|
Comparative effectiveness and safety of medications for type 2 diabetes: an update including new drugs and 2-drug combinations |
Bennett WL, Maruther NM, Singh S, Segal JB, Wilson LM, Chatterjee R, Marinopoulos SS, Puhan MA, Ranasinghe P, Block L, Nicholson WK, Hutfless S, Bass EB, Bolen S |
|
|
CRD summary This review concluded that evidence supported use of metformin as a first-line agent for treatment of type 2 diabetes. This was a generally well-conducted review in which a vulnerability to publication bias and limitations of the evidence base were acknowledged. Provided these are borne in mind, the conclusions are likely to be reliable. Authors' objectives To assess the benefits and harms of metformin, second-generation sulfonylureas, thiazolidinediones, meglitinides, dipeptidyl peptidase (DPP-4) inhibitors and glucagon-like peptide-1 receptor agonists as monotherapy and in combination for the treatment of adults with type-2 diabetes. An update review. Searching MEDLINE, EMBASE and Cochrane Central Register of Controlled Trials were searched from inception to April 2010. The MEDLINE search for long-term clinical outcomes was updated to December 2010. Fifteen relevant journals were searched. Reference lists of included studies and relevant reviews were searched. Experts in the field were contacted. Public health websites were searched. Pharmaceutical companies were contacted. Search terms were reported. Only studies reported in English were eligible for inclusion in the review. Study selection Studies that assessed head-to-head comparisons in adults aged over 18 years with type 2 diabetes who were not pregnant were eligible for inclusion. Any Food and Drug Administration (FDA) approved oral diabetes medication used as monotherapy or in a two-drug combination with metformin or a thiazolidinedione or insulin therapy in combination with oral medication was was included. Studies of α-glucosidase inhibitors were excluded from the review. Studies that assessed colesevelam were excluded. Studies were required to report major long-term clinical outcomes or at least one of the intermediate outcomes haemoglobin A1c (HbA1c), body weight, lipid levels, hypoglycaemia or one of a number of defined adverse events. Studies were required to enrol at least 40 patients and to have a minimum of three months follow-up.
Most of the included studies were conducted in Europe or USA. More than half of the studies assessed metformin or a thiazolidinedione in combination with another medication. Most studies excluded patients with significant comorbidities such as liver, kidney or cardiovascular disease. Many studies excluded older individuals.
Two reviewers independently selected the studies for inclusion in the review. Assessment of study quality Two reviewers independently assessed the studies for validity using the Jadad scale of up to five points based on criteria of randomisation, blinding and treatment of withdrawals and dropouts. Observational studies were assessed using items related to selection bias, treatments, outcome measurement, statistical methods and losses to follow-up. In all cases overall quality of studies was rated as good, fair or low. External validity was assessed by comparison with usual care of people with type 2 diabetes in USA. Overall evidence for each comparison was assessed using GRADE criteria. Data extraction Data to permit calculation of mean differences or odds ratios (OR) with 95% confidence intervals (CI) were extracted by one reviewer and checked by a second. Medications were grouped by drug class with the exception of rosiglitazone and pioglitazone, which were considered separately. Methods of synthesis Statistical synthesis of data was considered only where at least three trials with sufficient clinical homogeneity were available. Continuous outcome data were combined using DerSimonian and Laird random-effects model analyses to calculate weighted mean differences (WMD) with 95% CI. For the outcomes of congestive heart failure and ischaemic heart disease, pooled odds ratios were calculated; a Peto odds ratio was used for the outcome of hypoglycaemia.
Heterogeneity between studies was assessed using Χ2 and I2 (threshold 50%) statistics. Where significant heterogeneity was detected, either the studies were not pooled or meta-regression was employed in an attempt to explore the heterogeneity.
Sensitivity analyses omitted each individual study in turn. Publication bias was assessed using the Begg and Mazumdar test and Egger's test. Results of the review One hundred and sixty six studies were included in the review: 140 were RCTs and 26 were observational studies. Most pooled analyses were assessed as having a low or medium risk of bias; results for the validity of individual studies were not reported.
Metformin was more efficacious than DPP-4 inhibitors in reducing HbA1c as monotherapy (WMD -0.37, 95% CI -0.54 to -0.20; three studies) with both agents together more efficacious than monotherapy. There was low strength evidence for metformin plus GLP-1 agonist being more effective then metformin plus a DPP-4 inhibitor. Metformin decreased mean body weight relative to thiazolidinediones (WMD -2.6, 95% CI -4.1 to -1.2; eight studies), sulfonylureas (WMD -2.7, 95% CI -3.5 to -1.9; 12 studies) or DPP-4 inhibitors (WMD -1.4, 95% CI -1.8 to -1.0; three studies). Metformin showed positive effects on cholesterol levels with reduced low density lipoprotein levels compared with pioglitazone, sulfonylureas and DPP-4 inhibitors.
Long term clinical outcomes: The evidence for treatment effects on all long-term clinical outcomes was either low-strength or insufficient. A large RCT (n=4,360) found no differences over four years in all-cause mortality, cardiovascular mortality or morbidity and stroke in groups treated with metformin, rosiglitazone or glyburide. Another large RCT found non-inferiority of rosiglitazone plus metformin and of rosiglitazone plus sulfonylurea compared to metformin plus sulfonylurea for hospitalisation or death from cardiovascular disease. Trials with shorter durations also showed no differences between interventions. Observational studies typically showed benefits for metformin over sulfonylureas for all-cause mortality and cardiovascular mortality and morbidity. Short-term RCTs showed a statistically non-significant benefit for metformin versus metformin plus rosiglitazone in fatal and non-fatal ischaemic heart disease (OR 0.43, 95% CI: 0.17 to 1.10; seven RCTs).
Results for analyses of other intermediate and safety outcomes were reported.
There was some evidence of publication bias in the analysis of metformin alone compared with metformin plus a thiazolidinedione for the outcome of HbA1c and in the analysis of metformin versus rosiglitazone for the outcome of triglyceride levels. Authors' conclusions Evidence supported use of metformin as a first-line agent for the treatment of type 2 diabetes. Most two-drug combinations reduced HbA1c levels; some also increased risks of hypoglycaemia and other adverse events. CRD commentary The review question and inclusion criteria were clear. The authors searched three relevant databases and several additional sources, which reduced the chances of publication bias. Publication bias was assessed and evidence of it was detected in two of the analyses. The decision to limit the review to studies reported in English may have led to the introduction of language bias. The authors used methods designed to reduce reviewer bias and error at all stages of the review process and conducted appropriate validity assessments for both randomised and non-randomised studies. The meta-analyses used were reasonable. Heterogeneity was assessed and attempts were made to explore its sources.
The authors' conclusions reflected the results of the review and appear likely to be reliable. Implications of the review for practice and research Practice: The authors did not state any implications for practice.
Research: The authors stated a need for further research to evaluate long-term clinical outcomes in high risk subpopulations, such as different ethnic groups, older adults and patients with comorbidities. Funding Agency for Healthcare Research and Quality contract 290-02-0018. Bibliographic details Bennett WL, Maruther NM, Singh S, Segal JB, Wilson LM, Chatterjee R, Marinopoulos SS, Puhan MA, Ranasinghe P, Block L, Nicholson WK, Hutfless S, Bass EB, Bolen S. Comparative effectiveness and safety of medications for type 2 diabetes: an update including new drugs and 2-drug combinations. Annals of Internal Medicine 2011; 154: 602-613 Other publications of related interest Bennett WL, Wilson LM, Bolen S, Maruthur N, Singh S, Chatterjee R et al. Oral diabetes medications for adults with type 2 diabetes: an update. AHRQ Publication 11-EHC038-EF. 2011.
Bolen S, Wilson L, Vassy J, Feldman L, Yeh J, Marinopoulos S et al. Comparative effectiveness and safety of oral diabetes medications for adults with type 2 diabetes. Rockville MD: Agency for Healthcare Research and Quality. AHRQ publication 07-EHC010-EF. 2007.
Bolen S, Wilson L, Vassy J, Feldman L, Yeh J, Marinopoulos S et al. Systematic Review: comparative effectiveness and safety of oral medications for type 2 diabetes mellitus. Annals of Internal Medicine 2007; 147:386-399. Indexing Status Subject indexing assigned by NLM MeSH Body Weight /drug effects; Comparative Effectiveness Research; Diabetes Mellitus, Type 2 /blood /drug therapy; Drug Therapy, Combination; Hemoglobin A, Glycosylated /metabolism; Humans; Hypoglycemic Agents /adverse effects /therapeutic use; Lipids /blood; Publication Bias AccessionNumber 12011002635 Date bibliographic record published 11/05/2011 Date abstract record published 18/05/2011 Record Status This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn. |
|
|
|