A study published in the British Journal of Surgery has provided international validation of the St George’s Vascular Institute (SGVI) risk score, which has been developed to predict the risk of complications in patients undergoing endovascular aortic aneurysm repair (EVAR).
Alan Karthikesalingam (St George’s Vascular Institute, London, UK) and others report that “lifelong surveillance imaging” is considered mandatory in all patients who have undergone EVAR, but add that “there is considerable uncertainty regarding when and how surveillance should be performed”. Therefore, the authors explain, the SGVI risk score was developed to risk stratify patients (by identifying significant differences in individual aspects of aortic morphology) to provide greater insight into how often patients should undergo surveillance after EVAR. The aim of the present study was to validate the risk score, which has already been validated in two UK centres, in an international setting.
Karthikesalingam et al used the risk score to group patients from the ENGAGE registry, which was a postmarket study of patients (1,207 patients with a complete dataset) who underwent elective EVAR with Medtronic’s Endurant stent graft system at 79 centres in 30 countries, into low-risk (86.08% of the study population) and high-risk categories. The primary outcome was the development of endograft complications or the requirement for secondary reintervention.
The authors found that high-risk patients were more likely to be male (p<0.001), have a “considerably greater” angulation of the aneurysm neck (p<0.0001) with wider neck diameter measurements (p<0.0001), and significantly greater iliac tortuosity (p<0.0001) than low-risk patients. Karthikesalingam et al state: “Freedom from endograft complications was 87.9%, 84.8%, and 77.9% in the low-risk group and 77.9%, 74%, 69.6% in the high-risk groups at one, two, and three years respectively (p=0.0117)”. They add that, in multivariate Cox proportional hazards modelling, the GSVI risk score was the only “significant morphological predictor of endograft complications (p<0.001), which were not independently associated with aneurysm neck (p=0.664), proximal non-aneurysm neck diameter (p=0.380), infrarenal neck angulation (p=0.838), or iliac tortuosity (p=0.910).” Furthermore, the percentage of patients who remained free from reinterventions was significantly in the low-risk group (p<0.001) at one, two, and three years compared with the high-risk group.
Using data from the study population, the author simulated outcomes for 1,207,000 patients. They found that using current surveillance intervals (nine scans in five years) would only identify 25.53% of complications in low-risk patients and 27.44% of complications in high-risk patients. “A risk-stratified approach, increasingly frequency in the high-risk group to 12 scans in five years and reducing frequency in the low-risk groups to eight scans in five years, would improve the detection of complication to 37.25% in the high-risk group and 25.56% in the low-risk group,” Karthikesalingam et al comment.
They conclude: “This study has demonstrated that it is possible to stratify the risk of key long-term outcomes after EVAR, based on routinely available preoperative morphology. This might impact both patient selection and postoperative surveillance. Further research should be prospectively assess the impact of risk-stratified surveillance after EVAR.”