We read with interest the article titled “Outcome of Undersized Drug-Eluting Stents for Percutaneous Coronary Intervention of Saphenous Vein Graft Lesions” by Hong et al. Considering the potential clinical implications of this attractive study, addressing some methodologic issues would be appreciated.

Despite the retrospective nature of the study, it is a quite large series of interventions of saphenous graft lesions with groups well-balanced for baseline characteristics. The main observed difference is, as expected by design, the stent diameter/average reference lumen diameter. However, the angiographic reference diameter was also different: group I, 4.2 ± 0.7 mm; group II, 3.7 ± 0.6 mm; and group III, 3 ± 0.5 mm (p <0.001). Also, the inflation pressure was significantly different among the groups. Perhaps these differences played a role in the final result? Previous publications have highlighted the relation between larger stents/greater pressure and the incidence of distal embolization and no-reflow. The Society of Cardiac Angiography and Interventions has recommended the use of devices not requiring high pressure to be deployed in saphenous vein grafts.

A discrepancy in the rates of stent malapposition between the abstract data and Figure 2 is noted. In the abstract, we read that “no significant differences were found in the incidence of stent malapposition among the 3 groups (group I, 21%; group II, 42%; and group III, 52%; p = 0.001).” However, Figure 2 reflects a 14%, 7%, and 5% incidence for groups I, II, and III, respectively (p = 0.11). Erratum? The latter results seem more plausible because lower stent/artery ratios are associated with a greater incidence of malapposition. If these data are confirmed, the rate of stent malapposition would be significantly greater in group I (14%) than in groups II and III (6%, p = 0.043).

The rate of major adverse cardiac events is similar to that of previously published studies, including a very low incidence of myocardial infarction. However, the authors do not detail the incidence of (late) stent thrombosis: do the 3 cases of myocardial infarction (assumed not to be procedure related) stand for episodes of stent thrombosis? Because one of the major fears of leaving stents not fully apposed is the greater incidence of stent thrombosis, this point should be specifically addressed.

As the authors state in their last sentence, “we compared many parameters in the present study according to the ratio of the drug-eluting stent and saphenous vein graft sizes, not by small drug-eluting stents versus large drug-eluting stents.” Thus, do oversized stents (in smaller grafts) fare worse than the rest? To support this theory, we analyzed the differences in postpercutaneous coronary intervention creatinine kinase elevation greater than 3 times the upper limit of normal. No significant difference was observed in the incidence of this complication between groups I (6%) and II (9%, p = 0.5). However, the difference reached statistical significance when we compared groups I and III (p = 0.014) and showed a strong trend when groups II and III were compared (p = 0.06). It seems a bad idea to oversize the stents in saphenous graft lesions as we usually do in native vessels. We should choose undersized stents. However, by how much? Which would be the best stent size to use compared to the angiographic reference diameter: slightly smaller or really smaller?