5b. Hypoglycaemia

5b.1 Background and Evidence Review

Pathophysiology and trial data

SGLT 2 inhibition does not increase the risk of hypoglycaemia when used in isolation or when combined with metformin, pioglitazone, dipeptidyl peptidase-4 inhibitors (DPP-4i) or glucagon-like peptide-1 receptor agonist (GLP-1RA) therapy (20). Results from the large placebo-controlled trials suggest they may even help reduce risk of hypoglycaemia. DECLARE-TIMI 58 and DAPA-CKD found that severe hypoglycaemia (plasma glucose <70 mg/dL or <3.9 mmol/L or hypoglycaemia requiring assistance) was more frequent in placebo than in the SGLT 2 inhibitor groups (8, 21), whilst results from EMPA-REG OUTCOME (22), VERTIS CV (23) and the other trials conducted in CKD populations (CREDENCE, EMPA-KIDNEY and SCORED) found no significant excess of hypoglycaemia (1, 7, 24). The CREDENCE trial reported all hypoglycaemia (of any severity) and found no excess in its population of people with type 2 DM and albuminuric diabetic nephropathy (hazard ratio [HR]=0.92, 95%CI 0.77-1.11; absolute difference -4.6 per 1000 person years) (7). DAPA-CKD reported a reduction in the risk of severe hypoglycaemia (14/2149 vs 28/2149: an absolute excess among those allocated placebo of about 2.7 per 1000 person years) (8). Despite the additional effect of sotagliflozin inhibiting SGLT-1, there was no suggestion of increased risk of severe hypoglycaemia in SCORED (53/5291 vs 55/5286) (24). EMPA-KIDNEY reported no difference in severe hypoglycaemia between groups (77/3304 vs 77/3305) (1). It is possible that reductions in doses of other hypoglycaemia-inducing therapies (i.e. insulin/sulphonylurea [SU]) to mitigate against perceived risk of hypoglycaemia associated with an additional glycaemia-lowering agent may explain these modest benefits on severe hypoglycaemia observed in the placebo-controlled large SGLT 2 inhibitor trials. No incidence of severe hypoglycaemia was reported by the 5877 people without DM included in EMPEROR- REDUCED, DAPA-HF or DAPA-CKD (DAPA-CKD included 1398 people with CKD without DM) (8, 12, 13). Among people without diabetes recruited to EMPEROR-PRESERVED (representing 51% of the total population) there were 4 cases of hypoglycaemia episodes requiring assistance (2/1531 vs 2/1518 in empagliflozin and placebo arms respectively) (25). DELIVER reported similar incidence of major hypoglycaemia in the treatment and placebo groups (6/3126 vs 7/3127 respectively), but data from the DELIVER trial stratified by diabetes status are not currently available (2).

Insulin in combination with SGLT 2 inhibition

Insulin therapy is associated with an increased risk of hypoglycaemia in people with DM. Meta-analyses and observational data suggest that, when added to insulin therapy, SGLT 2 inhibition does not increase this risk of hypoglycaemia following a 10-20% reduction in total daily insulin dose (26, 27). Insulin doses were also reduced by up to 20% to prevent hypoglycaemia in people with type 1 DM in the SGLT 2 inhibitor arm of the DEPICT trials (28, 29). Any further insulin dose reduction (beyond 20%) should be cautious and targeted at avoiding hypoglycaemia (30), as excessive insulin dose reduction may increase risk of DKA. Those with more labile blood glucose control may benefit from discussion with the diabetes team for consideration of ketone monitoring (see section 5a).

Sulphonylurea (SU) and meglitinide in combination with SGLT 2 inhibition

Insulin secretagogues, whether used as monotherapy or in combination with other glucose-lowering drugs, are associated with an increased risk of hypoglycaemia. Meta-analyses of relatively short-term phase 3 placebo-controlled studies found an excess risk of hypoglycaemia when SGLT 2 inhibitors are added to metformin and SU (Odds Ratio=1.75, 95%CI 1.43-2.15) (31), but did not consider the impact of CKD. Conversely, the large clinical outcome trials found no excess risk of severe hypoglycaemia with SGLT 2 inhibition (with several reporting reductions in risk). This has generated some clinical uncertainty leading to variation in clinical practice, with some clinicians recommending SU doses are reduced when starting SGLT 2 inhibitors, and others proposing SUs should be stopped altogether (30). It should be remembered that in people with an eGFR <45 mL/min/1.73m², SGLT 2 inhibition has only modest effects on glucose lowering (32). There is very little evidence for the use of SGLT 2 inhibitors in combination with meglitinides. However, as the risk of hypoglycaemia with meglitinide use is increased in people with advanced CKD (33), we do recommend consideration of meglitinide dose reductions when initiating SGLT 2 inhibitors.

5b.2 Recommendation for Implementation

1. We recommend considering reducing the dose of insulin/SUs/meglitinides when initiating SGLT 2 inhibitors to reduce the risk of hypoglycaemia (Grade 1C).
2. We recommend that when initiating SGLT 2 inhibitors in people taking SUs (e.g. gliclazide) or meglitinides (e.g. repaglinide) when the HbA1c <58 mmol/mol AND eGFR >45 mL/min/1.73m², consider reducing dose of SU or meglitinide by 50% to reduce risk of hypoglycaemia (Grade 1C).
3. We recommend that when starting SGLT 2 inhibitors in people taking insulin when the HbA1c <58 mmol/mol AND eGFR >45 mL/min/1.73m², consider reducing the insulin dose by 20% to avoid hypoglycaemia (Grade 1C).
4. We recommend that when starting SGLT 2 inhibitors in people taking only metformin ± pioglitazone ± DPP-4i/gliptins or GLP-1RA therapy, no dosage adjustment is necessary (Grade 1C).

Rationale: SGLT 2 inhibitors are effective drugs at reducing hyperglycaemia when they are used in people with preserved kidney function (e.g. eGFR >60 mL/min/1.73m²), however, their glycaemic effectiveness reduces as the eGFR declines. Where a treatment for DM carries a risk of hypoglycaemia (such as SUs and insulin use), the addition of an SGLT 2 inhibitor may potentiate that risk, particularly if baseline glycaemic control is reasonable at the time of initiation of treatment. There is no evidence that SGLT 2 inhibitors cause significant hypoglycaemia on their own or in addition with DM medicines that are not associated with hypoglycaemia.

5b.3 Audit Measures

1. The proportion of people on Insulin/SUs with HbA1c <58 mmol/mol and eGFR >45 mL/min/1.73m², whose therapy was appropriately reduced when initiating SGLT 2 inhibitors.