numero rivista e pagine: HSR Proceedings in Intensive Care and Cardiovascular Anesthesia 2012; 4(2): 109-118
PDF version

Dilatation of the ascending aorta in patients with congenitally bicuspid aortic valves

Authors: F. Robicsek*1, R.F. Padera Jr2, M.J. Thubrikar3

1The Department of Thoracic and Cardiovascular Surgery, Carolinas Medical Center, Charlotte, North Carolina
2The Department of Pathology, Brigham and Women's Hospital, Boston, Massachusetts
3Thubrikar Aortic Valve, Inc., Rapid City, SD

Corresponding author: * Corresponding author:
Francis Robicsek, MD, PhD
1001 Blythe Blvd., Suite 300,
Charlotte, NC 28203
E-mail: frobicsek@carolinashealthcare.org



The cause of ascending aortic dilatation occurring in patients with congenitally bicuspid aortic valves was investigated.


Flow patterns through human aortic roots with congenitally bicuspid aortic valves as well as through porcine constricted aortas were studied in a left heart simulator. Vibration was recorded as a measure of turbulence in the post-stenotic segment. Histological changes in fetal aortas with isolated congenitally bicuspid aortic valves were compared to fetal aortas with congenitally bicuspid aortic valves and hypoplastic left hearts, as well as to normal fetal aortas with tricuspid aortic valves.


Congenitally bicuspid aortic valves were anatomically stenotic even in the absence of pressure gradients and without history of relevant symptoms. Histology of the aortic wall in isolated fetal congenitally bicuspid aortic valves was similar to that of fetal aortas with normal tri-leaflet aortic valves, but was abnormal if congenitally bicuspid aortic valves was associated with other cardiovascular anomalies. Flow studies revealed that turbulence and vibration in the post-stenotic aortic segments generated by the stenosis were proportional to the degree of the narrowing.


Congenitally bicuspid aortic valves are inherently stenotic, asymmetrical, generate turbulence and vibration. This not only leads to early failure but also to injury of the ascending aortic wall and ascending aortic dilatation. The more progressive form of ascending aortic dilatation occurs in patients where congenitally bicuspid aortic valves is combined with other inborn anomalies and may require a radical procedure (replacement).


Keywords: bicuspid aortic valve, aortic aneurysm


Congenitally bicuspid aortic valve (CBAV) is the most common inborn cardiovascular anomaly [1, 2] with a prevalence of 0.9% to 2% in the general population [3]. Besides the potential for the early degeneration of the valve itself, CBAV may also lead to ascending aortic dilatation (AAD), dissection and/or rupture [4, 5].
The cause of AAD associated with CBAV has been the subject of lively debate in which the participants may be divided into two principal groups:

  1. those who believe that AAD is caused by flow patterns induced by the abnormal valve [6,7,8,9,10];
  2. those who postulate that the AAD is caused by genetic faults which are also responsible for the CBAV itself [11, 12]. The principal support to the latter view is that pathological changes and abnormal gene expressions in the aortic wall are often present in patients with CBAV even in the absence of clinical [13] or hemodynamic [14] evidence of stenosis.

The purpose of this paper is to present data that reinforces our original stance [6,7,8] that the changes in the ascending aorta in patients with isolated CBAV are not inborn, but are caused by lifelong bombardment of the ascending aortic wall by turbulence. In patients with CBAV and associated congenital anomalies, this process may indeed be exaggerated by congenital weakness of the ascending aortic wall.



Functional morphology. In vitro experiments were carried out on three cryopreserved then thawed CBAVs obtained from human cadavers with no history of heart disease. Leaflet function was studied in a left-heart simulator (Vivitro Systems, Vancouver, Canada) primed with 38% glycerol, with flow rates set to 2.5-4.0 L/min, aortic pressure at 120/80 mmHg, and a pulse rate of 72 beats/min. The flow rates were measured using electro-magnetic probes (Carolina Medical Electronics). Pressure gradients were determined with appropriate transducers (Transpac IV, Abbott, IL). The function of the valves was visually observed and recorded using axially positioned cameras on both the aortic (500 frames/sec and 30 frames/sec) and ventricular aspects (30 frames/sec). The shape and size of the valve and leaflet movements were studied using intravascular ultrasound (IVUS) for the short axis and an external ultrasound probe for the long-axis views.
Flow Studies. Neutrally buoyant seeds were injected into the human CBAVs as well as through six fresh, non-valved porcine thoracic aortas. Using 2D external ultrasound, patterns of the reattachment of the flow-jet to the aortic wall and zones of turbulence were recorded. In the porcine aortas with approximately 20 mm diameters, different degrees of stenoses were created using sections of 3 mm wide Dacron tape. The diameter of the stenoses varied from 7 mm to 2 mm, with induced peak systolic gradients from 20 to 56 mmHg. The porcine aortas were also mounted in the left heart simulator and photographed using a 500 frames/sec cine-camera applying either pulsatile or non-pulsatile flow. Pressure and flow wave-forms were monitored [15]. The flow patterns through the CBAV were also studied by creating a computer model of bileaflet valves with eccentric orifices, different leaflet lengths and uneven commissure attachments. The model was constructed using Fluent software, as described previously [15].
Vibration Experiments. Vibration induced by and used as a measure of turbulence was studied in porcine aortas using a lightweight accelerometer (Model PCB 309A, PCB Piezotronics, Depew, NY) with the sensor glued to the surface of the aorta. The probe was placed 1, 2, 3, 5, and 7 cm distal to the stenosis respectively with the degree of stenosis unchanged. In a second series, the probe was left at each of the locations while the degree of stenosis was varied to create systolic pressure gradients of 20, 38, and 56 mmHg and compared to a background level of vibrations with no stenosis. The sensor output was 5 mV/g, where “g” represents acceleration due to gravity (9.8m/s2). The sampling frequency of the accelerometer was 1000 Hz. The frequency of the vibration was analyzed using fast-Fourier transform analysis (FFT).
Histology. Histological analysis of the ascending aortic wall was done in four fetal hearts with CBAV of gestational ages 19, 20, 21, and 25 weeks and compared to four age-matched controls with normal trileaflet aortic valves (TAV). Two of the four hearts with CBAV had no other anomalies; the other two had hypoplastic left hearts and mild ascending aortic hypoplasia.



In functional morphological studies, Valve A (10 y. o. male) when viewed from the aortic aspect, revealed one leaflet with a prominent raphe, and a prominence near the center of the coaptation line on the other leaflet. The presence of one large and one smaller sinus was obvious. The line of approximation was arched and off-center, resulting in an eccentric orifice. The shape of the fully open orifice was elliptical. Ultrasound imaging also revealed a curved line of closure. When measured via IVUS, the orifice area appeared moderately narrowed (5 – 10%). Valve B (24 y. o. male) had a tiny rudimentary cusp between the two leaflets. In the short axis view, at sinus level, the aorta was elliptical. The line of closure was straight, but the leaflet-coaptation was curved. The open orifice area measured by IVUS was about 75% of that of the annular orifice. Valve C (24 y. o. male), had conjoined leaflets acting as one fused by a prominent raphe and free edge longer than its counterpart (Figures 1 A, B).



Figure 1

Three congenitally bicuspid aortic roots obtained from young individuals who had no symptoms of heart disease and died of non-cardiac causes shown in full diastole (a) and in full systole (b). None of the roots had any measurable trans-valvular pressure gradients. Note the morphological stenosis of various degrees in every specimen. Casts obtained in full diastole (c) shows creases and wrinkles in full diastole.


Both cusps were thin and pliable, with exception of the area toward the raphe. This caused radial shortening. The attachment for the conjoined cusp was flat and scalloped for the other. Measured by IVUS, the orifice was about 55% of the annular area (Figure 2).



Figure 2

Valve C. Ultrasound image showing asymmetrical, elliptical and narrow (57% of the annulus area) leaflet opening.


There was no measurable pressure gradient through any of the valves.
These observations confirmed the wide range of anatomy of the CBAV, especially the asymmetrical geometry of the sinuses. The eccentricity and the stenotic nature of the CBAV are evident even in the absence of clinical symptoms and measurable pressure gradients. Visualization of flow patterns in constricted porcine aortas showed that the flow jet exits the stenosis, creating turbulence and recirculation near the reattachment points to the aortic wall (Figures 3 A, B, C).



Figure 3

2D ultrasound images of flow through a vascular stenosis show:

  1. The initial jet flow through the stenosis.
  2. The development of a recirculation zone.
  3. The region of reattachment to the aortic wall.


 The tighter the stenosis, the further downstream was the reattachment of flow. Similar turbulence could also be seen in the glass model and was also demonstrated in our earlier experiments (Figure 4).



Figure 4

Extension of "functional" stenosis past the post-stenotic area (a) and formation of eddy-currents indicated by residual dye in a hydraulic model. Stenosis with post-stenotic dilatation (c) also shows extended "functional stenosis and increase in eddy-currents and turbulence" (from Robicsek F, et al: The post-stenotic dilatation of great vessels Acta Med Scand 1954; 155: 481-484).


Compared to normal TAV aortic valves of similar annular dimensions, the orifice of the digital CBAV model was about 50% smaller. By varying the degree of eccentricity, we observed that the more eccentric the model, the more the flow path deviated from the center resulting in increased stress on the longer leaflet.
Under nonpulsatile flow conditions with the CBAV fully opened, the orifice was not only asymmetrical, but also stenotic. This created recirculation of eddy-currents with peak velocity concentrated at the leaflets’ free edge. Unlike a normal TAV, the eddy-currents were not restricted to the sinuses, but extended distally, and were directed toward the right (Figure 5).



Figure 5

Symmetrical opening (shaded, clover-shaped area showing maximal leaflet opening) of the trileaflet aortic valves with three even sinuses and leaflets expanding beyond the circumference of the annulus (dotted reversed cone) and the asymmetrical and incomplete (shaded ellipse) opening of the congenitally bicuspid aortic valve with two uneven sinuses. In the latter, the flow-jet impacts the aortic wall.

TAV = trileaflet aortic valves ; CBAV = congenitally bicuspid aortic valve


Notably, abnormal flow patterns also occur in patients with stenosis of their initially normal TAV. It was reported that in these patients the resulting AAD is symmetrical, “pear” or “teardrop” shaped, while in patients with CBAV, the AAD is convex toward the right antero-lateral aspect, where degenerative media changes were also found to be the most severe [16]. These observations agree with our own, in which we found stress-overload at the same location in patients with “normally” functioning CBAVs with no pressure gradients or regurgitation [15].
Vibrations were the highest 2-3 cm distal to the stenosis, and diminished at a distance of about 5 cm. The dominant frequency was about 110 Hz within a wide range of 20-300 Hz. Higher frequencies were also found, but with less intensity. The amplitude increased proportionally with the pressure gradient.
The higher was the pressure gradient, the further downstream was the maximum amplitude of vibration (Figure 6).



Figure 6

Vibration recordings at 1, 2, 3, and 5 cm from the stenosis. Strongest signal occurs at a frequency of 110 Hz. Vibration is higher at 1 and 2 cm from the stenosis when the pressure gradient is 38 mmHg. Vibration is higher at 3 and 5 cm when the gradient is 56 mmHg.


The intensity of vibrations was highest during systolic ejection. Proportional to increased pressure gradient, the point of reattachment of the flow jet, considered to be the force generating vibration, moved distally, indicating that maximum vibration occurs at the site of reattachment of the flow-jet to the aortic wall. With change in pressure and velocity, the point of reattachment moved back and forth, defining the length of aortic segment that was exposed to the most intense vibration (Figures 7-9).



Figure 7

Different degrees of aortic stenoses, formation of eddy currents, flow reattachment regions and different magnitudes of vibration, leading to development of poststenotic dilatation.



Figure 8

Digital re-creation of the flow pattern through bicuspid aortic root and aortic arch.



Figure 9

Shear-stress values measured in an aortic arch with a bicuspid aortic valve. Note the high values on the right-lateral aspect of the ascending aorta.


Histological studies of fetal aortas with normal TAV showed normal microscopic anatomy (Figure 8).
Histological pictures of the ascending aortas from tri-leaflet and hearts with fetal hearts with CBAV, but no other cardiovascular anomalies, were also within normal range (Figure 10,11).



Figure 10

Ascending aorta of a fetus with normal tri-leaflet aortic valve. Note the orderly arrangement of smooth muscle cells of the media, thin intima and diffuse, homogeneous elastin staining throughout the entire thickness of the media. 2.5 X magnification.

  1. Hematoxylin eosin stain.
  2. Movat pentachrome stain.



Figure 11

Ascending aorta of a fetus with congenitally bicuspid aortic valve as the only cardiovascular abnormality. It represents a histological picture similar to those of a fetal aorta with a normal tri-leaflet aortic valve. 10 X magnification.

  1. Hematoxylin eosin stain.
  2. Movat pentachrome stain.


The ascending aortas from the hearts with CBAV and hypoplastic left hearts showed mild intimal thickening, medial hypertrophy and focal elastic tissue disorganization (Figure 12).



Figure 12

Ascending aortic wall of a fetus with congenitally bicuspid aortic valve and hypoplastic left heart syndrome. Note the relative thickening of the aortic wall and focally disordered architecture on both the H&E and elastin stains in an area of relative stenosis of the aorta. 10 X magnification.

  1. Hematoxylin eosin stain.
  2. Movat pentachrome stain.


The distribution of the abnormalities seen in aortic media towards the intimal surface was suggestive that abnormal luminal fluid dynamics contributed to these changes. Microscopic and histological examination of the ascending aorta of a 5 year old boy with CABG, but no other cardiac anomalies showed intimal damage and disintegration of the elastic elements of the ascending aorta (Figure 13).



Figure 13

Ascending aorta of a 5 y.o. boy with congenitally bicuspid valve as the only cardiovascular anomaly. Note the intimal damage in the aortic root caused by turbulent flow (a) and the disintegration of the elastic elements in the ascending aortic wall (b) 25X magnification.

Movat pentachrome stain.



The curious phenomenon known as post-stenotic dilatation of arteries was already illustrated in the sixteenth century by the Italian anatomist Giuseppe Morgagni [17] and described in detail by Paget [18]; however, it was not studied clinically until 1916 when Halsted described post-stenotic dilatation caused by compression of the subclavian artery by the first rib, and also suggested that it may be caused by “paralysis of the vasomotor nerves and the occlusion of the vasa vasorum” [19]. Holman, in 1954, forwarded the theory that post-stenotic dilatation is caused by “deceleration of fluid, having passed through a constricted orifice, [which] results in a localized pressure increase” (Bernoulli’s law) [9].

Simultaneously, with his presentation we have demonstrated in both in vitro experiments and in clinical observations [6, 7] that blood streaming through stenoses does not decelerate in the immediate post-stenotic segment, but further down; ergo propter, and that there is no localized pressure increase which would explain the dilatation. We also demonstrated the birth of eddy-currents, turbulence and disintegration of elastic elements in the aortic media in the post-stenotic segment. From these findings, we have concluded that post-stenotic dilatation of the great vessels is caused by wall damage inflicted by the turbulent flow of blood streaming through stenotic areas [6, 7].
Since then our theory has been challenged by several authors who identified genetic rather than hemodynamic factors as a cause of AAD in patients with CBAV. One argument to support this view was that AAD and aortic wall changes are often seen in patients with other cardiovascular anomalies [19]. Another argument is that patients with CBAV may have anomalies not only in the wall of the ascending aorta, but also in the pulmonary artery. Yet, another argument in favor of the genetic theory is that changes in the ascending aortic wall resembling connective tissue disorders are frequently found in patients with CBAVs and AAD [21, 22].
It also has been suggested that apoptosis [23] and an increase in matrix metalloproteinases may also contribute to the development of AAD [19, 20, 24,25,26,27].
We acknowledge the above findings, but not the conclusions drawn. Hereditary tissue disorders, infection, atherosclerosis and most importantly, shear-stress of any origin [23, 28] may cause similar, if not identical, histological and biochemical changes of the aortic wall [29, 30]. Thus, the argument that aortic wall changes in adult patients with CBAV resemble those with inborn connective tissue disorders does not prove that these are congenital. In our studies, we were able to differentiate two forms of CBAV: patients with isolated CBAV and those with CBAV combined with other anomalies. In the former group, we found aortic wall changes to be absent at birth, indicating that in these patients AAD develops later and is due solely to turbulent flow. In the second group, with associated other inborn anomalies, the aortic wall changes are already present in the newborn, and may be exacerbated later by turbulence.
The question arises as to how issues of pathogenesis may influence the selection of particular surgical interventions on the ascending aorta in patients with isolated CBAV? Past studies [30] showed that patients with CBAV frequently develop dissection, even if AAD is moderate. Therefore, we agree with the recommendations that the issue of AAD should be surgically addressed if patients, especially those with isolated CBAV, are in need of aortic valve replacement.
Several surgeons, however, who believe that the wall of the ascending aorta is genetically defective in all patients with CBAV, also recommend that patients who are operated upon for symptomatic CBAV should undergo prophylactic ascending aortic replacement, even in absence of ascending aortic dilatation.

Others would not replace ascending aortas of normal size, but would graft those that are only moderately (4-4.5 cm in diameter) dilated [11, 32, 33].



We demonstrated that ascending aortic wall damages in patients with isolated CBAVs are absent at birth, but develop in later life due to turbulent flow rather than due to inborn connective tissue abnormality. In dilations of a moderate to medium degree, less radical approaches such as aortoplasty, which carries a lower mortality and morbidity rate [34], should be considered [11, 35].

One should also emphasize the key factors of a properly performed aortoplasty, such as a reduction in size to normal (about 3.5 cm in diameter) and, if external reinforcement is used, the wrap should be anchored at the sinus level and extended past the origin of the innominate artery [11, 35]. Dilated aortas >6 cm in diameter should be replaced regardless of presence or absence of associated anomalies.



  1. Roberts W C. The congenitally bicuspid aortic valve. A study of 85 autopsy cases Am J Cardiol 1970; 26: 72-83.
  2. Roberts W C. The structure of the aortic valve in clinically isolated aortic stenosis: an autopsy study of 162 patients over 15 years of age. Circulation 1970; 42: 91-97.
  3. Larson E W, Edwards W D. Risk factors for aortic dissection: a necropsy study of 161 cases. Am J Cardiol 1984; 53: 849-855.
  4. Stewart W J, King M E, Gilliam L D, et alii. Prevalence of aortic valve prolapse with bicuspid aortic valve and its relation to aortic regurgitation: a cross-sectional echocardiographic study. Am J Cardiol 1984; 54: 1277-82. 1984; 54: 1277-1282.
  5. Kuralay E, Demirkilic U, Ozal E, et alii. Surgical approach to ascending aorta in bicuspid aortic valve. J Card Surg 2003; 18: 173-80. 2003; 18: 173-180.
  6. Robicsek F. The post-stenotic dilatation. Acta Med Acad Sci 1954; Suppl.: 1-4.
  7. Robicsek F. Post-stenotic dilatation of the great vessels. Acta Med Scand 1955; 151: 481-485.
  8. Robicsek F, Sanger P W, Taylor F H, et alii. Pathogenesis and significance of post-stenotic dilatation in great vessels. Ann Surg 1958; 147: 835-844.
  9. Holman E. The obscure physiology of poststenotic dilatation; its relation to the development of aneurysms. J Thorac Surg 1954; 28: 109-133.
  10. Jarchow B H, Kincaid O W. Poststenotic dilatation of the ascending aorta: Its occurrence and significance as a roentgenologic sign of aortic stenosis. Proc Staff Meet Mayo Clin 1961; 36: 23-33.
  11. Borger M A, Preston M, Ivanov J J, et alii. Should the ascending aorta be replaced more frequently in patients with bicuspid aortic valve disease? J J Thorac Cardiovasc Surg 2004; 128: 677-683.
  12. Lindsay J Jr. Coarctation of the aortic, bicuspid valve and abnormal ascending aortic wall. Am J Cardiol 1988; 61: 182-184.
  13. Keane M G, Wiegers S E, Plappert T, et alii. Bicuspid aortic valves are associated with aortic dilatation out of proportion to coexistent valvular lesions. Circulation 2000; 102: III35-39. 19 Suppl.
  14. Nistri S, Sorbo M D, Marin M, et alii. Aortic root dilatation in young men with normally functioning bicuspid valves. Heart 1999; 82: 19-22.
  15. Robicsek F, Thubrikar M J, Cook J W, Fowler B. The congenitally bicuspid aortic valve: How does it function? Why does it fail? Ann Thorac Surg 2004; 77: 177-85. 2004; 77: 177-185.
  16. Cotrufo M, Agozzino L, De Feo M, et alii. Aortic valve dysfunction and dilated ascending aorta: A complex and controversial association. Ital Heart J 2003; 4: 589-595.
  17. Thiene G. G.B. Morgagni: De sedibus et causis morborum per anatomen idagatis. G Ital Cardiol 1985; 15: 1002-1003.
  18. Paget J. On obstruction of branches of the pulmonary artery. Med-Chir transcr. 1844; 27: 162-166.
  19. Halsted W S. An experimental study of circumscribed dilatation of an artery immediately distal to a partially occluding, and it bearing on the dilatation of the subclavian artery observed in certain cases of cervical rib. J Exp Med 1916; 24: 271-286.
  20. de Sa M, Moshkovitz Y, Butany J, David T E. Histologic abnormalities of the ascending aorta and pulmonary trunk in patients with bicuspid aortic valve disease: clinical relevance to the Ross procedure. J Thorac Cardiovasc Surg 1999; 118: 588-94. 1999; 118: 588-594.
  21. Niwa K, Perloff J K, Bhuta S M, et alii. Structural abnormalities of great arterial walls in congenital heart disease: light and electron microscopic analyses. Circulation 2001; 103: 393-400.
  22. Erdheim J. Medionecrosis aortae idiopathecia cystica. Virchows Arch 1930; 2138-2143.
  23. Bonderman D, Gharehbaghi-Schnell E, Wollenek G, et alii. Mechanism underlying aortic dilatation in congenital aortic valve malformation. Circulation 1999; 99: 2138-43. 1999; 99: 2138-2143.
  24. Francke U, Berg M A, Tynan K, et alii. A Gly1127Ser mutation in an EGF-like domain of the fibrillin-1 gene is a risk factor for ascending aortic aneurysm and dissection. Am J Hum Genet 1995; 56: 1287-1296.
  25. Sakomura Y, Yoshioka S, Ishizuka N, et alii. Apoptosis of medial smooth muscle in dilated ascending aortas in patients with Marfan’s syndrome and annulo-ectasia. Circulation 1997; 96: ii126-Suppl. 1.
  26. Boyum J, Fellinger E K, Schmoker J D, et alii. Matrix metalloproteinase activity in thoracic aortic aneurysms associated with bicuspid and tricuspid aortic valves. J Thorac Cardiovasc Surg 2004; 127: 686-691.
  27. Koullias G J, Korkolis D P, Ravichandran P, et alii. Tissue microarray detection of matrix metalloproteinases, in diseased tricuspid and bicuspid aortic valves with or without pathology of the ascending aorta. Eur J Cardiothorac Surg 2004; 26: 1098-1103.
  28. McKusick V A, Logue R B, Bahnson H T. Association of aortic valvular disease and cystic medial necrosis of the ascending aorta; report of four instances. Circulation 1957; 16: 188-194.
  29. Bluestein D, Rombod E, Gharib M. Vortex shedding as a mechanism for free emboli formation in mechanical heart valves. J Biomed Eng 2000; 122: 125-134.
  30. Carlson R G, Lillehei C W, Edwards J E. Cystic medial necrosis of the ascending aorta in relation to age and hypertension. Am J Cardiol 1970; 25: 411-415.
  31. Russo C F, Mazzetti S, Garatti A, et alii. Aortic complications after bicuspid aortic valve replacement: long-term results. Ann Thorac Surg 2002; 74: 1773-1776.
  32. Fedak P W, Verma S, David T E, et alii. Clinical and pathophysiological implications of a bicuspid aortic valve. Circulation 2002; 106: 900-904.
  33. Fedak P W M, Verma S, David T E, et alii. Vascular matrix remodeling in patients with bicuspid aortic valve malformations: implications for aortic dilatation. J Thorac Cardiovasc Surg 2003; 126: 797-806.
  34. Robicsek F, Cook J W, Reames M K Sr, Skipper E R. Size reduction ascending aortoplasty: is it dead or alive? J Thorac Cardiovasc Surg 2004; 128: 565-570.
  35. Lytle B. Should the ascending aorta be replaced more frequently in patients with bicuspid valve disease? J Thorac Cardiovasc 2004; 128: 677-683. In discussion to Borger MA, Preston M, Ivanov J, et al. .

Cite as: Robicsek F, Padera RF Jr, Thubrikar MJ. Dilatation of the ascending aorta in patients with congenitally bicuspid aortic valves. HSR Proceedings in Intensive Care and Cardiovascular Anesthesia 2012; 4 (2): 109-118


Acknoledgments: We thank Anne Gale for carefully copy-editing the manuscript.


Source of Support: Nil.


Conflict of interest: None declared.


Presented at the 2nd Expert Forum of the Roland Hetzer International Cardiothoracic and Vascular Surgery Society, 11 February 2012, Freiburg, Germany