| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 2002;123:1222-1223
© 2002 The American Association for Thoracic Surgery
Letters to the Editor |
Division of Cardiac Surgery, University of Toronto, Toronto, Ontario M5G 2C4, Canada
To the Editor:
We have previously demonstrated that individuals with bicuspid aortic valve malformations have significant medial abnormalities that may result in a higher incidence of pulmonary autograft dilatation after the Ross procedure.
1,2 Subsequent to this observation, Luciani and colleagues 3 attempted an important clinicopathologic correlation to explore this hypothesis in more detail. Although their rationale is excellent and their attempt laudable, the study has significant methodologic problems. These include an inadequate power/sample size, a biased and nonrepresentative sample, a nonblinded evaluation of the data, a lack of adequate controls, and an inadequate follow-up period. As such, we urge caution in interpreting this data and provide the readers of the Journal with a more balanced approach to the manuscript by Luciani and coworkers.
First, for a study to have adequate power to uncover significant differences between groups, the sample size must be sufficiently large. It is no surprise that the authors were unable to show a difference in autograft root dilatation in the bicuspid aortic valve group given their sample size of 9 patients. In fact, if the expected incidence of dilatation is 18%, as the authors suggest, the incidence of dilatation at 33% (3/9) in the bicuspid group is almost double that expected! Although this 2-fold higher incidence is statistically insignificant because of the small sample size, it is indeed a clinically relevant difference that deserves attention. As well, by our calculations, the incidence of preoperative aortic root dilatation would have been significantly greater in the bicuspid group (44%, 4/9) as compared with the tricuspid group (12%, 1/8) if at least 18 patients per group were studied. The smaller the actual difference between groups, the larger the sample size necessary to statistically demonstrate a difference. For example, to uncover an 80% difference (a large difference), at least 50 patients would be required per group. To uncover a smaller difference of 20% (a more clinically relevant difference given that the authors identify an incidence of postoperative dilatation of 18%), more than 750 patients would be required per group.
The authors' conclusion that there is no difference in outcome between the patients with bicuspid and tricuspid valves is at the least biased, if not faulty. The fact that this study is the "only available clinicopathologic study in patients undergoing pulmonary autograft root replacement," as suggested in defense of the numerous limitations of their study, is perhaps striking evidence that others may have reserved judgment until sufficient numbers have been assessed.
Second, the small sample of patients is likely nonrepresentative of the population from which they are drawn and therefore does not describe a majority of patients who undergo the Ross procedure. The authors indicate that 50 patients were available for their study but only 17 (9 bicuspid, 8 tricuspid) were selected in total. The rationale for the selection process of this minority of patients from their potential patient pool remains elusive. Small, nonrandom samples are subject to significant bias that precludes useful conclusions. Further, the nonblinded analysis of tissue samples aggravates the inherent bias of this study.
Third, it is surprising that the authors fail to demonstrate the pulmonary artery lesions that we previously associated with bicuspid aortic valves. 2,3 In our sample of 20 patients with bicuspid valves, not only were the degenerative changes in the pulmonary artery and aorta of these patients more prevalent, but they were considerably more severe as assessed by independent blinded examiners with expertise in cardiac pathology. Our pathologic findings are in keeping with comparable histologic studies. 4,5 The argument that our previous data can be attributed to the effect of the secondary valvular lesions alone is weak and unsubstantiated. Recent studies suggest that medial abnormalities associated with congenital valvular malformations are present irrespective of the functional state of the valve favoring an inherent and ongoing process of adverse vessel wall remodeling. 4
Without demonstrating the expected pathologic condition, no clinicopathological correlation could be performed by Luciani and colleagues. The authors failed to describe their histologic findings in sufficient detail, and no images are offered for the reader to acknowledge. The authors did not grade the lesions as to their severity (as have most similar studies), making comparison of their data to previous studies impossible. By only noting the prevalence of lesions and not their severity, they have also seriously limited the sensitivity of their method to detect differences and, as such, they failed in their attempt. Adding a positive control (such as a sample of Marfan aorta with significant degenerative changes) and blinding the observers would provide evidence that the examiners, if lesions are present, are indeed able to demonstrate the expected changes (note that no changes were seen even in the patient with a bicuspid pulmonary valve). This is questionable as the results of this study are at odds with a growing body of evidence.
The authors argue that the younger age of their patients may account for the apparent discrepancy in the lack of pathologic findings as compared with our previous work. Although this is certainly possible, it does not rule out the likely possibility that the bicuspid aorta and pulmonary artery have an accelerated process of vessel degeneration in the face of hemodynamic stress, such as a pulmonary autograft in the aortic position. The short period of follow-up prevents the authors from a vigorous analysis of this important issue.
We are currently investigating possible mechanisms of adverse arterial wall remodeling in the setting of congenitally malformed aortic valves. Some of these potential mechanisms may include inherent fibrillin-1 defects leading to elastin fragmentation, increased matrix metalloproteinase activity, and extracellular matrix degradation and disorganization resulting in increased arterial dilatation, aneurysm formation, and dissection. 6 As well, disruptions in the nitric oxide synthase pathway have been associated with an increased incidence of bicuspid aortic valves and subsequent arterial dilatation and aneurysm formation. 7,8 Despite the results of Luciani and coworkers, we maintain that the bicuspid aortic valve is associated with an underlying pathologic process in the aorta and pulmonary artery, which has critical implications with respect to procedures such as the pulmonary autograft (Ross) procedure.
12/8/124394
doi:10.1067/mtc.2002.124394
References
This article has been cited by other articles:
|
|
 |
|
  G. R. Veldtman, H. M. Connolly, T. A. Orszulak, J. A. Dearani, and H. V. Schaff Fate of Bicuspid Aortic Valves in Patients Undergoing Aortic Root Repair or Replacement for Aortic Root Enlargement Mayo Clin. Proc., March 1, 2006; 81(3): 322 - 326. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. G. Gleason, T. E. David, J. S. Coselli, J. W. Hammon Jr, and J. E. Bavaria St. Jude Medical Toronto biologic aortic root prosthesis: Early FDA phase II IDE study results Ann. Thorac. Surg., September 1, 2004; 78(3): 786 - 793. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
