SGLT2 inhibitor ketoacidosis risk persists throughout treatment, study finds

People with type 2 diabetes taking SGLT2 inhibitors can remain at risk of ketoacidosis throughout treatment, with certain patient characteristics associated with substantially higher risk, a major Scandinavian study has found.

Researchers from the Karolinska Institutet in Sweden, alongside colleagues from institutions across Sweden, Denmark and Norway, analysed more than 320,000 SGLT2 inhibitor treatment episodes and found that ketoacidosis was uncommon overall, occurring at a rate of 2.43 cases per 1,000 person-years.

However, the risk varied considerably between patients and, while it was highest shortly after starting an SGLT2 inhibitor, it persisted during longer-term treatment.

The findings published in The Lancet Diabetes & Endocrinology suggest clinicians should assess ketoacidosis risk throughout treatment rather than focusing only on the period immediately after an SGLT2 inhibitor is started.

Using nationwide health registers from Sweden, Denmark and Norway, researchers studied 322,597 treatment episodes involving 282,282 adults with type 2 diabetes.

The participants had a mean age of 63 years and 36.1 per cent of the treatment episodes involved women.

During a median follow-up of 1.3 years, the researchers recorded 1,452 ketoacidosis events.

Several patient characteristics were strongly associated with an increased risk.

People with an HbA1c of at least 83 mmol/mol had more than 15 times the odds of ketoacidosis compared with those with an HbA1c of 52 mmol/mol or lower.

Malnutrition and a previous episode of ketoacidosis were each associated with more than ten-fold higher odds, while a BMI below 20 kg/m² was associated with almost ten times the odds compared with a BMI between 20 and less than 25 kg/m².

Recent hypoglycaemia was also an important risk factor, associated with more than five times the odds of ketoacidosis.

Infection emerged as the most common event precipitating or occurring alongside ketoacidosis.

It was recorded in 31.8 per cent of ketoacidosis cases compared with 6.1 per cent of matched controls, corresponding to an odds ratio of 7.60.

The strongest associations among precipitating or co-occurring events were seen with alcohol intoxication, acute renal events, acute abdomen, stroke and major surgery.

The researchers cautioned, however, that associations involving some temporary exposures – particularly milder conditions – might have been overestimated because they could have been under-recorded among people in the control group.

Exploratory analyses also indicated that the associations identified in the study were largely risk factors for ketoacidosis among people with type 2 diabetes generally, rather than being specific to SGLT2 inhibitor treatment.

The study also highlighted substantial changes in diabetes treatment following an episode of ketoacidosis.

Among 842 people assessed one year after the event, 25.1 per cent remained on an SGLT2 inhibitor.

Meanwhile, insulin use increased from 31.9 per cent at the time of ketoacidosis to 73 per cent one year later.

The study was led by researchers at Karolinska Institutet in Sweden, alongside colleagues from institutions across Sweden, Denmark and Norway. It was funded by Region Stockholm, the Swedish Society of Medicine and Karolinska Institutet.

The researchers concluded that ketoacidosis risk during SGLT2 inhibitor treatment “varies greatly by patient characteristics” and is not limited to the period shortly after treatment begins.

They recommended risk should therefore be assessed throughout treatment and that people taking SGLT2 inhibitors should be advised to temporarily stop treatment during periods of acute illness and physiological stress.

The study used nationwide register data covering January 2013 to December 2021 in Sweden and Denmark and January 2013 to December 2022 in Norway.