Introduction
The case presented here offers us an opportunity to discuss an increasingly commonly confronted clinical challenge. Two issues are raised: how best to manage the patient’s valvular heart disease, including both the indications for surgical intervention and the prosthetic options, and how best to manage his aortic dilatation. Let us take them one at a time.
The patient has severe aortic regurgitation but a normal ejection fraction. Although we do not have his ventricular dimensions, he has a class I indication for intervention given his symptoms, and therefore proceeding to the operating room is clearly appropriate [1]. What is less clear is the best course of action once there. Should the valve be repaired or replaced? And if it is replaced, do we recommend a mechanical valve or a tissue valve?
There is increasing interest in repairing regurgitation bicuspid valves and a general sense that perhaps we are better at this now than in the past [2]. Absent calcification of the leaflets and significant doming of the conjoined leaflet, many of these valves can be repaired without excessive degrees of stenosis being created. The degree of doming and resulting stenosis will be related to the orientation of the commissures and the degree to which that orientation diverges from the ideal (but uncommon) of 180° as is associated with two leaflets of equal size; the more circumference subscribed by the attachment of the conjoined leaflet, the more doming and stenosis that will result. The most common mechanism of regurgitation is prolapse of the conjoined leaflet, and as the leaflets become increasingly asymmetric, the degree of stenosis created by shortening of the free edge increases and the trade-off between regurgitation and stenosis becomes unacceptable. Preoperative echocardiography is somewhat predictive, but ultimately reparability can be determined only in the operating room.
Despite the enthusiasm building in the surgical community for repair of regurgitation bicuspid aortic valves, we should be honest that the data on durability are still a bit thin. I advise patients that they can expect durability of 80–90% at 5–10 years [3, 4] depending on whether they are an optimist or a pessimist. In my practice, my threshold for repair is also very much linked to the patient’s feelings about reoperation and whether I am “competing” with a tissue or mechanical prosthetic alternative. If patients are more concerned about avoiding anticoagulation and not too bothered by the idea of reoperation as reflected by selection of a tissue valve as a second option, then my repair need only beat the durability of a bioprosthesis.
I also feel strongly that a formal discussion of all prosthetic options must be part of every consultation even if the patient hopes for preservation of the native valve. Again, the mood of late in the surgical community favors a tissue prosthesis [5, 6], but it can be fairly argued that the evidence base for this is weak at best. The guidelines are appropriately ambivalent in reference to a 50-year-old as presented here [1]. As would be expected, there are data supporting lower rates of major bleeding at the cost of higher reoperation rates, but perhaps less anticipated, without a difference in stroke or survival rates from statewide databases among patients of this age [6]. At the same time, there are long-term single institution data [5] as well as national database studies indicating superior survival for the 50-year-old patient with a mechanical prosthesis [7]. Of course, this choice must be nuanced by the individual’s preferences and lifestyle, including their profession, and in the case presented, the patient is a construction worker who may be particularly reluctant to be anticoagulated. Still he should be aware of these data, and if he opts for a tissue prosthesis, I would advise him to anticipate at least one reintervention in the future. He should be reassured that the operative risk of redo aortic valve replacement is low albeit not zero [8]. While it has become trendy to suggest that he will be a candidate for valve-in-valve transcatheter replacement, the data on this as an option for an individual as young as the patient presented here are not sufficient yet. We do not know the long-term durability of the transcatheter valves in individuals of this young age. We should also recognize the cognitive disconnect between our enthusiasm for prosthetic options with low transvalvular gradients in the first place and our willingness to stuff a transcatheter valve (or two) inside a stented prosthesis. Indeed, transvalvular gradients appear prohibitive for stented prostheses smaller than 23 mm in diameter [9].
Finally we should mention the options of a homograft or pulmonary autograft operation. It is clear now that the durability of a homograft is limited, and its use in this patient would likely commit him to another operation within the next 15 years. Furthermore, since homografts are almost uniformly implanted as freestanding roots, this would be a nontrivial reoperation. The pulmonary allograft operation remains an option, although it is most often applied in the pediatric population, in which growth is necessary and the advantages of this operation are greatest. Despite a randomized trial indicating superior survival for pulmonary allograft over homograft replacement [10], it is fair to say that the appetite for this operation has diminished significantly.
Equally controversial would be the management of his ascending aorta. At 4.2 cm it is clearly larger than would be expected in an individual of his age absent valvular heart disease [11]. Whether one wishes to apply the moniker “aneurysm” to this dilatation as done in the case description above is, I think, more a matter of style than science. My own preference is to refer to this as “enlargement” or “dilatation” since the word “aneurysm” is emotionally charged and communicates a subtle but very real suggestion that intervention is necessary, and once a patient has heard the term, it is difficult to “un-ring the bell.”
There are, in this patient’s circumstances, at least three arguments for intervention on the aorta. First, if the patient chooses to opt for bicuspid valve repair, there are some data to suggest that the durability of the repair will be superior if it supported by a valve-sparing root-type procedure if the annulus itself is dilated (as is most often the case when the valve is regurgitant) [12]. A second argument for intervention is the suggestion that aortic dilatation will progress over time [13, 14] and, if one wishes to perform the most definitive possible procedure, aortic replacement is indicated. This decision obviously would be nuanced by the question of the choice of prosthesis type should the patient opt for replacement. A stronger argument can be made for aortic replacement if the patient choses a mechanical prosthesis with the aim of minimizing the risk of reoperation. If, however, the patient choses a tissue prosthesis accepting the risk of reoperation, the redo aortic valve replacement will be simpler in the setting of a native aorta.
The third and most controversial argument for intervention on the aorta is as a prophylactic procedure to prevent aortic dissection. Aortic dissection remains a remarkably lethal condition, with prehospitalization mortality of at least 50%, and its prevention is a worthwhile goal [15]. Still, a prophylactic surgical procedure should be undertaken only when the risk of the operation is less than the risk of the natural history of the disease. Unfortunately, the quality of the data on both sides of this equation is poor. It is argued by many that adding an ascending aortic replacement to an aortic valve replacement or repair imposes little incremental risk in experienced hands [16], and dissection after aortic valve replacement for a bicuspid aortic valve has long been a recognized entity [17]. Accordingly, why not extirpate the threat?
Unfortunately, it seems most likely that, despite protestations to the contrary, there is no such thing as a “free lunch” regardless of the experience or skill of the surgeon. It stands to reason that any additional intervention adds some measurable complexity to a procedure, and accordingly some increase of risk. Even as a surgeon I view retrospective surgical studies with some circumspection given their inherent biases related not only to patient selection but also to publication. Few of us publish bad results. Furthermore, morbidities may be subtle. For example, despite suggestions that open distal anastomosis during aortic replacement can be performed under hypothermic circulatory arrest at no significant risk [18], I have informally polled surgical audiences and have yet to anyone who would volunteer for an episode of circulatory arrest regardless of temperature or time.
So much for the risk of intervention; what about the risk of aortic dissection? There are scant data on this point for a patient with a 4.2-cm ascending aorta, or for that matter, for almost any aortic diameter. It is quite clear that a 6-cm aorta is a dangerous structure, but the Yale University data on which most of our recommendations are founded included only 230 patients [19]. The recognition that dissection can occur at diameters less than 5 or 5.5 cm as reported by the International Registry of Acute Aortic Dissection [20] inspired an aggressive recommendation for aortic replacement between 4.0 and 5.0 cm for bicuspid aortic valve–associated aortic dilatation when the thoracic aortic guidelines were written [21]. More recent data, however, published from a number of centers, have suggested that the moderately dilated aorta likely has a very low risk of dissection [22–24]. Accordingly, the pendulum has swung back a bit as reflected in the 2014 valve disease guidelines [1] and the subsequent reconciliation document setting the threshold at 5.0–5.5 cm [25]. Particularly relevant to the case presented here, recent studies support further refinement of the recommendation on the basis of valve pathologic features, with the risk of aortic events likely higher among patients with regurgitant values than among patients with stenotic valves [26].
Of course, guidelines should be a starting point, not the final word, in our deliberations and discussions with patients [27]. We nuance our recommendations on the basis of our clinical experience. But these too are subject to important biases. As subspecialists, surgeons “live in the numerator” seeing those patients with complications but almost blind to those without. In addition, our outcome information is asymmetric. For aortic disease we have some knowledge of dissections that do and do not occur among patients in whom the aorta is left alone, but have no means of determining how many dissections would or would not have occurred among those aortas we replaced. We can therefore only estimate the benefit incurred by intervention.
With all this as background, we are challenged to make a recommendation to this specific patient. Independent of epistemological arguments concerning the knowable and unknowable, this man makes a living performing heavy manual labor. Despite academic arguments regarding his actual risk of dissection, my clinical experience has been that this echocardiographic finding has unfortunate implications for his insurability and, potentially, his employability as a laborer. If he seeks life insurance, it is unlikely that the actuary will overlook his aortic diameter. Equally his employer may be unwilling to incur potential liability given concerns about his aorta. It may well be that he needs his aorta replaced regardless of the data simply to allow him to continue to work and provide for his family.
So what operation would I suggest? After pressing the patient to decide for himself, I must admit that in the back of my mind I would be in favor of a valve-sparing root with valve repair and graft replacement of the ascending aorta, recognizing the degree to which this recommendation is based on intuition and not evidence. We are in the operating room because of his valve, and aortic intervention will take dissection off the table. My answer would be very different if his valve functioned normally, or if he did not yet have strong indications for valve surgery, in which case I would recommend watchful waiting as his risk of aortic dissection is likely less than the risk of cardiac surgery.