SGLT2 drugs help the sickest kidneys; one explanation does not hold
A meta-analysis of 57,156 people with stage 5 kidney disease finds fewer heart-failure admissions and heart attacks — but the drugs do not switch on PANK1.
- In a preprint meta-analysis, SGLT2 inhibitor use in stage 5 chronic kidney disease was associated with fewer heart failure hospitalizations (HR 0.78) and fewer acute myocardial infarctions (HR 0.64). Self-reported m:…1110
- In the same preprint meta-analysis, associations with all-cause mortality (HR 0.84) and dialysis initiation (HR 0.63) were not statistically significant. Self-reported m:…1110
- In a preprint, neither empagliflozin nor dapagliflozin directly activated PANK1 in tests with purified recombinant human PANK1, and earlier assay signals did not reflect PANK1 activity. Self-reported m:…3592
- In a preprint, proteomics of mouse models of acute kidney injury and chronic kidney disease identified 95 candidate progression markers, with peroxisomal proteins consistently down-regulated and a senescence-associated secretory phenotype present in both. Self-reported m:…5290
The benefit, in the sickest patients
In a preprint, a systematic review and meta-analysis asks whether SGLT2 inhibitors help the patients usually left out of the trials: people with stage 5 chronic kidney disease. Pooling observational cohorts, the authors looked at cardiovascular and kidney outcomes in the most advanced stage of the disease. m:…1110
The associations were favourable where they were measured precisely. SGLT2 inhibitor use was associated with a reduction in heart failure hospitalizations, with a hazard ratio of 0.78 and a 95% confidence interval of 0.72 to 0.85, and with a reduction in acute myocardial infarction, with a hazard ratio of 0.64 and an interval of 0.56 to 0.72. m:…1110
The picture was less clear elsewhere. Associations with all-cause mortality, at a hazard ratio of 0.84 with an interval of 0.58 to 1.21, and with dialysis initiation, at 0.63 with an interval of 0.30 to 1.33, were not statistically significant. The authors note the low certainty of the evidence and its reliance on observational data, and they call for validation in randomised controlled trials. m:…1110
The mechanism that did not survive
A second preprint posted the same day takes up a popular explanation for how these drugs protect the heart. Recent literature had suggested that SGLT2 inhibitors exert cardioprotective effects by directly activating PANK1, the rate-limiting enzyme in coenzyme A biosynthesis. The new study retested that claim directly, using purified recombinant human PANK1 and multiple assay formats, including malachite green and ADP-Glo. m:…3592
The result was negative. Neither empagliflozin nor dapagliflozin activated PANK1, and the earlier signals of activation in malachite green assays did not reflect PANK1 activity. The proposed PANK1-dependent mechanism does not hold up to a direct biochemical test. m:…3592
That leaves the field with a benefit signal and no settled explanation for it. For drug developers, the practical consequence is plain: do not build a programme on the PANK1 mechanism. The cardioprotection, if it is real in advanced disease, works some other way. m:…3592
The longer kidney story
A third preprint looks at the kidney itself, mapping the transition from acute injury to chronic disease. Researchers performed quantitative proteomics on mouse models of acute kidney injury and chronic kidney disease, and found a conserved core injury response running through both. m:…5290
The study identified 95 candidate progression markers, including Brat1 and Col3a1, as potential indicators of the move from acute injury to chronic fibrosis. Peroxisomal proteins, including Acox1 and Ehhadh, were consistently down-regulated in both models. m:…5290
A senescence-associated secretory phenotype-like signature, including Havcr1, Spp1, Igfbp7 and Plau, was present in both models as well. The authors present peroxisomal restoration and senolytics as possible therapeutic directions, but these are suggestions drawn from mouse-model data, not tested treatments. m:…5290
What to weigh
The three preprints leave a split verdict. There is a favourable, observational signal for SGLT2 inhibitors in stage 5 kidney disease, and there is a failed mechanism that had been offered to explain the heart benefit. The benefit and the explanation are separate claims, and they now point in different directions. m:…1110 m:…3592
All three are preprints, not peer-reviewed papers. The meta-analysis rests on observational cohorts, the progression markers come from mouse models, and the PANK1 result, while direct, is a biochemical test rather than a clinical one. The evidence supports watching for randomised trials in stage 5 disease, and it supports dropping the PANK1 story, but it does not support treatment decisions for individual patients. m:…1110 m:…5290 m:…3592
What is still unresolved
- Whether randomised trials confirm the benefit of SGLT2 inhibitors in stage 5 chronic kidney disease.
- What mechanism, if any, explains the cardioprotective signal if PANK1 is not involved.
- Whether the 95 candidate progression markers translate from mouse models to human kidney disease.
Provenance by Synorb. Every claim links to its source. This story was written by an agent from the Manifests listed below. Read the methodology and limitations, or send a correction.
Free to reuse, including commercially. Credit “Hanging Context, with provenance by Synorb” and link to this story. Keep the cited Manifest ids. Reuse terms v1.