ONTARGET
ALTITUDE
VA NEPHRON-D
N
25,620
8,561
1,448
Study groups
Ramipril 10 mg, Telmisartan 80 mg, combination of both
Aliskiren 300 mg versus placebo on top of ACEI or ARB
Lisinopril 10–40 mg versus placebo on top of losartan 100 mg
Primary composite end-point
Death from cardiovascular causes, myocardial infarction, stroke, or hospitalization for HF
Cardiovascular death or first occurrence of cardiac arrest with resuscitation; myocardial infarction; stroke; unplanned hospitalization for HF; ESRD, death attributable to kidney failure, or need for RRT with no dialysis or transplantation available or initiated; or doubling of serum creatinine
First occurrence of a change in the estimated GFR (a decline of ≥30 ml/min/1.73 m2 if the initial estimated GFR was ≥60 ml/min/1.73 m2 or a decline of ≥50 % if the initial estimated GFR was <60 ml/min/1.73 m2), ESRD, or death
Secondary renal composite end-point
Any dialysis, doubling of serum creatinine, or death
ESRD, death attributable to renal failure, or the need for RRT with no dialysis or transplantation available or initiated; or doubling of serum creatinine
First occurrence of a decline in the estimated GFR or ESRD
Follow-up (years)
4.7
2.8
2.2
Population characteristics
Age (years)
66.4
64.5
64.6
Gender (Male/Female, %)
73/27
68/32
99/1
BMI (kg/m2)
28.1
29.9
34.6
Hypertension (%)
69
94.5
100
Diabetes (%)
37
100
100
Mean baseline BP (mmHg)
142/82
137/74
137/73
Baseline estimated GFR (ml/min/1.73 m2)
73.6
57
53.7
Estimated GFR < 60 ml/min/1.73 m2 (%)
24
67.5
62
Urinary albumin excretion
Normo-albuminuria (%)
82.9
14.5
0
Micro-albuminuria (%)
13.1
25.6
0
Macro-albuminuria (%)
4
58.4
100
Main findings
Primary composite outcome: telmisartan was not inferior to ramipril; combination treatment was not superior to ramipril, and was associated with higher incidence of hypotension, hyperkalemia, and syncope. Renal outcomes: telmisartan had similar effects with ramipril; combination treatment increased by 2.2-fold the risk of dialysis for acute kidney failure compared with ramipril
Terminated early because of hypotension, hyperkalemia, and acute kidney failure in the aliskiren group
Terminated early due to elevated risk of hyperkalemia and acute kidney failure in the combination group
Cardiovascular composite: 11 % increase with aliskiren (HR 1.11; CI: 0.99–1.25, P = 0.09), driven by 2.4-fold higher risk of resuscitated cardiac arrest
Primary composite renal endpoint: no difference between the combination and mono-therapy groups (HR: 0.88; 95 % CI: 0.70–1.12, P = 0.30)
Renal composite: no difference between groups (HR 1.03; 95 % CI: 0.87–1.23, P = 0.74)
Another tool in our therapeutic armamentarium that was suggested to offer additional benefits towards reno-protection is the blockage of RAAS with the recently introduced direct renin inhibitor aliskiren. The effects of aliskiren on proteinuria were evaluated in the Aliskiren in the Evaluation of Proteinuria in Diabetes (AVOID) study, in which 599 patients with type 2 DM, hypertension and macro-albuminuria were randomly allocated to aliskiren (150 mg force-titrated to 300 mg daily) or placebo on top of losartan 100 mg and optimal BP-lowering therapy. After a mean follow-up of 6 months, addition of aliskiren to losartan along with standard antihypertensive therapy was associated with a 20 % higher decrease in ACR relative to placebo [53]; of note, these anti-proteinuric properties of aliskiren were shown to be independent from BP reduction [54].
On the basis of this beneficial impact of aliskiren on proteinuria in the AVOID study, the subsequent Aliskiren Trial in Type 2 Diabetes Using Cardiorenal Endpoints (ALTITUDE) trial evaluated the effect of adding aliskiren 300 mg daily relative to placebo on top of standard therapy with ACEI or ARB on a primary composite cardio-renal endpoint consisting of the time to cardiovascular death or a first occurrence of cardiac arrest with resuscitation, non-fatal myocardial infarction, nonfatal stroke, unplanned hospitalization for HF, ESRD, death attributable to CKD, or the need for renal-replacement therapy, or doubling of the baseline serum creatinine. ALTITUDE was prematurely stopped after the second interim efficacy analysis because of frequent occurrence of renal adverse events, hyperkalemia and hypotension in the aliskiren group [55]; this early termination of the ALTITUDE trial gained the great attention of the international nephrology community and was considered again as the end of dual RAAS inhibition. The publication of the full-report of the trial shed light on several issues [11]. All components of primary composite cardio-renal outcome of the ALTITUDE did not significantly differ between the aliskiren and placebo groups, with the exception of resuscitated cardiac arrest. With regards to the renal components of the primary outcome, events of doubling of serum creatinine were evenly distributed between groups, whereas the endpoints of ESRD, dialysis or death due to renal failure were also not significantly different. Moreover, a greater reduction by 1.3/0.6 mmHg in BP and a more effective decrease in ACR by 14 % was noted in the aliskiren relative to the placebo group. In contrast, the main differences noted between groups were the proportion of study participants with hyperkalemia (11.2 % vs. 7.2 %), and reported hypotension (12.1 % vs. 8.3 %) (P < 0.001 for both) [11].
A plausible explanation for the early termination of the ALTITUDE trial due to renal complications can be provided again by the careful evaluation of the characteristics of the population enrolled in the study (Table 3.1). Following the inclusion criteria of macro-albuminuria or eGFR 30–60 ml/min/1.73 m2 and micro-albuminuria or eGFR 30–60 ml/min/1.73 m2 and history of cardiovascular events, the study cohort of ALTITUDE was closer to that of cardiovascular and not renal outcome trials; further, the population characteristics of the ALTITUDE trial differed significantly from that of AVOID. In particular, the mean age of the ALTITUDE population was 65 years, 42 % had cardiovascular disease, 67 % had eGFR lower than 45 ml/min/1.73 m2, and only 58 % of patients were macro-albuminuric. Enrolled patients also had BP lower than 135/85 mmHg, or BP between 135/85 and 170/110 mmHg at first visits if treated with at least three BP-lowering agents; thus mean baseline BP was 137/74 mmHg and 69 % of participants were already taking diuretics in combination with ACEI or ARB. On this basis, a high proportion of patients were particularly prediposed to complications from aggressive BP reduction (indeed hypotension was more frequently recorded in elderly individuals and in patients treated with loop diuretics). Therefore, the serious renal complications that resulted in premature termination of the ALTITUDE trial can be directly attributed to aggressive RAAS blocking in susceptible individuals. In this context, ALTITUDE resembled ONTARGET in showing that double RAAS blockade may provide more harm than good in predisposed individuals with CKD [56].
Finally, efficacy and tolerability of dual RAAS inhibition for proteinuric nephropathy was also evaluated in the recently published Veterans Affairs Nephropathy in Diabetes (VA NEPHRON-D) study [10], in which 1,448 patients with type 2 DM, macro-albuminuria and eGFR ranging from 30 to 89.9 ml/min/1.73 m2 were randomly assigned to receive lisinospril (10–40 mg daily) or placebo on top of treatment with the ARB losartan at a dose of 100 mg per day. The primary composite renal endpoint of the study was the first occurrence of a change in eGFR (a decline of ≥30 ml/min/1.73 m2 if the baseline eGFR was ≥60 ml/min/1.73 m2 or a decline of ≥50 % if the baseline eGFR was <60 ml/min/1.73 m2), ESRD, or death. VA NEPHRON-D study was also prematurely terminated at a median follow-up period of 2.2 years due to safety concerns, as combined RAAS blockade elevated the risk of hyperkalemia in comparison with mono-therapy (6.3 vs 2.6 events per 100 person-years, P < 0.001), as well as the risk of acute renal failure (12.2 vs 6.7 events per 100 person-years, P < 0.001). The risk of the primary renal endpoint did not significantly differ between groups in the whole study cohort (HR with the combination therapy: 0.88; 95 % CI: 0.70–1.12, P = 0.30) and among pre-specified sub-groups (P > 0.10 for all interactions). With regards to the secondary endpoint of decline in eGFR or ESRD, a trend towards a benefit of the combination therapy was noted after about 6–12 months of treatment, but this effect was not sustained with longer follow-up (HR: 0.78; 95 % CI: 0.58–1.05, P = 0.10) [10].
The early termination of VA NEPHRON-D trial due to heightened risk of acute renal failure and hyperkalemia should be rather considered the definite end of dual RAAS blockade for renoprotection. This is mostly because VA NEPHRON-D trial included a relevant population, i.e. patients with diabetic nephropathy and macro-albuminuria, with a mean age of 64.5 years, of whom 23 % had coronary artery disease, 16 % had congestive HF and 30 % had eGFR lower than 45 ml/min/1.73 m2, whereas mean baseline BP of study participants was 137/73 mmHg with background treatment with 3 or more BP-lowering agents [10] (Table 3.1). Overall, as HF with decreased cardiac-output is a typical predisposing factor for acute renal failure, it seems reasonable to avoid dual ACEI/ARB combination for renoprotection in these patients. It may be argued by some that a benefit of dual blockade in retarding proteinuric CKD progression could be still present in healthier populations with little risk for acute renal failure with aggressive RAAS blocking (i.e., young individuals with proteinuric kidney disease, preserved renal function, absence of overt vascular disease and high-adherence to low dietary potassium intake). As specific treatments against proteinuric kidney disease are still lacking a study on such a population may appear in the future, but any results should be cautiously interpreted.
As in heart failure patients, combining an aldosterone-receptor-antagonist with ACEIs or ARBs is proposed as another treatment option for slowing kidney disease progression, since plasma aldosterone is increased in CKD and may independently promote renal damage [57]; further, blockade of RAAS with the use of ACEIs or ARBs does not necessarily lead to prolonged lowering in plasma aldosterone levels [58]. Several pilot studies in proteinuric CKD have revealed promising results that addition of spironolactone in patients with proteinuria already treated with ACEIs or ARBs provides an additional anti-proteinuric effect to that exerted by mono-therapy [59–62]. In parallel, administration of eplerenone in addition to standard antihypertensive treatment with an ACEI was associated with further reduction in UAE in patients with hypertension and left ventricular hypertrophy [63]. In a more recent study 81 patients with DM, hypertension, and macro-albuminuria already under treatment with lisinopril 80 mg were randomized to losartan 100 mg, spironolactone 25 mg or placebo [64]. After 48 weeks of follow-up, ACR was significantly reduced by 34 % in the spironolactone group and by 16.8 % in the losartan group as compared to placebo; ambulatory BP, CrCl, sodium and protein intake, and glycemic control did not significantly differ between groups. However, addition of sprironolactone and losartan to lisinopril resulted in elevated serum potassium levels (yet on average <5.2 meq/L in all groups). Use of low-dose spironolactone in addition to ACEIs or ARBs seems to be a promising treatment approach for proteinuric patients, but large outcome trials evaluating hard renal end-points should be performed before any treatment recommendations can be made. This is mostly due to the concern of heightened rates of hyperkalemia [65], as was also exemplified by large population studies [66] following the extended use of spironolactone for HF after the publication of the Randomized Aldactone Evaluation Study (RALES) [67].
Choice of Antihypertensive Treatment in Patients with Diabetes Mellitus
For more than two decades one of the most active area of research topics in the field of hypertension therapeutics was the possible effects of major antihypertensive drug classes on carbohydrate metabolism and the development of new-onset DM [14]. In this regard, numerous clinical studies have investigated the impact of antihypertensive agents on insulin sensitivity (IS) and glycemic control, whereas large outcome trials have explored the potential association between antihypertensive agents and incident DM. Another important issue is the possible harmful impact of antihypertensive therapy – related new-onset DM on cardiovascular outcomes [14]; thus, presence of DM should be taken into consideration in the choice of the appropriate BP-lowering therapy in patients with hypertension and HF.
Effects of Antihypertensive Agents on Insulin Sensitivity
Preliminary reports from older studies suggested that thiazide and loop diuretics (especially in higher doses, i.e. equivalent to ≥25 mg of hydrochlorothiazide) and conventional β-blockers deteriorate IS and glycemic control and elevate the risk of developing new-onset DM [68–73]. Data on β-blockers and their metabolic effects gain much more interest, as newer, vasodilating agents were proposed to have a much different metabolic profile than conventional ones [74]. As β-blockers remain first-line treatment for HF, these effects are of major importance when selecting the appropriate agents in these patients. In this regard, initial studies that compared the metabolic effect of a vasodilating β-blocker carvedilol with either metoprolol [75–77] or atenolol [78, 79] in patients with hypertension or hypertension and DM showed significant differences between groups, as treatment with conventional β-blockers resulted in reduced IS. This favorable effect was strongly supported by the results of the large Glycemic Effects in Diabetes Mellitus: Carvedilol-Metoprolol Comparison in Hypertensives (GEMINI) multi-center trial that compared effects of carvedilol to metoprolol treatment on IS in 1,235 hypertensive patients with type 2 DM already treated with an ACEI or an ARB. This study revealed a significant reduction in the Homeostasis Model Assessment-Insulin Resistance (HOMA-IR) index of about 9 % with carvedilol relative to metoprolol [77]. Nebivolol, another β 1 -blocker that has been reported to increase nitric oxide (NO) production from endothelium, thus resulting in peripheral vasodilatation [80], was not previously shown to significantly affect IS in patients with hypertension [79] and/or type 2 DM [81, 82]. These beneficial metabolic effects of nebivolol were confirmed in recent studies that showed improvement in glucose metabolism with nebivolol relative to metoprolol in patients with metabolic syndrome [83].
Studies evaluating the effects of ACEIs on metabolic profile suggest that use of these agents in hypertensive patients improves, or at least, does not deteriorate IS. Acute oral administration of 25 mg of captopril during euglycemic hyperinsulinemic clamp in patients with type 2 DM was shown to be associated with a significant increase in insulin-induced glucose disposal [84]. Similarly, subsequent studies with captopril, enalapril, cilazapril, and fosinopril showed improvement in IS and glycemic control in patients with hypertension and/or type 2 DM [69, 85, 86], whereas in other cases treatment with ACEIs was shown to have no impact on these parameters [82, 87, 88], suggesting that this antihypertensive drug class has a positive, or at least neutral effect on metabolic profile.
Similarly to ACEIs, the currently available data on the use of ARBs suggest that these agents may have a neutral or positive impact on IS. A beneficial metabolic effect is particularly applied to telmisartan, that has been proposed to exert unique insulin-sensitizing properties [89]. This is in relation to the fact that telmisartan was shown to activate the peroxisome proliferator-activated receptor (PPAR)–γ receptors, an action that was not proven for other ARBs [90]. Activation of the PPAR–γ receptors is well-documented that results in IS improvement, as it represents the main mechanism of action for thiazolidinendiones [91]. In a randomized study including patients with hypertension and impaired glucose tolerance or type 2 DM, treatment with telmisartan was associated with a 26 % decrease in the HOMA-IR index along with 8 % reduction in fasting glucose, 9 % lowering in HbA1c, and 10 % decrease in fasting insulin, in contrast to losartan, that had no influence on any of these parameters [92]. A mild beneficial effect of telmisartan on HOMA-IR index is also supported by subsequent clinical studies [93–95]. Furthermore, treatment with candesartan was also related to a significant decrease of about 25 % in insulin resistance [96]. It has to be noted, however, that the potentially beneficial effects of ARBs on metabolic profile have to be confirmed in larger studies including euglycemic hyperinsulinemic clamp measurements.
Effects of Antihypertensive Agents on New-Onset DM
Data from Hypertension Outcome Trials Evaluating New-Onset DM as a Secondary Outcome
During the previous 15 years, a number of large-scaled outcome trials attempted to investigate as a secondary outcome, the effect of various antihypertensive drug classes on the occurrence of new-onset DM. Most of these studies compared newer (i.e. ACEIs, ARBs, or calcium channel blockers, or combinations of those) with conventional antihypertensive regimens (i.e. thiazide diuretics or β-blockers or combinations). In the Captopril Prevention Project (CAPPP) trial, in which above 11,000 patients with hypertension were randomly assigned to captopril or conventional treatment with diuretics, β-blockers or both, captopril-based regimen was shown to be associated with significantly lower incidence of type 2 DM in comparison with the conventional regimen (14 % lower RR in the intention-to-treat and 21 % in the on-treatment analyses respectively) [97]. The Intervention as a Goal in Hypertension Treatment (INSIGHT) study randomized more than 6,000 hypertensive individuals to a regimen based on long-acting nifedipine or co-amilozide (hydrochlorothiazide plus amiloride) [98]. After a mean follow-up of about 4 years, the occurrence of new-onset DM was lower in patients receiving nifedipine than in those receiving diuretic-based treatment (4.3 % versus 5.6 %). The LIFE study randomized 9,193 patients with hypertension and left ventricular hypertrophy to receive treatment with losartan-based or atenolol-based antihypertensive regimen for at least 4 years. Among patients without DM at trial initiation, those who were randomized to the losartan-based regimen had an 25 % lower risk of developing new-onset DM throughout the study compared to those on atenolol-based treatment [99].
One of the largest outcome trials in hypertension to evaluate the impact of various antihypertensive drug classes on the incidence of new-onset DM was the ALLHAT trial, in which more than 33,000 hypertensive patients were randomly assigned to a chlorothalidone-based, an amlodipine-based or a lisinopril-based therapy, as discussed above. Among non-diabetic patients at baseline, chlorothalidone treatment was associated with significantly higher incidence of new-onset DM during the 4 years of follow-up than treatment with either amlodipine or lisinopril (11.8 % versus 9.8 % and 8.1 % respectively) [100]. In the Second Australian National Blood Pressure Study (ANBP2), above 6,000 patients aged 65–84 years were randomized to receive treatment with an ACEI (mainly enalapril) or a diuretic (mainly hydrochlorothiazide) [101]. After a median follow-up of 4.1 years, treatment with the ACEI-based regimen resulted in 31 % lower risk of developing new-onset DM compared to the diuretic-based regimen (DM incidence of 4.54 % vs 6.58 % respectively) [102]. The International Verapamil-Trandorapril Study (INVEST) randomized 22,576 hypertensive patients with coronary artery disease to receive verapamil-based or atenolol-based antihypertensive treatment. Again, non-diabetic patients at entry in the verapamil group had a 15 % lower risk of developing new-onset DM than subjects in the atenolol group after a median follow-up of 2.7 years [103]. In agreement with the findings of the INVEST trial, the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT), which assigned more than 19,000 hypertensive patients to an amlodipine plus perindopril arm or an atenolol plus bendroflumethiazide arm, showed that the incidence of new-onset DM was 30 % lower in the amlodipine than in the atenolol group after a median follow-up of 5.5 years [104].
In addition to the above, effects of antihypertensive agents on incidence of new-onset DM were investigated in other trials using placebo as a comparator. In the Heart Outcomes Prevention Evaluation (HOPE) trial, a total of 9,297 patients with an age above 55 years were randomly assigned to receive ramipril (10 mg daily) or placebo on top of their standard medication, which included other antihypertensive agents, for a mean observational period of 5 years. At trial completion, treatment with ramipril induced a 34 % reduction in the incidence of new-onset DM relative to placebo in originally non-diabetic study participants [105]. The Candesartan in Heart Failure Assessment of Reduction in Mortality and morbidity (CHARM) trial, (which had as main inclusion criterion symptomatic HF and not hypertension) revealed that treatment with the ARB candesartan resulted in a significant decrease of 19 % in the risk of new-onset DM as compared with placebo [106].
It has to be noted, however, that some studies did not show statistically significant elevation in the risk of new-onset DM with conventional or significant reductions with newer antihypertensive agents. In the Systolic Hypertension in the Elderly Program (SHEP) trial, patients with isolated systolic hypertension and above 60 years of age were randomized to stepped-care therapy with 12.5–25.0 mg per day of chlorthalidone or matching placebo aiming to reach goal BP (systolic BP <160 mmHg). When BP remained above the goal, study investigators were permitted by protocol to add atenolol or reserpine; thus, 32 % of patients randomly allocated to receive chlorthalidone were also administered atenolol [107]. After 4.5 years of follow-up, incidence of DM was slightly higher in the chlorthalidone group compared with the placebo (8.6 % vs 7.5 %) [107], a finding that was translated into a non-significant 20 % elevation in the risk of new-onset DM (RR: 1.2; 95 % CI: 0.9–1.5) with this treatment [108]. The Swedish Trial in Old Patients with Hypertension 2 (STOP-Hypertension 2) study, in which 6,614 elderly hypertensive patients were randomly allocated to conventional treatment (β-blockers or diuretics) or to either ACEI-based or calcium channel blocker (CCB)-based regimen, also reported negative results; along with the absence of significant difference in the composite primary cardiovascular outcome, no significant differences in incidence of new-onset DM between the three study groups were evident after about 4.5 years of follow-up [109]. In addition, the Nordic Diltiazem (NORDIL) study, in which 10,881 hypertensive patients were randomized to diltiazem, or diuretics and β-blockers, or both, a 13 % lower risk of developing new-onset DM was evident in the diltiazem group, but this difference did not reach statistical significance [110]. The Study on Cognition and Prognosis in the Elderly (SCOPE), conducted in about 5,000 elderly hypertensive subjects aged 70–89 years, showed also a non-significant decrease of 19 % in incidence of new-onset DM with candesartan relative to placebo [111].
The only hypertension outcome trial that compared the treatment effects of thiazide diuretics against that of β-blockers on incidence of new-onset DM was the Heart Attack Primary Prevention in Hypertension (HAPPHY) trial, published almost 25 years ago. In this trial, above 6,500 men aged 40–64 years with mild to moderate hypertension were randomized to receive a thiazide diuretic or a β-blocker. As exactly happened with all components of the composite primary endpoint, incidence of new-onset DM did not significantly differ between groups after 45 months of treatment [112]. Similarly, only one outcome trial provided a direct head-to-head comparison of the effect of agents from two newer antihypertensive classes; the Valsartan Antihypertensive Long-term Use Evaluation (VALUE) study randomized more than 15,000 hypertensive patients to the ARB valsartan or the CCB amlodipine for a mean follow-up period of 4.2 years [113]. Although no significant difference in the primary composite outcome between the two drugs was noted, treatment with valsartan resulted in a 23 % lower risk of new-onset DM in comparison with amlodipine [114].
Overall, the majority of the above mentioned randomized studies support a beneficial impact of newer (ACEIs, ARBs and CCBs) over older (diuretics, β-blockers) antihypertensive agents on the incidence of new-onset DM; it has to be noted, however, that these studies suffer from some methodological limitation that may affect the strength of their findings and thus, the conclusions that can be drawn from them. First, none of the aforementioned studies published so far reporting effects of antihypertensive treatment on new-onset DM had the incidence of DM as the primary trial endpoint [97–99, 101, 103, 104, 107–110, 113]. Second, in some of these studies an ACEI [97, 109] or a CCB [109, 110] was compared to conventional treatment consisting of either diuretics or β-blockers or their combination, and therefore an individual treatment effect of the later antihypertensive classes cannot be easily evaluated. In some studies, a great proportion of patients in the various arms were receiving background treatment with second-line antihypertensive agents that could also affect IS and glycemic control towards the same [101, 103, 104, 110] or the opposite direction [101, 103, 110–112], making again difficult the assessment of the net impact of the main agents on the incidence of DM. Further, the possibility of detection bias in those of the above studies that were open-label with blinded end-point assessment can not be excluded [97, 101, 103, 109, 110, 112], as DM may have been more intensively sought in patients receiving conventional treatment. In other cases, new-onset DM was not diagnosed with the most accurate methods, i.e. it was self reported [105]. Finally, in trials comparing active antihypertensive regimens, the observed differences [97–101, 103, 104] could be arisen either from a harmful effect of conventional agents or from a beneficial impact of the newer drugs, a fact not allowing firm conclusions to be drawn.
Data from Trials and Meta-analyses Investigating the Effects of Antihypertensive Agents on Metabolic Parameters and New-Onset DM
The impact of antihypertensive therapy on new-onset DM was also explored in clinical studies that specifically aimed to evaluate the effects of such drugs on the metabolic profile. In the Antihypertensive Treatment and Lipid Profile in a North of Sweden Efficacy Evaluation (ALPINE) study, 392 subjects with hypertension were randomly assigned to receive treatment with low-dose hydrochlorothiazide alone or combined with atenolol or treatment with the ARB candesartan alone or combined with felodipine. This study showed that treatment for 12 months with the diuretic-based regimen resulted in elevation of fasting plasma glucose and insulin, whereas no increase in these parameters was evident in patients who received candesartan; further, hydrochlorothiazide treatment was associated with significantly higher incidence of new-onset DM than treatment with the ARB candesartan (4.1 % vs 0.5 % respectively) [115].
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
