SGLT-2 Inhibitors and GLP-1 Receptor Agonists for Type 2 Diabetes

Adding SGLT-2 inhibitors and GLP-1 agonists to standard treatment reduces death, nonfatal heart attack, and severe kidney disease in Type 2 diabetes, but also increases adverse events

Benefits in NNT

SGLT-2 inhibitors:
38
1 in 38 high-risk people was helped (death prevented)
100
1 in 100 low-risk people was helped (death prevented)
71
1 in 71 high-risk people was helped (heart attack prevented)
143
1 in 143 low-risk people was helped (heart attack prevented)
No one was helped (no stroke prevented)
40
1 in 40 high-risk people was helped (end-stage kidney disease prevented)
333
1 in 333 low-risk people was helped (end-stage kidney disease prevented)
Average weight loss: 2 kg
GLP-1 receptor agonists:
59
1 in 59 was helped (death prevented, high risk)
125
1 in 125 was helped (death prevented, low risk)
111
1 in 111 was helped (heart attack prevented, high risk)
250
1 in 250 was helped (heart attack prevented, low risk)
59
1 in 59 was helped (stroke prevented, high risk)
111
1 in 111 was helped (stroke prevented, low risk)
40
1 in 40 was helped (end stage kidney disease prevented, high risk)
500
1 in 500 was helped (end stage kidney disease prevented, low risk)
Average weight loss: 1.5 kg
SGLT-2 inhibitors and GLP-1 receptor agonists versus metformin:
No difference in death, heart attack, stroke, or end-stage kidney disease
Average weight loss (0.19–4.46 kg more) compared to metformin
SGLT-2 inhibitors:
38
2.6% lower risk of death for high-risk people
100
1% lower risk of death for low-risk people
71
1.4% lower risk of heart attack for high-risk people
143
0.7% lower risk of heart attack for low-risk people
No one was helped (no stroke prevented)
40
2.5% lower risk of end-stage kidney disease for high-risk people
333
0.3% lower risk of end-stage kidney disease for low-risk people
Average weight loss: 2 kg
GLP-1 receptor agonists:
59
1.7% lower risk of death for high-risk people
125
0.8% lower risk of death for low-risk people
111
0.9% lower risk of heart attack for high-risk people
250
0.4% lower risk of heart attack for low-risk people
59
1.7% lower risk of stroke for high-risk people
111
0.9% lower risk of stroke for low-risk people
40
2.5% lower risk of end-stage kidney disease for high-risk people
500
0.2% lower risk of end-stage kidney disease for low-risk people
Average weight loss: 1.5 kg
SGLT-2 inhibitors and GLP-1 receptor agonists versus metformin:
No difference in death, heart attack, stroke, or end-stage kidney disease
Average weight loss (0.19–4.46 kg more) compared to metformin

Harms in NNT

SGLT-2 inhibitors:
7
1 in 7 was harmed (experienced a genital infection)
GLP-1 receptor agonists:
17
1 in 17 was harmed (experienced a severe gastrointestinal event)
SGLT-2 inhibitors:
7
14% higher risk of genital infections
GLP-1 receptor agonists:
17
6% higher risk of severe gastrointestinal events
View As:

Source

Davila E, McCormack J. sglt ‐2 inhibitors and glp ‐1 receptor agonists for type 2 diabetes. Academic Emergency Medicine. 2024;31(4):408-411.

Study Population: 421,346 patients with Type 2 diabetes, already on standard treatments, followed for 24 weeks or longer

Efficacy Endpoints

Death, nonfatal heart attack, nonfatal stroke, end-stage kidney disease, body weight change

Harm Endpoints

Severe hypoglycemia, severe gastrointestinal events, genital infection, amputation, ketoacidosis

Narrative

Type 2 diabetes is a condition that can affect many organs and can lead to serious complications. Recently, new classes of medications have been introduced for the treatment of Type 2 diabetes, including glucagon-like peptide 1 (GLP-1) receptor agonists and sodium-glucose transport protein 2 (SGLT-2) inhibitors. SGLT-2 medications increase the elimination of glucose and sodium in the urine by blocking the reuptake of filtered glucose in the kidney. GLP-1 receptor medications mimic the intestinal hormone incretin, increasing glucose-dependent endogenous insulin excretion. Both medications slow gastric emptying, decrease appetite, and regulate insulin and glucagon.1 Several trials have shown benefits prompting some guidelines to recommend these class of medications for patients with Type 2 diabetes (T2DM).1, 2, 3

The systematic review and network meta-analysis summarized here4 included 764 randomized trials testing SGLT-2 inhibitors or GLP-1 receptor agonists typically added to other antidiabetes medications. Trial groups received SGLT-2 or GLP-1 medications while control groups received placebos. However, both trial arms were on—and stayed on—standard background treatments that could include a variety of other medications (metformin, sulfonylureas, dipeptidyl peptidase-4 inhibitors, thiazolidinediones, alpha-glucosidase inhibitors, glitinides, or insulin).4 A total of 421,346 patients were involved in the 764 studies.

Outcomes of interest included death, nonfatal stroke, end-stage kidney disease, nonfatal heart attack, body weight change, severe hypoglycemia, severe gastrointestinal events, genital infection, ketoacidosis, amputation, and hyperkalemia. End-stage kidney disease is defined in studies as a glomerular filtration rate <15 mL/min (per 1.73 m2) or initiation of dialysis. Tables 1 and 2 show the results found in the review. Of note, the review reports medication effects according to a patient's baseline cardiovascular risk. This is because the magnitude of the effect of diabetes medications varies according to a person's chance of developing the problem the medication is aiming to prevent. For instance, among those already at little to no risk of having a stroke, a medication's ability to demonstrate a reduction in strokes is obviously small. For those who are at higher risk of future strokes, there is a greater possibility for improvement, and effective medications can have a greater impact. Effective medications therefore tend to have different impacts in people with different risks—the higher the risk, likely the greater the impact. The authors therefore analyzed and reported medication effects separately for people in each of the following risk categories: very low (fewer than three cardiovascular risk factors), low (three or more risk factors), moderate (patients who already had known cardiovascular disease), high (those with chronic kidney disease), and very high (known cardiovascular and kidney disease).4

Major harms included severe hypoglycemic episodes, severe gastrointestinal adverse events, genital infections, amputation, and ketoacidosis. SGLT-2 inhibitors increased genital infections (odds ratio [OR] 3.5, 95% confidence interval [CI] 3.0–4.0, absolute risk difference [ARD] 14%; NNH 7; high certainty). GLP-1 receptor agonists increased the risk of severe gastrointestinal symptoms (OR 2.5, 95% CI 1.2–5.0; ARD 6%; NNH 17; low certainty). There were no other differences in risks of harm between SGLT-2 inhibitors and GLP-1 receptor agonists.

Another recent review by Shi et al.5 conducted a similar analysis; however, it included trials comparing SGLT-2 and GLP-1 medications to standard treatments without a placebo control. After reviewing a number of original trials, we made the judgment that comparing these medications to placebo paints a more accurate picture of their effects, and we therefore used the Palmer meta-analysis as our primary source (though most relative risks in the two reviews are quite similar).4

The meta-analysis by Shi et al. also, however, included trials of the new medication tirzepatide which is a glucose-dependent insulinotropic polypeptide (GIP) plus a GLP-1 medication. For completeness we are including here the weight loss impact of tirzepatide (average 8.6 kg, 95% CI 7.8–9.4 kg) and its attendant increase in severe gastrointestinal symptoms (OR 4.6, 95% CI 1.9–11.1; ARD 12.5%; NNH 8; moderate certainty), estimates from the review by Shi et al.5 There was no statistical difference between tirzepatide and standard treatment for all other outcomes of interest.

Importantly, when any of these medication classes were compared directly to metformin, no additional benefit was seen other than greater weight loss of varying degrees (0.2–7.7 kg).4, 5, 6 However, patients on metformin had fewer genital infections and gastrointestinal adverse events.4, 5, 6 While head-to-head trials are comparatively few, a large population-based study of the new medications from 2022 appears to confirm their lack of benefit over metformin, showing almost 9000 on SGLT-2 medications, when matched to over 17,000 on metformin, had identical rates of heart attack, stroke, and death but higher rates of genital infection.7

Caveats

Notable limitations exist in these data. Background treatments varied from trial to trial, making precise comparisons impossible. A total of 687 of the included 764 trials were rated as high risk of bias in at least one of six domains.4 Many endpoints were surrogate outcomes or laboratory values; however, the huge number of studies meant there were enough reporting patient-oriented outcomes to overcome this. There are a paucity of studies comparing SGLT-2 to either dual GIP/GLP-1 or GLP-1 medications directly, prohibiting conclusions about one class over another. There were also significant differences in trial follow-up periods, which could result in over- or underestimating effects.

In summary, while SGLT-2 and GLP-1 medications appear to reduce mortality, heart attack, and end-stage kidney disease compared to many other classes, available evidence does not show them to be better than metformin, though some (particularly tirzapetide) may lead to greater weight loss. The newer medications also cause genital infections, particularly SGLT-2s, at a NNH of 7. The cost and harms of SGLT-2 and GLP-1 medications therefore should be balanced against weight loss and other effects compared to metformin.

We have assigned an NNT color recommendation of green (benefits > risk) for SGLT-2 inhibitors and GLP-1 receptor agonists in patients with Type 2 diabetes for improving most patient-centered outcomes over most medications (as well as in addition to a patient's current diabetes medication regimen). However, there are important increases in harms, particularly genital infections with SGLT-2 inhibitors and severe gastrointestinal adverse events with GLP-1 medications. Therefore, if exclusively comparing the medications to metformin we would assign a color of yellow, as more head-to-head trials are needed to determine if the new medications improve outcomes greater than other agents (for instance metformin).

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

Author

Esteban Davila, MD; James McCormack, BSc (Pharm), PharmD
Supervising Editors: Shahriar Zehtabchi, MD

Published/Updated

April 23, 2024

References: