, and medications were compared between AAA patients who were and were not prescribed metformin using chi square and Student t tests for categorical and continuous variables, respectively. Random intercept, random intercept and slope models and random intercept and slope interaction models with metformin prescription and follow up period as fixed effects and individual patients as random effects were developed, and the model with lowest Akaike information criterion (AIC) was selected for further analysis. The best model was adjusted for potential confounding risk factors including sex, diabetes, smoking, hypertension, antithrombotic agents (aspirin, other antiplatelet agents and anticoagulants), and statin prescription. The fitness of the model was assessed using marginal and conditional R2 values. The linearity assumptions of the model were tested using the normality of residuals (qq plots) and constant variance of residuals (homoscedasticity). Linear mixed effects were conducted using “nlme” package of R software version 3.4.4. All statistical tests were two sided and p values of ≤ .050 were considered to be statistically significant. All data were rounded up or down to their nearest single decimal point.

      Results

      Characteristics of the included studies

      A total of 270 citations were identified and ultimately eight studies including 35 240 metformin users and 118 313 non-metformin users were included (Fig. 1). Two studies were nested case control population studies,
      • Hsu C.Y.
      • Su Y.W.
      • Chen Y.T.
      • Tsai S.H.
      • Chang C.C.
      • Li S.Y.
      • et al.
      Association between use of oral-antidiabetic drugs and the risk of aortic aneurysm: a nested case-control analysis.
      ,
      • Kristensen K.L.
      • Pottegård A.
      • Hallas J.
      • Rasmussen L.M.
      • Lindholt J.S.
      Metformin treatment does not affect the risk of ruptured abdominal aortic aneurysms.
      three studies were retrospective cohort studies,
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      ,
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      ,
      • Sutton S.S.
      • Magagnoli J.
      • Cummings T.H.
      • Hardin J.W.
      Association between metformin and abdominal aortic aneurysm in diabetic and non-diabetic US veterans.
      and three studies were prospective cohort studies
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      ,
      • Golledge J.
      • Morris D.R.
      • Pinchbeck J.
      • Rowbotham S.
      • Jenkins J.
      • Bourke M.
      • et al.
      Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
      (Table 1).
      Figure 1
      Figure 1The Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) flow diagram for observational studies reporting the association between metformin prescription and abdominal aortic aneurysm (AAA) growth or events.
      Table 1Characteristics of included observational studies reporting the association between metformin prescription and abdominal aortic aneurysm (AAA) growth or events
      ReferenceStudy designCountryImaging modalityAAA diameter measurement methodDefinition of AAA eventsConfounders adjusted forAAA growth –mm/yAAA events
      Adjusted odds ratio or hazard ratio.
      p value for AAA growth/events
      Fujimura 2016
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      Retrosp cohort studyUSANRNRNRNR0.4±0.6 vs. 1.7±0.5
      Only participants with diabetes included.
      NR<.050
      Golledge 2017
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      (Cohort 1)
      Prosp cohort studyAUS, NZUS scansMax. orthogonal outer to outer anterior to posterior diameterNRSmoking, ischaemic heart disease, initial AAA diameter and sex1.0±2.7 vs. 1.6±2.5NR.030
      Golledge 2017
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      (Cohort 2)
      Prosp cohort studyAUS, NZContrast enhanced CTMax. of anterior to posterior or transverse orthogonal outer to outer diameterNRSmoking, ischaemic heart disease, initial AAA diameter, sex and prescription of beta blockers1.4±3.0 vs. 2.5±3.0NR.010
      Golledge 2017
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      (Cohort 3)
      Prosp cohort studyAUS, NZContrast enhanced CTCentrally read maximum of anterior to posterior or transverse orthogonal outer to outer diameterNRSmoking, ischaemic heart disease, initial AAA diameter and sex0.4±1.3 vs. 1.4±1.5NR.010
      Itoga 2019
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      Retrosp cohort studyUSACT and MRI scansNRNRMaximum likelihood estimates and unstructured covariance were used to account for the varying intervals between radiographic examinations and number of scans per patient. Baseline demographics including age, ethnicity, comorbidities, and medication use at time of AAA diagnosis were incorporated using an interaction term with time1.2±1.9 vs. 1.5±2.2NR<.001
      Unosson 2020
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      Prosp cohort studySWEUS scansLeading edge to leading edge anteroposterior diameter perpendicular to blood flowNRGender, age, initial AAA diameter, active smoking, kidney failure, hypertension, diuretics, beta blockers, calcium channel blockers, renin-angiotensin converting drugs, lipid lowering agents, antithrombotic drugs, insulin, sulfonylurea, and DPP4 inhibitors1.1±1.1 vs. 2.3±2.2NR<.001
      Hsu 2016
      • Hsu C.Y.
      • Su Y.W.
      • Chen Y.T.
      • Tsai S.H.
      • Chang C.C.
      • Li S.Y.
      • et al.
      Association between use of oral-antidiabetic drugs and the risk of aortic aneurysm: a nested case-control analysis.
      Nested case-control studyTWNNRNRHospital coded intact or ruptured TAA or AAAFollow up duration, age, sex, socioeconomic status, adapted Diabetes Complications Severity Index score, duration of diabetes mellitus, alpha blocker, ACE inhibitor or ARB, beta blocker, calcium channel blocker, diuretics, an antiplatelet agent, warfarin, statin, steroid, antidepressants, NSAID, and insulinNR0.7 (0.6–0.8)<.001
      Kristensen 2017
      • Kristensen K.L.
      • Pottegård A.
      • Hallas J.
      • Rasmussen L.M.
      • Lindholt J.S.
      Metformin treatment does not affect the risk of ruptured abdominal aortic aneurysms.
      Nested case control studyDNKNRNRHospital coded ruptured AAAAge, sex, calendar year, any previous diagnosis or treatment of chronic obstructive pulmonary disease, atherosclerotic disease (any of the following: coronary heart disease, cerebrovascular disease, peripheral atherosclerotic disease), renal failure and cancer. In addition, included any history of use of lipid lowering drugs, oral corticosteroids, NSAIDs, ACE inhibitors, platelet inhibitors and additional use of antidiabetic medication other than metforminNR0.8 (0.6–1.2)
      Only participants with diabetes included.
      NR
      Golledge 2019
      • Golledge J.
      • Morris D.R.
      • Pinchbeck J.
      • Rowbotham S.
      • Jenkins J.
      • Bourke M.
      • et al.
      Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
      Prosp cohort studyAUSCT scansNRProspectively collected combined incidence of AAA repair (open or endovascular) or mortality as a result of AAA ruptureAge, sex, smoking, initial AAA diameter, hypertension, ischaemic heart disease, statin, furosemide, ACE inhibitor and ARBNR0.6 (0.4–0.9).018
      Sutton 2020
      • Sutton S.S.
      • Magagnoli J.
      • Cummings T.H.
      • Hardin J.W.
      Association between metformin and abdominal aortic aneurysm in diabetic and non-diabetic US veterans.
      Retrosp cohort studyUSANRNRHospital coded open or endovascular AAA surgery or deathPropensity score matching was performed to minimise observable differences between cohortsNR0.8 (0.7–0.8)
      Only participants with diabetes included.
      NR
      Data are presented as mean ± standard deviation, or median (interquartile range). ACE = angiotensin converting enzyme; ARB = angiotensin receptor blockers; CI = confidence interval; CT = computed tomography; DPP4 = dipeptidyl peptidase-4; MRI = magnetic resonance imaging; NSAID = non-steroidal anti-inflammatory drug; NR = not reported; TAA = thoracic aortic aneurysm; USA = United States of America; AUS = Australia; NZ = New Zealand; TWN = Taiwan; DNK = Denmark; SWE = Sweden; Retrosp = retrospective; Prosp = prospective; Max. = maximum.
      Only participants with diabetes included.
      Adjusted odds ratio or hazard ratio.

      Risk of bias of included studies

      Three studies were deemed to be at low risk of bias,
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      ,
      • Golledge J.
      • Morris D.R.
      • Pinchbeck J.
      • Rowbotham S.
      • Jenkins J.
      • Bourke M.
      • et al.
      Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
      four at moderate risk of bias,
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      ,
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      ,
      • Sutton S.S.
      • Magagnoli J.
      • Cummings T.H.
      • Hardin J.W.
      Association between metformin and abdominal aortic aneurysm in diabetic and non-diabetic US veterans.
      ,
      • Kristensen K.L.
      • Pottegård A.
      • Hallas J.
      • Rasmussen L.M.
      • Lindholt J.S.
      Metformin treatment does not affect the risk of ruptured abdominal aortic aneurysms.
      and one at high risk of bias.
      • Hsu C.Y.
      • Su Y.W.
      • Chen Y.T.
      • Tsai S.H.
      • Chang C.C.
      • Li S.Y.
      • et al.
      Association between use of oral-antidiabetic drugs and the risk of aortic aneurysm: a nested case-control analysis.
      All included studies reported the study objective clearly, described the study design and participant characteristics. Seven studies adjusted the outcomes for confounding factors.
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      ,
      • Sutton S.S.
      • Magagnoli J.
      • Cummings T.H.
      • Hardin J.W.
      Association between metformin and abdominal aortic aneurysm in diabetic and non-diabetic US veterans.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      ,
      • Kristensen K.L.
      • Pottegård A.
      • Hallas J.
      • Rasmussen L.M.
      • Lindholt J.S.
      Metformin treatment does not affect the risk of ruptured abdominal aortic aneurysms.
      ,
      • Golledge J.
      • Morris D.R.
      • Pinchbeck J.
      • Rowbotham S.
      • Jenkins J.
      • Bourke M.
      • et al.
      Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
      Five studies reported the type of imaging method used,
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      ,
      • Sutton S.S.
      • Magagnoli J.
      • Cummings T.H.
      • Hardin J.W.
      Association between metformin and abdominal aortic aneurysm in diabetic and non-diabetic US veterans.
      ,
      • Golledge J.
      • Morris D.R.
      • Pinchbeck J.
      • Rowbotham S.
      • Jenkins J.
      • Bourke M.
      • et al.
      Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
      and two studies reported the method used for assessing the AAA diameter.
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      None of the studies reported a sample size estimation. Overall, assessment of all included studies indicated a moderate mean quality score of 64.5 ± 14.2%. Detailed scoring of each study against specific criteria is shown in Supplementary Table 3.

      Reported association between metformin prescription and abdominal aortic aneurysm growth

      Four studies that reported a total of six different cohorts reported the AAA growth rates
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      in 5 745 participants prescribed metformin and 10 370 not receiving metformin.
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      One study reported AAA growth rates in three different cohorts in which different imaging protocols were used
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      (Table 1). The risk factors of participants who were and were not prescribed metformin are shown in Table 2. Most risk factors were reported to be not statistically significantly different in people who were and were not prescribed metformin; however, diabetes and frequency of statin prescription were greater in people prescribed compared with those not prescribed metformin (Table 2). Where reported, or available from the corresponding authors, diabetes was diagnosed in 100% of participants who were prescribed metformin compared with a weighted mean of 59.4% ± 7.2% (n = 8 550) of those not prescribed metformin (Table 2). Mean annual AAA growth was significantly less in participants prescribed metformin than in those not receiving the medication in all six cohorts. Metformin was reported to be associated with statistically significantly reduced AAA growth after adjusting for confounders in five of the six cohorts that reported such an analysis (Table 1).
      Table 2Participant characteristics in all studies reporting abdominal aortic aneurysm (AAA) growth segregated according to the presence of diabetes and exposure to metformin
      ReferenceStudy designCohort groupsFollow up period – ynAgeMales – %Initial AAA diameter – mmEver smokedHTNCHDStatins
      Fujimura 2016Retrosp cohort study4.03±6.5
      Mean derived from the median data provided.
      NR82.740.4±1.249
      Non-active smokers, active smokers, and unknown smoking status combined.
      (83.0)
      49 (84.5)36 (62.1)37 (63.8)
      DM and metformin15
      DM and no metformin43
      No DM and no metformin0
      Golledge 2017 (Cohort 1)Prosp cohort study3.6±2.4
      DM and metformin11872.7±6.1109 (92.3)37.8±7.1100 (84.7)100 (84.7)57
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (48.3)
      97 (82.2)
      DM and no metformin9973.6±6.691 (91.9)37.7±6.287 (87.9)79 (79.8)54
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (54.5)
      64 (64.6)
      No DM and no metformin1 14074.1±6.31 021 (89.5)36.7±6.2716 (62.8)700 (61.4)494
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (43.3)
      497 (43.6)
      Golledge 2017 (Cohort 2)Prosp cohort study2.9±2.6
      DM and metformin3972.1±5.637 (94.9)41.5±7.539 (100.0)34 (87.2)21
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (53.8)
      36 (92.3)
      DM and no metformin3072.4±7.228 (93.3)42.2±7.928 (93.3)22 (73.3)18
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (60.0)
      23 (76.6)
      No DM and no metformin21872.8±8.1170 (77.9)40.6±7.2191 (87.6)141 (64.7)104
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (47.7)
      137 (62.8)
      Golledge 2017 (Cohort 3)Prosp cohort study1
      DM and metformin1672.3±6.015 (93.7)40.0±4.816 (100.0)12 (75)9
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (56.2)
      14 (87.5)
      DM and no metformin369.9±5.62 (66.6)37.0±2.73 (100.0)001 (33.3)
      No DM and no metformin3473.7±8.528 (82.3)41.0±4.230 (88.2)11 (32.3)13
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (38.2)
      20 (58.8)
      Itoga 2019Retrosp cohort study4.2±2.669.8±7.899.438.0±7.113 834
      Non-active smokers, active smokers, and unknown smoking status combined.
      (100.0)
      NR2 034 (14.7)9 463 (68.4)
      DM and metformin5 492
      DM and no metformin8 342
      No DM and no metformin0
      Unosson 2020Prosp cohort study3.2±1.7
      DM and metformin6568.5±5.460 (92.3)36.5±5.960 (92.3)55 (87.3)27 (43.5)57
      Lipid modifying agent.
      (87.7)
      DM and no metformin3370.1±6.930 (90.9)37.5±6.030 (90.0)30 (90.9)15 (45.5)29
      Lipid modifying agent.
      (87.9)
      No DM and no metformin42869.1±5.4426 (99.5)38.2±6.1356 (83.2)282 (66.7)139 (33.0)275
      Lipid modifying agent.
      (65.0)
      Data are presented as n (%) or mean±standard deviation unless stated otherwise. AAA = abdominal aortic aneurysm; CHD = coronary heart disease; HTN = hypertension; NR = not reported; DM = diabetes mellitus; Retrosp = retrospective; Prosp = prospective; Max. = maximum.
      Mean derived from the median data provided.
      Non-active smokers, active smokers, and unknown smoking status combined.
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      § Lipid modifying agent.

      Meta-analysis examining the association between metformin and abdominal aortic aneurysm growth

      A meta-analysis including 5 745 people prescribed and 10 370 people not prescribed metformin suggested that metformin was associated with a significant reduction in AAA growth, although there was a high degree of heterogeneity between studies (WMD 0.8 mm/year; 95% CI 0.5 – 1.1; I2 = 89%) (Fig. 2). Pooled mean AAA growth was statistically significantly slower in patients receiving metformin (1.2 ± 1.9 mm/year) compared with those who were not prescribed this drug (1.6 ± 2.3 mm/year; p = .005). Leave one out analyses suggested that the significance of the association was not influenced by any single study (Supplementary Fig. 1). The study reported by Fujimura et al. contributed to the largest effect size,
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      exclusion of which resulted in a WMD of 0.6 mm/year (95% CI 0.3 – 0.9, I2 = 76%).
      Figure 2
      Figure 2Forest plot for the weighted mean difference (MD) in annual abdominal aortic aneurysm (AAA) growth in a pooled patient population who were and were not prescribed metformin. SD = standard deviation; CI = confidence interval; a = Cohort 1; b = Cohort 2; c = Cohort 3.

      Reported association between metformin prescription and abdominal aortic aneurysm events

      Four studies examined AAA events in 29 495 participants prescribed metformin and 107 943 not receiving metformin.
      • Sutton S.S.
      • Magagnoli J.
      • Cummings T.H.
      • Hardin J.W.
      Association between metformin and abdominal aortic aneurysm in diabetic and non-diabetic US veterans.
      ,
      • Hsu C.Y.
      • Su Y.W.
      • Chen Y.T.
      • Tsai S.H.
      • Chang C.C.
      • Li S.Y.
      • et al.
      Association between use of oral-antidiabetic drugs and the risk of aortic aneurysm: a nested case-control analysis.
      • Kristensen K.L.
      • Pottegård A.
      • Hallas J.
      • Rasmussen L.M.
      • Lindholt J.S.
      Metformin treatment does not affect the risk of ruptured abdominal aortic aneurysms.
      • Golledge J.
      • Morris D.R.
      • Pinchbeck J.
      • Rowbotham S.
      • Jenkins J.
      • Bourke M.
      • et al.
      Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
      The definition of AAA events varied, with three studies using hospital coded data mixing intact and ruptured aortic aneurysm diagnoses and only one study prospectively following patients with small AAAs to record requirement for AAA surgery and AAA rupture related mortality (Table 1). The risk factors of participants who were and were not prescribed metformin are shown in Table 3. Most risk factors were similar in both groups; however, diabetes and statin prescription were more common in people prescribed, compared with those not prescribed metformin (Table 3). Risk of AAA events was statistically significantly lower in participants prescribed metformin compared with those not receiving this medication in all four cohorts. Analyses adjusted for potential confounding variables reported in all four studies found an independent significant association between metformin and a reduced risk of AAA events (Table 1).
      Table 3Participant characteristics in all studies reporting abdominal aortic aneurysm (AAA) events segregated according to the presence of diabetes and exposure to metformin
      ReferenceStudy designCohort groupsFollow up period – ynAgeMalesInitial AAA diameter – mmCurrently smokingHTNCHDStatins
      Hsu 2016Nested case control studyNR67.5±47.366.5NRNRNRNRNR
      DM and metformin3 599
      DM and no metformin5 337
      No DM and no metformin0
      Kristensen 2017Nested case control studyNRNRNRNRNR
      DM and metformin1 40672.9±7.5
      Mean derived from the median data provided.
      1 191738 (70.7)
      DM and no metformin2 57674.3±8.1
      Mean derived from the median data provided.
      2 1421 077 (47.8)
      No DM and no metformin0NR
      Golledge 2019Prospective cohort studyCases: 3.2±3.3;DM control: 2.2±2.9;No DM control: 2.5±3.0
      DM and metformin12972.4±6.5116 (89.9)43.3±10.335 (27.1)115 (89.1)64
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (49.6)
      116 (89.9)
      DM and no metformin10574.2±7.293 (88.6)45.3±10.016 (15.2)80 (76.2)57
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (54.3)
      74 (70.5)
      No DM and no metformin84673.6±8.0672 (79.4)46.6±11.5226 (26.7)626 (74.0)412
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      (48.7)
      531 (62.8)
      Sutton 2020Retrospective cohort study18NRNRNR
      DM and metformin24 36169.6±7.124 361 (100.0)6 559
      Smoking data, not reported whether current or ever smoked.
      (26.9)
      20 030 (82.2)
      DM and no metformin43 07372.3±7.843 073 (100.0)9 820
      Smoking data, not reported whether current or ever smoked.
      (22.8)
      33 885 (78.7)
      No DM and no metformin56 00670.3±7.956 006 (100.0)14 979
      Smoking data, not reported whether current or ever smoked.
      (26.7)
      44 745 (79.9)
      Data are presented as n (%) or mean±standard deviation. AAA = abdominal aortic aneurysm; CHD = coronary heart disease; HTN = hypertension; NR = not reported; DM = diabetes mellitus.
      Mean derived from the median data provided.
      CHD data were taken from ischaemic heart disease, congestive heart failure, or coronary artery disease.
      Smoking data, not reported whether current or ever smoked.

      Meta-analysis examining the association between metformin and abdominal aortic aneurysm events

      A meta-analysis of three eligible studies including 5 722 people prescribed and 7 294 people not prescribed metformin suggested that metformin prescription was associated with a significant reduction in the risk of AAA events (RR 0.60, 95% CI 0.50 – 0.71, p = .028), with a moderate degree of statistical heterogeneity (Fig. 3).
      • Hsu C.Y.
      • Su Y.W.
      • Chen Y.T.
      • Tsai S.H.
      • Chang C.C.
      • Li S.Y.
      • et al.
      Association between use of oral-antidiabetic drugs and the risk of aortic aneurysm: a nested case-control analysis.
      • Kristensen K.L.
      • Pottegård A.
      • Hallas J.
      • Rasmussen L.M.
      • Lindholt J.S.
      Metformin treatment does not affect the risk of ruptured abdominal aortic aneurysms.
      • Golledge J.
      • Morris D.R.
      • Pinchbeck J.
      • Rowbotham S.
      • Jenkins J.
      • Bourke M.
      • et al.
      Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
      Leave one out analyses showed that the findings were dependent on the inclusion of two studies, as removal of the studies authored by Hsu et al. (RR 0.65, 95% CI 0.12 – 3.62, I2 = 44%) or Golledge et al. (RR 0.62, 95% CI 0.10 – 3.68, I2 = 80%) led to metformin prescription no longer being statistically significantly associated with a reduced rate of AAA events
      • Hsu C.Y.
      • Su Y.W.
      • Chen Y.T.
      • Tsai S.H.
      • Chang C.C.
      • Li S.Y.
      • et al.
      Association between use of oral-antidiabetic drugs and the risk of aortic aneurysm: a nested case-control analysis.
      ,
      • Golledge J.
      • Morris D.R.
      • Pinchbeck J.
      • Rowbotham S.
      • Jenkins J.
      • Bourke M.
      • et al.
      Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
      (Supplementary Fig. 2). Visualisation of the L’Abbe plot supported a true reduction in AAA event risk among metformin users (Supplementary Fig. 3).
      Figure 3
      Figure 3Forest plot for the risk ratio (RR) of abdominal aortic aneurysm (AAA) events in patients who were and were not prescribed metformin. The plot suggests a reduced risk of AAA events associated with metformin use. CI = confidence interval.

      Sub-analysis restricted to participants with diabetes

      A sub-analysis limited to participants with diabetes from four eligible studies,
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      including 5 745 metformin users and 8 550 non-metformin users, suggested that metformin prescription was associated with a statistically significant reduction in AAA growth with a high level of heterogeneity between studies (WMD 0.7 mm/year; 95% CI 0.3 – 1.0; I2 = 85%) (Fig. 4). Among participants with diabetes, mean annual AAA growth was significantly slower in those receiving metformin (0.9 ± 0.4 mm/year) compared with those not (1.6 ± 0.4 mm/year; p < .001). Leave one out sensitivity analysis suggested that the significance of the association of metformin prescription with reduced AAA growth was not influenced by any single study (Supplementary Fig. 4). The highest effect size was contributed by the study of Fujimura et al.,
      • Fujimura N.
      • Xiong J.
      • Kettler E.B.
      • Xuan H.
      • Glover K.J.
      • Mell M.W.
      • et al.
      Metformin treatment status and abdominal aortic aneurysm disease progression.
      exclusion of which resulted in a WMD of 0.3 mm/year (95% CI 0.2 – 0.5). Between studies heterogeneity of the effect size was influenced by the study of Itoga et al.,
      • Itoga N.K.
      • Rothenberg K.A.
      • Suarez P.
      • Ho T.V.
      • Mell M.W.
      • Xu B.
      • et al.
      Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
      exclusion of which resulted in a WMD of 0.9 mm/year (95% CI 0.5 – 1.2; Supplementary Fig. 4) and reduced the heterogeneity (I2 = 49%). A sub-analysis limited to participants with diabetes was not possible for AAA events because of the small number of studies.
      Figure 4
      Figure 4Forest plot for the weighted mean difference (MD) in annual abdominal aortic aneurysm (AAA) growth in patients with diabetes who were and were not prescribed metformin. SD = standard deviation; CI = confidence interval; a = Cohort 1; b = Cohort 2; c = Cohort 3.

      Certainty of evidence

      The GRADE summary for the association between metformin and AAA growth and events is provided in the supplementary data and suggested a very low certainty of evidence of effect on both outcomes (Supplementary Table 2).

      Individual patient data analysis

      Individual patient data were provided from only two included studies,
      • Golledge J.
      • Moxon J.
      • Pinchbeck J.
      • Anderson G.
      • Rowbotham S.
      • Jenkins J.
      • et al.
      Association between metformin prescription and growth rates of abdominal aortic aneurysms.
      ,
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      which comprised a total of four cohorts with 238 patients prescribed metformin and 1 985 patients not prescribed metformin. Patients receiving metformin were statistically significantly older and more likely to have a history of smoking, hypertension, and statin and antithrombotic agent prescription compared with patients with no diabetes who were not prescribed metformin. In analyses restricted to patients with diabetes, only statin prescription was significantly more common in metformin users than non-users (Supplementary Table 4). In the random intercept unadjusted models, metformin prescription was associated with a statistically significantly slowed mean AAA growth by 0.5 mm/year (95% CI 0.2 – 0.7) over a mean follow up period of 3.5 ± 2.3 years compared with patients not prescribed metformin (Supplementary Table 5). Adjusting the model for potential confounding variables (sex, diabetes, hypertension, statins, smoking, and antithrombotic drugs) suggested that metformin prescription was associated with a significant mean slower overall AAA growth of 0.5 mm/year (95% CI 0.2 – 0.7) (Supplementary Table 6). The fixed and random effects variables included within the model accounted for 90% of the outcome variations (Supplementary Table 6). The model predicted that the initial mean AAA diameter of metformin and non-metformin patients were similar (38.1 mm, 95% CI 34.1 – 42.2 vs. 37.2 mm, 95% CI 33.2 – 41.2). At the end of follow up, patients prescribed metformin were predicted to have a mean AAA diameter of 41.8 mm (95% CI 37.7 – 45.8) compared with 42.4 mm (95% CI 38.3 – 46.4) in patients not receiving metformin (p < .001) (Supplementary Table 7). “Qqplot” and homoscedasticity plots confirmed the linearity assumptions of the model (Supplementary Figs. 5 and 6).

      Discussion

      This meta-analysis updates and extends a prior meta-analysis which suggested that metformin prescription was associated with reduced AAA growth. In the current analysis, which included 123 966 additional participants, metformin prescription was consistently associated with significantly reduced AAA growth in the main analysis and also in multiple sensitivity analyses, including those restricted to people with diabetes and a sub-analysis of individual patient data. The latter analyses are important as all metformin users had diabetes and thus accounting for this risk factor is key. A recent literature review suggested that although patients with diabetes in general exhibited slower AAA growth, evidence from clinical and experimental studies suggested that antidiabetic medications may have direct inhibitory effects on AAA pathology.
      • Patel K.
      • Zafar M.A.
      • Ziganshin B.A.
      • Elefteriades J.A.
      Diabetes mellitus: is it protective against aneurysm? a narrative review.
      This study found that metformin prescription was associated with a 40% relative reduction in the risk of clinically important events, although this finding was not robust in sensitivity analyses. Overall, these findings provide strong support for the clinical trials that have already started, or are about to commence, testing the efficacy of metformin to reduce AAA growth and clinically important events.
      • Dalman R.L.
      • Wanhainen A.
      • Mani K.
      • Modarai B.
      Top 10 candidate aortic disease trials.
      ,
      • Wanhainen A.
      • Unosson J.
      • Mani K.
      • Gottsäter A.
      The Metformin for abdominal aortic aneurysm growth inhibition (MAAAGI) trial.
      In animal models, metformin has been reported to activate AMPK to reduce aortic inflammation, inhibit extracellular matrix remodelling and modify vascular smooth muscle cell phenotype to limit aortic expansion.
      • McCarthy A.D.
      • Cortizo A.M.
      • Sedlinsky C.
      Metformin revisited: Does this regulator of AMP-activated protein kinase secondarily affect bone metabolism and prevent diabetic osteopathy.
      • Cameron A.R.
      • Morrison V.L.
      • Levin D.
      • Mohan M.
      • Forteath C.
      • Beall C.
      • et al.
      Anti-inflammatory effects of metformin irrespective of diabetes status.
      • Aljofan M.
      • Riethmacher D.
      Anticancer activity of metformin: a systematic review of the literature.
      An anti-inflammatory effect of metformin has also been observed in human plasma samples from AAA patients.
      • Unosson J.
      • Wågsäter D.
      • Bjarnegård N.
      • De Basso R.
      • Welander M.
      • Mani K.
      • et al.
      Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
      A recent animal study where AAA was induced by intraluminal porcine pancreatic elastase (PPE) perfusion in male Sprague Dawley rats also suggested that metformin inhibited AAA formation through activation of the AMPK/mTOR signalling pathway.
      • He J.
      • Li N.
      • Fan Y.
      • Zhao X.
      • Liu C.
      • Hu X.
      Metformin inhibits abdominal aortic aneurysm formation through the activation of the AMPK/mTOR signaling pathway.
      The current meta-analysis has a number of strengths and weaknesses that should be acknowledged. This is the most detailed meta-analysis examining the association between metformin and AAA growth, including quality assessment of studies, numerous sensitivity analyses and incorporating an analysis of the effect on clinically important events. Weaknesses include the relatively small number of studies, particularly in relation to AAA events. Also the definition of AAA events and included populations varied. Furthermore, there are currently no published randomised clinical trials testing metformin in people with AAA. All the studies included were observational and thus subjective to associated biases. An individual patient analysis suggested that the findings persisted after adjusting for confounding factors. In the main analysis, diabetes was not matched between participants prescribed metformin and those not receiving the medication. Even in the analysis limited to people with diabetes, it is possible that the severity of diabetes may not have been matched between metformin users and non-users. All patients within the included studies were prescribed multiple other medications apart from metformin. Doses of and adherence to metformin were not reported. Duration of diabetes also varied and was poorly reported (Supplementary Table 8). Thus, it is impossible to rule out that the effect attributed to metformin may have been driven by some other feature of diabetes, such as other antidiabetic medications, or extracellular matrix cross linking effects of hyperglycaemia. It is possible that the associations between metformin and slower AAA growth and reduced AAA events could be explained by confounders. Observational studies have reported that intensive medical management, such as statin prescription, is associated with reduced AAA growth and this might have contributed to the association between metformin and reduced AAA progression found in the current study.
      • Mathisen S.R.
      • Abdelnoor M.
      Beneficial effect of statins on total mortality in abdominal aortic aneurysm (AAA) repair.
      • Takagi H.
      • Matsui M.
      • Umemoto T.
      A meta-analysis of clinical studies of statins for prevention of abdominal aortic aneurysm expansion.
      • Salata K.
      • Syed M.
      • Hussain M.A.
      • de Mestral C.
      • Greco E.
      • Mamdani M.
      Statins reduce abdominal aortic aneurysm growth, rupture, and perioperative mortality: a systemic review and meta-analysis.
      • Karrowni W.
      • Dughman S.
      • Hajj G.P.
      • Miller Jr., F.J.
      Statin therapy reduces growth of abdominal aortic aneurysms.
      Randomised controlled trials are therefore needed to test the value of metformin in an unbiased way.
      In conclusion, this up to date meta-analysis suggests that metformin is associated with reduced AAA growth and an important reduction in the risk of clinically important events. This finding may be subject to confounding and thus randomised trials are needed to properly test the value of metformin for people with AAA.

      Conflict of interest

      None.

      Funding

      Funding from the National, Health and Medical Research Council ( 1180736 ), Royal Australasian College of Surgeons , Australian and New Zealand Society for Vascular Surgery , and the Queensland Government supported this work. JG holds a Practitioner Fellowship from the National Health and Medical Research Council (1117061) and a Senior Clinical Research Fellowship from the Queensland Government, Australia. The funders played no role in study design, conduct, data collection, analysis, and interpretation, and did not assist in preparation or review of this manuscript.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Sampson U.K.
        • Norman P.E.
        • Fowkes F.G.
        • Aboyans V.
        • Yanna S.
        • Harrell Jr., F.E.
        • et al.
        Global and regional burden of aortic dissection and aneurysms: mortality trends in 21 world regions, 1990 to 2010.
        Glob Heart. 2014; 9: 171-180
        • United Kingdom Small Aneurysm Trial P.
        • Powell J.T.
        • Brady A.R.
        • Brown L.C.
        • Fowkes F.G.
        • Greenhalgh R.M.
        • et al.
        Long-term outcomes of immediate repair compared with surveillance of small abdominal aortic aneurysms.
        N Engl J Med. 2002; 346: 1445-1452
        • Golledge J.
        • Moxon J.V.
        • Singh T.P.
        • Bown M.J.
        • Mani K.
        • Wanhainen A.
        Lack of an effective drug therapy for abdominal aortic aneurysm.
        J Int Med. 2020; 288: 6-22
        • Wanhainen A.
        • Verzini F.
        • Van Herzeele I.
        • Allaire E.
        • Bown M.
        • Cohnert T.
        • et al.
        Editor's choice - european society for vascular surgery (ESVS) 2019 clinical practice guidelines on the management of abdominal aorto-iliac artery aneurysms.
        Eur J Vasc Endovasc Surg. 2019; 57: 8-93
        • Golledge J.
        • Singh T.P.
        Effect of blood pressure lowering drugs and antibiotics on abdominal aortic aneurysm growth: a systematic review and meta-analysis.
        Heart, 2020 (heartjnl-2020-318192)
        • Golledge J.
        Abdominal aortic aneurysm: update on pathogenesis and medical treatments.
        Nat Rev Cardiol. 2019; 16: 225-242
        • He J.
        • Li N.
        • Fan Y.
        • Zhao X.
        • Liu C.
        • Hu X.
        Metformin inhibits abdominal aortic aneurysm formation through the activation of the AMPK/mTOR signaling pathway.
        J Vasc Res. 2021; : 1-11
        • Yang L.
        • Shen L.
        • Gao P.
        • Li G.
        • He Y.
        • Wang M.
        • et al.
        Effect of AMPK signal pathway on pathogenesis of abdominal aortic aneurysms.
        Oncotarget. 2017; 8: 92827-92840
        • Wang Z.
        • Guo J.
        • Han X.
        • Xue M.
        • Wang W.
        • Mi L.
        • et al.
        Metformin represses the pathophysiology of AAA by suppressing the activation of PI3K/AKT/mTOR/autophagy pathway in ApoE(-/-) mice.
        Cell Biosci. 2019; 9: 68
        • Golledge J.
        • Moxon J.
        • Pinchbeck J.
        • Anderson G.
        • Rowbotham S.
        • Jenkins J.
        • et al.
        Association between metformin prescription and growth rates of abdominal aortic aneurysms.
        Br J Surg. 2017; 104: 1486-1493
        • Itoga N.K.
        • Rothenberg K.A.
        • Suarez P.
        • Ho T.V.
        • Mell M.W.
        • Xu B.
        • et al.
        Metformin prescription status and abdominal aortic aneurysm disease progression in the U.S. veteran population.
        J Vasc Surg. 2019; 69: 710-716
        • Fujimura N.
        • Xiong J.
        • Kettler E.B.
        • Xuan H.
        • Glover K.J.
        • Mell M.W.
        • et al.
        Metformin treatment status and abdominal aortic aneurysm disease progression.
        J Vasc Surg. 2016; 64: 46-54
        • Vasamsetti S.B.
        • Karnewar S.
        • Kanugula A.K.
        • Thatipalli A.R.
        • Kumar J.M.
        • Kotamraju S.
        Metformin inhibits monocyte-to-macrophage differentiation via AMPK-mediated inhibition of STAT3 activation: potential role in atherosclerosis.
        Diabetes. 2015; 64: 2028-2041
        • Yu X.
        • Jiang D.
        • Wang J.
        • Wang R.
        • Chen T.
        • Wang K.
        • et al.
        Metformin prescription and aortic aneurysm: systematic review and meta-analysis.
        Heart. 2019; 105: 1351-1357
        • Sutton S.S.
        • Magagnoli J.
        • Cummings T.H.
        • Hardin J.W.
        Association between metformin and abdominal aortic aneurysm in diabetic and non-diabetic US veterans.
        J Investig Med. 2020; 68: 1015-1018
        • Unosson J.
        • Wågsäter D.
        • Bjarnegård N.
        • De Basso R.
        • Welander M.
        • Mani K.
        • et al.
        Metformin prescription associated with reduced abdominal aortic aneurysm growth rate and reduced chemokine expression in a swedish cohort.
        Ann Vasc Surg. 2021; 70: 425-433
        • Moher D.
        • Liberati A.
        • Tetzlaff J.
        • Altman D.G.
        Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement.
        Bmj. 2009; 339: b2535
        • Singh T.P.
        • Wong S.A.
        • Moxon J.V.
        • Gasser T.C.
        • Golledge J.
        Systematic review and meta-analysis of the association between intraluminal thrombus volume and abdominal aortic aneurysm rupture.
        J Vascr Surg. 2019; 70: 2065-2073
        • Higgins JPT T.J.
        • Chandler J.
        • Cumpston M.
        • Li T.
        • Page M.J.
        • Welch V.A.
        Sons J.W. Cochrane Handbook for Systematic Reviews of Interventions. 2019 (Chichester (UK))
        • Higgins J.P.T.
        • Thompson S.G.
        • Deeks J.J.
        • Altman D.G.
        Measuring inconsistency in meta-analyses.
        BMJ (Clinical research ed.). 2003; 327: 557-560
        • L'Abbe K.A.
        • Detsky A.S.
        • O'Rourke K.
        Meta-analysis in clinical research.
        Ann Intern Med. 1987; 107: 224-233
        • Hsu C.Y.
        • Su Y.W.
        • Chen Y.T.
        • Tsai S.H.
        • Chang C.C.
        • Li S.Y.
        • et al.
        Association between use of oral-antidiabetic drugs and the risk of aortic aneurysm: a nested case-control analysis.
        Cardiovasc Diabetol. 2016; 15: 125
        • Kristensen K.L.
        • Pottegård A.
        • Hallas J.
        • Rasmussen L.M.
        • Lindholt J.S.
        Metformin treatment does not affect the risk of ruptured abdominal aortic aneurysms.
        J Vasc Surg. 2017; 66: 768-774
        • Golledge J.
        • Morris D.R.
        • Pinchbeck J.
        • Rowbotham S.
        • Jenkins J.
        • Bourke M.
        • et al.
        Editor's choice – metformin prescription is associated with a reduction in the combined incidence of surgical repair and rupture related mortality in patients with abdominal aortic aneurysm.
        Eur J Vasc Endovasc Surg. 2019; 57: 94-101
        • Patel K.
        • Zafar M.A.
        • Ziganshin B.A.
        • Elefteriades J.A.
        Diabetes mellitus: is it protective against aneurysm? a narrative review.
        Cardiology. 2018; 141: 107-122
        • Dalman R.L.
        • Wanhainen A.
        • Mani K.
        • Modarai B.
        Top 10 candidate aortic disease trials.
        J Intern Med. 2020; 288: 23-37
        • Wanhainen A.
        • Unosson J.
        • Mani K.
        • Gottsäter A.
        The Metformin for abdominal aortic aneurysm growth inhibition (MAAAGI) trial.
        Eur J Vasc Endovasc Surg. 2021; 61: 710-711
        • McCarthy A.D.
        • Cortizo A.M.
        • Sedlinsky C.
        Metformin revisited: Does this regulator of AMP-activated protein kinase secondarily affect bone metabolism and prevent diabetic osteopathy.
        World J Diab. 2016; 7: 122-133
        • Cameron A.R.
        • Morrison V.L.
        • Levin D.
        • Mohan M.
        • Forteath C.
        • Beall C.
        • et al.
        Anti-inflammatory effects of metformin irrespective of diabetes status.
        Circulat Res. 2016; 119: 652-665
        • Aljofan M.
        • Riethmacher D.
        Anticancer activity of metformin: a systematic review of the literature.
        Fut Sci OA. 2019; 5: FSO410
        • Mathisen S.R.
        • Abdelnoor M.
        Beneficial effect of statins on total mortality in abdominal aortic aneurysm (AAA) repair.
        Vasc Med. 2017; 22: 406-410
        • Takagi H.
        • Matsui M.
        • Umemoto T.
        A meta-analysis of clinical studies of statins for prevention of abdominal aortic aneurysm expansion.
        J Vasc Surg. 2010; 52: 1675-1681
        • Salata K.
        • Syed M.
        • Hussain M.A.
        • de Mestral C.
        • Greco E.
        • Mamdani M.
        Statins reduce abdominal aortic aneurysm growth, rupture, and perioperative mortality: a systemic review and meta-analysis.
        J Vasc Surg. 2019; 69: 967
        • Karrowni W.
        • Dughman S.
        • Hajj G.P.
        • Miller Jr., F.J.
        Statin therapy reduces growth of abdominal aortic aneurysms.
        J Investigat Med: Off Public Am Federat Clinic Res. 2011; 59: 1239-1243

      Linked Article

      • Is the Era of Drug Therapy for Abdominal Aortic Aneurysm Drawing Near?
        European Journal of Vascular and Endovascular SurgeryVol. 62Issue 5
        • Preview
          Although abdominal aortic aneurysm (AAA) is a life threatening condition, there are limited clinical data on the potential effects of pharmacotherapy. With advances in preventive medicine and the implementation of screening programmes, AAAs are often discovered incidentally at an asymptomatic stage, allowing a therapeutic window with the potential of halting the disease’s progress. AAA has been associated with several atherosclerotic risk factors except for diabetes mellitus. Such a paradox has been proposed to be related to metformin prescription.
        • Full-Text
        • PDF
        Open Archive

      Comments

      Commenting Guidelines

      To submit a comment for a journal article, please use the space above and note the following:

      • We will review submitted comments as soon as possible, striving for within two business days.
      • This forum is intended for constructive dialogue. Comments that are commercial or promotional in nature, pertain to specific medical cases, are not relevant to the article for which they have been submitted, or are otherwise inappropriate will not be posted.
      • We require that commenters identify themselves with names and affiliations.
      • Comments must be in compliance with our Terms & Conditions.
      • Comments are not peer-reviewed.