Better Renal Outcomes with CCB over Diuretic in ACCOMPLISH, but Implications Debated

Hogan, Michelle

doi: 10.1097/01.NEP.0000370090.62355.21

Initial antihypertensive therapy with benazepril plus amlodipine led to fewer renal outcomes than did benazepril plus hydrochlorothiazide, reported a prespecified secondary analysis of the Avoiding Cardiovascular Events through Combination Therapy in Patients Living with Systolic Hypertension (ACCOMPLISH) trial published early online by the Lancet.

ACCOMPLISH was a randomized, double-blind trial of patients with hypertension and a high risk for cardiovascular events. It was stopped early, at a mean follow-up of 2.9 years, because benazepril, an angiotensin-converting enzyme (ACE) inhibitor, plus amlodipine, a calcium-channel blocker (CCB), better reduced cardiovascular events. The trial was funded by Novartis.

While the trial was praised for comparing prespecified renal outcomes between these two regimens, questions were raised over whether the difference in outcomes was caused by short-term hemodynamic changes and not long-term improvement in kidney structure.

“This is a secondary analysis of a major trial that challenges the conventional wisdom that hydrochlorothiazide in combination with ACE inhibition is superior to the ACE inhibitor combination with a calcium-channel blocker,” said Rajiv Agarwal, MD, Professor of Medicine at Indiana University, who was not involved in the study.

“The results now extend to the renal outcome. I don't think there are any questions about the data.

“The question is why did this happen, and I think the editorial brings this out nicely,” Dr. Agarwal said, referring to the editorial that accompanied the analysis.

The renal endpoint in the trial was a composite of doubling of serum creatinine or end-stage renal disease, but it was driven by doubling of serum creatinine, noted the editorial, which was written by Hiddo Lambers Heerspink, PharmD, PhD, and Dick de Zeeuw, MD, PhD, of the Department of Clinical Pharmacology at University Medical Center Groningen in Groningen, the Netherlands. There were few cases of progression to ESRD in the trial overall, as it was not done in a population of patients with advanced chronic kidney disease (CKD).

Amlodipine induces a short-term rise in GFR, and hydrochlorothiazide a short-term fall, certainly when given with a renin-angiotensin-aldosterone system inhibitor, the editorial noted. Hydrochlorothiazide is a thiazide diuretic.

“If you look at the acute change, there's a clear difference between the drugs,” Dr. Agarwal said. “The amlodipine-based regimen increases the estimated GFR; the hydrochlorothiazide-based regimen reduces the GFR. And that's where the story finishes. The long-term decline in GFR is identical in the two groups.”

In a chart based on data from ACCOMPLISH that was presented by Dr. Bakris and colleagues at a conference, Drs. Heerspink and de Zeeuw show a similar long-term slope between the two treatment groups following the different short-term effects.

Lead author George Bakris, MD, Professor of Medicine and Director of the Hypertension Center at the University of Chicago Pritzker School of Medicine, disagreed with the assessment that short-term hemodynamic changes were behind the result.

“First of all orthostatic pressures were checked, and orthostatic hypotension was not more pronounced in the group receiving diuretics,” Dr. Bakris said in a phone interview. “That would have been a hint for a volume depletion related to the diuretics that would have affected the change in creatinine.

“Second of all creatinines would not have doubled because of this. Usually the changes that you see in creatinines related to diuretics and ACE inhibitors being given together is a rise of 30, 40, even 50%, I'll give you 50 or 60%, but you're not going to see a doubling.

“And lastly all the other trials that have been done, including RENAAL [Reduction of Endpoints in Non-Insulin-Dependent Diabetes Mellitus with the Angiotensin II Antagonist Losartan Study], including IDNT [Irbesartan in Diabetic Nephropathy Trial], all had diuretic add-ons to the ARB [angiotensin receptor blocker]. So why is it acceptable for those trials to have that endpoint and not this one?”

The point that hemodynamic changes could be behind the difference in the renal outcome is of interest in a shorter term study, Dr. Bakris said.

“Any change in GFR slope or hemodynamic changes related to CCBs are short-lived and exist only for about 18 months. After 18 months the slope is actually much steeper in the CCB group,” he said, citing data from the African-American Study of Kidney Disease and Hypertension (AASK).

Back to Top | Article Outline

Study Design

Included in ACCOMPLISH were 11,506 patients in the United States, Sweden, Norway, Denmark, and Finland randomized between October 2003 and May 2005 in a one-to-one ratio to receive 20 mg of benazepril plus 5 mg of amlodipine (n=5,744) or 20 mg of benazepril plus 12.5 mg of hydrochlorothiazide (n=5,762).

Of the patients with baseline data for creatinine concentration, 1,093 had chronic kidney disease. And of these patients with CKD, 81.4% had normoalbuminuria or microalbuminuria.

Drug doses were force-titrated to reach the recommended blood pressure goals of less than 140/90 mmHg for patients without diabetes or chronic kidney disease, or less than 130/80 mmHg for patients with diabetes or CKD.

After dose adjustment, mean blood pressure was 131.6/73.3 mmHg in the benazepril-amlodipine group and 132.5/ 74.4 mmHg in the benazepril-hydrochlorothiazide group.

Back to Top | Article Outline

Renal Outcomes

There were 113 (2.0%) occurrences of the renal endpoint in the benazepril-plus-amlodipine group compared with 215 (3.7%) in the benazepril-plus-hydrochlorothiazide group.

Of these events, there were 105 (1.83%) instances of doubling of serum creatinine and 7 (0.12%) dialysis events in the benazepril-plus-amlodipine group, and 208 (3.61%) instances of doubling of serum creatinine and 13 (0.23%) dialysis events in the benazepril-plus-hydrochlorothiazide group.

Among the 7,640 patients age 65 or older, there was a trend towards a difference in dialysis outcomes between the two groups, as Dr. Bakris noted, but that trend did not reach statistical significance. In this patient population, there were 3 (0.8%) dialysis events in those on benazepril plus amlodipine and 10 (0.26%) in those on benazepril plus hydrochlorothiazide.

Back to Top | Article Outline

Diabetic Nephropathy

More than half of the patients with CKD in each treatment group had diabetic nephropathy. For these patients, there was no difference in the renal endpoint between the two arms. There were 16 (4.8%) events in the benazepril-plus-amlodipine group and 17 (5.5%) in the benazepril-plus-hydrochlorothiazide group.

“Maybe if you had 3,000 patients there would be a small difference, but my guess is they largely got the same treatment because I would venture to say that most of the people with diabetes and CKD were also getting diuretic therapy,” said Katherine R. Tuttle, MD, Professor of Medicine at the University of Washington School of Medicine and Medical and Scientific Director of Providence Medical Research Center. Dr. Tuttle was one of the ACCOMPLISH site investigators but not a member of the steering committee.

“They were getting loop diuretics, not hydrochlorothizide, but in this context I doubt it makes much difference. We allowed diuretics to be given to the ACE-calcium channel blocker group, and you're comparing them with a group getting ACE-thiazide diuretics—we wouldn't expect a lot of difference. The people in the ACE-thiazide diuretic group also got antihypertensives, although they didn't allow other calcium channel blockers.

“Another practical point here, especially in diabetic kidney disease, is there's a primacy of blood pressure control, so the most important thing we can do in those patients, and this is a strong clinical message, is control blood pressure to goal, and then the different classes of medicine are like icing on the cake.”

The lack of a difference in the renal disease endpoint among the patients with diabetic nephropathy could have been due to the lower levels of proteinuria in ACCOMPLISH relative to other trials, Dr. Bakris said.

Speaking about the general trial population, so not just the group with diabetic nephropathy, Dr. Bakris said 10% or 15% of patients received the loop diuretic furosemide.

“Number one, do I think that had an effect? I think the answer's no because we actually asked a statistician that question and I think his argument was too few patients overall got it so therefore it probably wasn't playing a role, and it's not that more people got it in one group than the other.

“The second issue is if you're going to make the argument that diuretics are going to cause volume depletion, well those people had edema—they got it regardless.”

Back to Top | Article Outline

Blood Pressure, Albuminuria

There was a difference in blood pressure between the two groups, Dr. Tuttle noted.

“Admittedly in absolute numbers it's really small—in the clinic no one's going to be excited about a 1-mL difference—but when you're looking at large numbers of people over a number of years, and these are repeated analyses, you look at it in an integrated fashion. I still don't think you can completely exclude the possibility that one group might have had a little bit better blood-pressure control.”

Dr. Tuttle also noted the effects of the regimens on albuminuria reduction, which ran counter to their effects on the renal endpoint.

In the 446 (76.2%) of patients with baseline albuminuria more than 33.9 mg/mmol who were assessed at the final visit, there was a reduction in urine albumin-to-creatinine ratio (UACR) from baseline in the benazepril-plus-hydrochlorothiazide group, with a median change of 63.8%, compared with a median reduction of 29.0% in the benazepril-plus-amlodipine group.

In the 409 (97.1%) of 421 patients with CKD who were assessed at the final visit, there also was a reduction in UACR from baseline in the benazepril-plus-hydrochlorothiazide group, with a median change of 26.8%, and an increase in the benazepril-plus-amlodipine group, with a median change of 2.9%.

“This study along now with a growing body of literature shows a discordance between albuminuria reduction and clinical outcomes,” Dr. Tuttle said.

“It's either the outcome measurement isn't accurate, like the editorialists are suggesting, or I think there's a big element of truth to the fact that albuminuria in all settings, including this one, may not be a very good marker of clinical outcomes.”

Despite the caveats she mentioned, Dr. Tuttle said the trial makes a contribution, providing the first comparison of kidney outcomes with the two regimens as initial therapy.

“I think the ACCOMPLISH trial overall is an exemplary clinical trial. It is rigorously performed, it represented a very constructive collaboration between industry and clinical investigators that I think was appropriate and as unbiased as these kinds of studies can be, so I think the study itself is really a model of how to do a clinical trial and work constructively between academia and industry.

“The other thing that I think they're doing that I admire greatly is learning as much as they can from the data they collected on these patients.”

Back to Top | Article Outline

More than Two Drugs

The clinical implications of ACCOMPLISH are limited, Dr. Tuttle said.

“I don't think that many people are going to be making treatment decisions based on this study because most patients in this category are going to get three drugs anyway, and it would have been nice if they actually showed us the number of drugs and the distribution of drugs in the group because most people were on more than two drugs for blood-pressure control.

“I think it maybe tells us that in the minority of people who you can treat with two drugs then using the calcium channel blocker-ACE combination isn't worse than giving the ACE-diuretic therapy, but I'm not sure it's better. That's important because all the guidelines say ACE-diuretic first.”

The study may have implications for cases where time is of the essence, Dr. Agarwal said.

“I think that one of the things that we can learn from the trial is if your patient is nearing dialysis and you still need to buy time for the patient to be away from dialysis without compromising blood-pressure control, then it might be reasonable for the patient to switch from a diuretic-based regimen to a dihydropyridine calcium channel blocker-based regimen. It'll buy you some GFR.

“Second is you probably don't want to start a diuretic-based regimen at the late stage of kidney disease when you're waiting for those crucial few months before the access matures and the patient goes on dialysis.”

Back to Top | Article Outline

Setting the Stage

The results of a single trial shouldn't change clinical practice, but further research is warranted, Dr. Bakris said.

“These results set the stage for a second trial. This clearly sends the message that diuretics are not the end-all, be-all and that this needs to be much more carefully looked at, especially since the benefit in this group was seen. The trial was stopped early because of overwhelming cardiovascular benefit.”

Dr. Agarwal echoed the need for further research of these regimens.

“I think it's an exciting idea. What they need to do is an outcomes-driven trial in people who truly have kidney disease. I would suggest that this trial be conducted in people with say Stage 4 kidney disease or at least Stage 3b kidney disease to look at true renal endpoints.”

Dr. Tuttle agreed that such a trial would have to be conducted in patients likely to have hard renal outcomes, like dialysis, kidney transplantation, and death.

“That trial will probably never be done because as a practical matter I'm not sure that the question is that critical to patient management, and I think it would require such a huge sample size and such long duration of therapy I don't think anyone would be too interested in doing it,” she said.

“I'm really glad they published this paper because we want to try to learn everything we can from these huge investments—not just dollars but the effort and all that the patients gave us by letting us learn from them.

“We want to query the data and learn as much as we can, but a lot of those things are not going to be conclusive. They'll be suggestive; they'll still help build a knowledge base, but they won't necessarily set the world on fire.”

© 2010 Lippincott Williams & Wilkins, Inc.