Author + information
- Received February 14, 2018
- Revision received June 14, 2018
- Accepted June 19, 2018
- Published online August 27, 2018.
- Cemil Izgi, MDa,
- Simon Newsome, MScb,
- Francisco Alpendurada, PhDa,c,
- Eva Nyktari, MDa,
- Maria Boutsikou, PhDa,
- John Pepper, MDc,
- Tom Treasure, MDd and
- Raad Mohiaddin, PhDa,c,∗ (, )@IzgiCemil@UCL@LSHTM@RBandH
- aCardiovascular Research Center & Cardiovascular Magnetic Resonance Unit, Royal Brompton Hospital, London, United Kingdom
- bDepartment of Medical Statistics, London School of Hygiene and Tropical Medicine, London, United Kingdom
- cNational Heart & Lung Institute, Imperial College, London, United Kingdom
- dClinical Operational Research Unit, University College London, London, United Kingdom
- ↵∗Address for correspondence:
Prof. Raad Mohiaddin, Cardiovascular Magnetic Resonance Unit, Royal Brompton Hospital, National Heart and Lung Institute, Imperial College, Sydney Street, London SW3 6NP, United Kingdom.
Background Personalized external aortic root support (PEARS) was introduced in 2004 for prevention of aortic root dilatation in Marfan patients. The individual's aortic root is replicated by 3-dimensional printing. A polymer mesh sleeve is manufactured, which is implanted with the aim to support and stabilize the aortic wall.
Objectives The aim of this study was to assess effectiveness of PEARS for prevention of aortic root dilatation in Marfan patients.
Methods A total of 24 consecutive Marfan patients operated 2004 to 2012 were prospectively monitored with magnetic resonance imaging. Following a pre-defined protocol, baseline and follow-up aorta measurements were made in a blinded random sequence.
Results The mean age of the patients was 33 ± 13.3 years (range: 16 to 58 years), and the mean aortic root diameter was 45 ± 2.8 mm (range: 41 to 52 mm). Follow-up was 6.3 ± 2.6 years. There was no increase in the aortic root and ascending aorta diameters, but there was a tendency toward reduction: annulus diameter 28.9 ± 2.3 mm to 28.5 ± 2.4 mm (change −0.39 mm, 95% confidence interval [CI]: −1.05 to 0.27 mm), sinus of Valsalva diameter 44.9 ± 2.9 mm to 44.5 ± 3.0 mm (change −0.37 mm, 95% CI: −1.23 to 0.51 mm), and ascending aorta diameter 32.4 ± 3.6 mm to 32.3 ± 3.7 mm (change −0.10 mm, 95% CI: −0.92 to 0.74 mm). In the same period, the descending aorta diameter increased from 22.9 ± 2.4 mm to 24.2 ± 3.0 mm (change 1.32 mm, 95% CI: 0.70 to 1.94 mm; p < 0.001) with a tendency toward increase in aortic arch diameter 24.1 ± 2.0 mm to 24.5 ± 2.8 mm (change 0.41 mm, 95% CI: −0.56 to 1.37 mm).
Conclusions PEARS is effective in stabilizing the aortic root and preventing its dilatation. It is a viable alternative for prevention of aortic root dissection in Marfan patients.
Dissection and rupture of the ascending aorta is the main cause of mortality in Marfan syndrome patients (1). The risk of dissection progressively increases with increasing aortic root size (1,2). Current approaches for prevention of aortic dissection in Marfan patients have been focused on prevention of aortic root dilatation with drugs and prophylactic surgery to replace the aortic root (1,2). The drug therapy with beta-blockers and angiotensin receptor blockers is aimed at slowing the rate of aortic dilatation. Although the drugs might slow the rate of dilatation, the aorta still dilates, and Marfan patients eventually undergo root replacement when a guideline-directed diameter threshold is reached (1).
Prophylactic replacement of the aortic root aims at replacing the vulnerable aortic root prone to dissection with a prosthetic graft. The techniques include replacement of the aortic root and the aortic valve with a composite valve conduit (Bentall operation) and the valve-sparing root replacement (VSRR) surgery (3). Bentall operation is now a straightforward technique with a low operative mortality; however, it involves replacement of the aortic valve, which is competent in most of the patients. This introduces mandatory lifelong anticoagulation if a mechanical valve is used or risk of more frequent reoperation if a bioprosthetic valve is used. The cumulative lifetime risk of prosthetic valve complications is substantial for younger patients having these operations. The VSRR techniques allow replacement of only the root with preservation of the native aortic valve. Therefore, the risk of thromboembolism and endocarditis are reduced; however, there are other concerns with VSRR. These techniques are technically more challenging with less standardization among centers. Apart from intraoperative challenges, there is a risk of significant aortic regurgitation and reoperation with VSRR (4). A recent multicenter study involving patients from centers with substantial expertise reported a 7% rate of significant aortic regurgitation at 1-year follow-up after VSRR in Marfan patients, suggesting that real-life concerns about durability of these procedures are valid (5). Moreover, sparing of the aortic valve may not always be possible, and there is probability of intraoperative conversion to aortic valve replacement.
Personalized external aortic root support (PEARS) surgery has been developed as an alternative surgical method to prevent dilatation of the aortic root in Marfan patients (6,7). It involves surgical implantation of an individualized mesh support around the aortic root and the ascending aorta. With 3-dimensional printing utilizing the imaging data, a plastic model of the aortic root unique to each patient is produced. This plastic model is then used as a form upon which the mesh is produced, which fits perfectly to each individual patient's aortic root shape. It is essential that this is placed proximal to the coronary arteries and reaches and is secured to the aortoventricular junction. Cardiopulmonary bypass is typically not needed during the surgery (8). The native aortic valve and the blood endothelium interface are preserved, and therefore, the prosthetic valve–related complications of the Bentall procedure and technical challenges and the risk of aortic regurgitation of VSRR are ameliorated. These advantages would allow operating early in the natural history of the disease when the aortic root diameter lies between 40 and 50 mm. This would alleviate the significant anxiety that Marfan patients are facing during the long duration of watchful monitoring until their aorta reaches the current size threshold recommended in the guidelines for aortic root replacement. The critical question is whether PEARS can be reproducibly implanted to restrain all the proximal aorta segments and is really effective in preventing aortic root dilatation with the ultimate target of eliminating the risk of dissection.
We have previously shown perioperative and procedural advantages of PEARS compared with root replacement (8) and the favorable clinical outcome of the patients during follow-up (9). We also showed in the preliminary reports of the technique that it keeps the aortic root size stable (10). The target in PEARS as a prophylactic operation is to stabilize the aortic root and prevent its dilatation, because the size of the aortic root is currently regarded as the main factor determining the risk of dissection. The aim of the present study is to assess the medium term effectiveness of the PEARS on prevention of aortic root dilatation.
Patients and the PEARS surgery
Following the conceptual and technical development period that was reported previously (6,7), PEARS surgery was first performed in 2004 following approval by the Royal Brompton Hospital Research and Ethics Committee (6). To date, >100 patients with Marfan syndrome have undergone this surgery, which is now offered by several other centers across Europe where the PEARS mesh support is commercially available (ExoVasc Personalized External Aortic Root Support, Exstent Limited, Tewkesbury, United Kingdom). The present study involves prospective data from the series of the first 27 consecutive Marfan patients who had PEARS operation for prevention of proximal aorta dilatation and dissection between May 2004 and July 2012 during the evaluation phase of this new surgical technique at the Royal Brompton Hospital. These patients had close follow-up of their aorta size to monitor the effectiveness of PEARS.
All patients were diagnosed with Marfan syndrome according to Ghent criteria. They were recruited from the aortopathy clinic of the hospital, which has a well-established aortic surgery program, with both the Bentall and the VSRR surgeries routinely being performed. Eligibility criteria for PEARS were an aortic root size of 40 to 55 mm and no or only mild aortic regurgitation (8). PEARS was developed as a prophylactic surgery to prevent dilatation of the aortic root at an early point in the natural history of Marfan syndrome. Therefore, as approved by the ethics committee, the root size to offer PEARS was lower than the recommended cutoff size for aortic root replacement, and 40 mm was adopted as the lower end of the range of aortic root size for PEARS. A diameter of 40 mm and above in an adult is generally agreed to be pathological. All Marfan patients satisfying the eligibility criteria were considered for the operation, and none were declined on the basis of any pre-specified criteria. The patients were fully informed that PEARS is a novel approach for prevention of aortic root dilatation and dissection in Marfan syndrome in a detailed discussion with the operating surgeon (J.P.). They were also fully informed about the Bentall and VSRR options. All patients provided written informed consent for the PEARS surgery. Technical aspects of the PEARS procedure has been published before (6,7,11).
This prospective study is aimed to test stability of the aortic root size following PEARS surgery based on measurements by cardiovascular magnetic resonance (CMR) (Figure 1). The study was registered and approved as a clinical audit by the Quality and Safety Department of the Royal Brompton Hospital. Three of the 27 patients were excluded; in 2 of them, the baseline and follow-up imaging was by computed tomography (metallic spinal roads causing significant artefacts in 1 patient, and severe claustrophobia in the other patient precluded imaging by CMR), and the follow-up imaging studies for the third patient were not available because the patient was living abroad, and his follow-up was not in our hospital. The remaining 24 patients had CMR examinations in our center before the operation, at 6 and 12 months after the operation, and wherever possible annually thereafter. In 2 of these 24 patients, the PEARS mesh was not fully covering the aortic root due to identifiable technical failures, and the aortic root in these patients with incomplete support dilated at these sites at clinical follow-up. These patients were taken as outliers. One of these 2 patients had intractable hypotension in the recovery ward after the operation and noted to have ST-segment changes in the inferior electrocardiogram leads along with hypokinesia of the right ventricle on the echocardiogram. Ischemia with compromise of right coronary flow was suspected. Therefore, the chest was urgently reopened, and the seam of the mesh support was opened to release any possible impingement on the coronary arteries. Normal laminar flow in both coronary ostia was confirmed with echo. The electrocardiogram changes immediately resolved with hemodynamic stability post-operatively. The aortic root dilated in the uncovered area at follow-up. In the second patient, there was localized dilatation of the right coronary cusp of the aortic root in the region of the right coronary ostia where the opening for the coronary ostia was inadvertently cut large, leaving this region not adequately supported. Pre-operative and follow-up aorta size comparison was repeated by both including and excluding (per protocol analysis) these 2 outlier patients.
Measurements of thoracic aorta size
CMR examinations were performed on 1.5-T scanners. For aorta measurements a batch of 120 anonymized CMR studies was formed. This included the baseline and the latest CMR studies of all the 24 patients as well as randomly selected studies of the patients acquired at any time during their follow-up to dilute the batch in an attempt to minimize any possible measurement bias. A single operator (C.I.) measured the aorta size on these individual anonymized studies following a stringent, pre-defined protocol. The following measurements were made: aortic annulus diameter, 3 diameters of the sinus of Valsalva measured at cusp to commissure (the largest and the mean of these 3 measurements used for comparison), ascending aorta diameter, aortic arch diameter, and descending aorta diameter. Additionally, cross-sectional area measurements were taken at the sinus of Valsalva, ascending aorta, and descending aorta to detect any asymmetrical changes in size such as a localized bulging that might be difficult to capture with diameter measurements. The presence and severity of aortic regurgitation was assessed visually on cine images and by flow mapping at the aortic root. Measurements from the pre-operative baseline studies were compared with those at the latest study in the follow-up by March 2016.
Another set of 30 anonymized studies from all the CMR studies of the patients were randomly selected for analysis of intraobserver variability. These 30 studies were duplicated, and aorta measurements were repeatedly done blindly and in a random order to define the intraobserver variability and the limits of maximum change in aorta sizes attainable to measurement variability.
Summary statistics were recorded as mean ± SD for continuous variables and frequency and percentage for binary variables.
Initially, aortic measurements before the operation and at the latest follow-up were compared by paired t-test. The presence or absence of aortic regurgitation before and after the surgery as a binary variable was compared by the McNemar test. All p values were 2-tailed, and a p value of <0.05 was considered significant.
Secondly, we aimed to test whether any of the changes in aortic diameters and cross-sectional areas during follow-up after successful implantation of the PEARS mesh were significantly different from normal measurement variability. In the absence of a control group, such an approach allowed a reasonable assessment of the effectiveness of PEARS for prevention of aortic root dilatation for 2 reasons: 1) the well-defined natural history of Marfan syndrome involves progressive dilatation of the aorta; and 2) in practice, any clinically significant change in aorta size at follow-up should exceed any changes attributable to measurement variability. Therefore, we calculated the mean of the absolute intraobserver change for each of the aorta measurements and defined the measurement variability range as the absolute value of this change in either direction. Any changes in the aorta size from the pre-operative to follow-up measurement where the 95% confidence interval was completely outside this measurement variability margins were significant, similar to an equivalence and noninferiority analysis. All statistical analyses were performed by one of the authors (S.N.) using Stata 14 (StataCorp, College Station, Texas).
The mean age of the 24 patients at the time of PEARS surgery was 33.1 ± 13.3 years (range: 16 to 58 years), and the mean of the largest aortic root diameter was 44.9 ± 2.8 mm (range 41 to 52 mm). Eight of the patients were female, and 16 patients were male. Mean duration to the latest follow-up study after the operation was 75.6 ± 31.3 months (6.3 ± 2.6 years) with 19 of the 24 patients (∼80%) completing at least 5 years follow-up after the operation.
Comparison of pre-operative and follow-up aorta measurements by both including and excluding the 2 outlier patients is shown in Table 1 and Figure 2. There was no significant increase in the aorta size at the levels of the annulus, aortic root (sinus of Valsalva), ascending aorta, and the arch. However, the size of the descending aorta increased significantly at follow-up as seen in the diameter and cross-sectional area measurements. There was no increase in the percentage of patients with mild aortic regurgitation, and in none of the patients was there an increase in the severity of aortic regurgitation or more than mild aortic regurgitation at follow-up.
Intraobserver measurement variability and the accepted maximum margins attributable to measurement variability used for comparing significance of changes in the aorta measurements are shown in Table 2. For measurements of aortic diameters at different levels, the maximum of the acceptable ranges for measurement variability was 1.2 mm (for the largest sinus of Valsalva diameter) suggesting robust reproducibility of aortic measurements.
Figure 3 shows comparison of changes in aortic diameter and cross-sectional area measurements against the maximum acceptable margin of measurement variability in the 22 patients in whom the PEARS mesh was fully fitted. There was a significant increase in the descending aorta diameter at follow-up, and an increase in aortic arch diameter could not be excluded. There was a tendency to a reduction in ascending aorta diameter. There was no significant increase in the size of the annulus and sinus of Valsalva; in fact, there was a tendency for a smaller annulus and sinus of Valsalva at long-term follow-up after PEARS surgery. The results were in keeping with a stable proximal aorta diameter (annulus, sinus of Valsalva, and probably the ascending aorta), which are supported by the PEARS sleeve, and an increase in the diameter of the distal aorta segments that were not supported by the sleeve (the descending aorta and probably the arch).
Apart from the comparison of the blinded measurements reported here, some details on the clinical follow-up of aorta diameters are worth noting. Throughout the follow-up of the patients, on the unblinded side-by-side direct comparison of the pre- and post-operative MR studies for clinical use, there were again no clinically significant changes detected in the proximal aorta diameters except for the 2 outlier patients. Similarly, in the other 2 patients from the original cohort of 27 patients in whom the follow-up was done by computed tomography, no significant changes in the aortic diameters were detected during follow-up. Two female patients each had an uneventful pregnancy after the PEARS operation without any significant changes in their aorta sizes (decision for pregnancy was at the discretion of the patients after full counseling covering risks of pregnancy in Marfan patients as usual).
In medium-term follow-up, this study demonstrates that PEARS keeps the aortic root size stable and prevents dilatation in Marfan patients (Central Illustration). At the same time, it is seen that the unsupported segments, the aortic arch and the descending aorta, remain prone to dilatation over time, and hence, close follow-up is mandatory for the continued risk of type B dissection, as is the case after Bentall and VSRR operations.
Prevention of aortic dissection is a major target in the care of Marfan patients. The risk increases with increased aortic root diameter, and preventing aortic root dilatation is an essential goal. Drugs have long been used in an attempt for delaying aortic root dilatation. The initial enthusiasm with losartan has largely been blunted by the disappointing negative results of the recent Marfan Sartan and other trials that showed no effect of losartan on the rate of aortic dilatation (12,13). The effectiveness of beta-blockers has not been tested in large clinical trials. Therefore, the current strategy for prevention of aortic root dissection depends mainly on aortic root replacement surgery. Bentall operation is now a very well-established technique with very low operative mortality. But replacing an essentially normal valve and leaving the patients with all the potential prosthetic valve–related complications in the early years of life is a major drawback. Added to this is the requirement for lifelong anticoagulation if a mechanical valve is used. VSRR operations have been developed to address these issues; however, they are technically more challenging. Excellent results in Marfan patients were reported from the centers that have pioneered the technique, with very low operative mortality rates and good long-term success (14). However, it is difficult to estimate generalizability of these results to other centers (15). The issues related to VSRR are perioperative risks, conversion to valve replacement, and finally, long-term durability and risk of aortic regurgitation at follow-up. The recent multicenter registry by Coselli et al. (5) showed a substantial risk of significant aortic regurgitation at a rate of 7% at 1-year follow-up after VSRR in Marfan patients. It is of note that centers involved in this registry had years of experience in VSRR. Similarly, the rate of moderate-to-severe aortic regurgitation following VSRR was reported to be 4.2% in another registry of Marfan patients from experienced centers (16). The meta-analysis by Benedetto et al. (3) comparing Bentall and VSRR surgeries in Marfan patients also showed a considerable reintervention rate of 1.3%/year after VSRR surgery in Marfan patients. It is interesting to note that in this meta-analysis, the rate of prosthetic valve–related complications was higher with Bentall and the reintervention rate was higher with the VSRR as expected, but overall, the composite valve-related event rate in both groups were comparable and not significantly different. This indicates an actual tradeoff between a lower thromboembolic and a higher reoperation rate with VSRR.
The present study proved stabilization of the aortic root size with PEARS in an attempt to prevent aortic dissection. Also, evidence from previous animal studies has proved homogenous and full incorporation of the PEARS sleeve to the aorta resulting in a true mesh/biological aortic wall composite with significantly increased tensile strength (17). Therefore, we believe that by strengthening the aortic wall and by preventing root dilatation, the risk of a root dissection will be minimized with PEARS, and even if dissection occurs, the enhanced tensile strength of the mesh/aortic tissue composite will likely prevent a rupture. Because the aortic root is fully restrained within the personalized mesh that is fully incorporated into the aortic wall, it is expected that the stability of the aortic root size shown in this study will be maintained lifelong because the external support essentially leaves no room for further dilatation. The annulus is also fully restrained, and the risk of aortic regurgitation will be minimal. There has been skepticism about extension of the PEARS deep into the aortoventricular level, but our results prove stability of annulus size after PEARS because the proximal seam of the mesh was closed completely encompassing the annulus.
Our results also show that adjusting the mesh around the coronary orifices is a critical step in PEARS surgery because the scenarios in the 2 patients with aortic root dilatation at follow-up were related to the coronary openings in the mesh sleeve. Accordingly, a large cut probably will leave part of the sinus unsupported and might lead to dilatation from the unsupported segment, and on the other hand, an insufficient opening will lead to impingement at the coronary orifice leading to myocardial ischemia. We have incorporated the lessons learned from these 2 cases in the evolving experience with the PEARS surgery, and the imaging dataset is now vigorously studied in every detail for planning the location and size of the coronary openings as a critical step in all the patients. Coronary buttons are also a vulnerable aspect in root replacement surgeries, and aneurysm or stenosis at button sites are rare, but well-defined, complications of both the Bentall and VSRR operations (18–20).
It must be emphasized that PEARS was not developed with an intention to compete with the already established Bentall and VSRR techniques. It was mainly to complement the available techniques and to expand the armamentarium and choices for prevention of aortic dissection for the Marfan patients to choose from, depending on the relative values they place on avoidance of anticoagulation and avoidance of reoperation. A surgical approach that is durable as the Bentall procedure, limiting annular dilatation as the re-implantation variety of VSRR (David operation) and preserving the physiological functionality of the aortic root by preserving the sinuses as the remodeling variety of VSRR (Yacoub operation) would be an ideal approach in Marfan patients. We believe that PEARS satisfies most of these expectations. It is a durable valve-sparing technique preventing aortic root dilatation without any increased risk of aortic regurgitation. The mesh sleeve has a proximal hem similar to an annuloplasty ring that is sewn deep into the aortoventricular junction, limiting any annular dilatation as proved by our results here. The tensile strength of the aortic wall is significantly increased. The shape of the aortic root and the sinuses is preserved; therefore, the physiology is much better maintained, and the excessive stress on the valve leaflets seen in reimplantation (21) and associated with aortic regurgitation is avoided. The physiological Windkessel function after PEARS is at least partially preserved (22); this is in contrast to the fully rigid tubular grafts without any elasticity used in conventional root replacement surgeries. Also, PEARS is different from the other sleeve techniques that utilize semirigid, low-porosity grafts that are tailored during the surgery to wrap the aorta (23,24). These semirigid grafts incorporate poorly into the aortic wall, leading to regions of buckling and even gaps between the graft and the aortic wall as shown in histological and imaging studies (25,26). Finally, another hypothetical advantage of PEARS is worth mentioning. In 1 of the Marfan patients where an autopsy material was evaluated 4.5 years after PEARS (mortality unrelated to aortic pathology ), it was seen that there was no elastolysis in the segments supported by the mesh, and the aorta was of normal appearance; whereas the unsupported arch displayed fragmentation of the elastic lamellae typical of Marfan syndrome. Thus, PEARS not only stabilizes and strengthens the aortic wall, but potentially may remedy the aortic wall in Marfan syndrome. It is hypothesized that the aortic root dilatation in Marfan syndrome is an end result of the interplay between the underlying genetic defects and the hemodynamic load mediated by mechanotransduction signaling pathways (28). Hence, a plausible explanation for this rather unexpected beneficial effect of PEARS might be blunting of this abnormal mechanotransduction signaling pathway associated with elastolysis in the strengthened aortic wall (26). This exciting observation from just a single patient merits further research.
It is obvious that absence of a control group is an important limitation of this study, which is a common inherent limitation in many surgical studies. Nevertheless, we believe that the stringent and blinded measurement protocol and comparison with what is best achievable (i.e., limits of measurement variability) provided robust data. Dilution of the actual sample of 24 pre- and post-studies also helped to limit measurement bias. It is worth noting that, as expected, the blinded measurements correctly identified the increased aorta size in the 2 patients in whom PEARS failed due to known technical issues. Similarly, blinded measurements of the aortic arch and the descending aorta showed what is biologically plausible and expected. Both of these segments were not covered and supported by the external sleeve, and our analysis showed significant increase in size of the descending aorta. Progressive dilatation of the descending aorta is already well-known after root replacement in Marfan patients (29). Unblinded, side-by-side measurements for clinical follow-up performed by an experienced aortic imaging specialist (R.M.) also demonstrated the stability of the aortic root size after PEARS. It might be questioned whether these patients would necessarily have aortic root dilatation if they had not had PEARS surgery. Although in the absence of a control group, this would be a legitimate argument, progressive dilatation of the aortic root is well-defined in Marfan syndrome. Indeed, in the recent Marfan Sartan trial, the aortic root dilated at a mean rate of 0.51 mm/year in the study group (12); therefore the follow-up duration of our study was substantial enough to expect a clinically detectable aortic root dilatation in our Marfan population. It should also be noted that all the patients referred for PEARS had clinically proven progressive aortic root dilatation before they were considered for the operation. The first patient to have PEARS had an aortic root dilated from 44 mm to 49 mm over 12 years and remained stable at 49 mm 12 years after having the PEARS operation (Figure 4).
PEARS keeps the aortic root size stable and prevents its dilatation in the medium-term follow-up and therefore appears as a viable option for prevention of aortic root dissection in Marfan patients.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: 3-Dimensional printing allows an individualized approach to external aortic root support for prevention of aortic root dilatation in patients with Marfan syndrome.
TRANSLATIONAL OUTLOOK: Longer-term follow-up studies are necessary to establish the effectiveness of personalized external aortic root support to prevent aortic dissection and rupture.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiovascular magnetic resonance
- computed tomography
- personalized external aortic root support
- valve-sparing root replacement
- Received February 14, 2018.
- Revision received June 14, 2018.
- Accepted June 19, 2018.
- 2018 American College of Cardiology Foundation
- Hiratzka L.F.,
- Bakris G.L.,
- Beckman J.A.,
- et al.
- Pyeritz R.E.
- Benedetto U.,
- Melina G.,
- Takkenberg J.J.,
- et al.
- Coselli J.S.,
- Volguina I.V.,
- LeMaire S.A.,
- et al.,
- Aortic Valve Operative Outcomes in Marfan Patients Study Group
- Treasure T.,
- Crowe S.,
- Chan K.M.,
- et al.
- Treasure T.,
- Takkenberg J.J.,
- Golesworthy T.,
- et al.
- David T.E.,
- David C.M.,
- Manlhiot C.,
- Colman J.,
- Crean A.M.,
- Bradley T.
- Miller D.C.
- Song H.K.,
- Preiss L.R.,
- Maslen C.L.,
- et al.
- Tasca G.,
- Selmi M.,
- Votta E.,
- et al.
- Dumanski A.,
- Nowicki R.,
- Kustrzycki W.
- Van Hoof L.,
- Verbrugghe P.,
- Verbeken E.,
- et al.
- Humphrey J.D.,
- Schwartz M.A.,
- Tellides G.,
- Milewicz D.M.
- Engelfriet P.M.,
- Boersma E.,
- Tijssen J.G.,
- Bouma B.J.,
- Mulder B.J.