Staged and hybrid approach in the treatment of complex aortic dissection

Multilevel aortic disease presents a formidable challenge for vascular surgeons. In the past, multilevel aortic surgery was performed simultaneously or subsequently. Single-stage intervention is thought to be associated with a high incidence of complications, and sequential repair requires several major surgical interventions. The rupture of residual lesion, however, may take place while waiting the second operation.1 Endovascular treatment is recognized as a viable alternative to the treatment of the diseased aortic areas especially in the high-risk patient, as it could markedly lower the incidence of complications compared with other procedures. However, thoracic aortic anatomy must be favorable to the placement of a stent-graft (SG). Sufficient proximal and distal landing zone is mandatory to deploy the SG and ensure a satisfactory exclusion result. Therefore, complex dissection with multiple-tear located extremely near the orifice of the visceral arteries is excluded from orthodox endovascular therapy. Here we present a case of multi-teared dissecting aneurysm treated by a four-staged hybrid surgery with placement of 5 SGs. CASE REPORT A 53-year-old male patient had an episode of aortic disease that could be traced back to his back pain 12 years ago. He was then misdiagnosed as having acute pancreatitis. When he had a routine ultrasound checkup one year later, he was found to have aortic abnormality. Further CT scan revealed an aortic dissection originated from the left subclavian artery down to the aortic bifurcation. He was thereafter referred to our center for further treatment. The patient was asymptomatic upon admission with his blood pressure well controlled. Considering the high risk relating to open surgery, the patient preferred to keep on medical therapy. He was re-admitted in June 2004 to undergo endovascular repair. CT scan after the readmission revealed an abdominal dissecting aneurysm which was resulted from a tear near the celiac artery. Since this hostile anatomy was then considered as a contraindication for endovascular treatment, the patient had to resume the medication with CT follow-up on a 6-month basis. Because CT and angiography (Figure 1) during the follow-up showed that the dissecting aneurysm was enlarging, the patient returned to seek surgical treatment in August 2005.Figure 1.: Angiography (A) and CT 3D reconstruction (B) demonstrate Debakey III aortic dissection, the false lumen of the descending aorta was thrombosised (C), the false lumen of the abdominal aorta was partly patent (D), the true lumen of the aorta near the bifurcation was compressed by the false lumen (E).A two-staged hybrid surgery was planned with visceral reconstruction as a primary intervention followed by a secondary stent-grafting to seal both the proximal and distal tears. The left kidney will be sacrificed if the function of the right kidney is sufficient after the bypass procedure. In October 2005, the patient received the first stage operation. A bifurcated prosthesis (14 cm × 7 cm, VASCUTEK) was sutured end to side from the right iliac artery to the superior mesenteric artery (SMA) and the common hepatic artery, with proximal ligation of the celiac artery and the SMA. Then the right renal artery was exposed and was found to be calcified in the main trunk unsuitable for anastomosis. Thus an right renal auto-transplantation had to be adopted via a retroperitoneal approach instead of a bypass reconstruction. The patient recovered well and post-operative CT scan (Figure 2) demonstrated the detail of the bypass and auto-transplanted kidney and dissected aorta. The renal function check revealed that the auto-transplanted kidney was moderately impaired, indicating that the left renal auto-transplantation was mandatory before stent-grafting. This was done in February 2006.Figure 2.: CT 3D reconstruction shows the bypass from right iliac artery to right hepatic and superior mesenteric artery was patent (big arrow), the appearance of the autotransplanted kidney was well (small arrow), J tube was patent (arrow head) (A, B). The cross section of the CT scan demonstrates double lumen in thoracicl aorta (C). The cross section of the CT scan demonstrates triple lumen in abdominal aorta near the origin of the celiac artery, part of the false lumen was patent, compressing the true lumen (dotted arrow) (D). The cross section of the CT scan demonstrates double lumen in near the bifurcation of the abdominal aorta, the bypass was patent (big arrow) (E).In April 2006, the patient received the first endovascular repair. Initially, 2 SGs were planed to seal both the proximal and distal tears in a single procedure. However, intra-operative angiography revealed, much different from that of the CT scan, that there were 3 tears in the upper, middle and lower parts of the descending aorta, resulting in a dissecting aneurysm as large as 7 cm in the proximal section. Recognizing the risk of paraplegia, we would only sealed the proximal tear with good result on completion angiogram. Two weeks later, the patient received another CT scan to locate the origin of the Adamkiewicz artery (ADA) at the T11 level. CT scan (Figure 3) demonstrated the detail of the bypass, the two auto-transplanted kidneys, and dissection.Figure 3.: The CT reconstruction after the deployment of the first stent graft shows that the true lumen below the lower part of the SG was small (small arrow), the false lumen was patent and wrap the true one (big arrow) (A, B), the cross section demonstrates the stent graft was in the position, and the false lumen was full of thrombus and compressing the true one (big dotted arrow) (C, D, E).Another two SGs were planed to seal the other two distal tears and carefully spared the origin of the ADA. The initial angiogram showed that the proximal SG was doing fine (Figure 4A, B), but another new tear was found between T9 and T10, just one vertebral below the distal end of the first SG. The false lumen was patent. Several intercostal arteries, including the major T11 one, were visible on angiogram. One SG (32 mm × 72 mm, 12 mm proximal bare stent, Microport) was deployed to cover the upper tear (Figure 4D). Then another SG (32 mm×80 mm, 40 mm proximal bare stent, Microport) was deployed to cover the lower one. Completion angiogram showed that the two SGs were right in position, sealing the two tears without compromising the ADA. Then an abdominal angiogram (Figure 4C) demonstrated a major tear located between the L1 and L2 levels, and the abdominal aorta had degenerated into a big dissecting aneurysm. Two SGs (30 mm×14 cm, endofit) were deployed between the upper edge of the T12 and L4 with 2 cm overlapping. Completion angiogram showed (Figure 4E, F) that the distal tear was successfully sealed; mild endoleak was found between the two SGs. All of the visceral arteries and the prosthetic bypasses were patent; the two auto-transplanted kidneys were well-visualized.Figure 4.: The angiography demonstrates the true lumen was small (small white arrow), the false lumen was patent and compressing the true one, the ADA related intercostals artery was clearly present (arrow head) (A, B, C). The second SG was deployed (D). The false lumen (big dotted arrow) subsided after 3 stent grafts were deployed (E). The bilateral auto-transplanted kidneys were good in shape (black arrow) (F).The patient recovered well after the fourth operation. The findings of CT scan 2 weeks later demonstrated that the true lumen was noticeably enlarged (Figure 5A, B); part of the false lumen was thrombosed and shrank (Figure 5C, D, E); mild endoleak could be seen between the two lower SGs (Figure 5D). The patient was subjected to a CT follow-up on a 3-month basis. The CTA (Figure 6A, B) at a 6-month follow-up found no endoleak in the false lumen. Upon a 9-month follow-up, the patient complained of petechia in the limbs. Blood tests showed a moderate to severe decrease in the levels of fibrinogen and platelet, and a moderate decrease in the hemoglobin level. The patient received several blood transfusions, erythropoietin injection, and oral medication to stimulate blood cell proliferation. No severe bleeding was observed during the follow-up. In the latest visit, blood tests revealed a baseline level of fibrinogen, platelet and hemoglobin. CT scan at this point demonstrated a continuous improvement of aortic morphology.Figure 5.: The CT reconstruction shows the appearance of the SGs were fine. Part of the false lumen was patent (arrow head), and the prosthetic vessel was well present and patent (black arrow) (A, B). The cross section manifests the true lumen of the upper part of thoracic aorta was enlarged, and the false lumen was almost disappeared (C), while the true lumen of the middle part of abdominal aorta was enlarged (small dotted arrow) (D). The in-stent lumen of the lower part of abdominal aorta was thoroughly expanded (small white arrow), and most of the false lumen was thrombosised (big white arrow) (E).Figure 6.: 3D CT reconstruction showed the appearance of the SGs were fine. Most of the false lumen was full of thrombosis (small arrow), and the prosthetic vessel was well present and patent (big arrow)DISCUSSION The morphology of dissecting aneurysm treated with medical therapy could change in a complex pattern. The diseased aorta will be degenerated aneurysm and occasionally a new acute dissection will occur on the previous one. In approximately 20% to 40% of patients with chronic aortic dissection, the descending thoracic and thoracoabdominal aorta would eventually become an aneurysm within 2 to 5 years.2 The estimated incidence of simultaneous thoracic and abdominal aortic disease is approximately 10%-20% of the patients.1,3 After a planned resection of only one aortic lesion, the 3- and 7-year survival rates were 48% and 12% respectively, compared with 32% and 2% for patients who received no treatment, but 64% and 34% for patients who underwent resection of two diseased areas.4 This patient with extensive aortic lesion was classified to be apparently in a high risk category. The aorta degenerated into a dissecting aneurysm with a diameter exceeding the 5 cm threshold for surgical intervention. The dissection of the aneurysm in this patient was rather complex. A total of 5 tears appeared in the course of the descending thoracic and abdominal aorta. Three of the 5 tears were close to the orifice of the visceral arteries. It is known that standard open surgery entails full prosthetic replacement of the descending thoracic and abdominal aorta plus visceral and intercostal reconstruction, which is technically demanding since it is associated with a high incidence of mortality and morbidity.5 Simultaneous abdominal and thoracic endovascular repair for multilevel aortic disease is feasible and could be a viable alternative in high-risk patients, who otherwise may not be candidates for conventional repair.6 However, endovascular treatment in one stage is not sufficient to treat some multilevel aortic diseases. Staged and hybrid approach is a novel attempt which involves the transposition of the great vessels in combination with endografting. It is effective to extend the landing zone, cut the operation time, minimize blood loss, and lower the risk of graft infection. Postoperative paraplegia or paraparesis is another problem in open surgery and endovascular treatment. The major cause of this complication is thought to be spinal cord ischemia during and after the procedure. A dominant anterior medullary artery in the thoracolumbar region, which is called the Adamkiewicz artery, supplies blood to the anterior spinal artery, feeding two thirds of the spinal cord perfusion.7 CTA is the most effective way to detect the ADA, which could achieve a success rate of 90% for the localization of the ADA while evaluating the entire aorta.8 Miyairi et al9 reported that the blood supply to the spine was not solely dependent on single segmental arteries, the collateral circulation was also very important. Generally, thoracic SGs overlie and exclude the intercostal arteries originating from the aneurysm and the proximal and distal landing zones. Thus, the coverage of a long thoracic aortic segment is thought to be a significant risk factor for spinal cord ischemia. Finally, ischemia could be precipitated by the late sealing of a type II endoleak or the late thrombosis of collateral pathways.10 In the repair of endovascular SG, paraplegia occurred in 3.7% of patients with thoracic aortic aneurysms11 and in 5.6% of patients undergoing open surgery for abdominal aortic aneurysm in combination with endovascular SG repair of thoracic aortic aneurysm.12 Hence, even with the availability of SGs, the increased risk of spinal cord ischemia and paraplegia in treating thoraco-abdominal aortic diseases should be considered when the best course of patient management is prescribed. The staged procedure could allow the development of collaterals for spinal cord blood supply.13 Hence, despite the deliberate reservation of the ADA, treating the patient with SG in a staged way is another way to avoid paraplegia. In our patient, bypassed visceral reconstruction was adopted to expand the SG landing zone in the first stage. A bypass was made from the right iliac artery to the celiac and SMA with a Y-shape prosthetic vessel to reconstruct the blood flow. We reconstructed the right renal artery and retained one kidney in this patient before the operation. The left renal artery was covered by the SG. Because the right renal artery, short and calcified, was unsuitable for anastomosis, auto-transplantation was adopted. The isotope nephrogram after operation indicated a mild to moderate injury to the transplanted kidney. To avoid the renal failure caused by the endovascular therapy, auto-transplantion of the left kidney in the second stage was performed. Several strategies were plotted for optimal renal protection by restricting renal medication, sufficient hydration before and after the procedure, and minimizing the volume of isotonic contrast medium. A baseline renal function was achieved after this procedure. Three tears were found in the descending thoracic aorta in the third stage. In view of the high risk of paraplegia caused by covering of the ADA after multiple stent-grafting, only the proximal tear was sealed by a single SG in the primary endovascular treatment. CT scan afterwards clarified the origin of the ADA. Finally, three SGs were deployed to seal the tears separately, sparing the origin of ADA. Forward “jumping” of the SG at deployment is a common phenomenon especially when the SG is short. A long SG with a long proximal bare stent is an option to secure accurate positioning. Bare stent could be deployed first and minimize forward “jumping”. In this patient, the two customized SGs to seal the abdominal lesion had a bare stent of 20 and 40 mm in length respectively. It was much convenient to use a long SG to cover the diseased portion of the aorta. SGs of 20 cm in length cover the extensive diseased aorta, thus avoiding the use of the trombone technique that frequently leads to endoleak. The overlapping between SGs increases the risk of device migration and type III endoleak. The commercial SGs such as TAG, Zenith and Valiant, which have a longer design, could be used in some conditions. In this patient, the dissected abdominal aorta was aneurismal with two tears. The lesion was 25 cm in length. As the maximal length of the SG was 20 cm, two SGs were used to seal the lesion. A post-operative CT scan showed mild type III endoleak arising from the two SGs. During the period of follow up, the endoleak decreased gradually. During the 6-month follow up, the levels of fibrinogen, platelet and hemoglobin decreased possibly because of constant thrombosis and thrombolysis in the sealed lumen after EVAR. In most situations, however, this process goes at a low level and is hardly noticeable. In this patient, the diseased artery was dissected and the abdominal part developed into a large aneurysm, with a profound consumption of coagulation components. With complete sealing of the false lumen, the coagulatory state would return to normal. It is feasible for the treatment of complex multi-tear dissection with the staged and hybrid method. The result of mid-term follow up in this patient was satisfactory. More study cases and longer follow-up are needed to determine the results and efficacy of the staged and hybrid method. We suggest the treatment of complex aortic dissection with corresponding staged and hybrid method according to specific characteristics of the lesion in each patient.