GLP-1 safety in neuroendocrine tumours

A review published on 1 August 2026 in the European Journal of Endocrinology sets out what is actually known about giving GLP-1 receptor agonists, the class that includes semaglutide, to people who have a neuroendocrine neoplasm. The authors, an endocrinology group at Erasmus MC and Erasmus MC Cancer Center in Rotterdam, conclude that the available human data does not support a generalised increase in these tumours with GLP-1 receptor agonist use, and that the gaps in the evidence are still wide enough to warrant caution and monitoring^[1].

Neuroendocrine neoplasms are tumours that arise from hormone-producing cells scattered through the gut, pancreas, lungs and endocrine glands. They are uncommon, they behave very differently from one subtype to the next, and anyone who has one is already under specialist care. This article is background for a conversation with that team. It is not a basis for a decision on your own.

What the boxed warning actually says

The US prescribing information for Ozempic carries a boxed warning about thyroid C-cell tumours, and the exact wording matters, because it is more careful than the way it usually gets repeated. The label states that in rodents, semaglutide causes dose-dependent and treatment-duration-dependent thyroid C-cell tumors at clinically relevant exposures. It then states that it is unknown whether Ozempic causes thyroid C-cell tumors, including medullary thyroid carcinoma (MTC), in humans, because the human relevance of the rodent finding has not been determined^[2].

The contraindication that follows is narrow and specific: a personal or family history of medullary thyroid carcinoma, or multiple endocrine neoplasia syndrome type 2 (MEN 2)^[2]. That is the whole of the tumour-related contraindication on the label. Pancreatic neuroendocrine tumours, small intestinal neuroendocrine tumours, neuroendocrine neoplasms as a general category and MEN type 1 are not named on it. Anything beyond MTC and MEN 2 is inference from biology, not a labelled restriction, and the two should not be reported as if they were the same thing.

Europe manages the same uncertainty differently. The European Medicines Agency assessment page for Ozempic carries no boxed warning of the US kind, and lists a post-authorisation database linkage study to evaluate the risk of medullary thyroid carcinoma among its ongoing safety measures^[3]. Two regulators, the same open question, two different ways of holding it open.

Why neuroendocrine tumours raise the question at all

The concern is not arbitrary. GLP-1 receptors are present in healthy tissue including the neurohypophysis, duodenal glands and pancreatic islets, and they are also expressed on neuroendocrine tumours themselves^[1].

Strongly enough, in fact, that the receptor is used to find the tumours. Insulinomas, a pancreatic neuroendocrine tumour, are located with PET tracers built from exendin-4, a GLP-1 analogue. A 2026 study in the Journal of Clinical Endocrinology and Metabolism reported that in 58 patients whose conventional imaging had been negative or inconclusive, GLP-1 receptor targeted PET/CT found the tumour in 72% of cases, with a sensitivity of 93.8% among those who went on to histological confirmation^[4]. A receptor you can image that reliably is a receptor that is abundantly there.

The preclinical layer sits on top of that. The review notes that GLP-1 receptor agonists have shown proliferative effects in cell lines representing pancreatic and small intestinal neuroendocrine tumours^[1]. Cell lines are not patients, and a receptor being present does not establish that stimulating it drives a tumour in a person. But that chain, receptor present, receptor stimulated, growth signal in a dish, is the reasoning behind the caution.

What the human evidence shows, and what it misses

Here the picture loosens. The review found that retrospective studies, randomised trials and cancer registries give conflicting results, and that together they do not support a generalised increase in neuroendocrine neoplasms with GLP-1 receptor agonist use. Some epidemiological work has even reported survival benefits associated with exposure to the class^[1].

The largest synthesis of randomised evidence points the same way without being about neuroendocrine tumours. A systematic review and meta-analysis in Annals of Internal Medicine pooled 48 trials and 94,245 participants and found that GLP-1 receptor agonists probably have little or no effect on thyroid cancer risk (odds ratio 1.37, 95% CI 0.82 to 2.31, moderate certainty)^[5].

Read the limitation alongside the result. The trials in that analysis were not designed to evaluate cancer outcomes and had short follow-up^[5]. It reports thyroid cancer as a whole rather than medullary thyroid carcinoma specifically, and it does not report neuroendocrine neoplasms as a category at all. Reassurance about one endpoint is not reassurance about a rarer, different one, and treating it that way is how a careful finding turns into a false comfort.

What we don't yet know

The review is direct about the holes. There is no long-term prospective study with neuroendocrine neoplasms as the clinical endpoint. Several tumour subtypes have no data at all. And the position of patients with multiple endocrine neoplasia type 1, a syndrome that predisposes to pancreatic and other neuroendocrine tumours, is unresolved. MEN 1 is not a labelled contraindication, and it is not settled in the evidence either^[1].

The authors' own conclusion is worth taking exactly as written. Starting a GLP-1 receptor agonist in a patient predisposed to neuroendocrine neoplasia, where a contraindication does not exist, should be approached cautiously with careful monitoring for long-term adverse effects^[1]. That is a call for care. It is neither a prohibition nor an all-clear, and it should not be reported as either.

Where this leaves a reader

This is a review, not new trial data. It gathers what is already published and marks where the map runs out. Nothing in it changes a label, and nothing in it settles the question it examines.

If you have a neuroendocrine neoplasm, or a genetic syndrome that predisposes you to one, the variables that decide this question sit with your treating team: your tumour subtype, your genetics, your other treatments, and the reason a GLP-1 receptor agonist is being considered at all. Bring the paper to them. The country-by-country regulatory position is on our semaglutide regulation notes, and the class picture, including tirzepatide, sits alongside it. Both are preparation for that conversation, not a replacement for it. Nothing here is a reason to start a medicine, stop one, or keep taking one.