Is the newest angiotensin-receptor blocker azilsartan medoxomil more efficacious in lowering blood pressure than the older ones? A systematic review and network meta-analysis.

Angiotensin-receptor blockers are often considered insufficiently efficacious in reducing blood pressure. However, newer angiotensin-receptor blockers may be more effective than the older ones. A network meta-analysis was performed to compare the efficacy of various angiotensin-receptor blockers in reducing office and ambulatory blood pressure in hypertensive patients. Relevant literature was searched from English and Chinese databases for randomized controlled trials involving angiotensin-receptor blockers in hypertension. Efficacy variables included systolic and diastolic blood pressure either in the office or on ambulatory blood pressure monitoring. Absolute blood pressure reductions at 6-12 weeks of treatment and their credible intervals were reported. A total of 34 publications provided adequate data for analysis (n = 14 859). In 28 studies on office systolic blood pressure (n = 12 731), against the common comparator valsartan 80 mg, the differences in systolic blood pressure were in favor of azilsartan medoxomil (20-80 mg), irbesartan (300 mg), olmesartan (20-40 mg), telmisartan (80 mg), and valsartan (160-320 mg), but not candesartan (8-16 mg), losartan (50-100 mg), irbesartan (150 mg), olmesartan (10 mg), and telmisartan (40 mg). The ranking plot shows that azilsartan medoxomil 80 mg had a possibility of 99% being the best in the class. Similar results were observed for office diastolic blood pressure and from 13 studies for 24-hour ambulatory systolic and diastolic blood pressure. In conclusion, angiotensin-receptor blockers had different blood pressure lowering efficacy. The newest angiotensin-receptor blocker azilsartan medoxomil at the dose of 80 mg seemed to be most efficacious in reducing both systolic and diastolic blood pressure in the office and on ambulatory measurement.

INTRODUCTION

Angiotensin‐receptor blockers are recommended by the current hypertension guidelines as one of the firstline antihypertensive drug classes. Angiotensin‐receptor blockers lower blood pressures through inhibiting the actions of angiotensin II and exhibit good tolerability. Until recently, a number of agents in this class are available for the treatment of hypertension. However, since the first angiotensin‐receptor blocker, losartan, became available, this class of drugs has been unsatisfied for their less or low potency in blood pressure lowering. Some newer agents had been seen more efficacious than the older ones, but eventually found that the dosage might not be equivalent. The more potent blood pressure lowering action was with a problem of more side effects.

Recent technological advances have led to the discovery of a structurally and chemically even newer agent, azilsartan medoxomil. In clinical studies, this new angiotensin‐receptor blocker showed strong blood pressure lowering efficacy with similar tolerability and side effects. This new drug has been compared with several but not all available older agents. We therefore performed the present network meta‐analysis in attempt to have an overview on this class of antihypertensive drugs in patients with hypertension, with the focus on the efficacy of azilsartan medoxomil at various dosages.

METHODS

Search strategy and selection of studies

We followed the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses extension statement for network meta‐analysis. Medline, Cochrane Library, China National Knowledge Infrastructure, and WANFANG databases were searched for randomized controlled trials published over the period from January 1995 to September 2018. We then screened literature according to the following criteria: 1) hypertension; 2) studies included various angiotensin‐receptor blockers either as intervention or comparator; and 3) office or ambulatory blood pressure changes as outcome measure. Angiotensin‐receptor blockers, such as azilsartan medoxomil, candesartan, eprosartan, irbesartan, losartan, olmesartan, telmisartan, and valsartan, were also included as a term of literature search. Medical Subject Headings were searched on Medline and Cochrane Library. The search strategy included Medical Subject Headings terms and keywords related to “hypertension”, “blood pressure”, and “angiotensin II type 1 receptor blocker”. The details of the search strategy are listed and described in the online Supplementary Materials (Table S1). We assessed all relevant English and Chinese articles for eligibility.

Outcomes

The primary outcome was the absolute change in office systolic and diastolic blood pressures from baseline. If ambulatory blood pressure data were reported, it was considered as a secondary outcome. The network meta‐analysis included both primary and secondary outcomes at six and 12 weeks of follow‐up. The 8‐week blood pressure measurement was used, wherever available. In case that 8‐week measurement was not available, the blood pressure measurement closest in time was reported, for example, at 10 weeks of follow‐up. The follow‐up time had to be at least four weeks of assigned treatment (medication and dosage).

Data extraction and quality assessment

Two investigators independently screened the title and abstract in the search engines to identify potentially relevant studies. The extracted data included study characteristics, patient characteristics, interventions, outcomes, and other relevant findings. We also searched the references of the identified articles for possible inclusion. The extracted data were cross‐checked by a third investigator. Discrepancies were resolved on the basis of a consensus approach. The quality of the extracted studies was assessed using the GeMTC package of R (version 3.4.4). Quality was defined in four broad categories: high, moderate, low, and very low. The quality of evidence was assessed on the basis of five domains: risk of bias, inconsistency, indirectness, imprecision, and publication bias. The Cochrane Collaboration's risk of bias assessment tool was used to assess the risk of bias at the study level.

Data synthesis and statistical analysis

The effects of various angiotensin‐receptor blockers on the changes in systolic and diastolic blood pressures were estimated for all individual studies. The network meta‐analysis combined both direct and indirect treatment comparisons. Pairwise meta‐analyses of all interventions were performed to evaluate the possible statistical heterogeneity of studies within each comparison. Valsartan 80 mg was chosen as the common comparator, because valsartan was the mostly used angiotensin‐receptor blocker in China and many other countries, and 80 mg was the initial dose.

The network meta‐analysis was conducted using a Bayesian random‐effects model through a Markov Chain Monte Carlo process using the GeMTC package of R (version 3.4.4). , Various angiotensin‐receptor blockers were ranked by their effects relative to baseline when the Markov Chain Monte Carlo process was implemented. Hierarchy between various angiotensin‐receptor blockers was estimated by the ranking probabilities of the frequency table of iteration results. This was estimated using the surface under the cumulative ranking curves. The rankings for outcomes were then combined and summarized in a cluster ranking plot. Heterogeneity was assessed using both the Cochrane Q test and the I2 statistic. The following criteria of the I2 statistic was used to derive heterogeneity of studies: 25%‐50%, 50%‐75%, and greater than 75% as low, moderate, and high heterogeneity, respectively. The Cochrane collaboration's risk of bias assessment tool was used to asses risk of bias. A meta‐regression model was used to assess the effect of study‐level duration of treatment and baseline blood pressure on the treatment effect. The analysis was performed using the R statistical software, version 3.4.4. A two‐sided P‐value ≤ 0.05 was considered statistically significant.

RESULTS

Study selection

We identified 1,626 records by searching the databases. The majority of the identified articles were from Cochrane (n = 681, 41.9%) and Medline (n = 603, 37.1%, Figure 1). A total of 342 (21.0%) Chinese articles were identified from the China National Knowledge Infrastructure and WANFANG database. After removal of 693 (42.6%) duplicates and screening abstracts, 212 (13.0%) articles were assessed for eligibility. From these assessed articles, 178 (82.1%) were removed, because of inappropriate intervention (n = 75, 35.4%), outcomes not of interest (n = 41, 19.3%), incomplete data (n = 18, 8.5%), and so on (n = 40, 18.9%). Finally, a total of 34 studies were included in the present analysis. , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,

FIGURE 1

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) diagram of the systematic review search strategy

Characteristics of the selected studies

Table 1 summarizes the characteristics of the included studies. The majority of the studies had a double‐blind design (n = 28, 82.4%), and included two (n = 22, 64.7%) to three or more treatment arms (n = 12, 35.3%). Losartan was the most commonly studied treatment (n = 24, 70.6%), followed by olmesartan (n = 14, 41.2%), telmisartan (n = 11, 32.4%), valsartan (n = 11, 32.4%), and candesartan (n = 6, 17.6%). Azilsartan medoxomil and irbesartan represented the remaining study treatments (n = 5, 14.7%). The study duration ranged from six weeks to one year. Our analysis only included data between six and 12 weeks from baseline.

TABLE 1

Characteristics of the included studies

Author, year of publication	Country	Sample size	Trial design	Treatment	Study duration	Efficacy variable c
Kalikar et al, 2017[ 40 ]	India	60	O	Olmesartan 20 mg Telmisartan 40 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
Perez et al, 2017[ 41 ]	US, Mexico, Argentina & Peru	449	DB	AM 5 mg; 10 mg; 20 mg; 40 mg; 80 mg Olmesartan 20 mg Placebo	8 weeks	Mean change in SBP and DBP from baseline ABPM as primary end point
Flack et al, 2012[ 38 ]	US	941	PD	Olmesartan 20‐40 mg Losartan 50‐100 mg	8 weeks	Mean change in SBP and DBP from baseline
Punzi et al, 2012[ 39 ]	US	945	PD	Olmesartan 20 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
Bakris et al, 2011[ 35 ]	US, Peru, Argentina, & Mexico	1275	DB	AM 20 mg; 40 mg; 80 mg Olmesartan 40 mg	6 weeks	Mean change in SBP and DBP from baseline
Weir et al, 2011[ 36 ]	US	941	PD	Olmesartan 20 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
White et al, 2011[ 37 ]	US, Guatemala, Mexico Peru, & Puerto Rico	1291	DB	AM 20 mg; 40 mg Valsartan 160 mg Olmesartan 20 mg	6 weeks	Mean change in SBP and DBP from baseline
Chrysant et al, 2008[ 33 ]	US	1940	DB	Olmesartan 10 mg; 20 mg; 40 mg Amlodipine 5 mg; 10 mg Olmesartan/amlodipine 10/5 mg; 20/5 mg; 40/5 mg; 10/10 mg; 20/10 mg; 40/10 mg Placebo	8 weeks	Mean change in SBP and DBP from baseline
Fogari et al, 2008[ 34 ]	Italy	126	PO	Olmesartan 20 mg Telmisartan 80 mg	8 weeks	Mean change in SBP and DBP from baseline
Bahadir et al, 2007[ 30 ]	Turkey	42	RCT	Losartan 50 mg Telmisartan 80 mg	8 weeks	Mean change in SBP and DBP from baseline
Giles et al, 2007[ 31 ]	US	723	DB, titration	Olmesartan 20 mg Losartan 50 mg Valsartan 80 mg	8 weeks	Mean change in SBP and DBP from baseline
Phillip et al, 2007[ 32 ]	Study1: 6 countries Study2: 10 countries/ regions	1911	DB	Amlodipine 2.5 mg; 5 mg Valsartan 40 mg; 80 mg; 160 mg; 320 mg	8 weeks	Mean change in SBP and DBP from baseline
				Amlodipine 10 mg Valsartan 160 mg; 320 mg
Baguet et al, 2006[ 28 ]	France	256	DB	Candesartan 8 mg Losartan 50 mg Placebo	6 weeks	Mean change in SBP and DBP from baseline ABPM as primary end point
Zhu et al, 2006[ 29 ]	China	287	DB, active	Olmesartan 20 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
Destro et al, 2005[ 25 ]	Italy	114	PO	Olmesartan 20 mg Valsartan 160 mg	8 weeks	Mean change in SBP and DBP from baseline ABPM as primary end point
Liau et al, 2005[ 26 ]	Taiwan & China	126	DB, active	Olmesartan 20 mg Losartan 50 mg	12 weeks	Mean change in SBP and DBP from baseline
Smith et al, 2005[ 27 ]	US	588	DB, active	Olmesartan 20 mg Losartan 50 mg Valsartan 80 mg Irbesartan 150 mg	8 weeks	Mean change SBP and DBP from baseline ABPM as primary end point
Calvo et al, 2004[ 20 ]	Spain	70	PO	Telmisartan 80 mg Valsartan 160 mg	12 weeks	Mean change in SBP and DBP from baseline
Ding et al, 2004[ 21 ]	Taiwan & China	61	DB	Telmisartan 40 mg Losartan 50 mg	6 weeks	Mean change in SBP and DBP from baseline
Lee et al, 2004[ 22 ]	Taiwan & China	180	DB	Telmisartan 40 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
White et al, 2004[ 23 ]	US & Canada	490	DB, forced titration	Telmisartan 40 mg Valsartan 80 mg	8 weeks	Mean change SBP and DBP from baseline ABPM as primary end point
Bai et al, 2002[ 19 , 24 ]	China	330	DB	Telmisartan 40 mg Losartan 50 mg	12 weeks	Mean change in SBP and DBP from baseline
Bakris et al, 2002[ 18 ]	US	426	PO	Telmisartan 80 mg Valsartan 80 mg	8 weeks	Mean change in SBP and DBP from baseline
Bakris et al, 2001[ 15 ]	US	654	DB, forced titration	Candesartan 16 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
Elliot et al, 2001[ 16 ]	US	495	DB	Losartan 50 mg Valsartan 80 mg	12 weeks	Mean change in SBP and DBP from baseline
Vidt et al, 2001[ 17 ]	US	611	DB, forced titration	Candesartan 16 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
Monterroso et al, 2000[ 14 ]	9 countries in Africa, Europe, & Latin America	187	DB	Losartan 50 mg Valsartan 80 mg	6 weeks	Mean change in SBP and DBP from baseline
Gradman et al, 1999[ 10 ]	US	332	DB	Candesartan 16 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
Hedner et al, 1999[ 11 ]	Sweden	1369	DB	Valsartan 80 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline
Lacourcière et al, 1999[ 12 ]	France	231	DB	Candesartan 8 mg Losartan 50 mg	8 weeks	Mean change in SBP and DBP from baseline ABPM as primary end point
Mallion et al, 1999[ 13 ]	France	223	DB	Placebo Losartan 50 mg Telmisartan 40 mg; 80 mg	6 weeks	Mean change in SBP and DBP from baseline
Andersson et al, 1998[ 8 ]	Sweden & Denmark	337	DB	Candesartan 8 mg; 16 mg Losartan 50 mg Placebo	8 weeks	Mean change in SBP and DBP from baseline
Kassler‐Taub et al, 1998[ 9 ]	US, Canada, Mexico, Argentina & Brazil	567	DB	Placebo Losartan 100 mg Irbesartan 150 mg; 300 mg	8 weeks	Mean change in SBP and DBP from baseline

Abbreviations: ABPM, ambulatory blood pressure monitoring; AM, azilsartan medoxomil; DB, double‐blinded placebo/actively controlled, O, open‐label; DBP, diastolic blood pressure; PD, prospective double‐blinded; PO, prospective open‐label blinded end point; RCT, randomized controlled trial; SBP, systolic blood pressure.

The studies are listed in the descending order of the publication year.

Belgium, Canada, France, Germany, Mexico, and the United States.

Egypt, France, Germany, Korea, Malaysia, Norway, Peru, Portugal, Spain, and Taiwan.

Clinic blood pressure in the sitting position, unless indicated otherwise.

Heterogeneity was observed between studies in patient population, age, gender, and body mass index (P ≤ .05). The 14,859 patients from 34 studies had a mean age of 45.7 to 60.1 years (Table 2), and included more men than women except for the study of Bahadir et al. Mean (±SD) values of systolic blood pressure at baseline ranged from 140.0 mm Hg (±12.7) to 170.5 mm Hg (±12.6) and diastolic blood pressure from 86.0 mm Hg (±10.1) to 104 mm Hg (±5.0).

TABLE 2

Baseline characteristics of the study participants

Author, year	Sample analyzed	Treatment	Demographics				Baseline Office Blood Pressure, mmHg
			Mean age, years	Male, %	BMI	Diabetes mellitus	Definition of hypertension a	SBP (SD)	DBP (SD)
Kalikar et al, 2017	57	Olmesartan 20 mg Telmisartan 40 mg Losartan 50 mg	46.2 48.26 49.94	65 63.2 66.7	—	—	SBP 140‐159 or DBP 90‐99	148.3 (6.07) 148.8 (6.02) 149.9 (3.84	95.05 (2.15) 94.53 (2.65) 93.44 (2.45)
Perez et al, 2017	183	AM 20 mg AM 40 mg AM 80 mg Olmesartan 20 mg	54.6 55.3 53.5 53.4	53 47 56 46	31.2 31.5 31.4 29.3	—	DBP 95‐114	140.3 (12.7) 141 (12.6) 141 (16.5) 141.3 (13.8)	86.6 (11.4) 86 (10.1) 86.5 (9.6) 88.4 (10.5)
Flack et al, 2012	922	Olmesartan 20‐40 mg (Stage 1‐2 hypertension)	50.1 52.4	47.5 56.6	33.2 32.2	12.8% 8.6%	SBP 140‐159 or DBP 90‐99 & SBP ≥ 160 or DBP ≥ 100	149.5 (6.7) 162 (9.4)	97.9 (1.3) 102.5 (4.0)
		Losartan 50‐100 mg (Stage 1‐2 hypertension)	51.4 52.4	55.2 54.9	31.5 32.6	11.2% 12.5%		149.7 (7.1) 161.8 (9.3)	97.6 (1.3) 102.7 (4.0)
Punzi et al, 2012	934	Olmesartan 20 mg (Naïve, non‐naïve)	48.6 52.5	64.2 51.1	32.8 32.4	2.1% 11.9% (Metabolic syndrome 37.9% vs 41.1%)	DBP 95‐114 or SBP < 180	157.4 (10.89) 158.4 (10.22)	101.8 (4.26) 100.9 (3.96)
		Losartan 50 mg (Naïve, non‐naïve)	49.3 52.8	56 54.8	31.8 32.4	3.3% 14.3% (Metabolic Syndrome 35.2% vs 39.7%)		156.3 (10.75) 158.8 (10.06)	101.1 (3.94) 101.3 (4.19)
Bakris et al, 2011	1109	AM 20 mg AM 40 mg AM 80 mg Olmesartan 40 mg	57.1 57.4 58.1 58.9	47 50.2 52.3 49.6	30.4 30.6 30 29.8	—	SBP 150‐179 & 24‐h SBP 130‐169	—	—
Weir et al, 2011	934	Olmesartan 40 mg Losartan 100 mg	51.7 52.1	53.8 55	32.5 32.3	9.9%, 12.2%	DBP 95‐114 & SBP < 180	158.2 (10.4) 158.3 (10.2)	101.1 (4.0) 101.3 (4.1)
White et al, 2011	1093	AM 40 mg AM 80 mg Valsartan 320 mg Olmesartan 40 mg	57 56 55 56	53 53 54 55	31.7 30.7 31.1 31.1	—	SBP 150‐179 & 24‐h SBP 130‐169	157 (13) 158 (12) 157 (13) 158 (13)	93 (11) 92 (11) 93 (10) 92 (9)
Chrysant et al, 2008	484	Olmesartan 10 mg Olmesartan 20 mg Olmesartan 40 mg	53.8 53.6 53.9	54 55.9 50.6	33.5 32.9 33.6	—	Not specified	162.9 (16.7) 164.1 (16.5) 162.8 (15.7)	101.8 (5.9) 101.5 (4.6) 101.2 (5.1)
Fogari et al, 2008	126	Olmesartan 20 mg Telmisartan 80 mg	60.1 59.9	54 55.6	25.6 25.4	—	DBP ≥ 90	170.5 (12.6) 170.1 (12.1)	104.1 (7.5) 103.7 (7.1)
Bahadir et al, 2007	42	Losartan 50 mg Telmisartan 80 mg	47.7 52.3	33.3 28.6	32.8 31.5	—	SBP 140‐169 or DBP 90‐109	144.3 (6) 149.1 (7.7)	94.8 (5.1) 94.8 (7.5)
Giles et al, 2007	596	Olmesartan 40 mg Losartan 100 mg Valsartan 160 mg	52.2 51.3 52.2	62.8 60.5 66	—	—	DBP 100‐114	155.4 (11.2) 155 (11.5) 154.3 (10.6)	103.5 (3.1) 103.6 (2.8) 163.9 (11)
Phillip et al, 2007	922	Study 1: Valsartan 320 mg Valsartan 160 mg Valsartan 80 mg Valsartan 40 mg	56.8 53 53.1 55	52.3 53.9 45.2 56.7	—	—	DBP 90‐109 at week −4 to −2 & DBP 95‐109 at randomization	154.6 (11.41) 152 (14.19) 153.2 (11.63) 153.7 (12.56)	99.3 (3.59) 98.9 (3.54) 99.2 (3.55) 99.2 (3.22)
		Study 2: Valsartan 320 mg Valsartan 160 mg	56.7 56.8	51.9 44.4	—	—		157.5 (11.5) 155.6 (11.3)	99.1 (3.6) 98.9 (3.3)
Baguet et al, 2006	176	Candesartan 8 mg Losartan 50 mg	54 54	60.9 55.1	—	—	DBP 95‐114	140 (14) 140 (16)	91 (10) 89 (9)
Zhu et al, 2006	253	Olmesartan 40 mg Losartan 100 mg	53.29 54.78	64.3 65.7	—	—	DBP 95‐109 or SBP < 180	149.43 (12.56) 148.75 (11.97) 146 (5.7)	99.21 (3.28) 99 (3.21) 90.8 (4.3)
Destro et al, 2005	107	Olmesartan 20 mg Valsartan 160 mg	‐	56.1	—	—	DBP 95‐109	146 (5.7) 146.4 (5.3)	90.8 (4.3) 90.7 (3.9)
Liau et al, 2005	106	Olmesartan 20 mg Losartan 50 mg	48.5 48.1	55.1 66.7	—	—	DBP 95 to 114	148.4 (11.6) 149.3 (12.4)	102.4 (4.3) 103.2 (4.3)
Smith et al, 2005	534	Olmesartan 20 mg Losartan 50 mg Valsartan 80 mg Irbesartan 150 mg	52.3 52 51.9 52.1	65.4 64.2 57.7 56.7	—	—	Daytime DBP 90‐120	152.1 152.3 152.2 151.6	94.3 95.2 94.7 94.8
Calvo et al, 2004	70	Telmisartan 80 mg Valsartan 160 mg	45.7 49.2	55.9 36.1	26.8 28	—	SBP 140‐179 or DBP 90‐109	151.5 (16.4) 156.5 (14.9)	89.3 (12.1) 91.9 (10.5)
Ding et al, 2004	78	Telmisartan 40 mg Losartan 50 mg	50.5 52.7	54.8 66.7	‐	—	DBP 95‐114 or SBP 140‐199	153.6 (11.5) 154.5 (12.2)	101.1 (5.5) 99.1 (4.8)
Lee et al, 2004	176	Telmisartan 80 mg Losartan 100 mg	52.1 54.3	54.6 68.9	25.8 26.2	—	Morning DBP 95‐114 or SBP 140‐209	153.5 (11.8) 155.1 (11.8	100.5 (4.8) 101.8 (5.1)
White et al, 2004	490	Telmisartan 80 mg Valsartan 160 mg	54 55	77 75	29 30	—	DBP 95‐109	150 (12) 149 (12)	94 (6) 93 (6)
Bai et al, 2002	330	Telmisartan 80 mg Losartan 100 mg	50.8 50.8	73.8 74.5	—	—	DBP 95‐109	148.9 (12.8) 150.6 (12.4)	99.3 (3.9) 100.2 (3.9)
Bakris et al, 2002	426	Telmisartan 80 mg Valsartan 80 mg	53.6 53.1	68.2 67.9	—	—	DBP 95‐114 or SBP 140‐199 & 24‐h SBP/DBP ≥ 130/85	157 (12.9) 157.1 (12.4)	100.3 (4.9) 101.2 (5.1)
Bakris et al, 2001	654	Candesartan 32 mg Losartan 100 mg	54.2 54.1	57.8 58.4	—	—	DBP 95‐114	152.6 (12.3) 152 (12.6)	100.1 (3.9) 99.9 (4.2)
Elliot et al, 2001	486	Losartan 50 mg Valsartan 80 mg	53 54	60 59	—	—	DBP 95‐114	‐	‐
Vidt et al, 2001	611	Candesartan 32 mg Losartan 100 mg	55.5 55.1	58.3 58.9	—	—	DBP 95‐114	153.6 (11.7) 151.5 (11.7)	100.4 (4.3) 97.8 (6.1)
Monterroso et al, 2000	187	Losartan 50 mg Valsartan 80 mg	54.6 54.1	58.1 48.9	—	—	DBP 95‐114	‐	100.8 (5.6) 100.8 (6.6)
Gradman et al, 1999	322	Candesartan 32 mg Losartan 100 mg	54 57	57 58	—	—	DBP 95‐114	152.9 154.1	100.3 100.5
Hedner et al, 1999	1096	Valsartan 160 mg Losartan 100 mg	55.7 54.9	56.8 56.7	—	—	DBP 95‐114	157 (16.3) 157.4 (15.9)	101.4 (4.6) 101.6 (5.1)
Lacourciere et al, 1999	215	Candesartan 16 mg Losartan 100 mg	55 55	62.1 62.6	—	—	DBP 95‐114 or SBP < 200 & 24‐h DBP ≥ 85	155.1 153	101.8 100.2
Mallion et al, 1999	167	Losartan 50 mg Telmisartan 40 mg Telmisartan 80 mg	56 58 57	58 67 65	29.1 28.5 29.3	—	DBP 95‐114 or SBP 140‐ 199 & 24‐h DBP > 85	162.4 (16.3) 161.9 (14.7) 164.2 (15.3)	100.7 (4.5) 100.8 (4.2) 101.8 (4.9)
Andersson et al, 1998	249	Candesartan 8 mg Candesartan 16 mg Losartan 50 mg	60 59 59	57 67 57	—	—	DBP 95‐114	169 (14) 168 (15) 168 (16)	102 (5) 103 (5) 104 (5)
Kassler‐Taub et al, 1998	394	Losartan 100 mg Irbesartan 150 mg Irbesartan 300 mg	55 53.1 55.6	50 54 57	—	—	DBP 95‐109	153.3 (15.5) 155.3 (16.2) 155.4 (16)	100.6 (4.4) 101.1 (4.6) 100.4 (4.5)

Abbreviations: AM, azilsartan medoxomil; DBP, diastolic blood pressure; SBP, systolic blood pressure; SD, standard deviation.

The studies are listed in the descending order of the publication year.

Clinic blood pressure in the sitting position, unless indicated otherwise.

Network geometry

Figure 2 shows the network of treatment comparisons between various angiotensin‐receptor blockers for systolic and diastolic blood pressure. For systolic blood pressure, 19 angiotensin‐receptor blockers and dosages were studied with 34 out of 171 potential pairwise direct comparisons. For diastolic blood pressure, 18 were studied with 31 out of 153 pairwise comparisons. These studied angiotensin‐receptor blockers and dosages were azilsartan medoxomil (20 mg, 40 mg and 80 mg), candesartan (8 mg, 16 mg and 32 mg), irbesartan (150 mg and 300 mg), losartan (50 mg and 100 mg), olmesartan (10 mg, 20 mg and 40 mg), telmisartan (40 mg and 80 mg), and valsartan (40 mg, 80 mg, 160 mg, and 320 mg).

FIGURE 2

Network of direct comparisons between various angiotensin‐receptor blockers for office (upper panels) and ambulatory (lower panels) systolic (left panels) and diastolic (right panels) blood pressure. Abbreviations: AZL‐M 20, AZL‐M 40, and AZL‐M 80 indicate azilsartan medoxomil 20, 40, and 80 mg daily, respectively; CAN 8, CAN 16, and CAN 32, candesartan 8, 16, and 32 mg daily, respectively; IRB 150 and IRB 300, irbesartan150 and 300 mg daily, respectively; LOS 50 and LOS 100, losartan 50 and 100 mg daily, respectively; OLM 10, OLM 20, and OLM 40, olmesartan 10, 20, and 40 mg daily, respectively; TEL 40 and TEL 80, telmisartan 40 and 80 mg daily, respectively; VAL 40, VAL 160, and VAL 320, valsartan 40, 160, and 320 mg daily, respectively

Risk of bias assessment

Table S2 shows the risk of bias assessment for the included studies. All 34 studies were either of uncertain or low risk of selection bias. The risk of bias was unclear for random sequence generation in 20 (58.8%) studies and unclear for allocation concealment in 24 (70.6%) studies (Figure S1). The risk of bias was low with blinding of outcome assessment in 31 (91.2%) studies; it was low with incomplete outcome data, selective reporting and other sources in all 34 studies. The risk of bias was high with blinding of participants in five (14.7%) studies, and with blinding of outcome assessment in two (5.9%) studies.

Efficacy evaluations

Figure 3A shows the differences in office systolic blood pressure reductions on various angiotensin‐receptor blockers from valsartan 80 mg, being in favor of azilsartan medoxomil (20 mg, 40 mg, and 80 mg), irbesartan 300 mg, olmesartan 20 mg and 40 mg, telmisartan 80 mg, and valsartan 160 mg and 320 mg. The ranking plots shows that azilsartan medoxomil 80 mg had a 99% probability of being the best in class for systolic blood pressure reduction (Figure S2A), followed by azilsartan medoxomil 40 mg (90%) and irbesartan 300 mg (85%). There was no difference between direct and indirect pairwise comparisons of mean office systolic blood pressure (Figure S3). There was no heterogeneity (Figure S5) except for the comparisons of telmisartan 80 mg with valsartan 80 mg (I2 = 53.2%) and 160 mg (I2 = 80.2%). Similar results were observed for diastolic blood pressure (Figure 3B, Tables S3 and Figures S2B and S6).

FIGURE 3

Absolute mean (95% credible interval, CRI) difference in office systolic (left panel) and diastolic (right panel) blood pressure for various angiotensin‐receptor blockers in comparison with valsartan 80 mg daily. For explanations on the abbreviations of drugs, see the legend to Figure 2

Similar results were also observed for 24‐hour ambulatory blood pressure (Figure 4 and Table S4). The ranking plots show that azilsartan medoxomil 80 mg had a 93% and 86% possibility of being the best in class for 24‐hour systolic and diastolic blood pressure reductions, respectively. Further adjustment for the duration of treatment and baseline blood pressure values did not materially change the results for either clinic or ambulatory blood pressure (data not shown).

FIGURE 4

Absolute mean (95% credible interval, CRI) difference in ambulatory systolic (left panel) and diastolic (right panel) blood pressure for various angiotensin‐receptor blockers in comparison with valsartan 80 mg daily. For explanations on the abbreviations of drugs, see the legend to Figure 2

DISCUSSION

The findings of our meta‐analysis was that azilsartan medoxomil was superior to the other angiotensin‐receptor blockers in lowering both office and ambulatory blood pressures. The mean difference in office systolic/diastolic blood pressure in comparison with valsartan 80 mg was −9.9/‐7.2 mmHg and −8.2/‐6.0 mmHg for azilsartan medoxomil 80 mg and 40 mg, respectively. The corresponding mean values for ambulatory systolic/diastolic blood pressure were −9.2/‐5.4 mmHg and −8.8/‐5.3 mmHg, respectively. Such differences are appreciable in the control of hypertension and in the prevention of cardiovascular events.

Our finding is in keeping with the results of three previously published meta‐analyses on the basis of randomized comparisons. , , In an analysis that included seven studies with four angiotensin‐receptor blockers, azilsartan medoxomil (including azilsartan) reduced systolic and diastolic blood pressure more than the other three angiotensin‐receptor blockers by an overall difference of −3.7 mmHg (95% confidence interval [CI]: −5.7, −1.7) and −2.9 mmHg (95% CI: −3.8, −1.9), respectively. In another analysis that also included four angiotensin‐receptor blockers (candesartan, irbesartan, losartan, and valsartan) but without azilsartan medoxomil, the four angiotensin‐receptor blockers had similar blood pressure lowering efficacy, with a maximum between‐drug difference of 1.4/0.7 mmHg in office systolic/diastolic blood pressure. In the third analysis that included 31 studies with six angiotensin‐receptor blockers (candesartan, irbesartan, losartan, olmesartan, telmisartan and valsartan), valsartan 160 or 320 mg per day was more effective in lowering blood pressure than losartan 100 mg per day and as effective as the other angiotensin‐receptor blockers.

The finding of our analysis is also consistent with the results of individual trials that involved azilsartan medoxomil (prodrug, 40‐80 mg approved in most countries) or azilsartan (active drug, 20‐40 mg approved in Japan). In all these trials, azilsartan medoxomil , or azilsartan showed significantly greater blood pressure lowering efficacy, regardless of the comparator angiotensin‐receptor blocker or the blood pressure measurement technique, whether clinic or ambulatory blood pressure. Indeed, in the study with the largest sample size (n = 1291), azilsartan medoxomil 80 mg per day was superior to valsartan 320 mg per day and olmesartan 40 mg. The corresponding placebo‐adjusted mean changes in 24‐hour systolic blood pressure were −14.3 mmHg, −10.0 mmHg (P < .001 vs. azilsartan medoxomil), and −11.7 mmHg (P = .009) from baseline, respectively. For office systolic blood pressure, azilsartan medoxomil at both lower (40 mg) and higher dosages (80 mg) were superior to the two comparator angiotensin‐receptor blockers.

Although the mechanisms for the between‐drug differences in blood pressure lowering effects within the same mode of action are not entirely understood, there are several possible explanations. A key mechanism that makes the difference between angiotensin‐receptor blockers in blood pressure lowering is whether the angiotensin type II receptor is surmountable or insurmountable. Insurmountable antagonist forms tight sartan‐receptor complexes, and makes the dissociation half‐life longer. Such a mechanism increases the duration of action and is potentially beneficial in cardiovascular protection and prevention. In a recent follow‐up study in patients with myocardial infarction and treatment of various angiotensin‐receptor blockers, the one‐year risk of major cardiovascular events (14.3% vs. 11.2%, P =.025) and mortality (13.3% vs. 11.4%, P =.031) was significantly higher in patients on surmountable than insurmountable angiotensin‐receptor blockers.

Our study has to be interpreted within the context of its limitations, First, network meta‐analysis disregards randomization of each individual trial. The results of the analysis may be confounded by the heterogeneity of the study participants at baseline, especially blood pressure and comorbid diseases. Second, our analysis included short‐term studies only. Whether such a short‐term difference would remain in long‐term studies, especially when combination therapy is initiated, requires further investigation. Third, our study did not evaluate safety profile. The choice of angiotensin‐receptor blockers could be influenced by tolerability and side effects more than the other classes of antihypertensive drugs. However, angiotensin‐receptor blockers in general have the best treatment persistence, because of their better tolerability and less side effects.

In conclusion, our study showed that angiotensin‐receptor blockers had different blood pressure lowering efficacy, with the newest one, azilsartan medoxomil, being most efficacious in reducing both office and ambulatory blood pressures, at the least during short‐term treatment (<12 weeks). With the increasing availability, azilsartan medoxomil might improve blood pressure control. Nonetheless, more evidence on this new angiotensin‐receptor blocker is still needed.

CONFLICT OF INTEREST

Ji‐Guang Wang reports having received lecture and consulting fees form Takeda and from Astra‐Zeneca, Novarits, Omron, and Servier. Tzung‐Dau Wang has received honoraria for lectures from AstraZeneca, Boehringer Ingelheim, Daiichi Sankyo, Novartis, Omron, Pfizer, and Sanofi. The other authors declared no conflict of interest.

AUTHOR CONTRIBUTIONS

Ji‐Guang Wang contributed to the conception of the study. Miao Zhang did literature search, acquired the data, performed statistical analyses, and together with Ji‐Guang Wang prepared the first draft of the manuscript. All authors critically commented and revised the manuscript and gave the final approval.

Supplementary Material

Click here for additional data file.