Antihypertensive treatment and renal protection: Is there a J‐curve relationship?

Abstract

A bidirectional relationship between hypertension and kidney disease, with one exacerbating the effect of the other, is well established. Elevated blood pressure (BP) is a well‐recognized, modifiable risk factor for cardiovascular (CV) disease as well as for development and progression of chronic kidney disease and, therefore, the identification of optimal BP target is a key issue in the management of renal patients. Recent large trials and real life cohort studies have indicated that below a definite BP value renal protection seems to plateau and too low levels may even be associated with a paradoxical increase in renal morbidity, thus reviving the debate about the so called BP ‐renal function J‐curve relationship. Existing evidence supports a systolic target around 130 mm Hg to combine both renal and CV protection and possibly lower levels in the presence of overt proteinuria.

1. INTRODUCTION

Blood pressure (BP) and kidney function may have an opposite effect on each other, with one exacerbating the effect of the other. Hypertension has traditionally been considered both a cause and a consequence of renal damage.1 In fact, the kidney by presiding over long‐term regulation of body fluid homeostasis plays a pivotal role in the fine tuning of volemia and by modulating the activity of the renin angiotensin aldosterone system (RAAS) participates in the regulation of systemic vascular resistance. The pathophysiological mechanisms underlying the relationship between BP and renal function provide the rationale for the exceedingly high prevalence of hypertension observed in patients with chronic kidney disease (CKD).2, 3 Unsurprisingly, it has recently been emphasized that renal impairment is the most common cause of treatment resistant hypertension, defined as the failure to achieve recommended BP values despite optimal combination of at least three different drugs including a diuretic.4 On the other hand, the severity of hypertension has been related to the risk of developing renal damage and even more so to the rate of progression toward end‐stage renal disease (ESRD) in renal patients.5

Antihypertensive treatment, especially with RAAS‐inhibiting drugs (RAAS‐i), is currently considered the most powerful tool to retard progression of renal disease, especially in patients with increased albuminuria, and has been shown to effectively prevent cardiovascular (CV) complications in this high‐risk set of patients.6 Identifying optimal BP target and implementing appropriated therapeutic strategies is therefore a key issue in the management of CKD patients. Results of recent randomized clinical trials (RCTs), as well as real world cohort studies, have also suggested that renal protection seems to plateau when systolic blood pressure (SBP) is reduced to values around 130 mm Hg and there may even be a paradoxical increase in renal morbidity at lower values (Table 1). This has revived the debate about the so‐called J‐curve relationship between BP and renal function.7 Furthermore, it is well known that specific clinical conditions such as the presence and degree of albuminuria not only entail an unfavorable long‐term prognosis but require individually‐tailored and generally more intensive therapeutic approaches. Moreover, paradoxical relationship seems to emerge from results of recent large clinical trials with regards to pharmacologic inhibition of the RAAS and renal outcome (Table 2). In fact, angiotensin converting enzyme inhibitors (ACE‐is) and angiotensin receptors blockers (ARBs) have an established record for renal protection and albuminuria reduction and are considered the antihypertensive drugs of choice in CKD patients. Nonetheless, it appears that more profound blockade of angiotensin II obtained by combining multiple RAAS‐i provides no additional benefit and possibly more adverse effects.

Table 1.

Is more intensive blood pressure good for the kidney?

Reference	Population	BP targets				BP achieved				Results	Major findings associated to intensive BP lowering
		Intensive		Standard		Intensive		Standard
Randomized clinical trial
Klahr et al.17(MDRD) (1994)	840 CKD, mean age 52 y, female 40% Study A (n = 585): GFR 25‐55; Study B (n = 255): GFR 13‐24; proteinuria level <10 g/d	MAP < 92		MAP < 107		126/77		134/81		Rate of change in GFR: NS	No additional renal benefit
Toto et al.41 (1995)	77 HT with CKD; mean age 56 y; female 37%; serum Cr 1.6‐7.0; GFR ≤70; proteinuria ≤2 g/d	DBP 65‐80		DBP 85‐95		133/81		138/87		Rate of decline in GFR: NS	No additional renal benefit
Turner et al.42 (UKPDS 38) (1998)	1148 Type 2 DM with HT; mean age 56 y, female 44%; albuminuria 21%	150/85		<180/85		144/82		154/87		Renal failure: HR 0.58; 95% CI: 0.15‐2.2; P = 0.29	No additional renal benefit
Ruilope et al.43 (HOT) (2001) and Hansson et al.44 (HOT) (1998)	18 790 HT with DBP 110‐115; mean age 62 y, female 47%; CKD 18%	DBP 80	DBP 85		DBP 90	140/81	141/83		143/85	Creatinine increase >30% to a value >2.0 mg/dL: NS	No additional renal benefit
Schrier et al.45(2002)	75 ADPKD with CKD and LVH; mean age 41 y; female 45%; CrCl >30; proteinuria ≤3 g/d	<120/80		135‐140/85‐90		MAP 90 ± 5		MAP 101 ± 4		Change in GFR: NS	No additional renal benefit
Wright Jr et al.18(AASK) (2002)	1094 CKD African Americans; mean age 55 y; female 39%; GFR 20‐70; proteinuria ≤2.5 g/d	MAP <92		MAP 102‐107		128/78		141/85		Rate of change in GFR: NS	No additional renal benefit
Ruggenenti et al.19 (REIN‐2) (2005)	339 CKD with proteinuria 1‐3 g/d and GFR <45, or proteinuria >3 g/d and GFR <70; mean age 54 y; female 26%	<130/80		DBP < 90		130/80		134/82		ESRD: NS	No additional renal benefit
JATOS Study Group,46(2008)	4419 with serum Cr <1.5; mean age 74 y; female 64%; CKD 57%	SBP < 140		SBP 140‐160		NA		NA		Change in GFR; doubled Cr or ESRD: NS	No additional renal benefit
Cushman et al.27 (ACCORD) (2010)	4733 Type 2 DM with HT; mean age 62 y; female 47%; CKD 37%	SBP < 120		SBP < 140		SBP 119.3		SBP 133.5		Renal failure: NSDevelopment of macroalbuminuria: NSDevelopment of microalbuminuria: HR 0.84; 95% CI: 0.72‐0.97; P = 0.019	No additional renal benefit except for reduction in the incidence of albuminuria
Benavente et al.47 (SPS3) (2013)	3020 recent, MRI‐defined symptomatic lacunar infarctions; mean age 63 y; female 37%; CKD 16%	SBP < 130		SBP 130‐149		127/70		137/76		eGFR decline >30%: HR 1.4; 95% CI: (1.1‐1.6)	Worse renal outcome
Schrier et al.48 (HALT‐PKD) (2014)	558 ADPKD with CKD; mean age 37 y; female 49%; GFR >60; proteinuria ≤0.5 g/d (Study A)	95‐110/60‐75		120‐130/70‐80		Difference: SBP, 13.4/DBP, 9.3				Annual % change in kidney volume: NS	No additional renal benefit
Wright Jr et al.23 (SPRINT) (2015)	9361 with GFR ≥20 and proteinuria <1 g/d; mean age 68 y; female 36%; CKD 28%	SBP < 120		SBP < 140		SBP 121.5		SBP 134.6		50% reduction in GFR or ESRD: NS	No additional renal benefit
Post‐hoc analysis in randomized clinical trial
Weber et al.49 (ACCOMPLISH) (2016)	6459 Diabetes with in‐treatment PAS > 110; mean age 67 y; female 42%; CKD 17%	<120	<130	<140	>140	116.3	125.5	134.4	150.4	Serum creatinine increase ≥50%:lowest event rates in the SBP 130‐139 but further reductions below 130 were associated with increases in this outcome	Worse renal outcome with very low BP values
	4246 Not Diabetes with in‐treatment PAS > 110; mean age 70 y; female 34%; CKD 19%					116.3	125.5	134.6	149.8
Beddhu et al.28 (SPRINT) (2017)	6662 SPRINT participants with an eGFR ≥60; mean age 66 y; female 34%; CKD 0%	SBP < 120		SBP < 140		The average between‐group difference in SBP after 6 mo was 15.0 mm Hg				>30% decrease in GFR to a value <60: HR 3.54, CI: 2.50‐5.02, P < 0.001	Worse renal outcome
Cheung et al.24 (SPRINT) 2017	2646 SPRINT participants with an eGFR < 60; mean age 72 y; female 40%CKD 100%	SBP < 120		SBP < 140		123/66		135/72		Composite of >50% decrease in eGFR from baseline or ESRD: NSAfter the initial 6 mo, the intensive group had a slightly higher rate of change in eGFR (20.47 vs 20.32 mL/min per 1.73 m2 per year; P = 0.03	No additional renal benefit
Beddhu et al.50 (SPRINT & ACCORD) (2017)	11 026 SPRINT and ACCORD participants with an eGFR ≥60; mean age 61 y in ACCORD and 64 in SPRINT; female 46% in ACCORD and 34% in SPRINT; CKD 0%	SBP < 120		SBP < 140		The average between group difference in SBP was 13.9 mm Hg in the ACCORD trial and 15.2 mm Hg in SPRINT				>30% decrease in GFR to a value <60: HR 3.49 (2.42‐5.03) in SPRINTHR 2.29 (1.89‐2.76) ACCORD	Worse renal outcome
Metanalysis
Emdin et al.51 (2015)	100 334 pts included in trials on BP lowering treatments in T2DM	Associations with 10–mm Hg Lower SBP								Renal failure: NSAlbuminuria: 0.83 CI: 0.79‐0.87, P < 0.01	No additional renal benefit except for reduction in the incidence of albuminuria
		Associations with 10–mm Hg Lower SBP in trials with Mean Achieved SBP in the Active Group ≥130 mm Hg								Renal failure: NSAlbuminuria: RR 0.71, CI: 0.64‐0.79, P < 0.01
		Associations with 10–mm Hg Lower SBP in trials with Mean Achieved SBP in the Active Group <130 mm Hg								Renal failure: NSAlbuminuria: RR 0.86, CI: 0.81‐0.90, P < 0.01
Xie et al.52 (2016)	19 trials (44 989 pts) with more intensive vs less intensiveBP‐lowering treatment; CKD 7%					133/76		140/81		ESRD: NS	No additional renal benefit
Ettehad et al.31 (2016)	Trials (613 815 pts) on BP lowering treatments									Renal failure: NS	No additional renal benefit
Tsai et al.26 (2017)	Trials (8127 pts) with more intensive vs less intensive blood pressure‐lowering treatment in ptz with CKD and no diabetes					Difference in mean systolic BP varied from 4 to 13 mm Hg at the end of the trial				Composite renal outcome (doubling of serum creatinine level, 50% reduction in GFR, ESRD): NS	No additional renal benefit

BP, blood pressure; CKD is defined as eGFR <60 or albuminuria where available; CI, confidence intervals; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; ESRD, end stage renal disease; HR, hazard ratio; HT, hypertension; MAP, mean arterial pressure; RR, relative risk; SBP, systolic blood pressure.

Table 2.

Dual renin‐angiotensin system blockade and nephroprotection

Reference	Population	Comparison groups	Mean follow‐up	BP achieved	Results	Adverse events
Randomized clinical trial
Mogensen et al.53(CALM) (2000)	197 pts with MA, Ht and T2DM; mean age 60 y, female 35%	ARB (Candesartan 16 mg) vs ACEi (lisinopril 20 mg) vs both (Candersartan 16 mg + Lisinopril 20 mg)	12 wk’ monotherapy with ARB or ACEi followed by 12 wk’ monotherapy or combination treatment	At 24 wk the mean reduction in SBP and DBP with combination treatment were significantly greater than that with ARB (P = 0.002 and P = 0.0003) or ACEi (P = 0.02 and P = 0.005)	UACR reduction with combination treatment (50%, P < 0.001) was greater than with ARB (24%, P = 0.05) and ACEi (39%, P < 0.001)	Creatinine clearance was slightly decreased over 24 wk in the groups treated with ACEi (adj. mean decrease 0.0835 mL/s, P = 0.04) and the combination treatment (0.0735 mL/s, P = 0.05) while was not affected in the group treated with ARB
Epstein et al.54 (2006)	268 pts with T2DM and UACR ≥50 mg/g; mean age 59 y, female 38%	ACEi (Enalapril 20 mg)+placebo vs ACEi (enalapril 20 mg) + ARA50 (eplerenone 50 mg) vs ACEi (enalapril 20 mg) + ARA100 (eplerenone 100 mg)	12 wk	SBP decreased significantly and similarly in all treatment groups	Treatment with ARA50 (41%) or ARA100 (48.4%) but not placebo (7.4%) significantly reduced UACR from baseline (both ARA groups, P < 0.001 vs placebo)	The incidences of sustained and severe hyperkalemia was similar in the three treatment arms and did not differ on the basis of quartile of estimated GFR
Bakris et al.55 (IMPROVE) (2007)	405 pats with Ht at high CV risk and with MA despite prior treatment with ACEi; mean age 66 y, female 38%;MA: 70%; Overt nephropathy: 30%	ACEi (Ramipril) + ARB (irbesartan) vs ACEi (ramipril) + placebo	20 wk	Reduction in both DBP (P = 0.019) and SBP (P = 0.047) was greater in patients receiving ACEi+ARB compared with those receiving ACEi + placebo	Change in AER: adj. week 20 baseline geometric ratios for ACEi+ARB and ACEi + placebo were similar (P = 0.540)	The incidence of AE was similar in both groups. (including hyperkalemia 2.5% of patients in both treatment groups)
Menne et al.56 (VALERIA) (2008)	133 pts with Ht and MA; mean age 58 y, female 38%	ACEi (lisinopril 40 mg) vs ARB (valsartan 320 mg) vs ACEi (Lisinopril 20 mg) + ARB (valsartan 320 mg)	30 wk	BP reduction was similar in the three groups	ACEi+ARB was more effective in reduction of microalbuminuria than monotherapy with either ACEi or ARB and the difference fron ACEi was stastically significant (P = 0.034).	All treatments were safe and well tolerated.
Parving et al.57 (AVOID) (2008)	599 pts with Ht and T2DM with nephropathy (UACR of >300 mg/g or >200 mg/g in pts receiving RAAS inhibitors); mean age 60 y, female 29%	DRI (Aliskiren 150‐300 mg) + ARB (Losartan 100 mg) vs placebo + ARB (Losartan 100 mg)	6 mo	In the DRI+ARB group was observed a small, not significant greater reduction of SBP (P = 0.07) and DBP (P = 0.08)	DRI reduced mean UACR more effectively as compared to placebo (P < 0.001), with a reduction of ≥50% in 24.7% of the patients treated with DRI as compared with 12.5% of those who received placebo (P < 0.001). The mean rate of decline in eGFR during the study period was 2.4 mL/min×1.73 m2 in the DRI group and 3.8 mL/min×1.73 m2 in the placebo group (P = 0.07)	The total numbers of AE and serious AE were similar in the groups
Mann et al.32 (ONTARGET) (2008)	25 620 pts ≥ 55 y with coronary, peripheral or cerebrovascular disease or diabetes with end‐organ damage; mean age 66 y, female 27%; eGFR < 60: 24%; MA: 13.1%; Macroalbuminuria: 4%	ARB (Telmisartan 80 mg) vs ACEi (Ramipril 10 mg) vs ARB (Telmisartan 80 mg) + ACEi (Ramipril 10 mg)	4.7 y	Pts in the ARB group and the ACEi+ARB group had slightly lower BP throughout the study period than did patients in the R group	The incidence of the composite primary renal outcome (all cause death, doubling serum creatinine, ESRD) was similar with ARB and ACEi but increased with ARB+ACEi (ACEi+ARB vs ACEi HR 1.09, P = 0.037)	The incidence of acute dialysis was similar with ARB and ACEi but increased with ARB+ACEi (ACEi+ARB vs ACEi HR 2.19, P = 0.020) Renal abnormalities and reason for permanent stopping medication were similar with ARB and ACEi but increased with ARB+ACEi (ARB+ACEi vs ACEi RR 1.33, P < 0.0001 and RR 1.58, P < 0.0050).
					The secondary renal endpoint (ESRD or doubling of serum creatinine) was similar with ARB and ACEi but was more frequent with ACEi+ARB (ACEi+ARB vs ACEi HR 1.24, P = 0.038)
					Development of MA or macroalbuminuria was similar with ARB and ACEi, reduced with ACEi+ARB (ACEi+ARB vs ACEi HR 0.88, P = 0.003)
Mehdi et al.58(2009)	81 pts with T2DM, Ht, and albuminuria (UACR ≥300 mg/g) in treatment with lisinopril (80 mg once daily); mean age 50 y, female 64%	ACEi (Lisinopril 80 mg) + placebo vs ACEi (lisinopril 80 mg) + ARB (losartan 100 mg) vs ACEi (lisinopril 80 mg) + ARA (spironolactone 25 mg)	48 wk	Both 24‐h ambulatory SBP, clinical SBP and clinical DBP decreased significantly from the baseline at 24 and 48 wk in all three groups, but equally	Compared with ACEi+placebo, UACR decreased by 34.0% (P = 0.007) in the group ACEi+ARA and by 16.8% (P = 0.20) in the group ACEi+ARB	Serum potassium level was significantly higher with the addition of either ARA or ARB
Imai et al.59 (ORIENT) (2011)	566 pts with T2DM and with overt nephropathy; mean age 69 y, female 31%	ARB (Olmesartan from 10 to 40 mg) + ACEi vs placebo+ACEi	3.2 y	Time‐averaged differences of systolic and diastolic BP were greater in ARB+ACEi group compared to placebo+ACEi (P  < 0.01)	There was no difference in the primary renal composite outcome (doubling of serum creatinine, ESRD and death) (ARB+ACEi vs placebo+ACEi HR 0.97 P  = 0.791, after adjustement for BP HR 1.02; P  =  0.852)	Hyperkalemia was more frequent in the ARB+ACEi group than the placebo+ACEi group (9.2% vs 5.3%)
					Olmesartan ARB+ACEi significantly decreased proteinuria (P  =  0.005) and rate of change of reciprocal serum creatinine
Parving et al.33 (ALTITUDE) (2012)	8561 pts with T2DM at high cardiorenal risk; mean age 64 y, female 32%;eGFR < 60%:68%; MA: 25%; Macroalbuminuria: 58%	DRI (Aliskiren 150‐300 mg) vs placebo on top of optimal treatment with an ACEi or an ARB	Prematurely stopped after a median follow up of 33 mo	SBP and DBP were lower with DRI (between‐group differences, 1.3 and 0.6 mm Hg, respectively)	The incidence of renal composite outcome (ESRD, death attributable to kidney failure, or the need for RRT with no dialysis or transplantation available or initiated; or a serum creatinine value that was at least double the baseline value and that exceeded the upper limit of the normal range (>0.9 mg/dL in women and >1.2 mg/dL in men), sustained for at least a month) was similar in the two groups (HR 1.03, P = 0.74). There were no significant differences between study groups for any component of the renal outcome	Hyperkalemia was significantly higher in the DRI group than in the placebo group (11.2% vs 7.2%), as was the proportion with reported hypotension (12.1% vs 8.3%) (P < 0.001 for both comparisons)
					Reduction in UACR was greater in DRI group (between‐group difference, 14%; 95% CI: 11‐17%)
Fried et al.34 (VA NEPHRON‐D) (2013)	1448 pts with T2DM and proteinuric CKD (1‐3 stages); mean age 64 y, female 28%;eGFR<60: 62%	ARB (Losartan 100 mg) + ACEi (Lisinopril 10‐40 mg) vs ARB (Losartan 100 mg)+placebo	Prematurely stopped owing to safety concerns after a median follow‐up of 2.2 y	After adjustment of the ACEi or placebo dose, the ARB + ACEi had a slightly lower blood pressure than the ARB+placebo group (within 2 mm Hg)	The incidence of renal composite outcome (first occurrence of a change in the eGFR, ie a decline of ≥30 mL/min× 1.73 m2 if the initial eGFR was ≥60 mL/min× 1.73 m2 or a decline of ≥50% if the initial eGFR was <60 mL/min× 1.73 m2, ESRD, or death) was similar in the two groups (ARB+ACEi vs ARB+placebo HR 0.88, P = 0.30)	ARB+ACEi increased the risk of hyperkalemia (P < 0.001) and AKI (P < 0.001)
Torres et al.60 (HALT‐PKD) (2014)	486 pts with autosomal dominant polycystic kidney disease with eGFR from 25 to 60 mL/min× 1.73 m2; mean age 48 y, female 52%	ACEi (lisinopril) + placebo vs ACEi (lisinopril) + an ARB (telmisartan), with the doses adjusted to achieve a BP of 110‐130/70‐80 mm Hg	5.2 y	The two treatments controlled BP similarly	There was no significant difference between the study groups in the incidence of the composite primary outcome (death, ESRD, or a 50% reduction from the baseline eGFR) (HR with ACEi+ARB, 1.08; 95% confidence interval, 0.82‐1.42)	AE, including hyperkalemia and AKI, were similar in the two groups
Bakris et al.61 (ARTS‐DN) (2015)	821 pts with T2DM and high or very high albuminuria who are receiving an ACEi or an ARB; mean age 64 y, female 22%UACR ≥300 mg/g: 36.7% eGFR ≤60: 40%	ARA (Finerenone from 1.25 to 25 mg) vs placebo on top of ACEi or an ARB	90 d	Post hoc analysis showed that no significant correlation across all treatment groups between the ratio of UACR and the change in SBP	The primary outcome, the placebo‐corrected mean ratio of the UACR at day 90 relative to baseline, was reduced in the ARA groups respect to placebo (P < 0.01 for all comparisons)	Increases in serum potassium of at least 5.6 mmol/L, in 12 of 821 patients (1.5%), all of whom were receiving ARA, leading to subsequent discontinuation of study treatment
					The incidences of a decrease in eGFR at least of 40% were similar in the placebo and ARA groups (all doses) and no occurrences of eGFR decreases of at least 57%
Retrospective cohort study
McAlister et al.62 (2011)	32 312 elderly pts who were new users of an ACEi, an ARB or a combination of both medications between May 1, 2002, and Dec 31, 2006	ACEi+ARB vs ACEi or ARB alone	6 mo	NA	The primary composite outcome (doubling of serum creatinine or development of ESRD requiring dialysis or all‐cause death within 6 mo among new users of combination therapy vs monotherapy) was more common among patients given ACEi+ARB (5.2 events per 1000 pts per month) than among patients given monotherapy (2.4 events per 1000 pts per month), even after multivariable adjustment (adj.HR 2.36, 95% CI: 1.51‐3.71)	Hyperkalemia was more common among patients who received combination therapy than among patients who received monotherapy (adju. HR 2.42, 95% CI: 1.36‐4.32)
Meta‐analysis and systematic review
Kunz et al.63 (2008)	571 pts with or without diabetes and with MA or proteinuria mean age 76 y eGFR ≤60:40.7%	ACEi+ARB vs ACEi or ARB alone	1‐4 mo5‐12 mo	NA	ACEi+ARB therapy reduced proteinuria more effectively than either agent alone in the short term while in the long term this was true only in comparison with ARB	Many of the smaller studies did not provide reliable data on AE
Navaneethan et al.64 (2009)	Pts already on RAS inhibition therapy (ACEi or ARB) with CKD stages 1‐4 with albuminuria or proteinuria secondary to both diabetic and nondiabetic CKD. Seven trials (372 patients) compared non‐selective ARA+ACEi and/or ARB to ACEi and/or ARB alone	Adding non selective ARA (spironolactone from 25 to 50 mg)		Nonselective ARA along with ACEi and/or ARB significantly reduced SBP and DBP	In comparison to ACEi and/or ARB alone nonselective ARA + ACEi and/or ARB significantly reduced 24 h proteinuria without an improvement in GFR	There was a significant increase in the risk of hyperkalemia with the addition of nonselective ARA to ACEi and/or ARB (RR 3.06, 95% CI: 1.26, 7.41)
Maione et al.65 (2011)	4969 pts with albuminuria (MA or macroalbuminuria) and ≥1 cardiovascular risk factor	Combination therapy (ACEi+ARB) vs each monotherapy	NA	NA	There was no significant reduction in the risk of ESRD or doubling of serum creatinine with combined therapy (ACEI+ARB) when compared to each monotherapy but a significant reduction in the risk of progression from MA to macroalbuminuria (RR 0.80, 95% CI: 0.69‐0.92) and no significant reduction in the risk of regression from MA to normoalbuminuria	ACEi+ARB was associated with comparable AE rates, except for a higher risk of cough compared with ARB alone and a higher risk of hypotension compared with ACEI alone
Susantitaphong et al.66 (2013)	4975 CKD patients with proteinuria or low eGFR (<60 mL/min×1.73 m2); mean age ranged from 25 to 66 y; eGFR < 60:48.5%	Combined (ACEI+ARB, ACEI or ARB+ARA, ACEI or ARB+DRI) vs. single RAS blockade (ACEI and ARB and ARA)	Ranged from 1 to 49 mo	Combined RAS blockade therapy was associated with absolute net decreases in SBP and DBP	Combined RAS blockade therapy was associated with a significant net decrease in eGFR (−1.8 mL/min×1.73 m2; P = 0.005), albuminuria (−90 mg/g of creatinine; P = 0.001), and proteinuria (−291 mg/g; P = 0.003). Combined RAS blockade therapy was associated with a 9.4% higher rate of regression to normoalbuminuria	Combined RAS blockade therapy was associated with a significant net increase in serum potassium level, a 3.4% higher rate of hyperkalemia, and a 4.6% higher rate of hypotension
Bolignano et al.67 (2014)	596 pts with CKD stages 1‐4 with MA or proteinuria	ARA (both selective (eplerenone) and non‐selective (spironolactone)) alone or in combination with ACEi or ARB vs other anti‐hypertensive strategies or placebo	NA	There was a significant reduction in both SBP and DBP with additional non‐selective aldosterone antagonist therapy	Compared with ACEi or ARB (or both), non‐selective ARA (spironolactone) combined with ACEi or ARB (or both) significantly reduced 24‐h protein excretion. The effect on eGFR was uncertain (9 studies, 528 participants; MD −2.55 mL/min/1.73 m², 95% CI: −5.67 to 0.51) as compared to ACEi or ARB	Spironolactone + ACE or ARB (or both) doubled the risk of hyperkalaemia and increased the risk of gynaecomastia compared to ACEi or ARB (or both)
Palmer et al.68 (2015)	43 256 adults with T2DM and kidney disease	Any orally administered blood pressure‐lowering agent alone or in combination vs a second blood pressure agent or combination, placebo, or control.	NA	ACEi+CCB treatment lowered DBP to a greater extent than did monotherapy with a CCB, ACE i, ARB, or β blocker	ESRD: ACEi+ARB was better than placebo (OR 0.77, 95% CI: 0.65‐0.92), CCB (OR 0.74, 95% CI: 0.56‐0.98) and DRI (OR 0.64, 95% CI: 0.44‐0.94). Doubling serum creatinine: ACEI+ARB was better than DRI and BB, similar to ACEi or ARB. Regression of albuminuria: ACEi+ARB was similar to ARB or ACEI alone	ACEi+ARB was associated with a borderline increases in estimated risks of hyperkaliemia (odds ratio 2.69, 95% CI: 0.97‐7.47) and AKI (OR 2.69, 0.98‐7.38)
Catalá‐López et al.69 (2016)	69 380 adults with T2DM; mean age 60 y, female 49%	ACEi or ARB as monotherapy vs combination of RAS blockers (ACEi, ARB and DRI)	3.2 y	NA	Progression of renal disease (doubling of serum creatinine + ESRD + all cause mortality) no RAS blockers combination was associated with any significant reduction of progression of renal disease respect to ACEi or ARB alone

ACEi, angiotensin converting enzyme inhibitors; Adj, adjusted; AE, adverse effects; AER, album excretion rate; AKI, acute kidney injury; ARA, aldosterone receptor antagonist; ARB, angiotensin receptor blockers; BP, blood pressure; CCB, calcium channel blockers; CV, cardiovascular; DBP, diastolic blood pressure; DRI, direct renin inhibitors; eGFR, estimated glomerular filtration rate; ESRD, end‐stage renal disease; Ht, hypertension; MA, microalbuminuria; NA, not available; pts, patients; RAS, renin angiotensin system; RRT, renal replacement therapy; SBP, systolic blood pressure; T2DM, type 2 diabetes mellitus; UACR, urinary albumin:creatinine ratio.

Moving from a narrative review of the literature, this article highlights current evidence about BP reduction and renal protection in CKD patients and suggests areas for further clinical research. It is argued that a J‐curve relationship links renal function to the degree of BP reduction as well as pharmacologic inhibition of the RAAS.

2. GLOMERULAR FILTRATION RATE (GFR) AND PROTEINURIA AS INTERMEDIATE END‐POINTS: WHICH ONE SHOULD WE USE TO PREDICT LONG‐TERM RENAL PROTECTION?

Because of the relatively long natural history of chronic renal diseases, evaluating the effectiveness of new therapeutic interventions necessitates the completion of long term, large clinical studies when traditional, hard end‐points such as ESRD or renal replacement therapy (RRT) are used. This has led clinical investigators to rely on surrogate end points and biomarkers in order to speed up drug development and contain the size and costs of clinical trials. While changes in albuminuria and in GFR have been proposed and widely used as surrogate end‐points to assess renal protection in clinical trials, both these biomarkers have specific limitations that should be kept in mind when results are to be translated to real world clinical practice. In fact, in the context of renal function worsening under antihypertensive treatment, changes in renal function and albuminuria may follow a divergent path, with a paradoxical improvement in albuminuria taking place along with reduction in GFR.8 Understanding the merit and limitations of changes in both GFR and albuminuria overtime as predictor of outcome is a relevant issue to account for and reconcile the apparent contrasting data on BP lowering and renal protection.

Clinical trials have clearly shown that the efficacy of RAAS‐is in retarding progression to ESRD is related to their ability to reduce albuminuria and this has led to the hypothesis that albuminuria should be a target for therapy.9 However, albuminuria it is not a necessary step in the causal pathway from CKD to ESRD, in contrast to another accepted intermediate end point such as the doubling of serum creatinine. In fact, many patients may progress to ESRD without ever showing any significant amount of albuminuria.10, 11

In principle, changes in GFR seem to be a better tool for clinical assessment of long‐term renal protection, and doubling of serum creatinine has traditionally been accepted as a reliable intermediate end point in renal trials.12 Nonetheless, early variations in renal hemodynamics induced by systemic BP changes and/or the concomitant use of certain drugs may confound results’ interpretation unless specific clinical conditions and long‐term follow‐up are verified. Thus, initiation (or titration) of RAAS‐i has been shown to cause an early drop in GFR, up to 30%‐50%, despite providing greater long‐term renal protection, which is directly proportional to the degree of baseline as well as residual albuminuria. More recently, new evidence from real world clinical studies seems to challenge the notion that early changes in GFR due to treatment induced variations of intra renal hemodynamics are to be taken as benign and possibly favorable indicators for long term prognosis, especially in the absence of albuminuria.13 In general, both pharmacological and non‐pharmacological factors that modify renal hemodynamics (including systemic BP changes) may have an impact on GFR but this can only be appropriately assessed in long term studies. As an example, Bardoxolone, an activator of the cytoprotective Nrf2 pathway, was able to increase GFR although with no long term benefit and increased risk.14 Thus, although a 30%‐40% reduction in GFR from baseline has been shown to reliably predict renal outcome,15 long term follow up is mandatory when interpreting GFR variations over time and the use of greater cut‐off, such as 57% reduction from the baseline value, corresponding to doubling of serum creatinine, are advised.16

3. IS MORE INTENSIVE BP REDUCTION GOOD FOR THE KIDNEY?

In the last years several trials tested whether CKD patients would benefit from a BP target lower than the one usually recommended in adults with hypertension (Table 1). Three prospective randomized trials have addressed the hypothesis that patients with nondiabetic CKD may benefit from a target SBP lower than 130 mm Hg 17, 18, 19 and generally reached the same negative conclusions in a medium term follow‐up. The results of all these trials suggested, in fact, that patients with nondiabetic CKD do not benefit from more intensive BP‐lowering with regard to adverse renal and CV outcomes, while only results of some post‐hoc analyses suggested a benefit of intensive BP treatment in the subgroups with significant proteinuria.20, 21 Notably, these trials were not designed and powered to evaluate the impact of lower BP on CV outcomes. A long‐term meta‐analysis combining data from MDRD and AASK studies has recently been published. Interestingly, over a two decades follow‐up period, strict BP control did not reduced the risk of ESRD although it significantly lowered the relative risk of death in CKD patients (0.87, 95% confidence interval [CI]: 0.76‐0.99), suggesting that this therapeutic strategy may convey greater benefit on global and CV rather than on renal outcome.22 These results are in line with those more recently reported by RCTs such as SPRINT, wherein an intensive therapeutic strategy aiming at SBP values below 120 mm Hg was not associated to any kidney benefit in patients with CKD while a higher rate of renal adverse effects was observed in the short term.23, 24 As detailed more recently in a post‐hoc analysis, the overall effect of intensive BP control on incident albuminuria was favorable, although it did not prevent eGFR decline among participants with CKD and it significantly worsened renal function among those without CKD at baseline.25 Moreover, the adverse effect of intensive treatment on acute kidney injury (AKI) was not significantly modified by eGFR and was consistently observed across eGFR groups with an overall HR of 1.65 (1.31‐2.08).25 Accordingly, in a recent systematic review26 including 9 RCTs with 8127 non diabetic patients with CKD and a median follow‐up of 3.3 years, intensive and standard BP control provided similar effects on composite renal outcome. In particular, intensive BP control did not show a significant difference on the annual rate of change in GFR, doubling of serum creatinine level or 50% reduction in GFR, or ESRD, except for nonblack patients and those with higher levels of proteinuria which showed a trend of lower risk of kidney disease progression with intensive BP‐lowering treatments. As for patients with diabetes, in the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial the intensive treatment arm, targeting SBP < 120 mm Hg did not result in better renal and CV outcome, although a trend to lower stroke incidence was evident as compared to standard BP target arm.27 A recent pooled analysis of the ACCORD and SPRINT trials further strengthens this concept, showing that lower BP target is consistently associated with worse renal outcome independent of the presence of diabetes and albuminuria. While the absolute risk of developing stage 3 CKD is highest in diabetic patients with albuminuria, those without albuminuria do show greater relative risk of losing renal function with low BP regimens.28 These findings suggest that a BP target around 130 mm Hg should be adopted if one wants to maximize both CV and renal protection.6

Interestingly, in a large, real‐life cohort study in patients with type 2 diabetes and resistant hypertension in Italy, the achievement and maintenance of recommended BP values (ie 75% of visits with BP <140/90 mm Hg) were associated with a significantly greater risk of developing stage 3 or greater CKD and/or a clinically relevant reduction in eGFR over a 4‐year follow‐up.29 Furthermore, in this cohort the relationship between achieved BP and renal function seemed to be J‐shaped, and a worse outcome was recorded at SBP values below 130 mm Hg.

Reasons for this seemingly paradoxical association include abnormalities in auto‐regulatory mechanisms associated with CKD, such as ageing, resistant hypertension or atherosclerosis. Pathologic changes in vascular structure, which impair renal adaptation to fluctuations in BP levels30 may explain the J‐shaped phenomenon observed in the studies detailed above. A recently published meta‐analysis, conducted on a very large sample of over 600 000 patients, confirmed that BP lowering significantly reduces vascular risk across a wide range of baseline BP levels and comorbidities but failed to show similar benefits in terms of renal protection.31

4. RAAS INHIBITION AND RENAL OUTCOME: A PARADOXICAL RELATIONSHIP?

A further troublesome issue with important clinical implications, when one considers that treatment with ACEis or ARBs is the antihypertensive treatment of choice in patients with CKD, is the relationship between renal function and the degree of pharmacologic inhibition of the RAAS.

Combining multiple drugs which act at different sites of the RAAS enzymatic cascade has an attractive pathophysiological rationale, as Angiotensin II may be generated through ACE‐independent pathways, a phenomenon known as “Angiotensin escape”. Accordingly, several small studies have shown additive antiproteinuric effects of combination treatment with an ACE‐I and an ARB. Along this line, exploratory trials have been performed to test whether more complete pharmacological RAAS blockade, sometimes with three drug combination regimens, may further improve renal outcome. While combination treatment has proved more effective than standard therapy (with a single RAAS agent) in lowering proteinuria in the short term, this approach has generally failed to provide long‐term renal and CV benefits. Results of the ONTARGET32 and ALTITUDE33 trials have led to the conclusion that long term renal risk could outweigh the benefits of greater albuminuria reduction associated with combination treatment, at least in the subset of patients with no or modest proteinuria. This hypothesis was further tested in the context of overt diabetic renal disease and albuminuria, in the VA Nepron‐D trial. This trial, however, was also prematurely stopped because of an increased risk of acute kidney injury and hyperkalemia, while no clinical benefit in terms of renal progression was observed.34 European and U.S. Drug Regulatory Agencies, do not currently recommend combination treatment with either an ACE‐I, an ARB or the DRI Aliskiren.

Thus, despite the fact that dual RAAS blockade has been demonstrated to provide greater proteinuria reduction, its use in CKD patients has been questioned. Although proteinuria is a well‐known risk factor for disease progression, its reduction under antihypertensive treatment has never been clearly associated to improvement of hard end‐points. On the other hand, in the last 10 years several studies have demonstrated that although AKI is often reversible, some patients may experience incomplete recovery, while others subsequently develop accelerated loss of renal function, resulting in an increased risk of CKD.35, 36

Moreover, the benefit of treatment with dual, sometimes triple RAAS blockade remains anecdotal at present, and only small retrospective studies have been published so far. Thus, pharmacologic inhibition of RAAS seems to follow a paradoxical pattern, similar to that described for BP reduction.

Overall, results of clinical studies support the view that modulation of RAAS rather than maximal inhibition is beneficial in order to slow down CKD progression (Table 2).

5. OPTIMAL BP VALUES FOR RENAL PROTECTION: AN INDIVIDUALIZED APPROACH BASED ON RENAL PHENOTYPE

The issue of ideal BP target in CKD patients is currently an area of considerable controversy as indicated by the heterogeneity of recommendations by several major scientific societies (Table 3). Pharmacologic intervention to reduce high BP in CKD patients should ideally achieve the dual goal to prevent CV events, the most frequent complication of CKD, and to delay progression of renal damage. This may sometimes pose peculiar clinical challenges. In fact, a comprehensive meta‐analysis of 13 randomized trials and a total of over 37 000 patients with diabetes37 suggests that a target SBP of 130‐135 mm Hg is acceptable and reduces cerebrovascular and total mortality. However, when more intensive goals are pursued heterogeneity in organ damage emerges, in that the risk of stroke continues to fall but there is no benefit as for renal events, while the risk of serious adverse events even increases.37

Table 3.

Target blood pressure for patients with CKD as recommended by International Guidelines

Society	Year	Population	Target BP
KDIGO38	2012	All adult with CKD	≤140/90 mm Hg; with albuminuria ≤130/80
ESH‐ESC70	2013	All adults with CKD	<140/90 mm Hg; in diabetic patients DBP < 85 mm Hg)
JNC 871	2014	All adults with CKD	<140/90 mm Hg
NICE72	2014	All adults with CKD	<140/90; with diabetes or albuminuria <130/80 mm Hg
AHA/ACC/CDC73	2014	All adults with CKD	<140/90 mm Hg
ASH/ISH74	2014	All adults with CKD	<140/90 mm Hg
NHFA75	2016	All adults with CKD	<140/90 mm Hg, if well tolerated aiming at <120/80 mm Hg has shown benefits
CHEP76	2016	All adults with CKD	Non diabetic CKD: <140/90 mm HgIn diabetes: <130/80 mm Hg
ADA77	2017	Adults with DM and CKD	<140/90 mm Hg; with albuminuria <130/80 (avoiding DBP<60‐70 mm Hg)
ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA39	2017	All adults with CKD	<130/80

AAPA, American Academy of Physician Assistants; ABC, Association of Black Cardiologists; ACC, American College of Cardiology; ACPM, American College of Preventive Medicine; ADA, American Diabetes Associations; AGS, American Geriatrics Society; AHA, American Heart Association; APhA, American Pharmacist Association; ASH, American Society of Hypertension; ASPC, American Society for Preventive Cardiology; CDC, Centers for Disease Control and Prevention; CHEP, Canadian Hypertension Education Program; ESC, European Society of Cardiology; ESH, European Society of Hypertension; ISH, International Society of Hypertension; JNC 8, eighth Joint National Committee; KDIGO, Kidney Disease: Improving Global Outcomes; NHFA, National Heart Foundation of Australia; NMA, National Medical Association; PCNA, Preventive Cardiovascular Nurses Association.

In 2012 the KDIGO BP working group published its clinical practice guidelines on the management of BP in CKD. On the basis of long‐term randomized controlled trials, meta‐analyses and also systematic reviews these Guidelines recommended a target BP value below 140/90 mm Hg for CKD patients without proteinuria and a lower BP target (ie <130/80 mm Hg) in subjects with moderate‐to‐severe albuminuria (ie urinary albumin to creatinine ratio >30 mg/g), independently of the presence of diabetes.38 As for the choice of antihypertensive drugs, these guidelines recommend ACE‐Is or ARBs for CKD patients with increased urinary albumin excretion. However, although the evidence linking the use of RAAS‐I and renal protection is not particularly strong in the absence of albuminuria, given the relatively high risk profile of CKD patients, ACE‐Is or ARBs should always be considered the treatment of choice, thanks to their undisputed CV benefit.

Later, in a commentary the NKF‐KDOQI argued that available literature is actually inconclusive and does not prove that a strict cut‐off value improves prognosis in adults with CKD. Furthermore, treatment approach or BP target may vary considerably on the basis of the underlying etiology and presenting phenotype. The recently published American Heart Association/American College of Cardiology (AHA/ACC) Guidelines, largely based on results from the SPRINT study, endorse a BP target below 130/80 for all patients, included those with CKD.39 This is reasonable in light of the need to provide both CV and renal protection in this high risk subgroup. On the other hand, one could speculate that, strictly speaking, BP target recommended by these Guidelines are in contrast with results from the SPRINT trial. Implicitly this choice may signal the awareness about possible discrepancies in the methodology of BP measurement between SPRINT and current literature. In fact, unattended automated BP measurements, as implemented in SPRINT, have been shown to underestimate traditional office BP measurements of about 15/8 mm Hg40 a finding which accounts for the more relaxed BP values adopted by the AHA/ACC final recommendations.

6. CONCLUSIONS

In conclusion, while BP reduction preferentially obtained by the use of ACE‐is or ARBs based treatment is the most powerful tool to retard progression toward ESRD in CKD patients, considerable controversy still remains with regards to optimal BP values. Existing evidence supports a systolic target around 130 mm Hg to combine both renal and CV protection and possibly lower levels in the presence of overt proteinuria. Lower BP levels may not be associated with further renal protection and possibly entail greater risk especially in patients without albuminuria. A similar paradoxical increase in renal morbidity seems to be associated with profound pharmacologic inhibition of the RAAS in CKD patients. It appears that RAAS activity should therefore be modulated, rather than completely inhibited to provide optimal renal protection. Future clinical studies should aim at clarifying current areas of uncertainty, such as the role of RAAS‐I vs non RAAS‐I and/or different BP target in non albuminuric CKD, the most common renal presentation in the real life clinical setting. In conclusion, evidence from available literature suggests that in CKD patients, an individually tailored therapeutic strategy should be implemented. BP targets and types of antihypertensive drugs should be chosen on the basis of global risk profile as well as on renal phenotype, balancing the risk/benefit ratio for both CV and renal outcome.

CONFLICT OF INTEREST

The authors report conflict of interest to disclose.

Viazzi F, Leoncini G, Grassi G, Pontremoli R. Antihypertensive treatment and renal protection: Is there a J‐curve relationship?. J Clin Hypertens. 2018;20:1560‐1574. 10.1111/jch.13396