In the United States, there are 9,000 new cases of acute aortic dissections each year. Compared to medical therapy, the current role of open surgical repair or endovascular treatment of acute and chronic descending aortic dissections (Type B) remains controversial given the high mortality rates for medical therapy and open surgical therapy. Established criteria for operative intervention include failure of medical therapy, progression of dissection, aneurysmal enlargement or rupture, or evidence of malperfusion.
Particularly in the high risk patient population, endograft stent exclusion is emerging as a promising alternative treatment for Type B dissections. The use of endografts can be performed with high technical success and results in decreased blood loss, shorter hospital and ICU stays, and decreased morbidity. In addition, preliminary reports of endograft treatment of thoracic aortic aneurysms (TAA) and Type B dissections demonstrate lower paraplegia and mortality rates compared to open surgical repair.
Single institution, retrospective review
We recently published a retrospective, single institution review10 of our endovascular experience in treating a cohort of 42 high risk patients having Type B dissections for the period August 1999 to March 2005. This patient cohort of high risk patients are poor surgical candidates for open surgical repair and were referred to our institution for emergency intervention. Both the AneuRx (n=5) and Talent (n=37) thoracic endografts (Medtronic Vascular; Santa Rosa, Calif) were deployed with the goal of excluding the proximal entry tear of the Type B dissection and promoting thrombosis of the false lumen.
Results
In our study, 42 patients (acute=25, chronic=17) were treated by endograft exclusion of Type B dissections and followed by serial thin cut three-dimensional (3D) CT reconstructions (Medical MetRx Solutions, West Lebanon, NH). Operative outcomes are summarized in Table I.
Treatment of acute dissections (220.7 ± 76 minutes) resulted in a statistically significant longer operative time (P = 0.014; 95% confidence interval [CI], 0.40-0.73) than did chronic dissections (149.9 ± 88.5 minutes) with greater estimated blood loss (892.0 ± 1011.3mL; median, 500mL) than for chronic dissections (443.8 ± 403.3mL; median, 350mL) (P = .17). The origin of the left subclavian artery was intentionally occluded in 11 patients (26%) without subsequent evidence of vascular compromise to the arm. Fenestration of the distal aortic dissection membrane was performed in a single semi-conscious patient with abdominal pain to prevent possible gut and limb ischemia.
Endografts were successfully placed in 58/60 (96.7%) of 60 procedures. A total of 60 procedures were performed including 42 primary interventions and 18 secondary interventions. We found the risk of reintervention in patients with acute dissections to be four times more likely compared to patients with chronic dissections (p = 0.01; 95 CI, 1.04-15.96). Also, the proximal entry site was identified intraoperatively by IVUS interrogation in all cases.
The procedural 30 day mortality was 6.7% (n=4) for 60 procedures with one late in-hospital death > 30 days from multisystem organ failure. Late deaths resulted from malignancy, respiratory failure (2), myocardial disease, and multisystem organ failure bringing the overall patient mortality to 21% (9/42). There was a 9.5% stroke rate (4/42) and a procedural stroke rate of 6.7% (4/60 procedures) and there was an association with coverage of the origin of the left subclavian artery and the development of a posterior circulation cerebral stroke (2/11). No instances of postoperative paraplegia were observed. Overall, major morbidity occurred in 17 (40%) of patients.
Effects of endograft exclusion on the false lumen
As shown in Table II, complete thrombosis of the false lumen at the level of the stent coverage occurred in 25 (61%) of 41 patients within 1 month and 15 (88%) of 17 patients at 12 months. There was complete thrombosis in 15 (60%) of 25 acute dissection patients by 1 month (60%) compared with 10 (63%) of 16 patients in the chronic dissections (P = NS). Of the available 17 CT scans with contrast at 12 months, complete thrombosis occurred in 9 of 11 acute dissections and 6 of 6 chronic dissections.
For both groups, 27 (64%) of the 42 patients (acute 17; chronic 10) exhibited complete volume regression of the false lumen (>95% decrease in blood flow (contrast) volume within the false lumen on CT scan). By 30 days, 14 (82%) of 17 patients in the acute dissections had completely regressed compared to 7 (70%) of 10 chronic dissections. Volume regression of the false lumen was 66.4% (acute) and 91.9% (chronic) at 6 months (P = .6). At 12 months, all remaining patients in the chronic dissections showed complete regression.
Changes in blood flow volume within the true lumen and total volume
Volumetric analysis showed comparable increases in the true lumen volumes of the acute dissections (32.1% and 41.7%) compared to chronic dissections (15.2% and 37.0%) at 6 and 12 months, respectively. Those patients who developed endoleaks or distal reperfusion of the false lumen requiring reintervention had much smaller changes in the true lumen volume (4.8% and 3.9%) at six and 12 months, respectively. We found the acute dissections that required reintervention demonstrated a progressive increase in total aortic volume (13.6% and 28.9%) at six and 12 months, respectively and increases in maximum aortic diameter.
Discussion
We have conducted an analysis of outcomes over time as changes in patient selection, technique and expertise has evolved. Since the time of our previous report,11 additional patients have been treated with one periprocedural mortality. The reintervention procedural rate in our recent cohort was 21.7% (5 out of 23 procedures) compared with 35.1% in the initial 24 patients (13/37). Previously, we reported a 13% patient mortality rate with a corresponding 8.1% procedural mortality (3/37) . More recently, patient mortality was 5.5% (1/18) with a procedural mortality of 4.3% (1/23).
As our experience with endograft exclusion of descending thoracic aortic dissections continues to evolve, we have adjusted our treatment strategies. For instance, we now choose endografts with dimensions similar to normal reference aortic diameter, ie., a non-dilated ascending or transverse arch. We expect volume regression to occur and size endografts to the anticipated final diameter following regression. In addition, we now attempt to attain 50% (5-6cm) of endograft overlap when deploying multiple 130 mm long devices to prevent junctional endoleaks. We continue to rely heavily on IVUS interrogation to determine the sizing and deployment of endografts and to confirm preoperative CT planning. Among its many uses, IVUS can locate the site of proximal entry tear, navigate the true lumen ensuring wire passage into the correct lumen, and even guide deployment without the need for angiography. Particularly with Type B dissections, IVUS can help establish adequacy of stent length coverage by demonstrating increased pulsatile flow in the true lumen and stagnant flow in the false lumen. Even when visceral organs are perfused by the false lumen, we rarely perform aortic fenestration procedures; instead, relying on IVUS to document adequate blood flow in the true lumen to prevent organ ischemia.
As shown by Dake2 and others, placement of endografts to exclude the proximal entry tear can promote thrombosis of the false lumen in most cases. We found >88% of patients treated by endograft exclusion to have complete thrombosis over the length of aorta covered by the endograft by one year. The majority having thrombosed the entire length of the dissection (60%) at one month. In theory, the placement of an endograft promotes increased flow through the true lumen and a dramatic decrease in flow into the false lumen. By following changes in the false lumen thrombosis, percent changes in true lumen volume, and changes in total aortic volume, we are able to identify those patients who are likely to have continued false lumen pressurization and the need for secondary reintervention.
Endograft exclusion required complete coverage of the entire length of the descending thoracic aorta from the left subclavian to the celiac axis in 15 (36%) out of 42 patients. And yet, no patient developed postoperative paraplegia. This is a marked improvement over open surgical repair of dissections when the risk of spinal cord ischemia is reported as 7 to 36%. Because of the low paraplegia rates associated with endograft treatment of aortic dissection, we do not routinely use prophylactive lumbar drain placement.
We routinely cover the left subclavian artery when the proximal entry tear is so close as to prevent adequate fixation (n = 11, 26%) of the proximal landing zone. In general, coverage of the left subclavian artery is well tolerated by patients due to the extensive collateral blood supply to the arm and we have not routinely performed carotid subclavian bypass in conjunction with left subclavian artery occlusion. Of interest, there were three patients who suffered a posterior circulation stroke. Two of three patients suffering a posterior circulation stroke required coverage of the origin of the left subclavian artery (2/11). Pre-intervention MRA did not predict the need for carotid to subclavian bypass due to the presence of an intracerebral anomaly or the presence of right vertebral or common carotid artery occlusive disease.
Conclusion
Preliminary experience with endografts to treat acute and chronic Type B dissections in a high risk patient cohort has shown promising initial results with lower mortality and paraplegia rates compared to open surgical repair. Stroke remains a significant risk, particularly if left subclavian artery coverage is required. As with all new technologies, long term results and device specific complications have yet to be determined and will require continued long term surveillance. A multicenter, randomized clinical study (INSTEAD trial) is currently in progress and Nienaber reports preliminary results of improved two year survival with endograft treatment (94.9%) compared to medical therapy (67.5%).
|
