Publications Reviewed
Peterzan A, Hardy R, Chaturvedi N, Hughes AD. Meta‐Analysis of Dose Response Relationships for Hydrochlorothiazide, Chlorthalidone, and Bendroflumethiazide on Blood Pressure, Serum Potassium and Urate. Hypertension. 2012;59:1104–1109.
Roush GC, Holford TR, Guddati AK. Chlorthalidone Compared with Hydrochlorothiazide in Reducing Cardiovascular Events: systemic Review and Network Meta‐Analysis. Hypertension. 2012;59:1110–1117.
Thiazide‐type diuretics were first approved in the United States for hypertension in 1957. Chlorthiazide was the first thiazide diuretic used in clinical practice, but today, more than a half‐century later, the two most commonly used thiazide‐type diuretics in the treatment of hypertension are hydrochlorthiazide (HCTZ) and chlorthalidone. HCTZ is readily available and included in most fixed‐dose antihypertensive combinations. At 12.5 mg/d to 25 mg/d, HCTZ is the most widely used thiazide‐like diuretic in current practice, despite the fact that chlorthalidone is generally considered to have greater potency and longer duration of action and has been the diuretic most commonly used in major landmark clinical outcome hypertension trials.
Summary of Papers and Comment
In order to determine the relative potency and potential for adverse events, Peterzan and colleagues performed a systemic meta‐analysis to examine the placebo‐adjusted dose‐response effect of various doses of HCTZ, chlorthalidone, and bendroflumethiazide monotherapy on blood pressure (BP) and electrolytes. According to this analysis, the estimated dose of each drug predicted to reduce systolic BP by 10 mm Hg was 1.4, 8.6, and 26.4 mg, respectively, for bendroflumethiazide, chlorthalidone, and HCTZ. This 3:1 ratio in BP‐lowering potency between chlorthalidone and HCTZ is consistent with previous reports. Similarly, the estimated doses of bendroflumethiazide, chlorthalidone, and HCTZ predicted to reduce serum potassium by 0.4 mmol/L were 4.2, 11.9, and 40.5 mg, respectively, a similar ratio to that seen with BP efficacy.
Importantly, while 25 mg of chlorthalidone lowered BP to a greater extent than 25 mg of HCTZ, there was no obvious difference in the maximum BP‐lowering effect of one drug over the other. In fact, increasing the dose of HCTZ to 50 mg (or more), a dose well above what is commonly used in both clinical practice and fixed‐dose combination tablets, led to reductions in BP similar to those seen with the most commonly used doses of chlorthalidone, 12.5 or 25 mg daily. As such, the greater potency of chlorthalidone in and of itself should not lead to its endorsement as the thiazide‐type diuretic of choice. These data do suggest, however, that if the prescribing clinician is willing to increase the dose of HCTZ above the doses commonly used in this country, dose optimization of HCTZ could be a reasonable alternative to changing to a more potent thiazide diuretic such as chlorthalidone in patients whose BP is not adequately controlled.
That said, it is the prevention of cardiovascular (CV) events, not only BP control, that is truly our penultimate goal in treating patients with hypertension. Since there are no trials that directly compare HCTZ with chlorthalidone in the reduction of CV events, Roush and colleagues undertook a meta‐analysis to determine the relative risk reduction associated with each medication after correcting for office systolic BP. While the technical aspects of the trial are quite complex (see below), the primary finding of this analysis concluded that when the reduction in office systolic BP was equal, the risk for CV events was about 20% lower with the use of chlorthalidone than with HCTZ. In this analysis, use of chlorthalidone led to similar reductions in CV events as was seen with other nondiuretic antihypertensive agents, while HCTZ was not as effective in reducing CV events as its nondiuretic comparators.
Given the nature of these analyses, the conclusions should be considered hypothesis‐generating rather than confirmatory. However, taken together, these two reviews support the hypothesis that while the difference in antihypertensive potency between HCTZ and chlorthalidone can potentially be overcome by increasing the dose of HCTZ to one higher than often used in clinical practice, given the same degree of office BP reduction, chlorthalidone remains superior in reducing CV events. In the absence of head‐to‐head clinical trial evidence, we believe that the available data calls for an individualized approach using the most appropriate thiazide diuretic and, importantly, the optimal dose of that diuretic for each hypertensive patient.
BP Dose Response Meta‐Analysis: Methods and Results
The stated aim of this systemic review was to examine the placebo‐adjusted dose response effect of thiazide and thiazide‐like diuretic monotherapy on BP and electrolyte changes. In order to do so, the authors collected data from clinical trials that met the following specific criteria: (1) had to be a double‐blind study of thiazide or thiazide‐type diuretics in people with hypertension; (2) have a parallel design; (3) have random allocation with at least 2 arms of therapy; (4) have a follow‐up of at least 4 weeks; (5) require a baseline washout of previous antihypertensive medications for at least 2 weeks; (6) include a placebo arm; and (7) require at least the measurement of BP and serum chemistry (electrolytes) as endpoints. According to the authors, the only thiazide or thiazide‐type medications that met these criteria were chlorthalidone, bendroflumethiazide, and HCTZ, so they were included in the analysis. Statistical analysis was performed to determine the dose‐response relationships for these 3 medications both for placebo‐corrected BP and biochemistries.
Of the 6477 abstracts screened, 26 trials met the inclusion criteria, including data from 4683 patients with more than 53 treatment arms that were included in the analysis. Doses of HCTZ <6.25 mg were found to have no significant effect on BP, but there were insufficient data to delineate the lower part of the dosing curve for bendroflumethiazide or chlorthalidone. For systolic BP, the estimated daily dose of each drug predicted to reduce systolic BP by 10 mm Hg was 1.4 mg for bendroflumethiazide, 8.6 mg for chlorthalidone, and 26.4 mg for HCTZ. There was no significant difference found in the maximum reduction in systolic BP between agents at the highest doses studied, approximately 10.1 mm Hg for >25 mg HCTZ, 15.5 mm Hg for >25 mg of chlorthalidone, and 14.2 mm Hg for >5 mg of bendroflumethiazide.
For diastolic BP, the estimated daily dose of each drug predicted to reduce diastolic BP by 4 mm Hg was any dose of bendroflumethiazide, 14.0 mg for chlorthalidone, and 20.8 mg for HCTZ. There was no significant difference found in the maximum reduction in diastolic BP between agents at the highest doses studied, approximately 3.7 mm Hg for >25 mg HCTZ, 4.6 mm Hg for >25 mg of chlorthalidone, and 7.3 mm Hg for >5 mg of bendroflumethiazide.
For serum potassium, the estimated doses required to reduce serum potassium by 0.4 mmol/L were 4.2 mg for bendroflumethiazide, 11.9 mg for chlorthalidone, and 40.5 mg for HCTZ. For urate, the dose of each drug predicted to increase urate by 36 mmol/L was 2.1 mg, 8.9 mg, and 12.3 mg for bendroflumethiazide, chlorthalidone, and HCTZ, respectively. Of note, no data were provided regarding the relative effects of these agents on serum sodium, plasma glucose, other electrolytes, or azotemia.
It is clear from these data that the potencies of these medications, both for BP reduction and their effects on serum potassium, are quite different. If the clinician is willing to use higher doses of HCTZ than commonly used in clinical practice and found in most fixed‐dose combination agents, approximately 50 mg daily, similar effects on BP and serum potassium can be achieved as seen with the most commonly used daily doses of chlorthalidone, 12.5 mg to 25 mg.
CV Events Meta‐Analysis: Methods and Results
The stated aim of this systemic review was to examine the relative efficacy of chlorthalidone and HCTZ in reducing CV events. Articles were included in the analysis if one arm used either HCTZ or chlorthalidone as the first‐step antihypertensive medication and if the primary endpoint was CV morbidity and mortality. Trials in which the principal aim was to examine other outcomes, such as left ventricular hypertrophy or effects on BP, were excluded. Two types of analyses were performed on applicable studies. The first was a drug‐adjusted analysis, where trials with a shared referent arm (a similar drug or class of drugs) were compared. The second analyzed office systolic BP, where the difference in office systolic BP between the diuretic arms was used as a continuous variable.
Overall, 9 studies representing 78,350 patients were included in the systemic review; this included 3 trials with 18,374 patients based on HCTZ use and 6 trials with 59,978 patients based on chlorthalidone use. In 8 of the 9 trials, the mean age was 67 to 72 years with roughly 50% of participants being women. Definitions of CV events were similar in all 9 trials.
Four of the 9 trials shared a common comparator and were used in the drug‐adjusted meta‐analysis. Chlorthalidone was superior to HCTZ in reducing CV events when compared with either an angiotensin‐converting enzyme inhibitor (relative risk [RR], 0.83; P=.014) or amolodipine (RR, 0.77; P<.0001). In the pooled analysis, the number of patients needed to treat (NNT) with chlorthalidone rather than HCTZ to prevent one CV event during 5 years was 27.
All 9 of the included trials were used in the office systolic BP–adjusted analysis. In this analysis, the incidence of CV events was strongly associated with the antihypertensive effect of both HCTZ and chlorthalidone. However, for any given difference in mean achieved office systolic BP, the risk of CV events in the chlorthalidone arms were lower than the risk in the HCTZ arms (RR, 0.82; P=.024). Of note, for the same reduction in office systolic BP, the risk for CV events was similar in the chlorthalidone and the nondiuretic arms (RR, 0.98; P=.589), but use of HCTZ was associated with an increased risk of CV events when compared with nondiuretic use (RR, 1.19; P=.021).
Of note, because few data are provided about the timing or technique of office BP measurement, it remains unclear whether the apparent superiority of chlorthalidone over HCTZ in reducing CV events, despite similar office BP reductions, is a result of its longer duration of action or some other mechanism.
Final Thoughts
Like all thiazide and thiazide‐type diuretics, HCTZ and chlorthalidone inhibit the NaCl co‐transporter in the distal tubule. Acute administration of these agents in high doses leads to significant naturesis. In the collecting duct, the additional luminal sodium drives reabsorption of sodium with potassium loss into the urine through separate channels to maintain electrical neutrality; this mechanism is at least partly under the influence of aldosterone. While the mechanism of naturesis is well established, the antihypertensive effect of thiazide and thiazide‐type diuretics occur at lower doses than what is needed for a significant diuretic effect, and the exact mechanism by which these medications lead to long‐term reductions in BP remains unclear. A better understanding of how these agents lower BP, at commonly prescribed doses, both as a class and as individual agents, would probably go a long way toward resolving the lingering debate as to whether one or the other should be considered the preferred diuretic of choice in the treatment of hypertension.
While the systemic reviews presented here certainly support the increased use of chlorthalidone, in the absence of a much needed randomized controlled clinical trial comparing the two agents for their effects on CV events, it remains unclear whether chlorthalidone should be the thiazide‐type diuretic of choice for all patients with hypertension. We believe chlorthalidone remains a reasonable therapeutic option for many patients, especially those who are not adequately controlled with HCTZ and in those with resistant hypertension.
Disclosures
Dr Basile wishes to disclose that he has served on the speakers' bureau and/or served as a consultant to Takeda Pharmaceuticals and Daiichi‐Sankyo, who market fixed‐dose antihypertensive agents that contain either chlorthalidone or HCTZ. Dr Bloch wishes to disclose that he has served on the speakers' bureau and/or served as a consultant to Takeda Pharmaceuticals, Daiichi‐Sankyo, and Novartis Pharmaceuticals, who market fixed‐dose antihypertensive agents that contain either chlorthalidone or HCTZ.