Effect of amiloride, or amiloride plus hydrochlorothiazide, versus hydrochlorothiazide on glucose tolerance and blood pressure (PATHWAY-3): a parallel-group, double-blind randomised phase 4 trial

Summary Background Potassium depletion by thiazide diuretics is associated with a rise in blood glucose. We assessed whether addition or substitution of a potassium-sparing diuretic, amiloride, to treatment with a thiazide can prevent glucose intolerance and improve blood pressure control. Methods We did a parallel-group, randomised, double-blind trial in 11 secondary and two primary care sites in the UK. Eligible patients were aged 18–80 years; had clinic systolic blood pressure of 140 mm Hg or higher and home systolic blood pressure of 130 mmHg or higher on permitted background drugs of angiotensin-converting enzyme inhibitors, angiotensin-receptor blockers, β blockers, calcium-channel blockers, or direct renin inhibitors (previously untreated patients were also eligible in specific circumstances); and had at least one component of the metabolic syndrome in addition to hypertension. Patients with known diabetes were excluded. Patients were randomly assigned (1:1:1) to 24 weeks of daily oral treatment with starting doses of 10 mg amiloride, 25 mg hydrochlorothiazide, or 5 mg amiloride plus 12·5 mg hydrochlorothiazide; all doses were doubled after 12 weeks. Random assignment was done via a central computer system. Both participants and investigators were masked to assignment. Our hierarchical primary endpoints, assessed on a modified intention-to-treat basis at 12 and 24 weeks, were the differences from baseline in blood glucose measured 2 h after a 75 g oral glucose tolerance test (OGTT), compared first between the hydrochlorothiazide and amiloride groups, and then between the hydrochlorothiazide and combination groups. A key secondary endpoint was change in home systolic blood pressure at 12 and 24 weeks. This trial is registered with ClinicalTrials.gov , number NCT00797862 , and the MHRA, Eudract number 2009-010068-41, and is now complete. Findings Between Nov 18, 2009, and Dec 15, 2014, 145 patients were randomly assigned to amiloride, 146 to hydrochlorothiazide, and 150 to the combination group. 132 participants in the amiloride group, 134 in the hydrochlorothiazide group, and 133 in the combination group were included in the modified intention-to-treat analysis. 2 h glucose concentrations after OGTT, averaged at 12 and 24 weeks, were significantly lower in the amiloride group than in the hydrochlorothiazide group (mean difference −0·55 mmol/L [95% CI −0·96 to −0·14]; p=0·0093) and in the combination group than in the hydrochlorothiazide group (−0·42