The Classics: IDNT

September 2001 was a strong month for Nephrology in NEJM. RENAAL, IDNT and the IRMA 2 were all published in the same issue that month.

IDNT, the Irbesartan Diabetic Nephropathy Trial compared an ARB to a CCB and placebo in DMII patients with CKD and proteinuria.

1715 patients were randomized from multiple centers around the world.

To be included in the study patients needed to be between 30-70 years old, have DMII and HTN (greater than 135/85), proteinuria (at least 900mg/24 hours), Cr 1.0-3.0 mg/dl in women and 1.2-3.0 mg/dl in men.

The study drug Irbesartan was started at 75mg and titrated to 300mg while Amlodipine was started at 2.5mg and titrated to 10mg per day.

The primary endpoint was doubling of serum Cr, ESRD (dialysis, transplantation or a Cr of 6.0 mg/dl) or death.

The secondary endpoint was a composite of death from cardiovascular causes, nonfatal MI, CHF resulting in hospitalization, a permanent neurologic deficit caused by a cerebrovascular event, or lower limb amputation above the ankle.

Mean followup time was 2.6 years.

Apart from there being slightly more men in the placebo group baseline characteristics were similar.


The ACRs were higher than in RENAAL with median values of 1.9 g/24hrs. An interesting little tidbit was that 24 hr urinary albumin was roughly 2/3rds that of total 24 hr urinary protein (did not know that).


Crs were around 1.7 mg/dl lower than the 1.9 mg/dl in RENAAL. Using demographic data MDRD eGFRs were approximately 44 ml/min/1.73m2 for men.

Blood pressures were also higher in IDNT than RENAAL with rough mean values of 160/87 vs 152/82 respectively.

During treatment the MAP was significantly higher (3.3 mmHg) in the placebo group when compared to the ARB and CCB groups (which were similar in MAP control).

The primary composite endpoint showed a statistically significant ARR of 6.4% and 8.5% and RR reduction of 20% and 23% when ARB was compared to placebo and amlodipine respectively. According to the authors adjustment for difference in BP did not alter these results.


The breakdown of the composite was as follows...

ARB reduced the risk of doubling of serum Cr when compared with both CCB and placebo. There were no statistically significant differences in ESRD or death from any cause between the groups.

Again impressive was overall just how at risk diabetics with CKD and significant proteinuria are with 22% doubling their Cr, 17% reaching ESRD and 15% dying over just 2.6 years.


The secondary outcome did not differ significantly between groups.

Proteinuria was on average reduced by 33% in the ARB group and the rate of decline in eGFR was 5.5 ml/min/1.73m2 vs 6.8 ml/min/1.73m2 and 6.5 ml/min/1.73m2 in the CCB and placebo groups respectively.

So nutshell...

Big international muticenter ARB vs CCB vs placebo trial with intention to treat analysis of diabetic nephropathy with macroalbuminuria showing reduction in the ARB group in the combined endpoint of mortality, ESRD or doubling of serum Cr driven by the reduction in doubling serum Cr.

Very similar to RENAAL with the added information that ARB is superior to amlodipine in this population.

Digging Deeper: RENAAL Reexamined

So I stumbled across this KI paper while watching a great lecture on ukidney given by Phil Mcfarlane at last year's Prevention in Renal Disease Conference in Toronto.

It's a really nice post hoc analysis of the RENAAL data looking at albuminuria's predictive value for events (ESRD, doubling of serum Cr, death).

See the previous post for what RENAAL was all about, patients with DMII and nephropathy who were randomized to ARB or placebo.

Here, de Zeeuw and colleagues asked...

1) Whether baseline albuminuria was predictive of events
2) Whether the degree of reduction was predictive of long term risk reduction for these events and finally
3) Whether the amount of albuminuria remaining after therapy conveyed risk similar to baseline levels.

1) Baseline albuminuria.


No shocker here, the more albumin you had in your urine the more likely you were to reach either the renal end point (ESRD, doubling of serum Cr, death) or ESRD. HRs are multivariate adjusted and calculated using an ACR less than 1.5 g/g as the referent.

What really struck me was that if you had a ACR of greater than 3.0 g/g, almost 100% of folks reached the renal end point at four years and nearly 80% got to ESRD.

2) % reduction of albuminuria at 6 months.

Exciting stuff here and very elegant figures. In terms of reduction in the composite end point you see a drop in risk the greater the reduction in albuminuria.


Around a 60% reduction in risk for those who achieved a drop of 60% or more in their albuminuria at 6 months.

When looking at ESRD in isolation the risk of progression vs protection as predicted by albuminuria is more marked.


Those with greater than a 40% increases in their albuminuria were upwards of two and a half more times more likely to need dialysis or transplantation when compared to those whose values were unchanged.

Likewise those who achieved the most dramatic reductions saw their risk drop by roughly 75% compared to unchanged values.

3) Residual albuminuria.

This is fascinating.


The curves of risk for a renal event at baseline and after six months based on the amount of albuminuria are virtually superimposable with lower levels associated with fewer events.

Interestingly when you subdivide the degree of albuminuria based on whether someone was on ARB or placebo and look at risk both at baseline and at six months you see that at six months for a given amount of albuminuria those on ARB fared no better than those on placebo.


The authors suggest that these findings taken together point towards albuminuria as a specific target for renoprotective therapy and that consideration should be given to further reducing albuminuria in patients on ARBs who continue to have high levels.

I love the idea that we can potentially do more for RENAAL style patients by driving albuminuria down as far as possible. To further lower proteinuria in a patient already on a ARB one could 1) use high dose ARB 2) add an ACEI 3) Add an aldosterone blocker 4) add a renin inhibitor or 5) add a nondihydropyridine CCB.

This is not without some controversy so with that in mind the ONTARGET trial will be next under the spotlight.

The Classics: RENAAL

Great name, RENAAL the Reduction of Endpoints in NIDDM with the Angiotensin II Antagonist Losartan study.

Lots of stuff packed into that name: industry sponsorship (Merck, the makers of Losartan), the hypothesis (ATII blockade would reduce endpoints) and the nephrologic bent of the study (two of the primary composite end point components were renal outcomes, doubling of serum Cr and reaching ESRD).

The trial was international with multiple participating centers. 1513 patients were randomized.

To be included you needed to be between 31 and 70 years old and have DMII with diabetic nephropathy defined as macroalbuminuria (spot urine albumin/Cr of >300 in mg/g) and a Cr between 1.3-3.0 mg/dl (higher Cr's were required for men of at least 60 kg's).

Exclusion criteria were DMI, non-diabetic kidney disease, an MI or CABG one month prior to enrollment, CVA or PTCA with 6 months, TIA within one year or any history of CHF.

Mean followup time was meant to be 4.5 years after the final enrolled patient but the study ended early (mean followup 3.4 years) after data emerged that RAAS blockade with ACE inhibitors provided cardiovascular protection to patients with CKD including those with DMII.

Baseline characteristics were similar between groups. A couple things did stand out to me on this reading that I hadn't appreciated before...


1) Patients were quite proteinuric. This was by no means a trial of DMII patients with microalbuminuria.

2) Most patients were CKD III. If you take the mean serum Cr's of 1.9 mg/dl in the two groups and plug them into the MDRD most participants (non-African American) using the mean age of 60 had mean eGFRs of around 39 ml/min/1.73m2.

Interestingly two standard deviations below yields a serum Cr of 0.9 mg/dl which would seem to fall outside the inclusion criteria. I suppose older women with values around this could fall into CKD III by MDRD and perhaps this is what drove it. I'll have to look back at the study description paper to sort this out.

Patients received antihypertensive regimens that included either Losartan started at 50mg and increased to 100mg qday or placebo. The breakdown by classes of medications used in each group is shown below...


You can see that slightly more of all the classes other than ARBs (and ACEs which were excluded) were used in the placebo group. Makes sense as at the end of the study there was no significant difference between the two groups in terms of BP (though there was a small trough time averaged difference in BP between the ARB and placebo group which did not significantly alter outcomes).

The primary outcome was the composite of doubling of serum Cr, ESRD (defined as need for prolonged dialysis or transplant) or death (defined as failure to call back Barry Brenner).

The secondary outcomes included a cardiovascular composite (MI, stroke, first hospitalization for angina or CHF, coronary or peripheral revascularization or death from any of the above), progression of renal disease and change in proteinuria.

Discontinuation of ARB and placebo due rise in serum Cr occurred in 1.5% and 1.2% of patients respectively. Similarly discontinuation due to hyperkalemia occurred in 1.1% and 0.5% of patients respectively. (Amazing how low these are in studies.)

Overall though, 53.5% and 46.5% of patients in the placebo and ARB groups stopped their study drug. Given this it's amazing how large effects remained in the intention to treat analysis.


The primary outcome occurred in 47.1% of the placebo group and 43.5% in the ARB group for an ARR of 3.6% and a RRR 16%. This was driven by the two renal endpoints as mortality was equal between the two group with 1 in 5 patients (!) in both groups passing on. The effect was larger when looking at those who actually stayed on rx with the RRR growing to 22%.

There was no difference in the cardiovascular composite with roughly 1/3rd of both groups experiencing one of these outcomes though there were significantly less 1st hospitalizations for heart failure in the ARB group.

The authors speculate that this may have been due to the relatively low cardiovascular risk produced by exclusion criteria which is a little tough to see when 1/3rd of patients are having an event in just 3.4 years of followup.

There was a significant reduction in the rate of GFR decline in the ARB when compared with placebo with the groups losing 4.4 ml/min/1.73m2 and 5.2 ml/min/1.73m2 per year respectively.

According to the discussion the above translates to roughly a 2 year delay to the time to need for dialysis or transplantation.

Proteinuria was also reduced by 35% in the ARB group.

So nutshell...

Big international muticenter ARB vs placebo trial with intention to treat analysis of diabetic nephropathy with macroalbuminuria showing reduction in a combined endpoint of mortality, ESRD or doubling of serum Cr driven by the two renal outcomes.