| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2003;126:191-199
© 2003 The American Association for Thoracic Surgery
Evolving technology |
a Department of Design, Engineering, and Production, Delft University of Technology, Delft, The Netherlands
b Heart Lung Center Utrecht, University Medical Center Utrecht, Utrecht, The Netherlands
revisions requested July 8, 2002; revisions received July 24, 2002 Received for publication April 28, 2002; accepted for publication September 24, 2002.
* Address for reprints: Cornelius Borst, MD, PhD, Professor of Experimental Cardiology, University Medical Center Utrecht (Room G02.523), Heart Lung Center Utrecht, PO Box 85500, 3508 GA Utrecht, The Netherlands
c.borst{at}hli.azu.nl
 |
Abstract |
|---|
 Top Abstract Materials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
METHODS: In the device-constructed anastomosis, blood-exposed non-intimal surface was estimated in all anastomosis configurations identified in truly new (ie, nonduplicate and nonrelated) patent applications and in anastomotic devices recently introduced by several institutions. In the sutured anastomosis, blood-exposed non-intimal surface area was estimated by analysis of previously investigated anastomoses. In 0 blood-exposed non-intimal surface anastomosis configurations, finite element modeling was used to calculate coronary wall stress.
RESULTS: By the end of 2001, 57 truly new applications for the distal coronary anastomosis had been published, categorized in 11 types of anastomosis configurations. The tissue blood-exposed non-intimal surface area (ie, non-intimal tissue surface area) ranged from 0 to 6 mm2. Approximate total blood-exposed non-intimal surface areas (ie, sum of tissue blood-exposed non-intimal surface and foreign body surface area) in recent devices are 80 mm2 (GraftConnector, Jomed, Helsingborg, Sweden); 33 mm2 (Magnetic Vascular Positioner rings, Ventrica, Inc, Fremont, Calif); 4.3 mm2 (distal connector of St Jude Medical, Inc, St Paul, Minn); and 0.3 mm2 (Crinoline frame, University Medical Center Utrecht/Delft University of Technology, The Netherlands). The sutured anastomoses, in contrast, contained approximately 1.3 mm2 blood-exposed non-intimal surface area. The mean peak porcine coronary wall stress in 0 blood-exposed non-intimal surface anastomosis configurations with greater than 90° arteriotomy edge eversion ranges from 0.4 to 0.8 N/mm2 compared with the mean porcine coronary tear stress of 0.8 N/mm2.
CONCLUSIONS: In recently introduced devices for clinical use, the total blood-exposed non-intimal surface area ranges from 4.3 to 80 mm2 compared with 1.3 mm2 in sutured anastomoses. The blood-exposed non-intimal surface area depends on anastomotic orifice size, wall thickness, and bonding components’ location and size. Deforming the coronary wall to most of the 0 blood-exposed non-intimal surface anastomosis configurations leads to dangerously high stress concentrations in the coronary arteriotomy corners.
In off-pump coronary artery bypass grafting, some surgeons find suturing of the anastomosis demanding and time-consuming. In thoracoscopic coronary surgery on the beating heart, anastomosis construction by suturing is exceedingly demanding and time-consuming. To replace manual suturing by a facilitated and accelerated vessel wall-bonding process, various anastomotic devices for the distal anastomosis have been developed1 and are being developed.2-6
In the year 2001 alone, 40 patents for end-to-side and side-to-side anastomotic devices were published in the patent literature (Figure 1). Although only 21 patents describe truly new ideas (ie, ideas that have not been published in duplicate or related patents), the total number of truly new patents that describe anastomotic designs, which can be used for the distal coronary anastomosis, exceeds 50. In the past 2 years, however, only 4 designs have evolved to prototype devices that have been reported in animal studies: the GraftConnector (Jomed, Helsingborg, Sweden),2,3 the Magnetic Vascular Positioner (MVP rings; Ventrica, Inc, Fremont, Calif),4 the St Jude Medical distal connector (St Jude Medical, Inc, St Paul, Minn),5 and the Crinoline frame developed in our institute (University Medical Center Utrecht/Delft University of Technology, The Netherlands).6 The first 3 devices are currently being tested in clinical trials.7-9 All 3 clinically tested devices display blood-exposed non-intimal surface (BENIS) within the anastomosis and thus conflict with the current surgical practice of intima-intima apposition in anastomosis suturing.10 BENIS may be the result of non-intimal wall layers exposed to blood (tissue-BENIS) or foreign body material (foreign-BENIS).
|
The perfect anastomosis would have 0 BENIS area. As a result, we further investigated the technical feasibility of anastomosis configurations without foreign-BENIS and tissue-BENIS. Zero BENIS area is obtained by proper deformation of both graft and recipient artery, necessary to expose only intima. We hypothesized that these deformations would induce wall stress that may exceed the threshold tear stress of the artery, leading to tears and leakage. Earlier, we studied deformation of the graft wall when it is everted around an anvil.11 The second aim of this study was to estimate coronary wall stress in anastomoses with 0 BENIS and to compare this stress with the ultimate tear stress of the coronary artery.
|
Materials and methods |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
Because the surface of non-intimal tissue, that is, media or adventitia (tissue-BENIS), relates to the anastomosis configuration (anastomosis type and wall-apposition), tissue-BENIS was calculated for all anastomosis configurations found in the patent literature. Because the surface of foreign body materials, such as metal or synthetics (foreign-BENIS), relates to the choice of device (in dimensions and location of the bonding components), foreign-BENIS was only estimated in anastomoses constructed with anastomotic devices introduced by several companies and institutions in the past 2 years.
Anastomotic devices
Tissue- and foreign-BENIS areas were estimated using reported and assumed frame dimensions; the formulas are listed in Appendix II. BENIS area in the sutured anastomosis was estimated from analysis of sutured anastomoses constructed in a previous study.12
Coronary wall stress
In anastomosis configurations with 0 BENIS area, coronary wall stress was estimated by finite element modeling (FEM) techniques. The modeled coronary dimensions (inner diameter: 1.8 mm; wall thickness: 0.6 mm) resembled the (porcine) coronary arteries from a study13 that was used to approximate the material model behavior (Appendix III). Distributed loads were applied to the edge of a longitudinal slit arteriotomy (length: 4.0 mm; slightly rounded corners to avoid nonrealistic stress values: 0.05 mm radius) in such a way that the coronary wall deformed to the shape required for 0 BENIS anastomosis configurations. Mean coronary wall peak stress was defined as mean principal peak Cauchy stress over the 6 FEM nodes with the highest stress values. The value for the threshold tear stress of the porcine coronary artery was derived from a previous study.14
To investigate the influence of the arteriotomy shape and the distribution of the loads that need to be applied to deform the coronary wall, 3 additional situations were analyzed for all 0 BENIS anastomosis configurations: initial longitudinal slit arteriotomy and intermittently applied loads along the arteriotomy edge that is connected to the graft (eg, anastomotic device pins); initial oval arteriotomy and distributed loads along the arteriotomy edge (cf, adhesive connection); and initial oval arteriotomy and intermittently applied loads along the arteriotomy rim.
|
Results |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
|
Anastomotic devices
In Figure 3,
sketches of anastomoses constructed with 4 recently developed anastomotic devices are shown, along with 2 conventionally sutured anastomoses. The GraftConnector of Jomed (Figure 3, C) belongs to the intima-edge, end-to-side configuration (although direct contact between the walls is blocked by the sheet). Both the MVP rings of Ventrica, Inc (Figure 3, D) and the distal connector of St Jude Medical (Figure 3, E) belong to the side-to-side, adventitia-adventitia configuration, although in the former there is no direct tissue contact. The Crinoline frame of our own institute (Figure 3, F) cannot be categorized in only 1 single configuration. It uses a combination of appositions (intima-intima at the cheeks and intima-edge at the heel and toe). The estimated areas of tissue- and foreign-BENIS in each facilitated anastomosis and in the conventionally sutured anastomosis are listed in Figure 3 as well.
|
|
|
|
Discussion |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
3.5 mm2) (Figure 3, B), even when bypass graft flow was less than 15 mL/min.15
Tissue-BENIS in recent devices
The tissue-BENIS area in the anastomosis constructed with the St Jude Medical distal connector (3.0 mm2) is larger than in the sutured anastomosis (1.0 mm2) but similar to the adventitia rim-exposure study (3.5 mm2).Owing to its larger anastomotic orifice, the MVP-ring system yields a larger tissue-BENIS area (6 mm2). The Crinoline frame of our institute and the GraftConnector result in an anastomosis with 0 tissue-BENIS, the latter because the polytetrafluoroethylene sheet covers the media-adventitia edge.
Foreign-BENIS in recent devices
A large foreign-BENIS area is present in the GraftConnector (80 mm2) because of the thin polytetrafluoroethylene sheet that covers the stent. The sandwich structure of rings in the MVP shows a fairly large foreign-BENIS area too (27 mm2). After implantation of both devices,3,4 anticoagulants were administered for a period. The foreign-BENIS area in the St Jude Medical distal connector (1.3 mm2) and the Crinoline (our institute) (0.3 mm2) are considerably smaller. Depending on the material properties, it is conceivable that the larger the foreign-BENIS area, the more pronounced the anticoagulation needs to be. As with tissue-BENIS, however, the relation between foreign-BENIS area and patency rate remains to be determined. Unfortunately, no inferences on the effect of foreign-BENIS area can be drawn from in-stent restenosis studies. Stenting is associated with wall tears and other injury induced by the dilation of the stenosis. Wall tears, in combination with shear stress, are more likely to influence patency than the foreign-BENIS area of the stent itself. However, recent technologic developments with drug-eluting coatings on stents bear the promise that both thrombogenicity and intimal hyperplasia may become minor problems in anastomotic devices, irrespective of the foreign-BENIS area.
Wall deformation
Although new developments in stent coatings, as well as a recent animal study,15 are challenging the assumptions underlying the gold standard of the sutured intima-intima apposition anastomosis, the most attractive anastomosis configuration remains the one with 0 tissue-BENIS and 0 foreign-BENIS area. In only 3 anastomosis configurations, BENIS might be avoided entirely: end-to-side intima-intima appositions 1, 2, and 3 (Figure 2, A-C). All other configurations have tissue-BENIS within the anastomosis or require bonding components that are at least partially located intraluminally (observation in patents).
The coronary deformation of the arteriotomy edge that is required in the end-to-side intima-intima apposition 2 (Figure 2, B) (and the side-to-side, edge-edge apposition, Figure 2, J) leads to a stress of at least 0.7 N/mm2 in the case of both distributed (Figure 4, B) and intermittently (Figure 4, C) applied loads. This value nearly equals the threshold tear stress (0.8 N/mm2),14 leaving no safety margin. Reshaping the arteriotomy to an oval shape will reduce the risk of tearing, because wall stress becomes lower (Figure 4, D). In some patients, however, the threshold tear stress may still be lower than the porcine mean value of 0.8 N/mm2, because in the porcine study,14 tear stress ranged from 0.65 N/mm2 to 1.10 N/mm2.
Limitations to this study
First, the tear stress threshold was based on experiments performed with healthy coronary arteries from young pigs.14 When 3 diseased human coronary artery segments were pressurized, they burst before 300 mm Hg was reached, whereas 8 porcine coronary arteries were challenged to 300 mm Hg without bursting (unpublished data). Unfortunately, in the literature, no tear stress values are available from diseased human coronary arteries. Second, the material model we used was based on compliant pig arteries.13 Human coronary arteries show more than 300% increase in stiffness when age increases from 8 to 19 to 60 to 80 years.16 Thus, in a diseased artery it is likely that relatively small wall deformation induces large wall stress. Third, owing to limitations in the FEM software package (Marc/Mentat; MSC Software Corporation, Palo Alto, Calif; see Appendix III), the model produced a stress that, at large deformations, is lower than would realistically occur. Fourth, we used dimensions from pig coronary arteries (inner diameter 1.8 mm, wall thickness 0.6 mm13) rather than dimensions from humans (inner diameter 1.0-2.5 mm, wall thickness 0.2-0.4 mm). The smaller the wall thickness, the lower the wall stress will be when bending the wall.
Taken together, true human stress values are likely to be higher than calculated in this study, whereas human threshold tear stress may be lower than observed in the pig. We infer that substantial deformation of the human arteriotomy rim without causing wall tears is unlikely.
Experience within our group with the One-Shot stapling device of US Surgical Corporation (Norwalk, Conn)17 illustrates the risk of arteriotomy edge eversion. Despite the favorable results of the FEM analysis shown in Figure 4, D (applicable to this device for which an oval arteriotomy is created), in the same porcine model, this 0 BENIS device created a local coronary dissection in 2 of 14 cases.17
In the remaining 0 BENIS configuration (end-to-side intima-intima apposition 1) (Figure 2, A), the coronary wall is only deformed slightly, resulting in acceptable stress (Figure 4, A). Results from a previous study,11 however, in which we investigated graft eversion around an anvil, demonstrated that everting the graft to the shape (Figure 2, A) is likely to cause unacceptably high stress in the graft. Eversion can only be achieved properly around a small-rimmed anvil (GraftConnector) or around no anvil at all (Crinoline frame; developed by our institute).
|
Conclusion |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
1.3 mm2). A major reason a 0 BENIS anastomosis is unlikely to be successful is that deforming the coronary wall to its required shape leads to stress concentrations in the corners of the coronary arteriotomy that are dangerously close to the threshold porcine coronary tear stress (0.8 N/mm2).
|
Appendix I |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
Intima-intima (Figure 2, A-C), Intima-edge (Figure 2, D), Edge-edge appositions (Figure 2, E, J):
Tissue blood-exposed non-intimal surface (BENIS) = 0 (by definition)
Adventitia-intima (Figure 2, F):
(Csmall + 2
x widthrim) x tvein < tissue-BENIS < (Clarge + 2 x widthrim) x tCA
(4.7 + 2 x 0.3) x 0.2 < tissue-BENIS < (9.7 + 2 x 0.3) x 0.3
1.3 mm2 < tissue-BENIS < 3.5 mm2
Intima-adventitia (Figure 2, G), Edge-adventitia apposition (Figure 2, H):
Csmall x tCA < tissue-BENIS < Clarge x tCA
4.7 x 0.3 < tissue-BENIS < 9.7 x 0.3
1.4 mm2 < tissue-BENIS < 2.9 mm2
Adventitia-adventitia apposition (Figure 2, I, K):Csmall x (tCA + tvein) < tissue-BENIS < Clarge x (tCA + tIMA)
4.7 x (0.3 + 0.2) < tissue-BENIS < 9.7 x (0.3 + 0.3)
2.4 mm2 < tissue-BENIS < 5.8 mm2
|
Appendix II |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
Conventional (Figure 3, A):
In histologic transversal sections (15 sections/anastomosis) of 22 conventionally sutured anastomoses from a previous study,12 the percentage of intima-intima apposition (XII) along the anastomotic line was defined as the sum of anastomosis cheek sections with no medial edge exposed to blood divided by the total sum of cheek sections. XII was 60% ± 20% (mean ± SD).
Tissue-BENIS = (1 - XII/100) x tCA x circumferencestudied anastomosis12 = (1 - 60/100) x 0.3 x 8.0twice length arteriotomy12 = 1.0 mm2
Foreign-BENIS = numbersuture loops12 x diametersuture x lengthloop in lumen12 = 13 x 0.05Prolene 8-0 x 0.5assumption = 0.3 mm2
Adventitia-rim anastomosis (Figure 3, B):
Tissue-BENIS = (widthrim + tCA) x circumferencestudied anastomosis15 = (0.2 + 0.3) x 7.0twice length arteriotomy15 = 3.5 mm2
Foreign-BENIS = numbersuture loops15 x thicknesssuture x lengthloop in lumen = 11.3 x 0.05Prolene 8-0 x 0.5assumption = 0.3 mm2
GraftConnector (Figure 3, C):
Tissue-BENIS = 0 mm2 (medial edge covered by sheet)
Foreign-BENIS = diameterCA(derived from published data)3 x x lengthstent (estimated from published figures)3 - Areaanastomotic orifice = 2.7 x x 10 - x x 2.5(tube diameter)2 = 79.9 mm2
MVP rings (Figure 3, D):
(Ring dimensions estimated from sketches19 and assumptions are based on its use in arteries with a 2.5 mm inner diameter.)
Tissue-BENIS = (tCA + tIMA) x inner circumferencering = (0.3 + 0.3) x 9.7circumference 2 x 4 mm ellipse = 5.8 mm2
Foreign-BENIS = 4 x thicknessring x inner circumferencering + 2 x arearing + 2 x thicknessring x outer circumferencering = 4 x 0.25 x 9.7circumference 2 x 4 mm ellipse + 2 x 5.5area 3 x 5 mm ellipse - area 2 x 4 mm ellipse + 2 x 0.25 x 12.8circumference 3 x 5 mm ellipse = 27.1 mm2
St Jude Medical distal connector (Figure 3, E):
Tissue-BENIS = (tCA + tvein) x x diameterorifice5 - Areaframe body = (0.3 + 0.2) x x 2.2 - 2.2diameter_orifice x 2number of struts x 0.1strut_width (assumption) = 3.0 mm2
Foreign-BENIS = numberclips x 2 x lengthclipleg x widthclipleg + Areaframe body = 6assumption x 2 x 0.7assumption x 0.1assumption + 2.2diameter_orifice x 2number of struts x 0.1strut_width (assumption)= 1.3 mm2
Crinoline frame (Figure 3, F):
Tissue-BENIS = 0 mm2
Foreign-BENIS = 4 x thicknesshook x lengthexposed in lumen= 4 x 0.15 x 0.5assumption= 0.3 mm2
|
Appendix III |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
However, this model loses accuracy at large deformations. In reality, the wall stress will be higher than calculated, which would only strengthen our conclusions. A least-squares fitting procedure was applied on the coronary circumferential stress-strain relation derived in a previous study on porcine coronary arteries,13 with the following results: C10 = 0.0024, C01 = 0.0043, and C30 = 0.0063.
All meshes were made of incompressible hexahedron elements. The proximal and distal ends of the coronary artery were restricted in the longitudinal direction. Because the coronary artery is embedded in surrounded tissue, the vertical movement of segments at the bottom of the artery was fixed. The loads on the arteriotomy edge were incrementally increased until the desired deformation (no arteriotomy edge eversion: orifice width = lumen diameter; 90° arteriotomy edge eversion: 90° rotation of the edge in heel/toe) was observed. This occurred under the following conditions:
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractMaterials and methodsResultsDiscussionConclusionAppendix IAppendix IIAppendix IIIReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Nagasao, J. Miyamoto, T. Tamaki, K. Ichihara, H. Jiang, T. Taguchi, R. Yozu, and T. Nakajima Stress distribution on the thorax after the Nuss procedure for pectus excavatum results in different patterns between adult and child patients. J. Thorac. Cardiovasc. Surg., December 1, 2007; 134(6): 1502 - 1507. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Ferrari, P. Tozzi, and L. K. von Segesser The Vascular Join: a new sutureless anastomotic device to perform end-to-end anastomosis. Preliminary results in an animal model Interactive CardioVascular and Thoracic Surgery, February 1, 2007; 6(1): 5 - 8. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Vicol, S. Eifert, M. Oberhoffer, P. Boekstegers, and B. Reichart Mid-term patency after magnetic coupling for distal bypass anastomosis in coronary surgery. Ann. Thorac. Surg., October 1, 2006; 82(4): 1452 - 1456. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Tozzi, E. Borghi, E. Haesler, G. Siniscalchi, A. Motti, D. Hayoz, and L. K. von Segesser Progress in cardiovascular anastomoses: will the vascular join replace Carrel's technique? Eur. J. Cardiothorac. Surg., September 1, 2006; 30(3): 425 - 430. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Gummert, S. Demertzis, K. Matschke, U. Kappert, M. Anssar, F. Siclari, V. Falk, E. L. Alderman, and W. Harringer Six-Month Angiographic Follow-Up of the PAS-Port II Clinical Trial Ann. Thorac. Surg., January 1, 2006; 81(1): 90 - 96. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. E. Matschke, J. F. Gummert, S. Demertzis, U. Kappert, M. B. Anssar, F. Siclari, V. Falk, E. L. Alderman, C. Detter, H. Reichenspurner, et al. The Cardica C-Port System: Clinical and angiographic evaluation of a new device for automated, compliant distal anastomoses in coronary artery bypass grafting surgery--A multicenter prospective clinical trial J. Thorac. Cardiovasc. Surg., December 1, 2005; 130(6): 1645 - 1652. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Boening, F. Schoeneich, A. Lichtenberg, E. Bagaev, J. T. Cremer, and U. Klima First clinical results with a 30{degrees} end-to-side coronary anastomosis coupler Eur. J. Cardiothorac. Surg., May 1, 2005; 27(5): 876 - 881. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Vicol, M. Oberhoffer, G. Nollert, S. Eifert, P. Boekstegers, B. Wintersperger, and B. Reichart First Clinical Experience With the HEARTSTRING, a Device for Proximal Anastomoses in Coronary Surgery Ann. Thorac. Surg., May 1, 2005; 79(5): 1732 - 1737. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Vicol, S. Eifert, M. Oberhoffer, P. Boekstegers, A. Knez, F. Christ, and B. Reichart Early Clinical Results With a Magnetic Connector for Distal Coronary Artery Anastomoses Ann. Thorac. Surg., May 1, 2005; 79(5): 1738 - 1742. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Wiklund, L.F. Bonilla, and E. Berglin A new mechanical connector for distal coronary artery anastomoses in coronary artery bypass grafting: A randomized, controlled study J. Thorac. Cardiovasc. Surg., January 1, 2005; 129(1): 146 - 150. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
