When treating aortic stenosis, the surgeon’s mission is to improve survival and alleviate symptoms by reducing or eliminating left ventricular outflow tract obstruction. Although logic suggests that it might be important to use the largest, most hemodynamically efficient prosthesis at aortic valve replacement, studies of the impact of prosthesis size and function on survival have generated controversy (1-5). In the patient with critical aortic stenosis, all available prostheses provide hemodynamic improvement over the diseased native aortic valve. But, as Dr. Kon points out, both geometric and functional assessments reveal clear differences between prostheses—and these differences may be particularly important in the patient with a “small” aortic root. In determining optimal surgical management in this situation, it is necessary to answer 3 questions: 1) What is a “small” aortic root?, 2) What is the impact of prosthesis-patient size on outcome after aortic valve replacement?, and 3) What surgical techniques should be employed to avoid the clinical syndrome of prosthesis-patient mismatch at aortic valve replacement?
Defining the “small” aortic root and prosthesis-patient size
When considering aortic valve replacement in a “small” aortic root, it is most appropriate to employ the concept of prosthesis-patient size, which incorporates measures of both valve size and patient size. Prosthesis-patient size may be expressed using geometric or functional measures of valve size. For geometric measures of prosthesis-patient size, prosthesis size is normalized to patient size by relating geometric dimensions of the prosthesis to body surface area (BSA) (1,2). This methodology is accurate (unbiased) and precise (reproducible), but does not describe specifically the function (transvalvular gradient) of the valve in a given patient at a particular time. Others have chosen to express prosthesis-patient size using the indexed in vivo effective orifice area (EOA), a functional assessment of prosthesis valve (4-6). However, both accurate and precise functional estimation of prosthesis size is elusive: it is related to patient factors, left ventricular outflow complex factors, time, measurement variability, inter-observer variability, method of computation, and myriad other factors. Nevertheless, the concept of indexed in vivo EOA does acknowledge the important interaction between prosthesis size and patient size. Using either geometric or functional assessments of prosthesis-patient size, values in a particular patient may be compared to those from a larger group or population, enabling identification of the patient with a small prosthesis-patient size.
Rather than focusing on prosthesis-patient size, much of the literature examines the issue of prosthesis-patient mismatch (4-8). These are closely related, but different concepts. No doubt a clinical syndrome of prosthesis-patient mismatch, as described by Rahimtoola, occasionally occurs, and such a patient “may be hemodynamically worse after valve replacement.” (9,10) Although Rahimtoola described a clinical diagnosis, over the years the definition of prosthesis-patient mismatch has ceased being a clinical diagnosis and has become arbitrarily inferred from postoperative echocardiographic velocity measurements. The challenge today is to relate the quantitative concept of prosthesis-patient size to the clinical syndrome of prosthesis-patient mismatch.
The impact of prosthesis-patient mismatch upon outcome
Dr. Kon has noticed that prosthesis-patient mismatch influences survival. Published reports confirm that at the small extremes of prosthesis-patient size, adverse clinical sequelae, or prosthesis-patient mismatch, occurs; this is most notable in its effect upon early mortality after aortic valve replacement (1, 4). In a study of 13,258 aortic valve replacements, Blackstone and coworkers found that 30-day mortality increased 1% to 2% when indexed orifice area decreased to less than 1.2 cm2/m2 after aortic valve replacement (1). They provided an algorithm to determine which patients were at risk of increased early mortality by virtue of small prosthesis-patient size. However, with nearly 70,000 years of patient follow-up, the latter study found no impact of prosthesis-patient size on intermediate or long-term survival, confirming other reported data (7,8).
Using in vivo EOA values from a variety of sources, Pibarot and coworkers determined that functional assessments of prosthesis-patient size also impact early survival after aortic valve replacement (4). Employing a somewhat arbitrary definition of prosthesis-patient mismatch, they found that both moderate and severe mismatch, defined as postoperative EOA less than or equal to 0.85 cm2/m2, reduced early survival after aortic valve replacement; although they did not examine EOA as a continuous variable, their data suggest that increasing severity of prosthesis-patient mismatch is associated with increased early mortality (4). In their study, the survival impact of severe prosthesis-patient mismatch was particularly pronounced in patients with reduced left ventricular function. Like Blackstone, they provide methodology to enable preoperative identification of patients at greatest risk of postoperative prosthesis-patient mismatch.
The preponderance of evidence suggests that very small prosthesis-patient size does influence early survival after aortic valve replacement (1,4,7). However, this is an important consideration in a relatively small proportion of patients having aortic valve replacement, and the magnitude of this effect is controversial (1, 8). Furthermore, variations in prosthesis-patient size have little impact on late survival (1,7,8).
These findings do not preclude an effect of prosthesis-patient size on functional status, particularly during exercise. In fact, recent data from Canada suggest that small prosthesis size is associated with late development of congestive heart failure (11). The impact of prosthesis-patient size on functional status after aortic valve replacement certainly warrants continued study.
Surgical techniques to avoid prosthesis-patient mismatch
Using geometric or functional reference values from the literature (1,4,6), it is possible to determine the prosthesis size required to avoid extremely small prosthesis-patient size in the individual patient. When the aortic root is small, the patient is usually small as well, and, in such instances, implantation of a relatively small prosthesis is unlikely to produce the clinical syndrome of prosthesis-patient mismatch. However, in the unusual circumstance of a large patient with a very small aortic root, it is reasonable to consider surgical options to avoid very small prosthesis-patient size and the potential for severe prosthesis-patient mismatch. In this rare situation, we favor aortic root enlargement followed by placement of a stented bioprosthesis or mechanical valve. Although Dr. Kon considers aortic root enlargement by the Manougian or Nicks techniques obsolete, these operations should not be discarded. They are relatively simple and quick, and generally enable implantation of a valve that is one size larger; in a patient at risk for prosthesis-patient mismatch, this may be enough to avoid sequelae attributable to this syndrome. Castro and coworkers achieved excellent results with a strategy of aortic root enlargement in patients identified to be at risk for prosthesis-patient mismatch (12). However, other experienced surgeons have noted increased operative mortality in patients having aortic root enlargement (13).
In contrast, when very small prosthesis-patient size is likely, Dr. Kon performs aortic root replacement with a stentless valve. This certainly provides a hemodynamically excellent result, although a randomized prospective trial failed to demonstrate superior hemodynamic indicies in comparision to stented valves up to 12 months after implantation. In the hands of very experienced surgeons like Dr. Kon, aortic root replacement is performed elegantly and safely, enabling extension of this operation to patients with poor left ventricular function. In contrast, it would be difficult to recommend such a complex strategy to a surgeon who performs only 1 or 2 aortic root replacements per year. Increased aortic cross clamp time and potential difficulties with bleeding and coronary button orientation might jeopardize results.
Summary
Prosthesis-patient mismatch is a real clinical syndrome; however, it is uncommon with modern prostheses and surgical techniques. Preoperative measurement of patient size and intraoperative sizing of the aortic root enable identification of patients at risk for prosthesis-patient mismatch before valve implantation. In the unusual circumstance that such a patient is identified, additional surgical techniques may be employed to optimize outcome.
Disclaimer
In responding to this presentation, we must acknowledge that we reside within the institution with the highest volume of valve operations on the North American continent. The practice at The Cleveland Clinic has not been to avoid using small sized valves in small aortic roots. Yet, hospital mortality for 881 isolated aortic valve replacements in the last 5 years was 1.2%, and for combined replacement and coronary artery bypass grafting 2% in 996 patients. In some instances, left ventricular outflow tract myectomy is performed for obstruction at that level; however, rarely is aortic root enlargement performed. We remain interested in performance of prostheses, but believe that available evidence suggests that other factors have a greater impact on long-term survival than prosthesis-patient size, however expressed