This Week in Endocrinology — Oct 7, 2026

This AudioScholar briefing is generated by artificial intelligence for healthcare professionals and trainees. It is not medical advice.

Welcome to This Week in Endocrinology. This week we're covering 10 notable papers spanning the reshaping of type 2 diabetes therapy, glucose monitoring and automation across both types of diabetes, and the management of prolactinoma around pregnancy. Let's dive in.

We start with type 2 diabetes, where the treatment philosophy itself has been formally rewritten. In Diabetes Care, Davies and colleagues present the 2026 consensus report from the American Diabetes Association and the European Association for the Study of Diabetes, the first update since 2022 [1]. The panel deliberately broadens the frame from controlling hyperglycaemia to holistic management of type 2 diabetes and the multiple long-term conditions that come with it. Lifestyle, sleep and the full twenty-four hour pattern of physical behaviour, psychological support, and weight management are positioned as foundations. The headline pharmacological shift is that the panel advocates earlier use of SGLT2 inhibitors and GLP-1-based therapies, potentially from diagnosis, for organ protection rather than glucose lowering alone, and it supports earlier combination of the two classes in people who also have cardiovascular disease, chronic kidney disease or heart failure. This is consensus guidance built on a systematic review of the literature rather than a new trial, but it carries the weight of the two main professional bodies on both sides of the Atlantic, and its closing message is that the tools to transform outcomes already exist and the gap is consistent, equitable implementation.

The drug pipeline feeding into that framework kept moving. In the New England Journal of Medicine, Wharton and colleagues report the phase 3 SYNCHRONIZE-2 trial of survodutide, a once-weekly glucagon and GLP-1 receptor dual agonist, in 752 adults with type 2 diabetes and a body-mass index of 27 or more [2]. At 76 weeks, weight fell by roughly eight percent on the 3.6 milligram dose and close to ten percent on the 6 milligram dose, against about four percent on placebo, and close to two thirds of participants on the higher dose lost at least five percent of their body weight. Glycated haemoglobin dropped by just under a percentage point on either dose from a baseline of 7.4 percent. Gastrointestinal effects were common, affecting around three quarters of participants on active drug, though mostly mild and transient. These weight losses look modest beside some agents in people without diabetes, a familiar pattern in diabetic populations, and the trial does not address hard outcomes. Meanwhile in Nature Medicine, Guo and colleagues report OUTSTAND-2, a phase 3 trial of safiglipron, an oral small-molecule GLP-1 receptor agonist taken without fasting restrictions, compared head to head with dapagliflozin in 810 metformin-treated adults across 98 sites in China [3]. All three safiglipron doses were non-inferior to dapagliflozin for glycated haemoglobin at 32 weeks, and the 90 milligram dose met the threshold for superiority, though the 60 milligram dose did not. More participants on safiglipron reached a glycated haemoglobin below seven percent, roughly 55 to 64 percent versus about 37 percent on dapagliflozin, while weight loss was broadly similar between the higher doses and the comparator. Gastrointestinal events and discontinuations were more frequent with safiglipron. The trial was confined to a Chinese population, so generalisability remains to be shown.

And a mechanistic trial helps explain why the consensus leans so heavily on GLP-1 therapy for organ protection. Also in Nature Medicine, Tuttle and colleagues randomised 106 people with type 2 diabetes and chronic kidney disease to weekly semaglutide 1 milligram or placebo for a year, combining kidney magnetic resonance imaging with biopsy and transcriptomics in subsets [4]. The coprimary imaging outcomes, kidney oxygenation, perfusion and inflammation, were not significantly changed. Secondary measures, however, showed lower renal artery resistance, stabilisation of a diffusion marker suggesting halted fibrosis progression, and pronounced effects on glomerular endothelial cells with fewer nearby immune cells. Because the primary endpoints were negative and the biopsy subsets were small, these mechanistic signals are hypothesis-generating rather than settled.

Our second theme is technology, and this week it spans insulin automation, digital coaching, and disease staging. In Diabetes Care, Levy and colleagues report the pivotal twiist-T2D trial, in which 307 adults with insulin-treated type 2 diabetes used the twiist automated insulin delivery system for 13 weeks [5]. Glycated haemoglobin fell by about 0.7 percentage points to 7.4 percent, with the largest gains in people starting at nine percent or higher, and time in range rose by about sixteen points to just under three quarters of the day. Importantly, the benefit held when the system was added on top of a GLP-1 receptor agonist or SGLT2 inhibitor. Hypoglycaemia stayed low, there was no severe hypoglycaemia, and one participant had diabetic ketoacidosis. Satisfaction and sleep improved. The caveats are a single-arm design and a short follow-up, so the size of the effect against an active comparator is unknown.

In Diabetologia, Freckmann and colleagues asked whether continuous glucose monitoring alone is enough for people not on insulin [6]. In a six-month open-label trial at 24 German sites, 327 adults with non-insulin-treated type 2 diabetes all used intermittent monitoring and were randomised to the glucura digital therapeutic, which delivers personalised lifestyle guidance, or a control app. Monitoring alone lowered glycated haemoglobin by about 0.3 points, but adding glucura lowered it by about 0.8, an adjusted difference of nearly half a point, and the intervention group roughly tripled the odds of reaching a target below seven percent. Weight loss was more than doubled, at about four percent, and diabetes distress improved. The trial was manufacturer-funded and unmasked, but it suggests that the glucose data may matter less than what patients are helped to do with it.

Monitoring is also reshaping type 1 diabetes before it becomes clinical. In the Journal of Clinical Endocrinology and Metabolism, Steck and colleagues pooled three prospective cohorts totalling 152 people with early-stage type 1 diabetes, of whom about a third progressed to stage 3 [7]. Time spent above 140 milligrams per decilitre on continuous monitoring performed essentially identically to the two-hour oral glucose tolerance test value for predicting progression at two years, and the authors propose time-above-140 thresholds that separate stages 1, 2 and 3. This is the largest such dataset, but it remains observational and modest in size, so the thresholds need external validation before they could replace the glucose tolerance test in staging protocols.

That staging work matters because it is exactly the window in which immunotherapy might act, and this week brought a sobering result. In The Lancet Diabetes and Endocrinology, Rosenzwajg and colleagues report DIABIL-2, a phase 2b trial in which 141 children and adults with newly diagnosed type 1 diabetes received low-dose interleukin-2 on two maintenance schedules or placebo for a year [8]. Regulatory T cells expanded robustly, confirming the drug hit its target, yet there was no difference in preserved C-peptide at twelve months, overall, by regimen, or by pubertal status. The treatment was well tolerated. The authors suggest the inflammatory milieu at diagnosis may overwhelm monotherapy, pointing towards earlier-stage disease or combinations, which is where tools like monitoring-based staging could eventually help select patients.

Our final theme is prolactinoma around pregnancy, with two papers in the Journal of Clinical Endocrinology and Metabolism. Karaca and Kelestimur offer an approach-to-the-patient review covering preconception, pregnancy and postpartum care [9]. They emphasise that physiological pituitary enlargement and rising prolactin complicate monitoring, that symptomatic tumour growth and, rarely, apoplexy are the key risks, and that both bromocriptine and cabergoline are considered safe in gestation, with cabergoline increasingly favoured for efficacy and tolerability. They note that breastfeeding is generally possible, with dopamine agonists withheld during lactation unless clinically needed. Alongside this, Miki and colleagues report a large prospective Japanese series of women with 310 prolactinomas treated with cabergoline doses individualised to keep nadir prolactin at the lowest achievable normal level [10]. Where tumours threatened the optic apparatus, medical debulking before conception restored at least five millimetres of clearance, and cabergoline was stopped at conception. Across 365 pregnancies there was no symptomatic tumour enlargement, and just under three quarters of women were in remission postpartum. Required doses varied widely between subgroups, with bromocriptine-resistant tumours needing the most, and nadir prolactin was the strongest predictor of tumour disappearance. No valvulopathy or impulse control disorders were seen. This is a single-programme cohort without a comparison arm, but it offers a concrete, outcome-directed framework that complements the review's more individualised guidance.

If you only have time for one paper this week, make it the ADA and EASD consensus report in Diabetes Care [1]. It formally moves organ-protective drug classes from second-line add-ons towards first-line consideration and recasts type 2 diabetes care around multimorbidity, reframing the question from how to lower glucose to how to change long-term outcomes.

Here is what this week's evidence adds up to in Endocrinology. First, the major diabetes societies now advocate earlier SGLT2 inhibitor and GLP-1-based therapy, potentially from diagnosis, as consensus guidance rather than new trial evidence. Second, newer incretin agents, the injectable dual agonist survodutide and the oral small molecule safiglipron, show solid phase 3 efficacy, though hard outcome data and broader populations are still lacking. Third, automated insulin delivery and digitally supported monitoring both lowered glucose meaningfully in type 2 diabetes, but in a single-arm and a manufacturer-funded open-label trial respectively. Fourth, continuous monitoring looks comparable to the glucose tolerance test for staging early type 1 diabetes, while low-dose interleukin-2 failed to preserve beta-cell function at diagnosis. And finally, observational data suggest that prolactin-guided cabergoline with preconception debulking can make pregnancy safe in prolactinoma, pending confirmation outside a single expert programme.

That's your roundup for This Week in Endocrinology. The full transcript and references are available on the episode page. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And if you have a paper you have been meaning to read, upload the PDF, or paste any link, at audioscholar dot C C. We will turn it into audio like this one, in any of thirty-one languages.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.