George L. Bakris, MD: Good day. I am Dr. George Bakris, Professor of Medicine and Director of the American Society of Hypertension (ASH) Comprehensive Hypertension Center at the University of Chicago. Today I am speaking with Dr. Rajiv Agarwal, who is Professor of Medicine at Indiana University in the section of Nephrology.
Rajiv, as many of you know, has published several seminal studies recently on topics that enlighten and improve clinical practice. Today we are going to be talking about a paper that was recently published in the American Journal of Nephrology and focused on the role of thiazide-like diuretics in people with stage 4 and even stage 5 chronic kidney disease (CKD).
The whole concept, historically, has been that once the glomerular filtration rate (GFR) falls below 40 mL/min/1.73 m -- some people would say below 50 mL/min/1.73 m -- the patient needs loop diuretics because thiazide diuretics are not very effective. As we know, there are thiazide and thiazide-like diuretics, and the most recent JNC 8 (Joint National Committee) hypertension guidelines,the NICE guidelinesin the United Kingdom, and the European guidelines all recommend thiazide-like diuretics -- chlorthalidone or indapamide -- rather than hydrochlorothiazide as the preferred diuretics, in general, for blood pressure management.
But for stage 4 CKD we essentially have no guidance and no comparative studies. In this area, no meaningful studies have been done and certainly nothing using ambulatory blood pressure monitoring. Rajiv's research has set the stage for those questions to be answered.
Good morning, Rajiv. Tell us about why you did the study and what is actually involved.
Designing the Study
Rajiv Agarwal, MD: George, thank you for inviting me to discuss this. I conducted this study because I wanted to generate preliminary data to support a randomized controlled trial in this area. Before you design any study, you have to look at what exists in the literature, and as you correctly pointed out, the quality of data in people with CKD is not great. Although many of the guidelines have recommended the use of thiazides in this population, they are on the borderline about using thiazides in people with more advanced kidney disease. Many of the guidelines are truly silent on this because the quality of the data is meager.
We began by reviewing 7 studies of thiazides in patients with CKD. We found that there may be some evidence that these drugs may work, even for people with advanced kidney disease. Next, we decided to use ambulatory blood pressure monitoring to figure out how these drugs can improve blood pressure.
We were ambitious initially. We decided we would include people with resistant hypertension, and after we had optimized the antihypertensive medication (using a diuretic plus 2 other drugs) and truly ruled in resistant hypertension and confirmed that with ambulatory monitoring, then we would give these patients open-label chlorthalidone. As it turns out, we had an incredibly difficult time finding people with resistant hypertension. We then decided to include people with uncontrolled or poorly controlled hypertension, and we were more successful in finding those people.
During the first 2 weeks of the trial, we standardized the regimen of medications. The treatment regimens included an angiotensin-converting enzyme (ACE) inhibitor, a diuretic, a beta-blocker, or a calcium channel blocker. Most people were on 2 or 3 drugs. After standardizing the regimen, we confirmed ongoing hypertension with 24-hour ambulatory blood pressure monitoring and then gave the participants chlorthalidone 25 mg once a day. After 1 month, if the patient's blood pressure -- not the ambulatory blood pressure but home blood pressure -- was still above 130/80 mm Hg, we titrated the chlorthalidone to 50 mg. If the patient was still poorly controlled, we took the dose up to 100 mg. So this was a forced titration study; if the blood pressure was not below 120/80 mm Hg, we could increase the medication dose. At the end of 12 weeks we repeated ambulatory blood pressure monitoring.
Surprising Findings
We found an astounding improvement in blood pressure. The greatest improvement, as measured by home blood pressure monitoring, occurred simply by adjusting the doses of medications during the first 2 weeks. But we also saw a quite remarkable improvement on top of a median of 2.7, 2.8 drugs, with a fairly good magnitude of blood pressure drop using ambulatory blood pressure monitoring.
We then looked at the mechanism of this improvement by measuring the renin, aldosterone, and brain natriuretic peptide (BNP) levels, and we found that the renin initially goes up and then later falls to more normal levels. We found that the BNP declined, even from a normal range to a lower level, and then it returned towards normal. These are the classic hemodynamic changes that you saw in studies performed with thiazides about 30 or 40 years ago.
We also performed a direct measurement of the body volume. We placed participants into an air displacement plethysmography machine called a BOD POD, which enabled us to measure the total body volume, much as we can measure the total body weight. The scale is sensitive to 1 g; participants had to wear swimsuits when we made the measurements, and when we put them in the BOD POD we found that, lo and behold, they had lost about 2 L of volume. We can measure the pitting pedal edema, but with the BOD POD we can measure the total body volume. We found a loss of about 2 L within 4 weeks, which was remarkable. Even with only 14 patients the P value was very small, which means that it was a highly reproducible improvement in body volume.
Thus, in most patients, we saw an improvement in volume, stimulation of renin, and drop in blood pressure, and this can be detected rather nicely using home blood pressure monitoring and confirmed with ambulatory blood pressure monitoring. This outcome exceeded our expectations -- that it would work so well, even in people with advanced kidney disease.
Two of our 14 participants had stage 3B kidney disease. Participants had to have at least 3B kidney disease to get into the study, but most had stage 4 kidney disease.
What About Hypokalemia?
Dr. Bakris: Most clinicians are afraid of hypokalemia when they use diuretics. You used fairly significant doses; admittedly these are people with GFRs below 30 mL/min/1.73 m, so potassium issues, if anything, should be a problem the other way. But was hypokalemia an issue?
Dr. Agarwal: Absolutely. These drugs are dangerous, and I think the physicians who treat patients ought to be very cautious when using these drugs. Even though these patients had advanced renal failure, hypokalemia was a big issue, and sometimes we had to supplement with huge doses of oral potassium just to keep potassium levels stable. Having kidney disease does not mean that you will not be hypokalemic. In fact, that is a big problem.
Dr. Bakris: How many of these people met the criteria for the definition of obesity?
Dr. Agarwal: I do not recall exactly how many, but I would say at least 80% of them were obese. However, this depends on whether we classify obesity on the basis of body mass index. We were using the BOD POD, and in that case, more than 25% body fat is considered obese. Using that definition, I do not recall anyone in the study who was not obese.
Dr. Bakris: That is important for the next point I am going to make. There is a lot of data, largely from David Calhoun's group and others, showing that use of spironolactone in low doses not only protects against potassium loss but potentiates the antihypertensive effect of chlorthalidone. I know you did not use spironolactone in this study, but could you comment about that as a way to avoid the need for large doses of potassium and at the same time to potentiate any antihypertensive effects at even lower doses of chlorthalidone?
Dr. Agarwal: You make a good point, George. If I were to do the study all over again, I would not start with 25 mg of chlorthalidone. I would probably start with 12.5 mg. I was thinking that, in people who have advanced renal failure, 12.5 mg might not work. But these patients still had adverse effects of these medications, including hypokalemia, hyponatremia, hyperuricemia, etc., and even increased, transient changes in creatinine, which suggests volume depletion. All of these are things that you recognize as pharmacodynamic effects of the drugs and that tell you the drug is working.
I believe 12.5 mg would be a better way to go, and combining with spironolactone would be a fantastic idea, especially if you are not hyperkalemic. If you are hyperkalemic, use this to treat hyperkalemia and continue to use the ACE inhibitor, but if the potassium level is on the low side, then instead of giving potassium supplements, add a low-dose spironolactone to a low-dose chlorthalidone and voilà -- you have a great drug here.
Two Final issues: Gout and Creatinine
Dr. Bakris: Right. Very good. I want to ask about 2 final issues. These people are prone to increases in creatinine and increases in uric acid. Gout precipitation is a major problem in this subgroup of patients. Did you have that problem? I know you looked at a small group, but did you have that problem? And if so, how did you deal with it?
Dr. Agarwal: In the protocol, we prespecified what side effects would be of great interest in this group, and every 4 weeks we prospectively looked at potassium, sodium, and uric acid. I can tell you that almost no one did not have an increase in uric acid, a decline in potassium, or a decline in sodium. You can almost use it as a test to assure that the patient is taking the drug. They always had a couple-milliequivalents drop in sodium and a slight increase in uric acid. It is just a matter of time before you see gout. Those are typical side effects that you would see with thiazide diuretic use, and the patients need to be carefully monitored.
It struck me that, in one way or the other, there were a lot of side effects even though they were biochemical and lab abnormalities. These told me that the drug was working (perhaps a bit too well), and if I did this study again, perhaps we'd need to back off on the dose.
Dr. Bakris: That is a very good point. Let me remind the audience that chlorthalidone was used a lot in the 1970s and '80s and it was indicted as killing people because of hypokalemia, but in those days they were using 100 mg of chlorthalidone. This reinforces your point.
Finally, would you clarify for the audience the whole issue about change in serum creatinine? Those of us who see these types of patients know that this is not only a volume issue but it can be related to dramatic improvements in blood pressure and a transient hemodynamic effect.
What advice would you give the primary care physician, or even the specialist, who is not that facile with this information as to managing these patients?
Dr. Agarwal: Changes in the creatinine levels need to be interpreted in light of the symptoms that the patient might experience. If you have an increase in creatinine and the patient is dizzy and complaining that "when I stand up, I am ready to fall," these are patients who typically are older and sicker, and I would back off the drug. But if there is simply an increase in creatinine, a 30% decline in estimated GFR (eGFR) is telling you that the drug is working. Remember that this drug is being added on top of ACE inhibitors or angiotensin receptor blockers, and if those classes of drugs were not well tolerated, then we always have a beta-blocker on board. So we have A or B always on board and then we are adding these drugs; therefore, an increase in creatinine is an expected side effect. It is almost like the drug is working, and I would not stop the drug just because creatinine goes up.
It should not be interpreted as acute kidney injury (AKI) because some might say, "Oh, my goodness, we have AKI. Let's stop the drug." This is not AKI. If the blood pressure improves, you would expect an increase in creatinine and you have to live with it. It is only a matter of time that creatinine will trend towards baseline once the volume homeostasis is restored.
Dr. Bakris: Very good. Rajiv, thank you very much for sharing your work with our audience. Have a good day, and thank you for listening.
Dr. Agarwal: Thank you very much, George, for inviting me to discuss this study with you.