European Journal of Cardio-Thoracic Surgery Advance Access published February 16, 2015
ORIGINAL ARTICLE
European Journal of Cardio-Thoracic Surgery (2015) 1–6 doi:10.1093/ejcts/ezv020
Post-mortem computed tomography and post-mortem computed tomography angiography following transcatheter aortic valve implantation† Beatrice Vogela,*, Axel Heinemannb, Helmut Gulbinsa, Hendrik Treedea, Hermann Reichenspurnera, Klaus Püschelb and Hermann Vogelb a b
Department of Cardiovascular Surgery, University Heart Center Hamburg, Hamburg, Germany Department of Legal Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
* Corresponding author. Department of Cardiovascular Surgery, University Heart Center Hamburg Martinistraße 52; 20246 Hamburg, Germany. Tel: +49-40-69669349; fax: +49-40-5366862; e-mail: [email protected] (B. Vogel). Received 14 September 2014; received in revised form 25 December 2014; accepted 29 December 2014
Abstract OBJECTIVES: At present, transcatheter aortic valve implantation (TAVI) is widely used. As with any interventional treatment, however, TAVI may also be accompanied by complications and may result in periprocedural mortality. This study aims to evaluate such complications and causes of death after TAVI. METHODS: The study included 32 deceased (59.4% female, n = 19, median age: 82 years) patients with TAVI, since 2008, in whom postmortem computed tomography (PMCT) and PMCT angiography were performed with the intention of identifying complications. RESULTS: Altogether, we registered bleeding (28.1%, 9/32), perforation and rupture (25%, 8/32), cerebral infarction (18.8%, 6/32), injury of the conduction system (3.1%, 1/32), insufficiency of the aortic (12.5%, 4/32) and the mitral valve (9.4%, 3/32) and of valve-in-valve procedures (9.4%, 3/32). Furthermore, there were findings due to cardiopulmonary resuscitation and intensive care. PMCT and PMCT angiography has advantages over autopsy. The demonstration of bleeding vessels, ruptures, the position of the implanted aortic valve and its effects on the mitral valve and its suspensions were more easily accessible by computed tomography-imaging display than by customary autopsy photo-documentation. CONCLUSIONS: After TAVI, PMCT and PMCT angiography successfully demonstrated the complications leading to death. PMCT and PMCT angiography contribute to the post-mortem analysis of causes of periprocedural death. Keywords: Transcatheter aortic valve implantation • Complications • Post-mortem computed tomography • Multiphase PMCT angiography
INTRODUCTION Intra- or post-procedural complications in transcatheter aortic valve implantation (TAVI) patients may lead to rapid deterioration of the patient with subsequent death. In these cases, it is not always possible to clarify the reason in the clinical situation. For the development of these interventional strategies, however, detection and visualization of the complications, especially the affection of anatomical structures within the heart is of great importance. During the last years, post-mortem computed tomography (PMCT) has become an additional and reliable tool of the pathologist [1, 2]. PMCT guides the autopsy and furnishes additional information that cannot be achieved by conventional autopsy. † Presented at the 28th Annual Meeting of the European Association for CardioThoracic Surgery, Milan, Italy, 11–15 October 2014.
It has been proposed that PMCT could even replace autopsy. This proposal has induced a discussion about the priority of both methods. The consequence is a rapid progress of PMCT and the development of new tools. One of these new tools is PMCT angiography. PMCT angiography is a rather new diagnostic tool used in forensic pathology to examine the cardiovascular situation. This method allows the visualization of blood flow at the time of death. However, there is a limiting factor: PMCT angiography is invasive. Therefore, the relatives must give their consent. The potential and value of PMCT angiography for the detection of coronary disease has been analysed systematically by the group of Grabherr et al. [3]; they showed the value of PMCT angiography in the analysis of vascular and cardiac pathology. Therefore, it seemed likely to employ PMCT and PMCT angiography for analysing the outcome of TAVI. Furthermore, it seemed likely that an analysis with PMCT and PMCT angiography of the deceased, who had undergone TAVI, offered the chance to visualize a larger
© The Author 2015. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ADULT CARDIAC
Cite this article as: Vogel B, Heinemann A, Gulbins H, Treede H, Reichenspurner H, Püschel K et al. Post-mortem computed tomography and post-mortem computed tomography angiography following transcatheter aortic valve implantation. Eur J Cardiothorac Surg 2015; doi:10.1093/ejcts/ezv020.
2
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery
number of serious and deadly complications than an analysis of the patients also treated with TAVI who had survived the procedure. This leads to the following question: Which complications and causes of death can be found in those deceased, who had undergone TAVI, by applying PMCT and PMCT angiography?
MATERIALS AND METHODS Since 2008, PMCT has been performed in the Department of Legal Medicine of the University of Hamburg. Until 2014, more than 5000 deceased were examined. Searching the PMCT archive, we found 32 deceased with prior TAVI (13 men, 19 women, age: 68–92 years, median age: 82 years). Thirty had died as in-patients during their hospital stay directly following the procedure, i.e. up to 3 weeks after the intervention; 2 had died later (one after 3, the other after 6 months). The Department has its own CT scanner: Until 2013, a 4 Scan MDCT (Multi Detector CT, MX 8000, Philips) was used, followed by a 16 MDCT (Philips Imation) in 2013. The examination parameters for the whole body computed tomography (CT) were: 1 mm scan, pitch 1.5, 180–230 mAs, 130 kV. Each of those 32 patients underwent PMCT; 8 of them had an additional PMCT angiography with oily contrast substance, and 1 with air for contrast. PMCT angiography with contrast substance was done as a multiphase PMCT angiography following the method of Grabherr et al. [3]: multiphase PMCT angiography has an arterial, a venous and a dynamic phase. Primarily, 1200 ml of an oily perfusate was injected into the femoral artery (arterial phase), and after a few minutes (necessary to perform a whole body CT) an equal quantity was injected into a femoral vein (venous phase). Finally, in the dynamic phase, a special pump simulates the circulation. A whole-body CT was obtained after each of those phases [3]. Consent had been obtained for PMCT, multiphase PMCT angiography and autopsy (25/32).
RESULTS The analysis of these 32 cases allowed one to differentiate between specific procedure-related and non-procedure-related lethal and non-lethal complications (Table 1).
Procedure-related lethal complications The most frequent pathological finding was haemorrhage (9/32: Pleura 4, pericardium 2, inguinal and pelvis 2, multiple sources 2). Multiphase PMCT angiography visualized the source of bleeding in all examined cases (Fig. 1A and B). In conventional autopsy, however, it was not always possible to identify the bleeding vessel; this could only be suggested by interpreting indirect findings such as haematoma or haemorrhagic effusions. Perforation of blood vessels and rupture of the myocardium with damage to the cardiac circulatory system (8/32) were other lethal complications after TAVI. PMCT combined with multiphase PMCT angiography made the site of the interruption visible: In 1 case, the balloon-dilatation of the aortic valve had induced a rupture, which caused a shunt between the left ventricle and the right atrium. The patient suffered from electromechanical dissociation and went into cardiogenic shock requiring resuscitation. In this clinical setting, the injury had not been diagnosed. In PMCT alone, a rupture was
Table 1: Procedure-related and non-procedure-related complications n Procedure-related lethal complicationsa Bleeding (pleura 4 with mediastinum, pericardium 2 with mediastinum, subarachnoid space 1, inguinal and pelvis 2, multiple sources 2) Perforation and rupture (aortic ring 4, myocardium 1, aorta 1, pelvic artery 2) Injury of the conduction system (due to rupture) Procedure-related non-lethal complicationsa Aortic valve insufficiency (failure of the implanted valve 2, paravalvular leakage 2, treatment with valve-in-valve procedures 2) Embolism of the implanted valve into the aorta and implantation of a second valve Overlap of the coronary ostia Deformation of the mitral valve Non-procedure-related lethal and non-lethal complicationsa: Cerebral infarction 5 (4 non-lethal), multiple organ failure 14, rib fractures 24, due to futile resuscitation 26, pulmonary arterial embolism 1, tracheal tube in wrong position 3, splenic infarction 1 Cerebral infarction Lethal Non-lethal Multiple organ failure etc a
9
8 1 3
1 4 2
1 4 14
Multiple findings in 1 deceased were enlisted.
suspected; multiphase PMCT angiography clearly visualized the rupture and the shunt (Fig 2). The site of the rupture allowed one to deduce an injury of the electrical conduction system, and trauma to the AV-node was suggested, later confirmed by autopsy.
Procedure-related non-lethal complications Compared with static autopsy, one important advantage of multiphase PMCT angiography was the possibility for dynamic functional assessment of the implanted valves: multiphase PMCT angiography visualized aortic regurgitation in 2 cases. One persisting paravalvular leakage had occurred after an index procedure, and in 2 patients, aortic regurgitation had occurred after a coaxial ‘valve-in-valve’ procedure (1 Edwards Sapien valve inside a CoreValve, and 1 CoreValve placement following a conventional aortic valve replacement). During the arterial phase of the multiphase PMCT angiography, contrast substance propagated into the left ventricle. The contrast substance passed between the outer border of the valves’ rim and the inner surface of the aorta, which induced the diagnosis of paravalvular leakage (Fig. 3A and B). In addition to dynamic functional assessment of the aortic valve, also an analysis of coexisting mitral valve regurgitation was possible by multiphase PMCT angiography. The interaction between the new aortic valve with the mitral valve papillary ligaments and muscles on one side, and deformation of the mitral valve leaflets on the other were visualized. This information could not be obtained by PMCT alone; and even autopsy had its limits. Obviously, PMCT and multiphase PMCT angiography done together visualized more relevant findings than autopsy alone: PMCT alone and combined with multiphase PMCT angiography could also be displayed repeatedly, which facilitated evaluation under different diagnostic assumptions. This proved to be very
3
ADULT CARDIAC
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery
Figure 1: Bleeding and TAVI ( JenaValve). PMCT (A) and multiphase PMCT angiography (B). (A) Blood (with layer formation) in the left chest. (B) Intercostal artery with contrast substance extravasation. MIP: maximum intensity projection; PMCT: post-mortem computed tomography; TAVI: transcatheter aortic valve implantation; MPMCTA: multiphase PMCT angiography.
Figure 2: Rupture of the aortic ring. Subaortic shunt from the left ventricle to the right atrium. MPMCTA: multiphase PMCT angiography; MIP: maximum intensity projection.
important, when discussing the initially not always obvious cause of complications and possibilities for their avoidance. In 4 cases, PMCT and multiphase PMCT angiography made evident that the implanted valve reached or overlapped with the ostia of the coronary arteries; and all 4 patients had pre-existing three-vessel coronary artery disease. Thus, the overlap of the ostia was discussed as a possible cause for lethal coronary artery perfusion. The grid of the implanted valve overlapping the coronary artery ostium together with coronary artery stenosis was considered a risk. However, multiphase PMCT angiography showed that the overlapping valve did not block the passage of the contrast substance into the coronary artery. Nevertheless, one has to bear in mind that, in the living, an overlapping valve could impede the placement of a coronary artery catheter and stent insertion during such procedures, if necessary, later on. Autopsy did not provide additional findings.
Non-procedure-related lethal and non-lethal complications Five deceased showed recent cerebral infarctions in PMCT, which, however, were considered non-lethal. One deceased had an
infarction of the entire left hemisphere with marked oedema and cerebral herniation with incarceration, which was considered the cause of death. All deceased had calcifications of the coronary arteries, easily visible on PMCT. All deceased also investigated by multiphase PMCT angiography had three-vessel disease. On two occasions, autopsy revealed a myocardial infarction corresponding to areas of limited blood supply. Both, PMCT and autopsy made multiple-organ failure apparent; and none of these two diagnostic procedures was superior to the other. PMCT visualized generalized oedema, pleural and peritoneal effusion and also fluid in the scrotum. Additionally, signs of delayed ante mortem renal excretion of CT contrast material (injected ante mortem) were seen, such as dense contrast in the gall bladder, in the renal parenchyma and in the meninges. Therefore, the differential diagnosis of the lethal outcome also had to include an immediate or delayed lethal contrast material reaction. PMCT easily showed the typical pattern of rib fractures due to previous cardiopulmonary resuscitation. In 3 cases, the endotracheal tube ended in one main bronchus, and in 1 of these 3 cases, the tube’s balloon blocked the other main bronchus. PMCT offered an easy proof of these findings and had obvious advantages compared with autopsy, usually done after tube removal. Once, PMCT visualized fulminant pulmonary embolism: a contrast-enhanced thrombus was visible at the tip of a central venous catheter and a similar structure in the left pulmonary artery (Fig. 4A–C). Ante mortem, the pulmonary embolism was not at all suspected. This finding proved that the prior TAVI procedure itself was not the cause of death.
DISCUSSION In our study of deceased patients following TAVI, we were able to reproduce the good diagnostic results of PMCT and especially of PMCT angiography as described by Grabherr et al. [3] for deceased patients with coexisting coronary heart disease. PMCT showed haemorrhage, calcifications, the type of the implanted valve and coronary artery disease. Multiphase PMCT angiography provided additional information; it not only localized the bleeding vessel but also allowed statements about function and dysfunction of the implanted device. These statements apply to the following:
4
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery
Figure 3: Paravalvular leakage after a valve-in-valve procedure. (A) Contrast substance passes between the external limit of the valves and the inner surface of the aorta. PMCT/MPMCT angiography. Arterial phase. (B) Contrast substance propagates into the left ventricle. MPMCTA: multiphase PMCT angiography.
Figure 4: TAVI with embolic-like migration of the implanted valve into the aorta and implantation of a second valve. Thrombus at the tip of a central venous catheter and an embolus in the left pulmonary artery. (A) Contrast medium-enhanced thrombus at the tip of the central venous catheter. (B) Contrast medium-enhanced thrombus the left pulmonary artery. PMCT: post-mortem computed tomography.
the aortic valve by visualizing contrast substance between the implanted valve and the aortic wall; the mitral valve with its leaflets and its suspension; the coronary ostia with the overlap of the grid of the implanted valve; and the coronary arteries with stenosis and obstruction. Compared with autopsy, PMCT and multiphase PMCT angiography had clear advantages in detecting functional disorders, whereas autopsy with macro- and microscopic evaluation was advantageous in detecting structural organ damage. Guided by multiphase PMCT angiography, autopsy also determined the type of a localized vascular obstruction, and proved a suspected myocardial infarction. The main advantage of conventional autopsy, however, lies in obtaining tissue for histology, and close macroscopic organ-evaluation. Thus, the detection of infarction and other organic alterations became possible and confirmed the prior diagnostic suspicion. In our series, the accuracy for multiphase PMCT angiography in detecting cardiac causes of death was apparent; this corresponded to the literature [1–4]. A combination of PMCT and multiphase PMCT angiography with autopsy led to optimal results [2]. This was especially true when PMCT and multiphase PMCT angiography guided the autopsy. Our observations indicated that PMCT and MPMCTA also combined with autopsy to furnish better results and more information about lethal outcome after TAVI than either
method alone. Neither method can serve as a gold standard. Furthermore, our observations allowed the conclusion that PMCT alone, but better yet, in combination with multiphase PMCT angiography is a good alternative to investigate an in-patient’s lethal outcome following TAVI [1, 5], when autopsy cannot be performed. Haemorrhage was the most frequent complication (9 of 32). For 6 cases, it was due to procedural manipulations (catheter insertion, dilatation of the aortic valve), twice due to cardiopulmonary resuscitation with rib fractures and vascular rupture, and once due to anticoagulation with subsequent subarachnoid bleeding. This corresponds to the literature, which confirmed that, after TAVI, haemorrhage of any kind was a life-threatening complication [6–9]. Furthermore, our results supported previous statements that, after a transapical approach, haemorrhage is the most important predictor of a life-threatening situation [10, 11]. Only MPCTA was able to visualize the source of bleeding in all cases directly, thus leading to a definite diagnosis. This was a clear advantage compared with conventional autopsy, where sometimes bleeding could only be suspected by indirect findings, such as haemorrhagic effusions or haematoma. Ruptures appeared in 8 of 32 deceased. A rupture itself seems to indicate a life-threatening situation; this seems even more valid for a rupture into preformed potentially free spaces, like the
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery
CONCLUSION PMCT and multiphase PMCT angiography of the deceased after TAVI furnish information about procedure- and non-procedurerelated complications and thus make an important contribution in this field. Compared with the living TAVI patient, in the deceased more serious and lethal complications became obvious and allowed detailed and comparative analysis. PMCT combined with multiphase PMCT angiography produced results at least as good as those of autopsy. Furthermore, compared with autopsy alone, they had the big advantage of permanent repeated display. Furthermore, multiphase PMCT angiography allowed the visualization of functional pathologies. The combination of all three
tools provides the physician with more relevant information than conventional autopsy alone.
Conflict of interest: none declared.
REFERENCES [1] Westphal SE, Apitzsch J, Penzkofer T, Mahnken AH, Knüchel R. Virtual CT autopsy in clinical pathology: feasibility in clinical autopsies. Virchows Arch 2012;461:211–9. [2] Westphal SE, Apitzsch JC, Penzkofer T, Kuhl CK, Mahnken AH, Knüchel R. Contrast-enhanced postmortem computed tomography in clinical pathology: enhanced value of 20 clinical autopsies. Hum Pathol 2014;45: 1813–23. [3] Grabherr S, Grimm J, Dominguez A, Vanhaebost J, Mangin P. Advances in post- mortem CT-angiography. Br J Radiol 2014;87: doi: 10.1259/bjr. 20130488. [4] Rubin JM, Avanzas P, del Valle R, Renilla A, Ríos E, Calvo D et al. Atrioventricular conduction disturbance characterization in transcatheter aortic valve implantation with the CoreValve prosthesis. Circ Cardiovasc Interv 2011;4:280–6. [5] Wichmann D, Heinemann A, Weinberg C, Vogel H, Hoepker WW, Grabherr S et al. Virtual autopsy with multiphase postmortem computed tomographic angiography versus traditional medical autopsy to investigate unexpected deaths of hospitalized patients: a cohort study. Ann Intern Med 2014;160:534–41. [6] Eggebrecht H, Schmermund A, Voigtländer T, Kahlert P, Erbel R, Mehta RH. Risk of stroke after transcatheter aortic valve implantation (TAVI): a meta-analysis of 10,037 published patients. EuroIntervention 2012;8: 129–38. [7] Pilgrim T, Stortecky S, Luterbacher F, Windecker S, Wenaweser P. Transcatheter aortic valve implantation and bleeding: incidence, predictors and prognosis. J Thromb Thrombolysis 2013;35:456–62. [8] Stortecky S, Wenaweser P, Diehm N, Pilgrim T, Huber C, Rosskopf AB et al. Percutaneous management of vascular complications in patients undergoing transcatheter aortic valve implantation. JACC Cardiovasc Interv 2012;5:515–24. [9] Van Mieghem NM, Tchetche D, Chieffo A, Dumonteil N, Messika-Zeitoun Dvan der Boon RM et al. Incidence, predictors, and implications of access site complications with transfemoral transcatheter aortic valve implantation. Am J Cardiol 2012;110:1361–7. [10] Borz B, Durand E, Godin M, Tron C, Canville A, Litzler PY et al. Incidence, predictors and impact of bleeding after transcatheter aortic valve implantation using the balloon-expandable Edwards prosthesis. Heart 2013;99: 860–5. [11] Wong DR, Ye J, Cheung A, Webb JG, Carere RG, Lichtenstein SV. Technical considerations to avoid pitfalls during transapical aortic valve implantation. J Thorac Cardiovasc Surg 2010;140:196–202. [12] Martínez MI, García HG, Calvar JA. Interventricular septum rupture after transcatheter aortic valve implantation. Eur Heart J 2012;33:190. [13] Neragi-Miandoab S, Michler RE. A Review of Most Relevant Complications of Transcatheter Aortic Valve Implantation. ISRN Cardiol 2013; doi: 10.1155/2013/956252. [14] Haldenwang PL, Bechtel M, Schlömicher M, Lindstaedt M, Strauch JT. Annular rupture leading to fatal complications in an elderly patient with calcified aortic and mitral annulus undergoing transapical aortic valve implantation. Thorac Cardiovasc Surg 2013;61:425–7. [15] Ferrari E, Rizzo E, Sulzer C, von Segesser LK. Unexpected left ventricular free- wall rupture following an aortic catheter-valve implantation. Eur J Cardiothorac Surg 2010;37:242–4. [16] Pasic M, Unbehaun A, Dreysse S, Buz S, Drews T, Kukucka M et al. Rupture of the device landing zone during transcatheter aortic valve implantation; a life- threatening but treatable complication. Circ Cardiovas Interv 2012; 5:424–32. [17] Détaint D, Lepage L, Himbert D, Borchert E, Messika-Zeitoun D, Iung B et al. Determinants of significant paravalvular regurgitation after transcatheter aortic valve: implantation impact of device and annulus discongruence. JACC Cardiovasc Interv 2009;2:821–7. [18] Tamburino C, Barbanti M, Capodanno D, Mignosa C, Gentile M, Aruta P et al. Comparison of complications and outcomes to one year of
ADULT CARDIAC
pleural space (haematothorax) or the pericardium (haematopericardium with pericardial tamponade). One observation concerned a shunt connecting the left ventricle with the right atrium. Observations about ruptures creating intracardiac shunts were rare in the literature [12, 13]; therefore, we classified our case as exceptional. The ruptures into the mediastinum and into the chest were life-threatening [14–16]. This was valid, too, for one observation of a rupture of the iliac artery. Aortic valve insufficiency (3 of 32) also seemed to be a risk factor. This applied both to failure of the implanted valve and to paravalvular leakage. This corresponds with previous communications [17, 18]. Additionally, paravalvular leakage has been described as a significant predictor of 1-year mortality [19, 20]. Implanting a second valve into the first was an option in our series. The observations in our series support the statement that, in planning for a TAVI, all potential complications have to be considered [21, 22]. MPMCTA showed the ability to detect regurgitations of the implanted valves, valvular as well as paravalvular leakages. This was not possible with conventional autopsy or PMCT alone. Additionally, regurgitations of the other cardiac valves could also be visualized; thus, MPMCTA had the advantage to make apparent the functional status of the heart ante mortem except for myocardial contractility. The position of the implanted valve could indicate potential dysfunctions. The upper margin of the valve cusps could reach or even cover the coronary arteries (n = 4); an impairment of perfusion was a possible consequence. However, in our series, such an impaired perfusion was never considered to be the cause of death. Nevertheless, in the living, the selective angiography of the coronary arteries and the insertion of stents could become a problem [14]. The base of the valve deformed the upper leaflet of the mitral valve and/or its suspension in two cases. Mitral valve insufficiency was proposed as a possible consequence; it was not considered the cause of death. The implantation of the valve could injure the conduction system. This was valid for the rupture. Death could be a possible outcome [14]. Therefore, the also reliable function of pacemakers was vital, which included the correct and stable position. Also AV-conduction blockage is a known serious complication, also in the living, not undergoing surgery [4, 23, 24]. Unfortunately, the imaging methods of today do not visualize the cardiac conduction system. This is also valid for PMCT, MPMCTA and even for postmortem magnetic resonance imaging (PMMRI). MPMCTA allows only in singular cases the visualization of a rupture of that cardiac tissue which contains the conduction system, and thus might permit to diagnose an interruption. This fact limits the attribution of cardiac deaths to injuries of the conducting system.
5
6
[19]
[20]
[21]
[22]
[23]
[24]
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery transcatheter aortic valve implantation versus surgical aortic valve replacement in patients with severe aortic stenosis. Am J Cardiol 2012;109:1487–93. Abdel-Wahab M, Zahn R, Horack M, Gerckens U, Schuler G, Sievert H et al. Aortic regurgitation after transcatheter aortic valve implantation: incidence and early outcome. Results from the German transcatheter aortic valve interventions registry. Heart 2011;97:899–906. Sinning JM, Hammerstingl C, Vasa-Nicotera M, Adenauer V, Lema Cachiguango SJ, Scheer AC et al. Aortic regurgitation index defines severity of peri-prosthetic regurgitation and predicts outcome in patients after transcatheter aortic valve implantation. J Am Coll Cardiol 2012;59: 1134–41. Rodés-Cabau J, Dumont E, Doyle D. ‘Valve-in-valve’ for the treatment of paravalvular leaks following transcatheter aortic valve implantation. Catheter Cardiovasc Interv 2009;74:1116–9. Takagi K, Latib A, Al-Lamee R, Mussardo M, Montorfano M, Maisano F et al. Predictors of moderate-to-severe paravalvular aortic regurgitation immediately after CoreValve implantation and the impact of postdilatation. Catheter Cardiovasc Interv 2011;78:432–43. Liang M, Devlin G, Pasupati S. The incidence of transcatheter aortic valve implantation-related heart block in self-expandable Medtronic CoreValve and balloon- expandable Edwards valves. J Invasive Cardiol 2012;24: 173–6. Roten L, Wenaweser P, Delacrétaz E, Heillige G, Stortecky S, Tanner H et al. Incidence and predictors of atrioventricular conduction impairment after transcatheter aortic valve implantation. Am J Cardiol 2010;106:1473–80.
APPENDIX. CONFERENCE DISCUSSION Scan to your mobile or go to http://www.oxfordjournals.org/page/6153/1 to search for the presentation on the EACTS library
Dr N. Al-Attar (Glasgow, UK): I have a number of questions and comments. You conclude that post-mortem CT and post-mortem CT angiography are useful in TAVI deaths to improve the knowledge and understanding of subclinical complications and consequences and to try to determine whether death was procedure-related. This is indeed quite true. Your series comprised 32 patients, but we don’t know how large this is in the overall experience in Hamburg. This was done over a five-year period and is a relatively small number of cases. There are a notable disproportionate number of women, and I would inquire as to the reason of this bias? In your results you state that bleeding is one of the major procedure-related lethal complications, but neither here nor in your paper do you give details regarding the source of bleeding. Likewise, there are no details on the sites of the perforation and rupture that you described. Your paper, undoubtedly, and your work help add to our understanding of TAVI and its potential hazards and complications, but I was wondering which patient would you send for a postmortem CT or postmortem CTA as compared to an autopsy. Dr Vogel: First you started with the number of patients. The patients were from the different heart centres in Hamburg and the neighbouring states of northern Germany. So we did the whole region of northern Germany.
The problem is that not every relative wants to have an autopsy. I mean, 60% of the patients got an autopsy, but the same problem is for MPMCTA: you need consent of the relatives, because it is quite invasive. So this is one limiting factor for the small numbers. I think you need a bigger number to get more information about serious complications after TAVI. We started to do a series of TAVI patients where we analysed every TAVI patient since the beginning of 2014, but before, we did it retrospectively. So we have a very small number of patients, this is the reason for that. The source of bleeding is quite different. We had a lot of ruptures of the vessels in the first years of 2008 to 2012, and afterwards we had new devices. So it was more a myocardial rupture and sometimes it is a rupture of the conduction system. You asked which deceased patient we should send. I think every patient, of course. It is a tool of quality control, and we do it in Hamburg and we try to do it with the other patients, but I think every centre should think about it. Dr T. Walther (Bad Nauheim, Germany): You said in your last sentence that you learn from every autopsy, and if you can do it in this way by CT, maybe less invasively than with an autopsy which relatives won’t like, then we can get some valuable information. I would like to know, just technically, how much time do you have if you want to do this angiography, because there is some coagulation in the dead body and so on. How much time do you have between the deaths until you can do that examination? Dr Vogel: The first three days. Dr Walther: And one other question. You showed some very interesting pictures. For example, the VSD could probably have been diagnosed by root angiography during the implantation or by echo, could you underline the value of your methods, put up a table and say, okay, this amount of diagnosis has been seen clinically already and we can prove it with CT, similar with autopsy, or this amount has not been seen beforehand with clinical methods before death of the patient. I think that would be very valuable for the manuscript and for the overall acceptance of this method. Dr Vogel: It wasn’t seen before. It was expected but it wasn’t proven before. Dr Walther: Well, this can happen, of course. Dr Vogel: I think that was the challenge. Dr D. Camboni (Regensburg, Germany): I had the same idea in our institution, and the problem is that it is impossible in our institution because we had to close the CT. So where do you have your CT and how do you do this practically? Dr Vogel: It is in the department of legal medicine. We try to investigate every deceased after cardiac surgery or intervention. We have a series of more than 500 patients now, but it is the region of northern Germany since 2008. Dr V. Bapat (London, UK): The TAVI patients can die because of non-TAVIrelated problems. So do you select these patients? Do you think this method is actually valuable if you are thinking of TAVI-related complications, because there is a risk that you are going to see a lot of things which may not be related to their death but which can actually raise more questions than give you answers? Dr Vogel: Yes, this is right. Most of the time you have more than one cause for death, so you have to differentiate, of course, and it is true for that. We try to get some more details to differentiate more. But you can say something about bleeding or severe complications. That is possible, sure. Dr Bapat: I think it is an excellent contribution to exclude complications with TAVI and at least erases those doubts from the minds of the relatives.
ORIGINAL ARTICLE
European Journal of Cardio-Thoracic Surgery (2015) 1–6 doi:10.1093/ejcts/ezv020
Post-mortem computed tomography and post-mortem computed tomography angiography following transcatheter aortic valve implantation† Beatrice Vogela,*, Axel Heinemannb, Helmut Gulbinsa, Hendrik Treedea, Hermann Reichenspurnera, Klaus Püschelb and Hermann Vogelb a b
Department of Cardiovascular Surgery, University Heart Center Hamburg, Hamburg, Germany Department of Legal Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany
* Corresponding author. Department of Cardiovascular Surgery, University Heart Center Hamburg Martinistraße 52; 20246 Hamburg, Germany. Tel: +49-40-69669349; fax: +49-40-5366862; e-mail: [email protected] (B. Vogel). Received 14 September 2014; received in revised form 25 December 2014; accepted 29 December 2014
Abstract OBJECTIVES: At present, transcatheter aortic valve implantation (TAVI) is widely used. As with any interventional treatment, however, TAVI may also be accompanied by complications and may result in periprocedural mortality. This study aims to evaluate such complications and causes of death after TAVI. METHODS: The study included 32 deceased (59.4% female, n = 19, median age: 82 years) patients with TAVI, since 2008, in whom postmortem computed tomography (PMCT) and PMCT angiography were performed with the intention of identifying complications. RESULTS: Altogether, we registered bleeding (28.1%, 9/32), perforation and rupture (25%, 8/32), cerebral infarction (18.8%, 6/32), injury of the conduction system (3.1%, 1/32), insufficiency of the aortic (12.5%, 4/32) and the mitral valve (9.4%, 3/32) and of valve-in-valve procedures (9.4%, 3/32). Furthermore, there were findings due to cardiopulmonary resuscitation and intensive care. PMCT and PMCT angiography has advantages over autopsy. The demonstration of bleeding vessels, ruptures, the position of the implanted aortic valve and its effects on the mitral valve and its suspensions were more easily accessible by computed tomography-imaging display than by customary autopsy photo-documentation. CONCLUSIONS: After TAVI, PMCT and PMCT angiography successfully demonstrated the complications leading to death. PMCT and PMCT angiography contribute to the post-mortem analysis of causes of periprocedural death. Keywords: Transcatheter aortic valve implantation • Complications • Post-mortem computed tomography • Multiphase PMCT angiography
INTRODUCTION Intra- or post-procedural complications in transcatheter aortic valve implantation (TAVI) patients may lead to rapid deterioration of the patient with subsequent death. In these cases, it is not always possible to clarify the reason in the clinical situation. For the development of these interventional strategies, however, detection and visualization of the complications, especially the affection of anatomical structures within the heart is of great importance. During the last years, post-mortem computed tomography (PMCT) has become an additional and reliable tool of the pathologist [1, 2]. PMCT guides the autopsy and furnishes additional information that cannot be achieved by conventional autopsy. † Presented at the 28th Annual Meeting of the European Association for CardioThoracic Surgery, Milan, Italy, 11–15 October 2014.
It has been proposed that PMCT could even replace autopsy. This proposal has induced a discussion about the priority of both methods. The consequence is a rapid progress of PMCT and the development of new tools. One of these new tools is PMCT angiography. PMCT angiography is a rather new diagnostic tool used in forensic pathology to examine the cardiovascular situation. This method allows the visualization of blood flow at the time of death. However, there is a limiting factor: PMCT angiography is invasive. Therefore, the relatives must give their consent. The potential and value of PMCT angiography for the detection of coronary disease has been analysed systematically by the group of Grabherr et al. [3]; they showed the value of PMCT angiography in the analysis of vascular and cardiac pathology. Therefore, it seemed likely to employ PMCT and PMCT angiography for analysing the outcome of TAVI. Furthermore, it seemed likely that an analysis with PMCT and PMCT angiography of the deceased, who had undergone TAVI, offered the chance to visualize a larger
© The Author 2015. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ADULT CARDIAC
Cite this article as: Vogel B, Heinemann A, Gulbins H, Treede H, Reichenspurner H, Püschel K et al. Post-mortem computed tomography and post-mortem computed tomography angiography following transcatheter aortic valve implantation. Eur J Cardiothorac Surg 2015; doi:10.1093/ejcts/ezv020.
2
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery
number of serious and deadly complications than an analysis of the patients also treated with TAVI who had survived the procedure. This leads to the following question: Which complications and causes of death can be found in those deceased, who had undergone TAVI, by applying PMCT and PMCT angiography?
MATERIALS AND METHODS Since 2008, PMCT has been performed in the Department of Legal Medicine of the University of Hamburg. Until 2014, more than 5000 deceased were examined. Searching the PMCT archive, we found 32 deceased with prior TAVI (13 men, 19 women, age: 68–92 years, median age: 82 years). Thirty had died as in-patients during their hospital stay directly following the procedure, i.e. up to 3 weeks after the intervention; 2 had died later (one after 3, the other after 6 months). The Department has its own CT scanner: Until 2013, a 4 Scan MDCT (Multi Detector CT, MX 8000, Philips) was used, followed by a 16 MDCT (Philips Imation) in 2013. The examination parameters for the whole body computed tomography (CT) were: 1 mm scan, pitch 1.5, 180–230 mAs, 130 kV. Each of those 32 patients underwent PMCT; 8 of them had an additional PMCT angiography with oily contrast substance, and 1 with air for contrast. PMCT angiography with contrast substance was done as a multiphase PMCT angiography following the method of Grabherr et al. [3]: multiphase PMCT angiography has an arterial, a venous and a dynamic phase. Primarily, 1200 ml of an oily perfusate was injected into the femoral artery (arterial phase), and after a few minutes (necessary to perform a whole body CT) an equal quantity was injected into a femoral vein (venous phase). Finally, in the dynamic phase, a special pump simulates the circulation. A whole-body CT was obtained after each of those phases [3]. Consent had been obtained for PMCT, multiphase PMCT angiography and autopsy (25/32).
RESULTS The analysis of these 32 cases allowed one to differentiate between specific procedure-related and non-procedure-related lethal and non-lethal complications (Table 1).
Procedure-related lethal complications The most frequent pathological finding was haemorrhage (9/32: Pleura 4, pericardium 2, inguinal and pelvis 2, multiple sources 2). Multiphase PMCT angiography visualized the source of bleeding in all examined cases (Fig. 1A and B). In conventional autopsy, however, it was not always possible to identify the bleeding vessel; this could only be suggested by interpreting indirect findings such as haematoma or haemorrhagic effusions. Perforation of blood vessels and rupture of the myocardium with damage to the cardiac circulatory system (8/32) were other lethal complications after TAVI. PMCT combined with multiphase PMCT angiography made the site of the interruption visible: In 1 case, the balloon-dilatation of the aortic valve had induced a rupture, which caused a shunt between the left ventricle and the right atrium. The patient suffered from electromechanical dissociation and went into cardiogenic shock requiring resuscitation. In this clinical setting, the injury had not been diagnosed. In PMCT alone, a rupture was
Table 1: Procedure-related and non-procedure-related complications n Procedure-related lethal complicationsa Bleeding (pleura 4 with mediastinum, pericardium 2 with mediastinum, subarachnoid space 1, inguinal and pelvis 2, multiple sources 2) Perforation and rupture (aortic ring 4, myocardium 1, aorta 1, pelvic artery 2) Injury of the conduction system (due to rupture) Procedure-related non-lethal complicationsa Aortic valve insufficiency (failure of the implanted valve 2, paravalvular leakage 2, treatment with valve-in-valve procedures 2) Embolism of the implanted valve into the aorta and implantation of a second valve Overlap of the coronary ostia Deformation of the mitral valve Non-procedure-related lethal and non-lethal complicationsa: Cerebral infarction 5 (4 non-lethal), multiple organ failure 14, rib fractures 24, due to futile resuscitation 26, pulmonary arterial embolism 1, tracheal tube in wrong position 3, splenic infarction 1 Cerebral infarction Lethal Non-lethal Multiple organ failure etc a
9
8 1 3
1 4 2
1 4 14
Multiple findings in 1 deceased were enlisted.
suspected; multiphase PMCT angiography clearly visualized the rupture and the shunt (Fig 2). The site of the rupture allowed one to deduce an injury of the electrical conduction system, and trauma to the AV-node was suggested, later confirmed by autopsy.
Procedure-related non-lethal complications Compared with static autopsy, one important advantage of multiphase PMCT angiography was the possibility for dynamic functional assessment of the implanted valves: multiphase PMCT angiography visualized aortic regurgitation in 2 cases. One persisting paravalvular leakage had occurred after an index procedure, and in 2 patients, aortic regurgitation had occurred after a coaxial ‘valve-in-valve’ procedure (1 Edwards Sapien valve inside a CoreValve, and 1 CoreValve placement following a conventional aortic valve replacement). During the arterial phase of the multiphase PMCT angiography, contrast substance propagated into the left ventricle. The contrast substance passed between the outer border of the valves’ rim and the inner surface of the aorta, which induced the diagnosis of paravalvular leakage (Fig. 3A and B). In addition to dynamic functional assessment of the aortic valve, also an analysis of coexisting mitral valve regurgitation was possible by multiphase PMCT angiography. The interaction between the new aortic valve with the mitral valve papillary ligaments and muscles on one side, and deformation of the mitral valve leaflets on the other were visualized. This information could not be obtained by PMCT alone; and even autopsy had its limits. Obviously, PMCT and multiphase PMCT angiography done together visualized more relevant findings than autopsy alone: PMCT alone and combined with multiphase PMCT angiography could also be displayed repeatedly, which facilitated evaluation under different diagnostic assumptions. This proved to be very
3
ADULT CARDIAC
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery
Figure 1: Bleeding and TAVI ( JenaValve). PMCT (A) and multiphase PMCT angiography (B). (A) Blood (with layer formation) in the left chest. (B) Intercostal artery with contrast substance extravasation. MIP: maximum intensity projection; PMCT: post-mortem computed tomography; TAVI: transcatheter aortic valve implantation; MPMCTA: multiphase PMCT angiography.
Figure 2: Rupture of the aortic ring. Subaortic shunt from the left ventricle to the right atrium. MPMCTA: multiphase PMCT angiography; MIP: maximum intensity projection.
important, when discussing the initially not always obvious cause of complications and possibilities for their avoidance. In 4 cases, PMCT and multiphase PMCT angiography made evident that the implanted valve reached or overlapped with the ostia of the coronary arteries; and all 4 patients had pre-existing three-vessel coronary artery disease. Thus, the overlap of the ostia was discussed as a possible cause for lethal coronary artery perfusion. The grid of the implanted valve overlapping the coronary artery ostium together with coronary artery stenosis was considered a risk. However, multiphase PMCT angiography showed that the overlapping valve did not block the passage of the contrast substance into the coronary artery. Nevertheless, one has to bear in mind that, in the living, an overlapping valve could impede the placement of a coronary artery catheter and stent insertion during such procedures, if necessary, later on. Autopsy did not provide additional findings.
Non-procedure-related lethal and non-lethal complications Five deceased showed recent cerebral infarctions in PMCT, which, however, were considered non-lethal. One deceased had an
infarction of the entire left hemisphere with marked oedema and cerebral herniation with incarceration, which was considered the cause of death. All deceased had calcifications of the coronary arteries, easily visible on PMCT. All deceased also investigated by multiphase PMCT angiography had three-vessel disease. On two occasions, autopsy revealed a myocardial infarction corresponding to areas of limited blood supply. Both, PMCT and autopsy made multiple-organ failure apparent; and none of these two diagnostic procedures was superior to the other. PMCT visualized generalized oedema, pleural and peritoneal effusion and also fluid in the scrotum. Additionally, signs of delayed ante mortem renal excretion of CT contrast material (injected ante mortem) were seen, such as dense contrast in the gall bladder, in the renal parenchyma and in the meninges. Therefore, the differential diagnosis of the lethal outcome also had to include an immediate or delayed lethal contrast material reaction. PMCT easily showed the typical pattern of rib fractures due to previous cardiopulmonary resuscitation. In 3 cases, the endotracheal tube ended in one main bronchus, and in 1 of these 3 cases, the tube’s balloon blocked the other main bronchus. PMCT offered an easy proof of these findings and had obvious advantages compared with autopsy, usually done after tube removal. Once, PMCT visualized fulminant pulmonary embolism: a contrast-enhanced thrombus was visible at the tip of a central venous catheter and a similar structure in the left pulmonary artery (Fig. 4A–C). Ante mortem, the pulmonary embolism was not at all suspected. This finding proved that the prior TAVI procedure itself was not the cause of death.
DISCUSSION In our study of deceased patients following TAVI, we were able to reproduce the good diagnostic results of PMCT and especially of PMCT angiography as described by Grabherr et al. [3] for deceased patients with coexisting coronary heart disease. PMCT showed haemorrhage, calcifications, the type of the implanted valve and coronary artery disease. Multiphase PMCT angiography provided additional information; it not only localized the bleeding vessel but also allowed statements about function and dysfunction of the implanted device. These statements apply to the following:
4
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery
Figure 3: Paravalvular leakage after a valve-in-valve procedure. (A) Contrast substance passes between the external limit of the valves and the inner surface of the aorta. PMCT/MPMCT angiography. Arterial phase. (B) Contrast substance propagates into the left ventricle. MPMCTA: multiphase PMCT angiography.
Figure 4: TAVI with embolic-like migration of the implanted valve into the aorta and implantation of a second valve. Thrombus at the tip of a central venous catheter and an embolus in the left pulmonary artery. (A) Contrast medium-enhanced thrombus at the tip of the central venous catheter. (B) Contrast medium-enhanced thrombus the left pulmonary artery. PMCT: post-mortem computed tomography.
the aortic valve by visualizing contrast substance between the implanted valve and the aortic wall; the mitral valve with its leaflets and its suspension; the coronary ostia with the overlap of the grid of the implanted valve; and the coronary arteries with stenosis and obstruction. Compared with autopsy, PMCT and multiphase PMCT angiography had clear advantages in detecting functional disorders, whereas autopsy with macro- and microscopic evaluation was advantageous in detecting structural organ damage. Guided by multiphase PMCT angiography, autopsy also determined the type of a localized vascular obstruction, and proved a suspected myocardial infarction. The main advantage of conventional autopsy, however, lies in obtaining tissue for histology, and close macroscopic organ-evaluation. Thus, the detection of infarction and other organic alterations became possible and confirmed the prior diagnostic suspicion. In our series, the accuracy for multiphase PMCT angiography in detecting cardiac causes of death was apparent; this corresponded to the literature [1–4]. A combination of PMCT and multiphase PMCT angiography with autopsy led to optimal results [2]. This was especially true when PMCT and multiphase PMCT angiography guided the autopsy. Our observations indicated that PMCT and MPMCTA also combined with autopsy to furnish better results and more information about lethal outcome after TAVI than either
method alone. Neither method can serve as a gold standard. Furthermore, our observations allowed the conclusion that PMCT alone, but better yet, in combination with multiphase PMCT angiography is a good alternative to investigate an in-patient’s lethal outcome following TAVI [1, 5], when autopsy cannot be performed. Haemorrhage was the most frequent complication (9 of 32). For 6 cases, it was due to procedural manipulations (catheter insertion, dilatation of the aortic valve), twice due to cardiopulmonary resuscitation with rib fractures and vascular rupture, and once due to anticoagulation with subsequent subarachnoid bleeding. This corresponds to the literature, which confirmed that, after TAVI, haemorrhage of any kind was a life-threatening complication [6–9]. Furthermore, our results supported previous statements that, after a transapical approach, haemorrhage is the most important predictor of a life-threatening situation [10, 11]. Only MPCTA was able to visualize the source of bleeding in all cases directly, thus leading to a definite diagnosis. This was a clear advantage compared with conventional autopsy, where sometimes bleeding could only be suspected by indirect findings, such as haemorrhagic effusions or haematoma. Ruptures appeared in 8 of 32 deceased. A rupture itself seems to indicate a life-threatening situation; this seems even more valid for a rupture into preformed potentially free spaces, like the
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery
CONCLUSION PMCT and multiphase PMCT angiography of the deceased after TAVI furnish information about procedure- and non-procedurerelated complications and thus make an important contribution in this field. Compared with the living TAVI patient, in the deceased more serious and lethal complications became obvious and allowed detailed and comparative analysis. PMCT combined with multiphase PMCT angiography produced results at least as good as those of autopsy. Furthermore, compared with autopsy alone, they had the big advantage of permanent repeated display. Furthermore, multiphase PMCT angiography allowed the visualization of functional pathologies. The combination of all three
tools provides the physician with more relevant information than conventional autopsy alone.
Conflict of interest: none declared.
REFERENCES [1] Westphal SE, Apitzsch J, Penzkofer T, Mahnken AH, Knüchel R. Virtual CT autopsy in clinical pathology: feasibility in clinical autopsies. Virchows Arch 2012;461:211–9. [2] Westphal SE, Apitzsch JC, Penzkofer T, Kuhl CK, Mahnken AH, Knüchel R. Contrast-enhanced postmortem computed tomography in clinical pathology: enhanced value of 20 clinical autopsies. Hum Pathol 2014;45: 1813–23. [3] Grabherr S, Grimm J, Dominguez A, Vanhaebost J, Mangin P. Advances in post- mortem CT-angiography. Br J Radiol 2014;87: doi: 10.1259/bjr. 20130488. [4] Rubin JM, Avanzas P, del Valle R, Renilla A, Ríos E, Calvo D et al. Atrioventricular conduction disturbance characterization in transcatheter aortic valve implantation with the CoreValve prosthesis. Circ Cardiovasc Interv 2011;4:280–6. [5] Wichmann D, Heinemann A, Weinberg C, Vogel H, Hoepker WW, Grabherr S et al. Virtual autopsy with multiphase postmortem computed tomographic angiography versus traditional medical autopsy to investigate unexpected deaths of hospitalized patients: a cohort study. Ann Intern Med 2014;160:534–41. [6] Eggebrecht H, Schmermund A, Voigtländer T, Kahlert P, Erbel R, Mehta RH. Risk of stroke after transcatheter aortic valve implantation (TAVI): a meta-analysis of 10,037 published patients. EuroIntervention 2012;8: 129–38. [7] Pilgrim T, Stortecky S, Luterbacher F, Windecker S, Wenaweser P. Transcatheter aortic valve implantation and bleeding: incidence, predictors and prognosis. J Thromb Thrombolysis 2013;35:456–62. [8] Stortecky S, Wenaweser P, Diehm N, Pilgrim T, Huber C, Rosskopf AB et al. Percutaneous management of vascular complications in patients undergoing transcatheter aortic valve implantation. JACC Cardiovasc Interv 2012;5:515–24. [9] Van Mieghem NM, Tchetche D, Chieffo A, Dumonteil N, Messika-Zeitoun Dvan der Boon RM et al. Incidence, predictors, and implications of access site complications with transfemoral transcatheter aortic valve implantation. Am J Cardiol 2012;110:1361–7. [10] Borz B, Durand E, Godin M, Tron C, Canville A, Litzler PY et al. Incidence, predictors and impact of bleeding after transcatheter aortic valve implantation using the balloon-expandable Edwards prosthesis. Heart 2013;99: 860–5. [11] Wong DR, Ye J, Cheung A, Webb JG, Carere RG, Lichtenstein SV. Technical considerations to avoid pitfalls during transapical aortic valve implantation. J Thorac Cardiovasc Surg 2010;140:196–202. [12] Martínez MI, García HG, Calvar JA. Interventricular septum rupture after transcatheter aortic valve implantation. Eur Heart J 2012;33:190. [13] Neragi-Miandoab S, Michler RE. A Review of Most Relevant Complications of Transcatheter Aortic Valve Implantation. ISRN Cardiol 2013; doi: 10.1155/2013/956252. [14] Haldenwang PL, Bechtel M, Schlömicher M, Lindstaedt M, Strauch JT. Annular rupture leading to fatal complications in an elderly patient with calcified aortic and mitral annulus undergoing transapical aortic valve implantation. Thorac Cardiovasc Surg 2013;61:425–7. [15] Ferrari E, Rizzo E, Sulzer C, von Segesser LK. Unexpected left ventricular free- wall rupture following an aortic catheter-valve implantation. Eur J Cardiothorac Surg 2010;37:242–4. [16] Pasic M, Unbehaun A, Dreysse S, Buz S, Drews T, Kukucka M et al. Rupture of the device landing zone during transcatheter aortic valve implantation; a life- threatening but treatable complication. Circ Cardiovas Interv 2012; 5:424–32. [17] Détaint D, Lepage L, Himbert D, Borchert E, Messika-Zeitoun D, Iung B et al. Determinants of significant paravalvular regurgitation after transcatheter aortic valve: implantation impact of device and annulus discongruence. JACC Cardiovasc Interv 2009;2:821–7. [18] Tamburino C, Barbanti M, Capodanno D, Mignosa C, Gentile M, Aruta P et al. Comparison of complications and outcomes to one year of
ADULT CARDIAC
pleural space (haematothorax) or the pericardium (haematopericardium with pericardial tamponade). One observation concerned a shunt connecting the left ventricle with the right atrium. Observations about ruptures creating intracardiac shunts were rare in the literature [12, 13]; therefore, we classified our case as exceptional. The ruptures into the mediastinum and into the chest were life-threatening [14–16]. This was valid, too, for one observation of a rupture of the iliac artery. Aortic valve insufficiency (3 of 32) also seemed to be a risk factor. This applied both to failure of the implanted valve and to paravalvular leakage. This corresponds with previous communications [17, 18]. Additionally, paravalvular leakage has been described as a significant predictor of 1-year mortality [19, 20]. Implanting a second valve into the first was an option in our series. The observations in our series support the statement that, in planning for a TAVI, all potential complications have to be considered [21, 22]. MPMCTA showed the ability to detect regurgitations of the implanted valves, valvular as well as paravalvular leakages. This was not possible with conventional autopsy or PMCT alone. Additionally, regurgitations of the other cardiac valves could also be visualized; thus, MPMCTA had the advantage to make apparent the functional status of the heart ante mortem except for myocardial contractility. The position of the implanted valve could indicate potential dysfunctions. The upper margin of the valve cusps could reach or even cover the coronary arteries (n = 4); an impairment of perfusion was a possible consequence. However, in our series, such an impaired perfusion was never considered to be the cause of death. Nevertheless, in the living, the selective angiography of the coronary arteries and the insertion of stents could become a problem [14]. The base of the valve deformed the upper leaflet of the mitral valve and/or its suspension in two cases. Mitral valve insufficiency was proposed as a possible consequence; it was not considered the cause of death. The implantation of the valve could injure the conduction system. This was valid for the rupture. Death could be a possible outcome [14]. Therefore, the also reliable function of pacemakers was vital, which included the correct and stable position. Also AV-conduction blockage is a known serious complication, also in the living, not undergoing surgery [4, 23, 24]. Unfortunately, the imaging methods of today do not visualize the cardiac conduction system. This is also valid for PMCT, MPMCTA and even for postmortem magnetic resonance imaging (PMMRI). MPMCTA allows only in singular cases the visualization of a rupture of that cardiac tissue which contains the conduction system, and thus might permit to diagnose an interruption. This fact limits the attribution of cardiac deaths to injuries of the conducting system.
5
6
[19]
[20]
[21]
[22]
[23]
[24]
B. Vogel et al. / European Journal of Cardio-Thoracic Surgery transcatheter aortic valve implantation versus surgical aortic valve replacement in patients with severe aortic stenosis. Am J Cardiol 2012;109:1487–93. Abdel-Wahab M, Zahn R, Horack M, Gerckens U, Schuler G, Sievert H et al. Aortic regurgitation after transcatheter aortic valve implantation: incidence and early outcome. Results from the German transcatheter aortic valve interventions registry. Heart 2011;97:899–906. Sinning JM, Hammerstingl C, Vasa-Nicotera M, Adenauer V, Lema Cachiguango SJ, Scheer AC et al. Aortic regurgitation index defines severity of peri-prosthetic regurgitation and predicts outcome in patients after transcatheter aortic valve implantation. J Am Coll Cardiol 2012;59: 1134–41. Rodés-Cabau J, Dumont E, Doyle D. ‘Valve-in-valve’ for the treatment of paravalvular leaks following transcatheter aortic valve implantation. Catheter Cardiovasc Interv 2009;74:1116–9. Takagi K, Latib A, Al-Lamee R, Mussardo M, Montorfano M, Maisano F et al. Predictors of moderate-to-severe paravalvular aortic regurgitation immediately after CoreValve implantation and the impact of postdilatation. Catheter Cardiovasc Interv 2011;78:432–43. Liang M, Devlin G, Pasupati S. The incidence of transcatheter aortic valve implantation-related heart block in self-expandable Medtronic CoreValve and balloon- expandable Edwards valves. J Invasive Cardiol 2012;24: 173–6. Roten L, Wenaweser P, Delacrétaz E, Heillige G, Stortecky S, Tanner H et al. Incidence and predictors of atrioventricular conduction impairment after transcatheter aortic valve implantation. Am J Cardiol 2010;106:1473–80.
APPENDIX. CONFERENCE DISCUSSION Scan to your mobile or go to http://www.oxfordjournals.org/page/6153/1 to search for the presentation on the EACTS library
Dr N. Al-Attar (Glasgow, UK): I have a number of questions and comments. You conclude that post-mortem CT and post-mortem CT angiography are useful in TAVI deaths to improve the knowledge and understanding of subclinical complications and consequences and to try to determine whether death was procedure-related. This is indeed quite true. Your series comprised 32 patients, but we don’t know how large this is in the overall experience in Hamburg. This was done over a five-year period and is a relatively small number of cases. There are a notable disproportionate number of women, and I would inquire as to the reason of this bias? In your results you state that bleeding is one of the major procedure-related lethal complications, but neither here nor in your paper do you give details regarding the source of bleeding. Likewise, there are no details on the sites of the perforation and rupture that you described. Your paper, undoubtedly, and your work help add to our understanding of TAVI and its potential hazards and complications, but I was wondering which patient would you send for a postmortem CT or postmortem CTA as compared to an autopsy. Dr Vogel: First you started with the number of patients. The patients were from the different heart centres in Hamburg and the neighbouring states of northern Germany. So we did the whole region of northern Germany.
The problem is that not every relative wants to have an autopsy. I mean, 60% of the patients got an autopsy, but the same problem is for MPMCTA: you need consent of the relatives, because it is quite invasive. So this is one limiting factor for the small numbers. I think you need a bigger number to get more information about serious complications after TAVI. We started to do a series of TAVI patients where we analysed every TAVI patient since the beginning of 2014, but before, we did it retrospectively. So we have a very small number of patients, this is the reason for that. The source of bleeding is quite different. We had a lot of ruptures of the vessels in the first years of 2008 to 2012, and afterwards we had new devices. So it was more a myocardial rupture and sometimes it is a rupture of the conduction system. You asked which deceased patient we should send. I think every patient, of course. It is a tool of quality control, and we do it in Hamburg and we try to do it with the other patients, but I think every centre should think about it. Dr T. Walther (Bad Nauheim, Germany): You said in your last sentence that you learn from every autopsy, and if you can do it in this way by CT, maybe less invasively than with an autopsy which relatives won’t like, then we can get some valuable information. I would like to know, just technically, how much time do you have if you want to do this angiography, because there is some coagulation in the dead body and so on. How much time do you have between the deaths until you can do that examination? Dr Vogel: The first three days. Dr Walther: And one other question. You showed some very interesting pictures. For example, the VSD could probably have been diagnosed by root angiography during the implantation or by echo, could you underline the value of your methods, put up a table and say, okay, this amount of diagnosis has been seen clinically already and we can prove it with CT, similar with autopsy, or this amount has not been seen beforehand with clinical methods before death of the patient. I think that would be very valuable for the manuscript and for the overall acceptance of this method. Dr Vogel: It wasn’t seen before. It was expected but it wasn’t proven before. Dr Walther: Well, this can happen, of course. Dr Vogel: I think that was the challenge. Dr D. Camboni (Regensburg, Germany): I had the same idea in our institution, and the problem is that it is impossible in our institution because we had to close the CT. So where do you have your CT and how do you do this practically? Dr Vogel: It is in the department of legal medicine. We try to investigate every deceased after cardiac surgery or intervention. We have a series of more than 500 patients now, but it is the region of northern Germany since 2008. Dr V. Bapat (London, UK): The TAVI patients can die because of non-TAVIrelated problems. So do you select these patients? Do you think this method is actually valuable if you are thinking of TAVI-related complications, because there is a risk that you are going to see a lot of things which may not be related to their death but which can actually raise more questions than give you answers? Dr Vogel: Yes, this is right. Most of the time you have more than one cause for death, so you have to differentiate, of course, and it is true for that. We try to get some more details to differentiate more. But you can say something about bleeding or severe complications. That is possible, sure. Dr Bapat: I think it is an excellent contribution to exclude complications with TAVI and at least erases those doubts from the minds of the relatives.
