The kidney versus the heart

Thinking has changed on acute kidney injury during heart failure decompensation.


A conference speaker rarely feels compelled to fall on his sword, but Steven Coca, DO, MS, literally did just that at the Heart Failure Society of America's 2019 annual meeting, held in Philadelphia in September.

It was a foam sword, whipped out at the end of a debate over the significance of kidney biomarkers during heart failure decongestion. Dr. Coca, a nephrologist and associate professor at the Icahn School of Medicine at Mount Sinai in New York City, helped raise the alarm about kidney damage associated with heart failure treatment in the early 2000s. And now he wants to turn it off.

Image by Getty Images
Image by Getty Images

“This is the tension we see amongst the hospitalists and the housestaff, etc. Everybody's saying, ‘Oh my God, we can't diurese them, we'll worsen the kidney function.’ But, of course, they're congested and need diuresis,” Dr. Coca said. “I have to blame myself for part of this.”

He and other experts speaking at the meeting looked at the balance of cardiac and kidney function and reviewed existing evidence on the subject, extending from what Dr. Coca called his “old Kool-Aid–drinking analyses” of a decade ago to the latest data.

Kidney worries

Many researchers and clinicians got concerned about acute kidney injury (AKI) when it became a hot research topic at the turn of the century.

“Over the last 20 years, literature has emerged showing that renal dysfunction, either baseline or worsening kidney function, is strongly associated with adverse outcomes. This isn't one of these really small associations. It turns out that renal dysfunction is actually one of the strongest predictors of adverse outcomes,” said Jeffrey Testani, MD, MTR, associate professor and director of heart failure research at Yale School of Medicine in New Haven, Conn.

For example, there was Dr. Coca's meta-analysis published in the American Journal of Kidney Diseases in 2009. “A 0.3, 0.4 change in creatinine were associated with a risk of in-hospital mortality and mortality a year later, showing that these small changes and various indicators of AKI were associated with risk of long-term mortality,” Dr. Coca explained.

In response to findings like these, researchers developed new classifications for kidney injury, and the hunt for a “troponin of the kidney” began. In addition to studying the impact of creatinine changes, researchers looked at biomarkers in urine like neutrophil gelatinase-associated lipocalin, kidney injury molecule-1, interleukin-18 and others. “In non-heart failure settings, several of these urinary biomarkers provided additional prognostic information when added to changes in serum creatinine,” said Dr. Coca.

Clinicians, meanwhile, started to focus on what they could do to protect the kidneys, for example, reducing use of drugs such as renin-angiotensin-aldosterone system (RAAS) antagonists.

“The problem is that renal dysfunction is often at odds with life-saving therapy,” said Dr. Testani. “RAAS antagonists provide a very concrete and powerful improvement in outcomes in our patients, and we've really come to appreciate that congestion is also a major driver of adverse outcomes, probably participating in the heart failure disease progression process.”

This would seem like an impossible dilemma—whether to prioritize the heart or the kidneys—except that more recent evidence provided a fairly definitive answer. Researchers, including Dr. Testani, took a more detailed look at changes in kidney function, specifically in patients hospitalized for heart failure, and how they compared to short- and long-term outcomes.

“You start to diurese them and their creatinine improves: This should predict fantastic outcomes, right? Well, actually, whenever we've looked at this, we've surprisingly found that [the prognosis with] improving renal function is often as bad, if not worse, than worsening renal function,” he said.

It turns out that AKI (or worsening renal function, as some experts now prefer to call it) is not a predictor of bad outcomes in all patients.

“If you experience worsening renal function and you are not adequately diuresed, these are the patients that tend to die more. But if that worsening renal function comes at the hands of very aggressive decongestion, these patients do not suffer a mortality disadvantage,” explained Meredith A. Brisco-Bacik, MD, MSCE, associate professor and director of heart failure clinical research at the Lewis Katz School of Medicine at Temple University in Philadelphia.

For example, in a 2011 study by Dr. Testani and colleagues, patients with decompensated heart failure whose systolic blood pressure went down along with their renal function did not have increased risk of mortality. Those patients were also the ones who received higher doses of oral vasodilators and thiazide diuretics and lost more weight, according to results published in the European Journal of Heart Failure.

“This has been a pretty consistent theme: If you do something good for the patient and the creatinine goes in the wrong direction, it tends to be a good thing,” said Dr. Testani. That “something good” can be evidenced by blood pressure reductions or hemoconcentration, both of which have been associated with worsening renal function and better survival in heart failure research.

However, patients in the 2011 study whose renal function got worse without their heart failure improving did have worse outcomes, and that helps explain the initial data that led to so much concern about kidney injury. “The patients with AKI clearly are sicker at baseline. It's a marker of severity of disease,” Dr. Coca said. The initial data on AKI were “severely confounded” by this association, he now believes.

Those newer kidney biomarkers that initially showed promise as prognostic indicators had similar problems when examined in patients with acute decompensated heart failure. Patients who had elevated levels while being aggressively diuresed for heart failure actually had better six-month survival than those with more normal markers, according to a study published by Drs. Coca and Testani in the May 8, 2018, Circulation.

These findings showed that the new biomarkers are not going to help in heart failure care, Dr. Coca said. “The clinicians were blinded to these. If actually we could measure these urinary kidney injury biomarkers in real time and report them to clinicians, outcomes would be even worse. People would hold more diuretics, and I think they would do more harm than good.”

Whether they call it “permissive AKI” (Dr. Coca) or “benign hemodynamic elevation of creatinine” (Dr. Testani), the experts agreed that these changes in kidney function resulting from heart failure treatment are not the problem they were once thought to be. Dr. Brisco-Bacik summed up their current perspective in pop-culture lingo: “All we need to do is keep calm and diurese on.”

When diuresis isn't working

Of course, there are exceptions. Sometimes heart failure patients' kidney dysfunction is severe enough, or diuresis ineffective enough, that clinicians should still worry, and these cases were the focus of the second half of the conference session.

Dan Negoianu, MD, an associate professor of clinical medicine and director of ICU nephrology and inpatient dialysis at the Perelman School of Medicine at the University of Pennsylvania in Philadelphia, offered advice on how to identify such situations. “We need to ask, ‘Is decongestion successful?’” he said.

If the answer is no, the next step is to ask why, advised Dr. Brisco-Bacik. “Diuretic resistance is often actually a problem of not enough diuretic as opposed to true diuretic resistance,” she said. Clinicians may forget how steep the diuretic dose-response curve is. “You can't just double the dose of the diuretic. You have to maybe 10 times the dose,” she said.

But if the patient has received plenty of furosemide and still isn't losing fluid, “then you may actually be dealing with true diuretic resistance,” she said. “Compensatory distal tubular reabsorption is probably the primary driver of most diuretic resistance seen in heart failure. So when you give that loop diuretic and this allows more sodium to potentially be excreted into the urine, the kidney says, ‘That's not actually what I would like to do,’ and there's enhanced distal reabsorption.”

One possible solution is to add a thiazide, such as metolazone. Dr. Brisco-Bacik and colleagues looked at use of this strategy in patients with acute decompensated heart failure in a study published in the Sept. 18, 2018, Journal of the American Heart Association. “We analyzed 14,000 hospital patients over an entire hospital system, and metolazone was used in about 10% of admissions,” she said. “There really was no rhyme or reason in terms of how thiazides were being used.”

They did find a pattern between treatments and outcomes. “It turns out loop diuretics were not associated with increased mortality at a high dose, but if you received metolazone, this was a statistically significant association with your risk of mortality,” Dr. Brisco-Bacik said.

But what if patients received both? “In that small group of patients receiving a high dose of diuretics and metolazone, there actually is not an association with increased mortality, and in fact, there might even be a suggestion of benefit,” she said.

The evidence leaves the best treatment course still somewhat uncertain, she said. “When you're thinking about what to do with the difficult-to-diurese decompensated heart failure patient, given what we know about metolazone on average, in most comers, it's probably safer to actually give a higher dose of diuretics instead.”

If that still doesn't work, then you may have hit diuretic failure. “It's probably time for some form of renal replacement therapy in true diuretic failure,” said Dr. Brisco-Bacik.

But when you take that step, don't expect a happy response from your nephrologist. “There's a reason why you see the fear in our eyes when you call to say you want to dialyze this advanced heart failure patient,” said Dr. Negoianu. Among the many concerns with putting such sick patients on dialysis, he focused on the practical ones, describing the risks of the available dialysis methods.

Tunneled hemodialysis catheters are the simplest method, but “they always get infected,” Dr. Negoianu said. “Ask your heart failure patient if what they really need in their life is sepsis.”

The arteriovenous (AV) fistula is the gold standard but has its own heart failure-specific problems. “They cause a big AV shunt. This is no joke and a big decision to make in somebody whose heart failure is so bad that you're desperate enough to call me,” he said.

Finally, there's peritoneal dialysis, which is a good option, but only for patients who can handle the logistics of doing it at home or have caregivers to help.

Given all these risks, Dr. Negoianu offered some advice that differed from the other experts'. “There's so much evidence from cross-sectional data showing the benefit of RAAS blockade in patients with moderately worsening renal function. And I have drunk that Kool-Aid, but we may be at risk of going too far,” he said. “The majority of the data comes from the patients that bump their creatinine only by 0.3.”

Dr. Negoianu also discussed the problem of confounding in this cross-sectional data. The decisions about vasodilators for patients in these were all made by their individual physicians, who would often stop these medications in their sickest patients. The use of these medications may just be a marker for how sick a given patient is, rather than the cause, he explained.

For heart failure patients with more serious renal dysfunction—those for whom dialysis might be considered—there's little available evidence, and more vasodilator therapy may not be the best way to go, argued Dr. Negoianu.

Given that lower blood pressure is associated with decreased response to diuretics, there is risk that vasodilators could stand in the way of decongestion. Particularly in patients with preserved ejection fraction, in whom there is little evidence of benefit from vasodilator therapy in the acute setting, he recommends considering a reduction in vasodilators when diuresis isn't working and the kidneys are failing.

“We once were too willing to stop these medications when things were going badly. We now have data showing that patients who lose a little bit of kidney function actually do well if you continue these medications, and so the pendulum is swinging toward continuing them,” he said. “However, we have almost no data regarding patients who lose so much function that they approach the need for dialysis. We may be at risk of letting the pendulum swing too far if we continue these medications no matter what.”