Key words
diabetic kidney disease - angiotensin-converting enzyme inhibitors - angiotensin receptor blockers - albuminuria - cardiovascular events - meta-analysis
Introduction
Diabetic kidney disease (DKD) occurs in almost 25–40% of patients
with diabetes mellitus (DM) within 20–25 years of the onset of disease, and
is responsible for end-stage kidney disease (ESKD) [1]
[2]. Most people with DKD have hypertension
and the blood pressure rises as albuminuria increases, however, some studies have
shown that the incidence of microalbuminuria among type 2 diabetic patients without
hypertension was approximately 40% [3]. And the increased blood pressure and microalbuminuria in normotensive
patients with type 2 diabetes were associated with an increased cardiovascular risk
[4].
Published guidelines recommend angiotensin-converting enzyme inhibitors (ACEIs) or
angiotensin II receptor blockers (ARBs) as the first line drugs for DKD patients
with hypertension to reduce cardiovascular risk, kidney failure, and death [5]. Recent studies also provide evidence that
treatment with ACEIs or ARBs reduce the urinary albumin excretion rate (UAER) and
retards the progression of kidney disease in normotensive diabetic patients [6]
[7]. Some studies did not find the trend
toward an antiproteinuric effect of ACEIs or ARBs [8]
[9], which may be due to the short duration of
observation or less of the patient included. This systematic review was therefore
undertaken to assess the effects of ACEIs and ARBs on kidney and cardiovascular
outcomes in normotensive patients with DKD.
Materials and Methods
This systematic review of the literature was undertaken according to the approach
recommended by the statement in PRISMA-P 2015. This meta-analysis included
randomized controlled trials (RCTs) in which ACEIs and ARBs was compared to placebo
or an alternative antihypertensive agent in patients with normotensive diabetic
kidney disease of various stages: microalbuminuria (albumin
excretion<30–300 mg/d) or macroalbuminuria (albumin
excretion>300 mg/d) with or without eGFR (estimated
glomerular filtration rate)
60 ml/min/1.73 m2.
Search strategy and search selection
Two authors independently searched from the following data sources without
language restriction: MEDLINE via Ovid (from 1950 to August 2018), EMBASE (from
1966 to August 2018), and Cochrane Library databases using the MeSH headings and
text words of all spellings of known ACE inhibitors and ARBs, RCTs, diabetic
kidney disease, kidney outcomes and cardiovascular events.
We included all available RCTs, which compared ACEIs/ARBs with placebo or
other antihypertensive agents on the effects of kidney outcomes (including
decrease in albuminuria, change of glomerular filtration rate), cardiovascular
outcomes (defined as a composite of fatal or nonfatal myocardial infarction,
angina, stroke, heart failure, and cardiovascular death), all-cause death, or
drug-related adverse events [including hyperkalemia (commonly defined as serum
potassium>5.5 mmol/l), cough, hypotension, allege, and
edema] in patients with normotensive DKD. All completed RCTs that assessed the
effects of ACEIs/ARBs compared with placebo or other antihypertensive
drugs in normotensive patients with diabetic kidney disease, and which reported
cardiovascular, renal or adverse outcomes, were eligible for inclusion. The
exclusion criteria for the following studies are: (i) studies on population who
suffered from DKD along with hypertension or other renal diseases. (ii) studies
in patients undergoing dialysis or kidney transplantation. (iii) studies that
fail to report the mean value or data necessary to estimate the standard
deviation (SD) of the primary efficacy outcome.
Data extraction and quality of evidence
Two authors extracted data using standard data extraction forms, which included
participants, interventions, comparisons, and outcomes. We used standard
criteria (Jadad) to assess the quality of the trials (randomization, concealment
of allocation, double blinding, withdraw and dropouts). Differences were
resolved by consultation with a third reviewer.
Statistical analysis
We calculated odd risk (OR) and 95% confidence interval (CI) for each
outcome by the random-effects model. For the continuous measurement of change of
GFR, blood pressure and albuminuria, we used the weighted mean difference
between groups. Regression analysis was conducted using the weighted mean
difference of the systolic blood pressure at baseline and the weighted mean
difference of the decrease of proteinuria to demonstrate that ACEIs/ARBs
have the function of decreasing proteinuria independent of lowering blood
pressure. Heterogeneity was analyzed beyond chance using the I2
statistic to describe the percentage of variability. For data with high
heterogeneity, sensitivity analysis was performed and based on the result we
also performed a subanalysis in which participants were grouped by type 1
diabetes (1DM) and type 2diabetes (2DM). A 2-sided p-value less than 0.05 was
considered statistically significant, and all statistical analyses were
performed using STATA, version 12.0 and Review Manager 5.1 software.
Results
Our original search yielded 5073 articles, 4779 citations were excluded based on
titles and abstracts. After a thorough and careful review, 13 trials which contained
1282 patients were included in our meta-analysis ([Fig. 1]).
Fig. 1 Process for identifying studies eligible for the
meta-analysis.
Of the contained 13 trials, 9 compared the efficacy of ACEIs versus placebo, 1
compared ARBs with placebo, and 3 studies compared ACEIs versus calcium channel
blocker (CCB) and placebo. There are twelve studies included patients with eGFR
greater than 60 ml/min/1.73 m2
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17] and eleven studies included patients with
microalbuminuria [6]
[7]
[8]
[9]
[11]
[12]
[13]
[14]
[15]
[16]
[17]. Eight studies enrolled patients with
type 1 DM [6]
[8]
[9]
[10]
[11]
[13]
[15]
[18], four studies enrolled patients with type
2 DM [7]
[12]
[14]
[16] and one study included patients with type
1 and type 2 DM [17]. Follow-up ranged from 1
to 6 years. The characteristics of the included studies are presented in [Table 1].
Table 1 Characteristics of studies in meta-analysis.
Studies
|
Intervention
|
Control group
|
Type of diabetes
|
Duration (years)
|
Total patients (n)
|
Mean age (years)
|
Men (%)
|
eGFR (ml⁄min/1.73 m2)
|
Scr (μmol/l)
|
AER (mg/d )
|
PRO (g/d)
|
Atlantis 2000
|
Ramipril 1.25/5 mg
|
placebo
|
1
|
2
|
134
|
40±12.0
|
70.9
|
104.3±26.1
|
_
|
107.2±65.8
|
_
|
Innovation 2008
|
Telmisartan 40/80 mg
|
placebo
|
2
|
1.3
|
163
|
60.76
|
77.9
|
96.1±27.9
|
70.4±17.8
|
ACR:168.45±46.64
|
_
|
Laffel 1995
|
Captopril 50 mg
|
placebo
|
1
|
2
|
143
|
32.7±8.6
|
70.6
|
>60
|
97.2±17.6
|
89.3±55.4
|
_
|
Bojestig 2001
|
Ramipril 1.25/5 mg
|
placebo
|
1
|
2
|
55
|
39.7±9.7
|
74.5
|
101.0±21.5
|
_
|
103
|
_
|
Crepaldi 1998
|
Lisinopril 10 mg
|
placebo or Nifedipine 10mg
|
1
|
3
|
66
|
37.5±10.4
|
67
|
114.3±18.8
|
87.0±13.5
|
116.2±65.1
|
_
|
Ahmad 1997
|
enalapril 10 mg
|
placebo
|
2
|
5
|
103
|
50.0±2.6
|
30
|
124±13.4
|
_
|
77.8±45.9
|
_
|
Ekstrand 1996
|
Captopril 50 mg
|
placebo
|
1
|
2
|
235
|
34.0±11.6
|
52.3
|
<60
|
185.6±52.9
|
_
|
1.8±2.4
|
Viberti 1994
|
Captopril 50 mg
|
placebo
|
1
|
2
|
92
|
33.7±10.3
|
55.4
|
130±29.9
|
64.5±10.6
|
55.6±8.2
|
_
|
Ravid 1993
|
Enalapril 10 mg
|
placebo
|
2
|
5
|
108
|
44.1±3.3
|
44.7
|
–
|
104.6±7.5
|
_
|
0.1±0.062
|
O’Donnell 1993
|
Lisinopril 10 mg
|
placebo
|
1and 2
|
1
|
32
|
48.7±14.4
|
71.8
|
121.4±41.6
|
_
|
83.6±61.4
|
_
|
MDNSG 2004
|
Perindopril 8 mg
|
Nifedipine 40 mg bid/placebo
|
1
|
6
|
77
|
52.8±2.6
|
63.6
|
92.4±7.8
|
79.4±4
|
58.7±2.9
|
_
|
MDNSG 2001
|
Perindopril 8 mg
|
Nifedipine 40 mg bid/placebo
|
2
|
2
|
42
|
30.8±4.7
|
51.5
|
103.2±6.3
|
84.9±5.8
|
63.9±3.3
|
–
|
Parving 1989
|
Captopril 25 mg
|
placebo
|
1
|
1
|
32
|
31
|
71.9
|
104.3±26.1
|
–
|
107.2±65.8
|
–
|
eGFR: Estimated glomerular filtration rate; Scr: Serum creatinine; AER:
Albumin excretion rate; PRO: Proteinuria.
Quantitative analysis
Trial quality was variable. Allocation concealment was adequate in 7 trials [6]
[7]
[12]
[13]
[14]
[15]
[16], inadequate in remaining 7 trials
[8]
[9]
[10]
[11]
[17]
[18]. Nine studies were double-blind [6]
[7]
[8]
[9]
[11]
[13]
[16]
[17]
[18], three were open [10]
[14]
[15], and only one was single-blind [12]. Twelve (85.7%) trials used an
intention to treat analysis. The summary of the risk of bias is presented in
[Table 2].
Table 2 Quality assessment for included trials (Modified Jadad
Score).
Studies
|
Randomization
|
Concealment of allocation
|
Double blinding
|
Withdraws and dropouts
|
Total
|
Atlantis 2000
|
2
|
1
|
1
|
1
|
5
|
Innovation 2008
|
1
|
0
|
2
|
1
|
4
|
Laffel 1995
|
1
|
0
|
1
|
1
|
3
|
Bojestig 2001
|
1
|
0
|
1
|
1
|
3
|
Crepald 1998
|
2
|
0
|
2
|
1
|
5
|
Ahmad 1997
|
1
|
1
|
2
|
1
|
5
|
Ekstrand 1996
|
1
|
0
|
2
|
1
|
4
|
Viberti 1994
|
1
|
0
|
2
|
1
|
4
|
Ravid 1993
|
2
|
0
|
2
|
0
|
4
|
O’Donnell 1993
|
1
|
0
|
2
|
1
|
4
|
MDNSG 2004
|
1
|
2
|
0
|
1
|
4
|
MDNSG 2001
|
1
|
2
|
0
|
1
|
4
|
Parving 1989
|
1
|
0
|
0
|
1
|
2
|
Decrease in albuminuria
Data regarding the effects of ACEI/ARB on decrease in albuminuria were
available from 7 trials [1]
[6]
[7]
[12]
[13]
[17]
[18], including 5 trials (n=664)
of ACEI compared with placebo or active control therapy, and one trial
(n=163) of ARBs compared with placebo. The average decrease in
albuminuria was 80.28 mg/d (95% CI,
–104.79 mg/d to –55.77 mg/d)
less in patients receiving ACEIs/ARBs than in placebo or active control
group patients (p<0.001) and the heterogeneity analysis showed
I2=97%, p<0.001 ([Fig. 2]).
Fig. 2 Effect of ACEIs or ARBs compared with placebo or other
active agents on albuminuria
According to the high heterogeneity, sensitivity analysis was conducted (Fig.
1S), and the result showed that the heterogeneity was mainly caused by
two studies [7]
[18]. Further study revealed the main
difference between them was the types of diabetes in participants. The
pathogenesis of type 1 and type 2 DM is different, so we conducted a subanalysis
according to type of diabetes in order to eliminate the effect of different
types of diabetes on the reduction of albuminuria in the treatment of patients
with ACEIs/ARBs. There were three studies reporting the decrease on
albuminuria in patients with type 1 and type 2 DM, respectively. Data
demonstrated the average decrease on albuminuria was 57.26 mg/d
(95% CI, –71.11 mg/d to
–43.40 mg/d) and 99.82 mg/d (95%
CI, –125.72 mg/d to
–73.92 mg/d), respectively in DKD patients with type 1DM
and 2DM, less in patients receiving ACEIs/ARBs than in placebo or active
control group patients. The results for subgroup differences showed that
decrease on albuminuria in DKD patients with type 2DM was more significantly
with ACEIs/ARBs (p=0.005, I²=87.6%;
[Fig. 3]).
Fig. 3 Sub-analysis for effect of ACEIs or ARBs compared with
placebo or other active agents on albuminuria.
Relationship of albuminuria and blood pressure
We also did analysis of the correlation between decreased levels of proteinuria
and blood pressure, and found there was no significant association between two
of them (R=–0.23, p=0.55). And meta-regression showed no
association between decrease level of albuminuria and systolic blood pressure at
baseline (p=0.323; [Fig. 4]).
Fig. 4 Meta-regression of decrease level of albuminuria on the
decrease level systolic blood pressure
Change of glomerular filtration rate
Five trials comparing ACEIs and placebo reported data on the change of glomerular
filtration rate, and the results showed no statistically significant reduction
in decline of GFR (MD 2.39 ml/min/1.73 m2,
95% CI –1.29 ml/min/1.73 m2 to
6.07 ml/min/1.73 m2; [Fig. 5]) [6]
[12]
[14]
[15]
[17] with no evidence of heterogeneity
(I2=49%, p=0.1). Only one study reported
three cases of ESKD in 17 patients with placebo and zero in 15 patients with
ACEIs treatment [10], and the data showed
no difference between two groups (OR 0.28, 95% CI 0.00 to 1524.23).
(Fig. 2S 3S)
Fig. 5 Effect of ACE-Is or ARBs compared with placebo on GFR.
Cardiovascular disease outcomes
Data in three studies [6]
[7]
[13] including 302 patients reported 32
cardiovascular disease events. Of the 172 patients treated with ACEIs there were
15 cardiovascular events (8.7%) and 17 events occurred in 130 patients
treated with placebo or active agents (13.1%). Overall, ACEIs and ARBs
therapy did not reduce cardiovascular events versus placebo or other
antihypertensive agents (OR 0.97, 95% CI 0.45 to 2.12) with no evidence
of heterogeneity (I2=0.0%, p=0.95; [Fig. 6]).
Fig. 6 Effect of ACE-Is or ARBs compared with placebo or other
active agents on cardiovascular disease outcomes.
Total mortality
Five studies reported 8 deaths in 335 patients with ACEIs treatment
(2.4%) and 3 deaths in 301 patients with placebo or active agents
therapy (1.0%) [6]
[9]
[10]
[11]
[18]. Overall, ACEI therapy did not reduce
total mortality in patients with normotensive diabetic kidney disease (OR 1.09,
0.16–7.20) with no evidence of heterogeneity
(I2=0%, p=1.0; [Fig. 7]).
Fig. 7 Effect of ACE-Is or ARBs compared with placebo or other
active agents on total mortality.
Adverse effects
There were eight trials (931 patients) reported at least 1 adverse event [7]
[8]
[9]
[11]
[12]
[16]
[17]
[18]. The data showed 33 adverse events
occurred in 446 patients with ACEIs treatment (7.4%) and 34 events in
550 patients with placebo (7.2%). Compared with control,
ACEIs/ARBs therapy did not clearly increase the risk of adverse effects
(OR 1.12, 95% CI 0.69–1.81; [Table 3]). Among all kinds of adverse effects, cough is the most
frequent occurrence, but the difference is not significant between two groups
(OR 1.19, 95% CI 0.66–2.12; [Table 3]), the same as hypotension and other adverse effects.
Table 3 Adverse events in the included RCTs.
Adverse events
|
Studies reporting
|
ACEIs/ARBs group n/n
|
Control group n/n
|
OR (95%CI)
|
p-Value
|
Total patients with adverse events
|
8
|
33/446
|
35/485
|
1.12 (0.69, 1.81)
|
0.64
|
Specific adverse events
|
|
|
|
|
|
Cough
|
7
|
26/392
|
21/376
|
1.19 (0.66, 2.12)
|
0.56
|
Hypotension
|
1
|
1/70
|
1/73
|
1.04 (0.07, 16.35)
|
0.97
|
Severe adverse
|
1
|
14/109
|
7/54
|
0.99 (0.42, 2.31)
|
0.98
|
OR: Odds ratio.
Risk of bias
The Funnel plots and Begg’s test applied to individual trials did not
disclose any publication bias (Begg's Test:
Pr>|z|=0.230; [Fig. 8]).
Fig. 8 Funnel plot of included studies in the analysis for effect
of ACEIs or ARBs compared with placebo or other active agents on
albuminuria.
Discussion
This meta-analysis including a total of thirteen studies with 1268 patients was
conducted to investigate the efficacy and safety of ACEIs/ARBs on renal and
cardiovascular outcomes in normotensive patients with DKD. There was an obvious
trend for a favorable effect for ACEIs/ARBs for decreasing of albuminuria, which was
independent of the degree of blood pressure drop and systolic blood pressure at
baseline. And the average level of decrease in albuminuria was more significantly in
DKD patients with 2DM. However, no significant difference was observed on the risk
of adverse effects. However, we did not find remarkable difference between the
ACEIs/ARBs and control groups regarding renal events, cardiovascular disease and
total mortality.
There are several previously published meta-analyses and systematic reviews assessing
the effects of RAAS blocking agents in patients with DKD, and found RAAS blocking
agents are able to significantly reduce albuminuria. However, little data are
available on the effect of early introduction of ACEIs/ARBs to normotensive patients
with DKD. The initial reduction in albuminuria induced by ACEIs/ARBs was again shown
in this study, which is consistent with previously published smaller trials of
ACEIs/ARBs in normotensive patients with DKD [6]
[7]
[12]
[13]
[16]
[17]
[18]. We found ACEIs/ARBs therapy
confers renal protective effects that are independent of changes in blood pressure
in normotensive patients with DKD. And in the post-hoc analysis of the INNOVATION
study [7], treatment with telmisartan not only
prevents the progression of microalbuminuria, but also reverts from microalbuminuria
to normoalbuminuria in Japanese normotensive type 2 diabetic patients. Early stages
of diabetes mellitus are characterized by increases in intracapsular pressure and
reduction in renal plasma flow despite the patients are normotensive [19]. Evidence from clinical trials and animal
experiments suggests that the effect of renin angiotensin system (RAS) inhibitors in
the kidney is to decrease efferent arteriolar resistance with resulting reduction of
intraglomerular capillary pressure [10]. In
addition, increased vascular biomarkers of ACE indicate a vasomotor disturbance in
this earlier stage of DKD (renal hyperfiltration). The glomerular hyperfiltration
would be ameliorated by RAS inhibitors therapy [20]
[21]. These findings provide a likely
explanation for the damaging effect of glomerular hypertension in the early stage of
DKD and for the beneficial effect of ACEIs/ARBs even in patients without
hypertension. We hypothesis that blood pressure drop may lead to glomerular
hypoperfusion in patients with lower initial systolic blood pressure. In the present
studies, adverse events were similar between two groups in terms of cough,
hypotension and severe adverse [7]
[8]
[9]
[11]
[12]
[16]
[17]
[18] which suggest that ACEIs/ARBs
treatment in early stage of DKD is safe and well tolerated.
The development of microalbuminuria in diabetes mellitus strongly predicts ESRD and
is associated with increased risk cardiovascular complications, as well as total
mortality. Next, we analyzed endpoints of the rate of decline in GFR, ESRD,
cardiovascular disease outcomes and total mortality, but found no significant
difference between two groups. The relation between ACEIs/ARBs and later
kidney events and cardiovascular disease outcomes in these patients has not been
established, possibly because of the inclusion of only early-stage diabetic
nephropathy and short follow-up periods. In all of contained studies, only one
reported three cases of ESRD in placebo group and none in ACEIs therapy [17]. The results demonstrate that ACEIs could
reduce the occurrence of ESRD, although the data showed no statistical significance.
It is not yet clear whether ACEIs/ARBs can permanently prevent the
deterioration of renal function. Further larger studies with longer follow-up times
are required to elucidate the cardiovascular and renal protective effects of
ACEIs/ARBs therapy in normotensive diabetic disease patients with incipient
nephropathy.
There are several limitations of our meta-analysis that are inherent to the studies
included. First, because of the lack of sufficient data, a subgroup analysis
exploring the impact of ACEIs/ARBs on proteinuria is not conducted. Second,
there is heterogeneity between different ACE inhibitors or ARBs, so different agents
might not have the same risk-benefit ratio in DKD patients with normotension. Third,
as most of the included RCTs were from developed western countries, there is a
scarcity of data from other countries, which has limited the possibility to
generalize the results.
Conclusion
This study demonstrated a reduction in albuminuria by RAAS blockade in normotensive
patients with DKD, especially with 2DM, and side effects did not differ among the
groups. More studies with longer follow-up times are required to elucidate the
cardiovascular and kidney protective effects of ACEIs/ARBs therapy in these
patients.