Is CT scanning post-EVAR worth the cost?


Following the passing of the Screening Abdominal Aortic Aneurysms Very Efficiently (SAAAVE) Act through Congress, which will institute a Medicare one-time ultrasound screening for abdominal aortic aneurysms (AAA), one of the contemporary debates at this year’s VEITH meeting was between Ali F AbuRahma, Professor of Surgery, Charleston Medical Center, WV, and Geoffrey Gilling-Smith, Liverpool, UK, who debated the role of screening post-EVAR. While AbuRahma discussed the preferred modality used to detect endoleaks, Gilling-Smith talked about whether routine CT scanning post-EVAR is needed at all.

AbuRahma studied the role of Computed Tomography (CT) versus Color Duplex Ultrasound (CDUS) in the follow-up surveillance of endovascular grafts for AAAs, comparing how good both modalities are in detecting endoleak and measuring AAA diameters after endovascular repair.

AbuRahma and his team examined all patients with AAAs who underwent endovascular repair with three commercially available devices (Ancure, AneuRx and Excluder). The follow-up protocol included serial CT and CDUS scans at one month and every six months thereafter. In particular, sensitivity, specificity, positive predictive value, negative predictive value, Kappa statistics and the method of Bland and Altman to assess 95% limits of agreement were assessed with CT as the gold standard. Paired and unpaired t tests and correlation coefficients were used to compare the methods.

One hundred and seventy eight patients were enrolled in the study (86 Ancure, 55 AneuRx, and 37 Excluder), with a mean age of 74 years. Patients were followed for anything from one to 53 months (mean: 16 months). In total, 367 paired studies (CT and CDUS) were analyzed and a total of 34 (19%) endoleaks were noted (26 early and eight late endoleaks). The sensitivity, specificity, positive predictive value, and negative predictive value of CDUS in detecting endoleaks as 68%, 99%, 85%, and 97% (K statistic = 0.73), respectively. However, AbuRahma reported that CDUS was more accurate in detecting type I endoleak than type II (88% versus 50%, p<0.05). He disclosed that the mean diameter of the AAA sac after repair was by CT versus 4.99cm according to CDUS (p=0.07). The results also showed that a total of 93% of paired studies were to within 5mm. Pre- to post-operative AAA size changes throughout follow-up were -0.60 mm for CT versus -0.58 mm for CDUS (p=0.78). AbuRahma concluded that, “although CDUS has good correlation to CT in measuring the size of AAAs, it has a lower sensitivity in detecting endoleak, particularly type II endoleak. Therefore, CT scans should remain the primary imaging for the diagnosis of endoleak”. However, Gilling-Smith questioned whether routine CT is really necessary post-EVAR, arguing that it is both inconvenient for the patient and adds significantly to the cost of endovascular repair. Indeed, cost remains the fundamental issue on both sides on the Atlantic. There is now objective evidence that endovascular repair is safer than open repair. EVAR can be performed with lower morbidity and mortality than open surgical repair and this survival advantage is maintained for at least four years after operation. An inherent disadvantage of endovascular repair however is that the aneurysm remains in situ and susceptible to late rupture should it cease to be isolated from the circulation. For this reason, the integrity of the repair must be monitored throughout the remainder of the patient’s life. “But those who are responsible for funding and purchasing healthcare baulk at the additional cost of endovascular repair…If endovascular repair is to be accepted as the treatment of choice, the issues of cost and patient acceptance must be addressed.” Therefore, Gilling-Smith said, surveillance protocols should be examined to ensure that they are effective, safe, acceptable to patients and cheap. He said that the most important risk factors for aneurysm rupture after endovascular repair are graft-related endoleak, migration, graft-limb dislocation, stent fracture and fabric tear. Since expansion of the aneurysm may be evidence of persistent or recurrent pressurisation of the aneurysm sac, surveillance should also include monitoring of aneurysm size. Historically, such surveillance protocols have relied on serial CT scanning, but this is expensive, time consuming and hazardous, exposing patients to a substantial cumulative radiation burden and risk of malignancy. Therefore, Gilling-Smith posed the question, can continued reliance on CT be justified? He answered the question by stating that the presence or absence of graft-related endoleak can be determined by duplex scanning. Furthermore, migration, impending dislocation of a graft limb and/or stent graft distortion and fracture can all be detected using plain abdominal radiography. “Arguably,” he continued, “the only justification for CT scanning is to determine whether or not the aneurysm is expanding.” Gilling-Smith and his team retrospectively analyzed their endovascular database in order to determine whether duplex scanning could also be employed to monitor aneurysm size. He reported data from 99 patients who had been followed for at least one year after operation and who had undergone both CT and CDUS throughout follow up. For each patient, CT and US measurements of maximum aneurysm diameter (MAD) were plotted and independently examined by two observers. At each follow-up interval, MAD was compared with first postoperative and most recent MAD to determine if the aneurysm was expanding, shrinking or stable. A > 5mm change was considered significant. CT and US findings were compared to determine level of agreement.

In three patients, CT scans showed expansion while CDUS scans did not. However, in each case, US revealed expansion at the next follow-up. No cause for expansion was identified or intervention required prior to CDUS diagnosis. In 18 patients, CDUS scanning revealed expansion while CT did not. In six of these, expansion was revealed on subsequent CT. In the remaining 78 patients CT and CDUS scanning were concordant. Gilling-Smith concluded, “In our institution, duplex scanning can reliably detect expansion of the aneurysm after endovascular repair. Routine CT scanning is therefore unnecessary.”

He stated that patients in his institution now undergo duplex scanning, plain abdominal radiography and baseline CT scanning one month after endovascular repair. Thereafter, patients undergo annual duplex scanning and plain abdominal radiography. CT scanning is only performed if duplex scanning is technically unsatisfactory, equivocal or reveals a problem.

As a result, Gilling-Smith believes he has resolved some of the issues surrounding post-EVAR screening. “The abolition of routine CT scanning has resulted in a significant saving so that the cost of endovascular repair plus surveillance up to four years is close to the cost of open surgical repair. Moreover, a reduction in the frequency of examinations combined with a very significant reduction in the cumulative dose of ionizing radiation are additional and important benefits for the patient.”