CARDIOVASCULAR JOURNAL OF AFRICA • Vol 21, No 1, January/February 2010
4
AFRICA
diagnosis.
12
Clinical trials involving diabetic patients with either
normo- or microalbuminuria will require trials of 10 to 20 years’
duration to demonstrate benefits of doubling in serum creatinine
and ESRD, and by definition must rely on surrogate markers of
albuminuria, which fortunately in a diabetic patient is a funda-
mental marker for staging.
12
Meta-analysis is generally considered the highest form of
evidence. In 2004, Strippoli
et al
. published a systematic review
of ACEIs and ARBs on mortality and renal outcomes in diabetic
nephropathy.
14
ARBs and ACEIs showed a highly significant
36% relative risk (RR) reduction in hard renal endpoints of
doubling of serum creatinine and ESRD. In addition, on surro-
gate endpoints there was a 65% RR reduction in progression
from microalbuminuria to macroalbuminuria, and a 3.42 greater
chance of ARBs causing a regression of microalbuminuria. In
an updated meta-analysis in 2008, Sarafidis
et al
. confirmed
the results of the previous meta-analysis.
15
ARBs independently
confirm renal protection in type 2 diabetes, and this has been
unequivocally demonstrated in the IRMA2, IDNT and RENAAL
studies.
16-18
However, in a recent study by Mauer
et al
., inhibiting the RAS
with either enalapril or losartan compared to placebo did not
slow nephropathy but appeared to benefit retinopathy in normo-
tensive patients with type 1 diabetes.
7
This was a meticulously
conducted study that included renal biopsies pre- and post-treat-
ment. However, the study must be interpreted with caution. The
sample size was small (about 85 patients in each group), and at
baseline the mesangial fractional volume and albumin excretion
rates were well within the normal ranges.
19
Although there was
significant increase in patients developing microalbuminuria in
the losartan group, the mean increase in albumin excretion and
mesangial fractional volume in this group was exceptionally
small (4
µ
g/min and 0.01, respectively) and the mean values
remained well within the normal range.
Non-diabetic CKD
Similar to diabetic nephropathy, proteinuria is an excellent
surrogate marker of increased risk of CKD and ESRD and may
contribute directly to progression.
11
In 2001, a meta-analysis
by Jafar
et al
. demonstrated that ACEIs reduced proteinuria by
0.46 g/day and ESRD by 31%.
20
Benefit was seen in patients
with greater urinary protein at baseline, but it was doubtful if
benefit extended to patients with urinary protein below 0.5 g/day.
However, as it is likely that patients with less proteinuria had a
slower rate of progression of renal disease, it becomes increas-
ingly difficult to show benefits on hard renal endpoints.
In 2008, Kunz
et al
. conducted a meta-analysis of combina-
tion therapy with ACEIs and ARBs in non-diabetic CKD and
showed that it further reduced proteinuria by 24 to 25% over
monotherapy,
21
but safety concerns remain. The issue is further
clouded by the authenticity of the COOPERATE study,
22
which
was purported to show that combination therapy with losartan
and trandolopril showed superior renal protection over individual
monotherapy.
23
Hypertension
Although hypertension, especially systolic BP, is strongly associ-
ated with CKD,
24
there is no evidence that BP lowering in non-
diabetic hypertensives without malignant hypertension is associ-
ated with renal protection. In a meta-analysis of 10 randomised,
controlled trials, treated hypertensives did not have a lower risk
of renal dysfunction.
25
On the basis of these results, should we
assume that BP will have no benefit on renal outcomes? Again, if
we understand that progression of CKD in non-malignant hyper-
tension is slow, of the order of 2 to 4 ml/min/year, it would take a
trial of 10 to 20 years to show benefits of antihypertensive therapy
on CKD, which is far beyond the expectations of a clinical trial.
There is conflicting evidence to support the contention that
ACEIs have renoprotective effects in patients with hypertension.
In the ALLHAT study, there was no difference between lisinopril
and chlorthalidone, which is the entirely expected result as the
majority of patients had normal GFR at baseline.
26
However,
in the AASK study, ramipril appeared to be better at slowing
the rate of decline in GFR compared to amlodipine in African-
American patients with established hypertensive nephroscle-
rosis, especially those with a urinary protein/creatinine ratio
above 0.22.
27
Patients in this study had established hypertensive
nephrosclerosis with a mean baseline serum creatinine that was
twice normal, making renal endpoints more realisable. In a study
by Wendy Hoy on Australian Aborigines, who have a very high
prevalence of CKD, aggressive antihypertensive therapy, using
the ACE inhibitor perindopril as a cornerstone, showed a 47%
reduction in renal death.
28
Patients at high cardiovascular risk
The HOPE study did not show benefits of ramipril compared to
placebo on hard renal endpoints in patients at high cardiovascular
risk. In diabetics there was less progression of albuminuria in the
ramipril arm.
29
Of note, most patients in this study had normal
renal function with a mean creatinine level of 93
µ
mol/l.
In the ON-TARGET study, progression of microalbuminu-
ria was reduced by telmisartan and ramipril and combination
therapy but there was a slightly greater reduction in GFR with
the combination treatment.
30
These differences in GFR over the
entire clinical trial time period were very small (–6.11, –4.12
and –2.82 ml/min in the combination, telmisartan and ramipril
arms, respectively, over a median of 56 months), and are in the
bounds of age-related decline and therefore of questionable clini-
cal significance. This result was also entirely predictable based
on physiology. More intense inhibition of the RAS will lower
GC pressure to a greater degree and reduce GFR and microalbu-
minuria.
Combination therapy also resulted in more cases of acute renal
failure, probably due to over-treatment of BP and the effects of
RAS blockade in patients with unrecognised renal artery stenosis
(the estimated prevalence was about 10% in this study).
Cardiac failure
RAS inhibitors are overwhelmingly beneficial for patients with
cardiac failure with reduced ejection fraction. Renal endpoints
are included for safety reasons and not to assess effects on the
progression of CKD. In patients with severe cardiac failure, renal
function may be critically dependent on renal auotoregulation
mediated by angiotensin II, especially if patients are receiving
high doses of loop diuretics. Therefore, it is anticipated that some
patients receiving RAS inhibitors in cardiac failure may suffer
an acute deterioration in renal function and have a doubling in
serum creatinine due to acute pre-renal failure, which is gener-
