The authors stated that, in addition to the studies included in the previous review, 78 studies (n=33,664) were identified for inclusion : 4 RCTs (6 publications), 17 non-randomised controlled trials, 22 comparative observational studies, 28 case series and 6 registry publications. They presented details of 78 studies in their table of included studies. However, in the table there were 6 RCTs, 16 non-randomised controlled trials, 20 comparative observational studies, 32 case series, and 4 case series that were registry publications.
The quality of three of the 4 included RCTs was high. The non-randomised controlled trials, comparative observational studies and case-series studies were poorer in quality and more prone to bias.
EVAR versus open repair.
EVAR was associated with a statistically significant lower 30-day mortality rate compared with open repair (3 RCTs; OR 0.33, 95% CI: 0.17, 0.64). This result was supported by data from 17 non-randomised controlled trials (OR 0.31, 95% CI: 0.25, 0.39). There was no significant heterogeneity between the studies. However, 2 RCTs that reported 2-year and 4-year follow-up indicated that the early survival benefit was lost by the end of the first year.
The primary conversion rate to open repair was 0.8% and 1.8% in the 2 RCTs, a mean of 2.4% (95% CI: 1.8, 3.0; range: 0 to 7.3) in the 12 non-randomised controlled trials, and a mean of 1.2% (95% CI: 1.0, 1.4) in the case series.
The early aneurysm rupture rate was reported as 0.2% in one non-randomised controlled trial. There was no significant difference in delayed rupture rates between EVAR and open repair (OR 5.0, 95% CI: 0.58, 42.94). Of the 6 non-randomised controlled trials, five reported a 0% rupture rate for EVAR and one reported a 1.1% rate.
Two RCTs reported significantly higher secondary intervention rates following EVAR compared with after open repair (OR 2.57, 95% CI: 1.70, 3.87; HR 2.9, 95% CI: 1.1, 6.2).
The most common technical adverse event following EVAR was a type II endoleak, which occurred in 19.6% of patients at less than 30 days, 12.9% at 1 year and 11.7% beyond 1 year. Pulmonary complications, incidence of blood loss and haemorrhagic events were significantly lower in the EVAR group than in the open repair group. There was no significant difference in the rates of cardiac events, lower limb ischaemia, renal impairment, graft infection, colonic ischaemia or local wound complications. Total hospital stay was reduced for the EVAR group compared with the open repair group (3 RCTs). Three RCTs and 8 non-randomised controlled trials demonstrated a significant reduction in intensive treatment unit stay after EVAR compared with after open repair.
Further results were reported.
EVAR in high-risk patients.
One RCT evaluated EVAR compared with no intervention in a population who were deemed to be unfit for open repair. This RCT reported no statistically significant difference in aneurysm-related deaths between EVAR and no intervention (12% versus 12.8%, respectively; HR 1.00, 95% CI: 0.54, 1.84), or total mortality rates (44.6% versus 39.5%) during the follow-up period. The rupture rate was 2% in the first 30 days after EVAR, and 12.2% in the no intervention group. Conversion to open repair occurred in 0.6% of cases during the primary procedure and 1.2% during the follow-up period. The complication rate following EVAR was 32.6%. At 4 years, 26% required at least one re-intervention following EVAR, compared with 4% in the no intervention group (HR 5.8, 95% CI: 2.4, 14.0, P<0.001); however, if crossovers from the no intervention group were considered secondary interventions, the rate for the no intervention group rose to 30%.