Facilitation of primary PCI with ReoPro: reply

The corrected TIMI frame count (CTFC) method has been prospectively validated as providing independent risk stratification above and beyond the conventional TIMI flow grades.1 In myocardial infarction studies including patients with an occluded infarct-related artery, a frame count of 100, a value that is the 99th percentile of that for patent vessels, is imputed to an occluded vessel.1 As recommended by Gibson et al.,1 a CTFC of 100 was scored for TIMI grades 0–1 flow of the infarct-related artery in the ReoPro-BRIDGING study,2 similar to that seen in other studies, e.g. the SPEED, TIMI 4, 10A, 10B, and 14 trials. A sample-size calculation for only those patients having TIMI flow grade 2 or 3 in a STEMI study with primary percutaneous coronary intervention (pPCI), where a majority of the patients has occluded infarct-related artery at the time of the qualifying angiography, would have resulted in a serious bias. The upstream therapy with abciximab prior to pPCI does not replace the thrombolytic therapy and pPCI; of course, the pre-pPCI CTFC was higher in our study than in thrombolysis studies. However, as we pointed out,2 no data concerning the CTFC values before and after abciximab treatment exist. Therefore, we used already published findings and our own data on the normal values and standard deviations, with the CTFC value as accepted difference between an infarct-related artery treated with thrombolysis and normal arteries. Thereby, we did not assume that abciximab therapy would result in 100% flow restoration. Comparisons of randomized studies with a similar design have revealed similar initial TIMI flow 3 grades in the respective control groups, i.e. 2% in the study of Zorman et al., 7.9% in ERAMI trial, and 7% in ReoPro-BRIDGING study. The combined TIMI flow 2+3 (not given separately) was 48% in REOMOBILE and 33% in ReoPro-BRIDGING studies. The TIMI flow scoring is limited by a relatively high interobserver variability and its categorical nature, as we pointed out in our study.2 If we recalculate the difference between the groups with regard to TIMI flow 3, also using the combined TIMI flow 2+3, as in other studies,3,4 the difference between the groups in ReoPro-BRIDGING study remains statistically significant, similar to that seen in the meta-analysis of Montalescot et al.,3 but different from the On-TIME study,4 where the TIMI 3 flow was not different, whereas the combined TIMI 2+3 flow was different between the groups. We are aware that our study was underpowered with regard to the clinical endpoint of the study,2 and that clinical conclusion based on surrogate parameters should be interpreted with caution. We are, however, sure that the primary endpoints of our study were assessed correctly, as cited earlier.