Ultrasound elastography to quantify average percent pressure-normalized strain reduction associated with different aortic endografts in 3D-printed hydrogel phantoms

Abstract

Objective

Strain has become a viable index for evaluating abdominal aortic aneurysm stability after endovascular aneurysm repair (EVAR). In addition, literature has shown that healthy aortic tissue requires a degree of strain to maintain homeostasis. This has led to the hypothesis that too much strain reduction conferred by a high degree of graft oversizing is detrimental to the aneurysm neck in the seal zone of abdominal aortic aneurysms after EVAR. We investigated this in a laboratory experiment by examining the effects that graft oversizing has on the pressure-normalized strain ($\overline{{\epsilon }_{\rho +}}$/pulse pressure [PP]) reduction using four different infrarenal EVAR endografts and our ultrasound elastography technique. Approximate graft oversizing percentages were 20% (30 mm phantom-graft combinations), 30% (28 mm phantom-graft combinations), and 50% (24 mm phantom-graft combinations).

Methods

Axisymmetric, 10% by mass polyvinyl alcohol phantoms were connected to a flow simulator. Ultrasound elastography was performed before and after implantation with the four different endografts: (1) 36 mm polyester/stainless steel, (2) 36 mm polyester/electropolished nitinol, (3) 35 mm polytetrafluoroethylene (PTFE)/nitinol, and (4) 36 mm nitinol/polyester/platinum-iridium. Five ultrasound cine loops were taken of each phantom-graft combination. They were analyzed over two different cardiac cycles (end-diastole to end-diastole), yielding a total of 10 maximum mean principal strain ($\overline{{\epsilon }_{\rho +}}$) values. $\overline{{\epsilon }_{\rho +}}$ was divided by pulse pressure to yield pressure-normalized strain ($\overline{{\epsilon }_{\rho +}}$/PP). An analysis of variance was performed for graft comparisons. We calculated the average percent $\overline{{\epsilon }_{\rho +}}$/PP reduction by manufacturer and percent oversizing. These values were used for linear regression analysis.

Results

Results from one-way analysis of variance showed a significant difference in $\overline{{\epsilon }_{\rho +}}$/PP between the empty phantom condition and all oversizing conditions for all graft manufacturers (F(3, 56) = 106.7 [graft A], 132.7 [graft B], 106.5 [graft C], 105.7 [graft D], P < .0001 for grafts A-D). There was a significant difference when comparing the 50% condition with the 30% and 20% conditions across all manufacturers by post hoc analysis (P < .0001). No significant difference was found when comparing the 20% and 30% oversizing conditions for any of the manufacturers or when comparing $\overline{{\epsilon }_{\rho +}}$/PP values across the manufacturers according to percent oversize. Linear regression demonstrated a significant positive correlation between the percent graft oversize and the all-graft average percent $\overline{{\epsilon }_{\rho +}}$/PP reduction ($R^2$ = 0.84, P < .0001).

Conclusions

This brief report suggests that a 10% increase in graft oversizing leads to an approximate 5.9% reduction in $\overline{{\epsilon }_{\rho +}}$/PP on average. Applied clinically, this increase may result in increased stiffness in axisymmetric vessels after EVAR. Further research is needed to determine if this is clinically significant.