Update in nephrology for 2023

Today I was honored to give Medicine Grand Rounds at MedStar Health in Baltimore. Here is a wlk through and bibliography for this talk.

Title slide. Slides are available for download at:

Keynote What’s New in Nephron Town (31.7 MB)

PowerPoint What’s New in Nephron Town (16.9 MB) Note: I create, rehearse and deliver the presentation in Keynote. The PowerPoint version is a simple export of the Keynote presentation and often looks like garbage. If you want to see the presentation as it was meant to be, use Keynote

PDF What’s New in Nephron Town (9.8 MB) Note: The PDF shows only the final build of the slides and often that is not very illustrative of the point that the slide was trying to make. If you want to see the presentation as it was meant to be, use Keynote.

Outline of the talk. I originally outlined a section on aldosterone synthase inhibitors and IgA nephropathy, but ran out of time preparing for the lectures (aldosterone synthase) and thought it was not a great topic for general medicine audience (IgA nephroathy).

CDC data: SGLT2i use among Patients with CKD and Diabetes

Safety and Effectiveness of Bexagliflozin in Patients With Type 2 Diabetes Mellitus and Stage 3a/3b CKD AJKD

NO STONE Study: Dhayat, N. A. NEJM, 388(9), 781–791. (PubMed)

Craig Brater’s review of diuretics in the NEJM from 1998 (NEJM)

ALLHAT Hypertension results in JAMA. (PubMed)

CLICK Trial (PubMed)

The Diuretic Comparison Project. A pragmatic trial of HCTZ v chlorthalidone. (PubMed)

Meta-Analysis Comparing Torsemide Versus Furosemide in Patients With Heart Failure (PubMed)

TRANSFORM-HF Randomized Clinical Trial (PubMed)

AIN is a common finding on kidney biopsies done for AKI (at least at the University of Chicago, and among the elderly) (PubMed)

Urine eosinophils are not at diagnosing AIN. (PubMed)

Identification and validation of urinary CXCL9 as a biomarker for diagnosis of acute interstitial nephritis (PubMed)

Drug-Induced Acute Interstitial Nephritis. (CJASN)

Excellent review by the two guys who identified the gene. APOL1 Nephropathy from Genetics to Clinical Applications (CJASN)

Inaxaplin for Proteinuric Kidney Disease in Persons with Two APOL1 Variants (PubMed)

Link to information about free genetic test for APOL1 at APOL1CKD.com

The CONVINCE Trial (NEJM)

Retrospective data on the speed of correction from Pennsylvania (PubMed)

Toronto data showing a lack of association between speed of correction and incidence of ODS (NEJM Evidence)

Boston data looking at risk of mortality by speed of correction in hyponatremia. (NEJM Evidence)

Multi-center data also showing increased mortality with slow correction of hyponatremia. (J of Crit Care)

Apocalypse Nephrology

This summer I was really optimistic. I was riding a #Flozin high that was being further buoyed by the successes of finerenone. Long dormant fields were yielding fresh life. How long has it been since we’ve had this much investigation into glomerulonephritis? So many trials. So many agents.

This is how I started a letter of rec I wrote this summer:

Can you feel it? Can you smell it? Yup, the winds of change have reached nephrology. Just as the most optimistic and sanguine nephrologists predicted, the new diagnostics, new therapies, and new hopes have blown in fresh enthusiasm for nephrology as a career. I am seeing more interns express interest in nephrology, more seniors asking for letters of recommendation, and more general interest for our specialty than I have seen in a score of years. Yes, the gains in the match results have been meager but change is coming and Dr. XXXXXX is a flag bearer for this renewed renal enthusiasm.

Well I was wrong. The winds of change were blowing us backward. After a good year for 2023’s starting class, this year’s match was a dud.

A. Pivert, Kurtis. 2023. “First Look: AY 2024 Match.” November 29, 2023. https://data.asn-online.org/posts/ay_2024_match.

All those unfilled positions often make the training harder for the fellows that do match.
And residents see how hard the training is.
And they see the long and unpredictable hours.
And hospitalist looks better and better.

I love nephrology. I want nephrology to be a vibrant, attractive specialty, but what we are doing, isn’t working.

We have added training capacity like mad without building the demand for the positions. Look at the number of people matching every year (yellow). It is pretty flat since 2014. But the number of positions, and unfilled positions has grown steadily (blue).

A. Pivert, Kurtis. 2023. “First Look: AY 2024 Match.” November 29, 2023. https://data.asn-online.org/posts/ay_2024_match.

This graph from Brian Carmody is interesting. Look where rheumatology and endocrinology are. Two specialties that are not lucrative. But they fill. Reliably and competitively. Lifestyle is important. This has been neglected in nephrology. We work hard, long, hours for sick patients. This should be an area we work on to remodel the specialty around new visions of work. Do we need to train nephro-hospitalists? Consider a one year fellowship with a focus on ICU and inpatient nephrology. Less clinic. Less GN. Less chronic dialysis. And this could be paired with an outpatient nephrology fellowship. Also one year. Modeled after endo or rheumatology with a clinic based schedule. Time dedicated to learning the skills and procedures to create and maintain dialysis access. If we removed the burden of inpatient rounding, fellows could master outpatient nephro and outpatient dialysis in a single year.

Splitting the roles would make the job easier and less chaotic.

Now look at ID the other end of the graph. ID and nephrology. Two specialties that can’t draw fellows. We need to fix this. This is an existential moment for the field.

Updates from Twitter

Poyan Mehr

I tend to agree with Joel. [this is the most important part] The discussion around compensation is valid but doesn’t explain the strong demand for lower-paying specialties. Besides, as an entrepreneurial nephrologist in private practice a 500-600k salary is not unusual. Money is not everything. Nephrology in the way we currently practice it is not attractive. And trainees see that. Let me explain:

Nephrologists remain proud generalists in an era of ever more complex diagnostic and treatment options. Other than dialysis, for the most part, we don’t own anything that our patients need. We need to beg other specialties to help us with fistula, catheter, biopsy, imaging, infusions, pheresis, and invasive volume assessment. Had we shown some pride, ownership, and support for clinical research as cardiology did, we would have now branched into several subspecialties and be more attractive, given the diverse opportunities catering to different preferences. Still, we also would have likely been more inventive in driving progress. Again, using cardiology as an example, cardiac imaging has advanced because cardiologists understand the clinical problems and have content expertise in finding technical solutions. Cardiac MR or CT wouldn’t be where it was if it were for the radiologists who are busy with 60 different body parts and would remain unaware of questions that would have never been asked.

But rather than creating a subspecialty track, nephrology dwells in the nostalgia of being a generalist. In 2023 still talking about the same electrolyte and acid base problems from 60 years ago with formulas validated in 10 med students.

We are risking to become a profession of mediocre generalists. The average rheumatologist is better at treating lupus nephritis than the average nephrologist. Not because they are the better doctors but because the nephrologist also has to master peritoneal dialysis, CRRT in ICU, post-transplant renal care, hypoNa with seizure, and on and on. How do we expect trainees observing us (jacks of all trades and master of none) and getting inspired??

Nephrology as a specialty needs to have either a radical makeover or will diminish and perish, with endo taking over electrolyte management, rheum autoimmune disorders, cardiology volume and BP, critical care CRRT, and so on. Some say that would be good so. Because it would be better for the patients.

And though this was not in response to this essay it is a good thought

New Tweetorial: Comparing diffusive versus convective clearance

I have been workshopping this one for awhile in my mind and today I carved out a few hours to create it.

It starts here

Part of the inspiration for this came from an epic message on the Channel Your Enthusiasm back channel by Roger Rodby

Here is the draft of the script.

Diffusive vs Convective clearance

I was teaching the third year medical students about acute kidney injury and the lecture begins with a brief history of extracorporeal dialysis for AKI. And I asked a student what extracorporeal dialysis was, and he correctly identified it as “dialysis outside the body.”

Since the ironclad law of the Socratic Method is that every correct answer is rewarded with another, harder, question, I replied, “Can you think of any example of intracorporeal dialysis?” The right answer is peritoneal dialysis, but he said , “The kidney?”

And off on a tangent we went…
Does the kidney even do dialysis? No. The kidney does not use diffusion to clean the blood. Clearance is provided by convection at the glomerulus. Plasma is squeezed through the slit diaphragms of the podocytes in the glomerulus but besides the lack of protein, the solute composition on both sides of that membrane is essentially identical.

The kidney does not clear the blood by diffusion, the defining characteristic of dialysis, but rather by convection. How does that work? Glad you asked. Take Creatinine. The creatinine on both sides of the podocyte is the same, 4.4 mg/dL in this example.

4.4 mg per dL x a GFR of 25 mL per minute x 1440 minutes in a day divided my 100 mL in a dL comes to 1584 mg of creatinine filtered.

That is just about the amount of creatinine produced by a typical person a day. 

So convective clearance can clear all of the creatinine produced everyday, the additional creatinine secreted in the proximal tubule is just gravy. 

What about sodium? 

138 mEq/L x a GFR of 25 mL per minute x 1440 minutes in a day divided by 1000 mL in a L comes to 4968 mE of creatinine filtered. 

This is a problem since we only consume around 100-200 mEq of sodium a day. So this where the tubules earn their stripes by reabsorbing all the excess filtered sodium to keep us from peeing ourselves to death.

So these two examples demonstrate an important principle of convective clearance, it is better for clearing things at a high concentration than at a low concentration. In fact, a GFR of 1 is enough to clear a typical sodium daily load.

138 x 1 ml/min x 1440 min/day divided by 1000 ml/L = 198 mEq/day

This why even a tiny residual renal function makes a huge difference in dialysis patients. 

But that same GFR of 1 would only clear 

4.4 x 1 ml/min x 1440 divided by 100 ml/dL = 63 mg of creatinine only about 4% of the daily creatinine load.*

*This calculation is highly dependant on the serum Cr concentration, which would be a lot higher than 4.4 if the GFR was 1, but since a GFR of 1 in incompatible with life, the patient would also be getting renal replacement therapy, so it is hard to know where the serum Cr would actually be.

So after explaining that the kidney didn’t actually do dialysis, or anything remotely close to dialysis. I asked if there was an organ that did do dialysis? Or, more specifrically, used diffusion for clearance.
Answers from the crowd: 

Liver > nope

Spleen > nope

Skin > nope

And finally, Lung? Yup.

The lung clears carbon dioxide from the body and absorbs oxygen by setting up a setting where the gasses move down their respective concentration gradients across a semipermeable membrane. You know, like dialysis.

A ventilator is not really like an artificial lung, in the way a dialysis machine replaces the core function of a kidney. It provides flow, but no clearance. We still are dependent on the alveolar membrane for oxygen absorption and carbon dioxide clearance. 

But ECMO is an artificial lung and fully replaces the alveoli and uses the principles of dialysis to clear carbon dioxide and move oxygen into the blood. So at some level, ECMO is closer to the lung than dialysis is to the kidney. 

One final note on this thread is in regards to dialysis and convection. The kidneys work by convective clearance but our primary means of replacing them is by diffusive clearance. However this summer we saw a randomized controlled trial of modifying dialysis to use convection rather than diffusion…and the result? Significant reduction in total mortality. 

We don’t get a lot of wins in dialysis, so when we get one, we pay attention.

The script isn’t exact because I have to do some edits to meet the character limits of tweets.

Here are the Keynote slides that I used to create the gifs.

Is the juice worth the squeeze?

I have a long-time patient who came in yesterday and we were reviewing their labs together and noted a tremendous success. The patient’s albuminuria has fallen 90% since 2018. This is due to adding and then maximizing losartan, adding dapagliflozin, and most recently starting finerenone.

Here is the albuminuria over time:

Over that time their eGFR looks like this:

How should we square the recent drop in eGFR with the reduction proteinuria? We typically switch to a risk based model, here are their Tangri KFRE scores over time:

I was quite surprised to see the KFRE rise over time despite the 90% reduction in albuminuria. I am not changing therapy. The eGFR is nearly intact and I am not going to tiny changes in creatinine (the most recent jump was less than 0.2 mg/dL) guide therapy decisions. I am going to continue to ride these medications which have repeatedly been shown to protect kidneys (and hearts).

OUWB Question: What’s going on with this mess?

I create all of my presentations in Keynote. In the past I was able to present them in Keynote but due to construction and having to give lectures in 156 of North Foundation I had to convert the presentations to PowerPoint. And then hope that PowerPoint on the PC was mostly like PowerPoint on the one true computer, the Mac. It works surprisingly well, but this slide is an outlier.

Here is what it looks like in PowerPoint on my iMac

This is better, but how does this slide fit into the lecture? It is a summary slide after I have discussed two of the causes of the maintenance of metabolic alkalosis. Here is the slide that introduces the four mechanisms for maintaining metabolic alkalosis

Then the lecture describes kidney failure (the easiest to unsderstand)

After that I cover the mechanisms by which hypokalemia maintains metabolic alkalosis in 7 slides as seen here:

The I do the same for chloride (volume) deficiency in 7 slides:

After going through that I noted similarities between the mechanisms of hypokalemia and chloride deficiency, the summary slide is designed to highlight those. Looking at that slide now, it doesn’t do such a good job of that. Does this pair of slides work better?

Here is the revised powerpoint for you to download

You cannot tell if a respiratory acid-base disorder from the ABG

This is a frequent cause of confusion. I know that I was confused by this when I was a young learner. And I believe the source of this confusion was garbled teaching from a resident that was still struggling with the concept.

Take this ABG:

pH: 7.26

PaCO2: 80

HCO3:
39

The Henderson-Hasselbalch variables are moving in discordant directions (pH down, pCO2 and HCO3 going up) so it is a respiratory disorder. The pH is decreased so this is a respiratory acidosis.

Now look at the compensation to see if there is a second primary disorder affecting compensation.

  • In acute respiratory acidosis the HCO3 rises 1 mEq/L for every 10 mmHg the CO2 rises.
  • In chronic respiratory acidosis the HCO3 rises 3 mEq/L for every 10 mmHg the CO2 rises.

This patient’s CO2 is 80, an increase of 40, so the HCO3 should rise 4 (4×1) if the respiratory acidosis is acute, yielding a bicarb of 28 (24+4). The actual bicarbonate is 39, too high, so there is an additional metabolic alkalosis if the respiratory acidosis is acute.

If the respiratory acidosis is chronic, to increase in CO2 of 40 should increase the bicarbonate by 12 (4×3), so a bicarb of 36 (24+12). The actual bicarbonate is 39, which is just outside of our ±2 so we’ll call it nearly appropriate with just a touch of metabolic acidosis.

The ABG can be “solved” with either acute or chronic respiratory acidosis. Patients cannot be diagnosed with Occam’s razor so the simpler explanation (chronic respiratory acidosis without the need for additional acid-base disorders) may not be the right one. In medicine we need to assume Hickam’s Dictum “A patient can have as many diseases as he damn well pleases.”

The ABG does not determine whether a patient has acute or chronic respiratory disorder, the physician must do that.

So what’s my beef with the two Bruces? Take a look at this question from chapter 8…

The simple acid-base disorders are:

  1. Metabolic acidosis
  2. Metabolic alkalosis
  3. Respiratory acidosis
  4. Respiratory alkalosis
  5. Respiratory acidosis
  6. Respiratory alkalosis

But this is answers the authors expect…

Since the question stipulates that these are simple acid-base disorders, one can extrapolate the acuity of the respiratory disorder by the degree the bicarb has adjusted, a large adjustment is chronic, and a smaller change is acute. But since patients don’t tell you if they have simple or complex acid-base disorders when the blood is drawn, this trains students to expect the ABG to provide information that it cannot provide.

Stupid book.

OUWB student question on ammonium production and potassium

How does hyperkalemia or even alkalosis suppress NH3 production?

So the mechanism is not completely worked out but the conventional explanation is…

Hypokalemia causes potassium to shift out of the cells. Hydrogen ions then move on the opposite direction into cells to maintain electroneutrality.

This causes intracellular acidosis. The intracellular acidosis causes the cells of the proximal tubule to erroneously concludes that there is systemic acidosis and so they up regfulate the production of NH4 to increase renal excretion of acid and produce new bicarbonate.

An interesting note about this happens in liver failure. Patients with liver failure are unable to metabolize ammonia and levels build up to high concentrations and can cause hepatic encephalopathy. A known risk factor for this is hypokalemia. The above paragraph provides an explanation for this.

Conclusion from a recent study on this phenomena (Ref)

Hyperkalemia works the same way but in reverse. Potassium shifts into the cells, This causes hydrogen to move out of the cells and causes intracellular alkalosis. The cell mistakes this intracellular alkalosis for systemic alkalosis and the last thing the kidney wants to do in systemic alkalosis is generate additional bicarb, so it down regulates this generation of ammonium.

Slide 97 from Monday’s lecture