Precious Bodily Fluids

At Renal Week I went to Mark Perazella’s talk on “Urine Microscopy for Diagnosis and Prognosis in Hospital-acquired AKI.” Perazella published the interesting paper in CJASN in 2008 on using urinalysis microscopy for prognosis. His talk focused on the same subject. The gist is that a trained observer of urine microscopy is not only more accurate at diagnosing the etiology than FENa and FEurea but can provide prognostic information. Additionally, urine microscopy performs as well as the novel biomarkers: KIM-1, NGAL and IL-18.

As part of his talk Perazella threw-up some slides with some surprising statistics on the diagnostic utility of FEUrea and FENa:

From Pepin AJKD 2007: 50:566-573

A sensitivity of 48% without diuretics? A specificity of 33% with diuretics? Terrible. A second study showed similarly terrible numbers:

From Darmon et al. Critical Care 2011: 15 R178.

Only a bit better than a coin toss.

I felt shamed. I had been teaching and using urinary acute renal failure indices for 20 years and to see that they sucked sucked, kinda hurt. But then as I thought about it, I remembered seeing primary data that was quite a bit more compelling. So I went back to the primary literature.

The fractional excretion of sodium was invented in 1976 by Carlos Hugo Espinel. His initial series was published in JAMA. In that study he found 100% sensitivity and specificity in 17 oliguric patients.

Holy shit! Shrier is a member of the Indiana Basketball Hall of Fame!

The study that is usually referenced as the mother of the fractional excretion of sodium comes from Mr. Sodium himself, Robert Schrier (fun fact: the founding member of my practice, Joseph Beals, was college roomates with Robert Schrier). This group found excellent performance of FENa in 85 patients:

Dr. Kohn is in the middle

FEUrea was invented by Kaplan and Kohn in 1992 (Fun fact: Dr. Kohn was one of my professors at University of Chicago) but the technique was not validated until 2002 when Cavounis et al prospectively looked at 102 patients with AKI. I remember reviewing the Cavounis article in journal club with Dr. Kohn in the room.

Results for using FENa to diagnose pre-renal disease (<1%)

• 92% sensitivity without diuretics
• 48% sensitivity with diuretics
• 76% sensitivity for the entire cohort

Results for FEUrea to diagnose pre-renal disease (<35%)

• 90% sensitivity without diuretics
• 88% sensitivity with diuretics
• 89% sensitivity for the entire cohort

Kim et al looked at 107 patients with acute kidney injury. Instead of pre-renal versus ATN they used transient versus persistent. They also looked at the performance of the AKI indices in the context of diuretics. (disclosure, I just read the abstract)

Results for using FENa to diagnose pre-renal disease (<1%)

• 96% sensitivity 100% specificity without diuretics
• 63% sensitivity 98% specificity with diuretics

Results for FEUrea to diagnose pre-renal disease (<30%)

• 92% sensitivity 87% specificity without diuretics
• 96% sensitivity 83% specificity with diuretics

Diskin et al looked at 100 oliguric patients and evaluated FENa and FEUrea. They used 40% for the line with FEUrea and 1% for FENa.

• FENa 54% sensitivity
• FEUrea 95% sensitivity

One perspective on the proposed changes to how the distribute kidneys fb.me/100dAwkxi
— Karol Franks (@okarol) November 12, 2012

Then I saw almost the exact same message was on my Facebook feed from The Kidney Group:

Twitter’s most followed Kidney Practice: the kidney group should be ashamed of this post.

Both posts point to the same article in US News and World Report about the proposed change to the transplant guidelines and describes the change as unethical.

To understand what is driving the desire to change the current system and to understand how the proposed system is superior, one needs to understand the shortcomings of the current way organs are distributed.

First, unlike hearts or livers, people can live for years on dialysis waiting for a transplant. This is why we have 75,000 people on the kidney transplant list. Tragically, only 16,843 were transplanted in 2010.

Look at the 15,000 people listed for the second or third time (USRDS)

Currently the most common reason for a kidney transplant to fail is death with a functioning graft. On the one hand, death with a functioning graft is an ideal outcome, this is how the vast majority of people without kidney disease die, i.e. they die with a kidney still capable of supporting life. But in the world of kidney transplant, death with a functioning graft is a measure of inefficiency. We live in a world with a vast undersupply of organs and not only are there not enough organs to go around, but an increasingly common reason for patients to be listed for a transplant is re-listing after their first transplant failed. If all of these patients had been matched to a kidney that better lined up with their life-expectency they would not be getting back in, an already overcrowded, line.

Over time total graft failure and return to dialysis/transplant has fallen but death with a functional graft has remained flat. Better graft matching could mean more people living with transplants. (USRDS)

Another problem with the current list is how wait-time is calculated. Currently time on the list starts from the time a patient is approved for a transplant. This seems fair until you see who is not getting referrals to transplant. There is data that shows that patients dialyzed in commercial dialysis facilities (the vast-majority of dialysis patients) are less likely to get referred for transplant. Other studies have pointed to minorities and patients without commercial insurance being less likely to be referred for transplant.

Michigan no longer uses that system for calculating wait-time. In Michigan, the clock starts the first day of dialysis. So a dialysis patient that takes three years to finally find his way to a transplant center and get listed would get credit for three years of wait time as soon as he was approved for transplant. This largely neutralizes any discrimination he would have suffered from being randomized to a bone-head nephrologist or living in a racist medical system. This is a smart change and it is used by the proposed transplant allocation system.

The current system is over 25 years old and suffers from limited medical knowledge available at the time it was created. The current system looks at factors that affect graft survival (HLA matching) but ignores factors associated with recipient survival (diabetes, age, etc).

We grade patients antibody levels (PRA) on a continuous scale from 1 to 100. Patients with higher levels have more difficulty finding matches. In order to help these patients with additional burdens in finding a match they are awarded points. Unfortunately the current system awards a fixed value if the level is over 80 and does not appreciate the continuous nature of this factor. The proposed system treats this as the continuous value it is.

Currently, there are a lot of discarded kidneys that could be used to benefit patients but the current system ranks kidneys as Standard Criteria or Extended Criteria. A system that recognized that not all recipients need 20 years of renal function would allow more of these marginal kidneys to be used.

It is interesting that Laine Ross’s article does not address any of the weaknesses of the current system. The reader is left with the impression that the current system is fine except for a vague mention of “inefficiencies.” She does not elaborate on what inefficiencies means but, make no mistake, it means we can get more people living with a transplant by changing how we allocate the organs. The new system plans on doing that, so if you agree with Laine you are arguing for less people living with a transplant.

The centerpiece of the proposed system is a system that ranks both recipients and the kidneys. The top 20% of candidates (by estimated years of life remaining) are matched to the top 20% of kidneys. Getting young people high quality kidneys that will last decades makes sense to me but apparently ethicists from the University of Chicago feel that putting 20 year old kidneys in 68 year old diabetics is a good use of a scarce resource. By giving young people long lasting kidneys we can avoid these patients losing their kidneys and getting re-listed. This will help reduce the total wait list.

The rank list also works on the bottom end of the scale, acceptable but marginal kidneys at the bottom of the scale, the worst 15%, are made available to patients on a voluntary basis. Older patients and patients who are failing dialysis can be desperate for a kidney. This is intended to expand the project that began with extended criteria donors and reduce the number of viable kidneys that are wasted.

Outside of the top 20% and the bottom 15% the system does not change. This new allocation is really just a revision to update and fix some weaknesses with the current system.

Ross calls out the 20% rule because it can not magically determine the exact time of death of everyone in the queue for a transplant. Seems like a pretty high bar to cross. The system that is proposed has a c-statistic of 0.72, which is where she get’s her:

…their model will get it right about 75 percent of the time.

The system is just as accurate as the system used to determine life expectancy for liver transplants and does a very good job of preventing the primary problem of putting kidneys with a short life expectancy in patients with a long life expectancy. The system does not need to be perfect to be better than the current system which is totally blind to this important issue. Interestingly, the Kidney Committee, looked at a better system to predict life-expectancy but it was so complex that people were suspicious. They settled on the current system because it is simple and does a good enough job. I agree that transparency is an important quality in any system that will be distributing kidneys.

Reading the whole article it looks like a hit-peice trying to score ethical points while lives hangs in the balance. This is the third attempt to revise the transplant allocation program. Ethicists have repeatedly deep six-ed previous attempts and while we try to find the perfect system we are trapped in a system that everyone agrees has significant ethical lapses and wastes both kidneys and years of transplant viability.

Update: The Kidney Group has removed the link to the article from their Facebook page. @okarol has blocked me on Twitter. No peep from U.S. News and World Report. This post made it to the RenalWEB homepage.

I have posted on the release of the EVOLVE data at Kidney Week 2012 at the eAJKD blog.

This is a big deal and when I read the @NEJM tweet it makes me mad.

Cinacalcet doesn’t lower death or major CV event risks in patients undergoing dialysis. nej.md/U4yeSv #kidneywk12
— NEJM (@NEJM) November 3, 2012

That tweet provides only the first sentence in what should be an important and longer discussion of the findings of EVOLVE. To completely discount EVOLVE as @fish2phil and @Nephroboy do…

@kidney_boy these are statistical machinations attempting to squeeze a diamond out of coal.
— Homer W. Smith (@fish2phil) November 4, 2012

It’s cack. “@kidney_boy: My first take on EVOLVE: biggest, longest, RCT ever undertaken in dialysis population ajkdblog.org/2012/11/04/kid…“
— Nephroboy (@nephroboy) November 4, 2012

…leaves the nephrologist in a bit of a bind. If we ignore studies that don’t meet their primary endpoint, what are we left with? Here is a list of ALL the randomized controlled trials in dialysis that use hard outcomes. This list ignores any study where the end-point is an improvement in a biochemical parameter, echographic finding, or vital sign:

1. NCDS from 1981: Positive. We learned that small molecule clearance is better than large molecule clearance
2. Normalization of hematocrit: Negative. No advantage and nearly significant disadvantage to normalizing the hemoglobin
3. HEMO: Negative. Definitively negative trial showing that small molecule clearance can not get any better with thrice weekly hemodialysis
4. ADAMEX Negative. Peritoneal dialysis version of HEMO, same result
5. AURORA Negative. Statin trial with rosuvastatin
6. 4D Negative. Statin trial with atorvastatin
7. SHARP Positive, kind of. Statin plus ezetimibe was able to reduce CV endpoints however they lumped dialysis and CKD patients together. The dialysis patients alone were not powered to find a difference and they did not. I think most nephrologists, in the face of 4D and AURORA are pretty skeptical of this data approach.
8. DCOR Negative. Trial of sevelamer versus calcium containing binders.
9. IDEAL Negative. Trial of early versus late initiation of dialysis. No advantage for early start. Some may argue that this is a CKD rather than dialysis study.
10. EVOLVE. Negative.

I bet if the skeptical nephrologist went to his dialysis patients and explaned to them:

USRDS 2012 Annual Data Report

• Dialysis patients have a 3 year survival of 50%.
• There is a drug that is already approved (i.e. not experimental) that recently has been shown to have an 89% chance of being able to reduce mortality by 17%

That every one of his patients would beg for the drug.

Final thoughts: The first time I heard about intention to treat was during the HIV epidemic. Some of drugs were so toxic, especially in the early days of HAART when patients were taking handfuls of pills, that significant number of patients weren’t able to tolerate them. I remember an ID doc expressing frustration that he needed to know if the drug would save the lives of the patients he had that could tolerate the drug. If they can’t tolerate it, he was going to stop the drug, but if they could tolerate it he needed to know if it slowed the virus. In the middle of the epidemic, there is no room for statistical purity.

Like that ID doctor, I want to know if the drug will work if the patient actually swallows the pill, and the answer to that is a definitive “Yes.” Hazard ratio for death of 0.83 P=0.009.

Disclosures: I was a principle investigator in EVOLVE, that means my practice was paid to recruit, and administer a bitch of a study that lasted longer than it was supposed to. It was a lot of work, a lot of meetings, a lot of signatures, a lot of responsibility, and not so much money. My name will not be among the authors.

My name on page 53 on the supplement

What do you do when your tentpole molecule fails it’s phase three clinical trial? I guess you don’t even build your booth at Renal Week.

At Kidney Week 2010 in Denver, Reata presented the results of BEAM, the phase II trial of bardoxolone as part of the late breaking clinical trial data. They shocked the world by showing increased GFR in diabetic nephropathy. The data was published in the NEJM in July 2011. But even before publishing they began work on the phase III trial, BEACON.

I can attest to the excitement that BEACON generated. I was not an investigator but I did enroll one patient in the trial and even had friends calling me trying to get loved ones in the trial. I outlined my experience with bardoxolone in the diabetic nephropathy talk that I gave to the Michigan State Medical Society (Keynote | PDF). On October 18th, Reata terminated BEACON due to “excess serious adverse events and mortality.”

Unfortunately, October 18th was only 2 weeks before the Start of KidneyWeek 2012, the premier nephrology conference in the U.S. Reata had already signed on with the American Society of Nephrology to be a Diamond Level Sponsor:

Not only were they a Diamond Sponsor, but they had bought top billing on the Abstract Book with a full page ad on the back cover:

The ad implores you join them at booth 1529. I went by there just to see how the company would try to spin this disaster. But there was no booth 1529. The map shows it between Amgen and Otsuka, but there is just a pharmacy and an EMR company in its place. It’s like the opposite of when Obi-Wan goes to Yoda trying to find the Planet Kamino.

Lost a Planet Master Obi-Wan Has

Poor Reata. Evidence-based medicine can be such a bitch.

I’d show pictures, but taking pictures of the booths is forbidden by ASN policy. Whisky Tango Foxtrot

Internists use normal saline.
Surgeons use lactated ringers.

Its a cultural difference, perpetuated by dogma.

Here’s how Burton Rose characterized Ringer’s solution in his classic Clinical Physiology of Acid-Base and Electrolyte disorders:

During the control phase, physicians were able to prescribe IV fluids per personal preference. During the experimental period, chloride rich solutions were restricted to specific clinical conditions: hyponatremia, traumatic brain injury and cerebral edema. Otherwise patients were given low chloride solutions: Hartman’s solution, plasma-Lye 148 and 4% albumin.

The primary outcome was the change in creatinine and incidence of AKI, using RIFLE criteria. Secondary outcomes included need for acute dialysis, length of of ICU and hospital stay and survival.

RESULTS

The study consisted of 1644 admissions to the ICU, 760 in the control period and 773 in the experimental period. The two cohorts were well matched with significant differences only with metabolic diagnosis being more common in the control period (7 vs 4.4%) and neurologic disorders being more common in the experimental period (6.2 vs 8.8%).

The difference in the fluids being used was dramatic and this resulted in significant differences in electrolyte exposure:

For the third year the nephrology blogging and tweeting community is going to get together to tip some back. This year we are meeting at the Barley Mash on Friday at 8:30. We had a lot of fun last year and its fun to meet the people you read and tweet in real life. If you find that you have an hour to spare, stop by.

600 5th Ave, San Diego, CA 92101

I am just back from Med 2.0. It was a great conference that did so many things right. One of the things that was right, was that every talk, every poster, every rapid fire session has its own web page. Here is what the top of those pages look like:

The square in the upper right, if you’r are not familiar, is a QR-code. All posters are supposed to use this label and all presentations are supposed to display this graphic on the title slide. When you scan the code with a smart phone (you can find an excellent free, QR-scanner for the iPhone here, life hacker endorsement of said scanner here), the smart phone will open up the web page for the talk or poster. Cool idea. Unfortunately, it was poorly executed by most presenters. Here is a typical title slide:

Using my iPhone from my seat in the middle of the auditorium, I had no chance of getting an accurate scan. Since I am one of those annoying conference participants who photographs every slide, I have found myself scanning my MacBook screen to get the links. Absurd, and really no easier than using the conference website to search for the presentation page. The right way to do it is the way John Ainsworth did it in his presentation on using Drupal to create a low tech, no software, SMS-notification system across 11 countries, 6 languages and 3 time zones:

More than a few times this week, I have found myself prescribing Jewish Penicillin, chicken noodle soup. CNS is rich in sodium so it is just about the best way to prescribe volume repletion in the out-patent arena. So i was delighted to see one of the dietician students projects on display in the hospital:

1720 mg of sodium per can of Swanson 100% Natural Chicken Broth (two 240 mL servings per can). So how does that stack up against old faithful, normal saline?

Normal saline has 154 mmol of sodium per liter or 3,542 mg of sodium per liter (154 mEq x 23 mg per mmol, the molecular weight of sodium), so a can of chicken soup is equivalent to about half a bag of normal saline.

Update: I don’t get many good comments on Blogger but I got this pitch perfect comment on Twitter:

@kidney_boy Nice. Didn’t fully appreciate that a bag of saline was 3.5gm Na! Makes me laugh thinking of the 2gm Na diet pts getting IVF!
— Aaron Logan (@pyknosis) September 15, 2012