Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Late complications after prosthetic replacement of the ascending aorta: what did we learn from routine magnetic resonance imaging follow-up?

Late complications after prosthetic replacement of the ascending aorta: what did we learn from... Abstract Objective: Advances in surgical technique have improved early survival after surgery of the ascending aorta. However, follow-up data document serious late complications, mainly evolutive peri-prosthetic false aneurysms. Magnetic resonance imaging (MRI) has proved to be highly effective for monitoring these complications. This study evaluates 10 years of experience with routine MRI for follow-up. Methods: Since January 1988, 114 patients with replacement of the ascending aorta either for type A acute dissection (group I, 45 patients) or aneurysms (group II, 69 patients) were followed up with annual MRI. Prosthetic replacement was either limited to supra-coronary ascending aorta (45%, 51/114) or extended to the aortic root and/or the aortic arch (55%). Biological glue was always utilized. MRI focused on peri-prosthetic haematoma, analyzing signal intensity changes and volume augmentation for early detection of false aneurysms, and on persistent residual dissection with or without evolutive aortic aneurysm distant to the prosthesis. Results: Peri-prosthetic hematomas were almost equally found in both groups (26 (58%) in group I and 42 (61%) in group II) and were detected within the first year. Peri-prosthetic false aneurysms developed in 15 patients (group I, seven; group II, eight) as a complication of pre-existing hematomas and were indicated for elective reoperation. Forty-three (96%) of patients in group I had persistent residual dissection. Five patients in group I and two in group II needed reoperation for evolutive aortic aneurysm. In total, 22 of 114 (19%) patients were reoperated on during follow-up (12 (27%) in group I and ten (15%) in group II). Operative mortality was 13% (3/22). Freedom from reoperation at 1 year/5 years was: group I, 93%/84%; group II, 98%/88%. Conclusion: Peri-prosthetic haematoma occurs equally after aneurysm or dissection repairs and is a pre-existing condition for peri-prosthetic false aneurysm; biological glue or extended repair do not prevent late complications. Long-term MRI follow-up allows successful elective reoperation for life-threatened but asymptomatic patients. Aorta, Aneurysm, Dissection, Peri-prosthetic haematoma, False aneurysm, Magnetic resonance imaging 1 Introduction Since the first successful repair reported by Cooley and DeBakey in 1956 [1], significant advances resulted in lower operative mortality for elective aneurysm repair as for emergency acute dissection surgery. During the last decade, the total number of patients at risk for development of late complications considerably increased. Therefore, reoperations on the thoracic ascending aorta has gained utmost surgical consideration. However, routine post-operative examination of a repaired thoracic aorta has not been as well emphasized as the evaluation of other common cardiac operation. Our group has previously documented follow-up data showing a high frequency of serious late complications in patients who underwent surgical treatment of acute Stanford type A aortic dissection [2]. We demonstrated that magnetic resonance imaging (MRI) has proved to be highly effective in detection and monitoring of peri-prosthetic haematoma (PPH), peri-prosthetic false aneurysm (PPFA) and evolutive aortic aneurysm (EAA) developing at distant location from the graft attachment. The purpose of this study was to evaluate routine MRI after surgical repair of acute dissection or non-dissecting aortic aneurysm and to assess the frequency of late complications in a larger group of patients who underwent a variety of surgical procedures. 2 Materials and methods Between January 1988 and December 1997, 114 patients who survived replacement of the ascending aorta either in emergency for Stanford type A aortic dissection (Group I; 45 patients), or electively for non-dissecting ascending aorta aneurysms (Group II; 69 patients), were followed up with routine MRI. 2.1 Group I: aortic dissection Forty-five patients survived emergency procedures that occurred within 4 h following their referral to our Centre and within 4 days from the onset of first acute events. There were 36 men and nine women (mean age, 55.9±17.7 years). Three patients had previous coronary surgery (more than 1 year before acute dissection), three had pre-existing ascending aorta aneurysm, and five were presenting Marfan syndrome. Employed surgical procedures are listed in Table 1 . Extension of prosthetic replacement to the aortic arch was more frequent since 1991, when a more radical approach was decided, consisting in a wider use of ‘open’ anastomosis technique that allows aortic arch exploration in search of additional intimal tears. We therefore found 47% (21/45) of patients presenting intimal tears in this aortic location and necessitating an extended repair. In some instances, intimal ruptures at the origin of innominate artery necessitated its transection followed by reimplantation in a separate graft. Table 1 Open in new tabDownload slide Type of procedures employed in group I (45 patients) and group II (69 patients) Table 1 Open in new tabDownload slide Type of procedures employed in group I (45 patients) and group II (69 patients) Five of eight patients who underwent Bentall procedures also had aortic arch surgery, consisting in two hemi-arch with innominate artery reimplantation and three total arch replacement. Bentall procedures were indicated in patients with Marfan disease and in cases with destruction of the sinus of Valsalva or tearing out of coronary ostia. Otherwise, tissue excision was achieved for adequate residual aortic wall remodelling and aortic valve coaptation. GRF (gelatine risorcine formaldehyde) biological glue was utilized in all instances, to obtain adequate proximal and distal stumps and optimal obliteration of the space between the dissected aortic layers, ultimately sutured with 4/0 monofilament after Teflon felt tissue reinforcement. 2.2 Group II: non-dissecting aneurysms In the same period of time, 69 patients survived elective surgery for non-dissecting aneurysm of the ascending aorta. There were 51 men and 18 women (mean age, 55.3±14.6 years). Employed surgical techniques are listed in Table 1. Patients undergoing aortic arch operations alone were excluded from this series. Interposition graft and aortic valve replacement (or re-replacement) were selected for patients with aortic valve disease (or prosthesis). This represented one-third of this group (23/69, 33%), including four patients bearing an old prosthesis. The entire root was replaced in 30 cases (30/69, 44%). The technique of coronary reimplantation varied from the standard technique pre-dating 1994, to the ‘button’ technique (13/30, 43%), associated with complete aortic transection followed by end-to-end anastomosis and using interrupted pledgeted stitches for annular implantation of the composite graft. More recently, the valve-sparing procedure, as described by David and colleagues [3] was performed in five selected young patients presenting moderate annulo-ectasia with preserved valvular structures. Impregnated Dacron grafts and composite grafts (St. Jude Medical, Minneapolis, MN) were utilized in all patients except one having a root replacement using an aortic homograft. In all cases, french glue or fibrin glue (ImmunoCo, Vienna, Austria) were employed as a spray or a simple layer over the suture lines. 2.3 MRI follow-up All patients were included in a routine follow-up program in which MRI was performed once during the first 3 post-operative months, and once a year thereafter. This sequence was possibly modified, depending on the results of previous examinations. A total of 325 MRI examinations were performed in this series (range, 1–8 examinations per patient; mean, 2.9 examinations). All patients had at least two MRI examinations. During follow-up, two patients were switched to computed tomography (CT) scan evaluation due to pacemaker implantation. No patient was lost to follow-up. At the beginning of this study, MRI findings were compared with those from angiography and computed tomography. First angiography, then CT were abandoned early, except for pre-surgical workshop in patients scheduled to undergo reoperation. MRI data were acquired with electrocardiographic gating in the transaxial and oblique sagittal planes parallel to the main axis of the aortic arch. Field of view was 38–40 cm, and section thickness varied from 5 to 7 mm. Whenever an intermediate signal intensity (SI) zone within the aorta or the Dacron graft prosthesis could not be definitely attributed to slow flow and thrombosis, multi-echo sequences were used to assess the SI and assess the blood-flow pattern in the suspected areas. Slow-flowing blood was characterized by SI enhancement on the second echo of the multi-echo sequence. Thrombus was characterized by no change in SI. Fast flowing blood was characterized by a signal void inside or outside the aortic lumen, demonstrating PPFA. After 1992, multi-echo sequences were replaced by rapid imaging with fast low-angle shot (TurboFLASH, Siemens, Erlangen, Germany) sequences after rapid intravenous injection of gadoterate meglumine (Guerbet, Roissy, France). Limited amounts of fibrin and/or small clots are commonly observed around the aortic prosthesis in the early post-operative period. Such an observation did not assess a PPH. On the other hand, a substantial increase in SI after injection of contrast material was attributed to a circulating blood flow around the prosthesis, suggestive of heterogeneous PPH with a high risk for PPFA, whereas unchanged SI was interpreted as non-complicated PPH. All PPH were therefore classified into the following two categories (Cat): Cat1PPH, moderate, homogeneous, with no SI change (Fig. 1) ; Cat2PPH, large, circumference greater than 10 mm, heterogeneous and/or enhanced SI after contrast injection (Fig. 2) . Fig. 1 Open in new tabDownload slide MRI showing a Cat1PPH: limited size, homogeneous, not circumferential, with no SI changes. Fig. 1 Open in new tabDownload slide MRI showing a Cat1PPH: limited size, homogeneous, not circumferential, with no SI changes. Fig. 2 Open in new tabDownload slide MRI showing a Cat2PPH: greater than 15 mm, heterogeneous, non-regressive, with SI changes. Fig. 2 Open in new tabDownload slide MRI showing a Cat2PPH: greater than 15 mm, heterogeneous, non-regressive, with SI changes. The peri-prosthesis anastomotic leaks which fed the PPFA were determined and compared with intraoperative findings (Fig. 3) . In the native aorta distant to the prosthesis, the diameter of the lumen was assessed and residual dissection with blood flow in the false lumen was searched (Fig. 4) . A diameter of 40 mm was the cut-off size for diagnosis of an aneurysm. An EAA was assessed as more than 10-mm increase within a 6-month period. Such a diagnosis needed to be confirmed over another 6-month period with repeated MRI examination before consideration for surgery. Fig. 3 Open in new tabDownload slide MRI showing PPFA with fast-flowing blood around the prosthesis from a leaking coronary suture. Fig. 3 Open in new tabDownload slide MRI showing PPFA with fast-flowing blood around the prosthesis from a leaking coronary suture. Fig. 4 Open in new tabDownload slide MRI showing persistent false lumen in the descending aorta, 8 years after arch repair. Fig. 4 Open in new tabDownload slide MRI showing persistent false lumen in the descending aorta, 8 years after arch repair. For statistical analysis, freedom from reoperation was calculated from the time of first operation to the end point of December 1997 (in months), using the Kaplan–Meier method. Significance was assigned at less than or equal to 5% probability (P-value less than 0.05). 3 Results 3.1 Group I 3.1.1 MRI follow-up MRI detected 26 PPH (26/45, 58%) divided into 12 Cat1PPH (12/45, 27%) which never complicated, and 14 Cat2PPH (14/45, 31%),which degenerated in seven PPFA as indicated in Table 2 . In two patients, PPFA was revealed as a fast progressing PPH within 3 months after surgery, and required rapid surgical management. In four patients also requiring reoperation, PPFA were only diagnosed after the third post-operative year. Cat2PPH could also improve as in four cases (4/14, 29%) showing regression in Cat1PPH. In three other patients presenting Cat2PPH, MRI follow-up showed evidence of circulating blood around the prosthesis without PPH volume augmentation. These patients did not undergo reoperation and were alive and stable at more than 2 years. Five patients presenting with EAA required reoperation, four with EAA distal to the aortic arch and one with an EAA developing in the sinuses of Valsalva. Table 2 Open in new tabDownload slide Post-operative complications detected with MRI in group I (dissections) Table 2 Open in new tabDownload slide Post-operative complications detected with MRI in group I (dissections) In addition to this, we observed non-evolutive aortic aneurysms distant to the prosthesis in six patients. Five of these six complications were demonstrated on MRI obtained during the first post-operative year, and no substantial change in morphology or size has been observed at follow-up up to 108 months. Two of these six patients died eventually, one from chronic renal failure and the other from cardiac failure not related to aortic valve dysfunction. Persistent residual dissection distal to the aortic graft with a patent false channel was almost constant at some degree, a majority of patients revealing various amounts of thrombus within the false channel. Total thrombosis and retraction of the false lumen occurred in only two patients, one in the first post-operative year, and another in the 18th post-operative month despite the presence of residual dissection 1 year earlier. 3.1.2 Reoperations (Tables 4 and 5) Table 4 Open in new tabDownload slide Reoperations for PPFA in group I (patients 1–7) and group II (patients 8–15)a Table 4 Open in new tabDownload slide Reoperations for PPFA in group I (patients 1–7) and group II (patients 8–15)a Table 5 Open in new tabDownload slide Reoperations for EAA in group I (patients 1 to 5) and group II (patients 6 and 7)a Table 5 Open in new tabDownload slide Reoperations for EAA in group I (patients 1 to 5) and group II (patients 6 and 7)a Twelve patients in this group were indicated for elective redo surgery, seven for PPFA and five for EAA. Ten of 12 patients were asymptomatic. Thoracic pain was reported in two early PPFA, necessitating reoperation at 2 months. One-year and 5-year freedom from reoperation for PPFA were 93 and 84%, respectively. Tables 4 and 5 indicate the delay in each case and procedures employed at each reoperation for PPFA and EAA, respectively. One patient (no. 5, Table 4) presented a PPFA in the first post-operative month related to bacterial contamination of the Dacron graft and died 2 days after surgery from septic shock. Another patient (no. 5, Table 5) died during reoperation for an EAA of the arch from massive bleeding and coagulopathy. Operative mortality was 17% (2/12) in this group. Intra-operative findings were always concordant with those at MRI. PPFA were fed by peri-prosthetic leaks coming from proximal graft attachment (seven cases), distal attachment (two cases), or coronary reimplantation (one case). 3.2 Group II 3.2.1 MRI Follow-up Forty-two PPH were observed during MRI follow-up (42/69, 61%), divided into 28 (28/69, 41%) regressive Cat1PPH and 14 (14/69, 20%) Cat2PPH, distributed as indicated in Table 3 . In eight patients, the PPFA occurred as an evolution of a pre-existing Cat2PPH, which could also be regressive as in four patients (4/14, 31%). Three of eight patients were revealed and operated on after the third post-operative year. In two other patients with stable but large Cat2PPH, a PPFA was suspected without absolute confirmation at further MRI. These patients were considered as ‘stable’ PPFA, although at risk for indication of reoperation. Table 3 Open in new tabDownload slide Post-operative complications detected with MRI in group II (aneurysms) Table 3 Open in new tabDownload slide Post-operative complications detected with MRI in group II (aneurysms) 3.2.2 Reoperations (Tables 4 and 5) Eight patients underwent elective reoperation for PPFA and two for EAA. One and 5-year freedom from reoperation for PPFA were 98 and 88%, respectively. In three patients, PPFA was detected after the third post-operative year. Tables 4 and 5 indicate the delay in each case, and the procedures employed at each reoperation and the outcome for PPFA and EAA, respectively. In one patient presenting PPFA (no. 13, Table 4), bacterial infection of the prosthesis was found 8 months after an ascending aorta+aortic valve replacement and indicated a root replacement using an aortic homograft. Operative mortality was 10% (1/10) in this group. Peri-prosthetic leaks came from proximal attachment of the graft (three cases), coronary reimplantation (five cases) or distal anastomosis (six cases). MRI findings were concordant with intra-operative observations. None of the five patients who underwent reoperation after Bentall procedure had a ‘button’ technique the first time. Two patients (2/5) presented a recurrent PPFA after a re-Bentall procedure, including one patient (no. 15, Table 4) who required a third operation 22 months after the second one for a recurrent right coronary anastomotic problem, and another patient (no. 8, Table 4) showing an early PPFA related to distal anastomosis but restrained from third-time surgery, due to the absence of MRI modification at 4 years. 3.3 Overall groups I and II In summary, when comparing MRI results between the two groups, we did not exhibit significant differences and we showed as many late complications after elective surgery for non-dissecting aneurysms as after emergency surgery for acute dissection. All Cat1PPH either regressed or remained stable; no false aneurysm occurred in this category. Twenty-five percent (28/114) of patients presented with a Cat2PPH which in 54% of cases (15/28) deteriorated in PPFA, including seven of 28 after the third post-operative year. In addition to this, EAA required reoperation in seven patients, and 13 other patients presented stable modifications of the thoracic aorta distant to the prosthesis. In total, 31% of patients (35/114) harboured significant morphologic changes of the thoracic aorta, including 22 patients who were reoperated during follow-up. Overall operative mortality was 13% (3/22). The length of time free from reoperation is presented in Fig. 5 . Fig. 5 Open in new tabDownload slide Freedom from reoperation (Kaplan–Meier method). Axis cut for graphical reasons. Fig. 5 Open in new tabDownload slide Freedom from reoperation (Kaplan–Meier method). Axis cut for graphical reasons. 4 Discussion Over the years surgical management for diseases of thoracic aorta has improved considerably, including earlier diagnosis and successive refinements in operative techniques. Therefore, in-hospital negative events have been reduced as well as operative mortality. More than 80% of patients can recover from emergency procedures for aortic dissection, and even more in elective surgery for aneurysm repair [4–6]. However, the underlying disease remains and residual fragility of aortic tissue exposes survivors to severe late complications and late deaths. Initial success rate supports the need for a careful follow-up of thoracic aorta morphology after surgery and for early detection of late pseudoaneurysms or aneurysms. This report revealed an unusual rate of late complications and reoperations in both groups of patients, particularly in the non-dissecting aneurysms group. We found that 35% of our patients harboured significant late morphologic changes of the thoracic aorta, either PPFA or EEA, requiring surgery in more than half of those cases. Other reports showed a lower rate for such events, but also relatively poor long-term results [7,8]. On another hand, late deaths have been reported in relation with rupture of the distal aortic arch [9] as well as unexplained sudden deaths. Type A aortic dissection has been associated with a 40% survival rate at 10 years. Most authors did not perform systematic MRI evaluation as we did, although several studies [10–13] have documented the superior performance of MRI in the demonstration of the thoracic aorta Asymptomatic complications have also been reported at angiography [14,15], but with a lower frequency, presumably because patients did not undergo routine screening examination. Because of its non-invasive nature, MRI should be considered as the gold standard for long-term follow-up including serial examinations of the entire thoracic aorta. In our experience, as for others [16], PPFA appeared as the most important late complication and surprisingly occurred after elective surgery for an aneurysm as well as after an emergency case of acute aortic dissection. In a vast majority of our patients, PPFA was not associated with any clinical symptom. In total, 25% of patients presented with a Cat2PPH which in 54% of cases deteriorated in PPFA. As opposed, none of the Cat1PPH complicated and therefore those may not be considered as pathological features. When analysing the volume and thickness of every PPH, we found that the first MRI contained predictive value. When the first post-operative MRI showed a PPH circumference greater than 10 mm or a non-circumferential PPH greater than 15 mm, the patient was at risk for a non-regressive Cat2PPH, and subsequently a PPFA. Our series exhibits 13% (5/38) of reoperations after a root replacement, without operative death. The incidence of reoperation after composite graft replacement has been reported as 6–12% [17]. Using angiography in late follow-up of Bentall procedure, Marvati and colleagues [14] reported three out of 15 patients presenting with PPFA that occurred predominantly at right coronary anastomosis. Even without routine evaluation, Kouchoukos and associates [18] reported nine PPFA out of 103 patients, and indicated that misdiagnosis was a cause for complex reoperative procedures with high mortality. MRI indicated to us the exact location of the dehiscence generating the PPFA. The coronary reimplantation failed more in group II (40%), whereas PPFA originated mainly from proximal Dacron graft attachment above the coronary arteries in group I (60%). It is indeed striking to note that all our patients reoperated for PPFA after root replacement underwent a first operation before we started to perform the ‘button’ technique for reattachment of the coronary arteries, and also before we abandoned prosthetic wrapping. Using this approach in the last 13 patients, we did not observe any late PPFA. Aneurysm wrap around the graft was extremely useful in achieving immediate hemostasis, but also caused excessive tension on all suture lines within the aneurysmal sac and induced leakage, large PPH and late PPFA. In six cases of our series, including five in group II, PPFA can be related to these factors and could have been eliminated by total root excision and tension-free coronary reimplantation at the first operation. Operative mortality (13%) at reoperation was comparable to the largest series of reoperations on the ascending aorta and aortic root, that ranged from 6 to 19%, as reported by Dougenis and colleagues [16]. We applied well-known principles such as femoral vessels cannulation and cardiopulmonary bypass before sternum re-entry in function of preoperative imaging, pericardial dissection limited to distal aorta and right atrium, coronary sinus perfusion and cold blood cardioplegia for myocardial protection, and selective antegrade cerebral perfusion technique. In two cases, profound cooling was initiated before sternotomy and arch repair was performed under retrograde cerebral perfusion into the superior vena cava, then flow was re-established in the antegrade direction through the new Dacron graft. Technically, debridement and aortic reconstruction were challenging at reoperation, due to the mediastinal tissue stiffness created by post-operative inflammatory reaction. In our experience, aortic homografts were helpful in patients having a second root replacement with delicate sutures on inflammatory and woody tissues (patients 3 and 6, Table 4). Finally, EAA in the native aorta proximal or distal to the graft has been commonly described after surgical treatment of aortic dissection and non-dissecting aneurysm. Although these complications occurred at distant locations and may not be obviously related to a technical failure of the suture lines, they may be consecutive to inadequate surgical management at the time of first operation. Many of the reoperations after type A acute dissection can be related to a conservative approach including incomplete resection of the sinuses of Vasalva. One recent report [19] emphasized a more radical proximal aorta resection. In contrast, a more aggressive approach beyond the ascending aorta, including ‘open’ anastomosis technique, may not be proven as beneficial. In spite of such a policy, we still observed in 95% of cases in group I the persistence of residual dissection and some severe cases of post-operative dilatation of the descending thoracic aorta that developed mainly from circulating false channels. Ergin and colleagues [19] and Barron and colleagues [12] showed, respectively, 47 and 72% of patients presenting residual patent false lumen, but these authors did not employ MRI for long-term follow-up. Similarly to our study, Guthaner and colleagues [13] demonstrated with an angiographic long-term study that a majority of patients (85%), whether symptomatic or not, continued to harbour patent distal false channel. In order to lower the frequency of persistent residual dissection and reduce the risk of late EAA, novel approaches to this important residual problem are necessary as combined endovascular treatments which aim to improve blood distribution into the subdiaphragmatic aortic branches and therefore organ perfusion. 5 Conclusion This work emphasizes the importance of indefinite surveillance after surgery of the ascending aorta. Periodic examination is possible by cardiovascular radiology specialists who are familiar with aortic diseases and surgical procedures. Routine MRI follow-up allows monitoring life-threatening complications such as PPFA in asymptomatic patients, by comparing the findings from the initial post-operative imaging with those from routine imaging in order to schedule reoperation, if necessary. Elective reoperation can provide satisfactory results in this particular group of patients, with low mortality and enhanced long-term survival. MRI surveillance can also help for critical analysis of technical failures. Adequate techniques for aortic root surgery, such as the button technique or extended root excision, are certainly influential on late results. In contrast, the late complications related to residual dissection distal to the prosthesis are not eliminated with more aggressive surgery and need novel approaches including endovascular therapy. ✩ Presented at the 13th Annual Meeting of the European Association for Cardio-thoracic Surgery, Glasgow, Scotland, UK, September 5–8, 1999. Dr H. Borst (Munich, Germany): I think this is an important contribution because up to now people have been concerned about the perfused perigraft space, but you are really talking about something else, hematomas occurring at the anastomosis. You didn't do any graft inclusion, did you, in this series? Dr Mesana: In the aneurysm group, at the beginning of our experience, about half of the Bentall procedures were performed with graft inclusion. Dr Borst: I have one question, and that is, how is the MRI different from a CT? Do you think they are equivalent? Dr Mesana: I think MRI is better. We published in Radiology with our colleagues from the radiology group that the CT scan was not as accurate on the root examination. CT scan was particularly limited in the presence of a valvular prosthesis. For examination of the descending aorta, we found at MRI that more than 90% of the patients operated on for Type A Stanford dissection presented had a residual dissection in the descending aorta. This is a higher frequency than all the previous reports, but it was assessed with MRI. Dr F. Wellens (Aalst, Belgium): Did you change your technical access to try to avoid the hematomas? Except for changing to an open aortic root anastomosis, do you have other tricks now, do you have another system for better drainage of the periaortic space? Dr Mesana: I believe that hematomas are always seen in the 3 months after the operation, either a little haematoma, or even larger. The glue had no impact on these hematomas and, as a matter of fact, from 3 years ago we don't use glue any more, even for aortic dissection, and we still have excellent results, even better than with the glue. The glue has just provided the surgeon with a higher level of confidence, but I don't believe that biological glue prevents late complications. We are more aggressive for distal repair on the arch, but this policy did not eliminate the problems distant to the prosthesis, because there is still re-entry in the descending aorta. We use more pledget-reinforced sutures than in the past, and we don't use that much glue. GRF glue in particular may be deleterious. In some cases where we reoperated we found some abnormalities around the prosthesis with large haematoma and a kind of tissue necrosis. We have reasonable experience and we did not use much formol, but I'm now a little concerned about an extended use of the glue. Dr Borst: Well, that's really very revolutionary, especially coming from France. Dr Mesana: Well, you know, France is big. I think Dr Bachet has published very good results. We are also pleased with our results; we have a very low operative mortality. I agree that glue is a very good adjunct to surgery, but I don't agree with surgeons who say that glue is a real technical revolution. As I said already, glue has just increased the confidence of surgeons in repairing the aortic dissection. Dr Borst: I agree with that. It certainly doesn't compensate for bad surgery. Dr Wellens: Were there correlations with biological markers, like increased sedimentation rate, CRP, and others? Were you looking at the biological markers of these patients? Dr Mesana: We did not investigate this and concentrated only on MRI. Dr D. Metras (Marseille, France): I have no experience about what you presented, but just a word about glue. We do many switch operations in the newborn. In the first, let's say, hundred, we were using glue, and in fact what it does, it obscures the small bleeding, and now we don't use glue any more. I agree with Dr Borst, better technique is better than glue which may be hiding the technical faults. Dr Mesana: I would just add that aprotinin plus good suture technique has probably a more significant impact than the glue itself. Dr Wellens: Could you identify the origin of the aneurysms, whether it was more the proximal or distal anastomosis in simple aneurysms or in the Bentall the coronary ostial suture line? Dr Mesana: It is a very important question. In aortic dissection, the peri-prosthetic leak was more frequent in the proximal site (80% of the cases out of 10 reoperations). I believe it's because we were not radical enough in the root repair in these patients. For the aneurysm group, we had a high number of PPFAs and hematomas, but I think it was in the early phase of the study when we didn't use the button technique, and half of these PPFAs occurred from the right coronary anastomosis; we confirm what anybody knows at present. With time, the rate peri-prosthetic false aneurysms will decrease with the use of the button technique, and with extension to root repair in aortic dissection. I am not sure about the open technique; maybe less peri-prosthetic false aneurysm, but still maybe some evolutive aneurysm distant to the prosthesis, because the disease is still present in the descending aorta. References [1] Cooley D.A. , Debakey M.E. . Resection of entire ascending aorta in fusiform aneurysm using cardiac bypass , J Am Med Assoc , 1956 , vol. 162 (pg. 1158 - 1159 ) Google Scholar Crossref Search ADS PubMed WorldCat [2] Gaubert J-Y. , Moulin G. , Mesana T. , Chagnaud C. , Caus T. , Delannoy L. , Blin D. , Bartoli J-M. , Kasbarian M. . Type A dissection of the thoracic aorta: use of MR imaging for long-term follow-up , Radiology , 1995 , vol. 196 (pg. 363 - 369 ) Google Scholar Crossref Search ADS PubMed WorldCat [3] David T.E. , Feindel C.M. , Bos J. . Repair of the aortic valve in patients with aortic insufficiency and aortic root aneurysm , J Thorac Cardiovasc Surg , 1995 , vol. 109 (pg. 345 - 352 ) Google Scholar Crossref Search ADS PubMed WorldCat [4] Bachet J. , Gigou F. , Laurian C. , Bical O. , Goudot B. , Guilmet D. . Four-year clinical experience with the gelatine-resorcine-formol biological glue in acute aortic dissection , J Thorac Cardiovasc Surg , 1982 , vol. 83 (pg. 212 - 217 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [5] Miller D.C. , Stinson E.B. , Oyer P.E. , Rossiter S.J. , Reitz B.A. , Griepp R.B. , Shumway N.E. . Operative treatment of aortic dissections: experience with 125 patients over a sixteen-year period , J Thorac Cardiovasc Surg , 1979 , vol. 78 (pg. 365 - 382 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [6] Crawford E.S. , Svenson L.G. , Coselli J.S. , Safi H.J. , Hess K.R. . Surgical treatment of aneurysm and/or dissection of the ascending aorta, transverse aortic arch, and ascending aorta and transverse aortic arch , J Thorac Cardiovasc Surg , 1989 , vol. 98 (pg. 659 - 674 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [7] Glower D.D. , Speier R.H. , White W.D. , Smith L.R. , Rankin J.S. , Wolfe W.G. . Management of long-term outcome of aortic dissections , Ann Surg , 1991 , vol. 111 (pg. 381 - 391 ) OpenURL Placeholder Text WorldCat [8] Haverich A. , Miller D.C. , Scott W.C. , Mitchell R.S. , Oyer P.E. , Stinson E.B. , Shumway N.E. . Acute and chronic aortic dissections: determinants of long-term outcome for operative survivors , Circulation , 1985 , vol. 72 Suppl 2 (pg. 22 - 34 ) OpenURL Placeholder Text WorldCat [9] Pansini S. , Gagiardotto P.V. , Pompei E. , Parisi F. , Bardi G. , Castenetto E. , Orzan F. , diSuma M. . Early and late risks factors in surgical treatment of acute type A aortic dissection , Ann Thorac Surg , 1998 , vol. 66 (pg. 779 - 784 ) Google Scholar Crossref Search ADS PubMed WorldCat [10] Amparo E.G. , Higgins C.B. , Hricak H. , Sollito R. . Aortic dissection: magnetic resonance imaging , Radiology , 1985 , vol. 155 (pg. 399 - 406 ) Google Scholar Crossref Search ADS PubMed WorldCat [11] Link K.M. , Lesko N.M. . The role of MR imaging in the evaluation of acquired diseases of the aorta , Am J Roentgenol , 1992 , vol. 158 (pg. 1115 - 1125 ) Google Scholar Crossref Search ADS WorldCat [12] Barron D.J. , Livesey S.A. , Brown I.W. , Delaney D.J. , Lamb R.K. , Monro J.L. . Twenty-year follow-up of acute type A dissection: the incidence and extent of distal aortic disease using magnetic resonance imaging , J Cardiac Surg , 1997 , vol. 12 (pg. 147 - 159 ) Google Scholar Crossref Search ADS WorldCat [13] Guthaner D.F. , Miller D.C. , Silvermann J.F. , Stinson E.B. , Wexler L. . Fate of false lumen following surgical repair of aortic dissections: an angiographic study , Radiology , 1979 , vol. 133 (pg. 1 - 8 ) Google Scholar Crossref Search ADS PubMed WorldCat [14] Marvasti M.A. , Parker F.B. , Randall P.A. , Witner G.A. . Composite graft replacement of the ascending aorta and aortic valve: late follow-up with intra-arterial digital subtraction angiography , J Thorac Cardiovasc Surg , 1988 , vol. 95 (pg. 924 - 928 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [15] Pugliese P. , Pessoto R. , Santini F. , Montalbano G. , Luciani G.B. , Mazzucco A. . Risk of late reoperations in patients with acute type A aortic dissection: impact of a more radical surgical approach , Eur J Cardio-thorac Surg , 1998 , vol. 13 (pg. 576 - 581 ) Google Scholar Crossref Search ADS WorldCat [16] Dougenis D. , Daily B.B. , Kouchoukos N.T. . Reoperations on the aortic root and ascending aorta , Ann Thorac Surg , 1997 , vol. 64 (pg. 986 - 992 ) Google Scholar Crossref Search ADS PubMed WorldCat [17] Crawford E.S. , Coselli J.S. , Safi H.J. . Stark J. , Pacifico A.D. . Reoperations for thoracic and thoracoabdominal aneurysms , Reoperations in cardiac surgery , 1989 Berlin Springer (pg. 361 - 381 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [18] Kouchoukos N.T. , Waering T.H. , Murphy S.F. , Perillo J.B. . Sixteen-year experience with aortic root replacement. Report of 172 operations , Ann Surg , 1991 , vol. 214 (pg. 308 - 318 ) Google Scholar Crossref Search ADS PubMed WorldCat [19] Ergin M.A. , Phillips R.A. , Galla J.D. , Lansman S.L. , Mendelson D.S. , Qintana C.S. , Griepp R.B. . Significance of distal false lumen after type A dissection repair , Ann Thorac Surg , 1994 , vol. 57 (pg. 820 - 825 ) Google Scholar Crossref Search ADS PubMed WorldCat © 2000 Elsevier Science B.V. All rights reserved. Elsevier Science B.V. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Cardio-Thoracic Surgery Oxford University Press

Late complications after prosthetic replacement of the ascending aorta: what did we learn from routine magnetic resonance imaging follow-up?

Loading next page...
 
/lp/oxford-university-press/late-complications-after-prosthetic-replacement-of-the-ascending-aorta-uxYFR0C00h

References (22)

Publisher
Oxford University Press
Copyright
© 2000 Elsevier Science B.V. All rights reserved.
Subject
Articles
ISSN
1010-7940
eISSN
1873-734X
DOI
10.1016/S1010-7940(00)00512-1
Publisher site
See Article on Publisher Site

Abstract

Abstract Objective: Advances in surgical technique have improved early survival after surgery of the ascending aorta. However, follow-up data document serious late complications, mainly evolutive peri-prosthetic false aneurysms. Magnetic resonance imaging (MRI) has proved to be highly effective for monitoring these complications. This study evaluates 10 years of experience with routine MRI for follow-up. Methods: Since January 1988, 114 patients with replacement of the ascending aorta either for type A acute dissection (group I, 45 patients) or aneurysms (group II, 69 patients) were followed up with annual MRI. Prosthetic replacement was either limited to supra-coronary ascending aorta (45%, 51/114) or extended to the aortic root and/or the aortic arch (55%). Biological glue was always utilized. MRI focused on peri-prosthetic haematoma, analyzing signal intensity changes and volume augmentation for early detection of false aneurysms, and on persistent residual dissection with or without evolutive aortic aneurysm distant to the prosthesis. Results: Peri-prosthetic hematomas were almost equally found in both groups (26 (58%) in group I and 42 (61%) in group II) and were detected within the first year. Peri-prosthetic false aneurysms developed in 15 patients (group I, seven; group II, eight) as a complication of pre-existing hematomas and were indicated for elective reoperation. Forty-three (96%) of patients in group I had persistent residual dissection. Five patients in group I and two in group II needed reoperation for evolutive aortic aneurysm. In total, 22 of 114 (19%) patients were reoperated on during follow-up (12 (27%) in group I and ten (15%) in group II). Operative mortality was 13% (3/22). Freedom from reoperation at 1 year/5 years was: group I, 93%/84%; group II, 98%/88%. Conclusion: Peri-prosthetic haematoma occurs equally after aneurysm or dissection repairs and is a pre-existing condition for peri-prosthetic false aneurysm; biological glue or extended repair do not prevent late complications. Long-term MRI follow-up allows successful elective reoperation for life-threatened but asymptomatic patients. Aorta, Aneurysm, Dissection, Peri-prosthetic haematoma, False aneurysm, Magnetic resonance imaging 1 Introduction Since the first successful repair reported by Cooley and DeBakey in 1956 [1], significant advances resulted in lower operative mortality for elective aneurysm repair as for emergency acute dissection surgery. During the last decade, the total number of patients at risk for development of late complications considerably increased. Therefore, reoperations on the thoracic ascending aorta has gained utmost surgical consideration. However, routine post-operative examination of a repaired thoracic aorta has not been as well emphasized as the evaluation of other common cardiac operation. Our group has previously documented follow-up data showing a high frequency of serious late complications in patients who underwent surgical treatment of acute Stanford type A aortic dissection [2]. We demonstrated that magnetic resonance imaging (MRI) has proved to be highly effective in detection and monitoring of peri-prosthetic haematoma (PPH), peri-prosthetic false aneurysm (PPFA) and evolutive aortic aneurysm (EAA) developing at distant location from the graft attachment. The purpose of this study was to evaluate routine MRI after surgical repair of acute dissection or non-dissecting aortic aneurysm and to assess the frequency of late complications in a larger group of patients who underwent a variety of surgical procedures. 2 Materials and methods Between January 1988 and December 1997, 114 patients who survived replacement of the ascending aorta either in emergency for Stanford type A aortic dissection (Group I; 45 patients), or electively for non-dissecting ascending aorta aneurysms (Group II; 69 patients), were followed up with routine MRI. 2.1 Group I: aortic dissection Forty-five patients survived emergency procedures that occurred within 4 h following their referral to our Centre and within 4 days from the onset of first acute events. There were 36 men and nine women (mean age, 55.9±17.7 years). Three patients had previous coronary surgery (more than 1 year before acute dissection), three had pre-existing ascending aorta aneurysm, and five were presenting Marfan syndrome. Employed surgical procedures are listed in Table 1 . Extension of prosthetic replacement to the aortic arch was more frequent since 1991, when a more radical approach was decided, consisting in a wider use of ‘open’ anastomosis technique that allows aortic arch exploration in search of additional intimal tears. We therefore found 47% (21/45) of patients presenting intimal tears in this aortic location and necessitating an extended repair. In some instances, intimal ruptures at the origin of innominate artery necessitated its transection followed by reimplantation in a separate graft. Table 1 Open in new tabDownload slide Type of procedures employed in group I (45 patients) and group II (69 patients) Table 1 Open in new tabDownload slide Type of procedures employed in group I (45 patients) and group II (69 patients) Five of eight patients who underwent Bentall procedures also had aortic arch surgery, consisting in two hemi-arch with innominate artery reimplantation and three total arch replacement. Bentall procedures were indicated in patients with Marfan disease and in cases with destruction of the sinus of Valsalva or tearing out of coronary ostia. Otherwise, tissue excision was achieved for adequate residual aortic wall remodelling and aortic valve coaptation. GRF (gelatine risorcine formaldehyde) biological glue was utilized in all instances, to obtain adequate proximal and distal stumps and optimal obliteration of the space between the dissected aortic layers, ultimately sutured with 4/0 monofilament after Teflon felt tissue reinforcement. 2.2 Group II: non-dissecting aneurysms In the same period of time, 69 patients survived elective surgery for non-dissecting aneurysm of the ascending aorta. There were 51 men and 18 women (mean age, 55.3±14.6 years). Employed surgical techniques are listed in Table 1. Patients undergoing aortic arch operations alone were excluded from this series. Interposition graft and aortic valve replacement (or re-replacement) were selected for patients with aortic valve disease (or prosthesis). This represented one-third of this group (23/69, 33%), including four patients bearing an old prosthesis. The entire root was replaced in 30 cases (30/69, 44%). The technique of coronary reimplantation varied from the standard technique pre-dating 1994, to the ‘button’ technique (13/30, 43%), associated with complete aortic transection followed by end-to-end anastomosis and using interrupted pledgeted stitches for annular implantation of the composite graft. More recently, the valve-sparing procedure, as described by David and colleagues [3] was performed in five selected young patients presenting moderate annulo-ectasia with preserved valvular structures. Impregnated Dacron grafts and composite grafts (St. Jude Medical, Minneapolis, MN) were utilized in all patients except one having a root replacement using an aortic homograft. In all cases, french glue or fibrin glue (ImmunoCo, Vienna, Austria) were employed as a spray or a simple layer over the suture lines. 2.3 MRI follow-up All patients were included in a routine follow-up program in which MRI was performed once during the first 3 post-operative months, and once a year thereafter. This sequence was possibly modified, depending on the results of previous examinations. A total of 325 MRI examinations were performed in this series (range, 1–8 examinations per patient; mean, 2.9 examinations). All patients had at least two MRI examinations. During follow-up, two patients were switched to computed tomography (CT) scan evaluation due to pacemaker implantation. No patient was lost to follow-up. At the beginning of this study, MRI findings were compared with those from angiography and computed tomography. First angiography, then CT were abandoned early, except for pre-surgical workshop in patients scheduled to undergo reoperation. MRI data were acquired with electrocardiographic gating in the transaxial and oblique sagittal planes parallel to the main axis of the aortic arch. Field of view was 38–40 cm, and section thickness varied from 5 to 7 mm. Whenever an intermediate signal intensity (SI) zone within the aorta or the Dacron graft prosthesis could not be definitely attributed to slow flow and thrombosis, multi-echo sequences were used to assess the SI and assess the blood-flow pattern in the suspected areas. Slow-flowing blood was characterized by SI enhancement on the second echo of the multi-echo sequence. Thrombus was characterized by no change in SI. Fast flowing blood was characterized by a signal void inside or outside the aortic lumen, demonstrating PPFA. After 1992, multi-echo sequences were replaced by rapid imaging with fast low-angle shot (TurboFLASH, Siemens, Erlangen, Germany) sequences after rapid intravenous injection of gadoterate meglumine (Guerbet, Roissy, France). Limited amounts of fibrin and/or small clots are commonly observed around the aortic prosthesis in the early post-operative period. Such an observation did not assess a PPH. On the other hand, a substantial increase in SI after injection of contrast material was attributed to a circulating blood flow around the prosthesis, suggestive of heterogeneous PPH with a high risk for PPFA, whereas unchanged SI was interpreted as non-complicated PPH. All PPH were therefore classified into the following two categories (Cat): Cat1PPH, moderate, homogeneous, with no SI change (Fig. 1) ; Cat2PPH, large, circumference greater than 10 mm, heterogeneous and/or enhanced SI after contrast injection (Fig. 2) . Fig. 1 Open in new tabDownload slide MRI showing a Cat1PPH: limited size, homogeneous, not circumferential, with no SI changes. Fig. 1 Open in new tabDownload slide MRI showing a Cat1PPH: limited size, homogeneous, not circumferential, with no SI changes. Fig. 2 Open in new tabDownload slide MRI showing a Cat2PPH: greater than 15 mm, heterogeneous, non-regressive, with SI changes. Fig. 2 Open in new tabDownload slide MRI showing a Cat2PPH: greater than 15 mm, heterogeneous, non-regressive, with SI changes. The peri-prosthesis anastomotic leaks which fed the PPFA were determined and compared with intraoperative findings (Fig. 3) . In the native aorta distant to the prosthesis, the diameter of the lumen was assessed and residual dissection with blood flow in the false lumen was searched (Fig. 4) . A diameter of 40 mm was the cut-off size for diagnosis of an aneurysm. An EAA was assessed as more than 10-mm increase within a 6-month period. Such a diagnosis needed to be confirmed over another 6-month period with repeated MRI examination before consideration for surgery. Fig. 3 Open in new tabDownload slide MRI showing PPFA with fast-flowing blood around the prosthesis from a leaking coronary suture. Fig. 3 Open in new tabDownload slide MRI showing PPFA with fast-flowing blood around the prosthesis from a leaking coronary suture. Fig. 4 Open in new tabDownload slide MRI showing persistent false lumen in the descending aorta, 8 years after arch repair. Fig. 4 Open in new tabDownload slide MRI showing persistent false lumen in the descending aorta, 8 years after arch repair. For statistical analysis, freedom from reoperation was calculated from the time of first operation to the end point of December 1997 (in months), using the Kaplan–Meier method. Significance was assigned at less than or equal to 5% probability (P-value less than 0.05). 3 Results 3.1 Group I 3.1.1 MRI follow-up MRI detected 26 PPH (26/45, 58%) divided into 12 Cat1PPH (12/45, 27%) which never complicated, and 14 Cat2PPH (14/45, 31%),which degenerated in seven PPFA as indicated in Table 2 . In two patients, PPFA was revealed as a fast progressing PPH within 3 months after surgery, and required rapid surgical management. In four patients also requiring reoperation, PPFA were only diagnosed after the third post-operative year. Cat2PPH could also improve as in four cases (4/14, 29%) showing regression in Cat1PPH. In three other patients presenting Cat2PPH, MRI follow-up showed evidence of circulating blood around the prosthesis without PPH volume augmentation. These patients did not undergo reoperation and were alive and stable at more than 2 years. Five patients presenting with EAA required reoperation, four with EAA distal to the aortic arch and one with an EAA developing in the sinuses of Valsalva. Table 2 Open in new tabDownload slide Post-operative complications detected with MRI in group I (dissections) Table 2 Open in new tabDownload slide Post-operative complications detected with MRI in group I (dissections) In addition to this, we observed non-evolutive aortic aneurysms distant to the prosthesis in six patients. Five of these six complications were demonstrated on MRI obtained during the first post-operative year, and no substantial change in morphology or size has been observed at follow-up up to 108 months. Two of these six patients died eventually, one from chronic renal failure and the other from cardiac failure not related to aortic valve dysfunction. Persistent residual dissection distal to the aortic graft with a patent false channel was almost constant at some degree, a majority of patients revealing various amounts of thrombus within the false channel. Total thrombosis and retraction of the false lumen occurred in only two patients, one in the first post-operative year, and another in the 18th post-operative month despite the presence of residual dissection 1 year earlier. 3.1.2 Reoperations (Tables 4 and 5) Table 4 Open in new tabDownload slide Reoperations for PPFA in group I (patients 1–7) and group II (patients 8–15)a Table 4 Open in new tabDownload slide Reoperations for PPFA in group I (patients 1–7) and group II (patients 8–15)a Table 5 Open in new tabDownload slide Reoperations for EAA in group I (patients 1 to 5) and group II (patients 6 and 7)a Table 5 Open in new tabDownload slide Reoperations for EAA in group I (patients 1 to 5) and group II (patients 6 and 7)a Twelve patients in this group were indicated for elective redo surgery, seven for PPFA and five for EAA. Ten of 12 patients were asymptomatic. Thoracic pain was reported in two early PPFA, necessitating reoperation at 2 months. One-year and 5-year freedom from reoperation for PPFA were 93 and 84%, respectively. Tables 4 and 5 indicate the delay in each case and procedures employed at each reoperation for PPFA and EAA, respectively. One patient (no. 5, Table 4) presented a PPFA in the first post-operative month related to bacterial contamination of the Dacron graft and died 2 days after surgery from septic shock. Another patient (no. 5, Table 5) died during reoperation for an EAA of the arch from massive bleeding and coagulopathy. Operative mortality was 17% (2/12) in this group. Intra-operative findings were always concordant with those at MRI. PPFA were fed by peri-prosthetic leaks coming from proximal graft attachment (seven cases), distal attachment (two cases), or coronary reimplantation (one case). 3.2 Group II 3.2.1 MRI Follow-up Forty-two PPH were observed during MRI follow-up (42/69, 61%), divided into 28 (28/69, 41%) regressive Cat1PPH and 14 (14/69, 20%) Cat2PPH, distributed as indicated in Table 3 . In eight patients, the PPFA occurred as an evolution of a pre-existing Cat2PPH, which could also be regressive as in four patients (4/14, 31%). Three of eight patients were revealed and operated on after the third post-operative year. In two other patients with stable but large Cat2PPH, a PPFA was suspected without absolute confirmation at further MRI. These patients were considered as ‘stable’ PPFA, although at risk for indication of reoperation. Table 3 Open in new tabDownload slide Post-operative complications detected with MRI in group II (aneurysms) Table 3 Open in new tabDownload slide Post-operative complications detected with MRI in group II (aneurysms) 3.2.2 Reoperations (Tables 4 and 5) Eight patients underwent elective reoperation for PPFA and two for EAA. One and 5-year freedom from reoperation for PPFA were 98 and 88%, respectively. In three patients, PPFA was detected after the third post-operative year. Tables 4 and 5 indicate the delay in each case, and the procedures employed at each reoperation and the outcome for PPFA and EAA, respectively. In one patient presenting PPFA (no. 13, Table 4), bacterial infection of the prosthesis was found 8 months after an ascending aorta+aortic valve replacement and indicated a root replacement using an aortic homograft. Operative mortality was 10% (1/10) in this group. Peri-prosthetic leaks came from proximal attachment of the graft (three cases), coronary reimplantation (five cases) or distal anastomosis (six cases). MRI findings were concordant with intra-operative observations. None of the five patients who underwent reoperation after Bentall procedure had a ‘button’ technique the first time. Two patients (2/5) presented a recurrent PPFA after a re-Bentall procedure, including one patient (no. 15, Table 4) who required a third operation 22 months after the second one for a recurrent right coronary anastomotic problem, and another patient (no. 8, Table 4) showing an early PPFA related to distal anastomosis but restrained from third-time surgery, due to the absence of MRI modification at 4 years. 3.3 Overall groups I and II In summary, when comparing MRI results between the two groups, we did not exhibit significant differences and we showed as many late complications after elective surgery for non-dissecting aneurysms as after emergency surgery for acute dissection. All Cat1PPH either regressed or remained stable; no false aneurysm occurred in this category. Twenty-five percent (28/114) of patients presented with a Cat2PPH which in 54% of cases (15/28) deteriorated in PPFA, including seven of 28 after the third post-operative year. In addition to this, EAA required reoperation in seven patients, and 13 other patients presented stable modifications of the thoracic aorta distant to the prosthesis. In total, 31% of patients (35/114) harboured significant morphologic changes of the thoracic aorta, including 22 patients who were reoperated during follow-up. Overall operative mortality was 13% (3/22). The length of time free from reoperation is presented in Fig. 5 . Fig. 5 Open in new tabDownload slide Freedom from reoperation (Kaplan–Meier method). Axis cut for graphical reasons. Fig. 5 Open in new tabDownload slide Freedom from reoperation (Kaplan–Meier method). Axis cut for graphical reasons. 4 Discussion Over the years surgical management for diseases of thoracic aorta has improved considerably, including earlier diagnosis and successive refinements in operative techniques. Therefore, in-hospital negative events have been reduced as well as operative mortality. More than 80% of patients can recover from emergency procedures for aortic dissection, and even more in elective surgery for aneurysm repair [4–6]. However, the underlying disease remains and residual fragility of aortic tissue exposes survivors to severe late complications and late deaths. Initial success rate supports the need for a careful follow-up of thoracic aorta morphology after surgery and for early detection of late pseudoaneurysms or aneurysms. This report revealed an unusual rate of late complications and reoperations in both groups of patients, particularly in the non-dissecting aneurysms group. We found that 35% of our patients harboured significant late morphologic changes of the thoracic aorta, either PPFA or EEA, requiring surgery in more than half of those cases. Other reports showed a lower rate for such events, but also relatively poor long-term results [7,8]. On another hand, late deaths have been reported in relation with rupture of the distal aortic arch [9] as well as unexplained sudden deaths. Type A aortic dissection has been associated with a 40% survival rate at 10 years. Most authors did not perform systematic MRI evaluation as we did, although several studies [10–13] have documented the superior performance of MRI in the demonstration of the thoracic aorta Asymptomatic complications have also been reported at angiography [14,15], but with a lower frequency, presumably because patients did not undergo routine screening examination. Because of its non-invasive nature, MRI should be considered as the gold standard for long-term follow-up including serial examinations of the entire thoracic aorta. In our experience, as for others [16], PPFA appeared as the most important late complication and surprisingly occurred after elective surgery for an aneurysm as well as after an emergency case of acute aortic dissection. In a vast majority of our patients, PPFA was not associated with any clinical symptom. In total, 25% of patients presented with a Cat2PPH which in 54% of cases deteriorated in PPFA. As opposed, none of the Cat1PPH complicated and therefore those may not be considered as pathological features. When analysing the volume and thickness of every PPH, we found that the first MRI contained predictive value. When the first post-operative MRI showed a PPH circumference greater than 10 mm or a non-circumferential PPH greater than 15 mm, the patient was at risk for a non-regressive Cat2PPH, and subsequently a PPFA. Our series exhibits 13% (5/38) of reoperations after a root replacement, without operative death. The incidence of reoperation after composite graft replacement has been reported as 6–12% [17]. Using angiography in late follow-up of Bentall procedure, Marvati and colleagues [14] reported three out of 15 patients presenting with PPFA that occurred predominantly at right coronary anastomosis. Even without routine evaluation, Kouchoukos and associates [18] reported nine PPFA out of 103 patients, and indicated that misdiagnosis was a cause for complex reoperative procedures with high mortality. MRI indicated to us the exact location of the dehiscence generating the PPFA. The coronary reimplantation failed more in group II (40%), whereas PPFA originated mainly from proximal Dacron graft attachment above the coronary arteries in group I (60%). It is indeed striking to note that all our patients reoperated for PPFA after root replacement underwent a first operation before we started to perform the ‘button’ technique for reattachment of the coronary arteries, and also before we abandoned prosthetic wrapping. Using this approach in the last 13 patients, we did not observe any late PPFA. Aneurysm wrap around the graft was extremely useful in achieving immediate hemostasis, but also caused excessive tension on all suture lines within the aneurysmal sac and induced leakage, large PPH and late PPFA. In six cases of our series, including five in group II, PPFA can be related to these factors and could have been eliminated by total root excision and tension-free coronary reimplantation at the first operation. Operative mortality (13%) at reoperation was comparable to the largest series of reoperations on the ascending aorta and aortic root, that ranged from 6 to 19%, as reported by Dougenis and colleagues [16]. We applied well-known principles such as femoral vessels cannulation and cardiopulmonary bypass before sternum re-entry in function of preoperative imaging, pericardial dissection limited to distal aorta and right atrium, coronary sinus perfusion and cold blood cardioplegia for myocardial protection, and selective antegrade cerebral perfusion technique. In two cases, profound cooling was initiated before sternotomy and arch repair was performed under retrograde cerebral perfusion into the superior vena cava, then flow was re-established in the antegrade direction through the new Dacron graft. Technically, debridement and aortic reconstruction were challenging at reoperation, due to the mediastinal tissue stiffness created by post-operative inflammatory reaction. In our experience, aortic homografts were helpful in patients having a second root replacement with delicate sutures on inflammatory and woody tissues (patients 3 and 6, Table 4). Finally, EAA in the native aorta proximal or distal to the graft has been commonly described after surgical treatment of aortic dissection and non-dissecting aneurysm. Although these complications occurred at distant locations and may not be obviously related to a technical failure of the suture lines, they may be consecutive to inadequate surgical management at the time of first operation. Many of the reoperations after type A acute dissection can be related to a conservative approach including incomplete resection of the sinuses of Vasalva. One recent report [19] emphasized a more radical proximal aorta resection. In contrast, a more aggressive approach beyond the ascending aorta, including ‘open’ anastomosis technique, may not be proven as beneficial. In spite of such a policy, we still observed in 95% of cases in group I the persistence of residual dissection and some severe cases of post-operative dilatation of the descending thoracic aorta that developed mainly from circulating false channels. Ergin and colleagues [19] and Barron and colleagues [12] showed, respectively, 47 and 72% of patients presenting residual patent false lumen, but these authors did not employ MRI for long-term follow-up. Similarly to our study, Guthaner and colleagues [13] demonstrated with an angiographic long-term study that a majority of patients (85%), whether symptomatic or not, continued to harbour patent distal false channel. In order to lower the frequency of persistent residual dissection and reduce the risk of late EAA, novel approaches to this important residual problem are necessary as combined endovascular treatments which aim to improve blood distribution into the subdiaphragmatic aortic branches and therefore organ perfusion. 5 Conclusion This work emphasizes the importance of indefinite surveillance after surgery of the ascending aorta. Periodic examination is possible by cardiovascular radiology specialists who are familiar with aortic diseases and surgical procedures. Routine MRI follow-up allows monitoring life-threatening complications such as PPFA in asymptomatic patients, by comparing the findings from the initial post-operative imaging with those from routine imaging in order to schedule reoperation, if necessary. Elective reoperation can provide satisfactory results in this particular group of patients, with low mortality and enhanced long-term survival. MRI surveillance can also help for critical analysis of technical failures. Adequate techniques for aortic root surgery, such as the button technique or extended root excision, are certainly influential on late results. In contrast, the late complications related to residual dissection distal to the prosthesis are not eliminated with more aggressive surgery and need novel approaches including endovascular therapy. ✩ Presented at the 13th Annual Meeting of the European Association for Cardio-thoracic Surgery, Glasgow, Scotland, UK, September 5–8, 1999. Dr H. Borst (Munich, Germany): I think this is an important contribution because up to now people have been concerned about the perfused perigraft space, but you are really talking about something else, hematomas occurring at the anastomosis. You didn't do any graft inclusion, did you, in this series? Dr Mesana: In the aneurysm group, at the beginning of our experience, about half of the Bentall procedures were performed with graft inclusion. Dr Borst: I have one question, and that is, how is the MRI different from a CT? Do you think they are equivalent? Dr Mesana: I think MRI is better. We published in Radiology with our colleagues from the radiology group that the CT scan was not as accurate on the root examination. CT scan was particularly limited in the presence of a valvular prosthesis. For examination of the descending aorta, we found at MRI that more than 90% of the patients operated on for Type A Stanford dissection presented had a residual dissection in the descending aorta. This is a higher frequency than all the previous reports, but it was assessed with MRI. Dr F. Wellens (Aalst, Belgium): Did you change your technical access to try to avoid the hematomas? Except for changing to an open aortic root anastomosis, do you have other tricks now, do you have another system for better drainage of the periaortic space? Dr Mesana: I believe that hematomas are always seen in the 3 months after the operation, either a little haematoma, or even larger. The glue had no impact on these hematomas and, as a matter of fact, from 3 years ago we don't use glue any more, even for aortic dissection, and we still have excellent results, even better than with the glue. The glue has just provided the surgeon with a higher level of confidence, but I don't believe that biological glue prevents late complications. We are more aggressive for distal repair on the arch, but this policy did not eliminate the problems distant to the prosthesis, because there is still re-entry in the descending aorta. We use more pledget-reinforced sutures than in the past, and we don't use that much glue. GRF glue in particular may be deleterious. In some cases where we reoperated we found some abnormalities around the prosthesis with large haematoma and a kind of tissue necrosis. We have reasonable experience and we did not use much formol, but I'm now a little concerned about an extended use of the glue. Dr Borst: Well, that's really very revolutionary, especially coming from France. Dr Mesana: Well, you know, France is big. I think Dr Bachet has published very good results. We are also pleased with our results; we have a very low operative mortality. I agree that glue is a very good adjunct to surgery, but I don't agree with surgeons who say that glue is a real technical revolution. As I said already, glue has just increased the confidence of surgeons in repairing the aortic dissection. Dr Borst: I agree with that. It certainly doesn't compensate for bad surgery. Dr Wellens: Were there correlations with biological markers, like increased sedimentation rate, CRP, and others? Were you looking at the biological markers of these patients? Dr Mesana: We did not investigate this and concentrated only on MRI. Dr D. Metras (Marseille, France): I have no experience about what you presented, but just a word about glue. We do many switch operations in the newborn. In the first, let's say, hundred, we were using glue, and in fact what it does, it obscures the small bleeding, and now we don't use glue any more. I agree with Dr Borst, better technique is better than glue which may be hiding the technical faults. Dr Mesana: I would just add that aprotinin plus good suture technique has probably a more significant impact than the glue itself. Dr Wellens: Could you identify the origin of the aneurysms, whether it was more the proximal or distal anastomosis in simple aneurysms or in the Bentall the coronary ostial suture line? Dr Mesana: It is a very important question. In aortic dissection, the peri-prosthetic leak was more frequent in the proximal site (80% of the cases out of 10 reoperations). I believe it's because we were not radical enough in the root repair in these patients. For the aneurysm group, we had a high number of PPFAs and hematomas, but I think it was in the early phase of the study when we didn't use the button technique, and half of these PPFAs occurred from the right coronary anastomosis; we confirm what anybody knows at present. With time, the rate peri-prosthetic false aneurysms will decrease with the use of the button technique, and with extension to root repair in aortic dissection. I am not sure about the open technique; maybe less peri-prosthetic false aneurysm, but still maybe some evolutive aneurysm distant to the prosthesis, because the disease is still present in the descending aorta. References [1] Cooley D.A. , Debakey M.E. . Resection of entire ascending aorta in fusiform aneurysm using cardiac bypass , J Am Med Assoc , 1956 , vol. 162 (pg. 1158 - 1159 ) Google Scholar Crossref Search ADS PubMed WorldCat [2] Gaubert J-Y. , Moulin G. , Mesana T. , Chagnaud C. , Caus T. , Delannoy L. , Blin D. , Bartoli J-M. , Kasbarian M. . Type A dissection of the thoracic aorta: use of MR imaging for long-term follow-up , Radiology , 1995 , vol. 196 (pg. 363 - 369 ) Google Scholar Crossref Search ADS PubMed WorldCat [3] David T.E. , Feindel C.M. , Bos J. . Repair of the aortic valve in patients with aortic insufficiency and aortic root aneurysm , J Thorac Cardiovasc Surg , 1995 , vol. 109 (pg. 345 - 352 ) Google Scholar Crossref Search ADS PubMed WorldCat [4] Bachet J. , Gigou F. , Laurian C. , Bical O. , Goudot B. , Guilmet D. . Four-year clinical experience with the gelatine-resorcine-formol biological glue in acute aortic dissection , J Thorac Cardiovasc Surg , 1982 , vol. 83 (pg. 212 - 217 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [5] Miller D.C. , Stinson E.B. , Oyer P.E. , Rossiter S.J. , Reitz B.A. , Griepp R.B. , Shumway N.E. . Operative treatment of aortic dissections: experience with 125 patients over a sixteen-year period , J Thorac Cardiovasc Surg , 1979 , vol. 78 (pg. 365 - 382 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [6] Crawford E.S. , Svenson L.G. , Coselli J.S. , Safi H.J. , Hess K.R. . Surgical treatment of aneurysm and/or dissection of the ascending aorta, transverse aortic arch, and ascending aorta and transverse aortic arch , J Thorac Cardiovasc Surg , 1989 , vol. 98 (pg. 659 - 674 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [7] Glower D.D. , Speier R.H. , White W.D. , Smith L.R. , Rankin J.S. , Wolfe W.G. . Management of long-term outcome of aortic dissections , Ann Surg , 1991 , vol. 111 (pg. 381 - 391 ) OpenURL Placeholder Text WorldCat [8] Haverich A. , Miller D.C. , Scott W.C. , Mitchell R.S. , Oyer P.E. , Stinson E.B. , Shumway N.E. . Acute and chronic aortic dissections: determinants of long-term outcome for operative survivors , Circulation , 1985 , vol. 72 Suppl 2 (pg. 22 - 34 ) OpenURL Placeholder Text WorldCat [9] Pansini S. , Gagiardotto P.V. , Pompei E. , Parisi F. , Bardi G. , Castenetto E. , Orzan F. , diSuma M. . Early and late risks factors in surgical treatment of acute type A aortic dissection , Ann Thorac Surg , 1998 , vol. 66 (pg. 779 - 784 ) Google Scholar Crossref Search ADS PubMed WorldCat [10] Amparo E.G. , Higgins C.B. , Hricak H. , Sollito R. . Aortic dissection: magnetic resonance imaging , Radiology , 1985 , vol. 155 (pg. 399 - 406 ) Google Scholar Crossref Search ADS PubMed WorldCat [11] Link K.M. , Lesko N.M. . The role of MR imaging in the evaluation of acquired diseases of the aorta , Am J Roentgenol , 1992 , vol. 158 (pg. 1115 - 1125 ) Google Scholar Crossref Search ADS WorldCat [12] Barron D.J. , Livesey S.A. , Brown I.W. , Delaney D.J. , Lamb R.K. , Monro J.L. . Twenty-year follow-up of acute type A dissection: the incidence and extent of distal aortic disease using magnetic resonance imaging , J Cardiac Surg , 1997 , vol. 12 (pg. 147 - 159 ) Google Scholar Crossref Search ADS WorldCat [13] Guthaner D.F. , Miller D.C. , Silvermann J.F. , Stinson E.B. , Wexler L. . Fate of false lumen following surgical repair of aortic dissections: an angiographic study , Radiology , 1979 , vol. 133 (pg. 1 - 8 ) Google Scholar Crossref Search ADS PubMed WorldCat [14] Marvasti M.A. , Parker F.B. , Randall P.A. , Witner G.A. . Composite graft replacement of the ascending aorta and aortic valve: late follow-up with intra-arterial digital subtraction angiography , J Thorac Cardiovasc Surg , 1988 , vol. 95 (pg. 924 - 928 ) Google Scholar PubMed OpenURL Placeholder Text WorldCat [15] Pugliese P. , Pessoto R. , Santini F. , Montalbano G. , Luciani G.B. , Mazzucco A. . Risk of late reoperations in patients with acute type A aortic dissection: impact of a more radical surgical approach , Eur J Cardio-thorac Surg , 1998 , vol. 13 (pg. 576 - 581 ) Google Scholar Crossref Search ADS WorldCat [16] Dougenis D. , Daily B.B. , Kouchoukos N.T. . Reoperations on the aortic root and ascending aorta , Ann Thorac Surg , 1997 , vol. 64 (pg. 986 - 992 ) Google Scholar Crossref Search ADS PubMed WorldCat [17] Crawford E.S. , Coselli J.S. , Safi H.J. . Stark J. , Pacifico A.D. . Reoperations for thoracic and thoracoabdominal aneurysms , Reoperations in cardiac surgery , 1989 Berlin Springer (pg. 361 - 381 ) Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC [18] Kouchoukos N.T. , Waering T.H. , Murphy S.F. , Perillo J.B. . Sixteen-year experience with aortic root replacement. Report of 172 operations , Ann Surg , 1991 , vol. 214 (pg. 308 - 318 ) Google Scholar Crossref Search ADS PubMed WorldCat [19] Ergin M.A. , Phillips R.A. , Galla J.D. , Lansman S.L. , Mendelson D.S. , Qintana C.S. , Griepp R.B. . Significance of distal false lumen after type A dissection repair , Ann Thorac Surg , 1994 , vol. 57 (pg. 820 - 825 ) Google Scholar Crossref Search ADS PubMed WorldCat © 2000 Elsevier Science B.V. All rights reserved. Elsevier Science B.V.

Journal

European Journal of Cardio-Thoracic SurgeryOxford University Press

Published: Sep 1, 2000

Keywords: Aorta Aneurysm Dissection Peri-prosthetic haematoma False aneurysm Magnetic resonance imaging

There are no references for this article.