Hypoglycaemia in patients with diabetes mellitus and renal impairment

With increasing evidence that certain glucose-lowering agents may have particular end-organ benefits,^1,2 it is becoming important to not only assess glucose-lowering efficacy of these drugs but also side effects such as hypoglycaemia. Hypoglycaemia is a major barrier to optimal glucose control, with intensive glycaemic control (with a lower target HbA1c) having been shown to be associated with an increased frequency of hypoglycaemic events. ³ This issue of hypoglycaemia is particularly prevalent in diabetes patients with renal impairment since reduced glomerular filtration rate (GFR) is commonly associated with increased risk of low blood glucose. Hypoglycaemic events, in turn, are associated with increased incidence of morbidity and mortality in diabetes patients with and without renal impairment, although a causal link remains controversial.^4–6 Indeed, it has been shown that chronic kidney disease (CKD) and hypoglycaemia have a multiplicative effect on all-cause mortality in diabetes, indicating that CKD increases the risk of hypoglycaemia-associated mortality. ⁷

The mechanisms responsible for the link between renal impairment and hypoglycaemic events are manifold and involve changes in glucose metabolism during renal impairment. First, whole-body insulin clearance declines during renal impairment, leading to an increase in the insulin half-life. With the tubular cells of the kidney playing a major role in the clearance of insulin, this is not surprising. The urinary insulin clearance, whose contribution to whole-body insulin clearance increases in renal impairment, mirrors the GFR, which itself progressively decreases as diabetic kidney disease progresses. Second, renal gluconeogenesis, which contributes a significant amount of the de novo glucose produced as part of the counter-regulatory response against hypoglycaemia, is altered in CKD. ⁸ Third, the pharmacodynamic and pharmacokinetic responses to glucose-lowering therapies are also altered in diabetes patients with CKD.

This relationship between hypoglycaemia and CKD has widespread implications for the pharmacological management of glycaemic control in patients with diabetes. Pharmacological treatments that clinicians have at their disposal include the biguanides, sulphonylureas, incretins [glucagon-like peptide (GLP)-1 analogues and dipeptidylpeptidase (DPP)-4 inhibitors], insulin and the sodium–glucose cotransporter (SGLT)-2 inhibitors. Of these medications, the sulphonylureas and insulin are associated with an increased risk of hypoglycaemic events and therefore the prescription of these drugs needs to be carefully managed, particularly in individuals with renal impairment.

Sulphonylureas

Sulphonylureas stimulate glucose-independent secretion of insulin, by blocking pancreatic K_ATP channels and causing the cell to depolarise and trigger a Ca²⁺-dependent signal transduction pathway leading to the secretion of insulin. Notwithstanding the fact that sulphonylureas increase the secretion of insulin, which itself has a longer half-life in patients with CKD, alterations in the pharmacokinetics of these drugs in CKD amplify the insulin-dependent risk of hypoglycaemia. Sulphonylureas are generally metabolised in the liver with the metabolites excreted by the kidneys. First-generation sulphonylureas (rarely used today) and glibenclamide are metabolised to pharmacologically active metabolites, while other second- and third-generation sulphonylureas are metabolised to inactive metabolites. Sulphonylureas producing inactive metabolites would have a lower theoretical risk of hypoglycaemia. However, this risk would still be high given that they increase the secretion of insulin, which itself has a longer half-life in CKD. Therefore, dose adjustment of these drugs should be considered in the setting of CKD.

Incretins (GLP-1 analogues and DPP-4 inhibitors)

Incretins, comprising GLP-1 analogues and DPP-4 inhibitors, amplify the glucose-dependent secretion of both insulin and glucagon and are not associated with hypoglycaemia. The relative contribution of the kidneys to excretion of these agents is variable. With respect to the DPP-4 inhibitors, only linagliptin ⁹ is cleared via renal-independent mechanisms. Thus, all other DPP-4 inhibitors must have dose reductions in subjects with reduced GFR, whereas the dose of linagliptin remains unchanged even in subjects with end-stage renal disease (ESRD). Among the GLP-1 analogues, liraglutide is cleared via renal-independent mechanisms ¹⁰ whereas exenatide is filtered at the glomerulus and degraded within the tubular cells of the kidney. ¹¹ Therefore, the effect of renal impairment on the clearance of this class of medications will be heterogeneous. While these agents alone should not affect the risk of a hypoglycaemic episode, they are nonetheless given in combination with other glucose-lowering drugs and could augment their risk of causing hypoglycaemia in the setting of CKD.

In particular, there is a link between incretin therapy and hypoglycaemia in patients already taking sulphonylureas. There is a theoretical basis behind this, with evidence that just as incretins amplify the glucose-dependent secretion of insulin, they also amplify the sulphonylurea-dependent secretion of insulin. In an isolated rat pancreas initially exposed to a low-glucose environment, GLP-1 stimulation did not lead to a significant degree of insulin secretion; however, when the pancreas was initially exposed to a low-glucose environment with a sulphonylurea added, subsequent GLP-1 stimulation led to a massive increase in insulin secretion. ¹² There is also empirical evidence that in type 2 diabetic subjects whose standard background therapy includes a sulphonylurea, treatment with sitagliptin or linagliptin increases the incidence of hypoglycaemic events.^13,14 Presumably, renal impairment would exacerbate the relationship between incretin therapy and hypoglycaemia in patients taking sulphonylureas, owing to impaired clearance of insulin and the glucose-lowering agents whose clearance is renally dependent. Thus, combining a sulphonylurea and an incretin in patients with CKD should be approached with caution.

SGLT-2 inhibitors

SGLT-2 inhibitors lower blood glucose by decreasing tubular reabsorption of glucose resulting in glycosuria. ¹⁵ A review published by Halimi and Vergès ¹⁵ addressed the incidence of hypoglycaemic events caused by SGLT-2 therapy in the presence of other anti-diabetic agents. It was concluded that SGLT-2 inhibitors do not cause a significant increase in hypoglycaemic events in type 2 diabetes patients in the presence of most combinations of other anti-diabetic agents, with the exception of sulphonylureas. There does not appear to be any corresponding data for type 2 diabetes subjects with CKD because these agents are not recommended in subjects with a GFR less than 45 mL/min. Nevertheless, with the data for the EMPA-REG study showing end-organ benefits in subjects with GFR between 30 and 45 mL/min, this class of drugs may ultimately be used in this population. Furthermore, it is theoretically possible that SGLT-2 inhibitors in combination with a sulphonylurea in such a population would be associated with a higher risk of hypoglycaemia. Therefore, if this class of drugs is to be used in this population, dose adjustment and careful monitoring for hypoglycaemic episodes by the clinician should be performed.

In summary, hypoglycaemia is commonly seen in diabetic individuals with a low GFR, particularly with the use of insulin and sulphonylureas. With increasing evidence of numerous adverse consequences of hypoglycaemia, as reported in the ADVANCE study, ⁴ the clinician needs to be cautious in their therapeutic approach towards optimising glycaemic control in the diabetic individuals with impaired renal function.

Recommendations

With increased risk of hypoglycaemia in type 2 diabetic patients, it is preferable to avoid, if at all possible, agents such as insulin and sulphonylureas. Metformin should certainly be used with caution but previous stricter conditions with respect to avoidance of this agent in subjects with low GFR have been relaxed. Indeed as is likely to appear in most guidelines over the next few years, subjects with GFR as low as 30 mL/min can now remain on metformin albeit using lower doses of this agent. Finally, SGLT-2 inhibitors are not generally used in subjects with reduced GFR due to modest hypoglycaemic effects. However, with recent positive effects on not only cardiac but also renal end points in such individuals with GFR between 30 and 60 mL/min as seen in the EMPA-REG study, a role for this class in type 2 diabetes subjects with reduced GFR needs to be reappraised but as yet cannot be recommended in routine clinical practice.