SGLT2 / SGLT1-2 inhibitors (dapagliflozin, sotagliflozin)

You will still read, in a lot of places, that SGLT2 inhibitors are "approved in Europe for type 1 diabetes." They are not, and they have not been since 2021. That sentence is the single most repeated stale fact about adjunct therapy in type 1, so it is worth putting the correction first: Europe licensed these drugs for type 1, and then the licences went away.

What they actually do. SGLT2 inhibitors block a protein in the kidney called the sodium-glucose co-transporter 2, whose day job is to reabsorb glucose from urine back into your blood. Block it, and surplus glucose leaves the body in urine instead. Sotagliflozin blocks SGLT1 as well — a transporter in the gut — which slows glucose absorption from a meal on top of the kidney effect. Either way the mechanism is insulin-independent: the drug lowers glucose without needing your (absent) beta cells to do anything, and without directly causing lows the way extra insulin does. They are tablets taken alongside insulin. Nobody stops injecting.

They work. This is the part people forget when they hear the safety story. Regulators do not hand out a type 1 indication for nothing, and two of them did: the European Commission licensed dapagliflozin as an insulin adjunct in type 1 diabetes in 2019, and the EMA authorised sotagliflozin on 26 April 2019 for adults with type 1 diabetes and a BMI of 27 or above who were not reaching target on insulin alone.¹ Japan's MHLW approved dapagliflozin for adult type 1 diabetes in March 2019, and that approval still stands.² Across the trial programmes the pattern is consistent: a modest HbA1c improvement, more time in range, less total daily insulin, and weight loss — the BMI threshold in the European licence is a tell that the weight and insulin-dose effects were considered part of the point.

And then there is the ketoacidosis. Diabetic ketoacidosis is listed as a common adverse reaction — meaning at least 1 in 100 people — in the type 1 studies of dapagliflozin.³ For a drug you take every day for decades, "common" is a serious word. But the raw frequency is not the real problem. The real problem is what kind of DKA it is.

Euglycemic DKA is the whole story. Normally, ketoacidosis announces itself: your glucose climbs, your CGM alarms, you see the number and you act. On an SGLT2 inhibitor, the kidney is dumping glucose the entire time — so glucose can stay in range while ketones rise to dangerous levels. Your meter reads 140 mg/dL (7.8 mmol/L). Your CGM line is flat. You feel unwell and cannot think why. That is euglycemic DKA, and the single defence against it is the one thing standard diabetes technology does not measure: ketones. Guidance for people using these drugs off-label in type 1 therefore centres on ketone testing, a sick-day plan, and testing on symptoms rather than on glucose numbers — decisions to make with your clinician, not from this page.

What Europe actually did — and what it did not say. On 25 October 2021, the marketing-authorisation holder withdrew dapagliflozin's type 1 indication across the EU and the UK, and a Direct Healthcare Professional Communication went out the following month: Forxiga 5mg should no longer be used for type 1 diabetes.³ Sotagliflozin's EU marketing authorisation followed, withdrawn on 22 March 2022 at the holder's own request, for commercial reasons.¹

Read that carefully, because it is routinely mis-told. Neither withdrawal was triggered by a new safety signal.³ No fresh evidence appeared showing the drugs were more dangerous than the labels already said. What happened is that companies concluded a type 1 indication carrying a common, hard-to-detect, potentially fatal complication was not worth maintaining — and, in Zynquista's case, was not commercially worth keeping at all. The honest summary is therefore not "it was found to be dangerous." It is "a known, manageable-but-serious risk stopped being worth managing at scale." Whether that was the right call is a legitimate argument, and people with type 1 were not really party to it.

So where can you actually get it, for type 1?

• Japan — yes. Dapagliflozin is approved as an insulin adjunct in adult type 1 diabetes.²
• United Arab Emirates — unclear. Lexicon's licensee Viatris obtained regulatory approval for sotagliflozin in the UAE, but neither Lexicon nor Viatris has publicly stated which indication it covers, and sotagliflozin is also licensed elsewhere for heart failure. We will not list it as a type 1 route until that is confirmed.⁴
• EU and UK — no. Both type 1 authorisations were withdrawn.³¹
• US — no. The FDA has never approved an SGLT2 or SGLT1-2 inhibitor for type 1 diabetes.
• Off-label, almost anywhere — in practice, yes. The drugs are stocked everywhere for type 2 diabetes and heart and kidney disease, and doctors do prescribe them off-label in type 1. That means no type 1 dosing licence, uncertain insurance coverage, and a prescriber carrying the risk personally. It is not the same thing as approval, and this site will not pretend it is.

The US question, still open. Lexicon says it expects to resubmit its sotagliflozin application for type 1 diabetes in mid-2026, with a possible approval before the end of the year — explicitly contingent on the independently run STENO1 study producing the DKA-incidence and exposure data the FDA demanded after rejecting the drug the first time.⁴ Two cautions. A resubmission is not an approval; and the FDA has already said no to this exact drug in this exact indication once. Treat the 2026 timeline as a company's plan, not a schedule.

What would change our mind. The hazard here is not really toxicity — it is invisibility. Every part of the DKA problem in this class comes down to ketones rising unseen behind a normal glucose reading. That is a measurement failure, and measurement failures are fixable. Abbott's Libre Duo reads glucose and ketones from one wearable; the broader case is laid out in our continuous ketone monitoring entry. It is no coincidence that several trials now running — at UC San Diego, HealthPartners and AdventHealth — are explicitly testing SGLT2 inhibitors alongside continuous ketone monitors, because that pairing is the experiment that matters. If a sensor can catch euglycemic DKA hours before you feel it, the risk calculus that got these drugs pulled from Europe changes, and a genuinely effective adjunct comes back into play.

Our read. The safety score here is 20 out of 100, and it is the lowest safety score of any adjunct on this site. That is not a verdict on whether the drug works — it plainly does — and it is not a claim that anyone taking one is being reckless. It is a statement about a specific, well-documented failure mode that a CGM cannot see, in a disease where people already carry a heavy monitoring load. If you take one of these drugs, euglycemic DKA is the thing to understand, and a normal glucose reading is not evidence that you are fine — discuss ketone monitoring and a sick-day plan with your diabetes team.

References

1. European Medicines Agency. Zynquista (sotagliflozin) — medicine overview. Authorised 26 April 2019 as an adjunct to insulin in adults with type 1 diabetes and BMI ≥27 inadequately controlled on insulin alone; EU marketing authorisation withdrawn 22 March 2022 at the request of the marketing-authorisation holder, for commercial reasons. https://www.ema.europa.eu/en/medicines/human/EPAR/zynquista ↩ ↩² ↩³

2. AstraZeneca. Forxiga approved in Japan for type-1 diabetes (press release, 27 Mar 2019). https://www.astrazeneca.com/media-centre/press-releases/2019/forxiga-approved-in-japan-for-type-1-diabetes-27032019.html ↩ ↩²

3. European Medicines Agency. Direct Healthcare Professional Communication: Forxiga (dapagliflozin) 5mg should no longer be used for the treatment of type 1 diabetes mellitus (Nov 2021). Indication withdrawn by the marketing-authorisation holder across the EU and UK on 25 October 2021; not prompted by a new safety signal; DKA is a common (≥1 in 100) adverse reaction in the type 1 studies. https://www.ema.europa.eu/en/medicines/dhpc/forxiga-dapagliflozin-5mg-should-no-longer-be-used-treatment-type-1-diabetes-mellitus — see also Diabetes UK's patient-facing explanation of the UK withdrawal: https://www.diabetes.org.uk/about-us/news-and-views/dapagliflozin-no-longer-used-for-type-1-diabetes ↩ ↩² ↩³ ↩⁴

4. Lexicon Pharmaceuticals. Lexicon Pharmaceuticals Provides a Business and Pipeline Update at the 44th Annual J.P. Morgan Healthcare Conference (12 Jan 2026). Sotagliflozin US NDA resubmission expected mid-2026 with potential approval before year-end 2026, contingent on the independently run STENO1 study. The release states that licensee Viatris obtained regulatory approval in the United Arab Emirates, without naming the indication. https://www.globenewswire.com/news-release/2026/01/12/3216775/0/en/Lexicon-Pharmaceuticals-Provides-a-Business-and-Pipeline-Update-at-the-44th-Annual-J-P-Morgan-Healthcare-Conference.html ↩ ↩²