SUSTAIN-6 Follow-Up: What Later Semaglutide Studies Confirmed

First, there was no conventional SUSTAIN-6 extension

The original SUSTAIN-6 publication reports 3,297 participants followed for a median of 2.1 years. Semaglutide reduced the primary composite of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke compared with placebo, with a hazard ratio of 0.74 and a 95% confidence interval of 0.58 to 0.95 (PubMed PMID 27633186).

After that trial period, the randomized comparison did not continue as a published open-label or blinded extension with the same cohort. This matters. A post hoc analysis can extract more information from the original data, but it does not add years of randomized exposure. A later semaglutide trial can test durability in another population, but it cannot show what happened to every SUSTAIN-6 participant after the protocol ended.

The right follow-up map therefore has three branches: reanalysis of SUSTAIN-6, later randomized trials, and current labeling. Keeping those branches separate prevents a common citation error in which SELECT or FLOW is described as if it were the SUSTAIN-6 extension.

What the original trial established

SUSTAIN-6 enrolled adults with type 2 diabetes and high cardiovascular risk. Participants received once-weekly subcutaneous semaglutide 0.5 mg or 1.0 mg, or matching placebo, in addition to standard care. The primary outcome occurred in 6.6% of the semaglutide group and 8.9% of the placebo group (PubMed PMID 27633186).

The composite result was positive, but individual components were less certain. Nonfatal stroke was lower with semaglutide; cardiovascular death was similar between groups. New or worsening nephropathy was lower, while adjudicated diabetic retinopathy complications were higher. For a full account of the design and endpoints, see the SUSTAIN-6 overview and results deep dive.

What the retinopathy follow-up analysis found

A dedicated paper examined retinopathy data across the SUSTAIN program and performed post hoc analyses of SUSTAIN-6. It concluded that much of the excess retinopathy-complication risk could be attributed to the magnitude and speed of HbA1c reduction during the first 16 weeks among patients with pre-existing diabetic retinopathy, poor baseline glycemic control, and insulin treatment (PubMed PMID 29178519).

That is a plausible mediation finding, not proof that semaglutide has no retinal risk. Post hoc mediation depends on measured variables and assumptions made after the trial result is known. Clinically, the signal supports identifying pre-existing retinopathy and avoiding casual reassurance when glucose is expected to improve quickly.

The current Ozempic prescribing information retains a diabetic-retinopathy warning and notes the SUSTAIN-6 event difference (FDA Ozempic label). Label retention is useful context because it shows how regulators translated the trial signal into prescribing information.

What later kidney analyses added

Later SUSTAIN-6 analyses asked narrower kidney questions:

| Analysis type | What it adds | What it cannot add | |---|---|---| | KDIGO risk-category analysis | compares treatment effects across baseline kidney-risk groups | new randomized follow-up time | | Incident kidney-disease analysis | tests development of albuminuria or reduced eGFR definitions | a prespecified dedicated kidney endpoint trial | | Pooled SUSTAIN-6 and PIONEER-6 analysis | increases events and examines eGFR slope | identity with either original trial's population | | FLOW trial | directly tests major kidney outcomes in type 2 diabetes with CKD | a continuation of SUSTAIN-6 participants |

A 2024 post hoc analysis reported kidney outcomes across KDIGO risk categories using the original SUSTAIN-6 data (PubMed PMID 39081763). Another examined primary prevention of diabetic kidney disease in participants who did not meet its kidney-disease definition at baseline (PubMed PMID 39188242). These analyses make the original kidney signal more interpretable, but post hoc status remains a limitation.

A pooled analysis of SUSTAIN-6 and PIONEER-6 suggested a more stable eGFR trajectory with semaglutide than placebo (PubMed PMID 36738891). Pooling can improve precision. It also blends injectable and oral semaglutide trials, so readers should not attribute every pooled result to SUSTAIN-6 alone.

FLOW answered the dedicated kidney question

FLOW was designed for adults with type 2 diabetes and chronic kidney disease, a more kidney-specific population than SUSTAIN-6. Its main publication reported a 24% lower risk of the primary composite kidney outcome with weekly semaglutide 1.0 mg versus placebo (PubMed PMID 38785209).

That result supports a kidney benefit in the FLOW population and is stronger for hard kidney outcomes than a secondary SUSTAIN-6 composite driven heavily by albuminuria. It still should not be called "long-term SUSTAIN-6 follow-up." Different eligibility criteria, trial years, background therapies, and endpoint definitions prevent that shortcut.

SELECT broadened the cardiovascular population

SELECT tested semaglutide 2.4 mg in 17,604 adults with pre-existing cardiovascular disease and overweight or obesity but without diabetes. Over a mean follow-up of 39.8 months, the primary cardiovascular endpoint occurred in 6.5% of semaglutide participants and 8.0% of placebo participants, hazard ratio 0.80 (PubMed PMID 37952131).

SELECT answers a question SUSTAIN-6 could not: whether cardiovascular benefit extends to a high-risk population without diabetes. It also adds longer mean observation than SUSTAIN-6. It does not directly validate the SUSTAIN-6 effect size because it used a different dose and population.

A prespecified SELECT kidney analysis reported fewer events in its composite kidney outcome with semaglutide (PubMed PMID 38796653). That finding complements FLOW, but SELECT was a cardiovascular outcomes trial rather than a dedicated CKD trial.

What "durable" can and cannot mean

The combined evidence shows repeated cardiovascular benefit across distinct semaglutide trials and kidney benefit in a dedicated CKD trial. That replication is more informative than an unsupported assumption that the original SUSTAIN-6 hazard ratio persists unchanged forever.

Durability at the population level does not answer what happens after an individual stops treatment. SUSTAIN-6 was not designed as a withdrawal trial. SELECT and FLOW evaluated assigned treatment during their protocols. Questions about post-discontinuation risk, adherence gaps, weight regain, and glycemic change require evidence designed around discontinuation rather than extrapolation from outcome trials.

How to read the follow-up evidence accurately

Use the original SUSTAIN-6 paper for its participants, 2.1-year follow-up, primary cardiovascular result, and original safety signals. Use the retinopathy analysis for its post hoc mediation interpretation. Use the kidney post hoc papers for subgroup or exploratory SUSTAIN-6 findings. Use FLOW for dedicated kidney outcomes and SELECT for cardiovascular outcomes in obesity without diabetes.

Separating the studies prevents a blended claim that "semaglutide research proves everything." Each result belongs to a defined population, dose, endpoint, and follow-up period. Our SUSTAIN-6 limitations review explains the remaining design constraints.