Last Month in Nephrology

When Bayes Confronts Fisher: thinking on the Fish oil (PISCES Trial)

Dialysis trials rarely surprise us, this is a ‘department of perennial disappointment’. So quite naturally, when a trial reports a massive 40% reduction in cardiovascular events, our first reaction is a mixture of excitement and suspicion.

For the general population, with regards to fish oil, major guidelines (AHA and ESC) emphasise individualised assessment, suggesting fatty fish intake over pills due to inconsistent trial results. n-3 fatty acid blood levels are lower in patients receiving hemodialysis than in the general population, and PISCES investigators hypothesized that replenishing this deficit will improve outcomes in patients with kidney failure. Trial recruited over 1200 patients on hemodialysis, from 16 sites across Canada and Australia, most over 60 yrs, about half in each arm were diabetic.

Primary end point was serious cardiovascular events, a composite of cardiovascular death (sudden and non-sudden cardiac death), nonfatal and fatal myocardial infarction, nonfatal and fatal stroke, and peripheral vascular disease leading to amputation. Here are the results: an astonishing 43% reduction in serious cardiovascular events.

Results have left most nephrologists awestruck: double blind trial, no industry conflicts, adequate power and almost flawless execution! We are not used to seeing massive CV benefits like this. HDF enthusiasts debating over the holy grail of “higher convective volumes” will be equally clueless and embarrassed, about the discovery of this ‘magic bullet’ containing fish oil, offering 43% reduction in the CV events (HDF claims about 22-31% reduction). At the same time, the mortality signal remains modest and statistically uncertain, with a hazard ratio of 0.89 and a confidence interval that crosses unity.

Reconciling these two results — impressive cardiovascular benefit and restrained survival effect — requires more than a glance at confidence intervals. As we were discussing these results, Dr Manjunath decided to understand these findings from a less used but often useful Bayesian point of view.

If we dissect the mortality result — an HR of 0.89 with a confidence interval of 0.73 to 1.01 — the discussion quickly follows a familiar pattern. Optimists argue that there is a trend toward benefit. Skeptics conclude that there is no mortality benefit. Fence-sitters quote the confidence interval and move on.

We count ourselves among the skeptics (not cynics), and we decided to invoke Dr. Bayes to help us decipher what the trial actually tells us. What follows here is a very simple understanding of this alternative statistical approach, we assure that most clinicians will be able to understand the concept.

The question we care about — but rarely ask

We usually ask:

Is mortality reduction statistically significant?

But the question we should be asking is different:

How likely is it that fish-oil supplementation meaningfully reduces mortality?

These are not the same questions. A 5% reduction in mortality and a 30% reduction in mortality both produce hazard ratios less than 1, but clinically they are worlds apart. If we are going to invoke Dr. Bayes, we must ask an honest question:

What is the probability that fish oil reduces mortality by an amount we would actually care about?

Let’s do a simple thought experiment. Let us go back in time — to a point before the PISCES trial was published. If we had asked our colleagues then: What would happen if we gave omega-3 fatty acid supplementation to dialysis patients? Would it reduce mortality?

Most of us would be surprised to see any mortality benefit, though small effects might be possible. These statements are perfectly acceptable over a cup of tea (or coffee, if you are so inclined).

But to invoke Dr. Bayes, we must translate this intuition into Bayesian statistic language which goes like this:

Zero effect is the most plausible value, but small effects are possible.

This was the prior belief we carried out our Bayesian analysis. This prior knowledge about the intervention (referred to as ‘prior’ henceforth) is crucial in Bayesian statistics unlike the conventional Frequentist approach where the current trial’s results alone are considered. For readers interested in the mathematical details, we have included the full calculations in a supplement(Click to download). We did not want to burden all readers with Bayesian notation and jargon in the main text.

What happens when belief (prior knowledge) meets the current results of PISCES trial?

Cardiovascular events

When we combine our skeptical prior (centered at no effect) with the strong cardiovascular signal from PISCES, the results are reassuring. The probability that fish oil reduces cardiovascular events at all is approximately 99.99%. More importantly, the probability of a ≥25% reduction in cardiovascular events is about 80%.

This is the moment to pause and revisit the headline HR of 0.57. Under Bayesian scrutiny, the effect size softens slightly, drifting toward 0.65–0.70, but the conclusion survives intact. The Bayesian verdict is clear:

The cardiovascular benefit appears real. The exact magnitude is uncertain, but the direction is not.

Far from weakening confidence, the Bayesian analysis strengthens the results of the trial.

Mortality: where caution is essential

Now let’s look at all-cause mortality results, and this is where we must tread carefully.

Our analysis shows that the probability of any mortality benefit (very small to large) is around 75%. However, the probability of a mortality reduction greater than 25% is between 0 and 1%!

Read that again.

It is essentially zero.

This means that if fish oil reduces mortality at all, it almost certainly does so by a small amount, not a dramatic one. This conclusion is not obvious from the confidence interval alone, but it becomes unavoidable when we ask the right question.

Clinically, this matters. Saying “the confidence interval crosses 1” is not helpful. Bayesian analysis allows us to state the result clearly:

A mortality benefit is plausible, but a large survival advantage is extremely unlikely.

This protects us from overselling a benefit that the trial does not support.

Does this change how we practice?

Yes — but in a measured way.

Fish oil deserves serious consideration for cardiovascular protection in dialysis patients
With fish-oil supplementation, expect fewer cardiovascular events, not miracles
Do not sell this as a magic bullet that improves survival

The next time a patient, a pharmaceutical representative, or a colleague who has not read this blog asks what PISCES means for clinical practice, a single sentence will suffice:

Fish oil almost certainly reduces cardiovascular events in dialysis patients, but if it reduces mortality, the effect is likely to be small and not dramatic.

We can assure you: that sentence is stronger than any P value.

When diclofenac saved a kidney

“Saturday evening syndrome” is a recurrent phenomenon in doctor’s life—when a last-minute emergency spoils your weekend plans. It was just like any other Saturday evening, except that I needed to finish seeing my clinic patients before 7:30 p.m. I had promised my daughters a post-dinner treat at their favorite ice-cream parlor. This commitment was long overdue; two previous plans had been cancelled because it rained like “foxes and wolves” last week (this phrase is an invention by my elder daughter who thought “cats and dogs” falls short to describe last week’s unprecedented downpour).

Well aware of my previous track record of keeping time, they started texting me from 7 p.m. onward, asking about my whereabouts. But this time, I finished the last consult at 7:25 p.m. and had just stepped out, when a yellow-black taxi rushed into the clinic premises. I asked my secretary to check the patient’s vitals and replied to the last SMS, “an urgent case, I’ll join late.”

Sayeeda (name changed), a 40-year-old lady, was obviously in distress, wincing with severe pain, her right hand moving from the right hypochondriac to the right flank, as if consoling her furious right kidney—the source of her pain. She was febrile, tachycardic, and had relative hypotension. A tender swelling was palpable in the right paraumbilical region, and her right flank was tender as well. She was recently discharged from a hospital as a case of “Acute Kidney Injury on CKD, ADPKD, Cyst infection.” She had been aware of her CKD for many years and reported no other major comorbidities, her mother died of a stroke at the age of 70 and was also diagnosed with CKD only during her terminal illness. Sayeeda’s kidney function had declined rather rapidly, likely related to her uncontrolled blood pressure, with serum creatinine stable between 2 and 3 mg/dL. Three similar episodes of cyst infection had occurred over the last six months, and she had been in and out of the hospital, exhausting all her family’s financial resources. Serum creatinine had increased to 6.3 mg/dL, acidosis was compensated, and electrolytes were normal.

More than elevated creatinine what bothered her most was severe pain, which was not controlled by round-the-clock paracetamol and intermittent oral tramadol that she was prescribed at last discharge. She was advised to undergo right nephrectomy by a urologist during that admission, with likely need of dialysis post-surgery. They had come to my clinic for a second opinion on this decision.

“It’s a demon inside —that I can feel with my own hand. Take it out, doctor, if that is what is required to alleviate my pain!”

Her bloodshot, sleep-deprived eyes conveyed not only the excruciating pain but also her helplessness, frustration, and a bit of anger over the futility of the treatment she had received so far. The severity of her pain would surely exceed the visual analogue scale limit. I suggested continuing parenteral antibiotics for the cyst infection and ‘diclofenac, 50mg, one tablet at bedtime for three days.’ This, of course, was with a lot of hesitation that any nephrologist would experience when prescribing a potent NSAID to patient with advanced CKD not yet on dialysis. They would arrange for a nurse to give her IV antibiotics at home. The consultation lasted 30 minutes.

It was 8:15 p.m. when my phone beeped with a message notification—an angry face emoji from my younger daughter. I rushed to mark my attendance at the scheduled ice-cream party. Iranian Pista Kulfi is my favorite; it has half the sugar content of most other options and a rich Persian flavor. I usually savor every bite, but that Saturday my taste buds were at a loss, as I kept wondering if an angry face emoji should be added at the top of the pain scale to indicate pain that Sayeeda was experiencing.

Her nurse called me the next day to inform me that, for the first time in a month, Sayeeda slept well and was undisturbed by pain—thanks to diclofenac. I was delighted to see her walking into the clinic with a smile. One question made me anxious: what might have impacted her kidney function in today’s labs—control of infection with antibiotics, or diclofenac? She was so happy that she didn’t even look at the reports to check her serum creatinine, which most patients do. We often gauge the trend in kidney function just by observing the expressions on patients and their family members’ faces. Her creatinine was 4 mg/dL, a significant improvement from her previous reading. Improvement was also evident in her relative hypotension, leukocytosis, and CRP. The antibiotic was continued for two weeks, and her kidney function further improved on subsequent follow-up, with serum creatinine settling between 3 and 4 mg/dL. The impending nephrectomy was deferred; diclofenac appeared to have “saved the kidney!”

NSAIDs are often listed in nephrologist’s patient education booklets as “drugs to avoid.” ‘A history of NSAID use’ is featured in all trainee case presentations. It is widely assumed that NSAIDs can produce CKD, accelerate progression of CKD, and cause acute kidney injury.

The fear of NSAIDs among nephrologists goes back to the ‘analgesic nephropathy’ first described in the 1950s by Spuhler (a clinical investigator) and Zollinger (a pathologist), both from Zurich, Switzerland, who reported a then-uncommon chronic kidney disease with an insidious, slowly progressive course ending in uremic death. These patients were consuming mixtures of phenacetin and several other NSAIDs for years. While acute (largely reversible/hemodynamic) decreases in GFR is a well-known effect of NSAIDs, this was once explored as a measure to control proteinuria. Although a variety of renal manifestations associated with NSAIDs have traditionally been emphasized by textbook tables, the absolute risk of these events for an individual patient may be far lower than previously believed. Is a blanket ban on the use of NSAIDs in CKD justified?

Older studies indicating nephrotoxicity from NSAIDs are limited by their retrospective nature, recall bias, lack of comparators, and confounding by indication. For example, in a study of over 12,065 adults (aged 20 years or older) from the National Health and Nutrition Examination Survey (1999–2004), NSAID use increased with worsening CKD stage: use reported by 2.5%, 2.5%, and 5.0% with no, mild, and moderate to severe CKD, respectively. Interestingly, disease awareness was not associated with reduced current NSAID use: 3.8% vs. 3.9%, aware vs. unaware (P = .979). It is possible that patients using NSAIDs chronically (despite the publicized harm) are more likely to have chronic pain conditions, chronic inflammatory disorders, and comorbid conditions likely to increase the risk of kidney function decline on their own, with or without NSAID exposure. This was highlighted by the Nurses’ Health Study, where acetaminophen use (over 3 kg in a lifetime!), and not NSAID use, was associated with kidney function decline. A preferential shift toward acetaminophen use may be due to a perceived risk by a population aware of NSAIDs’ risks. Such biases in observational studies cannot be controlled by even the most sophisticated statistical methods. The relationship between NSAID use and development of CKD is questioned by observational but well-conducted prospective cohort studies (here and here).

In a prospective rheumatoid arthritis cohort study involving 4,101 patients, after a mean follow-up of 3.2 years, among patients with eGFR >30, no difference in kidney function was noted between NSAID users and NSAID-naive cases (mean change in eGFR CG −0.87 mL/min/year, 95% CI −1.15 to −0.59) and ‘NSAID-naive’ (−0.67 mL/min/year, 95% CI −1.26 to −0.09, p=0.63). NSAIDs were an independent predictor for accelerated renal function decline only in patients with advanced baseline renal impairment (eGFR CG <30 mL/min).

The problem with the “No NSAIDs” policy is the increasing use of opioid analgesics and gabapentin-like drugs. There is no convincing evidence of safety for long-term opioid use in non-cancer pain; studies are limited to short-term trials with placebo comparators and exclusion of high-risk patients. The prescription opioid crisis and deaths involving tramadol abuse have significantly increased in the UK. In India, where tramadol-like drugs are available over the counter, we are already in the midst of the problem, with the only difference being a lack of reliable statistics on the harms of these drugs. There are reasons to believe that opioid manufacturers might have deliberately underplayed the risks and used several malpractices, such as large bonuses for sales representatives and coupons entitling new patients to free samples. The most commonly used drug in this category, probably the most notorious for patients with decreased kidney function, is known for serious side effects such as seizures and serotonin syndrome. CRIC Study Investigators reported a stronger association of opioids with adverse events than NSAIDs. Pregabalin and gabapentin are mostly ineffective for non-neuropathic pain, are known to cause somnolence, dizziness, and falls, and when combined with opioids, increase the risk of death.

Pain is an extremely common symptom in CKD, and uncontrolled pain significantly affects quality of life. A total ban on NSAIDs, along with rising opioid use, is completely unjustified in light of existing evidence. In patients with eGFR >30 mL/min, short courses of NSAIDs at reduced doses are unlikely to significantly impact kidney function. In patients with severe pain and eGFR <30 mL/min, a shared and individualized decision about the risks and benefits of NSAIDs should be made, rather than a strict prohibition on their use.

Albuminuria as a Surrogate Endpoint for ESKD: When Sophisticated Statistics Meet Uncomfortable Truths

If an examination is too difficult to pass, obviously wise approach is one of the two: students who consider themselves capable of the test shall only apply, and those who decide to appear study hard to crack the examination. This was commonsense wisdom, but things are changing in recent times. There is a possibility of getting through the examinations even if you can’t fulfil abovementioned two criteria. How?

Universities can now decide to lower the cutoffs for passing, making sure that most applicants can make it through. Now whether this is the right way is both a philosophical and practical debate. On one hand, you can ‘create’ as many professionals as you wish, but at the same time risking the overall lowering of the ‘minimum standard skills’ that are needed for safe and successful practice in these professions. You may be wondering whether last month in nephrology is infested by some bug that we are writing a post that appears to be irrelevant to the nephrology audience, but wait. Dr Manjunath is trying to give practical example of this in the following critical appraisal.

This blog is the result of an exercise that began as a journal club discussion of three major meta-analyses evaluating albuminuria as a surrogate for kidney failure. Here, we summarize the critical insights we uncovered during this process.

We started with a simple question: Can changes in albuminuria predict kidney failure outcomes in clinical trials? A positive answer to this million-dollar question could bring new drugs and more clinical trials to nephrology. Instead of waiting 5–10 years and spending over $100 million on traditional trials with hard endpoints like dialysis or transplantation, why not use albuminuria reduction as a surrogate, delivering answers in just 1–2 years at a fraction of the cost?

WHAT IS A SURROGATE ENDPOINT?
A surrogate endpoint is a lab test or measurement that researchers use to predict a clinical outcome without waiting for that outcome to actually happen. Example: Instead of waiting 10 years to see if a drug prevents heart attacks, researchers might use cholesterol reduction as a surrogate—assuming lower cholesterol means fewer heart attacks.
Appeal: Faster, cheaper trials that can bring treatments to patients sooner.
The Risk:What if the surrogate doesn’t actually predict the real outcome?

On the surface, tracking albuminuria seems like the wisest approach. The biological rationale is compelling: albuminuria reflects glomerular injury, correlates with disease progression, and responds to many effective therapies. There is a strong premise that albuminuria would be a good surrogate for ESKD. What could go wrong?

As it turns out, quite a lot.

Three Studies, One Uncomfortable Truth

Between 2018 and 2019, three major meta-analyses attempted to validate albuminuria as a surrogate for end-stage kidney disease (ESKD). Each study utilized increasingly sophisticated statistical methods, yet all revealed the same uncomfortable truth: the evidence isn’t ready for prime time (to borrow Dr. Tukaram’s words).

Study 1: Palmer et al. (AJKD 2018) — The Bayesian Bivariate Approach

This meta-analysis, involving 69,642 patients, attempted to correlate treatment effects on albuminuria with those on ESKD across multiple trials.

The surprising result: Only 3.4% of treatment comparisons showed significant ESKD reduction. The correlation was -0.41, with a 95% credible interval ranging from -0.98 to +0.88.

In plain terms: “We have no idea if albuminuria helps or harms kidney outcomes.” The R-squared was just 0.17, meaning albuminuria changes explained only 17% of the variation in kidney failure outcomes. The remaining 83%? Unexplained(GRADE assessment: Low).

UNDERSTANDING CORRELATION AND R-SQUARED
Correlation tells us if two things move together:
+1 = perfect positive relationship
– 0 = no relationship
-1 = perfect negative relationship

R-squared tells us how much one thing explains another:
R² = 0.17 means albuminuria explains 17% of kidney outcomes
R² = 0.83 would mean it explains 83% (much better!)
Palmer’s problem:Their correlation ranged from -0.98 to +0.88, meaning albuminuria could either strongly help or strongly harm—we just don’t know which.

Study 2: Heerspink et al. (Lancet D&E 2019) — The Individual Participant Data Meta-analysis

The investigators undertook a large consortium of individual patient data for surrogate validation: 29,979 patients from 41 trials. They used advanced Bayesian mixed-effects meta-regression with appropriate priors and state-of-the-art methodology (we admit: it took us quite some time to fully grasp the analysis).

The headline: “30% albuminuria reduction = 27% lower ESKD risk!” We thought we had finally found the holy grail, but our celebration was premature. Skeptics at heart, we dug deeper. Ignoring the uncertainty intervals is an unpardonable mistake when evaluating studies(see text box to understand Confidence interval). Reality set in: the 95% credible intervals were 5% to 45%. This means that a 30% reduction in albuminuria could lower ESKD risk by anywhere from 5% to 45%!

WHAT ARE CONFIDENCE INTERVALS?
A confidence interval shows the range where the true answer probably lies.
Example: “30% albuminuria reduction reduces kidney failure risk by 27% (95% CI: 5%-45%)”
Translation:The best guess is 27% reduction, but the true benefit could be as low as 5% or as high as 45%.
The problem: This range is so wide it’s clinically meaningless. How do you counsel a patient when the benefit could be anywhere from minimal to substantial?

A significant issue with this meta-analysis: the study obtained individual patient data from only 41 of 81 eligible studies. This 50% data-sharing rate introduces serious selection bias, as studies with positive results may be more likely to share data, skewing the whole analysis(GRADE assessment: Low).

Study 3: Coresh et al. (Lancet D&E 2019) — The Large Observational Cohort

Click here for Difference between Experimental studies versus Observational studies

OBSERVATIONAL vs EXPERIMENTAL STUDIES
Observational Study:”People who exercise more have fewer heart attacks”
– Shows association
– Can’t prove exercise prevents heart attacks
– Maybe healthy people just exercise more

Experimental Study (Clinical Trial): “We randomly assigned people to exercise vs no exercise and measured heart attacks”
– Can prove causation
– Tests if intervention actually works
For surrogate validation: We need experimental evidence that changing the surrogate actually changes outcomes.

The CKD Epidemiology Collaboration pooled data from about 700,000 participants across multiple cohorts, adjusting for regression dilution and other confounders.

The finding: a 30% albuminuria decrease equated to a 22% ESKD risk reduction (HR 0.78).

The limitation: this was purely observational data. In observational studies, association does not equal causation, and such relationships do not validate treatment surrogates(GRADE assessment: Low).

GRADE Assessment: The “Confidence Meter”
GRADE (Grading of Recommendations Assessment, Development and Evaluation) is a systematic way to evaluate how much we can trust research findings. It is a “confidence meter” that helps us decide whether to act on study results.
GRADE uses five key questions to assess evidence quality:
Risk of Bias: “Were the studies done properly?”
Inconsistency: “Do all studies point in the same direction?”
Imprecision: “Are the results precise enough to be useful?”
Indirectness: “Do the studies directly answer our question?”
Publication Bias: “Are we missing important studies?”

The grading process: Think of it like a report card. We start with an “A” (HIGH confidence) and subtract points for each problem we find. Serious problems drop us to “B” (MODERATE), then “C” (LOW), then “D” (VERY LOW). The more flaws we discover, the less we can trust the results.

Critical Flaws Overlooked by Peer Review

The Missing Data Bias: Investigators could get data from only half of eligible studies. This is not a minor limitation—it fundamentally threatens validity. Studies with weak or null albuminuria effects may have been less likely to share data.
The EMPA-REG Outlier Effect: Including the EMPA-REG OUTCOME trial gives R² = 0.47; excluding it, R² jumps to 0.72. When a single study shifts the conclusion from “weak-moderate” to “strong” association, the evidence is unstable—a house of cards.
The PROSPERO Problem: None of these studies were registered in PROSPERO before starting. For research intended to influence regulatory policy, this lack of transparency is concerning.

WHY STUDY REGISTRATION MATTERS?
Pre-registration: Researchers publicly post their analysis plan before starting.
Why it’s important:Prevents “cherry-picking” results or changing methods to get desired answers.
Example:If researchers say upfront “We’ll measure X, Y, and Z,” they can’t later ignore X and Y when they don’t support their hypothesis.
The problem:All three albuminuria studies were unregistered, making it impossible to know if they changed their approach after seeing preliminary results.

What Does a 30% Albuminuria Reduction Actually Mean for Clinicians?

Based on these three meta-analyses, if your patient achieves a 30% reduction in albuminuria:

Optimistic scenario: 45% reduction in kidney failure risk
Pessimistic scenario: 5% reduction, or even increased risk
Reality: We have no idea which scenario applies

For practicing nephrologists, the bottom line is to use albuminuria as a prognostic tool and treatment target—but do not assume that changes will directly translate into clinical benefit.

In summary: The next time someone claims that bigger datasets and fancier statistics will solve our surrogate endpoint problems, we won’t get excited. Instead, we’ll recall the albuminuria story: sometimes, sophisticated methods simply make the uncertainty more apparent—and that’s exactly what good science should do.

Optimising dialysis: can we stop racing against time?

What constitutes adequate dialysis for patients with end-stage kidney failure ? What dialysis goal will satisfy physicians, patients, and insurance providers? Should we focus on Kt/V, URR, nPCR, or the removal of middle molecules? Which therapeutic interventions could effectively reduce the high reported morbidity and mortality in the dialysis population? These are key questions for nephrologist.

Pivotal trials demonstrated that, beyond a certain threshold, increasing small solute clearance or enhanced clearance of the so-called middle molecules doesn’t offer much. This was clearly demonstrated by large trials in hemodialysis (HEMO, MPO, EGE) and CAPD (ADMEX, Hongkong trial). HDF is an attempt to further increase the clearances with the hope of improving otherwise static and not very reassuring outcomes of patients with kidney failure.

Although mortality rates have improved over decades, they still remain high. Will enhanced convective clearances fix this issue? While the CONVINCE trial shows it may, one should carefully consider the potential flaws that underlie this promise. Relatively healthy dialysis population (who could tolerate high convective volumes), lower than anticipated mortality rates decreasing the power to the study, important imbalances in the baseline characteristics (favoring HDF arm) and COVID 19 pandemic related deaths during the trial all make us feel that results need to be replicated before wide application. A cost utility analysis of HDF shows incremental cost of ∊27000 to 37000 per QALY (quality adjusted life year). If you kill the patient financially, that counts too!

One of the major limitations of the dialytic therapies is their intermittent nature which is only a far fetched imagination about “replacement” of kidney function. Intermittent therapies by their nature may not be adequate beyond reaching the arbitrary thresholds of clearances. In the initial era of dialysis, when membranes were low efficiency, therapy time was longer than that of the present day. Progressive shortening of the session duration, at least in part, was fuelled by the advent of urea kinetics and goals set by Kt/V. The idea, although not based on solid evidence, was quickly embraced not only by the hemodialysis community but was quite unnecessarily extrapolated to CAPD as well. With improved efficiency of the dialyzers, the target Kt/V could be delivered in a shorter time, falsely reassuring patients and health care providers. Dialysis adequacy assessment by Kt/V dominated the dialysis world for decades even when it was based on limited data and trials failed to show benefit of higher dialysis dose estimated by this ratio.

Even with lower dialysate and blood flow rates, why do patients on CAPD maintain well? Several theories were advanced to explain this curious phenomenon: middle molecule clearance, peak concentration hypothesis, and preservation of residual Kidney Function. Native kidney clearance can’t be equated to dialysis clearances, and nephron function is far too complex and intricate to be represented by urea kinetics. Two essential characteristics of native kidney clearance are their continuous nature and clearance of larger molecular weight solutes. Extending the duration of treatment by frequent hemodialysis offers several benefits as seen in FHN trials, indicating that extended dialysis delivers something beyond small solute clearance.

ISPD guidelines have radically moved away from Kt/V in their latest recommendation, which is a welcome step. Better late than never. Current ISPD guidelines advocate for a more patient-centered and less dictatorial approach to adequacy assessment. Why shouldn’t hemodialysis adopt a similar philosophy, prioritizing the patient’s needs over mere clearances?

After the initial improvement seen in the outcomes, CV disease has replaced the uremia as a major cause of death in this population and traditional approach to adequacy is oblivious to these complications of kidney failure. Even cardiologists have realised the crucial importance of therapy time and have advocated home based therapies with home hemodialysis and CAPD. Time seems to be a ‘non-negotiable’ factor in the delivery of adequate dialysis, and a slow but steady march (and not intermittent sprint) will help us win this race.

Given the fact that we may never be able to completely replace the kidney function by dialysis, and life expectancy will be limited by the comorbidities and CV disease, focus on mortality, although important, should not become an obsessive goal to be chased desperately. As an editorial puts it, there is more to life than not dying, and we need to keep this in mind when prescribing a dialysis regimen to our patient. While clearance target may still be a useful objective parameter, SONG HD has suggested several patient centered outcomes.

Dialysis adequacy has completed a circle: era when dialysis adequacy was assessed by looking at the patient, then focus shifted to urea, and now we need to move back to patient again. Probably another circle that is getting completed is of dialysis time: an era of low efficiency longer dialysis, to short high efficiency dialysis, and to back to longer dialysis therapies again.

Now that we buried and laid a tombstone to the concept of “chasing numbers”, we thought it was an opportune time to look at other advances in dialysis.

By coincidence, we came across the term: Expanded Hemodialysis (HDx). Medical representatives visiting our hospital claimed that HDx is a revolutionary concept in the history of dialysis. Intrigued but skeptical, we asked ourselves—what exactly is this?

expanded hemodialysis: what is it?

Expanded hemodialysis refers to the use of THERANOVA dialyzers—membranes designed with larger pores and a smaller internal diameter compared to conventional high-flux dialyzers. These design tweaks increase internal filtration and allow for greater removal of larger middle molecules. Although diffusion remains the primary mechanism, the dialyzer’s structure facilitates enhanced convective transport within the dialyzer (a process known as internal filtration, or water movement across the membrane within a closed circuit).

So far, so interesting. A dialyzer that mimics some of the benefits of online hemodiafiltration (HDF)—particularly improved middle molecule clearance—without requiring high convective volumes or special infrastructure? That piqued our interest. Especially since HDF, while promising, often needs excellent vascular access and higher blood flow rates.

But does HDx translate into better outcomes for patients?

We turned to the THREAD trial, the primary randomized controlled trial evaluating expanded HD with Theranova dialyzers. This multicenter study enrolled 80 incident dialysis patients (mean age ~63 years), randomized to receive either Theranova-based HDx or standard high-flux dialysis for 12 months.

Primary endpoint: change in residual renal function (mean GFR from urea and creatinine clearance).
Result: Over one year, the HDx group showed a slower decline in GFR:

Theranova group: –1.0 mL/min/1.73 m²/year
High-flux group: –2.4 mL/min/1.73 m²/year
Between-group difference: –1.4 mL/min/1.73 m²/year
(Least-squares mean difference: –1.4; 95% CI –2.4 to –0.5; p < 0.05)

Translated: HDx preserved kidney function by ~1.4 mL/min/1.73 m² more over one year than conventional HD. Also note the lower bound of 95%CI, the difference can be as low as -0.5ml/min/1.73m²

Interesting? Yes. Practice-changing? Not yet.

While the trial also looked at middle molecule clearance, that matters more in the realm of membrane science than clinical nephrology. For us, hard endpoints like mortality, hospitalizations, cardiovascular events, and quality of life matter more. One year follow up is too small to detect clinically meaningful changes.

Even residual renal function (RRF)—though important—is a surrogate outcome. But let’s give it the benefit of the doubt, assuming it may correlate with better long-term outcomes. Still, we had one question: Is a 1.4 mL/min/1.73 m²/year difference clinically meaningful?

Probably not. That’s not enough to justify a switch in practice, especially when therapy costs and logistical changes are considered.

Our enthusiasm further waned when we looked at absolute GFR values at discrete time points (baseline, 3, 6, and 12 months)—no significant difference between groups.

Confused? So were we. Here’s the key point:
The primary analysis used repeated measures mixed-effects modeling to compare the slope of GFR decline—not absolute GFR at fixed time points. Statistically, this approach showed a slower decline in the HDx group. But in real-world terms, there was no separation between the groups at the 12-month mark.

So, should you switch your patients to expanded HD?

Bottom line: HDx shows statistical significance—but lacks clear clinical significance. Until we see better evidence of improvements in hard outcomes, routine use of expanded HD is not justified.

We summarise the salient studies in dialysis in the table.

Strategy	Arms	comparison	Outcome	Studies Supporting
Increasing Kt/V	KT/V 1.25	KT/V 1.65	No improvement in mortality	HEMO Study
High-flux membranes	Low flux	High flux	No benefit	HEMO Study, MPO Study ,EGE study
Ultrapure dialysate	Standard	Ultrapure	no impact on mortality	EGE study
Hemodiafiltration (HDF)	High flux	HDF	Lower all-cause mortality	CONVINCE
Longer dialysis duration(increasing frequency)	3 times a week	6 times a week	Improved survival, better volume and BP control	FHN
Longer dialysis duration(increasing duration)	3 hours	More than 4 hours	Improved survival with longer	Cohort Study

So, What Should We Focus On?

The takeaway is simple: strategies like increasing Kt/V, changing membrane flux, or using ultrapure water are unlikely to impact survival. On the other hand, increasing dialysis frequency, and increasing duration, appear to offer survival benefits.

At the end of the day, the focus should be on what truly matters: improving quality of life, ensuring good nutrition, and maintaining fluid balance. Once we’ve tackled the main issues, we can look into Kt/V and similar measures if we have the time and resources. But really, chasing numbers shouldn’t be our priority.

The CPC Forum: Navigating dilemmas in Modern-Day Nephrology Practice KEM Hospital, Mumbai 26th April, 2025

The Clinico-Pathological Conference (CPC) is a case-based teaching-learning tool with a rich history, and remains an excellent exercise in clinical deductive reasoning. The origins of the fabled CPC can be traced back to Harvard Medical School, Boston (1898), forged in the crucible of cross-specialty interactions afforded by the illustrious campus, this time between a medical student, the future Professor of Physiology, Dr. Walter B. Cannon, and a law student. While supported by many reformists in medical education, the true renaissance occurred much later through Dr. Richard Cabot, who published the first ever series of CPCs from 1924 to 1935 in the Boston Medical and Surgical Journal, the erstwhile NEJM. That said, the enduring appeal of a CPC has been the measured, tactical progression from a patient’s presentation to a narrowed-down set of differential diagnoses, rather than it having to be a mere game of guessing the right answer. The CPC holds a passionate favour amongst many in the medical fraternity, and yet, to many others it represents a dying art, far strung from any need of revival.

One day in early February this year, the chief, Dr. Tukaram, put across a brilliant idea that left us, the team, squiggling with scholastic excitement. From the repertoire of everyday cases at KEMH, could we think of having a CPC conference of our own? Could we explore the format for a learning experience in nephrology of its kind? And it being the Centenary year (1926-2026) for our twin institutes- the Seth GS Medical College and the King Edward Memorial Hospital, could we kick off our department’s academic agenda for the year with this riveting meeting?

Before committing, we did give a good thought to the CPC format. Was it engrossing and effective enough? Relevant enough for the brave new world? In today’s thrilling era in nephrology, we often find ourselves turning to advanced diagnostics for the right answers. They form an essential toolkit for everyday diagnoses. With mass spectrometry steadily demystifying hidden molecular secrets, the barrage of potential biomarkers bolting towards finish-lines of approval, genetic analysis increasingly proving indispensable, and AI and equivalents here to stay (prefer it or not), can histopathology truly be the last word in a modern-day CPC? Shouldn’t the fantastic format evolve into a greater whodunnit, than the humble beginnings it had in the early years of the bygone century? Could we infuse the current zeitgeist and invigorate the CPC for a thrilling ride?

We rolled up our sleeves to turn this proposal into a reality. Did we have good cases? How many would make for a good program? How many would be too many? We launched into a collective scanning of our minds for intriguing case scenarios. Encounters which baffled us, called for investigations more than the kidney biopsy or autopsy, memorable for deeper clinico-pathological insights they sparked, in essence, cases that stayed with us long after.

We laid down the meter, divided an hour into segments of a CPC- primary presentation of the case by a presenter, case interpretation by a guest faculty, discussion by the moderators and audience, revealing of diagnosis, followed by a final round of discussion. Compelling case scenarios were shortlisted. Faculty invites were sent to the carefully curated list of prospective faculty-discussants and moderators. The august faculty, to our delight, accepted our invitations with great warmth! Weeks passed by as we discussed, debated and finalised each nitty-gritty of the event.

26th April, 2025 dawned soon enough. The spring-summer greens, in the entryway to GSMC’s heritage building, gleamed warmly under the golden sun. Yellow copper pod blossoms beautifully lay strewn on the college grounds in reciprocal frescos. The day’s proceedings commenced with Dr. Tukaram’s welcome address to the audience, both stationed in person and virtual, and an epilogue from the beloved, Dr. N. K. Hase, former head of the department and our patron.

The first set of chairpersons introducing the discussant for CPC Case 1– Dr. Bharat Shah, heralded the beginning of the day’s thrilling chase. He set out to tackle the mystery of a 33-year old kidney transplant recipient with unexplained progressive allograft dysfunction. A masterclass on making deductions in the face of uncertainty, he cracked the case by picking up ‘subtle (but persisting) tubular basement membrane thickening’ in the serial allograft biopsies provided to him. This was a rare involvement of the allograft by an MGRS (suspected recurrence). The case sparked an insightful discussion between the panelists, audience and the discussant, centered on the evolution of understanding of MGRS, utility of sophisticated techniques like immuno-phenotyping/flow-cytometry and the pervading disappointment with lack of effective therapies. CPC Case 1 served a quintessential modern-day diagnostic dilemma that mandates going beyond the prowess of a kidney biopsy for an answer.

CPC Case 2 was that of a 25-year old lady with rapidly progressive renal failure. The lady appeared to present with the well-known conundrum of malignant hypertension and renal-limited thrombotic microangiopathy. The puzzle here lay in connecting the dots uncovering the circumstance of glomerular injury. There were no definitive pointers in the history or clinical examination. Kidney biopsy showed an unexpected abundance of foam cells and a paradoxical presence of immunoglobulins and complement on immunofluorescence, defying the typical diagnosis. ‘Tough questions don’t last, but tough discussants do!’ Dr. Suceena Alexander brought to us a scholarly dissection of TMA occurring on the background of genetic glomerular disease, forcing us to expand the spectrum of both entities. A combination of histopathology (Dr. Amey Rojekar) and genetics (Dr. Anup Rawool) revealed the state of affairs, questioning our long-held views on pathogenicity of COL4A variants, while also highlighting the abstruse nature of complement dysfunction in glomerular disease.

CPC Case 3 was a wild ride of a 5-year old child with recurrent fever, pyuria and subnephrotic range proteinuria, who underwent a series of consultations for suspected UTI and unexplained renal involvement, which turned out to be granulomatous interstitial nephritis on kidney biopsy. Dr. Vaibhav Keskar, privy only to the findings of light microscopy of the biopsy, bravely charted the unknown seas, giving us an overview of renal involvement with each clue in the patient’s clinical course. However this was a true zebra! Finding zebra bodies on EM lead us to the diagnosis of Fabry’s disease, validated by the ɑ-galactosidase A enzyme test (the dried blood spot test) and genetics again!

‘We seek it here, we seek it there.’ We nephrologists basically seek it everywhere! Are we right in calling everything lupus/lupus-like? Do cryptic correlations lie just one layer beneath the garb of auto-immunity? CPC Case 4 made us question our beliefs when a 37-year old lady presented with unexplained anuria, and an assorted selection of ANAs on the immunoblot as the only faint clue to the unexpected histology- a membranous nephropathy, with necrotising vasculitis and cortical necrosis in a case of Sjogren’s syndrome (a tidy Occam’s razor situation). This was a ‘guess the kidney histopathology’ game like former times, and despite the treacherous pitch, Dr. Amit Langote, played a seasoned game at the googly! The panelists and the audience engaged in a spirited discussion, closing the case with many immediate take-aways for practice.

CPC Case 5 was the case of a 33-year old man with recurrent fever and nephrotic syndrome, who had ‘mild mesangial expansion’ on the kidney biopsy. Dr. Balan launched into the investigation by first ruling out differentials of ‘mild mesangial expansion’, and reaching to AA amyloidosis. Then he moved to a dissection of periodic fever syndromesautoimmune, autoinflammatory, and rheumatic. He kept us alive with speculation and in a masterful display of acumen and showmanship, uncovered the perpetrator very correctlythe Muckle Wells syndrome! Praises rang in from the panel and the floor for the brilliant Sherlock Holmes act, also hailed as evocative of an NEJM CPC by a virtual attendee. Kudos to his thorough approach, for he also discussed the most likely genetic variant that could be encountered in this particular patient! Now that’s the kind of clinico-pathological correlation we were hinting at when we promised this conference!

In a formidable moment after conclusion of case 5, we saw a sincere expression of gratitude from the patient’s mother, showing us how our decisions directly shape our patients’ worlds. She thanked Dr. N. K. Hase, the original Sherlock Holmes, who solved this case in real time in 2015. It being the esoteric diagnosis that it was, the patient had been extensively investigated and treated overseas, without any success. An encounter with Dr. N. K. Hase led to a presumptive treatment with colchicine, and further work-up in collaboration with the National Amyloid Centre, London, where the ensuing genetic report clinched the diagnosis. The patient was started on Anakinra, a transformational therapy in cyropyrinopathies, and remains symptom-free and in renal remission till date, a decade after the initial diagnosis.

The last segment of the conference was a Panel Discussion on two seemingly everyday cases in nephrology, with perplexing hints towards deeper pathological connections. Dr. N. K. Hase, Dr. Satish Balan, Dr. Suceena Alexander, Dr. Vinant Bhargava, and Dr. Anup Rawool gave their perspectives on

Case 1: Genetic variants in steroid-resistant nephrotic syndrome (LMX1B under the scanner)

Case 2: Role of complement in an intriguing overlap between MPGN and IgAN

This was a fitting closure, consolidating the message of the conference- Clinico-Pathological Correlation hides in plain sight. In our quotidian hustle at the patient’s bedside, it is up to us to decode the layers and contribute to advancing the scientific understanding of the disease at hand.

It was thrilling to have most of the audience stay with us right until the end of the proceedings! We closed the conference well in time, with a swell of gratitude towards our august faculty and audience for their enthusiastic participation throughout. A selection of books were gifted as souvenirs from the conference to the invited faculty. And we have reasons to believe that these books were loved as much by the receivers as well!

Amongst other things, the Herophilus Lecture Hall (Anatomy) proved to be a piece de nostalgia for many in the faculty and audience, a befitting opening venue for the Centenary Year celebrations of our department. The famous anatomist Andreas Vesalius’s thinking skeleton, pondering over a memento mori– a skull, is essential GSMC lore. On the day of the conference, many more got to appreciate this renaissance art-work. dense with meaning, proclaiming “Genius alone lives, all else is mortal”. (Pic 1 courtesy: Dr. Vinant Bhargava)

Dizzy with the day’s work, we may have ruined our one, single shot at getting a decent instagram-worthy picture of the forum! In hindsight, that stands as the only regret from the conference day that we haven’t been able to override. Here’s some sincere hoping that we get to do an encore someday!

Until next time!

-Sayali B. Thakare

Link to the recording: https://vimeo.com/1081895295/5eb0692fe7?ts=0&share=copy

Malignant Hypertension: Navigating Ambiguity of Definitions and Challenges in Management

32-year-old gentleman is referred to you by an ophthalmologist, who saw him for visual blurring. His BP is 230/120 mmHg, and serum creatinine is elevated to 2.1mg/dl. Urinalysis shows trace/1+ protein, and bland sediments. Kidneys appears normal on ultrasound except mild increase in the cortical echogenicity. A kidney biopsy reveals ischemic looking glomeruli, severe vascular intimal hyperplasia, onion skinning and one of the vessels shows thrombus. This is a classic phenotype of ‘Malignant Hypertension (MHTN)’. MHTN is supposedly a rare form of hypertension but remains a regular encounter in our clinical practice and some recent reports even indicate that incidence might have recently increased.

When the term ‘malignant hypertension’ was first used (Volhard and Fahr in 1914), it implied steady downhill course and death within months, justifying the word ‘malignant’. Severely elevated blood pressure, typically above 120 diastolic with associated systemic endothelial dysfunction manifesting as grade 3/4 retinopathy, and AKI characterized the initial presentation of these patients. Most malignant tumors today can have a reasonable prospect of cure/control, thanks to advances in cancer diagnostics and therapeutics, can we say the same about malignant hypertension after 50 years of its initial description? Answer is probably, ‘we don’t know’.

Fouad Amraoui et al reported the outcome of 120 patients with malignant HTN and documented poor outcomes for this condition: the annual incidence of all-cause mortality per 100 patient-years was higher in MHT patients (2.6) compared with normotensive (0.2) and hypertensive (0.5) controls (both P<.01). All-cause mortality was 10% at 5 years in patients with a mean age of 44 years, while 20% needed dialysis. Interestingly, patients with MHTN had a more favourable cardiovascular risk profile when compared with hypertensive controls, indicating that excess risk of death was driven mainly by renal involvement.

‘Acute hypertensive microangiopathy’ -a term proposed by 2018 European consensus appear to describe the old and rather scary term ‘malignant hypertension’ better. As was the case with acute kidney injury, problem starts with definition, which impedes the progress in the field. We aren’t sure if everyone talks about the same condition when they say, ‘malignant hypertension’. ISH, NICE, ESC/ESH, and a European definition of the term varies from each other, with one common component being diastolic BP elevation to >120mmHg. Mandatory need of papilledema is also questioned by some if microvascular injury is apparent in any other organ. Latest AHA guideline have shown their heightened sensitivity to the BP management by lowering the BP cutoff to classify someone as hypertensive (and thereby declaring half the world’s population as diseased!), however, they are remarkably silent about the issue of malignant hypertension reflecting a major oversight and misguided attention by experts. Primary hypertension is not a disease in itself but the risk factor for CVD and associated mortality, but MHTN is probably that rare scenario where hypertension itself becomes the disease and becomes immediate threat to vital organs and life.

Endothelial injury by severe sheer stress, vasoconstriction leading to end-organ ischemia, severe activation of renin angiotensin aldosterone system (RAS), activation of procoagulant and proinflammatory cascades-all join hands to perpetuates organ damage in this situation. However, central to the pathophysiology of MHTN is intense activation of RAS, and in their seminal work, Laragh noted that the aldosterone levels recorded in these patients often exceeded the levels seen in primary hyperaldosteronism. My endocrinologist colleague (we often debate as team Laragh versus Conn) call this disease ‘kidney seizure’ which seems appropriate, as in many of these cases, after initial bang, hypertension curiously becomes less severe with time, which coincides with decrease in plasma renin activity but persistently elevated aldosterone levels. Laragh termed this phenomenon ‘tertiary hyperaldosteronism’ hypothesizing that elevated renin drives aldosterone producing cells into an “autonomous mode’ and cells continue to fire even after renin has stopped commanding them.

Most of these patients are not biopsied unless a suggestion of proliferative GN is high. This is probably due to severe HTN which increases the bleeding risk and the fact that biopsy findings are unlikely to change immediate management. Characteristic histological features include severe vascular changes, glomerular ischemia, and sometimes thrombotic microangiopathy.

While reviewing the literature about this condition, it’s fascinating to see the perseverance and hard work of the nephrologists in earlier times who tried to elaborate the pathophysiology of the disease. It is equally astonishing to see the lack of major research activity in this area in recent times. A recent publication in JAMA caught our attention. In a prospective cohort study investigators wanted to see if there are different kidney outcomes among patients diagnosed with malignant hypertension (mHTN)–associated thrombotic microangiopathy (TMA) who received sacubitril/valsartan(ARNI) compared with angiotensin-converting enzyme inhibitor/angiotensin receptor blocker (ACEI/ARB) therapy. They found that with 217 consecutive patients with mHTN-associated TMA found that ARNI was associated with a shorter time to achieve a composite of kidney recovery outcomes, including a 50% decrease in serum creatinine level, a decrease in serum creatinine level to the reference range, or kidney survival free from dialysis for more than 1 month compared with ACEI/ARB therapy. They concluded that sacubitril/valsartan could be a promising therapeutic approach to improve kidney recovery in patients with mHTN-associated TMA. So far so good. But a closer look at the internal validity of the study gave raise to many questions.

We appraised the Population, Exposure, and Outcome (PECO) framework of the study and found several inconsistencies and possibility of major methodological errors.

Population Concerns: The study includes patients with malignant hypertension who underwent renal biopsy with a median serum creatinine level reported is 7.9 mg/dL. It is unclear when these renal biopsies were performed, as such elevated creatinine levels makes an early biopsy unlikely. Clarification is required regarding the timing of and the associated with biopsy in these patients.

Exposure Issues: The cohort was started on ARBs/ARNI within 72 hours of admission, with some patients already on dialysis. Initiating ARBs or ARNIs in patients with acute kidney injury (AKI) is not standard practice due to the potential risk of hyperkalemia and delaying recovery of AKI. Moreover, the availability of ARNI in 2008 is questionable. USFDA approved its use in 2015.

Outcome Ambiguities: The study uses a composite outcome, but fails to report the results of individual components, which could lead to misleading conclusions. Additionally, the concept of a “tertiary outcome” is mentioned without adequate explanation. The outcomes in a cohort study should be clear and specific. For instance, a 15% or 50% decrease in creatinine is difficult to measure without a predefined follow-up period, as is standard in randomized controlled trials. Was the outcome time to recovery, or the number of patients who recovered? This needs clarification.

Baseline Characteristics and Statistical Concerns: There are several oddities in the baseline characteristics. Fifty percent of patients were reportedly on sulodexide which is not a standard of care for this condition, and renal biopsies were performed on all consecutive patients, which seems impractical. The study mentions using the 2019 guidelines for managing malignant hypertension in 2008, as well as calculating eGFR using the CKD-EPI equation, which was not published until 2009. This is possible only if some investigotor has learnt the art of time travel! Calculating eGFR in AKI is not recommended.These anachronisms suggest either significant errors or potential data fabrication.

Follow-Up and Exposure Measurement: Follow-up details are unclear. If the exposure to ARNI/ARB is not consistent, then the outcomes may not be valid introducing classification bias.

The inconsistencies in the data suggest that the study may be unreliable, and peer review may have overlooked critical flaws.

There is significant knowledge gap in this area. It is not understood why some hypertensive patients progress to ‘acute hypertensive vasculopathy’? Some cases may be attributed to non-compliance to the antihypertensive drugs, but not all of such patients progress to MHTN. In addition to sheer stress on the vessel walls due to acutely elevated blood pressure, other factors like endothelial dysfunction, inflammation, complement pathway activation are considered important. There are striking similarities in the pathological features of MHTN and atypical HUS, and defective regulation of the complement pathway is of particular interest.

There are no evidence based guidelines for managing acute phase of MHTN. Usually clinicians resort to initial intravenous therapy for blood pressure control to a target level(usually within 8 to 24 hours) followed by oral drugs. We would like to bring to notice of our readers a strategy of using RAS inhibition in acute setting. In a cohort study, 168 patients with MHT were treated with ACEI: 1.25 mg ramipril with close monitoring of BP, followed by forced titration (2.5 mg at 6 h, 5 mg at 12 h, 10 mg at 24 h, and maximum dose of 10 mg twice daily for the second day)with a target SBP between 140 and 180 mmHg within the first 48 h. The 5-year survival without end stage renal disease was 90% in the cohort, similar to the results of other studies using intravenous treatments at the initial phase of MHT. RAS inhibitors could be the drug of choice in MHTN, yet they remain underutilized. It’s not uncommon to see patients on 5-6 different antihypertensives but lacking RAS inhibitors, likely due to an unfounded fear of using this class once kidney function declines beyond a certain threshold (which varies but is often overly conservative). We should be more proactive, leveraging hypokalemia as an opportunity rather than a barrier.

Clearly, MHTN awaits a better uniform definition, carefully designed cohort studies to understand risk factors and outcomes, and the basic science research to understand the molecular and genetic basis.

October 2024

Dawn of the precision medicine era in management of primary podocytopathies

“We recommend using cyclosporine or tacrolimus as initial second-line therapy for children with steroid-resistant nephrotic syndrome (1C)”, recommends KDIGO 2021 guidelines. “Rule of thirds” is a favorite question in examination, 1/3 patients will achieve complete remission, 1/3 will achieve partial response and 1/3 won’t respond (although a favorite of examiners, I’m still to find credible reference for these numbers!). Only difficulty with this seemingly easy proposition is that it is wisdom of hindsight. We can’t know beforehand which patient belong to the last 1/3, who will be harmed by trial-and-error approach. Children with cushingoid features, hypertrichosis, and steroid associated eye and bone disease is not an uncommon scenario in GN clinics.

Things will probably be better for these patients in coming years, this exciting paper in KI may well herald ‘the times they are changing!’. Primary podocytopathies are traditionally viewed as ‘pauci-immune’-no immunoglobulin/complement deposits on immunofluorescence, but this may not be true, and no IF staining doesn’t mean no immune deposits. In fact, many patients with primary podocytopathies have variable, weak staining of immunoglobulins, and at least in a subset of patients with this condition, it may not be nonspecific trapping. Discovery of autoantibodies targeting ‘nephrin’ in a subset of patients with MCD was a great leap forward in understanding of the autoimmune basis of the podocytopathies. International Society of Glomerular Disease team elegantly documented the role anti-nephrin antibodies in primary MCD/FSGS. In their large multicenter study involving 539 patients (357 adults and 182 children) and 117 controls, they demonstrated presence of these antibodies in 46 of the 105 (44%) with minimal change disease, 7 of 74 (9%) with primary focal segmental glomerulosclerosis, 94 of 182 children with idiopathic nephrotic syndrome, only in rare cases among the patients with other conditions. Further, they found strong correlation between the clinical disease severity (proteinuria) and antibodies. To rule out the possibility of antibody formation as a consequence of podocyte damage (rather than being a cause of podocytopathy), they found development of anti-nephrin autoantibodies and nephrotic syndrome in mice that was immunized with nephrin. The localization of IgG at the slit diaphragm and its partial colocalization with nephrin indicated antigen–antibody binding and such binding increased phosphorylation of nephrin at tyrosine residue Y1191 (this corresponds to Y1176 in human nephrin). Consequent to increased phosphorylation of nephrin, cytoskeletal reorganization take place in podocytes. Downstream processes ultimately producing proteinuria were elucidated by proteomic data which suggested an increase in nephrin endocytosis and disruption of the slit diaphragm.

While pathogenesis involving antigen antibody interaction is similar to other GNs like membranous nephropathy, crucial difference is lack of deposits on traditional immunofluorescence microscopy. This may have to do with different downstream pathways activated in primary podocytopathies (endocytosis of the slit diaphragm proteins and resultant disruption of its function), versus complement pathway activation and damage in other GNs.

Valentina Raglianti et al in their latest paper in KI report a cohort of 128 pediatric patients with MCD/FSGS. Anti-nephrin IgG ELISA assay in the serum and stimulated emission depletion microscopy of kidney biopsies showed IgG-nephrin co-localization only in 77.8% of cases (this was still lower in 44.4% in cohort of 48 adult patients with MCD or FSGS), suggesting role of other target antigens in the slit diaphragm region. Such patients (anti slit antibody positive, exact antigen remains to be characterized), were more likely to have severe nephrosis, have less chronic changes on biopsy, more likely to be secondary steroid resistant than primary (87 versus 27%), more likely to respond to second line immunosuppression (92% versus 20%) and have best prognosis amongst SRNS (irrespective of histological pattern of MCD or FSGS). None of the patients with anti-slit antibodies developed renal failure. The only anti-slit positive patient who was multidrug resistant was a diagnosed case of Nail-Patella syndrome who developed proteinuria recurrence after kidney transplantation.

This study is a long-awaited step forward in refining treatment decisions in steroid resistant primary podocytopathies and high-resolution confocal microscopy of the kidney biopsies and serum testing for anti-slit antibodies (in addition to anti nephrin ab) is useful tool in management. Although limited by small sample size and bias towards SRNS patients (but that’s what bothers us the most), this is beginning of an exciting era in the personalized management of primary podocytopathies. Large-scale evaluation of this approach is urgently needed.

These two research articles in basic science defining the role of anti-nephrin and anti-slit antibodies created a lot of buzz in nephrology circles. Adding to this excitement was a randomized control trial in treatment of MCD patients with severe oedema. Did the trial live up to our expectations?

SPINning the Stats: When a Low Power Study Goes for a Victory Lap

A single-center, open-label, parallel-arm randomized controlled trial from China randomized 117 patients with Minimal Change Disease to receive conventional oral steroids or 2 weeks of intravenous methylprednisolone followed by oral steroids.

Remission (complete and partial) in the sequential group was higher than in the oral group after treatment for 2 weeks and 4 weeks (P=0.032, P=0.027). Complete remission rate was higher in the sequential group than in the oral group (63.3% vs. 41.5%, P=0.031) after treatment for 2 weeks. The time to achieve complete remission was shorter in the sequential group than in the oral group, with a statistically significant difference (14.0 days, 95% CI [13.5 to 14.5] vs. 16.0 days, 95% CI [12.7 to 19.3], P= 0.024).

The authors concluded that highly edematous Minimal Change Nephrotic Syndrome (MCNS) patients who were treated with sequential intravenous methylprednisolone therapy followed by oral prednisone achieved earlier remission than those treated with the conventional oral prednisone regimen.

At first glance, the RCT looks all good and fine. But stay with us and read on—we will take you through some of the issues with the trial.

Misleading title: WHO defines adults as people above the age of 19 years. The editors of NDT, or the authors, may have a different definition. They enrolled patients above 14 years old but chose the title as “Adults”!

Internal validity of the trial: We found some serious concerns about the internal validity of the trial. We list a few of them below.

Did intervention and control groups start with the same prognosis?	Yes
Were patients randomized?	Yes
Was randomization concealed?	No
Were patients in the study groups similar with respect to known prognostic factors?	Yes
Was prognostic balance maintained as the study progressed	Yes
To what extent was the study blinded?	No
Were the groups prognostically balanced at the study’s completion?	YES
Was follow up complete?	13.5% loss to follow up
Were patients analyzed in the groups to which they were randomized?	Yes
Was the trial stopped early?	May be (enrolled 117 patients only against a planned 534)

Gender, age, and GFR of trial participants:Ninety percent of those enrolled were male. The number of patients below 18 years of age is not mentioned in the manuscript. The GFR of enrolled patients appears in 2 rows with different values—why is that so? see the highlighted part in table below

Results are not presented in a manner for practical use: We found it very difficult to decipher what the “PRIMARY OUTCOME” of the trial was and when it was supposed to be measured. This was not defined in the “TRIAL REGISTRY”. Check it out in the registry here.

Since the time point at which the outcome is measured was not defined, the authors performed tests at 1 week, 2 weeks, 4 weeks, 8 weeks, and 12 weeks. They found that the results were positive at 2 and 4 weeks, and hence decided to consider those results as positive. This is a classic case of “cherry-picking, HARKing, and SPIN”. Dr. Tukaram, founder of LMIN, remarked: “Hit with an arrow first, then draw the target wherever the arrow hits.”

Well, the problems do not end there. There is no mention of treatment effect or effect estimate anywhere. Usually, study results are expressed as “Relative Risk along with Confidence Interval”. We went through the manuscript multiple times—alas! We could not locate this vital information. We decided to calculate these values ourselves, but we did not know how many patients were included in the final analysis. We thought it was 117 patients, but to our dismay, it was 102. See the tables below:

Table 1. Calculated with n= 117

	Remission	No remission	Total
Sequential	36	21	57
Oral	25	35	60

P value 0.0264

Table 2. Calculated with n= 102

	Remission	No remission	Total
Sequential	31	18	49
Oral	22	31	53

P value 0.0314

Adverse events and harms: There are major concerns here:

1. Hepatotoxicity: 14 of 102 patients developed hepatotoxicity. This is unusual. How would a DSMB allow a study to continue with 13.5% of participants developing hepatotoxicity?

2. Vision loss: 4% of participants in the intervention group experienced vision loss. This is something to be worried about.

3. Purple stripes: More than 50% of trial participants had “purple stripes.” What does this even mean? Are the authors referring to “striae”, a side effect of steroid therapy?

Underpowered study: The planned sample size was 534, but the final study enrolled only 117, and 15 were lost to follow-up. The “post-hoc power” of the study is approximately 51.2%, which is below the conventional threshold of 80% for sufficient power.

Unusual causes of relapse: Table 3 of the manuscript lists some unusual reasons for relapse. We were amused to find that “herbs” caused relapse in 8% of cases, “exhaustion” caused relapse in 17% of cases, “smoking” in 8%, and “stay up” in 13%. What is “stay up”? We don’t know! We give up—please let us know if you decipher it.

Take home message: The study is underpowered to detect significant differences between the treatment arms. Multiple testing and obscuring the effect estimate in the publication further diminish confidence in the results. This trial could serve as yet another case study for a hypothetical book titled: “SPIN-vincible: How to Win a Trial Without Enough Data.”

Are HIF PHI safe in anemia of chronic kidney disease?

Background: The efficacy of HIF-PHI inhibitors in treatment of anemia of chronic kidney disease has been proven in many randomized control studies. But, there are concerns about the safety of these drugs. The safety concerns are particularly more in CKD-ND. This prompted US FDA not to approve this drug in this patient population.

In August 2024, NEJM-EVIDENCE published a systematic review and meta-analysis (SRMA) which concluded that there is no evidence of difference in long-term cardiovascular safety profile of HIF-PHI and ESA in adults with dialysis dependent CKD and adults with non-dialysis dependent CKD. (emphasis is ours). This is quite a strong conclusion. Take a moment here and read the conclusion of 2022 Cochrane systematic review on HIF-PHI: HIF stabilizer management of anemia had uncertain effects on CV death, fatigue, death (any cause), CV outcomes, and kidney failure compared to placebo or ESAs. These discrepant conclusions piqued our interest. What are we missing here? We decided to do a critical appraisal of the NEJM-EVIDENCE SRMA. We wanted to see if the SRMA will stand up to the review standards. We used AMSTAR 2 framework to appraise the review. The table 1 below summaries the appraisal.

TABLE 1 AMSTAR-2 Items

Item	Item type	Description	Responses
1	Noncritical	Did the research questions and inclusion criteria for the review include the components of PICO?	Yes
2	Critical	Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review and did the report justify any significant deviations from the protocol?	No
3	Noncritical	Did the review authors explain their selection of the study designs for inclusion in the review?	Yes
4	Critical	Did the review authors use a comprehensive literature search strategy?	Yes
5	Noncritical	Did the review authors perform study selection in duplicate?	Yes
6	Noncritical	Did the review authors perform data extraction in duplicate?	Yes
7	Critical	Did the review authors provide a list of excluded studies and justify the exclusions?	Yes
8	Noncritical	Did the review authors describe the included studies in adequate detail?	Yes
9	Critical	Did the review authors use a satisfactory technique for assessing the risk of bias (RoB) in individual studies that were included in the review?	Yes
10	Noncritical	Did the review authors report on the sources of funding for the studies included in the review?	Yes
11	Critical	If meta-analysis was performed, did the review authors use appropriate methods for statistical combination of results?	No
12	Noncritical	If meta-analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta-analysis or other evidence synthesis?	Yes
13	Critical	Did the review authors account for RoB in individual studies when interpreting/ discussing the results of the review?	Yes
14	Noncritical	Did the review authors provide a satisfactory explanation for, and discussion of, any heterogeneity observed in the results of the review?	Yes
15	Critical	If they performed quantitative synthesis, did the review authors carry out an adequate investigation of publication bias (small study bias) and discuss its likely impact on the results of the review?	No
16	Noncritical	Did the review authors report any potential sources of conflict of interest, including any funding they received for conducting the review?	Yes
	Overall	AMSTAR-2 criteria for overall confidence -High confidence: ≤1 non-critical weaknesses -Moderate confidence: ≥2 non-critical weaknesses, with no critical weaknesses -Low confidence: 1 critical weakness, with or without non-critical weaknesses –Critically low confidence: ≥2 critical weaknesses, with or without non-critical weaknesses	*Critically low confidence*

The above table shows that the SRMA has many critical issues, and overall confidence is CRITICALLY LOW. Honestly, we were surprised to see these many problems in a SRMA published in NEJM-EVIDENCE. In fact, we had to ask each other(and other experts whom we knew) if we were missing something. How can so many critical issues be ignored by the reviewers of a prestigious journal.

You need not believe us upfront. You can read our explanation below and decide for yourself if the SRMA has been conducted with methodological rigor.

Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review and did the report justify any significant deviations from the protocol?

No.

Explanation: The protocol states that all chronic kidney disease (CKD) patients would be analyzed together, with dialysis-dependent and non-dialysis patients being analyzed in subgroups. However, in the actual review, the analyses were conducted separately for dialysis-dependent and non-dialysis CKD patients, with no combined analysis of all CKD patients. This SUBGROUP analysis has been published as primary outcomes. This represents a deviation from the protocol, as the original plan to analyze all CKD patients together was not followed in the final review. This change could have implications for the interpretation of the findings, as the separation of these groups may lead to different conclusions than if they were analyzed together.

There were many other subgroup analyses planned in the protocol. But the authors have not to performed subgroup analyses and have not given any valid reason. The authors say they were “unable” to perform subgroup analysis!

2. If meta-analysis was performed, did the review authors use appropriate methods for statistical combination of results?

No.

Before you read further see this FOREST PLOT.

This forest plot is from the supplement of the study. In figure S12, we notice “NA” for PROTECT trial. This means to say data was NOT AVAILABLE. But with no data available, the authors have calculated Risk Ratio of 1.10 with 95% CI of 0.96 to 1.27. They also go ahead and give a weight of 34.47% to a study which is contributing no data at all. It does not stop here. See the third row as well. PROTECT ESA-treated study also contributes no data but provides a RR of 0.90 and weight of 32.45%. Let this sink in. Two studies which contribute absolutely no data to composite kidney outcome are given 67% weightage.

This does not inspire confidence in the way the statistical analyses have been conducted. For more examples and fun, see the supplementary appendix and have a look at figure S5, S7, S8, S21, and S24.

3. If they performed quantitative synthesis, did the review authors carry out an adequate investigation of publication bias (small study bias) and discuss its likely impact on the results of the review?

The review included 25 studies in the final analysis. But the authors have not assessed publication bias. GRADE analysis requires assessment of publication bias. We are not sure how GRADE analysis was done without this data.

This was not all. Now let us look at some other errors in data collection and analysis. When the meta-analysis software is fed with wrong data, the results will also be wrong, something like-garbage in garbage out (GIGO). Now refer the table below for the demonstration

Look at the numbers highlighted in yellow. The authors of systematic review report a 100% drug discontinuation rate in the DIALOGUE 3 and DIALOGUE 5 trials, which seems to be an oversight. After all, who would conduct a trial with a 100% discontinuation rate?

Typically, data for meta-analysis is collected in duplicate to avoid errors. The situation here seems like a comedy of errors—eight authors missed it, the reviewers missed it, and even the editors missed it. A quick check of the trial data shows that, according to the DIALOGUE 5 trialists, “A larger proportion of patients withdrew from molidustat treatment (23% [13/57]) than from epoetin treatment (7% [2/30]) (6).”

These were some of the problems in this SRMA. Though we understand that the authors have put in lot of efforts in conducting this SRMA, we are concerned about the process of the review. We want to emphasize that articles published in best of the journals may not be immune to serious errors. There is a need to strengthen the peer-review process so that the published studies which inform clinical practice live up to certain minimum standards.

Conclusion: The results of this systematic review are better ignored. The safety of HIF-PHI in CKD-ND is still not proven.

(We would like to acknowledge contributions of Dr Indu Rao and Dr Ravindra Prabhu from Department of Nephrology, Kasturba Medical College, Manipal for helping us in the critical appraisal of the review)

May 2024

Impact of clinical practice guidelines on hospital admissions and mortality in people living with kidney disease

The goal of clinical practice guidelines (CPG) is to lessen the burden of disease on the community and the individuals by reducing hospitalization, morbidity, and mortality. Whether the CPGs in chronic kidney disease are achieving this goal or not is an important question. Whether guidelines are getting translated into better outcomes for patients is the question answered by a cluster RCT by ICD-Pieces Study Group.

In this open-label, cluster-randomized trial, the investigators assigned 11,182 patients with the kidney-dysfunction triad who were being treated at 141 primary care clinics either to receive guideline-based interventions or to receive usual care. The primary outcome was hospitalization for any cause at 1 year. Secondary outcomes included emergency department visits, readmissions, cardiovascular events, dialysis, and death.

If you have not read the trial earlier, we want you to pause a bit and think before reading further. What do you think would have been the results of the trial? If you are like us, you would be thinking that CPGs would have made a SIGNIFICANT difference in the outcomes. Well, only one of us-Dr Tukaram- was skeptical regarding the effect of CPGs.

The results were alarming! The hospitalization rate at 1 year was 20.7% (95% confidence interval [CI], 19.7 to 21.8) in the intervention group and 21.1% (95% CI, 20.1 to 22.2) in the usual-care group (between group difference, 0.4 percentage points; P=0.58). The risks of emergency department visits, readmissions, cardiovascular events, dialysis, or death from any cause were similar in the two groups. The risk of adverse events was also similar in the trial groups, except for acute kidney injury, which was observed in more patients in the intervention group (12.7% vs. 11.3%).

Let that sink in! CPGs did not make any difference in outcomes. In fact, adverse events like AKI increased in the intervention group.

The trial has several strengths. It has a pragmatic study design with a large patient number. The intervention used a personalized algorithm (based on the patient’s electronic health record) to identify patients and practice facilitators to assist providers in delivering guideline-based interventions

One may argue that this may not be the correct way to deliver evidence based interventions. But to refute the results of this study, there is a need for another pragmatic trial which has more efficient delivery of CPGs. Until then, we must make do with the fact that CPGs may not be a panacea. The tapestry of combined wisdom gleaned from the trials, scripted by experts, and etched in gold may not deliver what we hope for.

Sparsentan reduces proteinuria in FSGS and IgAN, but is that enough?

Glomerular diseases are the leading cause of ESRD in the developing world where diabetes is fast catching up or it would be more apt to say that diabetes has already caught up and GN hasn’t loosened its grip a bit. After a long period of quiescence, the field of GN management is finally witnessing long awaited controlled trials, testing various treatments strategies to prevent or delay kidney failure in this population.

About 50% of patients with primary FSGS don’t attain remission with available options and have high risk of ESRD. Modulation of the common downstream injury pathways involving hypertension, glomerular hyperfiltration, vasoconstriction, and fibrosis, has been the cornerstone of the treatment. SGLT2i and MRAs were important additions to the RAS inhibition in recent years.

A dual endothelin–angiotensin receptor antagonist, sparsentan, has shown promising reduction in the proteinuria in a phase 2 study, and was evaluated in the phase 3, DUPLEX clinical trial involving 371 patients across 240 centers (mainly USA). At the median follow-up of 2 years, primary efficacy endpoint -the eGFR slope at the time of the final analysis at 108 weeks- was similar in two groups: the between-group difference in total slope (day 1 to week 108) was 0.3 ml per minute (95% CI −1.7 to 2.4), and the between-group difference in the slope from week 6 to week 108 (i.e., chronic slope) was 0.9 ml per minute (95% CI, −1.3 to 3.0). The mean change in eGFR from baseline to week 112 was −10.4 ml per minute with sparsentan and −12.1 ml per minute per 1.73 m2 with irbesartan (difference, 1.8 ml per min, 95% CI, −1.4 to 4.9). Proteinuria reduction at prespecified interim analysis conducted at 36wks was more with sparsentan than irbesartan: partial remission of proteinuria was 42.0% in the sparsentan group and 26.0% in the irbesartan group (P=0.009).

Trial failed to show the impact of the drug on the primary endpoint of kidney function but demonstrated a reduction in proteinuria. Limited representation of blacks, reliance on surrogate endpoints of eGFR change and proteinuria reduction, limited duration of follow up, high discontinuation rate in both the arms are significant limitations. Very few patients had features or nephrosis (edema, hypoalbuminemia) and unfortunately, almost ⅓ of the patients enrolled didn’t have electron microscopy available, implying that this was a heterogeneous population inadvertently including many patients with secondary and genetic FSGS.

FSGS as a cause of ESRD has increased tenfold in the United States (0.2% in 1980s to 2.3% in 2000), unlike the rest of the world where such increases are not observed. This is likely to do with the misclassification of histological lesions of FSGS (which can be practically observed in kidney disease affecting almost any of the glomerular, tubulointerstitial or vascular compartments of the kidney) as primary FSGS. Other background information on the disease like severity of associated HTN, response to the prior immunosuppressive treatments, trajectory of eGFR before enrolment, important histological characteristics like glomerular, tubulointerstitial and vascular chronicity changes- which would significantly affect the natural history of the disease irrespective of the intervention -is not available. Authors hope that longer follow up may discover more substantial benefit on eGFR but it is equally likely to reveal side effects of sparsentan.

Sparsetan was also studied in IgA nephropathy (with proteinuria >1gm and eGFR 30-90ml), in a phase 3, PROTECT trial, randomizing over 400 patients to sparsetan or irbesartan. The primary endpoint was proteinuria change between treatment groups at 36 weeks favored sparsentan :change from baseline in UPCR was 40% lower in the sparsentan group than in the irbesartan group (−42·8%, 95% CI −49·8 to −35·0, with sparsentan versus −4·4%, −15·8 to 8·7, with irbesartan; geometric least-squares mean ratio 0·60, 95% CI 0·50 to 0·72). Patients in the sparsentan group had a slower rate of eGFR decline than those in the irbesartan group: eGFR chronic 2-year slope (weeks 6–110) was −2·7 mL/min per year versus −3·8 mL/min per year (difference 1·1 mL/min 95% CI 0·1 to 2·1; p=0·037); total 2-year slope (day 1–week 110) was −2·9 mL/min per year versus −3·9 mL/min per year (difference 1·0 mL/min per year, 95% CI −0·03 to 1·94; p=0·058).

While DUPLEX investigators remain equivocal about the benefits of sparsentan as an option to renoprotection, those of PROTECT trial seem to be convinced. The least important part of the research paper is ‘discussion’, especially when written by medical writers appointed by sponsors. But it can sometimes reveal important biases. Authors of the PROTECT trial express their disappointment twice in this section by stating “P value for total eGFR slope narrowly missed statistical significance” rather than simply stating that there was no significant difference in eGFR. While both trials are important and welcome addition to the scarce evidence base to manage common GNs, conclusions should be viewed alongside the important limitations. Interestingly, more than half of the IgAN patients in the PROTECT trial were normotensive; this is unlike most patients with IgAN especially when other features like significant proteinuria, decreased GFR, and hematuria were present. High discontinuation (23% in irbesartan and 14% in irbesartan), higher use of rescue treatment in the irbesartan group (did they have higher disease activity/severity?) are other bothersome observations. Similar to DUPLEX, detailed reporting of the histological severity (MEST-C) or International IgA nephropathy score, information on treatment and responses prior to randomization are missing. Higher incidence of dizziness and hypotension were reported in the sparsentan arm of the PROTECT trial. Side effects attributed to fluid retention: edema, hypertension and heart failure were not observed in the trial. This might be related to the specific safety of sparsentan (unlike other endothelin receptor blockers where these are reported in >15% patients). However, high discontinuation rate in both the arms of the trial make this hypothesis less than convincing. One needs to be watchful about hepatotoxicity although no such signal was observed in the PROTECT or DUPLEX trial. Of note, pivotal trials of ERAs also failed to pick up this safety signal which emerged only during post marketing surveillance. Finally, the perpetual question that bothers me is this: why can’t trials spending so much money measure BP and GFR by gold standards-24hr ABPM and measured GFR respectively? Larger studies with longer follow up are needed to better clarify efficacy and safety of this drug in FSGS and IgAN.

Beginning of a new era in GN management?

An interesting development in the therapeutics of IgA is here with the publication of the APRIL trial. This is arguably the first example of the immunosuppressive agent specifically developed for a GN (IgAN) which until now were borrowed/ repurposed from rheumatology or oncology. APRIL-a proliferation-inducing ligand- is a member of the TNF α superfamily, regulates IgA production by B-cells and is implicated in the pathogenesis of IgAN. Sibeprenlimab (VIS649) is a humanized IgG2 monoclonal antibody that binds to APRIL and produce reversible and dose dependent decrease in the serum levels of IgA, galactose-deficient IgA1, IgG, IgM, and APRIL.

In this phase 2 study involving 155 patients with IgAN (proteinuria >1gm in 24 hour urine and eGFR >30ml/min on standard of care) safety and efficacy of sibeprenlimab at various doses (2, 4, or 8 mg per kilogram of body weight) was compared to placebo with regards to primary endpoint of change in proteinuria at 12 months. A significant reduction in proteinuria was noted: geometric mean ratio reduction (±SE) from baseline in the 24-hour urinary protein-to-creatinine ratio was 47.2±8.2%, 58.8±6.1%, 62.0±5.7%, and 20.0±12.6% in the sibeprenlimab 2-mg, 4-mg, and 8-mg groups and the placebo group, respectively. Similarly eGFR was better presented with drug: the least-squares mean (±SE) change from baseline in eGFR was −2.7±1.8, 0.2±1.7, −1.5±1.8, and −7.4±1.8 ml per minute in the sibeprenlimab 2-mg, 4-mg, and 8-mg groups and the placebo group, respectively. Higher incidence of nasopharyngitis and upper respiratory tract infection was noted but major infections were not observed.

Proteinuria reduction at 12 months was maintained at 16 months (4months after treatment discontinuation) in 4mg and 8mg groups but started to increase to baseline in the 2mg group. APRIL, IgA and Gal deficient IgA levels showed significant decline and APRIL levels returned to pretreatment levels by 16 months, indicating need for further repeated administration of the drug. Lack of detailed histological assessment at baseline is a notable limitation of the study, nevertheless, initial results are highly encouraging and phase 3 evaluation of this drug is eagerly awaited.

Inaxaplin for APOL1 associated FSGS

About 13% of black individuals have high risk APOL1 genotype, with biallelic G1 and/or G2 haplotype and contributes to 70 % excess ESRD risk in them. Inaxaplin, is a small molecule inhibitor of APOL1 channel function, and has shown activity in an APOL1 transgenic mouse model of proteinuric kidney disease. In 13 participants with two APOL1 variants and biopsy-proven FSGS, this molecule showed remarkable decrease in the proteinuria: the mean change from the baseline urinary protein-to-creatinine ratio at week 13 was −47.6% (95% confidence interval, −60.0 to −31.3).

Small sample size, short duration of follow up and lack of placebo-are important limitations. FSGS in patients who are homozygous (or compound heterozygous) for two APOL1 variants is rare, and progresses fast with a very low likelihood of spontaneous remission of proteinuria. Results are preliminary but exciting and further evaluation of this therapy is ongoing in phase 3 trial.

Risk of hair straightening products

About 60 percent of the world population has curly or wavy hair (and among South Asians, it’s nearly 85 percent). Many desire to have straight and long hairs. Skin straightening traditionally involved use of formalin (Brazilian blow out) which was declared a health hazard for both the person treated and treatment provider. This led to the emergence of alternatives like glycolic acid for this purpose. A recent report from Israel highlighted the havoc this hair straightening chemical can create in the kidney. 26 cases of AKI (2 cases had recurrent AKI) were reported, over 3 year (2019-2022), from 14 centres. All were female in their 20s. Dialysis was needed in 3 two patients had recurrent AKI. Kidney biopsy was performed in 7 cases: all 7 had acute tubular injury, 6/7 had calcium oxalate deposition. A recent NEJM report described a case of AKI following use of glyoxylic acid for hair straightening. Investigators went on to elegantly demonstrate the underlying mechanism of AKI by a case control (5 mice in each group) experiment in mice. Secondary oxalosis was induced by glyoxylic acid application- demonstrated by 4D computerized tomography and histopathology. Possible predictor might be local burning/itching sensation or ulcers after the treatment. Given the wide popularity of such treatments, problem may be a tip of iceberg and we need to maintain high degree of suspicion in appropriate settings. Typically, hairs go back to their curly style after 3 months of such treatment, same can’t be said about elevated serum creatinine and then AKI becomes CKD.

False positive urine ethanol test

In the small village where I was born, a certain tribe had exceptional skills of cheap ethanol production that could cater to the needs of half a dozen villages surrounding our’s. I vividly remember those women carrying bright white bricks of ammonium chloride (navsagar) which is supposed to activate yeasts for fermentation and subsequent country liquor production. The British categorised these tribes as “criminal” which was repealed after independence. But, this had made little difference, and any robbery or major crime in the district would automatically lead to arrest of the several men from these tribes. Police would drag their sniffer dogs (even when they were pointing in opposite directions) to the village outskirts where these people lived. Things have significantly changed now and several young men from these tribes hold reputable positions in police and administration now.

The reason I went back in the memory lane is a recent case report of a gentleman in 60s who was being investigated by the city probation office and was about to be sent to jail for illegal alcohol use. He claimed to be sober for at least the last 10 months, but a positive urine toxicology screen for ethanol proved him a blatant liar. Primary care physician requested another test at his clinic, that turned out to be negative and investigation into this disparity revealed something that all of us should be aware about. Urine was sent outside for testing after 24 hours of storage without refrigeration at the probation office. During this time ‘glucose’ in the urine was fermented to ethanol by the ‘microbes’, -two ingredients needed for ethanol production. Who put them in urine? Man was on SGLT2i.

That’s why patients love their primary care physicians, they treat and can sometimes prevent you from going to jail.

Protein restriction in Chronic Kidney Disease

Wines get better with time-they say-but the same can’t be said with old hypotheses unsupported by good evidence. Hypothesis in question is that of dietary protein restriction in management of CKD. With the expanding armamentarium of CKD, when nephrologists were almost saying goodbye, the protein restriction debate was sparked by the latest KDOQI (2020) guidelines, by recommending Very Low Protein Diet (VLPD) (0.3-0.5gm/kg/day) with ketoanalogues in patients with pre dialysis CKD (1A recommendation). Debate was colored further by the NEJM review article as many doctors believe that the letters printed on those glossy color pages of the journal are as sacrosanct as religious texts of Bhagavad Gita or New Testament. For academic atheists like me, however, it’s a difficult pill to digest.

Absence of enough evidence, possibility of harm, old, non-contemporary nature of the evidence and better things that one can try in patients with CKD are the principal reasons why the idea of protein restriction is useless in current practice. We recently reflected upon this issue in a review which just underlines the famous quote by F Parsons-all a low protein diet does is to shrink the patient down to the size of his kidneys. VLPD may postpone the decision to start dialysis for a few months while increasing the risk of malnutrition and death in the bargain.

Obesity and metabolic syndrome are precursors of the two most important CKD risk factors-diabetes and hypertension. They elevate CV risk manifolds and mitigation of this risk doesn’t need complex dietary plans but simply the restriction of salt, sugar and fat. This trio arguably underlies the major death toll, far greater than that of all the wars and natural calamities put together. Most of our patients don’t want to be bothered much about these: we know that, just talk about a diet for creatinine! Why we prefer to remain blind to the obvious and keep searching for illusory solutions in the diet is beyond comprehension. Probably, we feel so badly about salt, sugar and fat, that we simply love them!

CKD Diet: with a pinch of humor

Visit to the nephrologist is incomplete without the discussion of diet-that’s what most patients in India believe. Many patient families hold a strong opinion (opinions are not the monopoly of guideline experts) that diet is far more important than medications, optimum blood pressure control, physical activity and moderation of caloric intake. As the menu of questions concerning CKD diet is far more diverse than that of an average Gujarati Thali, these diet talks exceed the typical time needed for the actual consultation.

These sessions are greatly enjoyed by patients and families alike but are often irritating to the doctors and for the next patient waiting outside the chamber. As if this was not enough, even other specialist colleagues tell patients to ‘get the diet clarified by the kidney doctor’ as if the food eaten is metabolized by the kidney and nephrologists are in possession of additional qualification in nutrition science! I usually teach such colleagues a lesson by making sure that patient doesn’t see them back ever again(just kidding!).

So, let’s see what CKD diet is. Only thing that is clearly known about this diet is that it’s never the same for 2 different patients. Take for instance, the happy couple of Hansaben and Maheshbhai, whom I saw in OPD at the beginning of my practice. Maheshbhai, a gentleman in his 80s, with well controlled diabetes and hypertension, was rushed to my office after discovering that his eGFR has dropped to 60. Hansaben was a lady of discipline and had taken over management of the kitchen after Maheshbhai got to know about his kidney problem. After assessing his case, I assured the couple that the kidney problem isn’t that serious, and you can be less restrictive about his diet.

This information came as a disappointment and relief for Hansaben and Maheshbhai respectively. Maheshbhai was specifically fond of the ‘tomato-uttapam’ which (in addition to other tomato containing foods) was banned completely in his diet for over a year. When I said that tomatoes won’t harm the kidney, Maheshbhai was in tears, reflecting how deeply he loved that tomato uttapam. I was very happy to have helped him, but this joy was short-lived as Maheshbhai stopped seeing me thereafter. Surely Hansaben wasn’t happy with my ignorance about the intricacies of the CKD diet. However, Maheshbhai, although now under the care of a different nephrologist, sees me once in a while (of course secretly) just to say thank you, while returning back from that famous South Indian food joint!

Maheshbhai got his tomato uttapam back, but I lost a patient. What is the use of the knowledge or the facts about the diet if that is going to take away your patients? Instances like these were so annoying in my initial practice that I was on the verge of being burnt out. But a senior nephrology colleague saved my life. I happened to attend his interesting lecture which literally enlightened me about CKD diet.

First of all, he said, you have to actually come to share the belief that the “CKD diet” is the most important aspect of clinical practice although all of us are well aware that it’s sham. This is in the interest of maintaining your sanity in the practice and keeping your clientele. Second, it is best to remain passive and patient in these discussions, saying less and listening more. This, however, doesn’t mean that you appear unsure/uncertain. A rule of thumb in practice is that the certainty of your advice should generally be inversely proportional to the strength of evidence on which it is based. Whether bajra roti is better than jowar roti or if seeds should be taken out of the capsicum or boiling dal at 100 degree centigrade for 20 min would be better than 80 degrees for 40 min, baby brinjal versus giant brinjal,-whatever the question may be, it’s far more important to be sure than correct. Don’t ever say that ‘either of these is ok’; as this will leave the patients restless and invite 10 more questions that were previously not on his/her question list (yes, like home BP chart, patients carry this always). Careful listening to each case will eventually help you understand the “CKD diet” of your patient-each one may have a little similarity-but is always unique, thanks to the rich vegetation and flora of our country.

He emphasized the need to understand this area of our practice as a blessing in disguise. For example, a patient with diabetic nephropathy, experiencing steady decline in kidney function in spite of the best of your efforts. This is not a very comfortable scenario, when a patient, with his anxious daughter and son in law, are staring at you (when you are cluelessly going through all his previous records) with expressions showing variable proportions of hope, fear and anger. A nephrologist often feels cornered in this situation, with no clue as to how to break the ice, when daughter offers a respite by recalling that father ate 3 samosas at an anniversary party last week. You are relieved (that’s why I love this snack!), for at least another month until the next follow up.

Finally, he narrated a story of a very senior physician who held very rigid and peculiar views about food and health, which in nutshell, summarized the CKD diet. The man in question had a roaring practice and saw hundreds of patients in a day. One of his recommendations was complete avoidance of bottle gourds/Dudhi. Being a very authoritarian figure, no one dared to question his peculiar dietary recommendations, and patients and doctors remained clueless about the rationale, which was to be inadvertently revealed by alcohol. In a party where he was celebrating his 80th birthday, after a couple of drinks, one of his students dared to ask him this question, “What is wrong with Dudhi in diet?”

“Nothing, I just hate Dudhi!”

Now, you can replace Dudhi with Chicken, Capsicum, Tomato or any damn thing and come with a customized CKD diet of your own. Moral of the story is that dietary recommendations are driven more by such personal biases than credible scientific evidence.

May you be in private or academic practice, there is no getting away from the devil-CKD diet. So, as don Carlene has famously said, “Keep your friends close, but enemies closer”.

Happy Holi to all!

Share this:

Share this:

Three Studies, One Uncomfortable Truth

Study 1: Palmer et al. (AJKD 2018) — The Bayesian Bivariate Approach

Study 2: Heerspink et al. (Lancet D&E 2019) — The Individual Participant Data Meta-analysis

Study 3: Coresh et al. (Lancet D&E 2019) — The Large Observational Cohort

Critical Flaws Overlooked by Peer Review

What Does a 30% Albuminuria Reduction Actually Mean for Clinicians?

Share this:

expanded hemodialysis: what is it?

Share this:

Share this:

Share this:

Share this:

Share this:

Share this:

Share this: