乌鸦传媒

Chlorthalidone didn't best hydrochlorothiazide (HCTZ) for reducing cardiovascular risk in the treatment of hypertension, a pragmatic VA trial showed.

For the primary outcome of major adverse cardiovascular events and non-cancer deaths, the more potent diuretic showed no hint of reduction versus HCTZ (10.4% vs 10.0%, HR 1.04, 95% CI 0.94-1.16), reported Areef Ishani, MD, of the Minneapolis VA Health Care System, at the American Heart Association (AHA) annual meeting.

As expected, chlorthalidone was associated with more hypokalemia (6.0% vs 4.4%, HR 1.38, 95% CI 1.19-1.60).

Unexpectedly, though, subgroup analysis turned up a significant disadvantage to chlorthalidone in patients with a prior myocardial infarction (MI) or stroke (P=0.002 for interaction unadjusted and P=0.035 after multiplicity adjustment).

"It's difficult to understand how to place this in the context of an overall negative trial," Ishani said.

Prior studies had suggested better cardiovascular outcomes with chlorthalidone versus HCTZ, potentially due to better 24-hour blood pressure control and pleiotropic effects; whereas recent observational trials had pointed to no cardiovascular benefit but greater risk of hypokalemia, acute kidney injury, and chronic kidney disease.

AHA session discussant Daniel Levy, MD, of Mercy Medical Center in Baltimore, noted that the trial addressed a question with a great deal of clinical uncertainty for these two commonly used diuretics.

What to do with the VA trial results "is a question we've been struggling with," Ishani told 乌鸦传媒 at an AHA press conference. "Either of the drugs is appropriate to use for the treatment of hypertension."

For the subgroup analysis, he added: "I think this is a discussion with your patient on how they want to hedge their bets. Because the drugs are so similar, if they want to take one drug or the other because of the subgroup [analysis findings], that's a conversation to have. But we should focus on future trials looking specifically at that subgroup."

AHA press conference discussant Biykem Bozkurt, MD, PhD, of Baylor College of Medicine in Houston, cautioned that it's possible the subgroup findings were due to chance alone, or possibly due to differential effects in sicker patients.

The trial included 13,523 patients, ages 65 and older, who were identified in the electronic health records (EHR) of consenting VA primary care physicians as being already on 25 or 50 mg of HCTZ (94% on 25 mg). These patients were randomly assigned to either stay on that treatment or to be switched open-label to a corresponding dose of chlorthalidone (12.5 or 25 mg). Care was entirely in the hands of the treating physician thereafter and outcomes were followed for an average of 2.4 years through the VA EHR, Medicare claims, and the National Death Index.

"Prior studies suggested the benefits were all with the higher dose," Ishani noted. "But it turns out the world has voted. We had 4,000 clinicians, and they're all using the low dose; and it was a definitively negative study."

No differences emerged for individual components of the primary endpoint between chlorthalidone and HCTZ:

• MI: 2.1% in both groups (HR 1.01, 95% CI 0.80-1.28)
• Stroke: 1.2% in both groups (HR 1.00, 95% CI 0.74-1.36)
• Hospitalized heart failure: 3.6% vs 3.4% (HR 1.04, 95% CI 0.87-1.25)
• Unstable angina with urgent coronary revascularization: 0.3% vs 0.2% (HR 1.54, 95% CI 0.77-3.10)
• Non-cancer deaths: 5.3% vs 5.2% (HR 1.01, 95% CI 0.88-1.17)

Ishani noted several factors in favor of generalizability -- that 68% of primary care physicians agreed to participate; the well-treated population with a baseline 139 mm Hg systolic blood pressure; and 45% of patients being in rural areas -- but also noted that 97% of participants were men and 15% were African American.

Levy raised the question of generalizability to women and to middle-age and younger patients, and pointed out that the trial also selected out patients with favorable responses to HCTZ by including only those already on the drug.

Still, he called these "compelling results, even though a few questions remain."

Notably, "I think this is the first time in my career I heard two successive presentations using the word pragmatic ... a trial design that never could never have been performed by industry," said Levy, referring to the TRANSFORM-HF trial that preceded Ishani's presentation.

Comment