Eplerenona para la hipertensión
Resumen
Antecedentes
La eplerenona es un bloqueador de los receptores de aldosterona, derivada químicamente de la espironolactona. En Canadá, se indica el uso como tratamiento adyuvante para reducir la mortalidad en los pacientes con insuficiencia cardíaca que presentan disfunción sistólica ventricular izquierda e insuficiencia cardíaca crónica sistólica de clase II según la New York Heart Association (NYHA). También se usa como tratamiento adyuvante para los pacientes con insuficiencia cardíaca después del infarto de miocardio. Además, está indicada para el tratamiento de la hipertensión esencial leve y moderada en los pacientes que no pueden ser tratados de manera adecuada con otros agentes. Es importante determinar la repercusión clínica de todos los hipotensores, incluidos los antagonistas de la aldosterona, para apoyar el uso continuo en la hipertensión esencial. En ninguna revisión sistemática anterior se ha evaluado el efecto de la eplerenona sobre la morbilidad cardiovascular, la mortalidad y la magnitud de la disminución de presión arterial en los pacientes con hipertensión.
Objetivos
Evaluar los efectos de la monoterapia con eplerenona versus placebo para la hipertensión primaria en adultos. Los resultados de interés fueron: la mortalidad por todas las causas, los eventos cardiovasculares (infarto de miocardio mortal o no mortal), los eventos cerebrovasculares (accidentes cerebrovasculares mortales o no mortales), los eventos adversos o los retiros debidos a los eventos adversos y la presión arterial sistólica y diastólica.
Métodos de búsqueda
Se hicieron búsquedas en el registro especializado del Grupo Cochrane de Hipertensión (Cochrane Hypertension Specialised Register), CENTRAL, MEDLINE, Embase, y en dos registros de ensayos hasta 3 marzo 2016. Se buscaron las referencias de los estudios recuperados para identificar cualquier estudio omitido en la búsqueda inicial. También se buscaron los datos no publicados mediante contacto con los autores correspondientes de los estudios incluidos y las compañías farmacéuticas que participaron en los estudios sobre monoterapia con eplerenona en la hipertensión primaria. En la búsqueda no hubo restricciones de idioma.
Criterios de selección
Se seleccionaron los ensayos controlados con placebo aleatorios que estudiaran pacientes adultos con hipertensión primaria. Se excluyeron los estudios de pacientes con hipertensión secundaria o hipertensión gestacional y los estudios en que los pacientes recibían múltiples hipotensores.
Obtención y análisis de los datos
Tres autores de la revisión, de forma independiente, revisaron los resultados de búsqueda para encontrar estudios que cumplieran los criterios. Tres autores de la revisión, de forma independiente, extrajeron los datos y evaluaron la calidad de los ensayos utilizando un formulario estandarizado de extracción de datos. Un cuarto autor de la revisión independiente resolvió discrepancias o desacuerdos. Se realizó la extracción de datos y la síntesis usando un formato estandarizado en Covidence. Se realizó el análisis de datos utilizando Review Manager 5.
Resultados principales
Un total de 1437 pacientes adultos participaron en los cinco estudios aleatorios de grupos paralelos, con una duración del tratamiento que varió de 8 a 16 semanas. Las dosis diarias de eplerenona variaron entre 25 mg y 400 mg. El metanálisis de estos estudios mostró una reducción de la presión arterial sistólica de 9,21 mmHg (IC del 95%: −11,08 a −7,34; I2 = 58%) y una reducción de la presión diastólica de 4,18 mmHg (IC del 95%: − 5,03 a −3,33; I2 = 0%) (pruebas de calidad moderada).
Puede haber un efecto de dosis respuesta de la eplerenona en la reducción de la presión arterial sistólica en dosis de 400 mg/día. Sin embargo, este resultado es incierto, ya que se basa en un único estudio incluido con evidencia de muy baja calidad. En general, no parece que haya una dosis respuesta clínicamente de reducción de la presión arterial sistólica ni diastólica con las dosis de eplerenona de 50 mg a 400 mg por día. No pareció que hubiera diferencias en el número de pacientes que se retiraron por los eventos adversos ni en el número de pacientes con al menos un evento adverso en el grupo de eplerenona en comparación con el placebo. Sin embargo, sólo tres de los cinco estudios incluidos informaron eventos adversos. La mayoría de los estudios incluidos fueron de calidad moderada, ya que el riesgo de sesgo en los dominios múltiples se consideró como poco claro en la evaluación del “Riesgo de sesgo”.
Conclusiones de los autores
La eplerenona en dosis de 50 a 200 mg/día reduce la presión arterial en los pacientes con hipertensión primaria en 9,21 mmHg sistólicos y 4,18 mmHg diastólicos en comparación con el placebo, sin diferencia en el efecto entre las dosis de 50 mg/día a 200 mg/día. Una dosis de 25 mg/día no produjo una reducción estadísticamente significativa de la presión arterial sistólica ni diastólica, y la evidencia para las dosis superiores a 200 mg/día es insuficiente. No existe en la actualidad evidencia disponible para determinar el efecto de la eplerenona sobre resultados clínicamente significativos, como la mortalidad o la morbilidad en los pacientes hipertensos. La evidencia disponible sobre efectos secundarios es insuficiente y de mala calidad, lo que hace imposible establecer conclusiones acerca del daño potencial asociado con el tratamiento con eplerenona en pacientes hipertensos.
PICO
Resumen en términos sencillos
Eplerenona para la hipertensión
Pregunta de la revisión
El objetivo de esta revisión fue determinar la efectividad de la eplerenona para reducir la presión arterial, el perfil de efectos secundarios y la repercusión sobre resultados clínicamente significativos, como la mortalidad y la morbilidad.
Antecedentes
Los médicos han usado la eplerenona para el tratamiento de la hipertensión desde 2002. Es importante determinar la repercusión clínica de todos los hipotensores usados para apoyar el uso continuo en pacientes con hipertensión esencial. Se efectuaron búsquedas en las bases de datos múltiples y se hallaron cinco estudios aptos en 1437 pacientes que recibieron eplerenona o ninguna medicación de manera aleatoria.
Características de los estudios
Las dosis de eplerenona usada en estos estudios variaron de 25 mg a 400 mg por día. Estos estudios siguieron a los pacientes por 8 a 16 semanas mientras eran tratados. Ninguno de los estudios informó resultados clínicamente significativos de la eplerenona, por ejemplo, si la eplerenona puede reducir los ataques cardíacos, el accidente cerebrovascular o la muerte en comparación con el placebo. Sólo tres de los cinco estudios informaron sobre los efectos secundarios.
Resultados clave
No hay en la actualidad evidencia de que la eplerenona tenga un efecto beneficioso sobre la esperanza de vida ni las complicaciones relacionadas con la hipertensión (p.ej. ataque cardíaco, accidente cerebrovascular). La evidencia sobre el riesgo de efectos secundarios con la eplerenona es limitada y de calidad deficiente; es difícil estimar el grado de daño potencial con la eplerenona versus placebo. Este metanálisis indica que la eplerenona en dosis de 50 a 200 mg/día reduce la presión arterial sistólica en cerca de 9 mmHg y la diastólica en 4 mmHg, en comparación con ninguna medicación.
Calidad de la evidencia
La calidad de los cinco ensayos incluidos se consideró moderada, ya que los autores no describieron de manera exhaustiva parte de su metodología.
Authors' conclusions
Summary of findings
| Eplerenone compared with placebo for primary hypertension | |||
| Patient or population: adults with primary hypertension Settings: primary care Intervention: eplerenone (25‐400 mg) Comparison: placebo | |||
| Outcomes | Relative effect | No of participants | Quality of the evidence |
| Effect on systolic blood pressure (mean difference in systolic blood pressure) | (MD −9.21 mmHg, 95% CI −11.08 to −7.34) | 1437 | ⊕⊕⊕⊝ |
| Effect on diastolic blood pressure (mean difference in diastolic blood pressure) | (MD −4.18 mmHg, 95% CI −5.03 to −3.33) | 1437 | ⊕⊕⊕⊝ |
| Any adverse event | OR 1.07 (0.82 to 1.41) | 1105 (3) | ⊕⊕⊝⊝ |
| Adverse event leading to withdrawal | OR 1.10 (0.47 to 2.55) | 1105 (3) | ⊕⊕⊝⊝ |
| CI: confidence interval; OR: odds ratio. | |||
| GRADE Working Group grades of evidence | |||
| aDowngraded due to high risk of bias in many of the studies. | |||
Background
Description of the condition
Hypertension is associated with structural changes in the heart and blood vessels, which can lead to cardiovascular mortality and morbidity (e.g. cardiovascular disease, stroke, peripheral vascular disease, and renal disease). Hypertension is generally defined as having a systolic blood pressure (SBP) of 140 mmHg or more, a diastolic blood pressure (DBP) of 90 mmHg or more, or both (Arguedas 2009; CHEP 2015). For every 20 mmHg increase in SBP and 10 mmHg increase in DBP (through the range of 115/75 to 185/115 mmHg) in people aged 40 to 70 years, the risk of cardiovascular disease‐related morbidity doubles (JNC 7). Epidemiologic studies have shown increased blood pressure to be associated with increased incidence of stroke, ischemic heart disease, and other vascular mortality (Lewington 2002). While blood pressure is a surrogate goal of therapy for the prevention of hypertension‐associated target‐organ damage (DiPiro 2014), no research has convincingly shown that lowering blood pressure below the target value of 140/90 mmHg reduces cardiovascular morbidity and mortality (Arguedas 2009). At present, there is no definitive threshold to predict the blood pressure above which treatment provides more benefit than harm (Arguedas 2009). In very elderly individuals aged 75 years or older, there is no established association between SBP and all‐cause mortality (Banach 2014). Interestingly, there is some evidence that blood pressure values below 140/70 mmHg are associated with excess mortality in individuals aged 85 and over (Van Bemmel 2006). Current recommendations on target blood pressures are mostly based on expert opinion and cannot be generalized to all age groups (NCGC 2011). This variability underscores the importance of finding safe and effective antihypertensive agents that demonstrate a proven benefit for improving hard clinical outcomes (morbidity and mortality), rather than chasing arbitrary blood pressure values.
Description of the intervention
Eplerenone is an aldosterone receptor blocker that is chemically derived from spironolactone. In a recent Cochrane Review, spironolactone, considered fourth‐line therapy for hypertension in patients already treated with multiple medications, was shown to reduce systolic/diastolic blood pressure by approximately 20/7 mmHg compared to placebo. However, the review found no evidence on the effect of spironolactone on clinical outcomes in hypertensive patients (Batterink 2010). This review attempts to define the effect of eplerenone, another aldosterone antagonist, on hard clinical outcomes in hypertensive patients.
Compared to spironolactone, eplerenone exhibits less affinity for androgen, progesterone, and glucocorticoid receptors (Katzung 2012). As a result, many authors consider that it reduces the risk of gynaecomastia, menstrual irregularities, and sexual dysfunction. Eplerenone undergoes extensive hepatic metabolism, primarily through the enzyme cytochrome P450 3A4 (CYP3A4) to inactive metabolites, so clinicians should avoid co‐administration of potent CYP3A4 inhibitors (Muldowney 2009). Eplerenone itself neither induces nor inhibits CYP3A4 and does not alter the pharmacokinetics of co‐administered substrates or inducers of CYP3A4, highly protein‐bound drugs, or highly renally cleared drugs. As a result, dosage adjustments of other drugs are generally not necessary. Less than 5% of the administered dose is cleared unchanged in the urine and faeces. Current data suggest that adverse effects are generally rare and mild with eplerenone therapy. These include hyperkalemia, dizziness, elevation in serum creatinine, diarrhea, cough, fatigue, dyslipidemia, abdominal pain, and albuminuria (Muldowney 2009). Hyperkalemia is the most significant adverse effect of eplerenone and may result in fatal cardiac arrhythmias if serum potassium levels are inadequately monitored. Patients with declining renal function are at greater risk for hyperkalemia. Evidence also suggests that diabetic patients with proteinuria may be at increased risk of high serum potassium levels. No data is currently available for the use of eplerenone in patients with severe hepatic dysfunction, although those with mild to moderate impairment (Child‐Pugh class B) do not require dosage adjustments (eCPS 2014).
Eplerenone is contraindicated in people with known hypersensitivity to the drug or its excipients, clinically significant hyperkalemia or serum potassium of more than 5.0 mmol/L at initiation of treatment, severe hepatic dysfunction (Child‐Pugh class C), moderate to severe renal dysfunction (eGFR < 50 mL/min/1.73 m2), type 2 diabetes with microalbuminuria, and in people taking potassium‐sparing diuretics, potassium supplements, or potent CYP3A4 inhibitors (eCPS 2014).
In Canada, eplerenone is indicated for use as adjunctive therapy to reduce mortality for heart failure patients with New York Heart Association (NYHA) class II systolic chronic heart failure and left ventricular systolic dysfunction, and as adjunctive therapy for patients with heart failure following myocardial infarction. Additionally, it is indicated for treating mild and moderate essential hypertension in patients who cannot be treated adequately with other agents (eCPS 2014). The Canadian Hypertension Education Program (CHEP) recommends using aldosterone antagonists as adjuncts for patients with systolic dysfunction (ejection fraction < 40%) and recent cardiovascular hospitalization, acute myocardial infarction, elevated brain‐type natriuretic peptide (BNP) or N‐terminal pro b‐type BNP (NT‐proBNP) level, or NYHA class II to IV symptoms (CHEP 2015). The American Society of Hypertension recommends the addition of aldosterone antagonists when standard three‐drug regimens (angiotensin‐converting enzyme inhibitor or angiotensin receptor blocker/calcium channel blocker/diuretic) are ineffective in treatment‐resistant patients (Weber 2014).
How the intervention might work
The renin‐angiotensin‐aldosterone system (RAAS) plays a key role in the pathophysiology of hypertension. Aldosterone increases blood pressure by inducing sodium reabsorption, vascular remodeling, and endothelial dysfunction, and possibly through other genomic and non‐genomic effects (eCPS 2014). The role of aldosterone itself has been established not only in hypertension secondary to hyperaldosteronism, but also in primary, or essential, hypertension. There is also some evidence that non‐hypertensive patients with high‐normal aldosterone levels are at increased risk for developing elevated blood pressure (Jansen 2009). Eplerenone is an aldosterone antagonist and competes with aldosterone for binding to the mineralocorticoid receptor. In the late distal and cortical collecting tubules of the kidney, this antagonism will lead to decreased activation of sodium channels and decreased numbers of sodium/potassium ATPase pumps. The net effect is diuresis: increased sodium and, as a consequence, water excretion (Katzung 2012).
Aldosterone antagonists may also have a role in treatment‐resistant hypertension. Non‐responders to typical antihypertensive therapies may benefit from a trial of an aldosterone antagonist (Weber 2014). Prolonged treatment with angiotensin‐converting enzyme (ACE) inhibitors or angiotensin receptor blockers (ARBs), which initially lower serum aldosterone levels, may result in a rebound increase in serum aldosterone to higher than pre‐treatment levels. This is referred to as 'aldosterone escape' or 'aldosterone breakthrough' and could be a key mechanism in resistance to ACE inhibitors and ARBs (Jansen 2009).
Aldosterone has also been implicated in blood pressure‐independent adverse events, including vascular inflammation and cardiac and perivascular fibrosis, which may promote end‐organ damage. Evidence suggests that patients with hyperaldosteronism may be at increased risk for stroke, non‐fatal myocardial infarction, atrial fibrillation, elevated left ventricular mass, and arterial wall stiffness versus matched controls with essential hypertension (Jansen 2009). Use of an aldosterone antagonist may thus reduce these pathologies in at‐risk patients.
Why it is important to do this review
Hypertension can be difficult to manage. Patients may be intolerant to antihypertensive medications, and many require multiple classes of medications to control their blood pressure. More importantly, lowering blood pressure below 140/90 mmHg is not a validated surrogate outcome for clinically significant endpoints such as mortality and morbidity. It is important to determine the clinical impact of all antihypertensive medications, including aldosterone antagonists, to support their continued use in essential hypertension. Previous systematic reviews have not evaluated the effect of eplerenone on cardiovascular morbidity, mortality, and magnitude of blood pressure lowering in people with hypertension (Wright 2009).
Objectives
To assess the effects of eplerenone monotherapy versus placebo for primary hypertension in adults. Outcomes of interest were all‐cause mortality, cardiovascular events (fatal or non‐fatal myocardial infarction), cerebrovascular events (fatal or non fatal strokes), adverse events or withdrawals due to adverse events, and systolic and diastolic blood pressure.
Methods
Criteria for considering studies for this review
Types of studies
All randomized controlled trials that compare oral eplerenone monotherapy to placebo.
Types of participants
Trials including participants of both sexes older than 18 years of age with primary hypertension defined by SBP greater than 140 mmHg, DBP greater than 90 mmHg, or both, with no known secondary cause for the high blood pressure. We included trials involving participants with and without co‐morbidities.
Types of interventions
The intervention of interest was oral eplerenone monotherapy (at any dose). The comparative intervention was placebo.
Types of outcome measures
Primary outcomes
-
All‐cause mortality
-
Cardiovascular mortality
-
Non‐cardiovascular mortality
-
Number of patients experiencing at least one serious adverse event
-
Fatal and non‐fatal, disabling stroke
-
Fatal and non‐fatal myocardial infarction
Secondary outcomes
-
Number of patients with at least one adverse event
-
Number of patients who withdrew due to adverse events
-
Change blood pressure
-
Change in systolic blood pressure
-
Change in diastolic blood pressure
-
Search methods for identification of studies
Electronic searches
We searched the following databases for primary studies.
-
Cochrane Hypertension Specialised Register (searched 3 March 2016).
-
Cochrane Register of Controlled Trials (CENTRAL; 2016, Issue 3) via the Cochrane Register of Studies Online (searched 3 March 2016).
-
MEDLINE Ovid (1946 to 3 March 2016).
-
Embase Ovid (1974 to 3 March 2016).
-
US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov; searched 3 May 2016).
The Cochrane Hypertension Specialised Register includes controlled trials from searches of CENTRAL, MEDLINE, Embase, CAB Abstracts, CINAHL, Food Science and Technology Abstracts (FSTA), Global Health, LILACS,ProQuest Dissertations & Theses, PsycINFO, the Web of Science, and the WHO International Clinical Trials Registry Platform (ICTRP).
We searched electronic databases using a strategy combining the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity‐maximizing version (2008 revision) with selected medical subject headings (MeSH) and free text words. We used no language restrictions and translated the MEDLINE search strategy into the other databases using the appropriate controlled vocabulary as applicable. We present full strategies in Appendix 1.
Searching other resources
We also tried to collect information from the following additional resources.
-
World Health Organization International Clinical Trials Registry Platform (apps.who.int/trialsearch).
-
Reference lists of all papers and relevant reviews identified.
-
Authors of relevant papers, regarding any further published or unpublished work.
-
Authors of trials reporting incomplete information, to obtain the missing information.
-
ISI Web of Science for papers that cite studies included in the review.
-
European Medicines Agency Database.
-
United States Food and Drug Administration (FDA) Database.
-
Manufacturer of eplerenone.
-
Database of Abstracts of Reviews of Effectiveness (DARE), for related reviews.
Data collection and analysis
Selection of studies
Three review authors (SM, TT, MW) independently assessed all identified studies to determine whether they met the predefined inclusion criteria. We reviewed all references identified in the search that mentioned the use of eplerenone and included them if they met the criteria. We documented reasons for excluding studies and resolved any differences regarding inclusion through discussion with a fourth independent review author.
Data extraction and management
Once we identified all studies, three independent review authors examined those that fulfilled the inclusion criteria in detail. We used a web‐based systematic review program, Covidence, to create a standardized data extraction form.
Assessment of risk of bias in included studies
We assessed the following parameters.
-
Method of random sequence generation.
-
Method of allocation concealment.
-
Blinding to treatment allocation (including healthcare provider, assessor, and patient).
-
Blinding of outcome assessment.
-
Incomplete outcome data.
-
Selective outcome reporting.
-
Funding of clinical trial.
We resolved any differences in interpretation of the data through consensus. If additional information was required, we contacted the original authors of the study and then reassessed the study once the missing information was available, where possible. Three review authors independently collected study characteristics and the outcome measures of interest using a pre‐formulated data extraction sheet on Covidence. We collected all data, regardless of compliance or completion of follow‐up, in order to allow for intention‐to‐treat analysis.
Measures of treatment effect
For evaluation of the primary outcomes (all‐cause mortality, cardiovascular events, and cerebrovascular events), we planned to record the total number of patients with at least one event within each trial, calculating proportions for these dichotomous outcomes and presenting comparisons between groups as relative risk ratios (RRs) with corresponding 95% confidence intervals (CIs).
We planned to combine data for blood pressure reduction using a weighted mean difference method. This combines a weight based on the number of individuals in the trial and the within‐study variance. If the trial did not report the within‐study variance for decrease in blood pressure, we imputed the standard deviation (SD) from the average SD from the other trials.
We first performed all analyses using a fixed‐effect model.
Unit of analysis issues
We used data from all patients individually randomized to each intervention in the analyses, taking care to identify situations in which studies had censored/excluded data and to differentiate data presented as the total number of events or the total number of patients with a first event. We contacted authors for clarification if necessary. We intended to calculate proportions for the first relevant event that occurred for each patient randomized to a particular intervention.
Dealing with missing data
In general if there were missing data, we contacted the authors of the study for clarification and documented this process. If we could not obtain the SD of the change in blood pressure, we imputed the value (using SD of the change data from other similar trials).
Assessment of heterogeneity
We assessed heterogeneity across the studies using the I2 statistic (defining important heterogeneity at a threshold of 30% to 60%) and the Chi2 statistic (with statistical significance set at P < 0.10). Where we did not detect heterogeneity, we used a fixed‐effect model. Otherwise, we used a random‐effects model to determine if the effects of eplerenone changed depending on the analysis model. More importantly, we planned to explore clinical and methodological sources of heterogeneity, considering characteristics like: baseline risk factors for the outcomes of interest, duration of studies, age, race, and sex distribution of participants across the studies. Based on the exploration of sources of heterogeneity, we made decisions about the appropriateness of meta‐analyzing data.
Assessment of reporting biases
In the event that we assumed that missing data represented a poor outcome, or where we imputed data, we planned to carry out sensitivity analyses to see if results were sensitive to the assumptions being made. However, since we did not find any clinical outcome data, this type of imputation was not necessary.
Data synthesis
We used Cochrane Review Manager 5 (RevMan 5) software , for all data analyses. We based quantitative analyses of outcomes on intention‐to‐treat results. We used RRs and the fixed‐effect or random‐effects model (depending on heterogeneity results) to combine outcomes across trials. We calculated absolute risk reduction (ARR) as risk difference × 100, and the numbers needed to treat for an additional beneficial outcome (NNTB) as 1/risk difference for all dichotomous outcomes. We pooled data for blood pressure reduction using a weighted mean difference method with standard deviations.
Subgroup analysis and investigation of heterogeneity
We planned to perform the following four subgroup analyses.
-
Trials of less than 6 months' duration, of 6 to 12 months' duration, and of more than 12 months' duration.
-
Effect of ethnic group on blood pressure and adverse events.
-
Effect of age on blood pressure and adverse events.
-
Effect of dose on blood pressure and adverse events.
Sensitivity analysis
We planned to carry out sensitivity analyses in order to test for the robustness of the results. We intended to analyze the following categories separately.
-
Trials without proper randomization compared to those with proper randomization.
-
Trials performed without proper allocation concealment versus those with proper allocation concealment.
-
Unblinded versus blinded trials.
-
Inclusion versus exclusion of data from trials where we imputed blood pressure standard deviations.
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies.
Results of the search
The search strategy identified 307 citations in CENTRAL, MEDLINE, and Embase. Following a review of their titles and abstracts, we excluded 229 citations that obviously did not meet our inclusion criteria and selected 78 studies for further review. We reviewed the full text of these 78 studies and excluded 73 that did not meet our inclusion criteria.
Of the six citations that met our inclusion criteria, one proved to be a duplicate publication, leaving five unique trials that met our inclusion criteria. See Figure 1.
Study flow diagram.
Included studies
Refer to Characteristics of included studies for further details regarding the studies.
All five included studies were parallel‐group, randomized controlled trials. Together, they involved a total of 1437 participants, who received treatment for 8 to 16 weeks. The daily doses of eplerenone ranged from 25 mg to 400 mg. Participants' average age was 54 years, and there were slightly more men than women (64%). The mean baseline blood pressure (BP) in the eplerenone group was 153/101 mmHg, compared to 152/100 mmHg in the placebo group.
Excluded studies
See: Characteristics of excluded studies.
We excluded 73 trials for the following reasons.
-
Not a randomized controlled trial (k = 32).
-
Did not study patients with essential hypertension (k = 16).
-
Treatment arm was not eplerenone monotherapy (k = 9).
-
Compared eplerenone with other antihypertensive agents (k = 7).
-
Did not study any of our primary or secondary outcomes of interest (k = 5).
-
Terminated study on clinicaltrials.gov; efforts to contact the responsible party were fruitless in obtaining usable information for the meta‐analysis (k = 1, NCT01373086).
-
Completed study on clinicaltrials.gov but with no published results; efforts to contact the responsible party were fruitless in obtaining usable information for the meta‐analysis (k = 1, NCT02345044).
-
Unpublished trial not providing enough information to be used in the analysis (k = 1, Trial 015 1999); found through the FDA medical reviews on the approval of eplerenone.
Risk of bias in included studies
We assessed the risk of bias on the basis of five major criteria: allocation concealment, blinding, completeness of outcome data addressed, presence of selective reporting, and other potential sources of bias (see Figure 2). Three review authors (SM, TT, MW) independently performed the 'Risk of bias' assessment on Covidence, resolving any differences in interpretation of the data through consensus with a fourth review author (AT). If additional information was required, we contacted the original author of the study and reassessed the study in light of the availability of missing information. We summarize our contact with the original authors of the five included studies in Figure 3.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Summary of data extraction and contact with corresponding authors
Allocation
Calhoun 2011 reported performing allocation by generating randomization sequences using an interactive voice‐response system provider. A validated system automated the random assignment of participant numbers to randomization numbers, linking them to the different treatment groups and the medication numbers on treatment packs. Study drugs were identical in packaging, labeling, schedule of administration, appearance, taste, and odor in order to conceal the nature of the treatment.
Flack 2003 and Saruta 2004 did not describe methods of allocation concealment; the trial reports noted that the order in which participants received eplerenone and placebo was randomized; however, authors did not describe the method for randomization or respond to our attempts to contact them.
Weinberger 2002 allocated all participants at the same time as randomization after the run‐in period by using a randomized computer‐generated schedule for sequence generation.
White 2003 stated that the study coordinator performed allocation concealment via an interactive voice response system. Through email contact, the primary author of White 2003 clarified details on the randomization of the participants, which investigators accomplished using standard multicenter block randomization techniques.
Blinding
Flack 2003, Saruta 2004, and Weinberger 2002 described blinding methods poorly, stating that they conducted double‐blind trials but without describing how blinding was maintained beyond use of a placebo pill. Thus, the information available for assessing the appropriateness of blinding in these studies was inadequate.
Only two of the five studies adequately described blinding of participants and outcome assessors (Calhoun 2011; White 2003). Calhoun 2011 stated that participants, investigators, outcome assessors, and data analysts remained blinded to treatment assignments. Moreover, study drugs were identical in packaging, labeling, schedule of administration, appearance, taste, and odor in order to conceal the nature of the treatment.
Additional correspondence with White 2003 served to clarify details on the blinding of the study participants and outcome assessors. White 2003 stated both study participants and site staff were blinded to treatment assignment. The study medication was in bottles of unidentified investigational product and was sent from a central location, so the outcome assessors had no way of knowing the randomization medication or code.
Incomplete outcome data
Only one of the five included studies adequately reported how they dealt with incomplete outcome data (Saruta 2004). In this case, Saruta 2004 excluded one participant in the eplerenone 50 mg group from the efficacy analysis due to withdrawal of consent. The results in a group of 49 participants are unlikely to be affected by missing data from 1 participant.
However, there was a high risk of bias in Flack 2003. Four participants in the placebo and 8 participants in the eplerenone group had no post baseline assessment and were not included in the efficacy analysis. In addition, 41% of participants withdrew from the placebo group, compared to 26% in the eplerenone group, and authors did not describe details of imputing missing values using the last observation carried forward method. The high rates of withdrawal in the placebo group may cause smaller mean changes in blood pressure and thus exaggerate the effects seen in the eplerenone group.
The remaining three included studies were at unclear risk of bias for incomplete outcome data reporting. In White 2003, there was insufficient reporting of incomplete outcome data. Other than treatment failure, the authors did not specify the other reasons for withdrawals in each treatment group and how they accounted for missing data in the primary outcomes. There was not enough information to assess the risk of bias due to incomplete outcome data.
In Weinberger 2002, eight participants were not included in the efficacy analyses. In addition, 39 participants did not complete the study. The authors of Weinberger 2002 did not specify how withdrawals were distributed across groups or how authors dealt with missing data.
In Calhoun 2011, the percentage of withdrawals from treatment groups ranged from 6.8% to 13.0%. The authors did not thoroughly document reasons for withdrawal but noted that they conducted all efficacy analyses with the full analysis set and imputed missing measurements by carrying forward the last available observation. However, there was not enough available information to assess the risk of bias from the missing data.
Selective reporting
In most of the included studies, there did not seem be selective reporting of outcomes (Calhoun 2011; Flack 2003; Saruta 2004).
Calhoun 2011 reported all outcomes specified in their protocol registered on ClinicalTrials.gov. Flack 2003 reported all primary study outcomes but failed to report some secondary endpoints and serious adverse events. Weinberger 2002 and White 2003 reported all outcomes specified in the Methods section. However, there was no protocol available to assess the a priori design.
It is worth noting that none of the included studies reported clinically meaningful outcomes (i.e. any of our primary outcomes) related to hypertension. Thus, there may be a high risk of selective reporting based on the omission of outcomes that should inform clinical practice and be collected in clinical trials involving human participants (e.g. all‐cause mortality, serious adverse events).
Other potential sources of bias
Four of the five included studies received industry funding. Pharmacia Corporation, which Pfizer purchased in 2002, financed Weinberger 2002, and Pfizer Inc funded Saruta 2004. Novartis funded Calhoun 2011, which academic authors and Novartis Parma designed together. Novartis was also responsible for collecting data from investigational sites to create the clinical database for the data analysis. As per the additional information provided by the corresponding author, Pharmacia Corporation sponsored White 2003.
Flack 2003 did not report the source of funding.
Effects of interventions
See: Summary of findings for the main comparison
See: summary of findings Table for the main comparison.
Primary outcomes
Unfortunately, none of the included studies reported results for the following clinical outcomes: all‐cause mortality, cardiovascular mortality, non‐cardiovascular mortality, serious adverse events, fatal and non‐fatal myocardial infarction, or fatal and non‐fatal stroke.
Secondary outcomes
Number of participants with at least one adverse event
There was no difference in the incidence of participants experiencing any adverse event in three trials(Flack 2003; Weinberger 2002; White 2003; Analysis 1.1; Figure 4). However, we could not stratify these results based on dose, and the potential harms associated with increased daily doses of eplerenone are uncertain. We did not specifically assess rates of hyperkalemia in this review but will do so in future updates.
Forest plot of comparison: 29 Eplerenone Monotherapy vs Placebo, outcome: Any Adverse Event.
Number of participants who withdrew due to adverse events
There was no difference in the incidence of adverse events leading to discontinuation in the eplerenone group compared to placebo (Flack 2003; Weinberger 2002; White 2003; Analysis 1.2).
Change in blood pressure
Four studies reported mean systolic and diastolic blood pressures from participants who were seated (Calhoun 2011; Saruta 2004; Weinberger 2002; White 2003). Flack 2003 measured participants blood pressure while standing/supine, which could have raised/lowered the reading with respect to the other studies. Ideally, all studies would have reported blood pressures measured in the same position.
White 2003 studied four different daily doses of eplerenone (25 mg, 50 mg, 100 mg, and 200 mg). We used the mean change from baseline in seated blood pressure in our analysis, and we extracted data from Figure 1 in the published study report. The authors did not specify whether the error bars presented in Figure 1 in the published study report were standard deviations or standard errors. Using graphing methods, we estimated the variances in Figure 1 based on the pictorial error bars. Comparing these variances with the other studies, we assumed they were standard errors, as the numbers were similar. We felt this to be an appropriate estimate, as authors reported changes from baseline in daytime and nighttime blood pressures in Table 2 in the published study report as standard errors with similar numbers. When inputting our data for analysis, we distributed the sample size of the placebo group equally among the four different treatment groups.
Weinberger 2002 studied three different daily doses of eplerenone (50 mg, 100 mg, and 400 mg). The authors reported the change from baseline in systolic and diastolic blood pressure in seated position in Figure 1 in the published study report. Using the same assumptions and methods described in White 2003, we estimated the variance numbers from the error bars in Figure 1 in the published study report and assumed them to be standard errors in our calculations. Again, we distributed the sample size of the placebo group equally among the three different treatment groups when inputting our data for analysis.
Saruta 2004 studied three different daily doses of eplerenone (50 mg, 100 mg, and 200 mg). The mean change in cuff seated systolic and diastolic blood pressure from baseline was extracted from Figure 1 in the published study report. We estimated the variance as we did in White 2003 and Weinberger 2002. We felt this to be an appropriate estimate, as adjusted mean change in pulse pressure, presented in Table 2 in the published study report, showed variance with similar numbers and were reported as standard errors. Similarly to White 2003 and Weinberger 2002, we distributed the sample size of the placebo group equally among the three different treatment groups when inputting our data for analysis.
Flack 2003 reported mean changes in systolic and diastolic blood pressure but did not specify the position or method of blood pressure monitoring. Flack 2003 studied eplerenone 50 mg/day against placebo. We used the data presented in Table 3 in the published study report of the mean changes in SBP and DBP for placebo and eplerenone for all participants. Calhoun 2011 reported mean changes in sitting systolic and diastolic blood pressures, comparing the daily dose of 100 mg against placebo. Using the data presented in Figure 3 in the published study report, we used graphing methods in the bar graph to estimate standard errors. We did not need to distribute the sample size of the placebo group in Flack 2003 and Calhoun 2011 because only one eplerenone treatment group was used in their analyses.
We performed meta‐analysis of the blood pressure lowering effects of eplerenone versus placebo for the five included studies (Calhoun 2011; Flack 2003; Saruta 2004; Weinberger 2002; White 2003). There were a total of 1437 participants receiving eplerenone or placebo in parallel group randomized controlled trials. The duration that participants were in these trials ranged from 8 to 16 weeks.
Change in systolic blood pressure
The analysis of the mean difference in SBP showed that eplerenone reduced SBP by 9.21 mmHg (95% CI −11.08 to −7.34; P < 0.001; Analysis 1.3; Figure 5). There appears to be moderate heterogeneity in the results (I2 = 58%).
Forest plot of comparison: 1 Eplerenone monotherapy vs placebo, outcome: 1.3 Systolic blood pressure.
Change in diastolic blood pressure
The analysis of the mean difference in DBP found that eplerenone reduced DBP by 4.18 mmHg (95% CI −5.03 to −3.33; P < 0.001; Analysis 1.4; Figure 6). There was no statistical heterogeneity in the results (I2 = 0%) when using a fixed‐effect model. We summarize the results in the summary of findings Table for the main comparison.
Forest plot of comparison: 1 Eplerenone monotherapy vs placebo, outcome: 1.4 Diastolic blood pressure.
Subgroup analyses
Because all five included studies were of short duration, there was insufficient evidence to perform a subgroup analysis of trials of less than 6 months' duration, 6 to 12 months' duration, and more than 12 months' duration. There was also no information provided on the blood pressure lowering effects of participants in different age groups, so a subgroup analysis was not possible. Only one study analyzed the effect of ethnicity on blood pressure (Flack 2003), so we did not perform a subgroup analysis on its influence as an effect modifier.
However, due to the variation in doses used in the included studies, we could stratify blood pressure data by dose in the meta‐analysis in order to examine the possibility of a dose response. For mean changes of both systolic and diastolic blood pressure, eplerenone 25 mg/day did not have a statistically significantly effect.
For changes in systolic blood pressure, the confidence intervals overlapped at eplerenone doses of 50 mg/day to 200 mg/day. However, eplerenone 400 mg/day decreased mean systolic blood pressure by 16.5 mmHg (95% CI −20.23 to −12.78; Analysis 1.3.5), which does not overlap with any other doses studied. This may lead us to believe eplerenone does exhibit a dose response at doses of 400 mg/day or higher. However, these data were only available from Weinberger 2002 as none of the other authors explored doses higher than 200 mg/day. Since only one study at high risk of bias investigated this dose, we cannot be confident of this effect until other studies replicate it.
The confidence intervals for change in DBP all overlapped at eplerenone doses of 50 mg/day to 400 mg/day. Thus, there is no apparent dose response. In other words, doses of 400 mg/day will not lower diastolic blood pressure more than doses of 50 mg/day (Analysis 2.2).
Interestingly, when we exclude the results from Weinberger 2002's analysis with 400 mg/day, there is no heterogeneity (I2 = 5%) with our results in changes in systolic blood pressure. In addition, when we exclude the 100 mg daily dose of Weinberger 2002, the I2 decreases to 0%. There may be some unknown factor that is contributing to the heterogeneity of the results we found with changes in systolic blood pressure that is inherent in Weinberger 2002 study. We could not identify any apparent reason for heterogeneity based on our analysis of risk of bias, characteristics of included participants, or specifics of interventions tested. When we changed the analysis from a fixed‐effect model to a random‐effects model to test the robustness of the finding, the point estimate and confidence interval for SBP reduction did not change substantially (−9.21 mmHg, 95% CI −11.08 to −7.34; Analysis 1.3).
We conducted additional dose response comparisons based on dose comparisons within trials (Analysis 2.1; Analysis 2.2). Readers should be cautious of drawing definitive conclusions from these comparisons, as at best, three trials made the same dose‐to‐dose comparisons and at worst, one trial compared a particular dose‐to‐dose comparison. With these qualifications in mind, we can make some general observations.
SBP direct dose comparisons
See Analysis 2.1.
-
It appears that 100 mg/day of eplerenone reduces SBP more than 50 mg/day eplerenone by 4.27 mmHg (95% CI −5.94 to −2.61), based on three trials with low statistical heterogeneity. These differences in SBP are within the variability of BP measurements and are unlikely to be clinically important (Musini 2009).
-
It is possible that 400 mg/day of eplerenone reduces SBP more than 50 mg/day or 100 mg/day of eplerenone based on one experiment at unclear risk of bias. These differences in SBP are within the variability of BP measurements and are unlikely to be clinically important (Musini 2009).
-
The difference in SBP reduction between 200 mg/day and 50 mg/day of eplerenone is unclear. One experiment demonstrated a significant difference while another experiment demonstrated no difference in SBP reductions with 200 mg/day versus 50 mg/day (high statistical heterogeneity).
-
Two experiments showed no statistical difference in SBP reduction between 200 mg/day and 100 mg/day of eplerenone (no statistical heterogeneity).
DBP direct dose comparisons
See Analysis 2.2.
-
It appears that 100 mg/day of eplerenone reduces DBP by 1.74 mmHg more than 50 mg/day eplerenone (Analysis 2.1), based on three trials with low statistical heterogeneity. These differences in SBP are within the variability of BP measurements and are unlikely to be clinically important (Musini 2009).
-
It is possible that 400 mg/day of eplerenone reduces DBP more than 50 mg/day or 100 mg/day of eplerenone based on one experiment with unclear risk of bias. These differences in SBP are within the variability of BP measurements and are unlikely to be clinically important (Musini 2009).
-
The difference in DBP reduction between 200 mg/day and 50 mg/day of eplerenone is unclear. One experiment demonstrated a numerical difference (not statistically significant) while another experiment demonstrated no difference in lowering SBP with 200 mg/day versus 50 mg/day (high statistical heterogeneity).
-
Two experiments showed no statistical difference in SBP reduction between 200 mg/day and 100 mg/day of eplerenone (no statistical heterogeneity).
-
One experiment showed no difference when comparing 50 mg/day, 100 mg/day, or 200 mg/day versus 25 mg/day.
Sensitivity analyses
In our sensitivity analysis, only including studies with proper randomization, proper allocation concealment, and proper blinding yielded similar results with respect to mean changes in systolic and diastolic blood pressure, although the number of trials was small, so there may be important differences in treatment effects based on risk of bias. Therefore, we do not present any of these results, as they did not contribute to a difference in our conclusions. We did not impute any of the blood pressure standard deviations, so a sensitivity analysis for this domain was not necessary.
Discussion
Summary of main results
There is insufficient evidence to draw any conclusions on the effects of eplerenone versus placebo for mortality, morbidity, or serious adverse events, as none of the included studies reported on any clinically meaningful outcomes.
Of the three studies that had adequate adverse event reporting, there were no differences in the incidence of participants experiencing any adverse event or adverse events leading to withdrawal in the eplerenone group compared to placebo (Figure 4). However, we could not analyze these results based on dose, and the potential harms related to increased daily doses of eplerenone are unknown.
Meta‐analysis of the five parallel‐group randomized controlled trials found a reduction in systolic blood pressure of 9.21 mmHg (95% CI −11.08 to −7.34; P < 0.001) and a reduction in diastolic blood pressure of 4.18 mmHg (95% CI −5.03 to −3.33; P < 0.001) (Figure 5; Figure 6). These results were statistically significant. There was no evidence of heterogeneity between the studies in changes in diastolic blood pressure. We found moderate heterogeneity between the studies in changes in systolic blood pressure. There may be a dose‐response effect with eplerenone at a dose of 400 mg/day; however, this finding is based on a single study (Weinberger 2002). The confidence intervals around the mean end‐of‐study blood pressure for doses ranging from 50 mg/day to 200 mg/day all overlapped. Thus, it appears that doses of more than 50 mg/day do not produce further reductions in systolic or diastolic blood pressure.
Overall completeness and applicability of evidence
While we attempted to contact authors of all five included studies, we only obtained further study information for White 2003. Contact with authors of White 2003 did not yield further information on clinical outcomes of mortality, morbidity, or serious adverse events. However, we were able to obtain further detail on blinding, allocation concealment, and randomization to better assess the risk of bias in the study. There was an underreporting of adverse effects in two of the studies (Calhoun 2011; Saruta 2004), which did not report the number of participants who withdrew due to adverse events or the number of participants with at least one adverse event.
It is interesting to note that the baseline mean BP was 153/101 mmHg in the intervention group and 152/100 mmHg in the placebo group across the included studies. This level of BP corresponds to the categorization of mild hypertension. A previous review (Diao 2014) suggested that there is no clear evidence that treatment of mild hypertension leads to reductions in the risk of morbidity and mortality.
Quality of the evidence
The effect sizes for reducing blood pressure that we calculated could be overestimates based on the unclear blinding used in three of the five included studies. Inappropriate blinding of outcome assessors, study participants, or both could exaggerate the effects of eplerenone on blood pressure by over‐reporting expected or desirable results. Only three included trials adequately reported on the number of participants who withdrew due to adverse events and the number of participants with at least one adverse event. Lastly, the reporting of adverse events was not stratified by dose. Therefore, we could not assess the harm due to eplerenone or based on its dose. This would be of particular interest to determine whether doses of eplerenone of 400 mg/day contributed to more significant harm.
Potential biases in the review process
Three independent review authors assessed the studies, and a fourth validated their judgement when any differences in interpretation arose. We undertook this process to screen 307 titles and abstracts as well as 78 full‐text articles, and to extract data from the five included studies for qualitative and quantitative analysis. None of the authors in this review have any conflicts of interest to declare.
Agreements and disagreements with other studies or reviews
The authors of this review are aware of one published review assessing the use of monotherapy eplerenone for treating essential hypertension (Pelliccia 2014). That review found statistically significant decreases in systolic blood pressure of 8.1 mmHg (95% CI −8.2 to −8.0) and in diastolic blood pressure of 4.1 mmHg (95% CI −4.1 to −4.0 mmHg). These results are very similar; however, Pelliccia 2014 included Krum 2002, which compared eplerenone plus ACE inhibitor or an angiotensin receptor blocker combination versus placebo plus ACE inhibitor or an angiotensin receptor blocker combination; this was not eplerenone monotherapy, so we excluded it from our review.
In a meta‐analysis of spironolactone versus placebo for hypertension, Batterink 2010 found that spironolactone lowered SBP by 20.09 mmHg and DBP by 6.75 mmHg. In a meta‐analysis of ACE inhibitors versus placebo, maximal blood pressure lowering for the ACE inhibitor class of drugs was −7.68 mmHg (95% CI ‐8.45, ‐6.91) for SBP and −4.59 mmHg (95% CI −4.99, −4.19) for DBP (Heran 2009). Another review found that beta‐blockers reduced systolic and diastolic blood pressure by approximately 11 mmHg and 6 mmHg, respectively, compared to placebo (Wiysonge 2017). Indirect evidence suggests that eplerenone lowers blood pressure to a lesser extent than spironolactone but to a similar extent as ACE inhibitors and beta‐blockers.
Study flow diagram.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Summary of data extraction and contact with corresponding authors
Forest plot of comparison: 29 Eplerenone Monotherapy vs Placebo, outcome: Any Adverse Event.
Forest plot of comparison: 1 Eplerenone monotherapy vs placebo, outcome: 1.3 Systolic blood pressure.
Forest plot of comparison: 1 Eplerenone monotherapy vs placebo, outcome: 1.4 Diastolic blood pressure.
Comparison 1 Eplerenone monotherapy vs placebo, Outcome 1 Any adverse event.
Comparison 1 Eplerenone monotherapy vs placebo, Outcome 2 Adverse event leading to withdrawal.
Comparison 1 Eplerenone monotherapy vs placebo, Outcome 3 Systolic blood pressure.
Comparison 1 Eplerenone monotherapy vs placebo, Outcome 4 Diastolic blood pressure.
Comparison 2 Eplerenone direct dose comparisons, Outcome 1 Systolic dose response.
Comparison 2 Eplerenone direct dose comparisons, Outcome 2 Diastolic dose response.
| Eplerenone compared with placebo for primary hypertension | |||
| Patient or population: adults with primary hypertension Settings: primary care Intervention: eplerenone (25‐400 mg) Comparison: placebo | |||
| Outcomes | Relative effect | No of participants | Quality of the evidence |
| Effect on systolic blood pressure (mean difference in systolic blood pressure) | (MD −9.21 mmHg, 95% CI −11.08 to −7.34) | 1437 | ⊕⊕⊕⊝ |
| Effect on diastolic blood pressure (mean difference in diastolic blood pressure) | (MD −4.18 mmHg, 95% CI −5.03 to −3.33) | 1437 | ⊕⊕⊕⊝ |
| Any adverse event | OR 1.07 (0.82 to 1.41) | 1105 (3) | ⊕⊕⊝⊝ |
| Adverse event leading to withdrawal | OR 1.10 (0.47 to 2.55) | 1105 (3) | ⊕⊕⊝⊝ |
| CI: confidence interval; OR: odds ratio. | |||
| GRADE Working Group grades of evidence | |||
| aDowngraded due to high risk of bias in many of the studies. | |||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Any adverse event Show forest plot | 3 | 1121 | Odds Ratio (M‐H, Fixed, 95% CI) | 1.07 [0.82, 1.41] |
| 2 Adverse event leading to withdrawal Show forest plot | 3 | 1132 | Odds Ratio (M‐H, Fixed, 95% CI) | 1.10 [0.47, 2.55] |
| 3 Systolic blood pressure Show forest plot | 5 | 1437 | Mean Difference (IV, Random, 95% CI) | ‐9.21 [‐11.08, ‐7.34] |
| 3.1 25 mg/day | 1 | 66 | Mean Difference (IV, Random, 95% CI) | ‐5.7 [‐11.89, 0.49] |
| 3.2 50 mg/day | 4 | 645 | Mean Difference (IV, Random, 95% CI) | ‐7.40 [‐9.59, ‐5.22] |
| 3.3 100 mg/day | 4 | 433 | Mean Difference (IV, Random, 95% CI) | ‐10.21 [‐12.37, ‐8.05] |
| 3.4 200 mg/day | 2 | 172 | Mean Difference (IV, Random, 95% CI) | ‐8.59 [‐11.57, ‐5.62] |
| 3.5 400 mg/day | 1 | 121 | Mean Difference (IV, Random, 95% CI) | ‐16.51 [‐20.23, ‐12.78] |
| 4 Diastolic blood pressure Show forest plot | 5 | 1437 | Mean Difference (IV, Fixed, 95% CI) | ‐4.18 [‐5.03, ‐3.33] |
| 4.1 25 mg/day | 1 | 66 | Mean Difference (IV, Fixed, 95% CI) | ‐2.0 [‐5.75, 1.75] |
| 4.2 50 mg/day | 4 | 645 | Mean Difference (IV, Fixed, 95% CI) | ‐3.73 [‐4.98, ‐2.48] |
| 4.3 100 mg/day | 4 | 433 | Mean Difference (IV, Fixed, 95% CI) | ‐4.64 [‐6.18, ‐3.10] |
| 4.4 200 mg/day | 2 | 172 | Mean Difference (IV, Fixed, 95% CI) | ‐4.13 [‐6.44, ‐1.83] |
| 4.5 400 mg/day | 1 | 121 | Mean Difference (IV, Fixed, 95% CI) | ‐7.69 [‐11.42, ‐3.97] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Systolic dose response Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 1.1 50 mg/day vs 25 mg/day | 1 | 128 | Mean Difference (IV, Fixed, 95% CI) | ‐1.0 [‐5.78, 3.78] |
| 1.2 100 mg/day vs 25 mg/day | 1 | 133 | Mean Difference (IV, Fixed, 95% CI) | ‐4.7 [‐9.72, 0.32] |
| 1.3 200 mg/day vs 25 mg/day | 1 | 132 | Mean Difference (IV, Fixed, 95% CI) | ‐3.10 [‐7.67, 1.47] |
| 1.4 100 mg/day vs 50 mg/day | 3 | 445 | Mean Difference (IV, Fixed, 95% CI) | ‐4.27 [‐5.94, ‐2.61] |
| 1.5 200 mg/day vs 50 mg/day | 2 | 234 | Mean Difference (IV, Fixed, 95% CI) | ‐5.33 [‐7.87, ‐2.78] |
| 1.6 400 mg/day vs 50 mg/day | 1 | 213 | Mean Difference (IV, Fixed, 95% CI) | ‐8.64 [‐10.63, ‐6.66] |
| 1.7 200 mg/day vs 100 mg/day | 2 | 269 | Mean Difference (IV, Fixed, 95% CI) | 0.12 [‐2.38, 2.63] |
| 1.8 400 mg/day vs 100 mg/day | 1 | 207 | Mean Difference (IV, Fixed, 95% CI) | ‐5.02 [‐7.00, ‐3.03] |
| 2 Diastolic dose response Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 2.1 50 mg/day vs 25 mg/day | 1 | 128 | Mean Difference (IV, Fixed, 95% CI) | ‐0.90 [‐3.19, 1.39] |
| 2.2 100 mg/day vs 25 mg/day | 1 | 133 | Mean Difference (IV, Fixed, 95% CI) | ‐2.60 [‐5.11, ‐0.09] |
| 2.3 200 mg/day vs 25 mg/day | 1 | 132 | Mean Difference (IV, Fixed, 95% CI) | ‐1.70 [‐3.99, 0.59] |
| 2.4 100 mg/day vs 50 mg/day | 3 | 477 | Mean Difference (IV, Fixed, 95% CI) | ‐1.74 [‐2.88, ‐0.61] |
| 2.5 200 mg/day vs 50 mg/day | 2 | 266 | Mean Difference (IV, Fixed, 95% CI) | ‐1.26 [‐2.66, 0.15] |
| 2.6 400 mg/day vs 50 mg/day | 1 | 213 | Mean Difference (IV, Fixed, 95% CI) | ‐4.34 [‐6.30, ‐2.39] |
| 2.7 200 mg/day vs 100 mg/day | 2 | 269 | Mean Difference (IV, Fixed, 95% CI) | 0.41 [‐1.20, 2.02] |
| 2.8 400 mg/day vs 100 mg/day | 1 | 207 | Mean Difference (IV, Fixed, 95% CI) | ‐2.61 [‐4.59, ‐0.63] |
