Is it Time to Review the Role of Thiazide-like Diuretics in the Treatment of Hypertension?

Diuretics such as bendroflumethiazide and hydrochlorothiazide have
been used for decades in the management of hypertension. They are
seen as older, less fashionable drugs that can have a nega&ve impact on
some metabolic markers in that they can raise blood glucose, cholesterol
and uric acid levels. Newer an&hypertensives, such as the calciumchannel blocker amlodipine, are more metabolically neutral.

In uncomplicated hypertension, the UK’s Na&onal Ins&tute for Health
and Care Excellence (NICE) recommends that the first an&hypertensive
that should be used either an angiotensin-conver&ng enzyme (ACE)
inhibitor or a calcium-channel blocker (CCB) depending on the age or
race of the pa&ent. If an ACE inhibitor is prescribed and is not tolerated
(for example, because of cough), a low-cost angiotensin-II receptor
blocker (ARB) should be considered. (1)

A Cochrane review of 24 studies, involving 58,040 pa&ents with high
blood pressure, looked at four different an&hypertensive drug classes
(thiazide diure&cs, calcium channel blockers, beta-blockers and ACE
inhibitors). The review concluded, “that most of the evidence
demonstrated that first-line low-dose diure&cs reduced mortality, strokes
and heart aWacks. No other drug class improved health outcomes beWer
than low-dose thiazides.” (2)

Likeable Differences
Low dose diure&cs are thought to reduce peripheral vascular resistance
and inhibit the re-absorp&on of sodium from the distal convoluted
tubule. There are subtle differences in the type of diure&cs available.
Bendroflumethiazide and hydrochlorothiazide are thiazide-type diure&cs
while indapamide and chlortalidone are thiazide-like diure&cs. Thiazidelike diure&cs tend to have a longer elimina&on half-life and possibly
beWer 24-hour blood pressure control. They are also potent inhibitors of
platelet carbonic anhydrase ac&vity reducing catecholamine induced
platelet aggrega&on significantly. (3) This is a different mechanism to
that seen in an&platelet drugs such as aspirin and clopidogrel.

Are We Making Any Progress?
The an&platelet ac&vity in thiazide-like diure&cs sounds interes&ng in
theory but how does this relate to clinical prac&ce. PROGRESS was a
placebo-controlled double-blind study involving 6,105 pa&ents with a
history of ischaemic or haemorrhagic stroke or transient ischaemic aWack
within the previous 5 years (4). One group of pa&ents was randomised to
receive the ACE inhibitor perindopril 4mg once daily or single placebo
while another group received a combina&on of perindopril 4mg plus
2.5mg of the thiazide-like diure&c indapamide or double placebo. The
primary outcome was total stroke (fatal and non- fatal).
A`er a mean follow-up 3.9 years, the group taking perindopril only
produced no sta&s&cally significant reduc&on in strokes compared to
placebo. The combina&on group of perindopril and indapamide saw a
sta&s&cally significant reduc&on in stroke incidence from 14.4% in the
placebo to 8.4% the combina&on group. This gives an absolute risk
reduc&on of 6% and an impressive numbers needed to treat (NNT) of 17.
So for every seventeen pa&ents gedng the addi&onal indapamide, one
stroke was prevented over 3.9 years.
What was also found was that the benefits occurred whether the pa&ent
had hypertension or not when they were enrolled into the study. It is not
known whether the benefit seen with combina&on therapy is due to
indapamide alone or to an addi&ve or synergis&c effect of indapamide
with perindopril. What is clear is that the benefit is not aWributable to
perindopril alone. Could the reduc&on in strokes be due to indapamide’s
addi&onal an&platelet ac&vity? This is unique property that is not seen
with the other different types of anihypertensives.

What About Type 2 Diabe8cs?

As discussed, thiazide-like diure&cs can increase blood glucose levels and this could be a poten&al issue in type 2 diabe&cs. The ADVANCE study that looked at over 10,000 pa&ents with type 2 diabetes and one other risk factor for a cardiovascular (CV) event.

Pa&ents where randomised into four groups

1) Intensive blood glucose lowering (including the addi&on of gliclazide MR) to achieve a target HbA1c of less than or equal to 6.5% and the addi&on of the BP lowering combina&on of perinodpril 2mg and indapamide 0.625mg

2) Standard glucose lowering therapy HbA1c 7.3% plus perindopril and indapamide

3) Intensive blood glucose lowering therapy and placebo

4) Standard glucose lowering therapy and placebo

A`er a mean follow up of 5 years, there were no differences in macrovascular outcomes between the pa&ents with the intensive blood glucose control and standard control (HR 0.94 [0.84–1.06]; P = 0.32). There was some reduc&on in nephropathy (HR 0.79; 95% [0.66-0.93}; p=0006) with intensive control but no other benefits. There was an increased risk of hypoglycaemia with intensive treatment (HR1.86 [1.42– 2.40]; P < 0.001). (5)

Those that received indapamide and perindopril saw a reduc&on in death from cardiovascular disease (HR 0.82 [0.68-0.98]; p=0.03) and death from any cause (HR 0.86 [0.75-0.98]; p=0.03) (6)

Along with other major studies, it confirms the importance of blood pressure control in type 2 diabe&cs but also the long-term safety and benefits of thiazide-like diure&cs in pa&ents with type 2 diabetes.

Major Studies ALLHAT was a landmark randomised controlled study involving 42,448 pa&ents with hypertension and one other risk factor. The pa&ents were randomised to receive chlortalidone (thiazide-like diure&c), amlodipine, lisinopril or doxazosin. Other drugs were added un&l a target BP of <140/90 mmHg was reached. The primary end point was CHD death or non-fatal myocardial infarc&on. Secondary outcome measures included all-cause mortality, stroke, and combined cardiovascular disease (CHD death, nonfatal MI, stroke, angina, coronary revasculariza&on, conges&ve heart failure (CHF) and peripheral arterial disease). (7)

There were no differences in primary end points for amlodipine vs chlortalidone. Secondary outcomes were similar except for a 38% higher rate of CHF with amlodipine (P<0.001) numbers needed to harm (NNH) of 50. This equated to one extra case of CHF with every fi`y pa&ents taking amlodipine compared to those taking a thiazide-like diure&c over the six years of the study.

For lisinopril vs chlortalidone there were no differences in primary end
points but a 15% higher risk for stroke (P=0.02) and a 10% higher risk of
combined CVD (P= 0.001) for those on lisinopril. This equates to a
NNH=42 for combined CVD over 6 years.

Doxazosin Issues

The doxazosin arm of the ALLHAT study had to be discon&nued early
(a`er 3.3 years).

There was no difference between the doxazosin and chlortalidone
groups in the primary outcomes of fatal CHD or non-fatal MI or the
secondary outcome of all-cause mortality. There was a higher risk of
stroke in the doxazosin group vs chlortalidone (4.2% vs 3.6%; P=0.04;
number needed to harm [NNH]=167).

The main reason why the doxazosin arm was discon&nued early was the significant increase in the number of cases of CHF compared to chlortalidone (8.1% vs 4.4%; P <0.001; NNH=27). This means that over 3.3 years for every 27 pa&ents taking doxazosin we should expect one extra case of heart failure compared to a low dose thiazide-like diure&c. This is surprising considering doxazosin’s neutral effects on metabolic markers compared to thiazide-like diure&cs. This highlights why we need to base clinical decisions on evidence from robust studies and not surrogate markers. (8)

Clearly, ALLHAT challenged the assump&on that the most important
parameter in trea&ng hypertension is lowering blood pressure, rather
than the drug by which blood pressure is lowered.

An Accomplished Study?

The ACCOMPLISH study was a randomized, double blind trial involving
11,506 pa&ents with hypertension who were at high risk of
cardiovascular events. They received treatment with benazepril plus
amlodipine or benazepril plus hydrochlorothiazide. The primary
endpoint was the composite of death from cardiovascular causes,
nonfatal myocardial infarc&on, nonfatal stroke, hospitalisa&on for
angina, resuscita&on a`er sudden cardiac arrest, and coronary
revasculariza&on. (9)

There were 552 primary-outcome events in the benazepril-amlodipine group (9.6%) and 679 in the benazepril-hydrochlorothiazide group (11.8%). This gives absolute risk reduc&on with benazepril-amlodipine therapy of 2.2%, NNT=45 (HR, 0.80, [CI] 0.72 to 0.90; P<0.001).

It was concluded that the benazepril-amlodipine combina&on was
superior to the benazepril-hydrochlorothiazide combina&on in reducing
cardiovascular events in pa&ents with hypertension. The ACCOMPLISH
trial was the only trial addressing combina&on therapies on outcomes,
which is important give that in clinical prac&ce many hypertensive’s end
up on combina&on therapies.

Several points need to be highlighted when considering this study when
developing any guidelines. The ACE inhibitor used in the ACCOMPLISH
study is not one that is used in clinical prac&ce in countries such as the
UK. It has a poor evidence base when compared to an ACE inhibitor such
as ramipril. The diure&c used in this study was hydrochlorothiazide,
which is a thiazide-type diure&c and not a thiazide-like diure&c, which
were used in PROGRESS and ALLHAT studies. As we have discussed, there
are subtle differences.

Thiazides and Heart Failure

Hypertension is a major risk factor in the development of heart failure.
A meta-analysis of 26 trials involving 223,313 pa&ents with hypertension
looked at which treatment was the most effec&ve in reducing the
development of heart failure. (10)

The main findings were that all an&hypertensive therapies are beWer
than placebo in the preven&on of heart failure, with the excep&on of
alpha-blockers (eg. doxazosin). Diure&c-based treatments were the most
effec&ve first-line an&hypertensive treatment for preven&ng heart
failure. Calcium channel blockers and beta-blockers where significantly
inferior to diure&cs. The reduc&on in HF incidence associated with
different an&hypertensive strategies was independent of the degree of
blood pressure reduc&on.

None of the hypertensive drugs were found to be clearly superior to
others in terms of cardiovascular mortality. Unfortunately, the
ACCOMPLISH trial did not provide data on the incidence of HF with
combina&on therapies although the results from this meta-analysis
would support the use of diure&cs and ACE inhibitors or ARB drugs alone
or in combina&on for preven&on of heart failure in pa&ents with
hypertension.

One point that needs to be considered from the meta-analysis is that
both thiazide-type and thiazide-like are include together under the
group diure&c-based treatment. A really useful subgroup analysis would
be to look at thiazide-like diure&cs versus other an&hypertensive drug
groups.

Discussion

There are number things we need to consider when trea&ng pa&ents
with uncomplicated hypertension. There can be a tendency to chase
surrogate markers (blood pressure, HbA1c, cholesterol etc.) in general
prac&ce. Primary endpoint data from robust randomised controlled trials
should be helping drive clinical decisions and guidelines.

Chasing surrogate markers and targets can lead to the use of drugs with
poor outcome data such as doxazosin rather than drugs such as thiazidelike diure&cs, which have been shown in numerous studies to be
superior in reducing cardiovascular events. A further example of this
would be the use of bendroflumethiazide that has limited primary
endpoint data, compared to indapamide, but is s&ll used due to its blood
pressure lowering proper&es.

A recently published study (11) looked at newly diagnosed hypertensive
pa&ents in the UAE. Different classes of an&hypertensive drugs were
prescribed as monotherapy (71.5%) or combina&on therapy (28.4%).The
most frequently used an&hypertensive drugs were angiotensinconvertase inhibitors/angiotensin receptor blockers (ACEi/ARBs) (31.4%),
followed by calcium channel blockers (CCBs) 17.5%) beta-blockers
(15.4%) diure&cs (6%) and alpha-blockers (0.8%). Furthermore, the
propor&on prescribed a combina&on of ACEi/ARBs with CCBs or diure&cs
was 18.1% and 10.2%, respec&vely.

There is a danger of missing out on therapies that have addi&onal
proper&es to their original licensed indica&ons that could have an impact
on clinical outcomes. Perhaps we need to follow other countries such as
the US where the prescribing of thiazide-like diure&cs is much higher
compared to the UAE.

References

1. Na&onal Ins&tute for Health and Care Excellence. Hypertension in
   adults: diagnosis and management. CG127. November 2016. Available
   from: hWps://www.nice.org.uk/guidance/cg127

2. Wright JM, et al. First-line drugs for hypertension. Cochrane Database
   of Systema;c Reviews 2018, Issue 4. Art. No.: CD001841. DOI:
   10.1002/14651858.CD001841.pub3

3. Cooney D, et al. Diure&cs for hypertension: Hydrochlorothiazide or
   chlorthalidone? Cleveland Clinic Journal of Medicine. 2015
   August;82(8):527-533

4. PROGRESS Collabora&ve Group. Randomised trial of a perindoprilbased blood-pressure-lowering regimen among 6,105 individuals with
   previous stroke or transient ischaemic aWack. Lancet. 2001;358:1033–41.

5. Heller S, et al. A summary of the ADVANCE trial. Diabetes Care. 2009
   Nov; 32(Suppl 2): S357–S361.

6. Patel A, et al. Effects of a fixed combina&on of perindopril and
   indapamide on macrovascular and microvascular outcomes in pa&ents
   with type 2 diabetes mellitus (the ADVANCE trial): a randomised
   controlled trial. Lancet. 2007 Sep 8;370(9590):829-40.

7. The ALLHAT Officers and Coordinators for the ALLHAT. Collabora&ve
   research group major outcomes in high-risk hypertensive pa&ents
   randomized to angiotensin-conver&ng enzyme inhibitor or calcium
   channel blocker vs diure&c. JAMA. 2002;288(23):2981-2997

8. Koval PG, et al. How effec&ve is doxazosin compared with
   chlorthalidone in the treatment of hypertension? J Fam Pract. 2000
   Jul;49(7):597-8.

9. Jamerson K, et al. Benazepril plus Amlodipine or Hydrochlorothiazide
   for Hypertension in High-Risk Pa&ents. N Engl J Med 2008;
   359:2417-2428

10. ScaireWa S, et al. An&hypertensive treatment and development of
    heart failure in hypertension: a Bayesian network meta-analysis of
    studies in pa&ents with hypertension and high cardiovascular risk. Arch
    Intern Med. 2011 Mar 14;171(5):384-94

11. Bhagavathula, A.S, et al. Hypertension Control and GuidelineRecommended Target Blood Pressure Goal Achievement at an Early
    Stage of Hypertension in the UAE. J. Clin. Med. 2022, 11, 47.

The article writer :

Ian Jackson is a UK registered clinical pharmacist with over 30 years’ experience. Roles have included Senior Lecturer in Pharmacy and Therapeutics at the UK-based University of Lincoln School of Medicine and School of Pharmacy (8 years). As well as working as an NHS Prescribing Advisor for over 20 years. Ian trained as a medical herbalist and has developed a clinical module on evidenced based natural therapies for a leading medical school. 

E-MAIL: ian.pharmacist@gmail.com