Semaglutide safety by eGFR in FLOW: fewer serious adverse events than placebo

The study

This is a prespecified secondary analysis of FLOW (NCT03819153), published in Kidney International. FLOW was a double-blind, randomised, placebo-controlled trial in people with type 2 diabetes (T2D) and chronic kidney disease (CKD). It tested subcutaneous semaglutide 1.0 mg once weekly against placebo. The primary outcome was a kidney composite: a fall in eGFR of 50% or more, kidney failure, or death from kidney or cardiovascular causes.

This paper looks only at safety, split by baseline eGFR. Safety data were collected selectively. They covered serious adverse events (SAEs), adverse events (AEs) that led to stopping treatment, and prespecified AEs of special interest (AESIs). The authors call it a descriptive assessment. The systematically collected data are reported as numbers by eGFR group.

What they found

Of 3533 participants, 1767 received semaglutide and 1766 received placebo.

• Serious adverse events: These were reported less often with semaglutide than placebo in every baseline eGFR subgroup. At eGFR 60 or more ml/min per 1.73 m2 the figures were 45.1% versus 50.4%. At eGFR under 30 they were 56.4% versus 60.4%.
• Types of SAE: SAEs related to infections and infestations, cardiac disorders, respiratory, thoracic and mediastinal disorders, and metabolism and nutrition disorders were generally less frequent with semaglutide, whatever the baseline eGFR.
• Stopping treatment for any AE: This was more common at lower baseline eGFR in both arms. At eGFR 60 or more it was 11.2% with semaglutide and 11.3% with placebo. At eGFR under 30 it was 18.8% with semaglutide and 18.7% with placebo.
• Stopping for gastrointestinal AEs: This was reported more often with semaglutide (4.5%) than placebo (1.1%), regardless of baseline eGFR.
• AESIs: The authors report that AESI frequencies were similar between treatment groups across eGFR subgroups.

The authors conclude that no unexpected safety concerns were identified across the range of eGFR in FLOW.

How much weight to give it

The parent trial is randomised, double-blind and placebo-controlled. This safety analysis is weaker than the main trial. The authors describe it as descriptive, and the abstract does not report formal statistical tests, confidence intervals or p values for these comparisons. The differences in SAEs are therefore best read as patterns, not proven effects.

Safety data were collected selectively, so events outside the collected categories may not be captured. The abstract gives full figures for only the highest and lowest eGFR groups. It does not report funding or follow-up length. The authors' conclusion of no unexpected safety concerns applies to this descriptive review. It does not rule out smaller or rarer harms.

The lower SAE frequencies with semaglutide should not be read as a claim that the drug prevents these events. The clearest excess with semaglutide was in discontinuations due to gastrointestinal AEs.

For your clinic

This analysis adds descriptive safety information for semaglutide across the eGFR range in FLOW. The authors identified no unexpected safety concerns, including in people with eGFR under 30 ml/min per 1.73 m2. This was a descriptive analysis, so it should be read with care.

Two findings are worth noting. Stopping treatment because of AEs was more common at lower baseline eGFR in both arms. Stopping because of gastrointestinal AEs was more frequent with semaglutide (4.5%) than placebo (1.1%), regardless of baseline eGFR.

The abstract gives figures for only two eGFR subgroups. Read the full paper, the main FLOW publication and current guidelines before changing practice. Treatment decisions for individual patients remain a matter for the treating clinician.