Download Population Science/Epidemiology

Population Science/Epidemiology
Chlorthalidone Reduces Cardiovascular Events Compared
With Hydrochlorothiazide
A Retrospective Cohort Analysis
Michael P. Dorsch, Brenda W. Gillespie, Steven R. Erickson, Barry E. Bleske, Alan B. Weder
See Editorial Commentary, pp 665– 666
Downloaded from http://hyper.ahajournals.org/ by guest on April 30, 2017
Abstract—There is significant controversy around whether chlorthalidone (CTD) is superior to hydrochlorothiazide
(HCTZ) in hypertension management. The objective of this analysis was to evaluate the effects of CTD compared with
HCTZ on cardiovascular event (CVE) rates. We performed a retrospective observational cohort study from the Multiple
Risk Factor Intervention Trial data set from the National Heart, Lung, and Blood Institute. The Multiple Risk Factor
Intervention Trial was a cardiovascular primary prevention trial where participants were men 35 to 57 years of age
enrolled and followed beginning in 1973. CVEs were measured yearly, and time to event was assessed by Cox
regression. Systolic blood pressure, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein
cholesterol, triglyceride, potassium, glucose, and uric acid were measured yearly. The difference between groups was
evaluated by repeated-measures mixed modeling, and each model was adjusted for predictors of each variable. CVEs
were significantly lower in those on CTD (adjusted hazard ratio: 0.51 [95% CI: 0.43 to 0.61]; P⬍0.0001) and on HCTZ
(adjusted hazard ratio: 0.65 [95% CI: 0.55 to 0.75]; P⬍0.0001) compared with those who took neither drug. When
comparing the 2 drugs, CTD had significantly fewer CVEs compared with HCTZ (P⫽0.0016). CTD displayed
significantly lower SBP (P⬍0.0001), lower total cholesterol (P⬍0.0001), lower low-density lipoprotein cholesterol
(P⫽0.0009), lower potassium (P⫽0.0003), and higher uric acid (P⬍0.0001) over time compared with HCTZ. In
conclusion, both HCTZ and CTD reduce CVEs compared with neither drug. When comparing both drugs, CTD reduces
CVEs more than HCTZ, suggesting that CTD may be the preferred thiazide-type diuretic for hypertension in patients
at high risk of CVEs. (Hypertension. 2011;57:689-694.) ● Online Data Supplement
Key Words: chlorthalidone 䡲 hydrochlorothiazide 䡲 hypertension 䡲 thiazide diuretic 䡲 antihypertensive agents
T
pamide. The choice between HCTZ and CTD for the treatment of hypertension is controversial and has recently been a
topic of editorials, meta-analyses, and reviews.3–7 CTD may
have superior 24-hour blood pressure control compared with
HCTZ.8 There are also several studies that demonstrate the
efficacy of both CTD and HCTZ,9 –15 but one recent clinical
trial has led some to question the effectiveness of HCTZ.16
The Multiple Risk Factor Intervention Trial (MRFIT) was
a large primary prevention trial that began in 1973.17 It was
the first study to hint that CTD may be superior to HCTZ. In
1980, the MRFIT Policy Advisory Board changed the hypertension treatment protocol, recommending CTD over HCTZ
for initial hypertension therapy.13 This was done because the
coronary heart disease (CHD) mortality of special intervention clinics using HCTZ was 44% higher than other clinics
(P⫽0.23). The MRFIT data set provides the opportunity to
hiazide-type diuretics are recommended as a primary
choice for the treatment of hypertension.1 In the case of
uncomplicated hypertension, without the presence of comorbid conditions that would benefit from another class of
antihypertensive drug, the thiazide-type diuretics are considered one of the initial agents to use. If not used as initial
therapy, thiazide-type diuretics are recommended as the next
agent added to existing therapy, should more aggressive
treatment be desired.
Clinicians have several options when prescribing a
thiazide-type diuretic. The most commonly prescribed
thiazide-type diuretic is hydrochlorothiazide (HCTZ).2 This
agent is listed 9 different times as HCTZ or as 1 of 2 agents
in a combination drug in the 2008 top 200 drugs prescribed in
the United States. Other thiazide-type diuretics include
chlorthalidone (CTD), chlorothiazide, metolazone, and inda-
Received August 18, 2010; first decision September 7, 2010; revision accepted January 31, 2011.
From the University of Michigan Health System (M.P.D.), Ann Arbor, MI; School of Public Health (B.W.G.), College of Pharmacy (S.R.E., B.E.B.),
and Medical School (A.B.W.), University of Michigan, Ann Arbor, MI.
This article was prepared using a limited access data set obtained from the National Heart, Lung, and Blood Institute and does not necessarily reflect
the opinions or views of the MRFIT or the National Heart, Lung, and Blood Institute.
Correspondence to Michael P. Dorsch, University of Michigan Health Systems, B2D301 SPC 5008, 1500 East Medical Center Dr, Ann Arbor, MI
48109-5008. E-mail [email protected]
© 2011 American Heart Association, Inc.
Hypertension is available at http://hyper.ahajournals.org
DOI: 10.1161/HYPERTENSIONAHA.110.161505
689
690
Hypertension
April 2011
directly compare HCTZ and CTD, because the documentation of these drugs was done at each follow-up visit. The
objective of the current analysis was to evaluate the cardiovascular end points between patients who took HCTZ and
CTD. The secondary objective was to compare the change in
systolic blood pressure (SBP), total cholesterol (TC), lowdensity lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, triglyceride, potassium, glucose, and
uric acid of patients in the 2 groups.
Methods
Design Overview
Downloaded from http://hyper.ahajournals.org/ by guest on April 30, 2017
We performed a retrospective observational cohort analysis comparing the clinical effects of CTD with HCTZ. Data for this analysis
were collected as part of the MRFIT, which was a randomized
cardiovascular primary prevention trial in 12 866 men, 35 to 57 years
of age, who were enrolled and followed beginning in 1973. The
study tested whether a multifaceted intervention program could
reduce mortality from CHD. The specifics of the study have been
published previously.17 Briefly, patients were randomized to a
special intervention group consisting of treatment for hypertension,
smoking cessation counseling, and dietary guidance or to usual care
from their healthcare provider with follow-up of 7 years. The
hypertension intervention involved a stepped-care approach, which
began with HCTZ or CTD at either 50 or 100 mg daily, with weight
and sodium reduction. The steps that followed were to add antiadrenergic drug therapy, an arteriolar vasodilator, and then guanethidine in step 4. The goal blood pressure was a diastolic blood pressure
of 89 mm Hg or 10-mm Hg less than the average diastolic blood
pressure at that visit, whichever is lower. The use of CTD and HCTZ
was assessed by interview at each follow-up visit for 7 years, along
with many other clinical variables. Our analysis was approved by the
University of Michigan Institutional Review Board.
Patient Population
Eligible patients for the MRFIT were in the upper 15% of risk for
death from CHD based on risk factors of elevated cholesterol,
elevated diastolic blood pressure, and cigarette smoking from the
Framingham Heart Study. Patients also could not have had preexisting, definite clinical CHD before entering the study.
Exposure Definition
The use of CTD or HCTZ was considered first line therapy for the
hypertension intervention in the MRFIT. Use of either CTD or
HCTZ was documented yearly. It was assumed that if the patient
indicated taking the diuretic at the follow-up assessment, they had
taken the medication in the year before that visit. This exposure was
then used as a time-dependent variable to include all of the follow-up
time points in which patients were exposed to either drug. Based on
the time-dependent nature of this analysis, an individual patient
could cross over from one drug group to the other over time. If a
patient reported stopping either drug during the follow-up, then those
time points were defined as “stopped drug.” This definition of
exposure was undertaken to capture the cross over that is seen in the
data set. Figure 1 provides further explanation. For the secondary
outcomes, the exposure was defined as those patients who stated
adherence to either CTD or HCTZ for 6 of the 7 yearly visits.
Outcome Measures
The full Outcome Measures section defining each end point is
described in the online Data Supplement (please see http://hyper.
ahajournals.org). The primary outcome of this analysis was cardiovascular events that were adjudicated and prespecified in the MRFIT
data set. Nonfatal events consisted of clinical myocardial infarction
(MI), MI determined by annual ECG, stroke, coronary artery bypass
surgery, ECG-defined left ventricular hypertrophy, heart failure,
angina determined by Rose questionnaire, and peripheral artery
Year 1
Year 2
Year 3
Year 4
Figure 1. Exposure definition for the primary analysis. The
dashed line is CTD, the dotted line is HCTZ, and the solid line is
when one of the drugs was stopped. Patients were not entered
into the data set until the first exposure to CTD or HCTZ. As
seen, this example patient did not enter into the data set until
year 2. Year 2 exposure was considered CTD. During year 3,
the patient stopped taking CTD and was considered stopped
drug for year 3. Year 4 was then considered HCTZ.
occlusive disease. Nonfatal CVEs were documented at the yearly
follow-up in the MRFIT data set; the specific dates of each event
were not recorded. Cause-specific mortality is based on death
certificates coded by trained nosologists using the International
Classification of Diseases, 9th Revision. The study personnel adjudicated these cardiovascular deaths to be caused by coronary
surgery, congestive heart failure, sudden death, MI, hypertension
with renal failure, hypertension with left ventricular failure, stroke,
and other cardiovascular disease. Secondary outcomes were the
change in SBP, TC, LDL cholesterol, HDL cholesterol, triglyceride,
potassium, glucose, and uric acid at each time point of follow-up
between CTD and HCTZ.
Statistical Analysis
The primary outcome was the time from first follow-up documenting
CTD or HCTZ exposure during the study to the first cardiovascular
event. A time-to-event analysis was performed using Cox regression,
stratifying by the MRFIT follow-up clinic. The covariates of primary
interest were the prescription of CTD, HCTZ, or neither drug,
entered in the model as time-dependent covariates. This method
allowed us to define exposure at each time point for a specific patient
and examine the events when patients were in that exposure time.
Models were adjusted for baseline age, race, smoking status, MRFIT
randomized group, diuretic dose, SBP, LDL, HDL, and baseline
hypertension treatment, which are either known predictors of CVEs
or had significant treatment imbalance at baseline. Patients without
an event were censored at last follow-up. Secondary analyses were
performed to investigate differences in the mean levels of SBP, TC,
LDL, HDL, triglyceride, potassium, glucose, and uric acid between
those on CTD versus HCTZ over time. Mixed-model analyses
controlling for baseline predictors for each variable were used with
a random slope and intercept for each patient. Tests for a difference
between those on CTD versus HCTZ were performed overall, as well
as at each time point. All of the statistical analyses were performed
using SAS software version 9.2 (SAS Institute, Inc).
Results
Of the 12 866 patients in the MRFIT, 6441 were initially
prescribed either CTD (N⫽2392) or HCTZ (N⫽4049) with a
median follow-up of 6 years. Seventy-five percent of the
initial HCTZ patients and 76% of the initial CTD patients
crossed over into either the other drug group or the drugstopped group. Of the initial HCTZ patients, 29% crossed
over into the CTD group, whereas 37% of the initial CTD
patients crossed over into the HCTZ group sometime during
follow-up. Overall, 46% of the time patients were in the
HCTZ group, 32% in the CTD group, and 23% in the
stopped-drug group, accounting for 33 614 years of exposure
among the 6441 patients. Baseline characteristics of patients
based on the initial diuretic are presented in Table 1. There
were no statistically significant baseline differences in demographics, cardiovascular risk factors, and metabolic parame-
Dorsch et al
Table 1. Baseline Demographics for Those Initially Receiving
CTD and HCTZ
Variables
HCTZ
(N⫽4049)
CTD
(N⫽2392)
Age, y
46.9⫾5.9
46.7⫾5.7
0.1234
2740 (67.7)
HTN
MI
Angina
DM
1563 (65.3)
0.0647
8 (0.2)
5 (0.21)
0.9832
19 (0.47)
10 (0.42)
0.2113
77 (3.2)
0.9126
125 (3.1)
LVH on ECG
P
8 (0.2)
4 (0.17)
0.7854
Smoker
2242 (55.4)
1346 (56.3)
0.4827
Total cholesterol, mg/dL
249.5⫾36.4
249.3⫾35.8
0.8072
HDL, mg/dL
42.6⫾11.6
42.9⫾12.4
0.3779
LDL, mg/dL
156.5⫾35.2
156.8⫾35.6
0.7427
SBP, mm Hg
142.1⫾13.7
142.7⫾13.1
0.1119
Downloaded from http://hyper.ahajournals.org/ by guest on April 30, 2017
Serum creatinine, mg/dL
1.11⫾0.152
Glucose, mg/dL
99.6⫾15.7
Potassium, mEq/L
1.10⫾0.152
100.3⫾14.5
0.2497
0.0986
4.4⫾0.5
4.4⫾0.5
0.9959
27.8⫾3.5
27.8⫾3.4
0.7393
⬎50 mg
1149 (28.4)
1145 (47.9)
⬍0.0001
ⱕ50 mg
2900 (71.6)
1247 (52.1)
White
3535 (87.3)
2116 (88.5)
0.1718
Black
385 (9.5)
213 (8.9)
0.4198
BMI, kg/m2
Dose
Race
Other
129 (3.2)
63 (2.6)
1814 (44.8)
2007 (83.9)
⬍0.0001
Any HTN medication
1406 (34.7)
697 (29.1)
⬍0.0001
Diuretic
1291 (31.2)
627 (26.2)
⬍0.0001
3 (0.07)
3 (0.13)
0.2249
564 (13.9)
242 (10.1)
⬍0.0001
MRFIT intervention, n (%)
0.2080
Medications
Ganglionic blockers
Methyldopa
HTN indicates hypertension; DM, diabetes mellitus; LVH, left ventricular
hypertrophy; SBP, systolic blood pressure; BMI, body mass index. Values are
represented as mean⫾SD or n (%). P values were calculated by Student t test
for age, total cholesterol, HDL, LDL, SBP, serum creatinine, glucose, potassium,
and BMI, whereas all others were calculated by ␹2 or Fisher exact test.
ters. However, significantly more CTD patients were randomized to the MRFIT intervention than the control and were
taking higher doses of a thiazide-type diuretic. The HCTZ
group had a significantly greater proportion of patients
receiving hypertensive drug therapy at baseline. A total of
1244 cardiovascular events were documented over the 7 years
of follow-up.
Cardiovascular Events
CVEs were significantly lower in those on CTD (adjusted
hazard ratio [aHR]: 0.51 [95% CI: 0.43 to 0.61]; P⬍0.0001)
and on HCTZ (aHR: 0.65 [95% CI: 0.55 to 0.75]; P⬍0.0001)
compared with those on neither drug, adjusting for baseline
age, race, smoking status, MRFIT randomized intervention
group, diuretic dose, SBP, LDL, HDL, and hypertension
treatment. When comparing the 2 drugs, those on CTD had
significantly fewer CVEs compared with those on HCTZ
CTD Reduces CVEs Compared With HCTZ
691
(aHR: 0.79 [95% CI: 0.68 to 0.92]; P⫽0.0016). Figure 2
displays the adjusted event-free probabilities for each group
over time for a subject with the covariate values specified in
the Figure 2 legend. The adjusted HR for each individual
component of the composite primary outcome is represented
in Table 2. The individual events that contributed most
importantly to the difference between HCTZ and CTD were
clinical MI, ECG MI, coronary artery bypass, Rose angina,
and peripheral artery occlusive disease. The unadjusted estimates of the individual components of the composite primary
outcome are available in the online Data Supplement Table
S1 (please see http://hyper.ahajournals.org).
Interactions among the drug variables, the MRFIT intervention groups, and time were tested. The MRFIT intervention significantly altered the effects of the drugs (drug by
study intervention interaction P⫽0.0003). The MRFIT intervention group enhanced the effect of CTD compared with no
drug in the intervention group (aHR: 0.47 [95% CI: 0.38 to
0.60]; P⬍0.0001), and the corresponding effect of HCTZ was
similar to the main effects compared with no drug (aHR: 0.67
[95% CI: 0.53 to 0.84]; P⫽0.0006). Within the MRFIT
intervention, those on CTD had fewer events than those on
HCTZ, but the difference was not statistically significant
(aHR: 0.79 [95% CI: 0.59 to 1.08]; P⫽0.15). There was also
a significant interaction among drug, MRFIT intervention,
and time. Over time, the initial hazard reduction seen with
both drug groups diminished significantly (HCTZ by time
interaction aHR: 1.19 [95% CI: 1.08 to 1.32], P⫽0.0004;
CTD by time interaction aHR: 1.15 [95% CI: 1.03 to 1.27],
P⫽0.01). Those in the MRFIT intervention group receiving
CTD had significant reductions in CVE compared with those
on neither drug at year 1 (aHR: 0.29 [95% CI: 0.19 to 0.44];
P⬍0.0001) and 3 (aHR: 0.38 [95% CI: 0.29 to 0.50];
P⬍0.0001). Those in the MRFIT intervention group receiving HCTZ had significant reductions in CVE compared with
those on neither drug at year 1 (aHR: 0.38 [95% CI: 0.26 to
0.56]; P⬍0.0001) and 3 (aHR: 0.54 [95% CI: 0.42 to 0.70];
P⬍0.0001).
Differential Effects of CTD Compared
With HCTZ
Figure 3 shows each variable over time adjusted for baseline
predictors with asterisks delineating the time points that are
significantly different between groups. In the longitudinal
models comparing CTD with HCTZ over time, CTD patients
displayed significantly lower SBP (overall P⬍0.0001), TC
(overall P⬍0.0001), LDL (overall P⫽0.0009), and serum
potassium (overall P⫽0.0003) and higher uric acid over time
(overall P⬍0.0001) compared with HCTZ. Glucose over time
(overall P⫽0.1595) and triglyceride over time (overall
P⫽0.2648) did not differ between the groups. The specific
values are available in the online Data Supplement Table S2.
Discussion
Based on a post hoc analysis of a primary prevention study,
currently the only study that allows for direct comparison
between HCTZ and CTD on cardiovascular outcomes, we
were able to demonstrate that CTD significantly reduced
CVEs as compared with HCTZ. This finding is analogous to
692
Hypertension
April 2011
1.0
0.9
Event Free Probability
0.8
0.7
0.85
0.6
0.80
0.5
0.75
0.4
0.70
0.3
6.0
6.5
7.0
0.2
CTD
HCTZ
Drug Stopped
0.1
CTD versus Stopped Drug HR 0.51; 95% CI 0.43-0.61; p<0.0001
HCTZ versus Stopped Drug HR 0.64; 95% CI 0.55-0.75; p<0.0001
CTD versus HCTZ HR 0.79; 95% CI 0.68-0.92; p=0.0016
0
Downloaded from http://hyper.ahajournals.org/ by guest on April 30, 2017
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
Time (years)
Figure 2. Adjusted event-free probability of CVEs. The hazard ratio was estimated using the Cox model. Adjusted estimates were controlled for by age, race, smoking status, MRFIT randomized group, diuretic dose, SBP, LDL, HDL, and baseline hypertension treatment.
the morality data after a 10.5-year follow-up from MRFIT.13
In March 1980, the CHD mortality of special intervention
clinics using HCTZ was 44% higher than other clinics
(P⫽0.23). This led to a change in study protocol to the use of
CTD as the thiazide-type diuretic of choice. From April 1980,
the rate of CHD mortality in those clinics decreased by 28%
(P⫽0.04, for comparison of percentage difference in mortality in the 2 time points).
The results from our study further enhance the notion that
CTD may be the preferred thiazide-type diuretic for treatment
of hypertension. Obviously, one question that needs to be
addressed is why CTD showed better efficacy in reducing
cardiovascular events as compared with HCTZ when both
agents work in a pharmacologically similar fashion. Based on
our findings, an answer may be that CTD at the doses studied
is a more potent antihypertensive agent resulting in greater
blood pressure lowering and better outcomes. Recent data in
Table 2.
the literature support our findings that CTD appears to lower
SBP more than HCTZ.8
Another reason why patients in the CTD group had
improved outcomes may relate to metabolic profile of the
drugs. When evaluating the metabolic profile, the CTD group
had lower LDL and glucose values as compared with the
HCTZ group. The lower the LDL levels the greater the
chance of reducing cardiovascular events. Similarly, impaired
glucose tolerance may also contribute to cardiovascular
disease. The trends in glucose values were higher in the
HCTZ group compared with the CTD group. The reason for
the improved metabolic profile seen in the CTD group is not
known. However, more CTD patients were in the study
intervention group where lifestyle interventions were performed. We controlled for this intervention within our models, but there may have been residual effects that were not
accounted for by that covariate. In contrast to these positive
aHRs for Each Component of the Composite End Point
CTD vs Drug Stopped
Variables
HR (95% CI)
P
HCTZ vs Drug Stopped
HR (95% CI)
P
CTD vs HCTZ, P
Cardiovascular death
0.33 (0.13 to 0.82)
0.0174
0.38 (0.17 to 0.84)
0.0163
0.6959
Clinical MI
0.13 (0.08 to 0.21)
⬍0.0001
0.34 (0.24 to 0.47)
⬍0.0001
0.0001
0.0103
ECG MI
0.26 (0.16 to 0.42)
⬍0.0001
0.49 (0.33 to 0.71)
0.0002
Stroke
0.55 (0.21 to 1.43)
0.2192
0.73 (0.32 to 1.67)
0.4493
0.4837
CABG
0.19 (0.12 to 0.30)
⬍0.0001
0.45 (0.33 to 0.62)
⬍0.0001
0.0001
ECG LVH
0.65 (0.38 to 1.10)
0.1088
0.62 (0.39 to 0.97)
0.0367
0.8114
Rose angina
0.52 (0.40 to 0.67)
⬍0.0001
0.71 (0.57 to 0.88)
0.0019
0.0029
CHF
0.28 (0.03 to 2.87)
0.2854
0.25 (0.06 to 1.045)
0.0575
0.9229
PAOD
0.75 (0.44 to 1.27)
0.2765
1.24 (0.78 to 1.95)
0.3662
0.0218
The hazard ratios were estimated using the Cox model. All of the variables were controlled for by age, race,
smoking status, MRFIT randomized group, diuretic dose, SBP, LDL, HDL, and baseline HTN treatment. There can be
overlap in events. CABG indicates coronary artery bypass; CHF, congestive heart failure; PAOD, peripheral artery
occlusive disease; LVH, left ventricular hypertrophy.
Dorsch et al
A
144
142
*
D 162
HCTZ
CTD
Overall p<0.0001
CTD Reduces CVEs Compared With HCTZ
140
HCTZ
CTD
Overall p=0.0009
160
693
158
138
SBP (mmHg)
*
134
132
130
*
128
*
*
LDL (mg/dL)
156
136
*
126
150
148
122
144
120
142
118
140
0
112
1
2
3
4
Years
5
6
0
7
E
HCTZ
CTD
4.6
1
2
3
Years
4
5
6
Overall p=0.0003
4.5
HCTZ
CTD
4.4
4.3
Potassium (mEq/L)
*
108
106
104
102
*
4.2
4.1
*
*
*
*
*
3
4
Years
5
6
4.0
*
3.9
3.8
3.7
100
3.6
3.5
98
3.4
0
255
1
2
3
4
Years
5
6
0
7
F
Overall p<0.0001
HCTZ
CTD
*
245
240
7.7
1
*
*
235
7
HCTZ
CTD
Overall p=0.0001
*
7.5
*
2
7.6
250
Uric Acid (mg/dL)
C
Total Cholesterol (mg/dL)
Downloaded from http://hyper.ahajournals.org/ by guest on April 30, 2017
Overall p=0.1595
110
Glucose (mg/dL)
152
146
124
B
*
154
*
230
7.4
*
7.3
7.2
225
7.1
220
7.0
215
0
1
2
3
4
Years
5
6
7
0
1
2
3
4
Years
5
6
7
Figure 3. Effects of CTD vs HCTZ over time. Adjusted values were estimated using mixed models and represented as mean⫾SEM.
Asterisks (*) represent P⬍0.05. The overall P value in each figure is the drug-by-time interaction. SBP was adjusted by age, diuretic
dose, diagnosis of hypertension, and MRFIT randomization group. Glucose was adjusted by age, diuretic dose, body mass index, and
diagnosis of diabetes. TC and LDL were adjusted by diuretic dose and the MRFIT randomization group. Potassium was adjusted by
age, diuretic dose, and MRFIT randomization group. Uric acid was adjusted by diuretic dose and the diagnosis of diabetes mellitus.
findings, the CTD group had lower potassium levels and
higher uric acid levels. The reason for these findings may
again relate to the potency of CTD and the doses used in the
study. Despite the lower potassium levels and higher uric acid
levels, increases in events were not seen in the CTD group.
These findings are consistent with a previous MRFIT subgroup analysis in patients with baseline rest ECG abnormalities, suggesting that hypokalemia did not contributed to
694
Hypertension
April 2011
Downloaded from http://hyper.ahajournals.org/ by guest on April 30, 2017
coronary artery disease mortality rate in the special intervention group at 48 months of follow-up.12
Other reasons for improved outcomes with CTD may relate
to differences in pharmacokinetic and pharmacodynamic
properties between CTD and HCTZ. With regard to pharmacokinetics, the half-life of CTD (45 to 60 hours) is much
longer than the half-life of HCTZ (16 to 24 hours), and CTD
has a larger volume of distribution. In addition, CTD is
thought to be 1.5 to 2.0 times more potent than HCTZ, with
24-hour ambulatory blood pressure monitoring demonstrating
lower blood pressure with CTD during the nighttime hours.8
Furthermore, a recent meta-analysis also suggested that CTD
has greater effects on systolic blood pressure reductions as
compared with HCTZ.5 Greater systolic blood pressure reductions in concert with more consistent effects over 24 hours
with the use of CTD may have contributed to our study
findings.
There are inherent limitations in an observational study
design, in that we cannot exclude unmeasured confounding,
selection bias, and information bias that may have occurred
during the study. Unmeasured confounding could exist,
because the treating physician determined antihypertensive
drug selection, and we could not determine why HCTZ or
CTD was chosen as initial therapy or why patients switched
from one drug to another. We have attempted to reduce
selection bias by comparing the HCTZ and CTD groups in a
time-dependent fashion to a time period when people stopped
taking either drug; we believe that this gives more strength to
the conclusions of this analysis. Another limitation to this
analysis is that we could access only the data collected in
MRFIT. The nonfatal events were only documented during
each yearly follow-up so the exact time of each event is not
known. We hope that defining drug exposure for the year
before the documented event limits bias for each group
studied. Also, data on why patients stopped taking CTD or
HCTZ are unavailable and may pose information bias.
Despite these limitations, our finding that both the HCTZ and
CTD groups have better outcomes as compared with the
stopped drug group gives a measure confidence in our
approach to the data.
Perspectives
Our study demonstrates that both HCTZ and CTD reduce
CVEs compared with neither drug. When comparing both
drugs, CTD reduces CVEs more than HCTZ, suggesting that
CTD may be the preferred thiazide-type diuretic for hypertension in patients at high risk of cardiovascular events.
Prospective randomized, controlled trials are needed to confirm the effects on clinical end points, because this class of
medications is so commonly prescribed for hypertension.
Sources of Funding
The MRFIT Study was conducted and supported by the National
Institutes of Health/National Heart, Lung, and Blood Institute in
collaboration with the MRFIT Study Investigators. This manuscript
was prepared using a limited access dataset obtained from the
NHLBI and does not necessarily reflect the opinions or views of the
MRFIT or the NHLBI.
Disclosures
None.
References
1. Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL
Jr, Jones DW, Materson BJ, Oparil S, Wright JT Jr, Roccella EJ, for the
Joint National Committee on Prevention, Detection, Evaluation, and
Treatment of High Blood Pressure; National Heart, Lung, and Blood
Institute; National High Blood Pressure Education Program Coordinating
Committee. Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure.
Hypertension. 2003;42:1206 –1252.
2. Lamb E. Top 200 prescription drugs of 2008. Pharmacy Times Web site.
May 2009. Available at: http://www.pharmacytimes.com/issue/
pharmacy/2009/2009-05/RxFocusTop200Drugs-0509. Accessed April
10, 2010.
3. Pitt B. The role of chlorthalidone in patients with high-risk hypertension.
J Clin Hypertens (Greenwich). 2009;11:460 – 461.
4. Ernst ME, Carter BL, Basile JN. All thiazide-like diuretics are not
chlorthalidone: putting the ACCOMPLISH Study into perspective. J Clin
Hypertens (Greenwich). 2009;11:5–10.
5. Ernst ME, Carter BL, Zheng S, Grimm RH Jr. Meta-analysis of doseresponse characteristics of hydrochlorothiazide and chlorthalidone:
effects on systolic blood pressure and potassium. Am J Hypertens. 2010;
23:440 – 446.
6. Sica DA. Chlorthalidone: a renaissance in use? Expert Opin Pharmacother. 2009;10:2037–2039.
7. Kaplan NM. The choice of thiazide diuretics: why chlorthalidone may
replace hydrochlorothiazide. Hypertension. 2009;54:951–953.
8. Ernst ME, Carter BL, Goerdt CJ, Steffensmeier JJ, Phillips BB, Zimmerman MB, Bergus GR. Comparative antihypertensive effects of hydrochlorothiazide and chlorthalidone on ambulatory and office blood
pressure. Hypertension. 2006;47:352–358.
9. Amery A, Birkenhger W, Brixko P, Bulpitt C, Clement D, Deruyttere M,
De Schaepdryver A, Dollery C, Fagard R, Forette F. Mortality and
morbidity results from the European Working Party on High Blood
Pressure in the Elderly Trial. The Lancet. 1985;1:1349 –1354.
10. ALLHAT Officers and Coordinators for the ALLHAT Collaborative
Research Group. The Antihypertensive and Lipid-Lowering Treatment to
Prevent Heart Attack Trial: major outcomes in high-risk hypertensive
patients randomized to angiotensin-converting enzyme inhibitor or
calcium channel blocker vs diuretic. JAMA. 2002;288:2981–2997.
11. Hypertension Detection and Follow-up Program Cooperative Group.
Five-year findings of the hypertension detection and follow-up program:
reduction in mortality of persons with high blood pressure, including mild
hypertension. JAMA. 1979;242:2562–2571.
12. Multiple Risk Factor Intervention Trial Research Group. Baseline rest
electrocardiographic abnormalities, antihypertensive treatment, and mortality in the multiple risk factor intervention trial. Am J Cardiol. 1985;
55:1–15.
13. Multiple Risk Factor Intervention Trial Research Group. Mortality after
10 1/2 years for hypertensive participants in the Multiple Risk Factor
Intervention Trial. Circulation. 1990;82:1616 –1628.
14. SHEP cooperative research group. Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Program
(SHEP). JAMA. 1991;265:3255–3264.
15. MRC working party. Medical research council trial of treatment of hypertension in older adults: principal results. BMJ. 1992;304:405– 412.
16. Jamerson K, Weber M, Bakris G, Dahlf B, Pitt B, Shi V, Hester A, Gupte
J, Gatlin M, Velazquez E. Benazepril plus amlodipine or hydrochlorothiazide for hypertension in high-risk patients. N Engl J Med. 2008;359:
2417–2428.
17. Multiple Risk Factor Intervention Trial Research Group. Coronary heart
disease death, nonfatal acute myocardial infarction and other clinical
outcomes in the multiple risk factor intervention trial. Am J Cardiol.
1986;58:1–13.
Chlorthalidone Reduces Cardiovascular Events Compared With Hydrochlorothiazide: A
Retrospective Cohort Analysis
Michael P. Dorsch, Brenda W. Gillespie, Steven R. Erickson, Barry E. Bleske and Alan B.
Weder
Downloaded from http://hyper.ahajournals.org/ by guest on April 30, 2017
Hypertension. 2011;57:689-694; originally published online March 7, 2011;
doi: 10.1161/HYPERTENSIONAHA.110.161505
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2011 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://hyper.ahajournals.org/content/57/4/689
Data Supplement (unedited) at:
http://hyper.ahajournals.org/content/suppl/2011/03/04/HYPERTENSIONAHA.110.161505.DC1
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located,
click Request Permissions in the middle column of the Web page under Services. Further information about
this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Hypertension is online at:
http://hyper.ahajournals.org//subscriptions/
ONLINE DATA SUPPLEMENT
Chlorthalidone reduces cardiovascular events compared to hydrochlorothiazide: A
retrospective cohort analysis.
Short title: CTD reduces CV events compared to HCTZ
Michael P. Dorsch, Pharm.D., M.S., *‡ Brenda W. Gillespie, Ph.D., † Steven R. Erickson,
Pharm.D., *‡ Barry E. Bleske, Pharm.D., *‡ Alan B. Weder, M.D. *§
From *University of Michigan Health System, †University of Michigan School of Public Health,
‡University of Michigan College of Pharmacy, and §University of Michigan Medical School,
Ann Arbor, MI, USA.
Reprint requests and correspondence:
Michael P. Dorsch, Pharm.D., M.S.
University of Michigan Health Systems
B2D301 SPC 5008
1500 East Medical Center Drive
Ann Arbor, MI 48109-5008
Email: [email protected]
Phone: 734-763-5576
Fax 734-936-7027
1
Methods
Outcome Measures
The primary outcome of this analysis was cardiovascular events that were adjudicated
and pre-specified in the MRFIT dataset. Non-fatal events consisted of clinical myocardial
infarction (MI), MI determined by annual electrocardiogram (ECG), stroke, coronary artery
bypass (CABG) surgery, ECG defined left ventricular hypertrophy (LVH), heart failure (HF),
angina determined by Rose questionnaire, and peripheral artery occlusive disease (PAOD).
Clinical MI was defined as an admission for MI based on review of hospital records. The
records, which included admission sheet, discharge summary, ECGs and laboratory reports were
forwarded to the coordinating center where nurses abstracted them. If a cardiovascular diagnosis
was recorded in the chart, the records were reviewed by a cardiologist who evaluated the data in
3 categories of evidence for MI: history, serum enzyme levels, and ECGs. Each category was
classified as strong, moderate, weak, or no evidence of MI. Definite MI required strong evidence
in at least one category and strong or moderate evidence in a second category. The cardiologist
diagnosis was compared to a computer algorithm of abstracted data. Discordant cases were
assessed by two cardiologists, who reached a consensus on the diagnosis. The diagnosis of ECG
MI required both of the following conditions: change in Minnesota Code for Q-QS patterns
within any lead groups and significant change from the last screening ECG as judged by two
cardiologists. ECG defined LVH was done by measurements made on standard 12-lead ECG.
R-wave amplitude greater than 26 mm in either leads V5 or V6 or R-wave amplitude in V5 or
V6 plus S or QS amplitude in V1 greater than 35 mm was diagnostic of ECG LVH. Stroke was
diagnosed if hemiplegia or hemiparesis was present on physical exam. HF was defined by the
presence of at least 2 major and 2 minor criteria. Major criteria were paroxysmal nocturnal
2
dyspnea, distended neck veins, rales in the presence of unexplained dyspnea, and S3 gallop.
Minor criteria consisted of bilateral ankle edema, dyspnea on ordinary exertion, hepatomegaly,
decrease in vital capacity by one-third from maximum record, and tachycardia. The Rose
questionnaire, a standardized measure of angina, was performed to assess angina. A patient
experiencing angina would respond that they have chest pain when walking up a hill, in a
hurrying, or at a normal pace, if they stop or slow down when they get chest pain, if it goes away
within 10 minutes of stopping or slowing down, and if the pain is located left sternum or left
arm. PAOD was determined on physical exam if the patient had an absent or diminished femoral
pulse with the presence of either a bruit, ischemic ulcers, surgery for peripheral arterial disease,
or amputation of the lower limb. Non-fatal CV events were documented at the yearly follow-up
in the MRFIT dataset; the specific dates of each event were not recorded. Cause-specific
mortality is based on death certificates coded by trained nosologists using the Ninth Revision of
the International Classification of Diseases (ICD-9). ICD-9 rubrics 410-414 and 429.2 are used
to identify CHD deaths. ICD-9 code 429.2 identifies "cardiovascular disease, unspecified." The
study personnel adjudicated these CV deaths to be caused by coronary surgery, congestive heart
failure, sudden death, myocardial infarction, hypertension with renal failure, hypertension with
LV failure, stroke, and other cardiovascular disease. Secondary outcomes were the change in
SBP, TC, LDL cholesterol, HDL cholesterol, triglyceride, potassium, glucose, and uric acid at
each time point of follow up between CTD and HCTZ. The standard mercury sphygmanometer
was used to measure blood pressure. Sitting blood pressure was taken in a quiet room at each
follow up visit.
3
Table S1: Unadjusted hazard ratios for each component of the composite endpoint
CTD vs
Variable
CTD vs Drug Stopped
HCTZ vs Drug Stopped
HCTZ
HR (95% CI)
p-value
HR (95% CI)
p-value
p-value
CV Death
0.31 (0.13-0.74)
0.0083
0.38 (0.18-0.83)
0.0143
0.5640
Clinical MI
0.13 (0.08-0.21)
<0.0001
0.35 (0.25-0.47)
<0.0001
<0.0001
ECG MI
0.30 (0.18-0.48)
<0.0001
0.51 (0.35-0.74)
0.0004
0.0221
Stroke
0.83 (0.32-2.15)
0.6961
1.21 (0.51-2.88)
0.6695
0.3115
CABG
0.21 (0.14-0.33)
<0.0001
0.54 (0.40-0.73)
<0.0001
<0.0001
ECG LVH
0.98 (0.60-1.59)
0.93
1.06 (0.68-1.63)
0.8029
0.7127
Rose Angina
0.66 (0.51-0.84)
0.0008
0.98 (0.79-1.22)
0.8504
<0.0001
CHF
0.13 (0.02-1.11)
0.062
0.28 (0.07-1.11)
0.0699
0.4829
PAOD
0.87 (0.52-1.46)
0.5873
1.56 (0.99-2.47)
0.0558
0.0038
4
Table S2: Values Represented in Figure 3. Effects of CTD versus HCTZ over time
Adjusted values were estimated using mixed models. SBP was adjusted by age, diuretic dose,
diagnosis of HTN, and MRFIT randomization group. Glucose was adjusted by age, diuretic
dose, BMI, and diagnosis of diabetes. TC and LDL were adjusted by diuretic dose and the
MRFIT randomization group. Potassium was adjusted by age, diuretic dose and MRFIT
randomization group. Uric acid was adjusted by diuretic dose and the diagnosis of diabetes.
SEM = Standard Error of the Mean
S2A:
SBP (mmHg)
Year
0
1
2
3
4
5
6
7
HCTZ
mean
135.98
132.81
130.53
128.43
127.97
127.83
127.51
128.28
CTD
SEM
0.6429
0.5024
0.5164
0.5172
0.5213
0.5314
0.5456
0.6364
mean
140.82
130.49
126.72
125.75
125.9
125.48
126.84
127.26
p-value
SEM
0.6724
0.5325
0.5487
0.5499
0.5521
0.5623
0.578
0.7066
<.0001
0.0022
<.0001
0.0004
0.0061
0.0022
0.3949
0.2824
S2B:
Glucose (mg/dL)
Year
0
1
2
3
4
5
6
7
HCTZ
mean
SEM
102.51
0.8348
99.1265
0.6629
100.11
0.7179
101.9
0.7656
104.01
0.8245
106.62
0.8965
108.72
0.9758
109.67
1.1412
CTD
mean
103.29
99.0201
99.6077
100.84
102.36
103.59
106.25
107.13
p-value
SEM
0.8717
0.7045
0.7632
0.8128
0.8738
0.9497
1.03
1.24
0.4302
0.916
0.6335
0.3438
0.1711
0.0202
0.0812
0.1326
5
S2C:
TC (mg/dL)
Year
0
1
2
3
4
5
6
7
HCTZ
mean
249.6
241.48
240.06
237.88
236.46
234.75
234.66
233.59
CTD
SEM
1.7067
1.4749
1.5111
1.5239
1.5396
1.5637
1.5925
1.7529
mean
251.67
235.17
234.6
234.1
231.93
228.39
231.98
231
p-value
SEM
1.8039
1.5758
1.6155
1.6294
1.6431
1.6677
1.6994
1.9245
0.3724
0.0066
0.0189
0.1066
0.0555
0.0079
0.2697
0.3377
S2D:
LDL (mg/dL)
Year
0
2
4
6
HCTZ
mean
157.17
151.46
147.27
146.84
CTD
SEM
2.5295
1.6976
1.7091
1.7341
mean
156.74
145.52
144.01
143.88
p-value
SEM
2.5901
1.794
1.7938
1.819
0.8475
0.0069
0.1382
0.1837
S2E:
Potassium (mEq/L)
Year
0
1
2
3
4
5
6
7
HCTZ
mean
3.9499
4.1664
4.1218
4.0824
4.0739
4.0509
4.0292
3.9829
CTD
SEM
0.02541
0.01997
0.02037
0.02038
0.02037
0.02076
0.02121
0.02553
mean
3.9232
3.9019
3.8511
3.8505
3.7919
3.7553
3.7502
3.7401
p-value
SEM
0.03056
0.026
0.02602
0.02583
0.02612
0.02658
0.02684
0.03565
0.4126
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
<.0001
S2F:
Uric Acid (mg/dL)
Year
HCTZ
mean
CTD
SEM
mean
p-value
SEM
6
0
1
2
3
4
5
6
7
7.306
7.3174
7.2391
7.1241
7.1835
7.1534
7.1942
7.1777
0.05755
0.04897
0.04839
0.04854
0.04922
0.05057
0.0525
0.06341
7.2998
7.5923
7.4402
7.2728
7.2222
7.2796
7.1941
7.2273
0.06016
0.05225
0.05249
0.05271
0.05326
0.05427
0.05584
0.07
0.9327
0.0001
0.0046
0.0371
0.5922
0.0882
0.9993
0.5998
7