Overall Study Question
The purpose of this study was to determine the effects of a blood-pressure lowering regimen (perindopril plus indapamide vs. perindopril alone) on the risk of stroke and other major vascular events among patients with a history of stroke or transient ischemic attack (TIA). The 6,105 patients included were those with a history of stroke or TIA within the past 5 years. They could have no clinical indication for an ACE inhibitor, and no contraindication to an ACE inhibitor. There were no blood pressure-related criteria for entering the study, however patients with uncontrolled hypertension were recommended to receive antihypertensive therapy before entry into the trial. Patients were also recommended to be clinically stable for 2 weeks prior to entry into the study.
The primary study outcome was fatal or non-fatal stroke. Secondary outcomes included a combined endpoint of fatal or disabling stroke; total major vascular events (composite of non-fatal stroke, non-fatal myocardial infarction, or death due to any vascular cause); total and cause-specific death; and hospital admissions.
Potentially eligible patients entered a 4-week pre-randomization run-in period during which they received open-label perindopril (2 mg PO daily for 2 weeks, then increased to 4 mg PO daily for 2 weeks). If they adhered to and tolerated this regimen, they were randomly assigned to continued active therapy or matching placebo. In the active therapy group, patients received perindopril 4 mg PO daily alone or perindopril 4 mg PO daily plus indapamide 2.5 mg PO daily. In the placebo group, patients received either double placebo or single placebo. This was a “flexible” regimen, meaning that prior to randomization, the responsible physician was asked about his or her intentions with respect to treatment intensity. Those patients whom the physician found no specific indication for or contraindication to treatment with a diuretic were randomized to either double active therapy (perindopril plus indapamide) or double placebo. The remaining patients received either single active therapy (perindopril) or single placebo. This flexibility allowed the authors to stratify by the intention to use single therapy or combination therapy.
Are the Results of the Study Valid?
1. Was assignment of patients randomized?
Partially. Although patients were randomized to either placebo or active therapy, there was no randomization between single or combination therapy.
2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?
Of the 7,121 patients who entered the run-in phase, 1016 (14%) were either ineligible or withdrew. The main reasons for withdrawal were dizziness or hypotension (3.4%), cough (2.7%), other suspected intolerance (2.3%) or patient’s decision (2.0%).
Of the remaining 6,105 patients, 3,544 were randomized to placebo and 3,051 to active therapy. Two patients in the active group and one patient in the placebo group were lost to follow-up.
3. Were patients, their clinicians, and study personnel ‘blind’ to treatment?
Partially. While patients and investigators appear to have been blinded as to whether or not a patient was receiving active therapy or placebo, they do not appear to have been blinded as to whether patients were receiving single or combination therapy. Aside from the use of identical appearing placebo tablets, specific blinding procedures for all study personnel are not elaborated upon.
4. Were the groups similar at the start of the trial?
The active treatment and placebo groups appeared to be similar. However, patients in the combination group tended to be younger, were more likely to be men, had higher blood pressures at entry, were more likely to have a history of hypertension, were more likely to have coronary heart disease, and were recruited sooner after their qualifying cerebrovascular event.
5. Aside from the experimental intervention, were the groups treated equally?
Yes. Aside from the pre-specified stratification into combination therapy or single therapy, the groups appear to have been treated similarly.
What were the Results?
1. How large was the treatment effect?
For the primary outcome, 307 patients in the active group (10%) had a stroke during follow-up as compared to 420 (14%) patients in the placebo group (p < 0.0001). There was no evidence of heterogeneity in the size of the hazard ratios between subgroups defined by type of stroke (ischemic vs. hemorrhagic) or the time interval between initial stroke and enrolment (<6 months or 6 months to 5 years).
For secondary outcomes, there were fewer fatal and disabling strokes in the active group compared to placebo (Relative risk reduction [RRR] 33% with 95% confidence interval [CI] 15 to 46; Event rates [ER] 4% vs. 6%). Total major vascular events (composite of MI, stroke, and vascular death) were also reduced in the active group (RRR 26%, 95% CI 16 to 34, ER 15% vs. 20%). There was no difference in total mortality, however the proportion of patients admitted to hospital was statistically significantly lower in the active group (RRR 9%, 95% CI 1 to 15; ER 41% vs. 44%).
The active combination (perindopril + indapamide) group also experienced a reduction in stroke risk compared to placebo combination (RRR 43%, 95%CI 30 to 54, ER 8.5% vs. 14.4%). The stroke rate in the active single therapy (perindopril) group did not differ from the placebo single therapy group (RRR 5%, 95%CI -19 to 23, ER 12.3% vs. 12.8%). As well, patients who received combination therapy had lower rates of any major cardiovascular event when compared to placebo (RRR 40%, 95% CI 29 to 49, ER 13% vs. 21%), whereas those receiving single therapy did not (RRR 4%, 95% CI -15 to 20, ER17.7% vs. 18.5%). The authors report significant heterogeneity (P < 0.001) in the sizes of these treatment effects.
The active combination (perindopril + indapamide) group also experienced a reduction in stroke risk compared to placebo combination (RRR 43%, 95% CI 30 to 54, ER 8.5% vs. 14.4%). The stroke rate in the active single therapy (perindopril) group did not differ from the placebo single therapy group (RRR 5%, 95% CI -19 to 23, ER 12.3% vs. 12.8%). As well, patients who received combination therapy had lower rates of any major cardiovascular event when compared to placebo (RRR 40%, 95% CI 29 to 49, ER 13% vs. 21%), whereas those receiving single therapy did not (RRR 4%, 95% CI -15 to 20, ER 17.7% vs. 18.5%). The authors report significant heterogeneity (P < 0.001) in the sizes of these treatment effects.
The overall average difference in blood pressure reduction (systolic/diastolic) was 9/4 mm Hg, favouring those in the active treatment group as compared with those in the placebo group. Those assigned to combination therapy had a greater reduction in blood pressure (12.3/5.0 mm Hg) compared to those assigned single therapy (4.9/2.8 mm Hg).
The reductions in stroke risk and major cardiovascular events were of similar size in both hypertensive (SBP ³160 mm Hg or DBP ³90 mm Hg) and non-hypertensive patients. Again, this risk was significantly reduced in patients on combination therapy, and not in the single therapy group.
2. How precise was the estimate of the treatment effect?
The risk of stroke was reduced by an absolute reduction of 4% when comparing active therapy to placebo (i.e. 10% vs. 14%). My calculation of the 95% confidence intervals for this ARR is 2% – 6% (NNT 17 – 50).
Will the Results Help Me in Caring for My Patients?
1. Can the results be applied to my patient care?
The patients enrolled are representative of stroke patients and the results appear to have been appropriately reported.
2. Were all clinically important outcomes considered?
Although the efficacy endpoints appear to be fairly complete, the authors report very little information on safety endpoints (i.e. drug-related side effects). In a trial where 22% of patients who completed the 4-week run-in/weed-out discontinued the study medications prematurely, this information would have been valuable.
3. Are the likely treatment benefits worth the potential harms and costs?
By the end of the study, 1350 (22%) of the patients had prematurely discontinued all study tablets. The overall proportion of discontinuation was similar in both active and placebo groups. The main reasons for discontinuation were participant’s decision, cough, hypotension, and heart failure requiring treatment with an ACE inhibitor or diuretic. There were more discontinuations for cough and hypotension in the active treatment group. There is no mention as to whether patients tolerated single therapy better than combination therapy. Other safety data is not included in the publication.
For those patients who did tolerate the study medications, the therapeutic benefit appears to be associated with using combination therapy, as opposed to single agent therapy only. The cost of perindopril or indapamide is not significantly different than that of other ACE inhibitors or diuretics (e.g. hydrochlorothiazide) commonly used to treat hypertension post-stroke. However, the cost of diuretics is considerably less than that of ACE inhibitors. This may be an important consideration when looking at treating all post-stroke patients, both hypertensive and non-hypertensive.
The PROGRESS study adds another piece of information to the growing literature on ACE inhibitors and diuretics for prevention of cerebrovascular events. What is striking about this particular study is that the results suggest that the indapamide was responsible for the benefit seen in the combination group. One would have expected a reduction in stroke risk in the perindopril group based on the observed drop in blood pressure, however this was not the case. Unfortunately, the study was not designed to determine if this benefit truly was the work of indapamide as there was no indapamide-alone group.
From previous studies, we know that the evidence for diuretics in stroke reduction is fairly consistent. In the Post-stroke Antihypertensive Treatment Study (PATS), indapamide 2.5 mg PO daily lowered blood pressure about 5/2 mm Hg and demonstrated a 29% relative reduction in stroke recurrence. In the Systolic Hypertension in the Elderly Program (SHEP), treatment of isolated systolic hypertension with chlorthalidone 25 mg PO daily resulted in a 36% relative reduction in total stroke frequency.
On the other hand, evidence for the use of ACE inhibitors has not been as consistent. The Captopril Prevention Project trial demonstrated that captopril lead to an increased risk for stroke compared to diuretics, b-blockers, or both. In the HOPE study, ramipril 10 mg PO daily was associated with a 1.5% absolute reduction in stroke overall. However, when examining the subgroups, it appears that the benefit was seen in patients without previous stroke/TIA, as ramipril provided no statistically significant benefit in patients with previous stroke/TIA.
Another interesting point is the heterogeneity between the combination and single therapy groups. Patients who received the benefit were younger, male, had higher baseline blood pressures, and were more likely to have coronary heart disease. One could argue that combining the results of these two very different patient populations should not be done, especially considering the confounding role that indapamide (which has previously shown efficacy in stroke reduction) may have played.
So where does all of this leave clinicians and patients? The authors suggest that treatment with perindopril and indapamide should be considered routinely for patients with a history of stroke or TIA, regardless of blood pressure. Overall, it appears that intensive blood pressure lowering, even in “normotensive” patients may reduce stroke risk. However, without a direct comparison of diuretic versus diuretic plus ACE inhibitor, we really can’t be certain that addition of an ACE inhibitor will substantially increase the benefit over a diuretic alone. Considering the cost of implementing combination therapy in all post-stroke patients, this information would be valuable to have.
In summary, PROGRESS suggests that all stroke patients, regardless of hypertensive status, should be initiated on a diuretic. Addition of an ACE inhibitor may be necessary to achieve blood pressure goals or if certain comorbid illnesses (e.g. diabetes) exist; however, when one considers all of the available ACE inhibitor evidence to date, these agents may not actually confer any additional benefit in stroke reduction as compared to a diuretic alone.
PROGRESS Collaborative Group. Randomised trial of a perindopril-based blood-pressure-lowering regimen among 6105 individuals with previous stroke or transient ischaemic attack. Lancet 2001;358:1033-41