Original Article
Iatrogenic intraoperative type A aortic dissection following cardiac surgery
Asian Cardiovascular & Thoracic Annals 2015, Vol. 23(1) 31–35 ß The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0218492314531140 aan.sagepub.com
Pradeep Narayan, Gianni D Angelini and Alan J Bryan
Abstract Background: An increase in the incidence of intraoperative aortic dissection has been reported recently, attributed to the increasingly elderly patient population undergoing cardiac surgery and more off-pump coronary artery bypass. We performed this study to examine current trends, identify risk factors for iatrogenic dissection, and compare iatrogenic intraoperative aortic dissection with spontaneous aortic dissection. Methods: The 15,144 consecutive patients who underwent cardiac surgery from April 1999 to April 2011 were studied retrospectively on data collected prospectively. Results: Iatrogenic type A aortic dissection following cardiac surgery was diagnosed intraoperatively in 7 (0.04%) patients. Of the 4784 patients who had off-pump coronary artery bypass, only 2 (0.04%) developed iatrogenic intraoperative aortic dissection. Patients in the iatrogenic aortic dissection group were older by a decade (median age 72 vs. 62 years, p ¼ 0.01). The cannulation site in conventional coronary artery bypass grafting and injury by the side-biting clamp in off-pump coronary artery bypass were the most common causes of dissection. Atheromatous disease was identified at the site of cannulation in 5 (71.4%) of the 7 cases. Conclusions: Intraoperative aortic dissection remains a rare and unpredictable complication of cardiac surgery, with worse outcomes than spontaneous aortic dissection. Increased age and atheromatous disease at the site of cannulation are significant risk factors for iatrogenic dissection. In this series, off-pump coronary artery bypass did not appear to be a risk factor for iatrogenic aortic dissection.
Keywords Aged, Aneurysm, dissecting, Aortic aneurysm, Coronary artery bypass, Iatrogenic disease, Intraoperative complications
Introduction Iatrogenic intraoperative aortic dissection in patients undergoing cardiac surgery is a rare but potentially fatal condition. The incidence of intraoperative aortic dissection over the last 3 decades has remained largely unchanged, and according to published data, remains at 0.06%–0.23%.1–6 Recently, an increase in the incidence of intraoperative aortic dissection has been reported secondary to an increase in off-pump surgery.7 According to the International Registry of Acute Aortic Dissection data, type A aortic dissection secondary to cardiac surgery constituted 1.1% (18/723) of all type A aortic dissections.8 The overall mortality in type A intraoperative aortic dissection is reported to be 24% to 43%.9–11 However, it has been shown that if the condition is diagnosed intraoperatively, mortality is considerably lower (20%–33%) compared to when it
is diagnosed in the postoperative period when the mortality is as high as 50%–78%.4 Studies have shown that operative mortality in iatrogenic type A intraoperative aortic dissection is higher but not significantly different from the mortality of spontaneous aortic dissection.4 The mortality of spontaneous aortic dissection has been reported as 15%–30%.12–16 To examine current trends in the occurrence of intraoperative aortic dissection, and to assess whether the introduction of offpump coronary artery bypass (OPCAB) has led to an increase in the incidence of intraoperative aortic
Bristol Heart Institute, Bristol Royal Infirmary, Bristol, UK Corresponding author: Pradeep Narayan, Bristol Heart Institute, Bristol Royal Infirmary, Bristol BS2 8HW, UK. Email: [email protected]
Downloaded from aan.sagepub.com at KAI NAN UNIV on March 2, 2015
32
Asian Cardiovascular & Thoracic Annals 23(1)
dissection, we carried out this retrospective study on our cardiac surgical database. We also aimed to compare the outcomes of intraoperative aortic dissections and spontaneous type A aortic dissections, and sought to assess potential risk factors responsible for iatrogenic intraoperative aortic dissection.
Patients and methods Data was collected prospectively on all patients undergoing cardiac surgery, and entered into a database (Patient Analysis and Tracking System; Dendrite Clinical Systems, London, UK). A series of 15,144 consecutive patients undergoing cardiac surgery from April 1999 to April 2011 was studied. Patients who developed type A intraoperative aortic dissection were identified and compared with those who had spontaneous aortic dissection. Patients with iatrogenic aortic dissection presenting days or weeks after the operation or following percutaneous interventions were excluded from the study. Only patients in whom a tube graft was used to replace the aorta were included; patch repairs for localized injury were excluded. Definitions of operative priority, premorbid conditions, and postoperative complications were those used in the National Adult Cardiac Database and accepted by the Society for Cardiothoracic Surgery in Great Britain & Ireland (www.scts.org). In cases where cardiopulmonary bypass (CPB) was employed, cannulation of the ascending aorta was performed in the majority. While there were individual differences, there was a broad consensus on the techniques of cannulation. Two 4/0 pursestring sutures were placed on the distal ascending aorta in a circular fashion after finger palpation to identify a plaque-free area. The adventitia was dissected over the area, and an oblique incision was made after ensuring that the systemic pressure was within acceptable limits. The antegrade cardioplegia line was inserted after dissecting the adventitia, and a 4/0 pursestring suture was placed in a
triangular or quadrangular fashion. Digital palpation alone was used to identify plaques in the ascending aorta. Proximal anastomosis was performed in all cases, using a side-biting clamp. In OPCAB, care was taken that the side-biting clamp was applied only when the systemic pressure was around 100 mm Hg. Details of our off-pump technique have been reported previously.17 Detailed operative, perfusion, and anesthetic techniques for repair of aortic dissections at our institution have also been described previously in detail.18 Baseline and operative characteristics were compared using the chi-square or Fisher’s exact test (categorical variables) or the Wilcoxon rank sum test (continuous variables). All analyses were carried out using Stat-View 2002 software package (SAS Institute Inc., Cary, NC, USA).
Results Iatrogenic type A aortic dissection was diagnosed intraoperatively in 7 (0.04%) of the 15,144 patients. The causes of intraoperative dissection and initial operative strategy are given in Table 1. Baseline characteristics were broadly similar in the 2 groups, but the patients who suffered aortic dissection were older by a decade (median age 72 vs. 62 years, p ¼ 0.01; Table 2). No arch replacements were necessary in the iatrogenic group, whereas 31 (15.5%) arch replacements were performed in the group presenting with spontaneous dissection. Reoperation rates for bleeding, the incidence of neurological complications, and length of hospital stay were similar in both groups; however, the intensive care unit stay was significantly longer in patients with iatrogenic dissection (11 vs. 3 days, p ¼ 0.03; Table 2). Of the 4784 OPCAB cases, only 2 (0.04%) developed intraoperative aortic dissection. Comparing off-pump with on-pump procedures, OPCAB was not a risk factor for aortic dissection in our study population (odds ratio ¼ 1.15, 95% confidence interval: 0.22–5.90). In the group in which CPB was utilized in the initial
Table 1. Sites of intraoperative iatrogenic aortic dissection. Initial operation
Site of dissection
CABG (on-pump with cardioplegia) CABG (off-pump converted to on-pump with cardioplegia) CABG þ MVR (on-pump with cardioplegia) CABG (on-pump without cardioplegia) CABG (off-pump) CABG (off-pump Redo mitral valve surgery via right thoracotomy (previous CABG)
Cardioplegia cannulation site Ascending aortic cannulation site Ascending aortic cannulation site Ascending aortic cannulation site Side-biting clamp for proximal anastomosis Side-biting clamp for proximal anastomosis Femoral cannulation
CABG: coronary artery bypass grafting; MVR: mitral valve replacement.
Downloaded from aan.sagepub.com at KAI NAN UNIV on March 2, 2015
Narayan et al.
33
operation, the cannulation site was found to be the most common etiology for iatrogenic intraoperative aortic dissection (4/7 patients, 57.14%). The ascending aorta cannulation site was involved in 3 cases, and the femoral cannulation site led to retrograde type A aortic dissection in one patient; in this case, the femoral artery was cannulated because it was a reoperative case with patent grafts, and it was decided to carry out mitral valve replacement through a right thoracotomy incision. Aortic dissection occurred at the cardioplegia site in one patient. Atheromatous disease was identified at the site of cannulation in all 5 cases. Injury by the side-biting clamp was the etiology in both cases of intraoperative aortic dissection in the OPCAB group.
Discussion The findings of our study were: firstly the incidence of iatrogenic intraoperative aortic dissection was low and within the range reported by other studies. Secondly, the mortality in intraoperative aortic dissection was around 33% which was higher than our mortality for spontaneous type A aortic dissection (17.5%) during the same period, although this did not reach statistical significance. Finally, we did not see an increase in the incidence of iatrogenic dissection secondary to OPCAB, as reported in previous studies. It is also evident from our study that the patient profile of those suffering intraoperative aortic dissection was different from those suffering spontaneous type A aortic dissection, with the patients in the former category being significantly older, by almost a decade. This is in keeping with the International Registry of Acute Aortic Dissection data which also showed that patients with iatrogenic dissection were significantly older than those with spontaneous dissection.8 This could in part explain the higher mortality associated with intraoperative aortic dissection because age has been shown to be an
independent determinant of mortality in aortic dissection repair.19,20 It is important to note that these patients had another procedure before they underwent operative repair of aortic dissection, further increasing the operative risk. However, this is partly counterbalanced by the fact that those presenting with spontaneous dissection often have evidence of ongoing organ malperfusion which is a determinant of poor outcome.19 Moreover, there was a significant number of patients in the spontaneous aortic dissection group who had concomitant aortic arch replacement (n ¼ 31) in contrast to none in the iatrogenic group. We excluded patients whose dissection was diagnosed postoperatively because they constitute a different subset of patients with poorer outcomes, and there is a dramatic rise in mortality if the dissection is not diagnosed intraoperatively. While the mortality ranges from 20% to 33% when diagnosis is made intraoperatively, it increases to up to 78% when diagnosed late.1 The use of transesophageal echocardiography intraoperatively for the diagnosis and management of this complication is imperative for this reason alone. Transesophageal echocardiography is especially useful in diagnosing tears in the posterior wall, and several centers have reported an increase in identification of this complication and better outcomes when transesophageal echocardiography is used more liberally for assessment of intraoperative issues in cardiac surgery.6 Transesophageal echocardiography is also extremely useful to differentiate between dissection and subadventitial hematoma, and its overall utility in cardiac surgery has now been well established.5 In almost all cases where iatrogenic dissection occurred secondary to cannulation, an atheromatous plaque was encountered. This is again in keeping with the experience in other reports where a strong association between iatrogenic dissection and atheromatous disease was reported.5 Thus increased age and the presence of atheromatous
Table 2. Univariate comparison of intraoperative and spontaneous dissection. Variable
Intraoperative dissection (n ¼ 7)
Spontaneous dissection (n ¼ 200)
p value
Age (years) [Inter Quartile range] Circulatory arrest time (min) Bypass time (min) [range] Crossclamp time (min) [range] Postoperative renal failure ICU stay (days) [range] Postoperative hospital stay (days) [range] Hospital death
72 25 193 75 5 11 19 2
60 33 176 80 29 3 11 35
0.01 0.15 0.57 0.48
Iatrogenic intraoperative type A aortic dissection following cardiac surgery
Asian Cardiovascular & Thoracic Annals 2015, Vol. 23(1) 31–35 ß The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0218492314531140 aan.sagepub.com
Pradeep Narayan, Gianni D Angelini and Alan J Bryan
Abstract Background: An increase in the incidence of intraoperative aortic dissection has been reported recently, attributed to the increasingly elderly patient population undergoing cardiac surgery and more off-pump coronary artery bypass. We performed this study to examine current trends, identify risk factors for iatrogenic dissection, and compare iatrogenic intraoperative aortic dissection with spontaneous aortic dissection. Methods: The 15,144 consecutive patients who underwent cardiac surgery from April 1999 to April 2011 were studied retrospectively on data collected prospectively. Results: Iatrogenic type A aortic dissection following cardiac surgery was diagnosed intraoperatively in 7 (0.04%) patients. Of the 4784 patients who had off-pump coronary artery bypass, only 2 (0.04%) developed iatrogenic intraoperative aortic dissection. Patients in the iatrogenic aortic dissection group were older by a decade (median age 72 vs. 62 years, p ¼ 0.01). The cannulation site in conventional coronary artery bypass grafting and injury by the side-biting clamp in off-pump coronary artery bypass were the most common causes of dissection. Atheromatous disease was identified at the site of cannulation in 5 (71.4%) of the 7 cases. Conclusions: Intraoperative aortic dissection remains a rare and unpredictable complication of cardiac surgery, with worse outcomes than spontaneous aortic dissection. Increased age and atheromatous disease at the site of cannulation are significant risk factors for iatrogenic dissection. In this series, off-pump coronary artery bypass did not appear to be a risk factor for iatrogenic aortic dissection.
Keywords Aged, Aneurysm, dissecting, Aortic aneurysm, Coronary artery bypass, Iatrogenic disease, Intraoperative complications
Introduction Iatrogenic intraoperative aortic dissection in patients undergoing cardiac surgery is a rare but potentially fatal condition. The incidence of intraoperative aortic dissection over the last 3 decades has remained largely unchanged, and according to published data, remains at 0.06%–0.23%.1–6 Recently, an increase in the incidence of intraoperative aortic dissection has been reported secondary to an increase in off-pump surgery.7 According to the International Registry of Acute Aortic Dissection data, type A aortic dissection secondary to cardiac surgery constituted 1.1% (18/723) of all type A aortic dissections.8 The overall mortality in type A intraoperative aortic dissection is reported to be 24% to 43%.9–11 However, it has been shown that if the condition is diagnosed intraoperatively, mortality is considerably lower (20%–33%) compared to when it
is diagnosed in the postoperative period when the mortality is as high as 50%–78%.4 Studies have shown that operative mortality in iatrogenic type A intraoperative aortic dissection is higher but not significantly different from the mortality of spontaneous aortic dissection.4 The mortality of spontaneous aortic dissection has been reported as 15%–30%.12–16 To examine current trends in the occurrence of intraoperative aortic dissection, and to assess whether the introduction of offpump coronary artery bypass (OPCAB) has led to an increase in the incidence of intraoperative aortic
Bristol Heart Institute, Bristol Royal Infirmary, Bristol, UK Corresponding author: Pradeep Narayan, Bristol Heart Institute, Bristol Royal Infirmary, Bristol BS2 8HW, UK. Email: [email protected]
Downloaded from aan.sagepub.com at KAI NAN UNIV on March 2, 2015
32
Asian Cardiovascular & Thoracic Annals 23(1)
dissection, we carried out this retrospective study on our cardiac surgical database. We also aimed to compare the outcomes of intraoperative aortic dissections and spontaneous type A aortic dissections, and sought to assess potential risk factors responsible for iatrogenic intraoperative aortic dissection.
Patients and methods Data was collected prospectively on all patients undergoing cardiac surgery, and entered into a database (Patient Analysis and Tracking System; Dendrite Clinical Systems, London, UK). A series of 15,144 consecutive patients undergoing cardiac surgery from April 1999 to April 2011 was studied. Patients who developed type A intraoperative aortic dissection were identified and compared with those who had spontaneous aortic dissection. Patients with iatrogenic aortic dissection presenting days or weeks after the operation or following percutaneous interventions were excluded from the study. Only patients in whom a tube graft was used to replace the aorta were included; patch repairs for localized injury were excluded. Definitions of operative priority, premorbid conditions, and postoperative complications were those used in the National Adult Cardiac Database and accepted by the Society for Cardiothoracic Surgery in Great Britain & Ireland (www.scts.org). In cases where cardiopulmonary bypass (CPB) was employed, cannulation of the ascending aorta was performed in the majority. While there were individual differences, there was a broad consensus on the techniques of cannulation. Two 4/0 pursestring sutures were placed on the distal ascending aorta in a circular fashion after finger palpation to identify a plaque-free area. The adventitia was dissected over the area, and an oblique incision was made after ensuring that the systemic pressure was within acceptable limits. The antegrade cardioplegia line was inserted after dissecting the adventitia, and a 4/0 pursestring suture was placed in a
triangular or quadrangular fashion. Digital palpation alone was used to identify plaques in the ascending aorta. Proximal anastomosis was performed in all cases, using a side-biting clamp. In OPCAB, care was taken that the side-biting clamp was applied only when the systemic pressure was around 100 mm Hg. Details of our off-pump technique have been reported previously.17 Detailed operative, perfusion, and anesthetic techniques for repair of aortic dissections at our institution have also been described previously in detail.18 Baseline and operative characteristics were compared using the chi-square or Fisher’s exact test (categorical variables) or the Wilcoxon rank sum test (continuous variables). All analyses were carried out using Stat-View 2002 software package (SAS Institute Inc., Cary, NC, USA).
Results Iatrogenic type A aortic dissection was diagnosed intraoperatively in 7 (0.04%) of the 15,144 patients. The causes of intraoperative dissection and initial operative strategy are given in Table 1. Baseline characteristics were broadly similar in the 2 groups, but the patients who suffered aortic dissection were older by a decade (median age 72 vs. 62 years, p ¼ 0.01; Table 2). No arch replacements were necessary in the iatrogenic group, whereas 31 (15.5%) arch replacements were performed in the group presenting with spontaneous dissection. Reoperation rates for bleeding, the incidence of neurological complications, and length of hospital stay were similar in both groups; however, the intensive care unit stay was significantly longer in patients with iatrogenic dissection (11 vs. 3 days, p ¼ 0.03; Table 2). Of the 4784 OPCAB cases, only 2 (0.04%) developed intraoperative aortic dissection. Comparing off-pump with on-pump procedures, OPCAB was not a risk factor for aortic dissection in our study population (odds ratio ¼ 1.15, 95% confidence interval: 0.22–5.90). In the group in which CPB was utilized in the initial
Table 1. Sites of intraoperative iatrogenic aortic dissection. Initial operation
Site of dissection
CABG (on-pump with cardioplegia) CABG (off-pump converted to on-pump with cardioplegia) CABG þ MVR (on-pump with cardioplegia) CABG (on-pump without cardioplegia) CABG (off-pump) CABG (off-pump Redo mitral valve surgery via right thoracotomy (previous CABG)
Cardioplegia cannulation site Ascending aortic cannulation site Ascending aortic cannulation site Ascending aortic cannulation site Side-biting clamp for proximal anastomosis Side-biting clamp for proximal anastomosis Femoral cannulation
CABG: coronary artery bypass grafting; MVR: mitral valve replacement.
Downloaded from aan.sagepub.com at KAI NAN UNIV on March 2, 2015
Narayan et al.
33
operation, the cannulation site was found to be the most common etiology for iatrogenic intraoperative aortic dissection (4/7 patients, 57.14%). The ascending aorta cannulation site was involved in 3 cases, and the femoral cannulation site led to retrograde type A aortic dissection in one patient; in this case, the femoral artery was cannulated because it was a reoperative case with patent grafts, and it was decided to carry out mitral valve replacement through a right thoracotomy incision. Aortic dissection occurred at the cardioplegia site in one patient. Atheromatous disease was identified at the site of cannulation in all 5 cases. Injury by the side-biting clamp was the etiology in both cases of intraoperative aortic dissection in the OPCAB group.
Discussion The findings of our study were: firstly the incidence of iatrogenic intraoperative aortic dissection was low and within the range reported by other studies. Secondly, the mortality in intraoperative aortic dissection was around 33% which was higher than our mortality for spontaneous type A aortic dissection (17.5%) during the same period, although this did not reach statistical significance. Finally, we did not see an increase in the incidence of iatrogenic dissection secondary to OPCAB, as reported in previous studies. It is also evident from our study that the patient profile of those suffering intraoperative aortic dissection was different from those suffering spontaneous type A aortic dissection, with the patients in the former category being significantly older, by almost a decade. This is in keeping with the International Registry of Acute Aortic Dissection data which also showed that patients with iatrogenic dissection were significantly older than those with spontaneous dissection.8 This could in part explain the higher mortality associated with intraoperative aortic dissection because age has been shown to be an
independent determinant of mortality in aortic dissection repair.19,20 It is important to note that these patients had another procedure before they underwent operative repair of aortic dissection, further increasing the operative risk. However, this is partly counterbalanced by the fact that those presenting with spontaneous dissection often have evidence of ongoing organ malperfusion which is a determinant of poor outcome.19 Moreover, there was a significant number of patients in the spontaneous aortic dissection group who had concomitant aortic arch replacement (n ¼ 31) in contrast to none in the iatrogenic group. We excluded patients whose dissection was diagnosed postoperatively because they constitute a different subset of patients with poorer outcomes, and there is a dramatic rise in mortality if the dissection is not diagnosed intraoperatively. While the mortality ranges from 20% to 33% when diagnosis is made intraoperatively, it increases to up to 78% when diagnosed late.1 The use of transesophageal echocardiography intraoperatively for the diagnosis and management of this complication is imperative for this reason alone. Transesophageal echocardiography is especially useful in diagnosing tears in the posterior wall, and several centers have reported an increase in identification of this complication and better outcomes when transesophageal echocardiography is used more liberally for assessment of intraoperative issues in cardiac surgery.6 Transesophageal echocardiography is also extremely useful to differentiate between dissection and subadventitial hematoma, and its overall utility in cardiac surgery has now been well established.5 In almost all cases where iatrogenic dissection occurred secondary to cannulation, an atheromatous plaque was encountered. This is again in keeping with the experience in other reports where a strong association between iatrogenic dissection and atheromatous disease was reported.5 Thus increased age and the presence of atheromatous
Table 2. Univariate comparison of intraoperative and spontaneous dissection. Variable
Intraoperative dissection (n ¼ 7)
Spontaneous dissection (n ¼ 200)
p value
Age (years) [Inter Quartile range] Circulatory arrest time (min) Bypass time (min) [range] Crossclamp time (min) [range] Postoperative renal failure ICU stay (days) [range] Postoperative hospital stay (days) [range] Hospital death
72 25 193 75 5 11 19 2
60 33 176 80 29 3 11 35
0.01 0.15 0.57 0.48
