Researchers from the University Medical Center Groningen (Groningen, The Netherlands) and the University of Twente (Enschede, The Netherlands) have developed a predictive model that they claim could help preoperatively identify patients who are likely not to achieve adequate endograft attachment after endovascular aneurysm repair (EVAR).
Willemina van Veldhuizen (University Medical Centre Groningen) and colleagues, including members of the Virtual Stent Research Group, presented their findings in a paper published online ahead of print. European Journal of Vascular and Endovascular Surgery (EJVES).
By background, the authors first highlight the association between challenging infrarenal aortic neck characteristics, such as short attachment (minimum attachment length <10 mm circumference) on the first postoperative computed tomography angiogram (CTA), and an increased risk of type Ia endoleak after EVAR.The research team's objective was to develop a model to predict postoperative minimum attachment length in patients with abdominal aortic aneurysm (AAA) based on preoperative geometry.
Van Veldhuizen others Note that a statistical shape model was developed to obtain the principal component scores.The dataset included a total of 147 patients, of which 93 were treated with standard EVAR without complications and 54 were patients who were treated with EVAR but experienced late type Ia endoleak.
Regarding methodology, the authors detailed that they obtained the infrarenal minimum attachment length from the first postoperative CTA and dichotomized it into a length of less than 10 mm and a length of 10 mm or more. The researchers then used the principal component scores that were statistically significant between the minimum attachment length groups as inputs for five classification models, after which they determined the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity of each classification model.
Van Veldhuizen et al. report: Abes Of the 147 patients studied, 24 had an infrarenal minimum attachment length of less than 10 mm and 123 had a minimum attachment length of 10 mm or more.
The authors state that the gradient boosting model yielded the highest AUC of 0.77, and 114 (78%) patients were correctly classified using this model, with sensitivity (correctly predicted attachments less than 10 mm) and specificity (correctly predicted attachments ≥ 10 mm) of 0.7 and 0.79, respectively.
Van Veldhuizen others In summary, their presented model can predict the binarized minimum attachment length of the endograft to the infrarenal aortic neck for the treatment of AAA on the first postoperative CTA scan with 78% accuracy. Because a minimum attachment length of less than 10 mm is associated with a higher risk of type Ia endoleak, the researchers claim that their model “may help vascular specialists in the preoperative phase to accurately identify patients who are likely to fail to achieve adequate attachment after EVAR.”
In their discussion of the findings, the authors acknowledge a number of limitations of the study, including the fact that the data used were taken from different datasets, which may have introduced selection bias.
Looking forward, van Veldhuizen and colleagues suggest that the model “needs to be externally validated in consecutive patient series, including patients with early type Ia endoleaks, before being used as a patient-specific virtual stent placement tool in clinical practice.”
