chlorthalidone is better than hctz for hypertension

A recent analysis of 24-h ambulatory blood pressure monitoring (ABMP) compared hydrochlorothiazide (HCTZ) with chlorthalidone, finding poor overall blood pressure control with HCTZ (see htn chlorthal vs hctz jacc2016 in dropbox, or Pareek AK. J Am Coll Cardiol 2016;67:379​). details:

--54 Indian outpatients with stage 1 hypertension (140-159/90-99 mmHg, diagnosed at office visits and confirmed by 24-h ABPM) were randomized to chlorthalidone 6.25mg vs HCTZ 12.5mg vs a sustained released HCTZ (HCTZ-CR) 12.5mg. followed for 12-weeks [they chose the 6.25mg dose of chlorthalidone, since studies have found it to be 1.5-2.0 x more effective than HCTZ and have a much longer duration of action]

--mean age 46, 50% female, BMI 27, BP 148/93, 10% current smokers

--ABPM was done at baseline, 4 weeks, 12 weeks

--results:

    ​--significant decrease in daytime systolic and diastolic blood. at week 12:

        --HCTZ decreased 7.16/4.73

        --chlorthalidone decreased 12.11/8.4

        --HCTZ-CR decreased 7.88/6.32

    --but at week 12 of nighttime blood pressures

        --HCTZ decreased 4.87/3.62

        --chlorthalidone decreased 10.17/6.82

        --HCT-CR decreased 12.66/10.10

    --and NO DIFFERENCE when look at office based blood pressures: all were about 15/8 mmHg lower

    ​--adverse events: few and not major. hypokalemia in 2 on HCTZ-CR, 1 on HCTZ and 1 on chlorthalidone; increased uric acid in 3 on HCTZ-CR, 4 on HCTZ, and 2 on chlorthalidone. no diffs in lipids

so, this brings up several issues:

--HCTZ is the most commonly prescribed antihypertensive in the world, and had >134.1 million prescriptions in 2008, and more than 48 million are written for monotherapy.

--there are essentially no clinical outcome data on HCTZ 12.5 mg, a very commonly used dose now and one supported by the JNCs (eg JNC8 "recommends thiazide or thiazide-like diuretics", and for HCTZ, "using 12.5 to 25 mg/d as initial dose, or chlorthalidone 12.5mg/d"). the old studies finding actual clinical efficacy for thiazides were done using much higher doses (and, of note, in a few studies using the 50-100mg dosages of HCTZ, there were increases in primary cardiac arrests). in the MRFIT study in the early 1980s both chlorthalidone and HCTZ were used, and in the course of the study those on HCTZ were switched to chlorthalidone because of increased cardiovascular mortality in the HCTZ group.

--there are even some data that combinations of other antihypertensives with low dose HCTZ are inferior to combinations with calcium-channel blockers (CCBs): eg the ACCOMPLISH trial found that there was a morbidity/mortality advantage of the combination benazepril/amlodipine over benazepril/HCTZ

--there are med adherence data suggesting that HCTZ is among the worst (eg, see htn med compliance 2010 in dropbox, or Friedman O. Am J Med 2010: 123; 173)

--i sent out another article by Messerli several years ago, which looked at data on 24-hour ABPM by several different antihypertensives, also finding that HCTZ faired poorly (see htn HCTZ Meta-anaylsis messerli am j cardiol 2011​ in dropbox, or Messerli F. J Am Coll Cardiol 2011; 57: 590). this meta-analysis found that HCTZ in doses of 12.5 to 25mg had 1/2 the blood pressure reduction of ACE-I, ARBs, b-blockers, and CCBs (6.5/4.5 mmHg, vs, respectively, 12.9/7.7, 13.3/7.8, 11.2/8.5, 11.0/8.1) though in-clinic BP recordings all have about the same BP reduction.

--one really big concern is that if we are only checking the daytime clinic blood pressure, low dose HCTZ seems to work fine, as it was in the above study, but the night-time efficacy is pretty bad (which is a real concern, especially given the high incidence of cerebrovasular and cardiovascular events in the early am!!)

--unfortunately in the US we have more limited choices than in the study: chlorthalidone is only available as 25mg (and, from my experience, the pills are way too small to cut further) and there is no extended release form for HCTZ

--one caveat here is that the vast majority of the anti-HCTZ articles involve Franz Messerli, who is a renowned hypertension researcher over the years, but there is always a little twinge of concern when the conclusions are largely related to one person (though a study by Freis ED in 1986 --see Magee PF, Freis ED. Hypertension 1986; 8: 135 (supplement 2) -- did conclude: "12.5 mg of HCTZ per day has no significant antihypertensive effect"). and the meta-analysis above did include studies done by other investigators (eg a Mayo Clinic study: Finkielman JD. Am J Hypertens. 2005; 18: 398)

--so, what should one do?

    --i have stopped using HCTZ 12.5-25 as an initial agent for hypertension for several years now, since the Messerli articles came out, and since there are so many good alternatives with 24-hour efficacy. and the Ontario study on medication adherence showed such low medication adherence rates. I am also really concerned that the good clinic-based blood pressure readings on low-dose HCTZ my be creating a delusion of its actual effectiveness, both for us and patients.

    --i initially tried chlorthalidone as first-line therapy, but the 25 mg tablet causes too much hypokalemia (there were a recent study showing that it really is no worse than HCTZ 25 mg, but i have had several patients on chlorthalidone 25mg with pretty striking hypokalemia)

    ​--in terms of pharmacokinetics, i think amlodipine is the best. it does well in the 24-hour ABPM studies, and there are other studies showing less blood pressure variability with it over ACE-I/ARB for example. and there have been several studies on blood pressure variability showing an increase in strokes in those on meds with more variability (includign ACE-I). if a person is on an ACE-I/ARB and needs a second agent, i usually will try a CCB instead of the combo pill with HCTZ

    --i would be happy to try chlorthalidone if it were available in the lower doses, and the above study with only 6.25mg looks pretty good to me. by the way, Messerli et al really single out HCTZ as the bad one, not indapamide. so that may be an option, though i have not used it.