Saturday, December 13, 2014

SIADH and lasix

I remember a time when I thought the treatment of chronic SIADH was going to be revolutionized by the vaptans. These small molecular ADH antagonists would interrupt the disease the precise mechanism of disease. I expected a Banting and Best like revolution. (If you have not seen the story of the discovery of insulin take a moment to watch the movie, Glory Enough for All, especially if you thought the greatest thing to come out of Canada was Tim Horton's)


The initial data was promising with convincing studies on conivaptan and tolvaptan, but something happened on the way to SIADH nirvana.

First the EVEREST trial went sideways. In heart failure:
  • Angiotensin 2 is elevated and blocking it prolongs life
  • The sympathetic nervous system is up-regulated and blocking it prolongs life
  • Aldosterone is elevated and blocking it prolongs life
  • ADH is elevated and blocking it doesn't do a damn thing

With no hope for a heart failure indication the drug was marketed solely as a treatment for hyponatremia where it was shown to be effective. The pitch was that doctors should not discharge people with hyponatremia and tolvaptan was faster and more effective than the previous standard of care. The drug was priced for short-term inpatient use at $300 a pill tolvaptan was a non-starter for chronic outpatient SIADH.
So generous of Otsuka to make the 30 mg dose the same price as the 15 mg

But I held out hope, I felt that as soon as the FDA licensed tolvaptan for ADPKD, the drug would be re-priced for chronic use and the price would come down. In fact during a TEMPO investigator meeting, an Otsuka executive hinted they would lower the price on approval (personal communication). However, despite being the only known treatment that slows the loss of renal function in autosomal dominant polycystic kidney disease, the FDA told Otsuka and the ADPKD community to pound sand.



Somewhere in there, Otsuka changed the labelling and limited tolvaptan to 30 days or less for hyponatremia, so the dream...is officially dead.

But my patients are still alive and they still have sodiums of 125. Demeclocycline, despite being a generic, is very expensive and not a good option. From the European Society of Intensive Care Medicine (ESICM), the European Society of Endocrinology (ESE) and the European Renal Association guidelines on hyponatremia:
The side effects reported for demeclocycline and lithium were such that we recommend not using them for any degree of hyponatraemia.
Fluid restriction, the cornerstone of therapy, is difficult to maintain and in severe cases is insufficient to correct hyponatremia (I'm thinking of patients with negative free water clearance). Urea has a good track record but I have not heard of it being used in the United States. Salt tablets can help, but often are inadequate to correct the hyponatremia.

On the list of possible treatments are loop diuretics. I have tried loops in hyponatremia on a number of occasions and though the math works, in my hands I have not found them to be effective. In the past, I have used loops in hospitalized patients with hyponatremia. The results have been underwhelming. But I know have a loop diuretic success story in a patient with significant but stable outpatient hyponatremia.


I met the patient when he was admitted to the ICU with mental status changes due to a sodium south of 120. This was not his first episode of hyponatremia. We corrected the sodium and restored normal mentation. We did a thorough work-up, looking for the etiology of the SIADH and despite some promising leads that turned into blind alleys, I am quite confident, now, that this is idiopathic SIADH.

During subsequent outpatient follow-up he had persistent hyponatremia with sodiums running in the high 120's. During this time, treatment consisted of salt tablets and fluid restriction. A couple of visits ago I added torsemide, and boom the two sodiums since have been 138 and 134.






Here is the sodium and urine osmolality over time. It plummets after the torsemide is started. This increases the free water clearance.

I also have data on the urine sodium, to get an idea of the electrolyte free water clearance. The change is not nearly as dramatic or convincing.

I have some of the data to calculate electrolyte free water clearance, but I'm missing urine volumes. We can determine the character of the urine from the following formula:

The following percentages represent the fraction of the urine volume which is electrolyte free water. The first three columns are negative, indicating that the urine the patient is producing has less than no free water. Urinating is more like drinking water as urination actually causes the sodium to fall, rather than rise. For a more in-depth explanation of the electrolyte free water calculation, check out this video.

It will be interesting to see if this improvement continues. I now believe that the reason I was underwhelmed when I used loop diuretics in the hospital is that I was working in the compressed time scale of inpatient medicine and only when you stretch the time horizon to months does the drug become effective. I think the reason it takes so long is that loop diuretics need to wash out the concentrated medullary interstitium, thus preventing the ADH from reabsorbing much water. A drug induced partial nephrogenic diabetes insipidus.


Friday, December 12, 2014

Yes, it's a little pocket of insanity but does it have something to teach us?

A few months ago there was a screen shot of a clinic note floating around the internet. It was, in the words of another nephrologist the most passive-aggressive nephrology consult note I have read in a long time:
Patient's sodium dropped further to 120 in the evening. He has had a precipitous drop that I suspect is due to over-diuresis, which does not seem to be a diagnosis within the lexicon of heart failure cardiologists. It is possible that he could have developed SIADH, through a drug side effect. In any case, we have reached the usual place where attempts to fix the heart have blithely interfered with renal physiology, and I am not willing to let his serum sodium decline into the 110s. If we give NS, and he has SIADH, we will worsen his serum sodium. We could use 3% NS, but he is not having mental status changes, yet, and this is bad form for a patient in heart failure. If he is volume depleted, and we use conivaptan, he could develop hypotension which would be difficult to fix. So I seem to have been finagled into ordering tolvaptan, which will hopefully prevent any further decrease tonight. Tolvaptan fixes a number and has not been shown to improve clinical outcomes with chronic use.
Clearly there is a large dose of crazy in the assessment and plan, but it highlights a number of real issues in hyponatremia. Let's dissect the note a bit and tease out the best parts.

The first clam that over-diuresis does not seem to be a diagnosis within the lexicon of heart failure cardiologists seems to be true. A brief survey of google finds a paucity of relevant hits for the phrase and most of those are from nephrologists or family practitioners. Given the frequency that I see patients suffering from this I was a bit shocked at these results.

The next sentence seems a bit preposterous, It is possible that she could have developed SIADH, through a drug side effect. Presuming that a patient with heart failure induced hyponatremia now has a second denovo disease seems a bit of a stretch, but we don't have access to the clinical data and so it is hard to determine if this is true. However the definition of SIADH requires that patients be euvolemic and judging from as much of the story as we know it seems like this patient is clinically hypervolemic. This rules out a clinical diagnosis of ADH, because the release of ADH in heart failure is due to physiological trigger for ADH, a decrease in perfusion. The disease of SIADH is specifically reserved for patients in which there is no physiologic stimuli for ADH release. The presence of heart failure and volume overload, definitionally rule out SIADH.

The next sentence is interesting: In any case, we have reached the usual place where attempts to fix the heart have blithely interfered with renal physiology, and I am not willing to let his serum sodium decline into the 110s. Diuretics increase water and sodium loss, but the cation content of the urine is almost always significantly lower than the plasma cation content, urinary sodium with loop diuretics is typically around 70 mmol/L. So use of loop diuretics cause loss of relatively more water than sodium and result in hypernatremia, except in heart failure. To understand why, one needs to understand electrolyte free water clearance (and an example of using it in the treatment in hypernaatremia is here). The higher the free water clearance, the less prone patients are to hyponatremia. Here are the calculations for electrolyte free water clearance for a patient with hyponatremia due to CHF before and after the addition of loop diuretics:
Before diuretics
In CHF, the patient is actually doing a pretty good job clearing free water. More than half of the urine output is electrolyte free water, the character of the urine is appropriate for correcting the hyponatremia. The problem is not the character of the urine but the amount. The patient just doesn't make enough urine to generate adequate electrolyte free water to account for the water the patient is drinking. Water restriction will be effective for these patients.

So, if the problem is an inadequate amount of urine the logical next step would be to increase the volume of urine with a diuretic:
With diuretics
Unfortunately, though the diuretic increases the volume of urine, it also changes the character of the urine. In this case, it dramatically increases the urine sodium content. This makes the urine almost completely ineffective at removing electrolyte free water and the net result is that the electrolyte free water clearances actually falls with the addition of the diuretic. This is the trap our poor nephrologist is raging against.

The next sentence of the rant: If we give NS, and he has SIADH, we will worsen his serum sodium. We could use 3% NS, but he is not having mental status changes, yet, and this is bad form for a patient in heart failure. I have a problem with this sentence and do not think it is well thought out. The nephrologists clearly believes this patient is over diuresed, we see this situations all the time and they respond briskly to additional fluids. Don't complain that the patient has been overdosed with diuretics and then refuse to provide the antidote for this overdose. I am skeptical of his theory that the patient has SIADH and it is a bit unreasonable to even give a trial of 0.9% saline. In regards to 3% saline, our good doctor might want to take a look at some of the data on the use of 3% saline in heart failure: SMAC-HF (PDF) or Seminars in Nephrology summary. While it is not standard of care the data are certainly intriguing and when the traditional approach is not helping the patient, as apparently is occurring in this patient, it may be worth a look.

The next sentence is regarding conivaptan: If he is volume depleted, and we use conivaptan, he could develop hypotension which would be difficult to fix. Conivaptan is a non-selective vasopressin antagonist, as opposed to tolvaptan which is a selective V2 receptor antagonist. Blocking V1 could cause hypotension as has been reported in multiple case reports. See this open access review.

So I seem to have been finagled into ordering tolvaptan, which will hopefully prevent any further decrease tonight. Tolvaptan fixes a number and has not been shown to improve clinical outcomes with chronic use. This complaint that tolvaptan fixes a number seems a bit obtuse for a nephrologists who was presumably consulted to fix a number. It also implies that tolvaptan is unique in that it has only been shown to fix a number. Well unfortunately, all the therapies of hyponatremia, outside of acute symptomatic hyponatremia have only been shown to fix a number. However given the profound morbidity associated with low sodium, it seems judicious to correct hyponatremia until data proves it is unhelpful. Additionally, to claim in one sentence that you refuse to allow the sodium to fall below 120 and in the next sentence to rail against a therapy that has only has been shown to fix a number seems to belie a profound lack of self awareness.

But keep fighting the good fight against the cardiologists, somebody needs to keep their egos in check. #PracticallySurgeons


Thursday, December 11, 2014

Last chance to vote

The longest tradition in the nephrology blogosphere, the Renal Fellow Network's Nephrology Story of the Year! For five years RFN has been posting the top stories of the year and for the last few years they have been off loading the work to the crowd. So do your duty and vote.

Go vote. 

Polls close tomorrow. 


And to the losers stuffing the ballot box for "Perivascular Gli1+ progenitors contribute to myofibroblast pool leading to fibrosis in multiple organs including kidney Cell Stem Cell" I will not stand for that!

I'm pulling for Dendritic cell isoketals activate T cells and promote hypertension as covered in NephJC.

Monday, December 8, 2014

The newest nephrology blog: Nephrology Tweetbook

Run by master tweeter Nikhil Shah, Nephrology Fellow at the University of Alberta, Nephrology Tweetbook is primarily a collection of educational tweets with, as far as I can tell, a single long form post on the use of What's App as an educational tool. Very interesting use of the app.

I'm not sure what he is using to post the tweets to blogger, but he would get better, most useful posts if he used the embed code from twitter.

This is what his posts look like:


No active links.

If he were to use the embed tool in twitter it would look like this:







It's a nice addition to the nephrology social media landscape.

Related Posts Plugin for WordPress, Blogger...