As you read this, you are probably wondering why the heck I am writing an editorial on complex aortic aneurysms when you know I am somewhere in the South focused on vascular trauma and limb ischemia. Well, no good deed goes unpunished. Patrick Chong presented the article, “Comparison of open surgery and endovascular techniques for juxtarenal and complex neck aortic aneurysms: The UK complex aneurysm study (UK-COMPASS) —perioperative and midterm outcomes,” to the VASCC Research Group WhatsApp group for comments, and the usual motley crew responded with differing impressions about the utility of the study. I suggested to Malachi Sheahan, medical editor, that he provide a commentary of the article in Vascular Specialist, and, Mal being Mal, suggested that I write the commentary. Thanks, my friend, and vascular bestie. I still love you, but I don’t like you anymore. Herein, I provide my analysis.
Registry data collected
This observational cohort study truly provides a snapshot of the evolving trends in the treatment of juxtarenal and complex neck abdominal aortic aneurysms (AAAs). The study looked at all UK patients undergoing treatment for AAAs between November 2017 and October 2019. Registry data were collected for patients with AAA ≥55mm in maximum transverse diameter and neck <10mm, and/or complex necks unsuitable for on-label standard endovascular aneurysm repair (EVAR). Aneurysms suitable for on-label standard EVAR, physician-modified endografts (PMEG), ruptured AAAs, and thoracic or thoracoabdominal aneurysms (TAAAs) were excluded. Treatment strategies for AAAs observed in the study included open surgical repair (OSR), fenestrated EVAR (FEVAR), combination of standard EVAR with adjuncts (endoluminal screws or parallel stent grafts) and off-label standard EVAR. The latter two cohorts were combined into one category due to low numbers (n=96) in the EVAR plus adjunct group.
Patients were stratified into six differing categories based on aneurysm neck anatomy and operative risk using the adjusted British Aneurysm Repair [BAR] score to calculate risk. Primary and secondary outcomes were compared for three differing anatomic considerations (0–4mm neck juxtarenal AAAs, 5–9 mm short neck AAAs, and ≥10mm neck/complex neck AAAs not suitable for on-label standard EVAR) and stratified for both high (highest quartile BAR score) and standard operative risk. The primary endpoint was death. Secondary endpoints included complications, secondary interventions for complications, and aneurysm-related death (defined as death from aneurysm-related complication or within 30 days of a secondary intervention).
The two primary comparisons of interest for this study were OSR vs. off-label EVAR and OSR vs. FEVAR. This study sought to address concerns that EVAR patients treated outside of instructions for use (IFU) fare poorer in the long term than similar patients undergoing OSR. As the patients were not randomized, propensity score stratification and regression adjustments were calculated to adjust for treatment biases related to patient baseline physiological fitness and aneurysm neck characteristics.
The Hospital Episode Statistics listed 8,994 patients as having undergone AAA repair in the UK during the study period. After exclusion criteria were applied, 1,916 patient datasets were reviewed.
Early outcomes
Nearly 3% of patients died in the peri-operative period. Not surprisingly, perioperative deaths were highest in the OSR group (4.5%), followed by FEVAR (2.2%, odds ratio [OR] 0.25, p≤0.001) and EVAR (1.2%, OR 0.24, p≤0.001). Death rates after OSR were higher for high-surgical-risk juxtarenal and short neck AAA patients (10.9% and 11.1%, respectively). Interestingly, death rate was highest (13.3%) in the high-operative-risk patients undergoing FEVAR for ≥10mm neck/complex AAAs. Perioperative complication rates were extremely high (55.4%) across the board and, again not surprisingly, occurred more frequently in high-risk patients and were highest (81.8%) in the OSR high-risk juxtarenal AAA cohort. Secondary early intervention rate was 7.7% for all comers and overall highest in the OSR group (8.9%), followed by FEVAR group (8.7%) and EVAR group (5.7%). One would not expect reintervention to be higher in the OSR than endovascular groups as they were in this study. Looking at each cohort separately, the early reintervention rate was highest (25%) in the high-operative-risk juxtrarenal AAA after EVAR.
So, some take home points: endovascular treatment of AAA is associated with lower perioperative morbidity and mortality and need for secondary interventions when compared with OSR. This is nothing new. FEVAR was associated with robust early perioperative mortality advantage, except for high-operative-risk ≥10mm neck/complex AAA patients. Treating patients off-label results in a quarter of these patients needing reintervention in the short term.
Midterm outcomes
Late all-cause death (excluding the perioperative deaths) was 17.8%, 24% after EVAR plus or minus adjuncts, 25.4% after FEVAR, and 9.7% after OSR. Adjusted risk of late death was significantly greater after EVAR (hazard ratio [HR] 2.18, 95% confidence interval [CI] 1.61–2.95; p<0.001) and FEVAR (HR 2.01, 95% CI 1.46–2.77; p<0.001) than after OSR. For the combined study population, as well as for short neck or complex neck AAA treatment, early survival benefit was lost and reversed during follow-up. The all-cause mortality rate after FEVAR and EVAR was double that of OSR by 3.5 years. Later secondary interventions were more frequent after endovascular AAA repair—FEVAR greater than EVAR—than after OSR.
Some take home points: patients selected for FEVAR and off-label EVAR died at two times the rate of patients undergoing OSR, and thus far there is not a good explanation for this.
So, what does this registry study tell me?
Its empirical evidence is as close to a randomized controlled trial as we can get for now since every patient who underwent AAA repair in the UK was included. Like the U.S., the UK has a varied practice for the treatment of juxtarenal, short neck and complex neck AAAs. In the UK, more than a third of AAA patients received off-label EVAR. Following off-label EVAR, 25% patients required reintervention within 30 days, and aneurysm-related death was 10 times higher compared with FEVAR and OSR for patients with juxtarenal AAAs.
For juxtarenal AAAs there is a robust perioperative mortality advantage to FEVAR over OSR, with equivalent survival up to midterm follow-up. Patients with short neck or complex neck AAAs lose their perioperative survival advantage midterm and have significantly worse survival compared with OSR patients. Higher midterm mortality is not explained by aneurysm-related mortality.
What are the implications of this study?
Avoid off-label use of EVAR when possible, especially in juxtarenal AAAs: it’s a no-brainer. FEVAR works well for 0–4mm neck/juxtarenal AAAs. For good-risk patients with ≥10mm necks/complex AAA necks, OSR is a viable option. Shocking!
I end in saying this: as vascular surgeons, we are obligated to offer our patients the most viable short- and long-term treatment option that they can tolerate. This means offering them OSR when appropriate. I believe the pendulum has swung too far in favor of endovascular treatment, not just for AAA care but also for the management of CLTI and other conditions. As such, I worry that, as a specialty, we are losing the art of open surgery when open surgery is the skillset that differentiates us from our interventional radiology and cardiology colleagues. Open surgical skills are essential to being a competent vascular surgeon. As a specialty, we need to ensure that our trainees, the future of our specialty, are learning and practicing the open surgical skills that they need to thrive and flourish as vascular surgeons. And that brings me back to why I am in the South. I am fortunate to still be practicing and teaching “the art of open surgery.” And for that I am grateful.
