Kidney outcomes: FLOW and the indirect route to the nephron

FLOW was the first trial built to test whether a GLP-1 medicine protects the kidney. It answered yes, clearly enough to be stopped early. How it does so is a good illustration of an organ that improves without carrying the receptor.

The trial

3,533 adults with type 2 diabetes and chronic kidney disease, defined by reduced eGFR with albuminuria, were randomised to semaglutide 1 mg weekly or placebo on top of standard care, which for most included renin-angiotensin system blockade. The primary endpoint was a composite of kidney failure (dialysis, transplant or eGFR below 15), a sustained 50% or greater fall in eGFR, or death from kidney or cardiovascular causes. Over a median 3.4 years it occurred in 18.7% of the semaglutide group and 23.2% of the placebo group, hazard ratio 0.76 (95% CI 0.66 to 0.88), P = 0.0003. The annual eGFR decline was slower by 1.16 mL/min/1.73 m². Major cardiovascular events fell 18% and death from any cause 20% (PubMed 38785209). An independent committee stopped the trial early on the strength of an interim analysis.

Why it must be indirect

Pyke's validated receptor mapping found GLP-1 receptor protein in the smooth muscle of some renal arterioles but not in tubular cells or glomeruli in human and monkey kidney (PubMed 24467746). The nephron's working cells do not hear the drug directly. The benefit therefore arrives through the kidney's environment, and the candidates are well characterised even if their relative weight is not.

The routes

Glucose, pressure, weight. All three fell on treatment and all three are established drivers of diabetic kidney disease progression. In FLOW the between-group differences were modest, in line with other outcome trials, and unlikely to explain the full effect alone.

Natriuresis. In 2004 Gutzwiller and colleagues infused GLP-1 into healthy men and insulin-resistant obese men and measured a rise in sodium excretion, attributed to reduced sodium reabsorption in the proximal tubule (PubMed 15181026). Less proximal sodium reabsorption increases sodium delivery to the macula densa, which through tubuloglomerular feedback constricts the afferent arteriole and lowers pressure inside the glomerulus. That is the same logic by which SGLT2 inhibitors protect the kidney, and it is a human finding, grade B, but whether it persists with chronic dosing and contributes to the outcome is not shown.

Inflammation and hemodynamics. C-reactive protein fell. Arteriolar smooth muscle carries the receptor. Both could change the glomerular environment; neither has been shown to mediate the outcome. Hammoud and Drucker's review treats the kidney mechanism as unresolved (PubMed 36509857), and so does this site: grade D.

What came before

LEADER, the liraglutide cardiovascular outcome trial in type 2 diabetes, reported a prespecified secondary renal composite with a hazard ratio of 0.78 (95% CI 0.67 to 0.92), driven mostly by less new persistent macroalbuminuria rather than by hard endpoints (PubMed 28854085). SUSTAIN-6 showed a similar albuminuria signal. FLOW was designed to test whether those hints translated into kidney failure, eGFR loss and death, and it did.

Where to go next

• The FLOW digest, with every prespecified secondary endpoint: FormBlends Research.
• Which kidney labs matter and how often, with reference ranges: GLP-1 Lab Guide.
• The receptor map that makes the indirect argument: GLP-1 receptor effects by organ.