Results
Technically there was no inadequate CDUS examination. There were 11
detected endoleaks (13.1%) with CTA and 7 with CDUS (8.3%). There was
one type IA, three type IB, two type III and five type II endoleaks
detected with CTA (Table 2). CDUS was not able to detect four type II
endoleaks, however, all type I and III endoleaks were detected. Figure
1-4 demonstrate the type I and III endoleaks in Blood flow image(BFi)
and Color Doppler modes. Among the modes of CDUS, eventhough BFi was
much more demonstrable, there was no superiority or sensitivity
difference. There were no leaks missed on CDUS requiring intervention
however, there was an insufficiency of CDUS in detecting low flow.
Eliminating this frailty, there was a very high correlation of aneurysm
sac diameter measurement between CTA and CDUS. For CDUS average aneurysm
sac diameter was 57.1±14.5 mm, for CTA 58.7±15.0 mm. There was a very
strong correlation of CTA and CDUS for diameter measurement and
endoleaks (r = 0.884 and r = 0.777 respectively, p < 0.001 for
both).
The mean difference from Bland-Altman analysis (Figure 5) was 1.5 ± 2.2
mm and limits of agreement were -2.85 mm (lower limit) and 5.93 mm
(upper limit). A small percentage of measurements were outside of the ±
2 SD ranges. The direction and magnitude of the mean differences were
similar between the two methods of measurement (CDUS and BTA). Linear
regression analysis represented no proportional bias (p=0.233).
The sensitivity and specificity of CDUS was 63.6% and 100%
respectively. The accuracy was 95.2%. Positive and negative predictive
values were 100% and 94.8% respectively. The missed type II endoleaks
were not considered to be clinically significant. They are still under
follow up for sac enlargement. The detected type II endoleak was written
as “suspected endoleak” because of a mobile thrombus seen inside the
aneurysm sac.