European Journal of Cardio-Thoracic Surgery Advance Access published February 19, 2014
Key success factors for thoracic endovascular aortic repair for non-acute Stanford type B aortic dissection† Tadashi Kitamuraa,*, Shinzo Toriia, Norihiko Okaa, Tetsuya Horaia, Kouki Nakashimaa, Keiichi Itatania, Sachi Koyamaa, Yosuke Haria, Haruna Arakia, Hajime Satob and Kagami Miyajia a b
Department of Cardiovascular Surgery, Kitasato University School of Medicine, Sagamihara, Kanagawa, Japan Department of Health Policy and Technology Assessment, National Institute of Public Health, Wako, Saitama, Japan
* Corresponding author. Department of Cardiovascular Surgery, Kitasato University School of Medicine, 1-15-1 Kitasato, Minami, Sagamihara, Kanagawa 252-0374, Japan. Tel: +81-42-7788111; fax: +81-42-7789840; e-mail: [email protected] (T. Kitamura). Received 13 September 2013; received in revised form 11 December 2013; accepted 26 December 2013
Abstract OBJECTIVES: We aimed to determine the key factors associated with successful early and late outcomes after thoracic endovascular aortic repair (TEVAR) for non-acute Stanford type B aortic dissection at our institution. METHODS: Inpatient and outpatient records were retrospectively reviewed. Patients operated on within 14 days after the onset of acute aortic dissection and those with rupture or malperfusion were excluded. RESULTS: Forty-five patients (mean age, 55.5 ± 13.1 years; 23–79 years) underwent 53 TEVAR operations for non-acute Stanford type B aortic dissection between 1998 and 2012. Thirty-four patients had a patent false lumen and 19 had an ulcer-like projection (ULP). No early mortality was observed. At late follow-up (7.5 ± 3.9 years) of the 45 patients, survival after the initial TEVAR was 100, 86 and 63%; freedom from aortic reintervention was 87, 73 and 59%; and freedom from open aortic surgery was 89, 84 and 73%, at 1, 5 and 10 years, respectively. Of 15 late deaths, 2 were due to aortic rupture and 2 were operative deaths associated with aortic surgery. Of the 34 patients with patent false lumens before TEVAR, 25 had their descending false lumens thrombosed; of these 25, 16 had remodelling of the descending aorta; and of these 16, 4 had complete obliteration of the false lumen of the entire aorta. By bivariate analysis, the site of the primary entry and age were significantly associated with thrombosis of the descending false lumen, maximum aortic diameter was associated with remodelling of the descending aorta, and absence of abdominal branches arising from the false lumen was associated with complete obliteration of the false lumen of the entire aorta. CONCLUSIONS: The early results of TEVAR for non-acute Stanford type B aortic dissection were favourable. However, for cases with patent false lumens, complete obliteration of the false lumen of the entire aorta was difficult to achieve. Absence of the primary entry at the outer curvature of the distal aortic arch, younger age, small aortic diameter and absence of the abdominal aortic branches arising from the false lumen were the key success factors. Keywords: Chronic type B aortic dissection • Aortic remodelling • Thoracic endovascular aortic repair
INTRODUCTION Uncomplicated chronic Stanford type B aortic dissection does not require intervention, and medical treatment with blood pressure control is normally indicated unless the condition is associated with aortic dilatation [1]. When the dissected aorta becomes dilated, intervention to the aneurysmal portion of the aorta is required. Conventional open surgery for chronic type B aortic dissection is still associated with high risk, with a reported in-hospital mortality of 4.5–18.8% [2, 3]. On the other hand, the advantage of thoracic endovascular aortic repair (TEVAR) over conventional open surgery in treating descending aortic aneurysm, especially in high-risk patients, has been well demonstrated [4]. Therefore, it is natural that † Presented at the 27th Annual Meeting of the European Association for CardioThoracic Surgery, Vienna, Austria, 5–9 October 2013.
there has been a drive to expand the indication of TEVAR to aortic dissection so that uncomplicated type B aortic dissection can be treated with stent grafting before the aorta becomes dilated and increases the risk of rupture. The goal of TEVAR here includes not only the improvement of survival but also the prevention of aortic dilatation to save patients from future high-risk surgery. In the INvestigation of STEnt Grafts in Aortic Dissection (INSTEAD) trial, which was conducted to compare TEVAR with medical treatment for stable type B aortic dissection at least 2 weeks after onset, no significant differences were observed in 2-year survival, aorta-related deaths or freedom from late aortic operation [5]. However, at the same time, the study proved that TEVAR had a positive impact on remodelling of the dissected descending aorta [6, 7]. The purpose of this study was to evaluate early and late outcomes after TEVAR for non-acute Stanford type B aortic dissection at our institution and to identify key success factors.
© The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
AORTIC SURGERY
ORIGINAL ARTICLE
European Journal of Cardio-Thoracic Surgery (2014) 1–6 doi:10.1093/ejcts/ezu012
2
T. Kitamura et al. / European Journal of Cardio-Thoracic Surgery
PATIENTS AND METHODS From 1998 to 2012, TEVAR was performed in 145 patients at Kitasato University Hospital. Among these, 82 were aortic dissection; 53 of which were non-acute cases. These 53 cases were selected for detailed analysis and were included in this study. Patients operated on within 14 days after onset, emergency patients with rupture or malperfusion, and Stanford type A dissection patients with prior proximal aortic surgery were excluded from this study. The patient medical records were retrospectively reviewed for preoperative demographics, diagnosis and operative findings including the length and size of the stent graft. The stent grafts used were handmade or Matsui-Kitamura (M-K) stent grafts [8]. Handmade stent grafts consisted of electrostatic-welded piano wire stents ( Japan Piano Wire Industry Standard no. 304) covered with woven Dacron grafts (Kitamura MFG Co., Ltd. Ishikawa, Japan). Of note, 0.4 mm stent wires were used for devices sized 30 mm or smaller, 0.45 mm for those sized 32–38 mm and 0.5 mm for those sized 40 mm or larger. Matsui-Kitamura stent grafts are made of a braided framework of a nitinol wire, which is covered with a woven polyester tube graft. Technical success was defined as the elimination of antegrade perfusion of the false lumen on the completion aortogram acquired before leaving the operating theatre. All of the patients underwent computed tomography (CT) angiography preoperatively, within 10 days after the procedure and in the late period (5.0 ± 4.3 years). CT images were used for the measurement of the true lumen diameter, aortic diameter and the site of the initial entry. Remodelling of the descending aorta was defined as thrombosis of the false lumen of the descending aorta without enlargement of the descending aortic diameter.
Statistical methods Stata 11.2 Special Edition (StataCorp, College Station, TX, USA) was used to perform statistical analysis. Continuous variables were presented as means ± standard deviations (SDs). Student’s t-test was used to analyse differences in continuous variables. Fisher’s exact test was used to compare categorical variables. Kaplan–Meier analysis was used to analyse survival. To examine the best models to predict the occurrence of thrombosis of the descending false lumen and remodelling of the descending aorta, logistic regression models were built with thrombosis/remodelling being a dependent variable. As explanatory variables, age, sex, site of the entry, maximum aortic diameter, origin of the abdominal branches, stent graft length and others were entered and removed at the level of P = 0.05 to produce the best predictive models. All P-values were two-sided, and differences were considered statistically significant for P < 0.05.
RESULTS Forty-five patients underwent 53 TEVAR operations for non-acute Stanford type B aortic dissection during the study period. Table 1 shows the operative data of the total of 53 TEVAR procedures. The false lumen of the aortic dissection was patent in 34 cases, and a thrombosed false lumen with an ulcer-like projection (ULP) was observed in 19 cases. The technical success rate was 100% and no early mortality was observed.
Table 1: (n = 53)
Description of the subjects: operative data
Male/female Marfan syndrome Age (years ± SD (range)) Interval from the onset of acute aortic dissection Prior TEVAR Haemodialysis Approach Femoral artery Iliac artery Abdominal aorta Pathology Patent false lumen ULP Mortality Morbidity Stroke New dialysis Prolonged ventilation Arterio-venous fistula Groin haematoma Paraplegia
50/3 0 55.5 ± 13.1 (23–79) 1.4 ± 2.8 years (15 days—15 years) 11 4 50 2 1 34 19 0 2 (3.8%) 1 1 1 1 0
The 45 patients were followed up for 7.5 ± 3.9 years (1.1–14.9 years), and the follow-up was 100% complete. Actuarial survival rates after the initial TEVAR were 100, 86 and 63%; rates of freedom from reintervention were 87, 73 and 59% and rates of freedom from open aortic surgery were 89, 84 and 73% at 1, 5 and 10 years, respectively (Fig. 1). There were 15 late deaths; 2 were caused by rupture of the dilated aorta, 2 by operative deaths following an open aortic surgery, 5 by cancer, 3 by pneumonia, 2 by stroke and 1 by gastrointestinal bleeding. Of the 34 patients who had patent false lumens before TEVAR, the primary entry was present in the outer curvature of the distal aortic arch in 14 patients, in the inner curvature of the distal aortic arch in 11 patients and in the descending aortic wall in 9 patients. The preoperative maximum aortic diameter was 44 ± 11 mm and the length of the stent graft was 115 ± 41 mm. The primary entry in the outer curvature was not significantly correlated with a larger aortic diameter (P = 0.082). Twelve patients had all 4 abdominal branches arising from the true lumen, 20 patients had 3 of the abdominal branches arising from the true lumen and 2 patients had 2 branches from the true lumen. Twenty-five patients of 34 (74%) had their descending false lumen thrombosed, 16 (47%) had remodelling of the descending aorta (Figs 2 and 3) and 4 (12%) had complete obliteration of the false lumen of the entire aorta during the study period. By bivariate analysis, it was found that absence of the primary entry at the outer curvature of the distal aortic arch was significantly associated with thrombosis of the descending false lumen, that a smaller aortic diameter was more likely to result in remodelling of the descending aorta and that all four abdominal branches arising from the true lumen was more likely to proceed to the obliteration of the false lumen of the entire aorta (Table 2). Stent graft length was not significantly correlated with late thrombosis of the false lumen. A combination of absence of the primary entry at the outer curvature and age is the best predictive model for thrombosis, while only the single variable, maximum aortic diameter, best predicts remodelling (Table 3). Since all cases with complete obliteration of the false lumen of the entire aorta had all four abdominal
3
AORTIC SURGERY
T. Kitamura et al. / European Journal of Cardio-Thoracic Surgery
Figure 2: Successful remodelling of the dissected descending aorta after TEVAR. (A) Preoperative CT image and (B) 2 years after TEVAR.
DISCUSSION
Figure 1: Kaplan–Meier analysis. (A) Actuarial survival after the initial TEVAR, (B) freedom from reintervention and (C) freedom from open aortic surgery.
branches arising from the true lumen, logistic regression analysis was not performed. Instead, the odds ratio was calculated [9]. For patients with all four abdominal branches arising from the true lumen, the odds of obliteration of the false lumen of the entire aorta are 2.95 times as large as the odds for those with one or more abdominal branches arising from the false lumen to produce obliteration (P = 0.011 by Fisher’s exact test). Analyses using a maximum aortic diameter of ≤35 vs >35 mm as a dichotomous variable resulted in more or less the same results. Of the 19 ULP patients, 5 (26%) developed recurrent ULPs during the study period, and 2 of the 5 were on haemodialysis (Fig. 4). Although maximum aortic diameter and preoperative haemodialysis were not significant factors in univariate analysis, and the statistical calculation was unreliable in multivariate logistic regression analysis, they did correlate with ULP recurrence.
A certain percentage of patients with uncomplicated Stanford type B aortic dissection develop late dilatation of the dissected descending and thoracoabdominal aorta because of a patent primary entry [10]. Consequently, patients with chronic type B aortic dissection often have to undergo graft replacement of the descending aorta followed by that of the thoracoabdominal aorta. If the entire descending and thoracoabdominal aorta is dilated, extensive replacement of the thoracoabdominal aorta is required in the first place. Surgery for chronic type B aortic dissection is associated with considerable risk [11]. According to the 2010 annual report by the Japanese Association for Thoracic Surgery, in-hospital mortality rate due to surgery for dissecting aortic aneurysm was 4.5% for replacement of the descending aorta and 18.8% for replacement of the thoracoabdominal aorta [2]. On the other hand, the advantage of TEVAR over conventional open surgery in treating descending aortic aneurysm has been well demonstrated [12], and TEVAR has been widely accepted as the optimal therapy for acute, complicated Stanford type B aortic dissection [13]. The goal of TEVAR for chronic type B aortic dissection includes not only improvement of prognosis but also prevention of aortic dilatation to save the patient from future surgical intervention. As shown in our study, aortic dissection heals after TEVAR in some cases, and there are even more cases in which remodelling of the descending aorta occurs, saving the patient from replacement of the descending aorta [6, 7]. The first step in healing of chronic type B aortic dissection with a patent false lumen is thrombosis of the descending false lumen.
4
T. Kitamura et al. / European Journal of Cardio-Thoracic Surgery
Figure 3: Thrombosis of the descending false lumen after TEVAR without remodelling. (A) Preoperative CT image, (B) 2 years after TEVAR and (C) 6 years after TEVAR. This patient underwent graft replacement of the thoracoabdominal aorta.
In our series, the primary entry at the outer curvature of the distal aortic arch was associated with a lower chance of thrombosis of the descending false lumen. This might be attributed to malapposition of the stent graft because of the acute angle of the aortic arch, to late distal dislodging of the outer circumference of the stent graft because of the expansion of the true lumen or to change in the aortic arch geometry after TEVAR [14]. The latest generation device may improve malapposition and increase the chance for descending false lumen thrombosis [15]. Progress in new techniques, including arch debranching [16] or the chimney graft technique [17], may result in better landing of the prosthesis, contributing to an improved thrombosis rate even for patients
with primary entry at the outer curvature of the distal aortic arch. However, the high stroke rate related to these new techniques remains a concern. Primary entry at the inner curvature of the distal aortic arch may be a good indication for early intervention with TEVAR because it has been reported that the primary entry at the inner curvature of the distal arch was associated with a higher risk of complication in acute type B aortic dissection [18, 19]. Successful thrombosis of the descending false lumen is not always followed by remodelling of the descending aorta. In the present study, 74% of the patients with a patent false lumen had their descending false lumen thrombosed after TEVAR but only 47% achieved remodelling of the descending aorta. A large aortic diameter had a negative impact on late descending remodelling. It has been reported that the rate at which the true lumen volume increases after TEVAR is not parallel to that of the reduction in false lumen volume; true lumen expansion plateaus 12 months after TEVAR, whereas false lumen volume reduction gradually continues beyond 12 months [20]. It is speculated that the false lumen with a large volume associated with a dissecting aneurysm having a large diameter would not yet be depressurized when the true lumen expansion plateaus, resulting in the failure of descending remodelling. Complete obliteration of the false lumen on top of descending remodelling, in other words, healing of the dissection, was achieved in only 12% of the cases with a patent false lumen in our series. However, this result does not mean that TEVAR is not an effective treatment option for chronic type B aortic dissection because remodelling of the descending aorta brought in by TEVAR may be able to prevent open descending aortic replacement and may delay the expansion of the thoracoabdominal aorta. Fenestrated endografting to cover the whole aorta may be realized in future, but the risk of spinal cord ischaemia remains a concern. TEVAR is certainly a promising option for chronic type B aortic dissection, but it should not be performed blindly, considering potential early and late complications, including retrograde type A aortic dissection, endoleaks, spinal cord ischaemia and intimal flap injury due to prosthesis [21–23]. As the number of TEVAR procedures has increased, an increasing number of open aortic operations with prior TEVAR have recently been performed [24, 25]. Stent grafts within the aorta can cause a nuisance at open reintervention [24], sometimes involving a non-negligible risk of flap injury during explanting and a risk of spinal cord ischaemia, which is particularly observed in emergency settings. Therefore, initial patient selection is very important. TEVAR can be used to treat chronic localized aortic dissection with a ULP as safely as an atherosclerotic descending aortic aneurysm and with the same reliable long-term outcomes. However, postoperative regular imaging is required to detect recurrent ULPs, especially in patients on haemodialysis. For chronic type B aortic dissection with a patent false lumen, absence of the primary entry at the outer curvature of the distal aortic arch, smaller aortic diameter and absence of the abdominal branches arising from the false lumen would be the key success factors. It is hoped that subgroup analysis will be performed regarding TEVAR vs medical treatment in this group of patients in a large randomized study. This study had some limitations. This was a retrospective study with a small number of patients, and the stent grafts used were mostly handmade, which meant that the quality may have varied. However, because of their non-uniformity, various prostheses with different sizes and lengths were included, enabling analysis of whether the size and length of the prosthesis were associated with success.
T. Kitamura et al. / European Journal of Cardio-Thoracic Surgery
Table 2: Factors for thrombosis of the descending false lumen, remodelling of the descending aorta and complete obliteration of the false lumen of the entire aorta for the cases with a patent false lumen (n = 34) Thrombosis
Sex (Female/male) Entry (inner curvature and descending/ outer curvature) All abdominal branches from the true lumen (no/yes) Stent graft length (
Key success factors for thoracic endovascular aortic repair for non-acute Stanford type B aortic dissection† Tadashi Kitamuraa,*, Shinzo Toriia, Norihiko Okaa, Tetsuya Horaia, Kouki Nakashimaa, Keiichi Itatania, Sachi Koyamaa, Yosuke Haria, Haruna Arakia, Hajime Satob and Kagami Miyajia a b
Department of Cardiovascular Surgery, Kitasato University School of Medicine, Sagamihara, Kanagawa, Japan Department of Health Policy and Technology Assessment, National Institute of Public Health, Wako, Saitama, Japan
* Corresponding author. Department of Cardiovascular Surgery, Kitasato University School of Medicine, 1-15-1 Kitasato, Minami, Sagamihara, Kanagawa 252-0374, Japan. Tel: +81-42-7788111; fax: +81-42-7789840; e-mail: [email protected] (T. Kitamura). Received 13 September 2013; received in revised form 11 December 2013; accepted 26 December 2013
Abstract OBJECTIVES: We aimed to determine the key factors associated with successful early and late outcomes after thoracic endovascular aortic repair (TEVAR) for non-acute Stanford type B aortic dissection at our institution. METHODS: Inpatient and outpatient records were retrospectively reviewed. Patients operated on within 14 days after the onset of acute aortic dissection and those with rupture or malperfusion were excluded. RESULTS: Forty-five patients (mean age, 55.5 ± 13.1 years; 23–79 years) underwent 53 TEVAR operations for non-acute Stanford type B aortic dissection between 1998 and 2012. Thirty-four patients had a patent false lumen and 19 had an ulcer-like projection (ULP). No early mortality was observed. At late follow-up (7.5 ± 3.9 years) of the 45 patients, survival after the initial TEVAR was 100, 86 and 63%; freedom from aortic reintervention was 87, 73 and 59%; and freedom from open aortic surgery was 89, 84 and 73%, at 1, 5 and 10 years, respectively. Of 15 late deaths, 2 were due to aortic rupture and 2 were operative deaths associated with aortic surgery. Of the 34 patients with patent false lumens before TEVAR, 25 had their descending false lumens thrombosed; of these 25, 16 had remodelling of the descending aorta; and of these 16, 4 had complete obliteration of the false lumen of the entire aorta. By bivariate analysis, the site of the primary entry and age were significantly associated with thrombosis of the descending false lumen, maximum aortic diameter was associated with remodelling of the descending aorta, and absence of abdominal branches arising from the false lumen was associated with complete obliteration of the false lumen of the entire aorta. CONCLUSIONS: The early results of TEVAR for non-acute Stanford type B aortic dissection were favourable. However, for cases with patent false lumens, complete obliteration of the false lumen of the entire aorta was difficult to achieve. Absence of the primary entry at the outer curvature of the distal aortic arch, younger age, small aortic diameter and absence of the abdominal aortic branches arising from the false lumen were the key success factors. Keywords: Chronic type B aortic dissection • Aortic remodelling • Thoracic endovascular aortic repair
INTRODUCTION Uncomplicated chronic Stanford type B aortic dissection does not require intervention, and medical treatment with blood pressure control is normally indicated unless the condition is associated with aortic dilatation [1]. When the dissected aorta becomes dilated, intervention to the aneurysmal portion of the aorta is required. Conventional open surgery for chronic type B aortic dissection is still associated with high risk, with a reported in-hospital mortality of 4.5–18.8% [2, 3]. On the other hand, the advantage of thoracic endovascular aortic repair (TEVAR) over conventional open surgery in treating descending aortic aneurysm, especially in high-risk patients, has been well demonstrated [4]. Therefore, it is natural that † Presented at the 27th Annual Meeting of the European Association for CardioThoracic Surgery, Vienna, Austria, 5–9 October 2013.
there has been a drive to expand the indication of TEVAR to aortic dissection so that uncomplicated type B aortic dissection can be treated with stent grafting before the aorta becomes dilated and increases the risk of rupture. The goal of TEVAR here includes not only the improvement of survival but also the prevention of aortic dilatation to save patients from future high-risk surgery. In the INvestigation of STEnt Grafts in Aortic Dissection (INSTEAD) trial, which was conducted to compare TEVAR with medical treatment for stable type B aortic dissection at least 2 weeks after onset, no significant differences were observed in 2-year survival, aorta-related deaths or freedom from late aortic operation [5]. However, at the same time, the study proved that TEVAR had a positive impact on remodelling of the dissected descending aorta [6, 7]. The purpose of this study was to evaluate early and late outcomes after TEVAR for non-acute Stanford type B aortic dissection at our institution and to identify key success factors.
© The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
AORTIC SURGERY
ORIGINAL ARTICLE
European Journal of Cardio-Thoracic Surgery (2014) 1–6 doi:10.1093/ejcts/ezu012
2
T. Kitamura et al. / European Journal of Cardio-Thoracic Surgery
PATIENTS AND METHODS From 1998 to 2012, TEVAR was performed in 145 patients at Kitasato University Hospital. Among these, 82 were aortic dissection; 53 of which were non-acute cases. These 53 cases were selected for detailed analysis and were included in this study. Patients operated on within 14 days after onset, emergency patients with rupture or malperfusion, and Stanford type A dissection patients with prior proximal aortic surgery were excluded from this study. The patient medical records were retrospectively reviewed for preoperative demographics, diagnosis and operative findings including the length and size of the stent graft. The stent grafts used were handmade or Matsui-Kitamura (M-K) stent grafts [8]. Handmade stent grafts consisted of electrostatic-welded piano wire stents ( Japan Piano Wire Industry Standard no. 304) covered with woven Dacron grafts (Kitamura MFG Co., Ltd. Ishikawa, Japan). Of note, 0.4 mm stent wires were used for devices sized 30 mm or smaller, 0.45 mm for those sized 32–38 mm and 0.5 mm for those sized 40 mm or larger. Matsui-Kitamura stent grafts are made of a braided framework of a nitinol wire, which is covered with a woven polyester tube graft. Technical success was defined as the elimination of antegrade perfusion of the false lumen on the completion aortogram acquired before leaving the operating theatre. All of the patients underwent computed tomography (CT) angiography preoperatively, within 10 days after the procedure and in the late period (5.0 ± 4.3 years). CT images were used for the measurement of the true lumen diameter, aortic diameter and the site of the initial entry. Remodelling of the descending aorta was defined as thrombosis of the false lumen of the descending aorta without enlargement of the descending aortic diameter.
Statistical methods Stata 11.2 Special Edition (StataCorp, College Station, TX, USA) was used to perform statistical analysis. Continuous variables were presented as means ± standard deviations (SDs). Student’s t-test was used to analyse differences in continuous variables. Fisher’s exact test was used to compare categorical variables. Kaplan–Meier analysis was used to analyse survival. To examine the best models to predict the occurrence of thrombosis of the descending false lumen and remodelling of the descending aorta, logistic regression models were built with thrombosis/remodelling being a dependent variable. As explanatory variables, age, sex, site of the entry, maximum aortic diameter, origin of the abdominal branches, stent graft length and others were entered and removed at the level of P = 0.05 to produce the best predictive models. All P-values were two-sided, and differences were considered statistically significant for P < 0.05.
RESULTS Forty-five patients underwent 53 TEVAR operations for non-acute Stanford type B aortic dissection during the study period. Table 1 shows the operative data of the total of 53 TEVAR procedures. The false lumen of the aortic dissection was patent in 34 cases, and a thrombosed false lumen with an ulcer-like projection (ULP) was observed in 19 cases. The technical success rate was 100% and no early mortality was observed.
Table 1: (n = 53)
Description of the subjects: operative data
Male/female Marfan syndrome Age (years ± SD (range)) Interval from the onset of acute aortic dissection Prior TEVAR Haemodialysis Approach Femoral artery Iliac artery Abdominal aorta Pathology Patent false lumen ULP Mortality Morbidity Stroke New dialysis Prolonged ventilation Arterio-venous fistula Groin haematoma Paraplegia
50/3 0 55.5 ± 13.1 (23–79) 1.4 ± 2.8 years (15 days—15 years) 11 4 50 2 1 34 19 0 2 (3.8%) 1 1 1 1 0
The 45 patients were followed up for 7.5 ± 3.9 years (1.1–14.9 years), and the follow-up was 100% complete. Actuarial survival rates after the initial TEVAR were 100, 86 and 63%; rates of freedom from reintervention were 87, 73 and 59% and rates of freedom from open aortic surgery were 89, 84 and 73% at 1, 5 and 10 years, respectively (Fig. 1). There were 15 late deaths; 2 were caused by rupture of the dilated aorta, 2 by operative deaths following an open aortic surgery, 5 by cancer, 3 by pneumonia, 2 by stroke and 1 by gastrointestinal bleeding. Of the 34 patients who had patent false lumens before TEVAR, the primary entry was present in the outer curvature of the distal aortic arch in 14 patients, in the inner curvature of the distal aortic arch in 11 patients and in the descending aortic wall in 9 patients. The preoperative maximum aortic diameter was 44 ± 11 mm and the length of the stent graft was 115 ± 41 mm. The primary entry in the outer curvature was not significantly correlated with a larger aortic diameter (P = 0.082). Twelve patients had all 4 abdominal branches arising from the true lumen, 20 patients had 3 of the abdominal branches arising from the true lumen and 2 patients had 2 branches from the true lumen. Twenty-five patients of 34 (74%) had their descending false lumen thrombosed, 16 (47%) had remodelling of the descending aorta (Figs 2 and 3) and 4 (12%) had complete obliteration of the false lumen of the entire aorta during the study period. By bivariate analysis, it was found that absence of the primary entry at the outer curvature of the distal aortic arch was significantly associated with thrombosis of the descending false lumen, that a smaller aortic diameter was more likely to result in remodelling of the descending aorta and that all four abdominal branches arising from the true lumen was more likely to proceed to the obliteration of the false lumen of the entire aorta (Table 2). Stent graft length was not significantly correlated with late thrombosis of the false lumen. A combination of absence of the primary entry at the outer curvature and age is the best predictive model for thrombosis, while only the single variable, maximum aortic diameter, best predicts remodelling (Table 3). Since all cases with complete obliteration of the false lumen of the entire aorta had all four abdominal
3
AORTIC SURGERY
T. Kitamura et al. / European Journal of Cardio-Thoracic Surgery
Figure 2: Successful remodelling of the dissected descending aorta after TEVAR. (A) Preoperative CT image and (B) 2 years after TEVAR.
DISCUSSION
Figure 1: Kaplan–Meier analysis. (A) Actuarial survival after the initial TEVAR, (B) freedom from reintervention and (C) freedom from open aortic surgery.
branches arising from the true lumen, logistic regression analysis was not performed. Instead, the odds ratio was calculated [9]. For patients with all four abdominal branches arising from the true lumen, the odds of obliteration of the false lumen of the entire aorta are 2.95 times as large as the odds for those with one or more abdominal branches arising from the false lumen to produce obliteration (P = 0.011 by Fisher’s exact test). Analyses using a maximum aortic diameter of ≤35 vs >35 mm as a dichotomous variable resulted in more or less the same results. Of the 19 ULP patients, 5 (26%) developed recurrent ULPs during the study period, and 2 of the 5 were on haemodialysis (Fig. 4). Although maximum aortic diameter and preoperative haemodialysis were not significant factors in univariate analysis, and the statistical calculation was unreliable in multivariate logistic regression analysis, they did correlate with ULP recurrence.
A certain percentage of patients with uncomplicated Stanford type B aortic dissection develop late dilatation of the dissected descending and thoracoabdominal aorta because of a patent primary entry [10]. Consequently, patients with chronic type B aortic dissection often have to undergo graft replacement of the descending aorta followed by that of the thoracoabdominal aorta. If the entire descending and thoracoabdominal aorta is dilated, extensive replacement of the thoracoabdominal aorta is required in the first place. Surgery for chronic type B aortic dissection is associated with considerable risk [11]. According to the 2010 annual report by the Japanese Association for Thoracic Surgery, in-hospital mortality rate due to surgery for dissecting aortic aneurysm was 4.5% for replacement of the descending aorta and 18.8% for replacement of the thoracoabdominal aorta [2]. On the other hand, the advantage of TEVAR over conventional open surgery in treating descending aortic aneurysm has been well demonstrated [12], and TEVAR has been widely accepted as the optimal therapy for acute, complicated Stanford type B aortic dissection [13]. The goal of TEVAR for chronic type B aortic dissection includes not only improvement of prognosis but also prevention of aortic dilatation to save the patient from future surgical intervention. As shown in our study, aortic dissection heals after TEVAR in some cases, and there are even more cases in which remodelling of the descending aorta occurs, saving the patient from replacement of the descending aorta [6, 7]. The first step in healing of chronic type B aortic dissection with a patent false lumen is thrombosis of the descending false lumen.
4
T. Kitamura et al. / European Journal of Cardio-Thoracic Surgery
Figure 3: Thrombosis of the descending false lumen after TEVAR without remodelling. (A) Preoperative CT image, (B) 2 years after TEVAR and (C) 6 years after TEVAR. This patient underwent graft replacement of the thoracoabdominal aorta.
In our series, the primary entry at the outer curvature of the distal aortic arch was associated with a lower chance of thrombosis of the descending false lumen. This might be attributed to malapposition of the stent graft because of the acute angle of the aortic arch, to late distal dislodging of the outer circumference of the stent graft because of the expansion of the true lumen or to change in the aortic arch geometry after TEVAR [14]. The latest generation device may improve malapposition and increase the chance for descending false lumen thrombosis [15]. Progress in new techniques, including arch debranching [16] or the chimney graft technique [17], may result in better landing of the prosthesis, contributing to an improved thrombosis rate even for patients
with primary entry at the outer curvature of the distal aortic arch. However, the high stroke rate related to these new techniques remains a concern. Primary entry at the inner curvature of the distal aortic arch may be a good indication for early intervention with TEVAR because it has been reported that the primary entry at the inner curvature of the distal arch was associated with a higher risk of complication in acute type B aortic dissection [18, 19]. Successful thrombosis of the descending false lumen is not always followed by remodelling of the descending aorta. In the present study, 74% of the patients with a patent false lumen had their descending false lumen thrombosed after TEVAR but only 47% achieved remodelling of the descending aorta. A large aortic diameter had a negative impact on late descending remodelling. It has been reported that the rate at which the true lumen volume increases after TEVAR is not parallel to that of the reduction in false lumen volume; true lumen expansion plateaus 12 months after TEVAR, whereas false lumen volume reduction gradually continues beyond 12 months [20]. It is speculated that the false lumen with a large volume associated with a dissecting aneurysm having a large diameter would not yet be depressurized when the true lumen expansion plateaus, resulting in the failure of descending remodelling. Complete obliteration of the false lumen on top of descending remodelling, in other words, healing of the dissection, was achieved in only 12% of the cases with a patent false lumen in our series. However, this result does not mean that TEVAR is not an effective treatment option for chronic type B aortic dissection because remodelling of the descending aorta brought in by TEVAR may be able to prevent open descending aortic replacement and may delay the expansion of the thoracoabdominal aorta. Fenestrated endografting to cover the whole aorta may be realized in future, but the risk of spinal cord ischaemia remains a concern. TEVAR is certainly a promising option for chronic type B aortic dissection, but it should not be performed blindly, considering potential early and late complications, including retrograde type A aortic dissection, endoleaks, spinal cord ischaemia and intimal flap injury due to prosthesis [21–23]. As the number of TEVAR procedures has increased, an increasing number of open aortic operations with prior TEVAR have recently been performed [24, 25]. Stent grafts within the aorta can cause a nuisance at open reintervention [24], sometimes involving a non-negligible risk of flap injury during explanting and a risk of spinal cord ischaemia, which is particularly observed in emergency settings. Therefore, initial patient selection is very important. TEVAR can be used to treat chronic localized aortic dissection with a ULP as safely as an atherosclerotic descending aortic aneurysm and with the same reliable long-term outcomes. However, postoperative regular imaging is required to detect recurrent ULPs, especially in patients on haemodialysis. For chronic type B aortic dissection with a patent false lumen, absence of the primary entry at the outer curvature of the distal aortic arch, smaller aortic diameter and absence of the abdominal branches arising from the false lumen would be the key success factors. It is hoped that subgroup analysis will be performed regarding TEVAR vs medical treatment in this group of patients in a large randomized study. This study had some limitations. This was a retrospective study with a small number of patients, and the stent grafts used were mostly handmade, which meant that the quality may have varied. However, because of their non-uniformity, various prostheses with different sizes and lengths were included, enabling analysis of whether the size and length of the prosthesis were associated with success.
T. Kitamura et al. / European Journal of Cardio-Thoracic Surgery
Table 2: Factors for thrombosis of the descending false lumen, remodelling of the descending aorta and complete obliteration of the false lumen of the entire aorta for the cases with a patent false lumen (n = 34) Thrombosis
Sex (Female/male) Entry (inner curvature and descending/ outer curvature) All abdominal branches from the true lumen (no/yes) Stent graft length (
