June 29, 2020
The largest network meta-analysis of diabetes drugs to date indirectly compared 21 glucose-lowering drugs for adults with type 2 diabetes in nine drug classes and 453 trials — to help guide therapy in a "challenging, complex landscape."
Or, in the words of an editorialist, to aid prescribers in selecting "the right diabetes medication for the right patient for the right outcome."
The analysis by Apostolos Tsapas, MD, PhD, Aristotle University of Thessaloniki, Greece, and colleagues was published online June 29 in Annals of Internal Medicine.
The researchers identified published English-language randomized trials of glucose-lowering drugs — including sodium-glucose cotransporter-2 (SGLT2) inhibitors and glucagon-like peptide-1 (GLP-1) agonists — that were at least 24 weeks long and reported glycemic, vascular, and mortality outcomes.
Drug-Naive Patients, Low or High CVD Risk
For drug-naive patients, the main findings and implications are:
- For drug-naive patients at low cardiovascular risk, the use of metformin seems justified.
- For drug-naive patients at increased cardiovascular risk, optimal initial treatment is unclear due to a lack of pertinent evidence.
"In drug-naive patients, all medications except for DPP-4 inhibitors were as efficacious as metformin in reducing hemoglobin A1c level," Tsapas and colleagues write.
In drug-naive patients at low cardiovascular risk, there were no differences in mortality and vascular outcomes among medications.
"We did not identify any trials exclusively recruiting drug-naive patients at increased cardiovascular risk," the researchers note.
Metformin Background Therapy, Low or High CVD Risk
For patients taking metformin, the main findings and implications are:
- For patients at low cardiovascular risk receiving metformin-based background therapy, vascular outcomes were similar for different therapies, so therapy should be chosen based on their effect on other efficacy and safety outcomes.
- For patients at increased cardiovascular risk receiving metformin-based background therapy, the optimal choice between specific GLP-1 agonists and SGLT2 inhibitors should be based on the cardiovascular profile of individual agents and guided by patients' personal preferences and therapeutic priorities.
Tsapas and colleagues elaborate that "when used as an add-on to metformin-based therapy, insulin regimens and GLP-1 [agonists] were the most efficacious in reducing hemoglobin A1c level, whereas sulfonylureas, basal–bolus insulin therapy, and premixed insulin increased the odds of severe hypoglycemia."
In patients at low cardiovascular risk receiving metformin-based background therapy, vascular outcomes with add-on therapies were similar to those with placebo.
In patients at increased cardiovascular risk receiving metformin-based background therapy, adding oral semaglutide, empagliflozin, or liraglutide reduced all-cause mortality and cardiovascular death, whereas adding extended-release exenatide or dapagliflozin only reduced all-cause mortality.
Both dulaglutide and subcutaneous semaglutide lowered the odds of stroke.
Subcutaneous semaglutide was associated with increased odds of diabetic retinopathy.
The incidence of amputation was increased with canagliflozin and reduced with liraglutide.
Compared with previous network meta-analyses, the current study was much larger (453 trials; 320,474 patients), and included "recently published cardiovascular or renal outcome trials and trials with novel agents, such as oral semaglutide," Tsapas and colleagues write.
However, study limitations include the fact that patients at low cardiovascular risk were a mixed population with unknown or variable cardiovascular risk.
Also, the definition of increased cardiovascular risk was inconsistent, and potential differences in baseline renal function may have affected the findings for end-stage renal disease.
And metformin background therapy differed in the individual trials.
Filling the Knowledge Gaps but Key Questions Remain
In the accompanying editorial, Christine G. Lee, MD, and William T. Cefalu, MD, from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), in Bethesda, Maryland, begin by acknowledging that "the landscape of diabetes care has grown increasingly complex" in recent years.
Guidelines recognize the need to consider patient differences (such as diabetes duration and risk of CVD, adverse drug effects, and comorbidities) to be able to set individualized treatment goals.
However, the cardiovascular outcome trials (CVOTs) for SGLT2 inhibitors and GLP-1 agonists show "benefits for both cardioprotection and renoprotection beyond glycemic control, which has made the selection of drugs for different therapeutic goals more challenging," especially since there is a paucity of head-to-head trials of these newer medications.
"A key clinical question that remains even today," the editorialists write, "is whether any of these newer medications can be used as initial monotherapy" and have better glycemic, CVD, and mortality outcomes than metformin.
And it is not clear whether patients at low cardiovascular risk would benefit more from an SGLT2 inhibitor or a GLP-1 agonist added to metformin.
Therefore, the large network meta-analysis by Tsapas and colleagues is useful in trying to fill this "knowledge gap." Among its findings, it confirmed that "diabetes drugs in all classes in both drug-naive and metformin-treated patients are more effective in lowering hemoglobin A1c levels than placebo," according to Lee and Cefalu.
The analysis also identified that for patients with type 2 diabetes who were receiving metformin and had a high risk of CVD, certain GLP-1 agonists and SGLT2 inhibitors significantly decreased the risks for all-cause and cardiovascular mortality.
However, although outcomes such as amputation, diabetic eye disease, and sexual dysfunction are important to patients, few trials collected these data.
"Clinical trials should attempt to enroll a more heterogeneous patient population that includes the full range of cardiovascular risk and diabetes duration and should try to categorize these factors similarly," Lee and Cefalu urge.
"Furthermore, the use of a core set of standard secondary outcomes that include patient-important outcomes would also facilitate comparisons for outcomes of greatest relevance to our patients."
39 Ongoing or Just Completed Unpublished Trials
Tsapas and colleagues add that an April 2020 search of ClinicalTrials.gov retrieved 39 ongoing or recently completed but unpublished pertinent trials that "are expected to strengthen the evidence base about the effects of GLP-1 [agonists] and SGLT2 inhibitors on vascular endpoints."
These studies include two ongoing CVOTs with oral semaglutide (NCT03914326) and dapagliflozin (NCT03982381) and the recently completed CVOT for ertugliflozin (NCT01986881), VERTIS-CV, the results of which were reported as part of the virtual American Diabetes Association (ADA) 80th Scientific Sessions.
There are three ongoing renal outcome trials with subcutaneous semaglutide (NCT03819153), dapagliflozin (NCT03036150), and empagliflozin (NCT03594110), and the renal outcome trials for dapagliflozin and empagliflozin also recruited patients without type 2 diabetes, provided they have chronic kidney disease.
And two ongoing trials with empagliflozin (NCT03057977 and NCT03057951) and one trial with dapagliflozin (NCT03619213) are evaluating the long-term effects of these agents on the composite outcome of cardiovascular death or hospitalization for heart failure in patients with heart failure with or without type 2 diabetes.
The study was funded by the European Foundation for the Study of Diabetes Patient-Centred Treatment to Support a Holistic Approach Toward Type 2 Diabetes (PACT) program, supported by an unrestricted educational grant from AstraZeneca. Disclosures for the authors and editorialists are listed with the article and editorial.