The Medical Briefing Nephrology
Atacicept in IgA nephropathy, continuous KRT in the ICU, AI and graft biopsies, hyponatremia
November 2026 edition (1/4)
The essentials in four articles
■ Glomerular disease
In the ORIGIN 3 trial, atacicept markedly slowed eGFR decline at 2 years, beyond its effect on proteinuria.
■ AKI and critical care
In a French trial of 554 patients, continuous kidney replacement therapy was associated with lower day-90 mortality than intermittent hemodialysis.
■ Transplantation
Artificial intelligence performs well at scoring Banff lesions on graft biopsies, without replacing the pathologist.
■ Clinical nephrology
Osmotic demyelination remains rare even after rapid correction of hyponatremia; caution still applies, with a correction rate tailored to the patient’s risk.
■ Glomerular disease
IgA nephropathy: atacicept confirms its efficacy in preserving kidney function at 2 years.
NEJM, phase III randomized trial
Atacicept Placebo Proteinuria reduction at 52 weeks
Annualized eGFR slope through week 104 (mL/min/1.73 m²/yr)
Kidney progression (composite) at week 104, HR 0.24
Key message. These results demonstrate a major slowing of eGFR decline with atacicept, beyond its antiproteinuric effect. |
Final efficacy results of the phase III ORIGIN 3 trial published in the New England Journal of Medicine (Barratt et al., October 2026).
Atacicept is a TACI-Fc fusion protein that simultaneously inhibits BAFF and APRIL, two cytokines involved in B-cell maturation and in the production of galactose-deficient IgA1 (Gd-IgA1), which drives the formation of the immune complexes responsible for IgA nephropathy.
Methods
Multicenter, randomized, double-blind, placebo-controlled trial.
428 patients analyzed:
Atacicept 150 mg/week subcutaneously (n = 214).
Placebo (n = 214).
Mean baseline eGFR: 64 mL/min/1.73 m².
Mean proteinuria: 2.2 g/day.
RAAS blocker in 98–99% of patients and SGLT2 inhibitor in about 61%.
The key secondary endpoints were the change in eGFR at 52 weeks and the annualized eGFR slope through 104 weeks.
Results
At 52 weeks:
Change in eGFR: −0.11 vs. −5.69 mL/min/1.73 m².
Between-group difference: +5.58 mL/min/1.73 m² (p < 0.001).
Reduction in proteinuria: 53.3% vs. 4.7%.
At 104 weeks:
Annualized eGFR slope: −0.60 vs. −5.60 mL/min/1.73 m²/year (p < 0.001).
Composite kidney progression endpoint: 5.1% vs. 17.8% (HR 0.24; 95% CI 0.12–0.48).
No dialysis, transplantation or death events in the atacicept group, vs. 7 in the placebo group.
Immunologically, Gd-IgA1 decreased by 66.1% at 52 weeks, vs. a 2.9% increase with placebo.
Safety
Adverse events: 74.8% vs. 78.0%.
Serious adverse events: 7.0% vs. 15.4%.
These results demonstrate a major slowing of eGFR decline with atacicept, beyond its antiproteinuric effect.
Therapies targeting BAFF and APRIL could durably modify the course of IgA nephropathy.
#Nephrology #IgAN #Atacicept #ClinicalTrials #KidneyDisease #NEJM
■ AKI and critical care
Acute kidney injury in the ICU: is continuous kidney replacement therapy associated with better survival than intermittent hemodialysis?
NEJM, multicenter randomized trial
The trial was stopped early because of an excess mortality signal in the intermittent group. Continuous KRT Intermittent hemodialysis MAKE90 (primary endpoint)
Mortality at day 90, HR 1.30
Key message. These results could change ICU practice in favor of continuous therapy as the initial modality in patients with severe AKI and multiple organ failure. |
This French trial, published in the New England Journal of Medicine, provides important results on the choice of kidney replacement therapy modality in ICU patients with severe acute kidney injury (AKI).
This multicenter, randomized, open-label, non-inferiority trial, conducted in 38 French ICUs, enrolled 554 patients, 535 of whom were included in the primary analysis. It compared intermittent hemodialysis with continuous kidney replacement therapy as the initial strategy.
The primary endpoint was a major adverse kidney event at day 90 (MAKE90), combining death, dialysis dependence or a persistent increase in serum creatinine > 25% from baseline.
Results at day 90:
MAKE90: 78.1% with intermittent hemodialysis vs. 71.6% with continuous therapy (adjusted difference: +6.28 points; 95% CI: −1.22 to 13.79).
Mortality: 63.1% vs. 53.5%, respectively (HR 1.30; 95% CI: 1.04–1.63).
Persistent kidney dysfunction among patients weaned from dialysis: 39.4% vs. 37.6%.
The trial was stopped early because of an excess mortality signal in the intermittent group.
Non-inferiority of intermittent hemodialysis was not demonstrated for the primary endpoint, nor was superiority of continuous therapy formally demonstrated for this composite endpoint.
However, the difference in mortality is an important clinical signal.
These results could change ICU practice in favor of continuous therapy as the initial modality in patients with severe AKI and multiple organ failure.
#Nephrology #CriticalCare #AKI #Dialysis #ICRAKI #RenalReplacementTherapy
■ Transplantation
Artificial intelligence and kidney graft biopsy: toward an automated Banff classification?
Kidney International Reports, review
Limitations. The main limitations remain staining heterogeneity, insufficient diversity of training datasets and lack of external validation. Key message. AI does not replace the pathologist, but could become an essential complementary tool to standardize biopsy interpretation and improve the follow-up of kidney transplant recipients. |
Since 1991, the Banff classification has been the international reference for the histological interpretation of kidney allograft biopsies. It relies on a semi-quantitative assessment of inflammatory, glomerular, vascular and tubulointerstitial lesions, with sometimes substantial interobserver variability.
This review published in Kidney International Reports summarizes the applications of artificial intelligence (AI), particularly deep learning, in kidney transplant pathology.
The reported performance is encouraging:
Glomerulosclerosis: segmentation algorithms achieve F1 scores of 90% to 98% in several studies.
Interstitial fibrosis and tubular atrophy (IFTA): models allow continuous quantification of lesions, with Dice coefficients of 0.80 to 0.84 and reproducibility comparable to that of pathologists.
Inflammation: one study of 423 biopsies reports AUCs > 0.94 for predicting the Banff ti score and > 0.86 for the ptc score.
Kidney prognosis: some fibrosis measurements are associated with graft survival, suggesting value beyond simple morphological classification.
The goal of the Banff Digital Pathology Working Group is to progressively develop automated assessment of lesions, notably glomerulitis, peritubular capillaritis, arteritis and tubulitis.
The main limitations remain staining heterogeneity, insufficient diversity of training datasets and lack of external validation.
The clinical value is considerable: moving from semi-quantitative appraisal to objective and reproducible measurement of lesions, potentially more sensitive for assessing their progression and prognostic value.
AI does not replace the pathologist, but could become an essential complementary tool to standardize biopsy interpretation and improve the follow-up of kidney transplant recipients.
#Nephrology #KidneyTransplantation #ArtificialIntelligence #Banff #DigitalPathology
■ Clinical nephrology
Hyponatremia: should it be corrected rapidly or slowly?
Clinical Kidney Journal, review
Correction targets
Key message. The clinical message therefore remains one of caution: treat the neurological emergency rapidly, then individualize the rate of correction according to the risk of demyelination, with close monitoring of serum sodium and urine output. |
The rate of correction of hyponatremia remains a subject of debate, between the risk of cerebral edema if correction is insufficient and that of osmotic demyelination syndrome (ODS) if correction is excessive.
A review published in Clinical Kidney Journal examines the pathophysiological data and recent studies that call into question the usual correction thresholds.
In the presence of severe neurological symptoms, administration of 3% hypertonic saline should rapidly raise serum sodium by 4 to 6 mmol/L within 1 to 2 hours, which is generally sufficient to reduce cerebral edema.
For chronic hyponatremia, the authors maintain the following targets:
Standard risk: target increase of 4 to 8 mmol/L/24 h, without exceeding 10 mmol/L/24 h or 18 mmol/L/48 h.
High risk of ODS: target of 4 to 6 mmol/L/24 h, maximum 8 mmol/L/24 h.
These recommendations are discussed in light of new observational data.
In a cohort of 22,858 patients, the incidence of ODS was only 0.05%, despite rapid correction in 17.7% of patients.
A meta-analysis of 16 studies including 11,811 patients even found an association between rapid correction and lower in-hospital mortality, without a statistically significant increase in ODS.
These results do not, however, support the conclusion of a causal benefit of rapid correction, notably because of confounding and the rarity of ODS.
The clinical message therefore remains one of caution: treat the neurological emergency rapidly, then individualize the rate of correction according to the risk of demyelination, with close monitoring of serum sodium and urine output.
#Nephrology #Hyponatremia #SIADH #FluidAndElectrolyteDisorders #InternalMedicine
Next week
IgA nephropathy: 52% reduction in proteinuria with povetacicept in the phase III RAINIER trial.
Sepsis-associated acute kidney injury: two biological subphenotypes with different outcomes and treatment responses.
Calciphylaxis and vitamin K: a new therapeutic perspective?
Complicated urinary tract infections: toward shorter antibiotic courses under the IDSA 2025 guidelines