7a. Type 1 diabetes mellitus (DM)

7a.1 Background

7a.1.1 Summary of trial evidence

Although previously marketed for use in type 1 diabetes mellitus (DM), there are currently no SGLT 2 inhibitors that are licensed for this indication in the UK. Although the National Institute for Health & Care Excellence (NICE) have issued a Technology Appraisal on dapagliflozin and sotagliflozin in people with type 1 DM and BMI >27 kg/m²,  the indication for use in type 1 DM for those of these medications has been withdrawn by the authorisation holder (see Section 4 for more information) (1, 2). The Association of British Clinical Diabetologists (ABCD) & Diabetes UK have released some guidance on the use of SGLT 2 inhibitors in type 1 DM, which highlights the need for closer monitoring and education in this population, particularly in relation to ketone monitoring (3).

Glycaemic control

There are a number of randomised controlled trials (RCTs) evaluating the use of SGLT 2 inhibitors in people with insufficiently controlled type 1 DM. The DEPICT-1 and -2 trials demonstrated that the addition of dapagliflozin (at either 5 mg and 10 mg daily doses) to existing insulin-based regimens reduces glycosylated haemoglobin (HbA1c) and body weight compared to placebo without provoking additional hypoglycaemia (4, 5). The EASE-1 trial examined the short-term (28 days) efficacy of adding empagliflozin to insulin and found that HbA1c was significantly reduced by between −3.8 to −5.4 mmol/mol (-0.35% to -0.49%; p<0.05) when compared with placebo (6). Similarly, a trial assessing canagliflozin in people with type 1 DM found that significantly more people achieved a reduction of >0.4% in HbA1c with 100 mg and 300 mg of canagliflozin versus placebo (36.9%, 41.4% vs 14.5%; p<0.001) (7). In summary, evidence from RCTs indicates that SGLT 2 inhibitors as an adjunct to insulin improves glycaemic control in people with insufficiently controlled type 1 DM.

Kidney outcomes

There are no large clinical trials reporting kidney outcomes of SGLT 2 inhibitors in people with type 1 DM. There are ad hoc analyses of larger trials that show benefit in reduction of urinary albumin-to-creatinine ratio (uACR). DEPICT-1 and -2 trials showed that after a year, the addition of dapagliflozin to insulin resulted in a dose-dependent reduction of uACR (≥3 mg/mmol) of −13.3% (95% confidence interval [CI] −37.2 to 19.8; for the 5mg dose) and −31.1% (95%CI −49.9% to −5.2%; for the 10 mg dose), compared to placebo (4, 5).

The inTandem-1 and -2 trials evaluated the effect of sotagliflozin versus placebo in people with type 1 DM. Ad hoc analysis of the inTandem-1 and -2 trials showed that in a subgroup of participants (n = 196) with mean elevated albuminuria (uACR ≥30 mg/g [3.4 mg/mmol]) at baseline, an initial dose-dependent reduction of mean albuminuria by −31.4% (SE 11.3; p=0.003) from baseline was observed for sotagliflozin 400 mg at week 24. This reduction was attenuated (and non-significantly different to placebo) at week 52 (−18.3% [SE 13.8]; p=0.18) (8, 9). Small trial size and short follow-up means there are no data available on regulator-accepted endpoints based on glomerular filtration rate (GFR) or need for kidney replacement therapy (or cardiovascular outcomes) in type 1 DM, even from the 1402 participant inTandem-3 trial (10).

Analyses of the effect of empagliflozin on uACR and eGFR slopes in the 68 participants with type 1 DM randomised into EMPA-KIDNEY are currently unavailable.

Safety

In the DEPICT-1 trial, insulin dose was reduced by an average (mean) of 14% in the first 2 weeks of starting dapagliflozin (4). In the DEPICT-2 trial, the reduction of insulin was between 10-11% (5). Both trials found no significant difference in the rates of hypoglycaemia between SGLT 2 inhibitors and placebo. During the first 7 days of the EASE-1 trial, insulin doses were kept as stable as possible, thereafter being adjusted according to glycaemic control. During these first 7 days, the rate of hypoglycaemia was higher in the 10 mg and 25 mg empagliflozin arm, than the 2.5 mg empagliflozin and placebo arm. After the first week, however, when insulin doses were adjusted, there was no significant difference in hypoglycaemia rates (6). See section 5b for more information on severe hypoglycaemia risk with SGLT 2 inhibition.

Whilst there were no diabetic ketoacidosis (DKA) events during the short EASE-1 trial (6), the DEPICT-1 and -2 trials demonstrated an increased risk of DKA in the SGLT 2 inhibitor group when compared to placebo (46.2 versus 12.7 events per 1000 person years) (4, 5). In an 18-week trial, the DKA rates were increased in the canagliflozin 100 mg or 300 mg, at 4.3% and 6.0% in comparison to 0% in the placebo group (7). Differences in the adjudication definitions of DKA and recruited populations can influence the cross-trial comparability and generalisability of these DKA rates from trials. It is not necessarily the case that one SGLT 2 inhibitor carries greater susceptibility for DKA than another. See section 5a for more details on DKA.

7a.1.2 Quality of the evidence

There are no data as yet, indicating whether SGLT 2 inhibitors provide kidney benefits in relation to regulatory accepted kidney endpoints based on GFR in people with type 1 DM. There are data that clearly demonstrate substantial excess absolute risk of DKA in people with type 1 DM. The lack of trial evidence on kidney outcomes in people with type 1 DM precludes direct comparisons of the potential balance of the definite risks versus the potential benefits.

7a.2 Recommendations for Implementation

1. We recommend that SGLT 2 inhibitors be initiated in people with type 1 DM, only under the strict direction of the diabetes team (Grade 1C).
2. We suggest considering referring people with type 1 DM to the specialist Diabetes team, for consideration of an SGLT 2 inhibitor, if they have an eGFR ≥20 mL/min/1.73m², and a uACR ≥25 mg/mmol despite being on maximum tolerated ACEi/ARB (Grade 2C).
3. We recommend all people with type 1 DM started on SGLT 2 inhibitors be provided with ketone monitoring, be advised on the signs and symptoms of DKA and to seek immediate medical advice if any of these symptoms develop or ketone levels are >0.6 mmol/L (Grade 1B).

Rationale: There is currently insufficient evidence to recommend the use of SGLT 2 inhibitors as an adjunct to existing therapies in the management of diabetic nephropathy in people with type 1 DM. Evidence of kidney benefits in people with type 2 DM makes this plausible but such results cannot be readily extrapolated to people with type 1 DM. Clinicians may wish to discuss treatment options with their patients and other specialists in cases where proteinuria persists despite current standard treatment. Thresholds for referral to the specialist diabetes team have been updated in the current version of this guideline in line with evidence available from trials of SGLT 2 inhibition irrespective of the presence or absence of diabetes. Furthermore, the grading for Recommendation number 2 has been increased from 2D to 2C, given evidence available from the InTandem3 trial (10).

7a.3 Clinical Research Recommendations

1. To establish whether the cardiovascular and kidney benefits of SGLT 2 inhibitors extend to those with type 1 DM
2. To establish the safety of SGLT 2 inhibitors in people with type 1 DM and chronic kidney disease