Reply to Totaro and Zattera.

We appreciate the comments of Totaro and Zattera [1]. As they correctly point out, surgical treatment of complicated active aortic valve endocarditis can be challenging, in particular, in the presence of perivalvular abscess cavities. In effect, such a scenario requires the treatment of each of the two components, i.e. valve destruction and perivalvular extension. Different surgical options exist; closure of the cavity with a patch and separate valve replacement is one, and we have used it in many patients over the past 20 years. We cannot confirm the authors’ observation that the insertion of a stented valve after closure of an abscess cavity is a relevant surgical challenge. Primary root and valve replacement with a homograft or a stentless bioprosthesis is another possibility [2, 3]. We have preferred this approach if extensive aortoventricular dehiscence was found intraoperatively and valve replacement was required. Fortunately, most of these procedures are successful if the surgical approach is adequate. Nevertheless, one has to keep in mind that active endocarditis is characterized by marked heterogeneity of patient risk profiles, details of endocarditic destruction and causative organisms. Along these lines, any interpretation of clinical results should be made carefully. The authors are to be congratulated for their success in 5 patients in whom a stentless prosthesis was used after the abscess cavity was closed with a patch. Nevertheless, the number is definitely too low to draw any conclusions; success in our mind does not warrant generalization, which the term Columbus’s egg implies. Misinterpretation can also occur in other directions; based on the treatment of 33 patients who underwent repair of active endocarditis [4], we observed that their survival was better compared with the survival of those who had replacement as the initial treatment. We pointed out that we could not determine with certainty whether this survival advantage was due to the operation, or rather, a consequence of patient characteristics. The authors also quote David in his editorial comment [5], stating that the probability of reoperation was markedly increased after repair. This statement also ignores some details, since repair stability was excellent after repair of tricuspid aortic valves, and repair failures were a point of concern only after reconstruction of bicuspid aortic valves with active endocarditis [4]. In summary, one should be careful in drawing conclusions from limited patient cohorts in such a heterogeneous disease. We will follow the activities of the authors with interest.