operative mortality. The proportion of patients with post-operative complications within 4 years of randomizationwas 41% in the EVAR group and 9% in the open repairgroup (HR 4.9; CI 3.5–6.8; p�0.0001). After 12 monthsthere was negligible difference in HRQL between the twogroups. The mean hospital costs per patient up to 4 yearswere UK£13,257 for the EVAR group versus £9946 for theopen repair group (mean difference £3311; SE 690).Conclusions: The researchers concluded that, comparedwith open repair, EVAR offers no advantage with respect toall-cause mortality or HRQL. They also concluded thatEVAR is more expensive than open repair and leads to agreater number of complications and reinterventions. Thespecific aneurysm-related mortality at 30 days was 3%better with EVAR than open repair, however. Finally, theinvestigators concluded that the continuing need for inter-ventions mandates ongoing patient surveillance, and thatlonger follow-up is needed for detailed cost-effectivenessassessment.Perspective: Two significant, randomized trials of endovas-cular versus open surgical repair of AAA reported nearlyidentical findings. Both the EVAR-1 study discussed hereand the DREAM study reported the same significantly lower30-day mortality for endovascular (EVAR) versus openrepair (approximately 3% lower in both studies). Now,both studies have longer-term follow-up, and they againreport nearly identical findings. Both DREAM and EVAR-1found no difference in mortality 2 years and 4 years afterrandomization, respectively. In the EVAR-1 study, thesmall difference in operative mortality persisted, but be-came insignificant in light of the much greater long-termmortality for patients with AAA. In DREAM, there was nodifference in overall long-term mortality. These studies sug-gest that the overall mortality risk for patients with AAA ismuch greater than that affected by any repair, whether openor endovascular. These results are unlikely to change, un-less longer-term follow-up data suggest that one repair issignificantly better than the other in preventing aneurysm-related death. The procedure-related mortality differencehas little or no impact on longer-term survival. This re-minds us that what procedure we perform for our patientsmatters less than how we care for their underlying athero-sclerosis. James Froehlich/Kim Eagle

Endovascular Aneurysm Repair and Outcome inPatients Unfit for Open Repair of Abdominal AorticAneurysm (EVAR Trial 2): Randomized ControlledTrialEVAR Trial Participants. Lancet 2005;365:2187–92.

Study Question: For patients with abdominal aortic aneu-rysm (AAA) of at least 5.5 cm, and who are medically unfitfor surgery, will endovascular repair (EVAR) versus nointervention reduce the risk of aneurysm-related death

from rupture and improve long-term survival and health-related quality of life (HRQL)?Methods: The investigators randomized 338 patients aged60 years or older who had aneurysms of at least 5.5 cm indiameter to receive either EVAR (n�166) or no interven-tion (n�172) at 31 participating hospitals in the UK. Theprimary end point was all-cause mortality, with secondaryend points of aneurysm-related mortality, health-relatedquality of life (HRQL), postoperative complications andhospital costs. Analyses were by intention to treat.Results: Of subjects randomized to EVAR, 144 of 166(87%) of those patients had an endograft implanted suc-cessfully, and 47 of 172 (27%) patients assigned no inter-vention underwent aneurysm exclusion, including 12 casesof open repair. Thus, 80% of patients adhered to protocol.The 30-day operative mortality in the EVAR group was 9%(13 of 150; 95%CI 5–15). If only elective cases were in-cluded, this operative mortality was reduced to 10 of 147(7%; 95%CI 3–12). The “no-intervention group” had arupture rate of 9.0 per 100 person-years (95%CI 6.0–13.5).No significant difference existed between the EVAR groupand the no-intervention group for all-cause mortality (haz-ard ratio [HR] 1.21; 95%CI 0.87–1.69; p�0.25). No differ-ence existed in aneurysm-related mortality. The overallreintervention rate was 11.5 per 100 person-years in theEVAR group and 1.8 per 100 person-years in the no-intervention group; by 4 years, 26% of patients in the EVARgroup had needed at least one reintervention comparedwith only 4% in the no-intervention group (HR 5.8; CI2.4–14.0; p�0.0001.) By posthoc per-protocol analysis,the HR for all-cause mortality was 1.07 (CI 0.75–1.52;p�0.70), which did not differ markedly from the analysisby intention to treat. The mean hospital costs per patientover 4 years were UK£13,632 in the EVAR group and£4983 in the no-intervention group (mean difference£8649, SE�1248), with no difference in HRQL scores.Conclusions: The researchers concluded that EVAR hadconsiderable 30-day operative mortality in patients be-lieved to be unfit for open repair of their aneurysm. EVARdid not improve overall survival or aneurysm-related mor-tality compared to “no-intervention” and was associatedwith a need for continued surveillance and reinterventions,at substantially increased cost. The investigators concludedthat both ongoing follow-up and improved fitness of pa-tients are priorities.Perspective: The promise of less invasive treatment for AAAwas the hope of avoiding the significant morbidity andmortality of open surgical repair, and thereby allowingtreatment of “high-risk” patients. Although endovascularrepair has significantly lower morbidity and mortality ini-tially, this study suggests that high-risk patients are justthat—even when undergoing minimally invasive proce-dures. (One important caveat in interpreting this study,however, is the unclear definition of “unfit” for surgery.)That said, the 30-day mortality in this study rivals that ofopen repair, and it makes less clear whether EVAR is appro-

ACC CURRENT JOURNAL REVIEW October 2005

53

priate in this population. More importantly, there seemedto be no benefit from EVAR compared with no-interventionin terms of aneurysm-related or all-cause mortality,whether by intention-to-treat or per-protocol analysis.Pending further study, the role of EVAR for AAA patientsdeemed “unfit” for surgery remains unclear. JamesFroehlich

Lifeline Registry of Endovascular Aneurysm Repair:Long-Term Primary Outcome MeasuresLifeline Registry of EVAR Publications Committee. J Vasc Surg2005;42:1–10.

Study Question: What is the long-term outcome after endo-vascular aneurysm repair (EVAR) of infrarenal abdominalaortic aneurysms (AAAs)?Methods: The researchers reviewed the primary outcomemeasures of patients treated with endovascular grafts (EG)in the Lifeline Registry of EVAR. The registry contains dataon 2664 EG patients and 334 open surgical control (SC)patients collected under four multicenter InvestigationalDevice Exemption (IDE) clinical trials that led to US Foodand Drug Administration (FDA) approval with mandatory5-year follow-up. Primary outcome measures included op-erative mortality, AAA-related death, all-cause mortality,aneurysm rupture and surgical conversion.Results: Pooled data from IDE clinical trials revealed thatEG patients were 3 years older (73�8 years) than SCpatients (70�8 years; p�0.01) and had significantly morecardiac comorbidities before treatment. However, there wasno difference in 30-day operative mortality between EG(1.7%) and SC (1.4%) (p�0.72). Both EG and SC weresuccessful in preventing rupture, with freedom from aneu-rysm rupture in 99.8% of EG and 100% of SC patients at 1year (p�0.51). Freedom from rupture remained at 99% inyears 1 to 6 after EG, with no increasing risk of late rupture.No significant difference existed in the AAA-related deathrate at 1 year between EG (98.2%) and SC (98.6%)(p�0.64). Freedom from AAA-related death remained at98% in years 1 to 6 after EG, with no increasing risk of lateAAA-related death. Kaplan-Meier analysis at 6 years re-vealed freedom from aneurysm rupture in 99%, freedomfrom AAA-related death in 98%, and freedom from surgicalconversion in 95% of EG patients. There was no differencein survival at 4 years between EG (74%) and SC (71%)(p�0.49). Overall EG patient survival at 5 years was 66%and at 6 years was 52%. Women had a higher risk ofrupture (2.4%) than did men (1.2%) (p�0.01) and a higherrate of surgical conversion (8.3%) than men (3.8%)(p�0.01) but had the same low AAA-related death rate(3.5%) as did men (2.1%) (p�0.16) at 5 years. Most sec-ondary interventional procedures (85%) were performed�30 days after EVAR. Freedom from secondary interven-tion was 84% at 1 year and 78% at 5 years.Conclusions: The investigators concluded that endovascularaneurysm repair using FDA-approved devices is a safe,

effective and durable treatment for anatomically suited pa-tients with infrarenal AAAs.Perspective: The analysis suggests that endovascular aneu-rysm repair using FDA-approved endografts is safe andeffective in preventing aneurysm rupture and avoidingAAA-related death in the great majority of patients under-going treatment. EVAR appears durable, with a low long-term risk of rupture and AAA death and a low likelihoodthat surgical conversion will be needed. Results of EVARremain favorable, even in the older higher-risk populationof patients who are poor candidates for surgery. Womenwho undergo endovascular aneurysm repair are at higherrisk of aneurysm rupture and surgical conversion than menbut exhibit the same low risk of aneurysm-related death asdo men. Overall, EVAR appears a safe, effective and durabletreatment for anatomically suited appropriate patients withinfrarenal AAAs. Debabrata Mukherjee

Genetic Analysis of Polymorphisms in BiologicallyRelevant Candidate Genes in Patients WithAbdominal Aortic AneurysmsOgata T, Shibamura H, Tromp G, et al. J Vasc Surg2005;41:1036 – 42.

Study Question: Are there polymorphisms in several candi-date genes encoding matrix metalloproteinases (MMPs),tissue inhibitors of metalloproteinases (TIMPs), TGF�-1,elastin or type III procollagen that can serve as geneticmarkers for a clinical susceptibility for abdominal aorticaneurysms (AAAs)?Methods: Genotyping for 14 previously identified polymor-phisms in 13 candidate genes for several MMPs, severalTIMPs, TGF�-1, elastin and type III procollagen was per-formed in 387 subjects with AAAs and 425 control subjectswithout AAAs. Results were analyzed and evaluated forsignificant associations between the particular genotypeand the existence of AAAs.Results: Of the 14 polymorphisms examined, only twopolymorphisms in the TIMP1 gene in male subjects withouta family history of AAA had a statistically significant asso-ciation. There was an association between MMP10 poly-morphism and gender in patients without a family historyof AAAs. There was an association between an elastin poly-morphism and a TIMP3 polymorphism and country oforigin in patients with a family history of AAAs.Conclusions: Genetic variation in the TIMP1, TIMP3,MMP10 and elastin genes may play a role in the occurrenceof AAA in some patients.Perspective: Abdominal aortic aneurysm is a common vas-cular disease that is quite treatable when identified beforethe aneurysm ruptures, but population screening is notthought to be cost-effective. Thus, many patients presentonly after AAAs rupture when the outcome of surgery ispoor. There is clearly a genetic influence on the incidence ofAAAs, though Mendelian inheritance patterns are not usu-

ACC CURRENT JOURNAL REVIEW October 2005

54