PROFESS: Combination therapy falls short of noninferiority vs Clopidogrel
|May 16, 2008||Susan Jeffrey|
In a factorial design, the trial also examined the effect of early blood-pressure lowering after a stroke using telmisartan (Micardis, Boehringer Ingelheim) vs placebo in the same patients and found there was no benefit with the addition of the angiotensin-receptor blocker in prevention of stroke recurrence, at least over the 2.5 years of follow-up in this trial.
Finally, a third analysis showed no neuroprotective effect of either dipyridamole or telmisartan that were on board when recurrent strokes did occur in this trial, despite suggestive results from previous preclinical studies.
The results, from the Prevention Regimen for Effectively Avoiding Second Strokes (PROFESS) trial, were presented here at the European Stroke Conference 2008.
The PROFESS trial used a factorial design to address
several questions in a large population of 20 332 patients from
695 sites in 35 countries. The patients included had all had a
noncardioembolic ischemic stroke within the previous 120 days and were
randomized to receive aspirin (25 mg) plus extended-release
dipyridamole (200 mg) twice daily or to receive clopidogrel (75 mg)
once daily. At the same time, patients were randomized to receive
either 80 mg per day of telmisartan or placebo.
In the first presentation here, Dr Ralph Sacco (University of Miami School of Medicine, FL) reported results for the
comparison of aspirin and extended-release (ER) dipyridamole vs
Current guidelines in Europe and the US recommend that for antiplatelet therapy after a stroke, aspirin, aspirin plus dipyridamole, and clopidogrel are options for prevention of stroke recurrence, but there are no direct comparisons of the latter agents available to guide choices, Sacco said. The ESPRIT and ESPS2 trials had previously compared aspirin vs aspirin and dipyridamole and had shown the combination to be more effective than aspirin alone.
The analysis used for this comparison was noninferiority first, then superiority, he explained, and the margin chosen for noninferiority was 1.075, or a 7.5% noninferiority difference.
In the end, the primary end point of recurrent stroke was not statistically significantly different between the groups, but they were not able to conclude that the combination was noninferior to clopidogrel. "The upper boundary of the effect estimate crossed our noninferiority margin (of 1.075), and therefore the trial could not conclude, statistically, noninferiority," Sacco said.
PROFESS: Primary outcome for comparison of aspirin plus extended-release dipyridamole vs clopidogrel
PROFESS: Type of first recurrent stroke by treatment group
The secondary outcome of combined stroke, MI, and
vascular death was "nearly identical" between groups; other end points,
including deaths and MIs, were not statistically different, with the
exception of new or worsening congestive heart failure, which was
significantly less frequent in the combination arm.
PROFESS: Secondary and tertiary outcomes for comparison of aspirin plus extended-release dipyridamole vs clopidogrel
Despite this increase in bleeds, when they combined
stroke recurrence and major hemorrhage into one end point reflecting a
benefit/risk relationship, there was no statistical difference between
PROFESS: Benefit/risk analysis
As expected, headache was more frequent with aspirin and extended-release dipyridamole but did not lead as often to permanent discontinuation, as was seen in other studies of this agent, Sacco noted. Headache led to discontinuation in 5.9% of the combination group vs 0.9% with clopidogrel. Dizziness, fainting, and migraine during the first six months of the study were also more frequent with the combination.
"In conclusion, for the antiplatelet part of this large
trial, we were not able to meet our prespecified noninferiority
criteria for the combination vs clopidogrel," Sacco said. The agents
had similar rates of recurrent stroke and the composite of stroke, MI,
or vascular death, he noted. Major hemorrhagic events, including
intracranial bleeds, were more frequent in the combination arm;
"however, the absolute risks were low and partially offset by fewer
ischemic events, primary outcomes," he concluded. "The net benefits and
risks were similar with the two agents."
Telmisartan vs placebo
In a separate presentation, Dr Salim Yusuf (McMaster University, Hamilton, ON) presented results of the telmisartan-vs-placebo comparison.
Previous studies such as the HOPE and PROGRESS trials had shown a benefit associated with using an ACE inhibitor initiated late after stroke, with or without a large reduction in blood pressure. No previous trials have looked at treatment early after a stroke, again using a blocker of the renin angiotensin system, telmisartan.
In this trial, though, the primary results for
recurrent stroke showed no statistically significant benefit with
treatment over placebo.
PROFESS: Recurrent stroke with telmisartan vs placebo
Similar findings were seen for the secondary end point of the composite of stroke, MI, and vascular death, where there was no significant difference overall, but a similar pattern emerged when the end point was analyzed before and after six months of treatment. There were also trends to lower rates of intracerebral hemorrhage and diabetes mellitus in patients receiving telmisartan, Yusuf said.
He speculated that with only 2.5 years of follow-up, the trial may have been too short to realize any potential benefit of treatment, since the previous HOPE trial, for example, had followed patients for four years. In addition, suboptimal adherence in the treatment group as well as competing use of other BP-lowering agents had reduced the blood-pressure differential between the groups, "which hurt our power," he said.
"What we need are longer trials, large trials, with greater blood-pressure lowering," Yusuf concluded.
Finally, Dr Hans-Christof Diener (University Duisburg-Essen, Essen, Germany) reported the results of an analysis looking for any potential neuroprotective effects of any of these agents among those who had a recurrent stroke during the study.
Previous preclinical data had suggested neuroprotective properties for dipyridamole, aspirin (although at very high doses), and angiotensin-receptor blockers in models of cerebral ischemia. The problem with previous neuroprotective trials—of which all of some 90 have failed to date—has been, among other things, the challenge of how to get the agent on board at the time of the stroke, Diener said.
To look for signals of neuroprotection with these agents in the PROFESS trial, they used the modified Rankin scale and the Barthel Index 3 months after the stroke occurred to compare functional outcomes after any recurrent stroke while patients were receiving study treatments.
"The result, unfortunately, was that there was no difference," Diener told a press conference here. There was no significant difference in functional outcomes on either scale for the comparison of aspirin plus extended-release dipyridamole vs clopidogrel (which has not been thought to have any neuroprotective effect) or comparing telmisartan vs placebo.
They also looked at cognitive impairment between groups using the Mini Mental State Examination (MMSE) and again found no difference between the groups in terms of how many patients were cognitively impaired after a recurrent stroke.
"There are two possible explanations" for these findings, Diener said. "The most likely explanation is there is no effect. The other possible explanation is that the observation at 2.5 years is not long enough. Perhaps you have to have treatment and observation times of 10 years or more to see a difference," he speculated.
Sacco, Diener, and Yusuf were cochairs of the PROFESS steering committee.
During the question period after the three presentations, a member of the audience asked the investigators whether, given the high dropout rate due to side effects and the higher bleeding risk with the combination of aspirin and extended-release dipyridamole, "Can we conclude that only clopidogrel would be the first choice for secondary prevention of stroke compared with aspirin and dipyridamole?"
Declining to answer, they directed him to a planned symposium where PROFESS was to be discussed by those not involved in the trial. The questioner, though, pressed for some conclusion.
"Like anything, guidelines are written by numerous groups, and I think until we can really digest all this data, it would be premature for us to make a conclusion," Sacco replied finally. "We are still working on the papers, we are still going through analyses, and then the guideline committees that are set up in Europe and the Americas will have to review the data and make decisions."
"I can answer that," Yusuf added wryly. "There is a range of reasonable conclusions and yours is one in that range, with wide confidence limits."
Sacco said that his own interpretation is that the two treatments are similar. "My current take is that we still have choices, these choices are up to us to decide, and if anything now we have direct comparisons showing equal efficacy between these agents."
The decision of which agent to use in individual cases,
he said, will depend on factors such as cost, tolerance of side effects
(eg, the increased risk for headache with the combination); concomitant
cardiovascular risk, since clopidogrel has been shown to be effective
in those who also have unstable angina or postangioplasty; and the
individual physician's familiarity with the agents.