Valve-Sparing Operation in Aortic Root Ectasia Hans-Joachim Schafers and Hans G. Borst
Aortic valve regurgitation caused by aortic root ectasia is a common finding.',2 The most common cause for this pathological complex is a diffuse degenerative process of connective tissue involving the media of the aortic wall, such as in Marfan's syndrome. Fragmentation and dissarray of elastic fibers, formally described as cystic media necrosis, leads to hyperelasticity and decreased mechanical stress resistance. In addition to Marfan's syndrome, root ectasia has also been observed in other patients with o r without apparent association to connective tissue disease." The risk of dissection o r rupture of the ascending aorta and left ventricular volume overload caused by aortic regurgitation define the need for surgical intervention in patients with advanced stages of the disease. Insertion of a valved conduit is still regarded the gold standard for treatment of root e c t a ~ i a . ~ ?However, ' despite favorable perioperative results, the typical long-term risks of alloprosthetic valve replacement remain a point of major concern. Thromboembolism and bleeding contribute to 75% of valve-related complications in mechanical aortic valve replacement. An annual incidence between 2% and 4% has been reported."8 Alternative concepts of valve replacement procedures without the need for anticoagulation do not seem to be ideal substitutes. Long-term experience with bioprosthetic xenografts has shown an unacceptably high rate of valve degeneration, particularly in younger patients.'." Homografts were originally thought to be the ideal valve substitute for valve o r root replacement in the younger patient. However, more recent reports have indicated a significant proportion of structural defects requiring reoperation within 10 years. Besides availability, limited long-term results, especially in the younger age group, have proven to be a major limiting factor. The pulmonary autograft (Ross operation), which seems to be the ideal valve substitute in young patients, must be considered inadequate in patients who have structural arterial wall disease by definition. Thus, at this time, there is no ideal biological replacement device for patients with aortic root ectasia. These considerations led David et all3 to design a surgical concept of valve-sparing replacement of the dilated ascending aorta. Preserving the native, morphologically intact aortic valve seems to have significant advantages over biological o r homograft valve replacement. The key principle of the operation consists of complete resection of all segments of the aorta prone to further aneurysmal dilatation; the aortic valve and associ-
38
ated fibrous parts of the aortic root are preserved and resuspended within a vascular graft. Compared with mechanical prostheses, the long-term risks and disadvantages of anticoagulation are avoided. Originally, this operation was proposed for elective correction of root ectasia. We have also used it in root ectasia in conjunction with acute o r chronic type I aortic dissection.
Indications for Surgery In most patients, the decision for surgical intervention is made on the basis of the diameters of the aortic root andlor ascending aorta. A diameter of more than 5 cm has been shown to be associated with an increased risk of perforation o r dissection and has been the standard cut-off point for decision making in replacement of the ascending aorta. Clinical observations indicate that, in patients with connective tissue disease (eg, Marfan syndrome) or familial history of aortic dissection, an ascending aortic diameter of more than 4 cm places the patient at an increased risk of acute dissection. In these patients, a diameter of 4 cm or greater may be considered as an adequate indication for surgery. In some patients, the degree of aortic valve regurgitation with increasing left ventricular overload indicates the need for an operation regardless of the degree of aortic dilatation. In acute type I dissection, the diagnosis per se defines the need for emergency intervention.
Diagnostic Procedures Transthoracic echocardiography is helpful as a screening examination. Not always can the aortic valve be assessed with adequate accuracy. The morphology of the valve and its leaflets are best studied by transoesophagial echocardiography. The degree of aortic regurgitation, diameter of the aortic root and ascending aorta, and morphology of the valve leaflets are carefully assessed. In patients over the age of 40 years, additional cardiac catheterization is necessary to rule out coronary artery disease. If aortic dilatation is limited to the root itself, the patient may not need further investigations. However, aortic dilatation extends into the arch, thoracic magnetic resonance ima@ng (MRI) or computed tomography (CT) assist in outlining the aortic pathology and planning the operative strategy. In acute dissection, transesophageal echocardiography will suffice in most instances to prove involvement of the ascending aorta. In addition, thoracic spiral CT scanning may be used, if the dissection membrane cannot be clearly shown.
Operative Techniques in Cardiac Q Thoracic Surgery, Vol 1, No 1 (July), 1996: pp 38-43
VALVE-SPARING OPERATION
39
40
SCHAFERS A N D BORST
SURGICAL TECHNIQUE
1
Most of the operative steps a r e more o r less identical regardless of the presence of type I dissection. The chest has heen opened via a midline sternotomy a n d the pericardium inc:ised. The aortic arch, including the origins of the supraaortic vessels, is dissected to determine the extent of aortic replacement and to determine the cannulation site. Distal aortic arch o r femoral artery a r e used for arterial cannulation, depending on the distal extent of aneurysmal disease; tht- right atrium is cannulated for venous drainage. A left ventricular vent catheter is inserted via the right superior pulmonary vein. With acute o r chronic dissection, the femoral artery is always used for arterial cannulation because reconstruction and/or partial replacement of the arch will be performed.
2
The aorta is cross-clamped inferior to the brachiocephalic trunk and opened via a longitudinal incision extended towards the noncoronary sinus to allow for exploration of the aortic root. I n view of a n anticipated cross-clamp time of 90 to 120 minutes, blood cardioplegia is used a n d administered directly into the coronary ostia. The infusion of cardioplegia is maintained throughout most of the procedure. Retrograde administration of cardioplegia via the coronary sinus may he u s c d alternatively. However, this will result in constant cdlection of blood in the operative field and is therefore not o u r preferred mode of administration. The aortic valve is now inspected carefully. To achieve a good long-term result, the valve should be tricuspid and the leaflets should not be overstretched. Small rents in the valve leaflets close to the commissures a r e encountered occasionally. These valves have shown histological signs of mucoid degeneration a n d a r e thus felt unsuitable for reconstruction. If the aortic valve does not seem suitable for reconstruction, the operation is continued as a composite replacement of valve a n d ascending aorta using the technique of Bentall and D e B ~ n o . ~
3
T h e aortir root is transsected subtotally a t the level of the sino-tuhular junction to allow for better exposure. Conimissural stitches (Prolene 4/0, Ethicon, Hamburg, Germany) a r e placed for traction. These sutures will subsequently he used for fixation of the commissures within the vascular graft. I n acute aortic dissection, the dissected layers of the aortic root a r e fused using gelatine-resorcineformaldehyde (GRF) adhesive at this time.
4
The aortic root is mobilized. I n the nondoronary perimeter, dissection procedes to the insertion line of the atria, corresponding to that of the anterior mitral leaflet. The dissection is carried into the right a n d left coronary sinus.
5
The sinuses a r e excised, leaving approximately 5 mm of aortic wall superior to the insertion lines of the valve leaflets. The valve now resem1)lrs a n aortic homograft prepared for free-hand insertion. Buttons of aortic wall a r e left around the coronary ostia with a rim of approximately 3 mm.
6
Mobilization of the aortic root is continued with improved visualization. Dissection is carried down to a horizontal line connecting the most inferior point of each sinus towards the right lateral aspect of the aortic root. Thus, in the commissure between the right coronary a n d noncoronary leaflets, the root is mobilized to the level of muscular septum and in the commissure hetween the left and noncoronary sinus to the insertion line of the anterior mitral leaflet. Only in the commissure between the right and left coronary leaflets does the dissection follow the insertion line of the valve. It is important to mobilize the right coronary sinus remnant to the most inferior point of valve insertion. Right ventricular muscle may adhere to the aortic wall and needs to be dissected to avoid asymmetry of the aortic root once the valve has been resuspended within the graft.
7
Horizontal mattress sutures (Ethibond 3-0, Ethicon, Hamhiirg, Germany) a r e placed along the plane of dissection. One suture each is placed in the central portion of each sinus remnant. Towards the right lateral aspect, additional sutures a r e placed in a horizontal plane. In the commissure between the right coronary a n d noncoronary sinus, careful placement of these sutures in the transition zone between the muscular a n d membranous septum avoids damage to the bundle of His. Towards the commissure between the right and left coronary sinus. the sutures follow the insertion of the valve leaflets.
8
The free height of all three aortic leaflets is measured. The size of the Dacron graft (Hemashield, Meadox Medical, Oakland, NJ)is estimated according to these measurements. Thirty to 40% of leaflet height is considered adequate for roaptation. Thus, 60% to 70% of leaflet height will be the internal radius of the neo-aortic root to be reconstructed. The thickness of the aortic wall is taken into consideration also. If leaflet height is 22 mm, the neo-root will have an internal radius of 13 mm (60% of leaflet height). Adding 1mm of radius for the thickness of the aortic wall, a 28 mm graft is chosen. It does not seem necessary to chose graft sizes of more than 30 mm in diameter.
V.4LVE-SPAHING OPERATION
4.1
42
9
Thv appropriate presealed Dacron graft is prepared. It 5 to 6 cm, which is sufficient for resuspension of the aortic valve a n d facilitates the subsequent operative steps. The geometry of the valve is studied carefully. If the three leaflets a r e of approximately equal size, three markings a t 120" angles a r e made on both ends of the graft. Corresponding to the commissure between the right and left coronary sinus, a triangle is excised to adjust the graft wnfiguration to the geometry of the inferior suture line. is cut to a length of
10
The mattress sutures placed previously a r e passed through the inferior rim of the Dacron graft, carefully maintaining the geometry of the aortic root. T h e commissural sutures a r e passed through the lumen of the graft. The aortic valve now comes to lie within the graft. The inferior sutures are tied. anchoring the graft to the aorto-ventricular junction.
11
The aortic valve commissures a r e fixed within the graft using the commissural stitches that were placed earlier. Attention is pait1 towards maintaining the geometry of the valve using the markings made in the graft that was made earlier. It is of equal importance to maintain the relative height of the commissures. This is achieved by placing each commissure ant1 the graft on similar tension while passing the suture transmurally through the graft. If the leaflets a r e of similar size, the commissures should lie in a horizontal plane. The valve is carefully inspected a n d checked for adequate cwaptation. If necessary, the position of the commissures needs to be rorrectetl a t this point.
12
The aortic wall remnants from which the cusps arise art' now sutured to the graft using a continuous transmural running suture line. One suture (Prolene 4/0)is used for each sinus, starting at the lowest point and suturing towards the commissures. Careful placement of these sutures is important for subsequent hemostasis of the aortic root. This step is ngain similar to the implantation of a n aortic homograft in a free-hand fashion.
SCHAFERS AND BORST
13
Once the latter suture lines have heen finished and all sutures have been tied, the valve geometry is again inspected for symmetry a n d leaflet coaptation. The graft is filled with saline to assess valve competence. Occasionally, prolapse of one of the leaflets is encountered. This can he corrected hy reducing the circumference of the free margin of the respective leaflet. A mattress suture (Prolene 510) buttressed with a pericardial pledget is used, as described by Trusler et aI,l4 for aortic valve reconstruction.
14
Both the right a n d left coronary ostia a r e anastomosed to the graft. Care is taken to place the sutures along the rim of the coronary ostia, thus effectively excluding the diseased aortic wall. Using this technique, we have not observed recurrent aneurysm formation at the site of coronary implantation.
15
If aortic ectasia extends to the distal ascending aorta o r aortic arch, o r in the presence of dissection, a second graft is anastomosed to the appropriate level of the distal aorta. For arch surgery, hypothermic circulatory arrest, with o r without retrograde cerebral perfusion, is used. I n acute type I dissection, the false arch lumen is obliterated using G R F glue. I n chronic dissection, this may be achieved hy including a n externally placed strip of Teflon felt (C.R. Bard, Bard Vascular, Haverhill, MA) into the distal anastomosis. Partial replacement of the arch is performed. The second graft can be clamped a n d extracorporeal circulation resumed in stan(lard fashion. The two grafts a r e shortened to the appropriate length a n d a r e anastomosed. The heart is deaired and coronary circulation is resumed.
I6
Remnants of the aortic wall a r e finally wrapped around the graft to shield it from possible mediastinal infection a n d to minimize the risk of entering the aorta in case of reoperation.
43
Y.4LVE-SPAHI N C ()PERATION
COMMENTS Operative Mortality In t.xperiencwl hands, mortality rates of 5% to 7% have h e n reportetl f o r replacement of the asrending aorta with n r without the root as the primary operation. T h e long-term ~tntic~oagulation-relate(l mortality in merhanic-al valve replacwnent is low in conjunc.tion with w m p s ite grafts. It has h e n o u r helief that any alternative pro(wlure should he associated with a n operative mortality not apprt.c-iahly higher than that of the s t a n d a r d operation. Results published until now indicate that the valve-sparing operation is indeed not associated with a higher operative risk. I n o u r present experienre with 27 o p r a t i o n s , we have not encountered perioperative mortality.
Results O u r experienw encwmpasses 27 operations performed over a Iwriotl of 3 years. All patients presented with significant ertasia of the aortic root a n d ascending aorta (45 to 90 nim) a n d aortic valve regurgitation (mean 2, 9 0, 8). I n eight patients, the presence of dissection o r thxtension of aneurysnial disease required additional arch surgery. Typical stigmata of Marfan's syndrome were ohserved in eight patients. T h e valve-sparing operation was possil)le in all patients in whom it was attrmpted. Mitral valve reconstruction was performed for severe niitral regurgitation in one patient. Aorticrross-clamp times have ranged from 85 to 150 minutes (mean 110 minutes). All patients survived the early posto1)erative lwriod. Ret.xploration for hleeding was required in two instances. All patients have remained in stable sinus rhythm. Late reoperations were required in three patients (6, 11. a n d 14 months after repair). One patient h a d h a d one episoclv of entlocarclitis M o r e his initial operation. This had I ) w n treated successfully I)y antihiotir therapy without aplm-ent s t r w t u r a l valve damage. This patient tlevelopecl recurrent endoc*artlitis6 months postoperatively with destruction of one aortic valve leaflet a n d tlrhiscence of the aorto-ventricular junction. Reoperation c.onsistecl of dehridement of a h rthstoration of a aorto-ventricular continuity by composite rt.l)lac.ement of the aortic. root. This patient died from pvrsistent sepsis a n d multiorgan failure. T h e two other patients were reoperated on for progressive aortic. valve regurgitation. I n one case, a n aortir valve was iiiil)lantect inside the ascentling aortic graft; the other patient underwent romposite graft replacement of the valve a n d root. Both recwvered uneventfully from reoperation.
+
Experience with this operation is still limited. Little is known regarding the long-term effects of aortic- valve resuspension within a t u b u l a r Dacron graft. I t has h e n argued that this key operative principle disregards the elasticity a n d dynamic: characteristics of the normal aortic root. I n addition, the normal geometry of the sinus of Valsalva is not maintained. On the other hand, it is unknown whether these considerations a r e merely theoretical o r will result in impaired long-term valve stahility. Although this operation will not be applicahle to every patient, it seem to he a valid option in a numher of patients with root ectasia a n d a morphologically in tact valve.
REFERENCES ~ isiilatvil 1. Giiinry TE. Davirs MJ. Parker DJ:Thr rtiiiliigy arid c o t t r s ~of scvt'rr aiirtic. rrgurgitaticin: A vlinival, pathtilogii.al an11 rc~liiwarcliograiiliir s t iiily . Br I f r a r t J 58:350-368. 1087 7 Olr~inI,, Sulwanianian K, Edwards W: S u r g i w l Imthology o f piirr tiortic -. insttft'icirncy: .4 stiiily 1,1225 i'asrs. May11C l i n k I'riw 50:835-841, I984 iilar pathology in Marfan synilromr-an overview, 3. in Ilrtzrr R , Grhlr I', Ennkrr .I ( 1 4 s ) :(:ariliiivascdar A s l i w t s id Marfan Synclriimr. Darnistailt. Crrtnany, Strinkcqif, 1005. 111) 25-32 iwiiplrti* rrplwrtiirtit iif thr 4 . Hrntall H, IIrBono A: A ascrnclitig aorta. Thorax 2 HO, I,akatin K. Marcin 6.1. 1.1 al: Svrial lorig-twin ;issi*ssmrnto f 5 . H~IIIOW thr natural histor? o f asyinptciniativ paiirnis with rhriitiir aortii. rrgitrgitatinn arid niirnial lrft v r n t r i i d a r systiilii. fiitwtiiiti. ( : i r i d a t i o n
84:1625-1635. 1991 Eilttiiiiiils 11: Thriiiiihitir and lilrriling i.iimlilii.atii,tis of priisthrtir lit-art v a l v c h . Ann Thorai. Sttrg 44:430-445, 1987 Horstkiittv D : Long-term f i i l l ~ i w - ~ iaftrtp hibart valvr rt~plaiwttrnt.Z / . Karcliol 75:147-152. 1086 l wnwr1i8 . Ilrirstkiittr I), Si.hitltr tf, Ijircks W, vt al: L l n r x p t ~ r t r i finclings irig tlirotiil~i~riiil~c~lii~ cc,iq)lir.atirins and atitiri,agiilatii)ti after romplrtr 10 year folliiw it11 of pativnts with S t . , I i t i l i * Mi*tlia.al prosthcsrs. J Htwrt Viilvr Dis 2:29-:<01. 1003 0 . Bloomfirlcl P, Whratlry DJ. Prrscott R J , t a t al: Twelve-year comparison o f a 13jork-Shilry iiirrhanical hrart valvr with porcine Iiiopriisthrsrs. N E ~ i g ,Il Mrcl 324:573-579, 1991 1 0 . Biirtliin TA, Miller DC, O y r r PE, et al: D u r a l d i t y of piirvinr valves at 6.
11.
12. 13.
14.
fil'trrti years in a rrprvwntativr North Antrrican Iiatirnt piliiilation. .I T h i x a e Cariliovasr Surg 104:238-251, 1902 Hamnirrmristrr KE, Srtlii GK, Htwilrrson WC, 1.1 al: Comparison iil ~ ~ i t t c o i i i c sin nirii 1 1 yrars after hrart valvr rrplaiwnrtit with a mrchanical valve or hioprofitht-sis. N Engl .I Me11 328: 1289.1296, 1903 Franckr U, Furthmayr H: Grnrs a r i d grnr Iirotliwts inviilvril in Marfan syndromr. Srtniti Thairac Cardiovase Siirg 5:3-10. 1003 Daviil TE. Friiiclrl CM: An aortiv valve-sliarinc qwratiiin f ~ ) patients r with aortic incotiiprtrnrr atid a~iriirystii of thr ascrtiilitig a o r t a . J T h ~ r a c(:ardiiwasc. Sirrg 103:017-021. 1002 Triisier GA, MIW* CAF, l i i i l t l I3SL: R r p t i r of v r n t r i i d i r srlital i l r f e i ~ t with aortir in~itffii.ii*tii.y. .I.Thiirav (:ariliovasv Sitrg 06: 304-398. 1973