Discussion
Endovascular repair is the treatment of choice for aneurysms or dissections of the thoracic descending aorta. When the disease extends to the arch, endovascular treatment is not a good option in most cases7 and open repair is challenging. The development of hybrid prostheses and the FET procedure8-10 has simplified the surgical treatment of these pathologies, allowing either a single-stage open procedure, or facilitating an open or endovascular second stage11,12. In acute and chronic dissections FET stabilizes the true lumen, covers the reentry sites in proximal descending thoracic aorta and promotes false lumen thrombosis and aortic remodeling, reducing the risk of distal aorta rupture and reinterventions3-5.
Two different hybrid prosthesis are available in Europe: The E-vita Open (Jotec GmbH, Hechingen, Germany ) and the Thoraflex (Vascutek, Terumo, Inchinnan, UK ). The E-vita Open is the prosthesis of choice at our center. Its main advantage is the intussuscepted position of the arch prosthesis, which facilitates the distal anastomosis. On the contrary, the lack of a side branch requires additional maneuvers to establish early reperfusion.
Authors from Essen have reported the greatest experience with E-vita Open prosthesis. They operated on 307 patients from 2005 to 2018. In-hospital mortality was 13.4%. Permanent cerebral and spinal cord deficit occurred in 7.2% and 2.9%, respectively13. Freedom from reintervention (open or endovascular) in distal aorta was 73% at 8 years. In acute type A aortic dissections, the false lumen remained thrombosed or resolved completely along the stent-graft aortic segment in 90% of patients. Authors from Hannover and Bologna have published their combined experience in 437 patients from 2007 to 2017. Three different hybrid prosthesis were used: the Chavan-Haverich (Hannover, Germany ), the E-vita Open and the Thoraflex14. In-hospital mortality was 14.9%. Permanent stroke and spinal cord injury rates were 10.8% and 5.5%, respectively. 23.1% of patients required an additional procedure, open or endovascular, during follow-up.
Although our series is still short, results are similar to those centers with greater experience. In-hospital mortality, permanent stroke and spinal cord injury were 7.1%, 3.6% and 1.8%, in spite of a high preoperative risk (type A acute dissections 32.1%, urgent surgery 37.5% and previous cardiac surgery in 19.6%). This is due in part to an adequate patient selection, a case-to-case procedure planning, a limited and consistent operating room team (surgeon, nursing, anesthesiology and perfusion) and the standardization of the procedure. In our opinion, FET procedure is more reproducible than standard surgery in complex thoracic aortic disease, enabling small groups like ours to have comparable outcomes to bigger centers.
Since 2013 surgical technique has evolved. We have moved the distal FET anastomosis to Zone 0-2, started to perfuse the LSA during lower body circulatory arrest and perfusion in the thoracoabdominal aorta is reestablished once distal anastomosis is performed. Tsagakis et al showed that the combination of Z2 anastomosis, LSA perfusion and early reperfusion reduced the risk of postoperative renal replacement therapy, prolonged intubation, re-exploration for bleeding and laryngeal nerve lesion13. In our series in-hospital mortality and permanent stroke were reduced to 3.2% and 0% in the last two years (group II), respectively. This is probably related not only to the technique improvement but also to the greater experience. Z0, 1 o 2 anastomosis is less technically demanding. In addition, it is easier to reach in case of bleeding, reduces the risk of laryngeal nerve injury and avoids left pleural cavity opening.
We switched to bilateral antegrade cerebral perfusion because it is an easy maneuver which warrants an adequate perfusion of both hemispheres15. Although a very low stroke rate has been published with unilateral perfusion16, concern remains when the duration of selective perfusion is more than 50 minutes17.
Z2 anastomosis is combined with debranching of the LSA with an 8-mm dacron graft or all three supra-aortic arteries with a trifurcated graft during cooling. In this way, FET procedure is performed under perfusion of all supra-aortic arteries, improving cerebral and spinal cord perfusion during lower body circulatory arrest. To date, we have been able to dissect the LSA during cooling in all cases. When anatomy is unfavorable with a displaced LSA origin, an extension of the sternotomy incision to the left neck is done in order to get a wide left carotid artery dissection and mobilization. Alternatively, an extra-anatomic bypass to axillary artery through second intercostal space can be performed3.
Other improvements in our surgical technique were the interposition of a graft for arterial cannulation and the use of crystalloid Bretschneider´s solution for myocardial protection. First, the innominate artery became our first-choice artery for cannulation, as additional incisions are not required, and it is easy to access. In addition, an 8mm dacron graft is now systematically interposed to diminish the risk of retrograde dissection and improve distal perfusion18 . Secondly, due to the complexity of this procedure, optimal myocardial protection plays a key role in achieving a successful outcome. Crystalloid Bretschneider´s cardioplegia allows a prolonged myocardial protection in a single dose19. Finally,
The growing experience and the shorter lower body circulatory arrest time (Group I 54±20min versus Group II 33±10min, p<0.001) made it possible to raise nasopharyngeal temperature (Group I 22±3ºC vs Group II 25±3ºC). Length of stent graft was significantly shorter in Group II (135.5±5mm versus 150±9mm), which is explained because acute dissections are more frequent in this group (45.2% versus 16%). In acute aortic dissections, shorter stent grafts (130mm) are enough to cover the reentry sites in proximal descending thoracic aorta and promote false lumen thrombosis.
We must not forget that FET is a major surgery which combines prolonged time on CPB, hypothermic circulatory arrest and is frequently performed in high risk patients, so it is not exempted of significant complications3-5,13,14. Permanent stroke and specially spinal cord ischemia are the most feared and devastating. In a recent meta-analysis, Preventza et al analyze the pooled outcome of 3154 patients from thirty-five studies who underwent a FET procedure. In-hospital mortality was 8.8%, and permanent stroke and spinal cord injury (SCI) rates were 7.6% and 4.7%, respectively. Stent length of 15cm or greater and coverage extending to T8 were associated with a higher risk of spinal cord ischemia. The proposed mechanism is the compromise of collateral flow from segmental arteries. Although its role is unquestioned, there must be other factors involved, because the incidence of SCI in endovascular treatment of thoracic aorta has a significant lower rate21. During open surgery, a period of spinal cord ischemia occurs during circulatory arrest. During SACP spinal cord is not perfused below T8-T9 and protection depends primarily on temperature22.On this basis, spinal cord is at risk in FET procedures at moderate (28º-32º) temperatures with long systemic circulatory arrest times (>60min). A third factor is an inadequate spinal cord perfusion in the postoperative period. Spinal cord suffers an ischemic insult during circulatory arrest and several segmental arteries are occluded, so a period of systemic hypotension can trigger the SCI23. It is critical to maintain an adequate perfusion pressure during perioperative period, raising mean blood pressure over 70mmHg. Spinal cord perfusion pressure can also be improved with the use of intraoperative CSF drainage24. Other factors involved in neurological injury are hypoxia, anemia, hyperglycemia and hyperthermia25.
Paraplegia occurred in one patient: a 67-year-old woman operated in an emergent situation because an acute type A aortic dissection. A 22x130mm E-vita Open prosthesis was implanted in zone 0. She was extubated after six hours without any neurological deficit. 24 hours after surgery she presented a sudden massive hemorrhage and sustained severe hypotension. She was transferred to the operating room and bleeding was controlled. When she was extubated, 48h later, paraplegia was present. Magnetic resonance imaging showed ischemic damage at T8 level. Despite hemodynamic and metabolic optimization and CSF drainage, paraplegia was permanent. Due to her small size (155 cm in height) the stent graft covered thoracic aorta until T9. Our hypothesis is that although collateral flow compensated initially the occlusion of segmental arteries, hemorrhage and subsequent sustained hypotension triggered the ischemia.
This study has several limitations. First, although data is collected prospectively, its analysis is retrospective. Second, it is a single center experience, which makes difficult to support the external validity of our results. Finally, the sample size is small. We have almost doubled the number of procedures since we started the FET program; but, still, the center volume is low as compared to others (Bologna, Essen or Hannover).