NSAIDs and Chronic kidney disease. A great post at the Renal Fellow Network

One of the standard pieces of advice I give patients regarding chronic kidney disease is to avoid NSAIDs. However, not infrequently, patients have co-morbidities that demand NSAIDs. This usually triggers a conversation with my patients where I describe how ibuprofen can cause acute renal failure (I just took care of a new patient who developed RIFLE stage: Failure from a couple of doses of Mobic on top of stepped up ibuprofen use). I then explain that we extrapolate from the acute renal failure that NSAIDs are probably not beneficial in CKD and are likely harmful.

Lisa J Cohen at The Renal Fellow Network has a nice post on the lack of hard data implicating NSAIDs in the progression of CKD:

What about chronic renal dysfunction following long-term NSAID intake? In today’s medical environment, the evidence is weak. Prospective cohort studies in the Physicians’ Health Study (Rexrode et al, JAMA 2001) and the Nurses’ Health Study (Curhan et al, Arch Int Med 2004) failed to show an association between even high levels of cumulative lifetime NSAID intake and decrease in renal function.

My concern about these studies is the effect, we physicians have on the outcomes. Telling kidney patients to avoid ibuprofen and other NSAIDs is standard fare in CKD care. I’m sure these patients tend to use less ibuprofen and more acetaminophen, just as liver patients probably do the opposite. So educated patients with CKD will avoid NSAIDs but regardless of NSAID intake they will have a much higher progression to kidney related endpoints than their peers without a diagnosis of CKD. The epidemiologist sees a large cohort of acetaminophen users (my CKD patients) ending up with renal failure and sees that people with CKD use very little ibuprofen and may infere that acetaminophen causes kidney failure and ibuprofen is protective.

These are the type of questions that CRIC should be able to answer.

A different view of the recession

One of my friends is an executive at a breast implant company. He told me that the recession obliterated the market for implants but now they are back.

2007 was the best year ever for breast implants
2008 was going great but the last two quarters fell apart so that the year was down slightly from 2007
2009 was a disaster with sales off 40% from ’08.

In the first quarter of 2010 sales are back up to near 2007 levels. So according to the breast implant index the recession is over!

– Posted using BlogPress from my iPhone

More ADPKD and sirolimus data: More definitive; less encouraging

Just a few weeks ago I was writing about the first patient data on the use of sirolimus in ADPKD. After years of being teased with promising animal data Perico et al. finally showed that sirolimus may help humans with ADPKD. That study was the first human data and so I forgave the fact that it was small, used a cross-over rather placebo design and was short in duration. And lastly it failed in its primary end point of total kidney volume but it showed less cyst volume (P=0.0558) and an increase in renal parenchymal volume (P=0.0089).

I had no idea that the next human data would be so quick in coming. Last week, the New England Journal of Medicine published two randomized, open label, controlled trials on the use of mTORs for ADPKD. Unfortunately, neither of them were encouraging.

The first article looked at everolimus, a newer mTOR inhibitor. It is by Walz et al. They looked at 392 patients and randomized to everolimus or placebo (double-blind) and maintained for 2 years. Patients had to have ADPKD and a GFR between 30 and 89 mL (Stage 2 and 3 CKD) or a GFR >89 and a kidney volume of 1,000 mL. Everolimus was titrated to keep trough levels between 3-8 ng/mL.

The primary end point was a change in kidney volume by MRI. The data showed decreased kidney volume with everolimus at 12 months but the advantage was no longer significant at 24 months.

Estimated GFR did not improve with everolimus:

Our linear regression model predicted a steep annual decline in the estimated GFR among patients receiving everolimus, owing to the significantly accelerated deterioration in renal function between months 6 and 18. However, the estimated GFR did not differ significantly between the everolimus group and the placebo group at 2 years.

The study also found a host of serious adverse events among the patients on everolimus. The ones found significantly more often with everolimus were:

  • Any
  • anemia
  • leukopenia
  • thrombocytopenia
  • stomatitis
  • diarrhea
  • folliculitis
  • hyperlipidemia
  • hypercholesterolemia
  • acne
  • angioedema
  • arthralgia
  • myalgia
  • ovarian cyst
  • epistaxis
  • peripheral edema

I was interested to see angioedema on the list as we had recently gone over a paper suggesting increased angioedema in transplant patients where sirolimus was implicated. During the discussion I had mentioned that this side effect would be important in the ADPKD trials. Walz et al. found a 5.6% rate of angioedema.

The second trial was by Serra et al. and was an 18 month, open-label, randomized controlled trial of 100 patients. They used sirolimus rather than everolimus. The study focused on patients with earlier disease, requiring a eGFR greater than 70 for enrollment. Patients had a 6 month run-in after enrollemnt where they needed to show 2% increase3 in kidney volume inorder to be randomized.

The results were not impressive:

And in their own words:

We found no significant difference in total kidney volume after 18 months of treatment with sirolimus, regardless of the patient’s age, sex, or albumin:creatinine ratio at randomization or whether they were receiving therapy with an ACE inhibitor or an ARB. Our estimate of the ratio of kidney volume in the sirolimus group to that in the control group rules out any clinically meaningful reduction in total kidney volume with the use of sirolimus.

I thought the accompanying editorial by Watnick and Germino was excellent. One of the primary points of the editorial was that this human, interventional data calls into question the use of change in kidney volume as being the ideal intermediate end-point.

second lecture of the year: acute kidney injury

This is a significant upgrade from the version I posted a couple of years ago. I put the lecture together right before the ATN trial was published. I finally got around to updating the presentation to include that data. I also updated the NGAL section and added some data on avoiding volume overload.

I used a number of the posts on the blog to allow me to rapidly update the presentation. I was pleased with how well my ATN commentary/review stood up.

“Hey, is that asystole?” from the introduction of The Checklist Manifesto

Its the first of July so that has me thinking about medical errors. In that vein here is an anecdote from Atul Gawande’s excellent Checklist Manifesto:

He told me about another patient, who was undergoing an operation to remove a cancer of his stomach when his heart suddenly stopped. John remembered looking up at the cardiac monitor and saying to the anesthesiologist, “Hey, is that asystole?” Asystole is total cessation of heart function. It looks like a flat line on the monitor, as if the monitor is not even hooked up to the patient.The anesthesiologist said, ‘A lead must have fallen off,” because it seemed impossible to believe that the patient’s heart had stopped. The man was in his late forties and had been perfectly healthy. The tumor was found almost by chance. He had gone to see his physician about something else, a cough perhaps, and mentioned he’d been having some heartburn, too. Well, not heartburn exactly. He felt like food sometimes got stuck in his esophagus and wouldn’t go down and that gave him heartburn. The doctor ordered an imaging test that required him to swallow a milky barium drink while standing in front of an X-ray machine. And there on the images it was: a fleshy mouse-size mass, near the top of the stomach, intermittently pressing up against the entrance like a stopper. It had been caught early. There were no signs of spread. The only known cure was surgery in this case a total gastrectomy, meaning removal of his entire stomach, a major four- hour undertaking.
The team members were halfway through the procedure. The cancer was out. There’d been no problems whatsoever. They were getting ready to reconstruct the patient’s digestive tract when the monitor went flat-line. It took them about five seconds to figure out that a lead had not fallen off. The anesthesi¬ologist could feel no pulse in the patient’s carotid artery His heart had stopped.
John tore the sterile drapes off the patient and started doing chest compressions, the patient’s intestines bulging in and out of his open abdomen with each push. A nurse called a Code Blue. 
John paused here in telling the story and asked me to suppose I was in his situation. “So, now, what would you do?” 
I tried to think it through. The asystole happened in the midst of major surgery. Therefore, massive blood loss would be at the top of my list. I would open fluids wide, I said, and look for bleeding. 
That’s what the anesthesiologist said, too. But John had the patient’s abdomen completely open. There was no bleeding, and he told the anesthesiologist so. 
“He couldn’t believe it,” John said. “He kept saying, “There must be massive bleeding! There must be massive bleeding!” But there was none. 
Lack of oxygen was also a possibility. I said I’d put the oxygen at 100 percent and check the airway. I’d also draw blood and send it for stat laboratory tests to rule out unusual abnormalities. 
John said they thought of that, too. The airway was fine. And as for the lab tests, they would take at least twenty minutes to get results, by which point it would be too late.
Could it be a collapsed lung—a pneumothorax? There were no signs of it. They listened with a stethoscope and heard good air movement on both sides of the chest. 
The cause therefore had to be a pulmonary embolism, I said—a blood clot must have traveled to the patient’s heart and plugged off his circulation. It’s rare, but patients with cancer undergoing major surgery are at risk, and if it happens there’s not much that can be done. One could give a bolus of epinephrine—adrenalin—to try to jump-start the heart, but it wouldn’t likely do much good. 
John said that his team had come to the same conclusion. After fifteen minutes of pumping up and down on the patient’s chest, the line on the screen still flat as death, the situation seemed hopeless. Among those who arrived to help, however, was a senior anesthesiologist who had been in the room when the patient was being put to sleep. When he left, nothing seemed remotely off-kilter. He kept thinking to himself, someone must have done something wrong. 
He asked the anesthesiologist in the room if he had done anything different in the fifteen minutes before the cardiac arrest. 
No. Wait. Yes. The patient had had a low potassium level on routine labs that were sent during the first part of the case, when all otherwise seemed fine, and the anesthesiologist had given him a dose of potassium to correct it. 
I was chagrined at having missed this possibility An abnormal level of potassium is a classic cause of asystole. It’s mentioned in every textbook. I couldn’t believe I overlooked it. Severely low potassium levels can stop the heart, in which case a corrective dose of potassium is the remedy. And too much potassium can stop the heart, as well—that’s how states execute prisoners. 
The senior anesthesiologist asked to see the potassium bag that had been hanging. Someone fished it out of the trash and that was when they figured it out. The anesthesiologist had used the wrong concentration of potassium, a concentration one hundred times higher than he’d intended. He had, in other words, given the patient a lethal overdose of potassium.
After so much time, it wasn’t clear whether the patient could be revived. It might well have been too late. But from that point on, they did everything they were supposed to do. They gave injections of insulin and glucose to lower the toxic potassium level. Knowing that the medications would take a good fifteen minutes to kick in—way too long—they also gave intravenous calcium and inhaled doses of a drug called albuterol, which act more quickly. The potassium levels dropped rapidly. And the patient’s heartbeat did indeed come back.
The surgical team was so shaken they weren’t sure they could finish the operation. They’d not only nearly killed the man but also failed to recognize how. They did finish the procedure, though. John went out and told the family what had happened. He and the patient were lucky. The man recovered—almost as if the whole episode had never occurred.