What ever happened to the PACE-MI trial? 

In March 2007, it was announced that a $15 million NIH grant was awarded to support the PACE-MI study (grant #5U01HL080416 from the National Heart, Lung, and Blood Institute).¹ The purpose of the initially randomized, controlled PACE-MI trial was to “evaluate the effectiveness of a pacemaker combined with beta-blocker therapy at improving the death rate and preventing subsequent heart attacks in individuals with bradycardia who have recently experienced a heart attack”.² Recruitment was anticipated to end in August 2011.² Almost eight years later, it is concerning to see what has happened to this multimillion-dollar taxpayer-funded trial. In 2009, the study was somehow converted to a completely different study — a registry with the stated purpose of analyzing beta-blocker dose-response effect on mortality over two years after MI.³ The new study had nothing to do with pacemakers. In 2013, the study PACE-MI was even given a new name: the Optimal Beta-blocker Therapy After Myocardial INfarction (OBTAIN) registry. 

Preliminary findings from the multicenter OBTAIN registry were presented at ACC.14 in March.⁴ The study compared survival curves in 6,682 patients for the following five beta-blocker dose groups based on the dose prescribed at the time of hospital discharge and its percentage of target dose: No beta-blockers, ≤12.5%, 12.6-25%, 26-50%, and >50% of target. The data showed that there was a statistically significant difference in the curves (p<0.001) with the worst survival among the group of patients who were discharged on no beta-blockers (approximately 80% at two years for the no beta-blocker group compared to 90% at two years for the remaining four groups). No surprise. 

The survival curves for the remaining groups in OBTAIN showed a slight progressive decrease in mortality when higher doses were prescribed. However, the differences between these curves was small, and assuming that the reported p-value of <0.001 is the result of a single log-rank test of all survival curves, the only conclusion that can be made based on the survival analysis is that the five survival curves are not the same. One cannot make any further conclusions when trying to split hairs and compare subsets of the five curves. Furthermore, multivariate analysis that included other clinical variables, that interestingly did not include body weight or heart failure status, showed that the individual hazard ratios for each beta-blocker dose group was only statistically significant, and marginally so, for the highest group (> 50%; p=0.05). Despite this, however, the conclusion was a surprising recommendation regarding the use of beta-blockers in patients discharged after an MI — “achieving the target doses used in prior randomized clinical trials is neither necessary nor advisable”. Really? These results should have no impact on current clinical practice.

The OBTAIN study is a prime example of the limitations of registries in medicine and raises concerns about publicly funded trials. Registries are non-randomized and cannot account for all potential covariates, including subjective patient characteristics such as frailty. The PACE-MI trial did not even account for objective potential covariates such as body weight and heart failure status. Another limitation of most registries is the inability to account for adjustments in medication doses over time. The dose of beta-blockers studied in the PACE-MI trial appears to have only been the dose at the time of discharge. Certainly, this dose changed during follow-up. Registries should be used to track quality, identify trends, and generate hypotheses rather than be used to challenge therapies that are already well-established on the basis of randomized clinical trials. PACE-MI also raises issues about federal funding of clinical trials. How was a $15 million, NIH-sponsored trial redesigned two years after it was started from a randomized controlled pacemaker trial to a registry about beta-blockers? Did the same NIH review committee that originally recommended funding for the randomized PACE-MI trial approve the study redesign? Was the funding amount adjusted to account for a completely different trial? These questions should be addressed in a peer-reviewed publication of the PACE-MI… OBTAIN registry.

References

1. $15 Million Grant Will Fund Study of Pacemaker/Beta Blocker Effect on Heart Patients. Northwestern University Feinberg School of Medicine. Available online at https://www.feinberg.northwestern.edu/news/2007/2007J-March/pacemaker.html. Accessed March 12, 2007. 
2. View of NCT00430612 on 2007_02_01. Clinicaltrials.gov. Available online at https://clinicaltrials.gov/archive/NCT00430612/2007_02_01. Accessed October 17, 2014.
3. View of NCT00430612 on 2013_07_012. Clinicaltrials.gov. Available online at https://clinicaltrials.gov/archive/NCT00430612/2013_07_12/changes. Accessed October 17, 2014.
4. ACC Scientific Sessions, March 2014.