Clement Lo, Tadashi Toyama, Ying Wang, Jin Lin, Yoichiro Hirakawa, Min Jun, Alan Cass, Carmel M Hawley, Helen Pilmore, Sunil V Badve, Vlado Perkovic, Sophia Zoungas

Diabetes is the commonest cause of chronic kidney disease (CKD). Both conditions commonly co-exist. Glucometabolic changes and concurrent dialysis in diabetes and CKD make glucose-lowering challenging, increasing the risk of hypoglycaemia. Glucose-lowering agents have been mainly studied in people with near-normal kidney function. It is important to characterise existing knowledge of glucose-lowering agents in CKD to guide treatment. To examine the efficacy and safety of insulin and other pharmacological interventions for lowering glucose levels in people with diabetes and CKD. We searched the Cochrane Kidney and Transplant Register of Studies up to 12 February 2018 through contact with the Information Specialist using search terms relevant to this review. Studies in the Register are identified through searches of CENTRAL, MEDLINE, and EMBASE, conference proceedings, the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov. All randomised controlled trials (RCTs) and quasi-RCTs looking at head-to-head comparisons of active regimens of glucose-lowering therapy or active regimen compared with placebo/standard care in people with diabetes and CKD (estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m2) were eligible. Four authors independently assessed study eligibility, risk of bias, and quality of data and performed data extraction. Continuous outcomes were expressed as post-treatment mean differences (MD). Adverse events were expressed as post-treatment absolute risk differences (RD). Dichotomous clinical outcomes were presented as risk ratios (RR) with 95% confidence intervals (CI). Forty-four studies (128 records, 13,036 participants) were included. Nine studies compared sodium glucose co-transporter-2 (SGLT2) inhibitors to placebo; 13 studies compared dipeptidyl peptidase-4 (DPP-4) inhibitors to placebo; 2 studies compared glucagon-like peptide-1 (GLP-1) agonists to placebo; 8 studies compared glitazones to no glitazone treatment; 1 study compared glinide to no glinide treatment; and 4 studies compared different types, doses or modes of administration of insulin. In addition, 2 studies compared sitagliptin to glipizide; and 1 study compared each of sitagliptin to insulin, glitazars to pioglitazone, vildagliptin to sitagliptin, linagliptin to voglibose, and albiglutide to sitagliptin. Most studies had a high risk of bias due to funding and attrition bias, and an unclear risk of detection bias.Compared to placebo, SGLT2 inhibitors probably reduce HbA1c (7 studies, 1092 participants: MD -0.29%, -0.38 to -0.19 (-3.2 mmol/mol, -4.2 to -2.2); I2 = 0%), fasting blood glucose (FBG) (5 studies, 855 participants: MD -0.48 mmol/L, -0.78 to -0.19; I2 = 0%), systolic blood pressure (BP) (7 studies, 1198 participants: MD -4.68 mmHg, -6.69 to -2.68; I2 = 40%), diastolic BP (6 studies, 1142 participants: MD -1.72 mmHg, -2.77 to -0.66; I2 = 0%), heart failure (3 studies, 2519 participants: RR 0.59, 0.41 to 0.87; I2 = 0%), and hyperkalaemia (4 studies, 2788 participants: RR 0.58, 0.42 to 0.81; I2 = 0%); but probably increase genital infections (7 studies, 3086 participants: RR 2.50, 1.52 to 4.11; I2 = 0%), and creatinine (4 studies, 848 participants: MD 3.82 μmol/L, 1.45 to 6.19; I2 = 16%) (all effects of moderate certainty evidence). SGLT2 inhibitors may reduce weight (5 studies, 1029 participants: MD -1.41 kg, -1.8 to -1.02; I2 = 28%) and albuminuria (MD -8.14 mg/mmol creatinine, -14.51 to -1.77; I2 = 11%; low certainty evidence). SGLT2 inhibitors may have little or no effect on the risk of cardiovascular death, hypoglycaemia, acute kidney injury (AKI), and urinary tract infection (low certainty evidence). It is uncertain whether SGLT2 inhibitors have any effect on death, end-stage kidney disease (ESKD), hypovolaemia, fractures, diabetic ketoacidosis, or discontinuation due to adverse effects (very low certainty evidence).Compared to placebo, DPP-4 inhibitors may reduce HbA1c (7 studies, 867 participants: MD -0.62%, -0.85 to -0.39 (-6.8 mmol/mol, -9.3 to -4.3); I2 = 59%) but may have little or no effect on FBG (low certainty evidence). DPP-4 inhibitors probably have little or no effect on cardiovascular death (2 studies, 5897 participants: RR 0.93, 0.77 to 1.11; I2 = 0%) and weight (2 studies, 210 participants: MD 0.16 kg, -0.58 to 0.90; I2 = 29%; moderate certainty evidence). Compared to placebo, DPP-4 inhibitors may have little or no effect on heart failure, upper respiratory tract infections, and liver impairment (low certainty evidence). Compared to placebo, it is uncertain whether DPP-4 inhibitors have any effect on eGFR, hypoglycaemia, pancreatitis, pancreatic cancer, or discontinuation due to adverse effects (very low certainty evidence).Compared to placebo, GLP-1 agonists probably reduce HbA1c (7 studies, 867 participants: MD -0.53%, -1.01 to -0.06 (-5.8 mmol/mol, -11.0 to -0.7); I2 = 41%; moderate certainty evidence) and may reduce weight (low certainty evidence). GLP-1 agonists may have little or no effect on eGFR, hypoglycaemia, or discontinuation due to adverse effects (low certainty evidence). It is uncertain whether GLP-1 agonists reduce FBG, increase gastrointestinal symptoms, or affect the risk of pancreatitis (very low certainty evidence).Compared to placebo, it is uncertain whether glitazones have any effect on HbA1c, FBG, death, weight, and risk of hypoglycaemia (very low certainty evidence).Compared to glipizide, sitagliptin probably reduces hypoglycaemia (2 studies, 551 participants: RR 0.40, 0.23 to 0.69; I2 = 0%; moderate certainty evidence). Compared to glipizide, sitagliptin may have had little or no effect on HbA1c, FBG, weight, and eGFR (low certainty evidence). Compared to glipizide, it is uncertain if sitagliptin has any effect on death or discontinuation due to adverse effects (very low certainty).For types, dosages or modes of administration of insulin and other head-to-head comparisons only individual studies were available so no conclusions could be made. Evidence concerning the efficacy and safety of glucose-lowering agents in diabetes and CKD is limited. SGLT2 inhibitors and GLP-1 agonists are probably efficacious for glucose-lowering and DPP-4 inhibitors may be efficacious for glucose-lowering. Additionally, SGLT2 inhibitors probably reduce BP, heart failure, and hyperkalaemia but increase genital infections, and slightly increase creatinine. The safety profile for GLP-1 agonists is uncertain. No further conclusions could be made for the other classes of glucose-lowering agents including insulin. More high quality studies are required to help guide therapeutic choice for glucose-lowering in diabetes and CKD.