doi: 10.1161/CIRCULATIONAHA.105.574954
(Circulation. 2005;112:e358.)
© 2005 American Heart Association, Inc.
Correspondence |
Letter Regarding Article by Mauri et al, "Late Loss in Lumen Diameter and Binary Restenosis for Drug-Eluting Stent Comparison"
Department of Cardiology, AZ Middelheim, Antwerp, Belgium
We read the article by Mauri et al1 and found it interesting. However, we raise some doubts related to the existence of methodological flaws in their analysis.
First, the 2 regression analyses reported by Mauri et al (one correlating the mean late loss in each study and the respective binary angiographic restenosis and the other correlating mean late loss and its standard deviation) included all the trials selected, merging data from bare-metal stents (BMS) and drug-eluting stents (DES). The authors correctly tested for an interaction for BMS versus DES, finding this term not significant. However, they combined data coming from different DES trials into one single group; thus, the relationships they found may not be present or may have a different shape if analyzed according to each type of DES. Indeed, much of the statistical weight of the relationships found by Mauri et al is dependent of BMS trials.
Second, Mauri et al failed to acknowledge the existence a bimodal pattern of late loss distribution in the BMS group. In fact, the more accurate analysis related to the pattern of late loss distribution after BMS implantation showed a bimodal distribution of late loss, ie, a distribution in which 2 values occur more frequently than the rest of the values.2 This finding suggested the potential existence of different "cohorts" of lesions, some less prone to restenosis and some others more prone to have a clinically significant late loss at follow-up.2 This issue is important because this distribution may be present also after DES implantation. All the trials assessing DES performance showed a right-skewed distribution of late loss. However, because of the low rate of restenosis, they were underpowered to detect a bimodal distribution of late loss after DES implantation, and in any case, no formal investigation of this issue has been carried out so far.
Thus, binary end points (like binary angiographic restenosis) should currently receive a relevant focus for the analysis of DES outcome, regardless of the fact that this implies trials with larger sample sizes. Further investigations are needed to appraise the real clinical impact of late loss.
 |
Acknowledgments |
|---|
None.
|
References |
|---|
 Top References Acknowledgments References |
|---|
1. Mauri L, Orav EJ, Kuntz RE. Late loss in lumen diameter and binary restenosis for drug-eluting stent comparison. Circulation. 2005; 111: 3435–3442.
2. Schomig A, Kastrati A, Elezi S, Schuhlen H, Dirschinger J, Dannegger F, Wilhelm M, Ulm K. Bimodal distribution of angiographic measures of restenosis six months after coronary stent placement. Circulation. 1997; 96: 3880–3887.
Brigham and Women’s Hospital, Harvard Medical School, Boston, Mass
We thank Dr Agostoni for his interest in our article,1 but we would like to challenge him on the basic assumptions of the mechanisms of coronary restenosis that he would take as axioms. Our analysis does not make the assumption of a bimodal distribution of the continuous restenosis end points, because we do not concur with such a result. We do not presume that the observed right-skewedness seen with some low mean late loss distributions is anything more than the shape change expected when the mean values approach zero, because the left side is constrained by the zero boundary with stents (with allowance for some measurement error). Moreover, our demonstration of monotonicity of the late loss–restenosis relationship does not rely on an assumption of normality or even unimodality, but only on the shape of the observed relationship from published trials. Whether right skew is intrinsic to the distribution or consequent to a bimodal distribution is irrelevant to our result.
The possibility raised by Dr Agostoni that each DES may have its own late loss–restenosis relationship cannot be refuted by the present study because such exploration would require numerous randomized trials per DES type. Furthermore, we find this premise unlikely. Indeed, the major result of our work was that the relationship between coronary stents (non-DES and DES) and restenosis metrics could be explained simply by sorting stents by their in-stent late loss. The major characteristic determining the propensity for restenosis of the individual stent types was the simple late loss scalar, obviating the need to invoke a more complex set of late loss–restenosis relationships unique to each stent or stent type.
We do agree that the gold standard for DES efficacy is binary, target lesion revascularization derived from randomized trials in which all confounders are nearly balanced by randomization. Late loss offers a more efficient valuation of antirestenosis stent performance that must be matched against patient restenosis risk. Recently published randomized comparisons of stents support our prediction that differences in late loss, generated by differences in antirestenosis capacity or sample restenosis risk, do translate to differences in target lesion revascularization rates when appropriately powered.2,3
Response
|
Acknowledgments |
|---|
|
TopReferencesAcknowledgments References |
|---|
Disclosures
None.
|
References |
|---|
|
TopReferencesAcknowledgments References |
|---|
1. Mauri L, Orav EJ, Kuntz RE. Late loss in lumen diameter and binary restenosis for drug-eluting stent comparison. Circulation. 2005; 111: 3435–3442.
2. Windecker S, Remondino A, Eberli FR, Jüni P, Räber L, Wenaweser P, Togni M, Billinger M, Tüller D, Seiler C, Roffi M, Corti R, Sütsch G, Maier W, Lüscher T, Hess OM, Egger M, Meier B. Sirolimus-eluting and paclitaxel-eluting stents for coronary revascularization. N Engl J Med. 2005; 353: 653–662.
3. Kastrati A, Dibra A, Eberle S, Stat D, Mehilli J, Suarez de Lezo J, Goy JJ, Ulm K, Schomig A. Sirolimus-eluting stents vs paclitaxel-eluting stents in patients with coronary artery disease: meta-analysis of randomized trials. JAMA. 2005; 294: 819–825.
Related Article:
Circulation 2005 112: 3667.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||