This Week in Nephrology — Aug 27, 2026

Welcome to This Week in Nephrology. This week we're covering 10 notable papers spanning dialysis decision-making and drug therapy at the extremes of kidney function, biomarkers and biopsy classification that sharpen how we phenotype glomerular disease, and risk stratification and preventive care in transplant candidates. Let's dive in.

Let's start with decisions at the dialysis interface, where two papers help us say when to start, how much, and when to stop. In the Journal of the American Society of Nephrology, McCoy and colleagues report an ancillary analysis of the LIBERATE-D trial intervention arm, addressing the everyday question of when a patient with acute kidney injury requiring dialysis can come off the machine [4]. Across 147 timed urine collections from 85 patients, with clinicians blinded to the results and dialysis given for prespecified indications, measured creatinine clearance discriminated better than urine output for whether dialysis would be needed over the following week. The striking finding is how low the bar can be: among collections with a creatinine clearance between 10 and 20 millilitres per minute, roughly four in five patients needed no dialysis in the subsequent week, and even between 5 and just under 10 millilitres per minute, close to half OF THOSE COLLECTIONS were followed by a dialysis-free week. In hemodynamically stable patients whose critical illness has resolved, we may be continuing intermittent hemodialysis longer than residual function requires. This was a modest single-trial sample with protocolised indications, so the numbers won't transplant perfectly to every unit, but the direction is clear enough to prompt a timed collection before assuming another week of treatments. On the chronic side, the Clinical Journal of the American Society of Nephrology reports a twelve-year single-centre experience with incremental peritoneal dialysis from Selwood and colleagues, covering 527 incident patients started on less-than-full-dose prescriptions [6]. At one year, about two thirds OF PATIENTS were still on their starting regimen, and more than a third had still not incremented by two years. Younger age, male sex, and lower albumin at the start predicted earlier escalation. Transfers to haemodialysis were uncommon and mortality was low, at roughly one event per nine person-years, which supports the safety of incremental prescribing and gives you concrete numbers to quote when a patient asks how long a lighter regimen might last. The authors built a simple clinical score from these predictors, though it awaits validation.

Staying with therapeutics in advanced kidney disease, the same journal publishes a matched retrospective cohort from Karpinski and colleagues asking whether GLP-1 receptor agonists retain value once patients reach end-stage kidney disease, a population excluded from the FLOW trial [3]. Nearly 2,500 adults who started thrice-weekly in-centre hemodialysis on a GLP-1 receptor agonist were matched one-to-one with non-users drawn from a large dialysis organisation. After matching and adjustment, users had about a 9 percent lower hospitalisation rate and about a 17 percent lower mortality rate. This is observational electronic health record data, and the users were notably different at baseline — almost all had diabetes, more had received predialysis nephrology care, and their body mass index was higher — so healthy-user and prescribing bias are real concerns. It is hypothesis-generating support for continuing these agents through the transition to dialysis rather than reflexively stopping them, but it is not trial evidence. Alongside that, the BMJ has published a state-of-the-art review on acute hyperkalaemia from Rech and colleagues, which is worth downloading precisely because practice remains so variable [2]. It walks through the dominant mechanisms — impaired renal potassium excretion from chronic kidney disease or renin-angiotensin-aldosterone system blockade, and transcellular shifts from acidosis, tissue breakdown, insulin deficiency, or hypertonicity — and synthesises contemporary management including the newer potassium binders and high-risk scenarios. For a nephrologist fielding emergency department calls, the practical value is having a single consistent framework for definitions, risk stratification, and the sequence of membrane stabilisation, shift, and removal.

The second theme this week is how we classify and detect glomerular disease, and here three papers push in the same direction: better phenotyping before treatment. Kidney International reports the outcomes of the 2025 International Society of Nephrology Forum on complement therapeutics in C3 glomerulopathy and IgA nephropathy, convened by Kavanagh and colleagues [1]. With alternative pathway complement inhibitors now succeeding in trials across C3 glomerulopathy, immune-complex membranoproliferative glomerulonephritis, and IgA nephropathy, the panel's focus is the real-world question of who to treat and how to monitor. The honest message is how much remains unsettled — how complement biomarkers, complement autoantibodies, genetics, and the biopsy itself should guide selection and duration of therapy. If you are starting to use these agents, this consensus document is the current map of both the evidence and its gaps. Also in Kidney International, Mohandes and colleagues use the TRIDENT cohort — 176 patients with diabetes undergoing clinically indicated biopsy — to revise the Renal Pathology Society classification of diabetic nephropathy [5]. The existing system discriminated only modestly, with no meaningful outcome separation between classes 3 and 4. Unsupervised clustering of light microscopy features produced a new high-risk class defined by visceral epithelial hyperplasia, which identified the most rapidly progressive group and improved two-year prediction modestly, with better net benefit on decision curve analysis. It was externally validated in about a hundred additional patients. The gain in raw discrimination is small, so treat this as refinement rather than reinvention — but if your pathologist reports visceral epithelial hyperplasia, that patient deserves closer surveillance.

Two biomarker papers round out that theme. In Nephrology Dialysis Transplantation, Rauen and colleagues describe a guanidinylated form of Y-box binding protein 1 measured by mass spectrometry, which was more prevalent in patients with high lupus disease activity and biopsy-proven nephritis, particularly proliferative classes three and four and in flares [8]. Sensitivity for lupus nephritis was around 85 percent with specificity near 70 percent, and circulating autoantibodies against the modified protein were detectable in both patients and lupus-prone mice. With 50 patients in the discovery cohort and 38 in validation, this is early-stage work, and specificity in that range will not replace a biopsy. Meanwhile, in the Clinical Journal of the American Society of Nephrology, Koh and colleagues examined growth differentiation factor 15 in nearly 32,000 UK Biobank participants without kidney disease at baseline [10]. Those in the highest quartile of this protein had close to double the risk of incident chronic kidney disease compared with the lowest. But here is the twist worth flagging: the Mendelian randomisation analysis pointed the opposite way, with genetically predicted higher levels associated with slightly higher estimated GFR. So the protein is a prognostic marker, plausibly a stress response rather than a causal driver, and this is a useful reminder not to translate a strong observational association into a therapeutic target.

Finally, transplant medicine offers two practical contributions. In the American Journal of Transplantation, Verhoeff and colleagues compared two frailty instruments in 953 kidney transplant candidates [9]. Prevalence differed dramatically — about 6 percent frail by the FRAIL questionnaire versus 24 percent by the Short Physical Performance Battery — and the two tools captured different phenotypes. Both independently predicted waitlist mortality, with the FRAIL scale carrying the stronger signal there, while the performance battery associated more with post-transplant outcomes including longer hospital stay and readmissions, and with objectively lower muscle density and higher visceral adiposity on computed tomography. The practical conclusion is that these are complementary, not interchangeable, and using both gives you a better sense of who needs prehabilitation. Alongside that, the American Society of Transplantation Infectious Diseases Community of Practice has released updated vaccination and safe living guidelines, led by Kumar [7]. Since the 2019 version, COVID-19 and respiratory syncytial virus vaccines have been added, newer pneumococcal, meningococcal and hepatitis B formulations are available, recombinant zoster vaccine has expanded age indications, and there are accumulating data supporting live-attenuated vaccines in selected paediatric and young adult recipients after transplant. Close contacts can receive most routine vaccines. The message for the pre-transplant clinic is to audit and complete vaccination before transplant, not after.

If you only have time for one paper this week, make it the dialysis discontinuation analysis in the Journal of the American Society of Nephrology [4]. It directly changes a bedside decision most of us make weekly, and it suggests our threshold for liberating stable patients from dialysis after acute kidney injury is set too high.

Here are the key takeaways from this week in Nephrology. First, in stable patients recovering from acute kidney injury, a timed urine creatinine clearance outperforms urine output for predicting dialysis need, and clearances in the 5 to 10 millilitres per minute range often permit discontinuation. Second, incremental peritoneal dialysis looks safe over twelve years of single-centre experience, with most patients staying on their starting prescription for at least a year. Third, GLP-1 receptor agonists continued into incident hemodialysis were associated with fewer hospitalisations and lower mortality, but this is observational and confounding by indication is likely. Fourth, in glomerular disease, better phenotyping is the theme — a new high-risk diabetic nephropathy class marked by visceral epithelial hyperplasia, a candidate serum biomarker for active lupus nephritis, and a consensus roadmap for complement inhibitors that is candid about what we still do not know. And fifth, in transplant candidates, use both a questionnaire-based and a performance-based frailty measure, and finish the vaccine schedule before transplant.

That's your roundup for This Week in Nephrology. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And think of one colleague — in any specialty — who never has time to keep up with the literature. Tell them about AudioScholar: a free ten-minute weekly for every specialty, to listen to in any podcast app, or to read at audioscholar dot C C.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.