Recent developments in GLP‐1RA therapy: A review of the latest evidence of efficacy and safety and differences within the class

As of 2021, 9 % of the global population is now affected by type 2 diabetes mellitus (T2DM). In the United Kingdom, 20% of National Health Service in-patients suffer from diabetes as a co-morbidity, while 10% of healthcare expenditure is linked to diabetes. Failed attempts to reduce the diabetes epidemic are due to an inability to halt rising levels of obesity and sedentary lifestyle. This has led to the development of an abundance of therapeutic agents to control hyperglycaemia and in the United States, there are now 12 different classes of glucose-lowering medication. The glucagon-like peptide 1 (GLP-1) receptor agonists (RAs) act in a glucose-dependent manner by enhancing insulin secretion and reducing production of glucagon. GLP-1RAs are peptides that are resistant to dipeptidyl peptidase and provide supra-physiological stimulation of the GLP-1 receptor. They also retard gastric emptying and increase satiety; these effects in tandem lead to weight loss in a substantial proportion of people with T2DM. Exenatide was the first GLP-1RA to be launched and was based on the exendin-4 molecule (isolated from Gila Monster lizard saliva) and is administered by twice-daily subcutaneous injection at meal times due to its short half-life.1 Subsequent GLP-1RAs were longer acting, once daily (OD) preparations (liraglutide and lixisenatide) administered without regard to meals.2, 3 Then in 2011, the European Medicines Agency (EMA) issued authorisation for the first long-acting GLP-1RA, a once-weekly (QW) version of exenatide.4 Currently, the most widely prescribed GLP-1RA is liraglutide, however, the convenience of less frequent dosing has led to an expansion of new prescriptions of the once-weekly GLP-1RAs.5 Unfortunately, exenatide extended-release (ER) is manufactured as microspheres suspended in a vehicle of medium-chain triglycerides, which requires vigorous shaking (>15 seconds) prior to injection.6 Injection-site swelling is also a prevalent issue due to the slow clearance of the co-formulated polymer, which does not resolve itself until several weeks following injection. As a result, exenatide ER is less frequently prescribed and this review will focus on dulaglutide and semaglutide. Dulaglutide was launched in 2014 and is a polypeptide analogue of human GLP-1 covalently linked to the Fc arms of human immunoglobulin G4.7 This structure reduces immunogenicity and is soluble allowing for simple administration of a clear solution. Dulaglutide is administered in four doses (0.75, 1.5, 3.0 and 4.5 mg QW) and the lower two doses have no need for up-titration. More recently, the long-acting GLP-1RA semaglutide was approved (20188). This molecule is 94% homologous with human GLP-1, differing by only two amino acids. There is an 18-carbon fatty di-acid chain attached to amino acid 26 of the molecule that provides strong binding to albumin and facilitates prolonged activity. A benefit of the long-acting GLP-1RA semaglutide is its administration as a clear solution that does not require resuspension. It is initiated at a dose of 0.25 mg QW titrated after 1 month to a maintenance dose of 0.5 mg with the option for further up-titration to 1.0 mg QW. The dulaglutide phase 3 clinical trial programme was known as AWARD (Assessment of Weekly AdministRation of LY2189265 [dulaglutide] in Diabetes) and assessed its glucose-lowering efficacy versus commonly prescribed second-line therapies. The spectrum of patients with T2DM was broad, extending from treatment-naïve to those needing insulin, and the primary end-point in each trial (AWARD 1-6) was a change in HbA1c.9-14 The primary endpoint was assessed at 26 weeks in AWARD 1 and 3 to 6 and at 52 weeks in AWARD 2 (versus insulin glargine U100). Doses of 0.75 and 1.5 mg QW dulaglutide were assessed and all studies had extended periods of observation (4-78 weeks) allowing for the collection of additional safety data. Both doses of dulaglutide achieved statistically superior glucose-lowering to the following active comparators: metformin; sitagliptin; exenatide and insulin glargine U100. The higher dose of dulaglutide was also non-inferior to liraglutide 1.8 mg OD. From a baseline HbA1c of 59.6 to 69.4 mmol/mol (7.6%-8.5%), the reductions in HbA1c were 7.8 to 17.4 mmol/mol (0.71%-1.59%) for dulaglutide 0.75 mg QW and 8.5 to 17.9 mmol/mol (0.78%-1.64%) for dulaglutide 1.5 mg QW. Post-launch, the dose of dulaglutide for glucose-lowering was increased following the AWARD 11 trial.15 This assessed the efficacy of dulaglutide 3.0 and 4.5 mg QW versus dulaglutide 1.5 mg QW and reported additional glucose-lowering of 1.9 mmol/mol (0.17%) and 3.7 mmol/mol (0.34%), respectively, which were both statistically significant after 36 weeks treatment. Phase 3 clinical trials of semaglutide were named SUSTAIN (Semaglutide Unabated Sustainability in Treatment of Type 2 diabetes) and included 8416 people with T2DM (SUSTAIN 1-5).16-20 Semaglutide was assessed in people across the spectrum of T2DM, from treatment-naïve, through combinations with oral glucose-lowering agents and insulin. All were randomised controlled trials (RCTs) and assessed HbA1c lowering over 30 to 56 weeks versus placebo, sitagliptin, exenatide ER and insulin glargine U100. The fall in HbA1c was significantly greater for both doses of semaglutide versus comparator and ranged from 13.2 to 16.0 mmol/mol (1.2%-1.5%) for 0.5 mg QW and 16.8 to 20.2 mmol/mol (1.6%-1.9%) for semaglutide 1.0 mg QW. An application to the Food and Drug Administration (FDA) for a 2.0 mg QW dosing of semaglutide for glucose-lowering was refused in March 2021 but resubmission is anticipated.21 The SUSTAIN 7 trial was a direct comparison of semaglutide with dulaglutide.22 Patients with T2DM and suboptimal glycaemic control on metformin were randomised to dulaglutide 0.75 or 1.5 mg QW versus semaglutide 0.5 or 1.0 mg QW. The mean BMI was 33.1 to 33.7 kg/m2 and diabetes duration 7.0 to 7.7 years. The mean baseline HbA1c was 66.1 to 67.2 mmol/mol (8.2%-8.3%) and fell significantly in all treatment cohorts after 40 weeks. Comparison of the low and high doses of each drug showed a significant advantage for semaglutide over dulaglutide (semaglutide 0.5 mg QW 16.5 mmol/mol (1.5%) versus dulaglutide 0.75 mg QW 12.1 mol/mol (1.1%) and semaglutide 1.0 mg QW 19.4 mmol/mol (1.8%) versus dulaglutide 1.5 mg 14.9 mmol/mol [1.4%]). Dulaglutide 1.5 mg QW and exenatide exhibited similar effects on bodyweight in AWARD-1 but patients receiving the 0.75 mg QW dose experienced a small weight gain compared with exenatide.9 In the AWARD-3 monotherapy study, there was similar weight loss with dulaglutide 1.5 mg QW compared with metformin but significantly less with the lower 0.75 mg QW dose.11 In AWARD-5, weight loss was significantly greater for both dulaglutide doses compared with sitagliptin 100 mg OD13 and in the two head-to-head comparisons with insulin glargine U100 (AWARD-2 and AWARD-4), there was a significant weight advantage for dulaglutide.10, 12 In AWARD-6, patients taking dulaglutide 1.5 mg QW showed significantly less weight loss versus liraglutide 1.8 mg OD.14 In the AWARD-11 trial, the higher doses of dulaglutide led to increased weight loss compared with 1.5 mg QW (0.9 Kg for 3.0 mg QW and 1.6 kg for 4.5 mg QW).15 Reduction in weight was a secondary endpoint in the SUSTAIN programme and weight loss with semaglutide 0.5 and 1.0 mg QW was significantly better than placebo, sitagliptin and insulin glargine U100.22 Semaglutide 0.5 mg QW was associated with 3.5 to 4.6 kg weight loss, equivalent to 3.7% to 4.8% reduction of baseline body mass. The 1.0 mg QW semaglutide dose achieved reductions of 4.5 to 6.5 kg (4.7%-7.0% baseline body mass). Significant weight loss was also seen when semaglutide was added to basal insulin therapy. Finally, in the head-to-head comparison of semaglutide versus dulaglutide, weight loss data showed highly significant findings for the comparison of the two doses of semaglutide versus dulaglutide, with reductions of 4.6 and 6.5 kg (semaglutide 0.5 and 1.0 mg QW) versus 2.3 and 3.0 kg (dulaglutide 0.75 mg and 1.5 mg QW), respectively.22 The clear advantage of semaglutide over dulaglutide in weight loss has led to the former being assessed in high dose as an anti-obesity agent.24 The Semaglutide Treatment Effect in People with obesity (STEP) programme is made up of five studies of overweight individuals (mean BMI 35.7-38.5 kg/m2). STEP 1 assessed the efficacy and safety of subcutaneous semaglutide 2.4 mg QW versus placebo in 1961 adults treated for 68 weeks.25 The trial met both co-primary endpoints with a statistically significant bodyweight reduction with semaglutide 2.4 mg QW compared to placebo. Semaglutide-treated subjects achieved a mean weight loss of 14.9% (from baseline bodyweight of 105.3 kg) compared to a 2.4% weight loss with placebo; 86.4% achieved weight loss >5% compared to 31.5% with placebo. The most common adverse events (AEs) among people treated with high-dose semaglutide were gastrointestinal (GI) and these were typically transient and mild or moderate in severity. STEP 2 assessed semaglutide 2.4 versus 1.0 mg QW and placebo in 1210 overweight or obese adults with type 2 diabetes again treated for 68 weeks.26 Estimated change in mean bodyweight was −9.6% with semaglutide 2.4 mg QW versus −3.4% with placebo; 68.8% of subjects on semaglutide 2.4 mg lost >5% of bodyweight compared with 28.5% on placebo (P < .0001). AEs were more frequent with semaglutide 2.4 and 1.0 mg QW compared with placebo (being 87.6%, 81.8% and 76.9%, respectively). STEP 3 compared the effects of semaglutide 2.4 mg QW with placebo for weight management in adults with overweight or obesity, as an adjunct to a low-calorie diet for 8 weeks and intensive behavioural therapy (made up of 30 counselling visits).27 A total of 611 participants were followed for 68 weeks with the same co-primary outcomes as in STEPs 1 and 2. The estimated mean bodyweight change from baseline was −16.0% for semaglutide versus −5.7% for placebo (P < .001); 86.6% of semaglutide participants lost >5% baseline bodyweight versus 47.6% who received placebo (P < .001). Treatment discontinuation due to GI side-effects occurred in 3.4% of semaglutide participants compared with 0% placebo. STEP 4 reported on 803 participants who received 20 weeks treatment with subcutaneous semaglutide, escalated to 2.4 mg QW who were then randomised to continued drug versus placebo (both with continued lifestyle intervention) for a further 48 weeks.28 The recruits who continued to receive semaglutide had a further mean weight loss of 7.9% over the second period of the trial whereas those who switched to placebo gained a mean of 6.9%, giving a difference of 14.8% in favour of semaglutide. This indicates that maintenance of semaglutide is necessary to continue weight reduction and avoid weight regain. As a result of the STEP programme data, the FDA has recently approved once-weekly semaglutide for weight loss; a licence for this indication has not yet been granted in the European Union. The SELECT study commenced in 2018 and is a randomised, double-blind, placebo-controlled trial to determine the impact of semaglutide 2.4 mg QW on cardiovascular outcomes in overweight or obese participants with cardiovascular disease who do not have diabetes.29 The primary endpoint is a composite of the major adverse cardiovascular events (MACE), cardiovascular death, non-fatal MI or non-fatal stroke (3-component MACE). The study is due to enrol around 17 500 participants with estimated completion in 2023 and is the first cardiovascular outcome trial (CVOT) in obesity powered to detect superiority of a therapeutic intervention.30 GLP-1RAs are known to raise the pulse rate and to reduce blood pressure (BP) although they are not approved as blood pressure lowering therapies.31, 32 Dulaglutide was assessed in a study using ambulatory BP and pulse monitoring over a 26-week period.33 Seven-hundred fifty-five subjects with T2DM received either dulaglutide 0.75 mg QW, dulaglutide 1.5 mg QW or placebo and were subjected to three 24-hour periods of ambulatory monitoring. After 26 weeks, the placebo-corrected systolic BP fell by 1.7 mmHg (dulaglutide 0.75 mg QW) and 2.7 mmHg (1.5 mg QW). Placebo-corrected increases in diastolic BP of 0.2 and 0.5 mmHg, respectively were observed. Heart rate increases of 1.3 beats per minute (BPM) and 3.5 BPM were seen at 26 weeks for the 0.75 and 1.5 mg dulaglutide doses. In the SUSTAIN phase 3 programme, therapy with semaglutide 0.5 mg QW was associated with a fall in systolic blood pressure (SBP) of 2.4 to 5.1 mmHg whilst the 1.0 mg QW dose led to a fall of 2.7 to 6.3 mmHg.23 This effect was significantly greater than that seen with comparators except for placebo in SUSTAIN 116 and dulaglutide in SUSTAIN 7.22 Diastolic blood pressure was also lowered although to a lesser extent, and generally not significantly different from comparators (apart from SUSTAIN 7 where semaglutide achieved a larger reduction than dulaglutide22). Across the programme, semaglutide caused an increase in pulse rate of 1 to 4 BPM. In SUSTAIN 7, semaglutide 1.0 mg QW showed a greater increase in heart rate than dulaglutide 1.5 mg QW (4.0 versus 2.4 BPM).23 Cardiovascular outcome trials (CVOTs) were mandated by the FDA in the United States in 2008 for new glucose-lowering medicines.34 This followed on from a controversy as to whether rosiglitazone, a thiazolidinedione, caused an increase in myocardial infarction in people with T2DM (who are already at a significantly increased risk).35 Both dulaglutide and semaglutide have been examined in placebo-controlled CVOTs. The “Researching CV Events with a Weekly INcretin in Diabetes” (REWIND) study was the CVOT for dulaglutide and it recruited 9901 people with T2DM, HbA1c < 9.5% (90 mmol/mol) and an estimated glomerular filtration (eGFR) rate ≥ 15 mL/min.36 The mean age of participants was 66 years, 55% were male and 69% did not have known cardiovascular disease. Participants were given dulaglutide 1.5 mg QW or placebo on top of standard of care. The primary end-point was the three-component MACE (as above). “Standard of care” is mandated by the FDA and requires optimal management of hypertension, LDL-cholesterol and use of anti-platelet therapies in diabetes CVOTs, as well as aiming for equivalent glucose-lowering in each of the trial arms.34 In 12% of dulaglutide participants, the primary endpoint was observed, versus 13% placebo, and resulted in a hazard ratio (HR) of 0.88 with confidence intervals (CI) between 0.79 and 0.99. Hence, the statistically significant superiority of dulaglutide was demonstrated (P = 0.026). The REWIND trial had the longest median follow-up and lowest placebo event rate of the diabetes CVOTs. This was because only a minority of trial participants (31%) had prior cardiovascular disease (defined as previous MI, ischaemic stroke, revascularisation, unstable angina with ECG changes or myocardial ischaemia on imaging or stress test). It is of note that the HR (0.87, CI 0.74-1.02) was identical in both the primary and secondary prevention cohorts, showing that the overall outcome was not driven by the results from high-risk subjects. This was in contrast to results from the LEADER CVOT trial for liraglutide, which appeared to show little benefit in lower-risk participants.37 Subcutaneous semaglutide was evaluated in the SUSTAIN 6 trial,38 which was designed as a non-inferiority (safety) study, to be followed by a larger superiority study following drug approval. SUSTAIN 6 was, therefore, much smaller and shorter than REWIND. Subjects were allocated to semaglutide 0.5 and 1.0 mg QW or volume-matched placebo, in addition to cardiovascular standard of care. SUSTAIN 6 enrolled 3297 people with T2DM at high CV risk and was both event and time-driven, patients having 2 years exposure to trial product. High CV risk was defined by two categories; age ≥ 50 years and an established CV disease (cerebrovascular disease, coronary heart disease, chronic kidney disease [stage 3 or higher], peripheral vascular disease or chronic heart failure of New York Heart Association classes II-III), or age > 60 years and one or more CV risk factors. The primary outcome was, once again, the three-component MACE, which occurred in 6.6% people receiving semaglutide versus 8.9% placebo (HR 0.74 (95% CI [CI] 0.58-0.95). This 26% reduction in HR for the primary endpoint demonstrated statistically significant CV safety for semaglutide (P < .001); a test for superiority was also significant (P < .02) although this was not pre-specified (because it was completely unexpected). For the reasons outlined below, the follow-on superiority CVOT of subcutaneous semaglutide was never undertaken. Both dulaglutide and semaglutide have demonstrated CV safety and superiority over placebo in CVOTs, albeit with differing levels of trial evidence. Both are now recommended as second-line therapy for glucose-lowering after metformin in people with T2DM and atherosclerotic CVD (ASCVD) or even as first-line, prior to metformin, in some cardiology guidelines.39 Analysis of the individual parts of the three-component MACE in both REWIND and SUSTAIN 6 show that non-fatal stroke was significantly reduced in these GLP-1RAs. In REWIND, dulaglutide 1.5 mg QW reduced the HR by 24% (HR 0.76; CI 0.61-0.95, P = .017)) and in SUSTAIN 6 the HR was 0.61 (CI 0.38-0.99, P = 0.04,36, 38). An exploratory analysis further examined stroke outcomes in REWIND.40 In the trial, strokes were categorised into three sub-groups; fatal or non-fatal, ischaemic, haemorrhagic, or undetermined. The modified Rankin scale was used to determine stroke severity. During the median follow-up of 5.4 years, 3.2% of 4949 participants assigned to dulaglutide (N = 158) and 4.1% of 4952 placebo participants (N = 205) suffered from a stroke (HR 0.76, CI 0.62-0.94; P = .010). Dulaglutide reduced the incidence of ischaemic stroke (HR 0.75, CI 0.59-0.94, P = .012), but did not impact the risk of haemorrhagic stroke (HR 1.05, CI 0.55-1.99; P = .89). Dulaglutide was also found to reduce the composite of non-fatal stroke or all-cause death (HR 0.88, CI 0.79-0.98; P = .017) and disabling stroke (HR 0.74, CI 0.56-0.99; P = .042). Treatment assignment did not affect the degree of disability post-stroke. The authors interpretation was that long-term use of dulaglutide may reduce clinically relevant ischaemic stroke in people with T2DM but did not affect stroke severity. Another exploratory analysis of REWIND focussed on cognitive impairment. In the study, both the Montreal Cognitive Assessment (MoCA) and Digit Symbol Substitution Test (DSST) were used to assess cognitive function.41 The exploratory primary cognitive outcome was the first occurrence of a follow-up score on MoCA or DSST that was 1.5 standard deviations or more below the baseline mean score in the participant's country. Baseline and follow-up scores were available for 8828 trial participants of whom 4456 received dulaglutide and 4372 placebo. In dulaglutide participants, the cognitive outcome occurred in 4.05 per 100 patient-years versus 4.35 per 100 patient-years in participants receiving placebo (HR 0.93, CI 0.85-1.02; P = .11). After a post hoc adjustment for individual standardised baseline scores, the HR for substantive cognitive impairment was reduced by 14% in those assigned dulaglutide (HR 0.86, CI 0.79-0.95; P = .0018) suggesting that long-term therapy might reduce cognitive impairment in people with T2DM. A post hoc analysis on pooled data from three CVOTs including two using semaglutide was performed to examine dementia-related AEs.42 A total of 15 820 patients with median follow-up of 3.6 years were included in this analysis and AEs were identified using standardised MedDRA (version 21.1) query for “dementia” narrow search terms. Across the three CVOTs, 15 GLP-1 RA-treated patients and 32 placebo-treated patients were identified with the development of dementia, giving a significant estimated HR of 0.47 (CI 0.25-0.86) in favour of the GLP-1 RA treatment. A placebo-controlled RCT of semaglutide in people with early Alzheimer's disease (EVOKE) is due to begin recruitment in 2021.43 There are small studies in Parkinsons disease (PD) suggesting that exenatide may improve motor impairment and a 2-year placebo-controlled RCT of exenatide ER (EXENATIDE-PD3) is currently recruiting patients.44-46 In-keeping with this, are studies suggesting that semaglutide may have a neuroprotective effect in animal models of PD. Zhang et al examined the neuroprotective effects of semaglutide in the MPTP mouse model and showed reduced motor impairment and changes consistent with protection of dopaminergic neurons.47 A clinical trial testing semaglutide in PD patients is planned.48 In common with other diabetes CVOTs, the REWIND trial investigators examined other efficacy and safety outcomes apart from the primary cardiovascular endpoints. An exploratory analysis of REWIND was simultaneously published with the main study to assess the effect of dulaglutide on the renal component of the composite microvascular outcome.49 This was defined as the first occurrence of new macroalbuminuria (UACR >33.9 mg/mmol [>300 mg/g]), a sustained decline in eGFR of 30% or more from baseline, or chronic renal replacement therapy. At baseline, 791 (7.9%) subjects had macroalbuminuria and mean eGFR was 76.9 mL/min/1.73 m2. During follow-up, the renal composite outcome occurred in 848 (17.1%) participants in the dulaglutide group and in 970 (19.6%) placebo participants (HR 0.85, CI 0.77-0.93; P = .0004). The major driver was new-onset macroalbuminuria (HR 0.77, CI 0.68-0.87; P < .0001), with HRs of 0.89 (CI 0.78-1.01; P = .066) for sustained decline in eGFR of 30% or more and 0.75 (CI 0.39-1.44; P = .39) for chronic renal replacement therapy. Although the HR for a sustained decline in eGFR of 30% was non-significant, the favourable trend led the authors to perform a series of sensitivity analyses. These showed that dulaglutide was associated with a significantly reduced incidence of a sustained eGFR decline of 40% or more (HR 0.70, CI 0.57-0.85) and 50% or more (HR 0.56, CI 0.41-0.76).49 These results are also supportive of the findings from the AWARD-7 trial.50 This was a 26-week study comparing dulaglutide 0.75 and 1.5 mg QW with insulin glargine U100 in participants with T2DM and moderate or severe CKD. The trial reported comparable glycaemic control, however, participants receiving dulaglutide experienced greater weight loss and less hypoglycaemia than those receiving insulin glargine. In addition, eGFR decline was mitigated and albuminuria was reduced: these benefits were most evident when the UACR exceeded 3.39 mg/mmol (30 mg/g). In SUSTAIN 6, renal microvascular outcomes were pre-specified secondary outcomes and there was a significant reduction of the composite renal endpoint (HR 0.64; CI 0.46-0.88; P = .005).38 This success was driven by a fall in new cases of persistent macroalbuminuria (2.5% versus 4.9%) whereas trial participants who experienced a doubling of serum creatinine and/or needed continuous renal replacement therapy was a small number and similar between groups. This improvement in renal composite outcome driven by reductions in albuminuria rather than “hard” clinical renal end-points is not unusual and has been seen with liraglutide and lixisenatide (as well as with linagliptin, a DPP4-inhibitor).37, 51, 52 To settle the question definitively regarding reno-protection, a placebo-controlled trial of semaglutide with primary renal endpoints is currently ongoing and expected to report in 2024.53 The primary endpoint for this study is time to first occurrence of a composite of persistent eGFR decline of greater than or equal to 50% from baseline, ESKD (eGFR <15 mL/min/1.73 m2, dialysis or transplantation), death from kidney disease or death from cardiovascular disease. A post hoc analysis of 1499 participants from AWARD-1, AWARD-5, AWARD-8 and AWARD-9 evaluated the effect of dulaglutide 1.5 mg QW on liver and metabolic parameters in people with T2DM and in a subgroup with possible non-alcoholic fatty liver disease (NAFLD) or non-alcoholic steatohepatitis (NASH).5, 9, 54-56 Thresholds of alanine aminotransferase (ALT) ≥30 IU/L in men and ≥ 19 IU/L in women were used to determine the NAFLD/NASH subgroup. In the overall population at 6 months, dulaglutide significantly reduced ALT, aspartate transaminase (AST) and γ-glutamyl transpeptidase (γGT) levels versus placebo. In the subgroup with presumed NAFLD/NASH, more pronounced reductions in ALT were observed with dulaglutide versus placebo (−8.8 vs −6.7 IU/L) whereas there was no significant difference between groups when the baseline ALT was in the normal range. The impact of dulaglutide on liver fat was subsequently assessed in an open-label, parallel-group RCT.57 Sixty-four adults with T2DM and MRI-derived proton density fat fraction-assessed liver fat content (LFC) ≥ 6.0% at baseline, were randomly assigned to receive dulaglutide QW for 24 weeks or usual care. Dulaglutide resulted in a control-corrected absolute change in LFC of −3.5% (CI −6.6, −0.4; P = .025) and relative change of −26.4% (−44.2, −8.6; P = .004). Dulaglutide-treated participants also showed a significant reduction in γGT levels. The authors suggested that dulaglutide could be considered for the early treatment of NAFLD in people with T2DM. Considering semaglutide liver data, analysis of ALT and high-sensitivity C-reactive protein (hsCRP) from SUSTAIN 6 and a 52-week weight management trial (semaglutide 0.05-0.4 mg/day) was undertaken.58 In weight management trial participants, 52% presented with elevated baseline ALT (defined as above) and end-of-treatment ALT reductions were seen in 6% to 21% (P < .05 for doses ≥ 0.2 mg/day)) and hsCRP reductions 25% to 43% vs placebo (P < .05 for 0.2 and 0.4 mg/day). Normalisation of elevated ALT occurred in 25% to 46% of weight management trial subjects versus 18% on placebo. Elevated baseline ALT was present in 41% of SUSTAIN 6 subjects. In these subjects, no significant ALT reduction was noted for semaglutide 0.5 mg QW whereas a significant 9% reduction versus placebo was seen for semaglutide 1.0 mg QW (P = .0024). Treatment ratios for changes in ALT and hsCRP were, however, not statistically significant after adjustment for weight change. Newsome et al conducted a 72-week, double-blind trial in people with biopsy-confirmed NASH and stage F1-3 liver fibrosis who were randomised to various doses of subcutaneous semaglutide OD or placebo.59 The primary end-point was resolution of NASH with no worsening of fibrosis and 320 people (230 with F2 or F3 fibrosis) were included. The percentage of subjects who satisfied the primary end-point was 36% to 59% of the semaglutide-treated patients and 17% placebo; this was significant (P < .001) for semaglutide 0.4 mg OD versus placebo. A confirmatory secondary end-point of improvement of at least one fibrosis stage occurred in 43% of semaglutide 0.4 mg OD patients and 33% of placebo (P = .48). The authors concluded that treatment with semaglutide resulted in a significantly higher percentage of patients with NASH resolution than placebo, however, without an improvement of fibrosis stage. The AEs seen with dulaglutide and semaglutide are mainly GI (nausea, vomiting and diarrhoea), as expected with this class of glucose-lowering therapies.60-62 In the SUSTAIN 7 direct comparison, GI AEs were reported in 43% to 44% of subjects receiving 0.5 and 1.0 mg QW semaglutide compared with 33% to 48% for 0.75 and 1.5 mg doses of dulaglutide.22 In AWARD-11, the higher doses of dulaglutide (3.0 and 4.5mh QW) had more GI AEs than the 1.5 mg QW comparator.15 GI AEs can lead to patient drop-out in clinical trials and poor adherence with GLP-1RA therapies in the real world. This can be partially addressed by a slow escalation of dose, smaller portion sizes, advising people to stop eating if they feel nauseated and the avoidance of spicey or fatty meals. The initial concerns regarding an increased risk of thyroid C-cell malignancy and pancreatic pathology (pancreatitis and pancreatic cancer) have not been substantiated for the GLP-1RA class. There is no clinically significant increase in risk of benign or malignant neoplasia with dulaglutide or semaglutide. Rates of non-malignant gall bladder pathology, are low with both drugs.61, 62 An unexpected safety signal from SUSTAIN 6 was an increase in diabetic retinopathy (DR) events.38 DR end-points were reported for more people randomised to semaglutide (50 subjects [3.0%]) than to placebo (29 subjects [1.8%], P = .02). Visbol et al subsequently conducted a detailed evaluation of DR data from the SUSTAIN clinical trial programme.63 No difference in DR AEs was found comparing semaglutide with comparator in the SUSTAIN 1 to 5 (and Japanese regulatory) trials. During these studies, it was noted that patients were excluded from recruitment if they were suffering with DR requiring active medical treatment and HbA1c limits were set at 86 to 91 mmol/mol (10.0%-10.5%), neither of which applied to SUSTAIN 6. They also reported that the impact of semaglutide on DR events could be a result of the speed and degree of HbA1c reduction during the first 16 weeks of the trial.64 Furthermore, the DR AEs were seen in individuals with pre-existing DR, high HbA1c at baseline and who were already receiving insulin. It is of interest that in REWIND, there was a 24% (albeit non-significant) increase in DR events with dulaglutide, the latter in a trial with an upper HbA1c inclusion limitation of 81 mmol/mol (9.5%).36 The EMA have requested an RCT to generate further evidence regarding DR and semaglutide. This trial (named FOCUS) is recruiting people with T2DM and an Early Treatment Diabetic Retinopathy Study (ETDRS) level of 10 to 75; evaluations are