Composite root replacement is an increasingly common procedure today nationally, although it remains one performed most frequently in a handful of high-volume centers by a relatively small number of surgeons. A relationship between volume and outcome has been demonstrated for this operation, like many others, although this need not necessarily be the case if we can share what we have learned effectively with one another and learn collectively. It is commonplace now to see single-center reports of truly remarkable operative results with mortality rates approaching zero, although more representative multicenter database reports suggest somewhat higher risk even in experienced centers.

It was not always so, however, with the first publication of successful root replacement by Hugh Bentall in 1968 as a successor to the Wheat operation being one of the true landmark publications in cardiac surgery. With the advent of preclotting of grafts—initially baked with blood or albumin—the “inclusion wrap” technique of hemostatically closing the aneurysm sac around the graft gave way to Kouchoukos’ freestanding “button” technique in the interest of reducing late anastomotic pseudoaneurysms. The demonstration by Vincent Gott and his team at The Hopkins that root replacement could be accomplished in Marfan patients with no mortalities among 44 elective procedures must surely be considered a milestone in establishing this operation as a viable prophylactic procedure for those at particular risk of aortic catastrophe.

Composite root replacement today is certainly indicated for treatment of aortic root disease in the presence of an unrepairable aortic valve. The exact place for root repair, however, is a little difficult to define. In addition to debate over the advisability or technical feasibility of aortic valve repair, an issue addressed by other authors in this text, the definition of “aortic root disease” is itself an arguable point. Clearly, those patients with well-defined syndromic aortic disease such as Marfan syndrome, Loeys–Dietz syndrome, SMAD mutations, and the like should have complete root replacement when presenting with aortic dissection or when root diameters reach accepted thresholds. It is also probably safe to say that the consensus would be in favor of composite root replacement (or valve-sparing root replacement) in nonsyndromic individuals presenting with
extensive dissection of the sinuses. Debate is greater around conditions such as bicuspid aortic valve-associated aortic disease where some have argued routine root replacement “even in the presence of a seemingly normal aortic root” while others question the very notion that we should exercise different thresholds for aortic replacement depending on valve morphology.

Challenged to define contraindications to root replacement, this author would argue that unless pressed by unusual circumstances such as dissection as noted above, extensive root destruction by endocarditis or given a very small annulus or outflow tract for which one is persuaded that a root replacement will afford the ability to implant a larger valve with superior hemodynamics, root replacement should not be performed in the absence of significant dilatation of the sinuses of Valsalva and displacement of the coronary arteries. It is in this setting that kinking of the reimplanted coronary arteries is the greatest risk with potentially catastrophic results. The surgeon is also well advised to remember that reoperation, if required for structural valve deterioration or endocarditis, is more difficult after root replacement than after isolated valve replacement or even a separate valve and ascending graft. I appreciate that some would argue in favor of routine biologic root replacement with a stentless xenograft, homograft, or Ross operation for the small annulus, but I would caution the occasional root surgeon that the risks increase as the operations get more complex.

Aneurysmal dilatation of the root will most often be identified by echocardiography or axial imaging such as computed tomography (CT) or magnetic resonance imaging (MRI). Echocardiography may be advisable to evaluate valve disease as well as ventricular function, although increasingly these data are available via MRI as well. Conversely, if the initial diagnosis is by echocardiography, axial imaging is generally a wise step to define the extent of the ascending aneurysm and identify any other aneurysmal disease. Knowledge of the distal extent and the possible need for circulatory arrest facilitates discussion with the patient and family as well as preparation by the operating room team.

Imaging of the coronary arteries by angiography or CT imaging is also advisable regardless of the patient’s age. Apart from identification of occlusive disease, the presence of an anomalous coronary course may be critical to a successful outcome. This is particularly relevant in the young patient with bicuspid aortic valve (BAV), a condition associated with anomalous coronary origin, when root replacement is being performed for only moderate enlargement as prophylaxis against dissection. As the occurrence of the latter is demonstrably uncommon, the operative risk associated with the procedure to prevent it must also be extremely low.

Patient positioning for root replacement is no different from that for other cardiac surgical procedures in that it may be performed via sternotomy or bilateral thoracosternotomy. It is important, however, to prepare in the field at least the upper portions of the lower extremities in case saphenous vein is required to manage coronary complications.

Once committed to composite root replacement, a decision needs to be made with regard to the type of prosthesis to be implanted, be it biologic or mechanical. If the choice has been the former, consideration must be made to the use of a human allograft (“homograft”) or stentless xenograft root. As reoperation for either of these options is particularly challenging, it is more common today for the biologic option to be managed with “on the table” construction of a biologic root prosthesis using a stented xenograft valve.

Either a porcine valve or a bovine pericardial prosthesis can be implanted within a Dacron graft 3 or 5 mm larger in diameter than the labeled size of the valve. I generally use the slightly larger graft as it is a bit easier to fit the prosthesis inside and because it gives
a bit more play in the graft for reimplantation of the coronary arteries. Some surgeons prefer to use a “Valsalva graft” for this same reason, although I have not routinely used one because of the added expense. There is no doubt, however, that such a graft makes coronary reimplantation a bit easier, especially in a reoperative setting where mobilization of the buttons is difficult, and in these instances I may opt for the Valsalva. I prefer to run a 4-0 Prolene bringing the sewing ring of the valve to the edge of the graft although this is arguably unnecessary since the subsequent valve sutures affixing the prosthesis to the surgical annulus will go through both sewing ring and graft and should be hemostatic. I just find it quick and convenient to run a stitch. Other surgeons have suggested implanting the valve a few millimeters up inside the graft to facilitate removal of a degenerated prosthesis, if necessary, without taking down the root, but I have not adopted this approach.

If a mechanical valve is preferred, prostheses manufactured with either a straight or Valsalva grafts are available, with the same argument being made in favor of the latter with regard to ease of coronary reimplantation.

In my practice, the first step in preparing the root after excision of the valve leaflets, debridement of the annulus, and sizing, is creation of the coronary buttons. I begin by transecting the aorta completely at the sinotubular junction and then approach the coronaries from the cut edge down. Since the course of the left coronary is away from the surgeon, I begin transecting the aorta on the surgeon’s side (Fig. 7.1) down below the level of the osteum and then medially. Provided coronary anatomy is normal, this should be a “safe” zone. This also provides a nice sense of the thickness of the aorta and provides a target endpoint for a second incision started at the cut edge of the aorta and then down on the side away from the surgeon. If this is a virgin root one can easily place one blade of the scissors outside the aorta and the other inside. If in a redo situation, more often I tend to use a cautery scoring the edge of the button from the inside of the aorta into the fat. Note that in the redo setting one does not need to mobilize extensively, but only divide enough to allow the root to open up and the extraneous wall to fall away from the button. This will allow for full-thickness bites through the wall when reimplanting the osteum.