MetaResearch

A peptide raised the share reaching a cell collection target to 92.5 percent

Wed, 09 Sep 2026 00:00:00 +0000

In 122 adults with a bone marrow cancer preparing for a transplant of their own stem cells, adding the peptide to the standard injection raised the share who collected enough cells in two sessions from 26.2 percent to 92.5 percent. A later comparison against the drug already in use found no advantage.

What was studied: a ring shaped synthetic peptide added to the standard growth factor injection before a person's own blood stem cells are collected for transplant, tested against an inactive substitute in a randomised double blind trial of 122 adults with a bone marrow cancer [1].

What it found: 92.5 percent of the peptide group collected the target number of cells within two collection sessions against 26.2 percent on the inactive substitute, and 88.8 percent reached it in a single session against 9.5 percent [1].

What it does not show: that anyone lived longer or had a better transplant, since the measurement was cells collected rather than what happened afterwards; nor that it beats the drug already used for the same job, since the trial did not include that drug and the first published comparison, in 60 patients at one hospital, found no advantage and more reactions needing treatment [1][2].

A randomised trial published in Nature Medicine in 2023 reported one of the wider gaps in recent haematology: 92.5 percent of one group met the trial's main measurement against 26.2 percent of the other. Three years later the first published comparison against the treatment already in routine use reported no advantage at all. Both are about the same peptide, and the distance between them is a lesson about what a large effect against a placebo does and does not establish [1][2].

What the procedure is

People with multiple myeloma, a cancer of the plasma cells in bone marrow, are often treated with a transplant of their own blood stem cells. The cells are collected first, then given back after high strength chemotherapy has cleared the marrow. To collect them, the cells have to be coaxed out of the bone marrow into the bloodstream, which is done with a growth factor injection over several days. The blood is then passed through a machine that separates out the stem cells, a procedure called apheresis, and returns everything else [1].

The procedure has a threshold. Transplant teams aim for around six million of a particular kind of stem cell for every kilogram of the patient's body weight. Many patients do not reach it with the growth factor alone and have to return for further sessions, each of which takes hours [1].

The peptide and the trial

Motixafortide is a cyclic peptide, a short chain of amino acids joined into a ring, that blocks a receptor called CXCR4. That receptor is part of what tethers stem cells inside the bone marrow, so blocking it releases them into the blood. The GENESIS trial enrolled 122 adults at 18 sites in five countries and assigned them by chance, two to one, to the peptide plus the growth factor or to an inactive substitute plus the growth factor. Neither patients nor staff knew which was which [1].

The main measurement was the share of patients who collected the target number of cells within two apheresis sessions. In the peptide group 92.5 percent did, against 26.2 percent in the placebo group, with a P value below 0.0001. The secondary measurement, reaching the target in a single session, was met by 88.8 percent against 9.5 percent [1].

Side effects were mostly local and short lived: pain at the injection site in 50 percent, redness in 27.5 percent and itching in 21.3 percent, all reported as mild or moderate. The peptide was approved in the United States on 11 September 2023, for use alongside the growth factor in this exact situation. It is developed by BioLineRx under licence from Biokine Therapeutics [1][3].

What the trial measured, and what it did not

The measurement is the number of cells collected. That is a real and useful thing to measure, because a patient who reaches the target in one session avoids returning for more, and because failing to collect enough can delay or prevent a transplant altogether. It is not a measurement of how the transplant went, how long the cancer stayed away, or how long anyone lived. The trial did not follow patients for those things, and no figure in it speaks to them [1].

The comparison also deserves attention. The trial compared the peptide plus growth factor against growth factor alone. It did not compare it against plerixafor, the drug already used for this job when the growth factor is not enough. Against an inactive substitute, the gap was enormous. Against the existing option, the trial says nothing, because the existing option was not in it [1].

The first head to head look

That gap in the record was partly filled in 2026, by a report in Transfusion from a single centre that compared 30 patients given the peptide with 30 matched patients given the older drug, one group followed forward and one drawn from records. The stem cell yields were similar. The time each session took was similar. Twenty seven patients in each group reached the target in one day [2].

What differed was tolerability and workload. Reactions needing treatment were substantially more common with the peptide, both at the injection site and throughout the body, and it required medicine given beforehand and monitoring afterwards. The centre concluded that the peptide was workable but not superior, kept the older drug as its default, and reserved the peptide for selected patients [2].

Sixty patients at one hospital, with one group taken from past records, is a weaker design than a randomised trial and cannot overturn one. It is also the only direct comparison published so far, and it points the opposite way from what the size of the placebo controlled result would lead a reader to expect.

Why both belong in the summary

A very large effect against a placebo establishes that the drug does what it was designed to do. It says nothing about where the drug sits among the options a clinic already has, and nothing about whether the patients ended up better off. Those are separate questions, and each one needs a trial built to answer it [1][2].

What would settle it

A randomised comparison against the existing drug, in enough patients to detect a difference, reporting what happened after the transplant rather than only what came out of the collection machine. Until that exists, the defensible reading of this record is that the peptide reliably releases stem cells into the blood, and that whether it is the better choice for a given patient remains open [1][2][3].

Sources

A self injected peptide beat placebo on a daily tasks score in 174 adults

Wed, 09 Sep 2026 00:00:00 +0000

In a 12 week randomised trial in a disease that weakens muscles, the peptide group improved 2.09 points more than the placebo group on an eight item scale of everyday tasks. Approvals in three regions followed, and no trial has yet compared it against the other new treatments.

What was studied: a ring shaped synthetic peptide that blocks one step of the immune system's complement cascade, given as a daily injection under the skin for 12 weeks to 174 adults with a disease that weakens muscles, half of them assigned by chance to an inactive substitute [1].

What it found: scores on an eight item scale of everyday tasks fell 4.39 points in the peptide group against 2.30 points on placebo, a difference of 2.09 points with a P value of 0.0004, and an open continuation study reported the improvement holding at 60 weeks [1][2].

What it does not show: how the peptide compares with the other treatments approved for the same disease, since no trial has put them head to head; nor how much of the 60 week picture is the treatment, since everyone in the continuation knew what they were taking and only people who had finished the first trial were in it [1][2].

A 12 week randomised trial published in Lancet Neurology in 2023 reported that a synthetic peptide improved how well people with generalised myasthenia gravis managed ordinary daily tasks. The result cleared its threshold comfortably, regulators in three regions accepted it, and the honest description of what it establishes is still narrower than the approvals suggest [1][3].

The disease and the target

Generalised myasthenia gravis is an autoimmune disease in which the body makes antibodies against the receptors that carry the signal from nerve to muscle. Muscles tire quickly and unpredictably: eyelids droop, chewing and swallowing become hard work, breathing can be affected. Most patients carry antibodies against the acetylcholine receptor, and those are the patients this trial enrolled [1].

The drug, zilucoplan, is a macrocyclic peptide, meaning a short chain of amino acids joined into a ring so that it holds its shape and survives longer in the body. It blocks complement component 5, one step in a chain of proteins the immune system uses to attack cells it has marked. Blocking that step is meant to stop the damage at the nerve to muscle junction. It is given as an injection under the skin that a patient does at home each day [1][3].

How the trial was run

RAISE was randomised, double blind and placebo controlled, run at 75 sites across Europe, Japan and North America. Of 239 people screened, 174 were enrolled and assigned by chance: 86 to the peptide and 88 to a matching inactive injection, for 12 weeks. Entry required a level of disability on two standard scales, so nobody in the trial had mild disease only [1].

The main measurement was the change at week 12 in the Myasthenia Gravis Activities of Daily Living score, an eight item questionnaire on things like talking, chewing, breathing and lifting the head, where a higher score means more difficulty. It is a patient reported measure, which is a strength, since it records what the person actually experiences, and a weakness, since it depends on the person not knowing which group they are in [1].

What it found

Scores fell by an average of 4.39 points in the peptide group and 2.30 points on placebo. The difference was 2.09 points, with a range of statistical uncertainty from 0.95 to 3.24 and a P value of 0.0004. The placebo group improving by more than two points on its own is worth noticing: it is a reminder of how much movement a trial of this kind records before any drug is counted [1].

Side effects occurred in 77 percent of the peptide group and 70 percent of the placebo group, most commonly bruising at the injection site, in 16 percent against 9 percent. Serious side effects and serious infections were similar in the two groups. One person died in each group; neither death was judged related to the study drug [1].

What happened after 12 weeks

People who finished the trial could enter an open continuation study called RAISE-XT, in which everyone received the peptide and everyone knew it. An interim analysis of 200 patients, published in 2024, reported that improvement continued to week 24 and held to week 60, and that people who had been on placebo improved within a week of switching [2].

An open continuation is the usual way to gather long term safety information and it cannot function as evidence of benefit. There is no comparison group, everyone knows what they are receiving, and only people who did well enough to continue are in it. In this one, 94 percent of patients reported some side effect, the most common being a worsening of the disease itself, in 26 percent [2].

A subgroup worth reading carefully

A prespecified analysis published in 2025 split the trial by whether patients had previously received immunoglobulin treatment or plasma exchange. The 54 who had not were less severely affected and had been diagnosed more recently. Their scores fell 4.22 points against 2.61 on placebo, against 4.93 against 2.94 in the group with previous treatment. The authors read this as support for using the drug earlier [4].

It is a descriptive analysis of subgroups within a trial that was sized for its whole population, so it generates a question rather than answering one. Fifty four people split across two groups is a small number to carry a conclusion about when to start a treatment.

What is missing from the record

Several new treatments for this disease were approved within a few years of each other, working through different mechanisms. None has been compared against another in a randomised trial. Every comparison currently available is indirect, drawn from separate trials with different patients and different entry criteria, which is the weakest form of evidence about which option suits whom [1][3].

The trial was funded by UCB Pharma, which develops the drug, and many of its investigators report consulting relationships with that company and with the makers of competing treatments. The paper states all of it. Approvals followed in Japan in September 2023, the United States in October 2023 and the European Union in December 2023 [1][3].

What would settle it

Head to head randomised trials against the other approved options, follow up long enough to describe what happens over years rather than one, and a blinded measure of daily function that does not depend on a person guessing their group. A 2.09 point difference over 12 weeks in 174 people is a real finding, reported carefully. It is a starting point in a record that is still thin [1][2][4].

Sources

An antibody carrying chemotherapy added five months in a 468 person trial

Wed, 09 Sep 2026 00:00:00 +0000

In advanced breast cancer of one type, median survival was 12.1 months against 6.7 months on the chemotherapy a doctor would otherwise have picked. In a later randomised trial in a different cancer, the same treatment did not extend life at all.

What was studied: a treatment built by attaching a chemotherapy drug to an antibody that finds a protein carried on the surface of many tumour cells, compared against a single chemotherapy drug chosen by the treating doctor, in a randomised trial of 468 adults with advanced breast cancer of one type [1].

What it found: median time before the cancer grew again was 5.6 months against 1.7 months, and median survival was 12.1 months against 6.7 months; a second randomised trial in 543 adults with a different breast cancer type found a smaller survival gap, 14.4 months against 11.2 months [1][2].

What it does not show: a cure, or a benefit wherever the same surface protein appears; a third randomised trial in 711 adults with advanced bladder and urinary tract cancer missed its main measurement, and severe side effects were more common on the antibody treatment in all three [1][2][3].

One of the clearest survival results in advanced breast cancer this decade came from a randomised trial of 468 people, published in the New England Journal of Medicine in 2021. The same treatment, tested later in a different cancer, failed. Both results belong in the same account, and reading them together is what the published record actually supports [1][3].

What the treatment is

Sacituzumab govitecan is an antibody drug conjugate. An antibody is a protein the immune system uses to recognise one specific target; here it recognises Trop-2, a protein carried on the surface of most breast cancer cells. Chemically bonded to that antibody is SN-38, a chemotherapy drug that damages an enzyme cancer cells need in order to copy their DNA. The antibody is the address and the chemotherapy is the parcel [1].

The idea is to deliver a drug that would be too harsh to give freely by attaching it to something that finds the tumour first. Whether the delivery is precise enough to matter is the question every trial of this design has to answer.

The first trial

The ASCENT trial enrolled adults whose breast cancer had returned or stopped responding, in the form called triple negative, meaning the tumour carries none of the three receptors that most breast cancer treatments are aimed at. All had already received taxane chemotherapy. Participants were assigned by chance to the antibody treatment or to a single chemotherapy drug of the treating doctor's choosing, which is what these patients would otherwise have been given [1].

Among the 468 participants without cancer that had spread to the brain, median time before the disease grew again was 5.6 months on the antibody treatment against 1.7 months on chemotherapy. Median survival was 12.1 months against 6.7 months. Tumours shrank measurably in 35 percent of the antibody group against 5 percent of the chemotherapy group. All three comparisons had a P value below 0.001, meaning a gap this large would turn up by chance less than once in a thousand times if the treatments were equal [1].

Severe side effects were more common. Low white cell counts of grade 3 or worse occurred in 51 percent of the antibody group against 33 percent on chemotherapy, and severe diarrhoea in 10 percent against under 1 percent. Three people in each group died of side effects; none of those deaths were attributed to the antibody treatment [1].

The second trial, and a smaller gap

A second randomised trial, TROPiCS-02, tested the same treatment in 543 adults with a more common form of breast cancer, one driven by hormone receptors and not by the HER2 protein, after several previous treatments had been used. Its final survival analysis, published in The Lancet in 2023, reported median survival of 14.4 months against 11.2 months, a difference of 3.2 months with a P value of 0.020 [2].

That trial was open label, meaning everyone knew which treatment they were receiving. Its measures of quality of life and fatigue also moved in favour of the antibody treatment. One person died of septic shock caused by an infection during a period of low white cell counts, and that death was judged to be related to the treatment [2].

The trial that missed

TROPiCS-04, published in Annals of Oncology in 2025, took the same treatment into advanced urothelial cancer, which arises in the bladder and urinary tract, in 711 people whose disease had already progressed through platinum chemotherapy and immunotherapy. The main measurement was survival, and the trial did not meet it: median survival was 10.3 months against 9.0 months, with a P value of 0.087 [3].

Tumours did shrink more often, in 23 percent against 14 percent, so the drug was doing something. What went the other way was safety. Severe side effects occurred in 67 percent of the antibody group against 35 percent on chemotherapy, and fatal side effects in 7 percent against 2 percent. Sixteen of the 25 deaths from side effects in the antibody group were infections during periods of low white cell counts, most of them early in treatment in people who already carried several risk factors for exactly that. The authors write that these early complications may have affected the result [3].

What the three together support

That a surface protein is present on a tumour is not by itself a reason to expect benefit. The same antibody, the same chemotherapy parcel and the same target produced a large survival gap in one cancer, a smaller one in another, and none in a third. Where a treatment sits in the sequence of therapies, how sick the patients already are and how well the side effects can be managed all moved the answer [1][2][3].

The first trial was funded by Immunomedics, the company that developed the drug. The later breast cancer trial was funded by Gilead Sciences. Both papers state this, which is how it should be, and it is the reason the failed trial belongs in any honest summary alongside the two that succeeded [1][2].

What would settle it

Trials that test the treatment earlier in the course of disease, comparisons against the newer options rather than against older chemotherapy, and reporting that separates what the drug does from what happens when a person is already too frail to absorb its side effects. The five month survival gap in the first trial is real and was measured carefully. It describes one group of patients, in one cancer, at one point in their treatment [1][3].

Sources

An antiviral shortened recovery in one trial and did not cut deaths in another

Wed, 09 Sep 2026 00:00:00 +0000

A blinded trial of 1,062 hospital patients found people recovered a median of five days sooner. An open trial that eventually randomised 8,275 found a difference in deaths too small to clear the usual threshold, except in one group.

What was studied: the same antiviral, given by drip to adults in hospital with COVID-19, across three published reports: a double blind placebo controlled trial of 1,062 patients and a large open trial that randomised 8,275 to the drug or to no trial drug [1][2][3].

What it found: in the blinded trial, median time to recovery was 10 days against 15 days; in the open trial, in-hospital deaths were 14.5 percent against 15.6 percent overall, which did not clear the usual threshold, and 11.9 percent against 13.5 percent among people not on a ventilator, which did [1][3].

What it does not show: that either reading cancels the other, since the two trials measured different things, one was blinded and one was not, and the care given alongside the drug changed between them; nor what the drug does against the virus circulating now [1][2][3].

Two of the largest trials run during the first year of the pandemic tested the same antiviral in the same kind of patient and were reported, in much of the coverage, as contradicting each other. They did not. They asked different questions, and the difference between the questions is most of the story [1][2].

The drug

Remdesivir is an antiviral developed by Gilead Sciences. It imitates one of the building blocks a virus uses to copy its genetic material, so that copying stalls when the imitation is used. It had shown activity against coronaviruses in laboratory work before 2020, which is why it was among the first candidates tested when the pandemic began. Emergency authorisation in the United States came in May 2020, followed by conditional approvals elsewhere [4].

The first trial and its measurement

ACTT-1 was a double blind randomised trial in adults in hospital with COVID-19 and signs of infection in the lower airways. Double blind means neither the patient nor the treating team knew who was receiving the drug and who was receiving an inactive drip. A total of 1,062 people were randomised, 541 to the antiviral and 521 to placebo [1].

Its main measurement was time to recovery, defined as leaving hospital or staying only for infection control reasons. Median time to recovery was 10 days on the antiviral against 15 days on placebo, with a P value below 0.001. Deaths by day 29 were 11.4 percent against 15.2 percent, a difference the trial was not sized to test on its own [1].

The second trial and its measurement

The World Health Organization's Solidarity trial asked a blunter question: does anyone die less often. It enrolled patients at hundreds of hospitals across dozens of countries and randomised them among whichever of four repurposed drugs were available locally, or to no trial drug. There were no placebos, so everyone knew what they were receiving [2].

The interim report, published in the New England Journal of Medicine in 2021, covered 11,330 adults. Among those randomised to the antiviral, 301 of 2,743 died against 303 of 2,708 in the matched control group, a rate ratio of 0.95 with a P value of 0.50. The report concluded that the four drugs it tested had little or no effect on deaths in hospital [2].

The final report, published in The Lancet in 2022, had 8,275 people randomised to the antiviral or its control. Overall, 14.5 percent of the antiviral group died against 15.6 percent of controls, a rate ratio of 0.91 with a P value of 0.12, which does not clear the usual threshold. Among those not already on a ventilator the figures were 11.9 percent against 13.5 percent, a rate ratio of 0.86 with a P value of 0.02, which does. Among those already ventilated there was no benefit [3].

Why the two readings sit together

Time to recovery and death are not the same measurement, and a drug can move one without moving the other enough to detect. A trial of 1,062 people can measure a five day difference in recovery and be far too small to settle a question about mortality. A trial of 8,275 can settle the mortality question and never record how quickly people got better [1][3].

Blinding differs too. In an open trial, knowing who received the drug can shape when a doctor discharges someone or moves them onto a ventilator. That is one reason the larger trial chose deaths as its measure: it is the outcome least open to being nudged [2][3].

The timing matters as well. The blinded trial ran first. By the time the larger one finished, standard care in hospital had changed, most importantly with the routine use of a cheap steroid in patients needing oxygen. A drug added to better background care has less room to show a difference [3].

What the record supports

The final Solidarity report also combined its own results with every other randomised trial of the same drug in hospital patients. Read that way, the honest summary is narrow: a modest reduction in deaths among patients who are in hospital but not yet ventilated, no benefit once a person is on a ventilator, and a faster recovery observed in the one large blinded trial that measured it [1][3].

That is a smaller claim than the early coverage made and a larger one than the later coverage allowed. Neither trial was wrong. They were built to answer different questions, and the answers only look contradictory if the questions are collapsed into one.

What this does not settle

None of these trials describes what the drug does against the versions of the virus circulating now, in a population with widespread immunity from vaccination and previous infection, or in people treated outside hospital. Every figure above comes from patients sick enough to be admitted, in 2020 and 2021, under the standard care of that time [1][2][3].

The general lesson outlives the drug. When two large trials of the same treatment appear to disagree, the first thing to check is not which one to believe. It is what each one measured, who was blinded, and what else was being given at the time [1][2][3].

Sources

No new HIV infections among 2,134 given a twice yearly injection

Wed, 09 Sep 2026 00:00:00 +0000

Two late stage randomised trials tested an injection given under the skin every 26 weeks against daily tablets. In the first, none of the 2,134 people in the injection group acquired the virus. In the second, two did.

What was studied: two late stage randomised trials of an injection given under the skin every 26 weeks to prevent HIV, one in 5,338 adolescent girls and young women in South Africa and Uganda, the other in 3,265 men, transgender people and gender diverse people at sites in North America, South America, Africa and Asia [1][2].

What it found: none of the 2,134 people in the injection group of the first trial acquired HIV, against 39 infections among 2,136 taking one daily tablet combination and 16 among 1,068 taking another; in the second trial there were 2 infections in the injection group against 9 in the daily tablet group [1][2].

What it does not show: how well the injection works against a group given nothing, since neither trial had one and the main comparison was an estimate of how much infection the same population would have had anyway; nor whether people keep to it outside a trial, where a report on 501 early recipients found starting delays and some stopping [1][2][3].

Two randomised trials published in the New England Journal of Medicine, one in 2024 and one in 2025, tested the same idea: an injection given under the skin twice a year to keep people from acquiring HIV. The first reported no infections at all in the injection group. The second reported two. Numbers that low are rare enough in prevention research that the design underneath them is worth reading carefully [1][2].

What the two trials tested

The drug is lenacapavir, given as an injection under the skin every 26 weeks. Both trials compared it against tablets that people already take daily for the same purpose, an approach called preexposure prophylaxis, meaning medicine taken before any contact with the virus rather than after [1][2].

The first trial, PURPOSE 1, enrolled adolescent girls and young women in South Africa and Uganda and assigned them by chance, in a two to two to one split, to the injection, to one daily tablet combination, or to a second daily tablet combination used as the active comparison. Everyone also received a matching inactive injection or tablet so that nobody knew which group they were in [1].

The second trial, PURPOSE 2, ran the same comparison in cisgender men, transgender women, transgender men and gender nonbinary people, at sites across North America, South America, Africa and Asia, assigning two people to the injection for every one assigned to a daily tablet [2].

What they found

In the first trial, among 5,338 participants who were HIV negative at the start, 55 people acquired the virus during the trial. None of them were in the injection group, which held 2,134 people. Thirty nine infections occurred among the 2,136 people on one tablet combination and 16 among the 1,068 on the other [1].

In the second trial, among 3,265 people in the main analysis, 2 infections occurred in the injection group and 9 in the daily tablet group. Expressed as a rate, that is 0.10 infections per 100 person years against 0.93. A person year is one person followed for one year, so a rate per 100 person years is the count expected if 100 people were watched for a year each [2].

The comparison that carries the result

Neither trial included a group given nothing. Withholding a prevention method that already works would not be acceptable, so the trials did something else: they estimated how much infection would have occurred in the same population without any prevention, by measuring the rate among everyone screened for entry. That estimate was 2.41 infections per 100 person years in the first trial and 2.37 in the second [1][2].

The headline efficacy figures are the injection rate set against that estimate. This is a reasonable way to run such a trial and it is also the softest joint in the argument, because the screened population is not the enrolled population and the estimate is built rather than observed. Both papers report the comparison against the daily tablets as well, which is the harder test, and the injection was lower on that comparison too [1][2].

What the trials do not settle

The first trial states plainly that adherence to both tablet combinations was low. That matters for how the result should be read. Part of what separated the groups is chemistry and part is the difference between an injection given twice a year at a clinic and a tablet somebody has to remember every morning. The trial cannot tell those two apart, and it was not built to [1].

Side effects were not absent. In the first trial, reactions at the injection site occurred in 68.8 percent of the injection group against 34.9 percent of those receiving the inactive injection, and 4 people, 0.2 percent, stopped the trial regimen because of them. In the second, 26 of 2,183 people in the injection group, 1.2 percent, stopped for the same reason, against 3 of 1,088 on tablets [1][2].

Both trials also ran in populations chosen for high background risk. That is what makes a prevention trial readable in a reasonable time, and it is also why the rates in either group cannot be carried over to a different population.

Between a trial and a clinic

A separate report, published in Clinical Infectious Diseases in 2026, followed the first 501 people to receive the injection outside a trial, across 17 sites in the United States. Starting it was often delayed by insurance processes. Early stopping was uncommon at 3.6 percent, but that is higher than the rate at which people left the trials, and it was linked to injections in the abdomen and to reactions that needed a clinical visit [3].

None of that contradicts the trials. It describes the gap that always exists between a result and its use, and it is the part of the record that a striking efficacy figure tends to crowd out.

Who ran them

Both trials were funded by Gilead Sciences, the company that makes the drug, and the author lists of both include company employees alongside academic investigators. That is stated in the papers themselves. It does not make a result wrong, and a trial that says who paid for it is behaving as it should; it sets how much independent replication a finding needs before it is treated as settled [1][2].

What would settle it

Longer follow up in populations that were not selected for high background risk, records of what happens when the injection is used in ordinary services rather than trial sites, and evidence on whether people return every 26 weeks over years rather than months. Two trials with almost no infections in one group is an unusually clean result. It is not the same as knowing what the same medicine does in a clinic that no trial is watching [1][2][3].

Sources

A swallowed peptide cut one kind of cholesterol 57.1 percent in a trial

Tue, 08 Sep 2026 00:00:00 +0000

The chain was joined into a ring so it could survive the gut, and 2,909 adults took it or an inactive tablet daily for 52 weeks. Whether it prevents heart attacks is a separate trial that does not report before late 2029.

What was studied: a ring shaped peptide, built to survive digestion and taken as a daily tablet, against an inactive tablet in 2,909 adults with raised cholesterol over 52 weeks [1].

What it found: by week 24 the treated group's low density lipoprotein cholesterol had fallen by an average of 57.1 percent while the comparison group's rose by 3.0 percent, and the report states that side effect rates did not appear to differ [1].

What it does not show: whether the medicine prevents heart attacks or strokes, which is the subject of a separate trial of roughly 14,550 people whose main measurement is not expected until November 2029 [3][4].

A peptide is a short chain of amino acids, and the reason peptide medicines are injected is that the gut is built to pull chains like that apart. Enlicitide is a peptide that is swallowed. Its chain is joined end to end into a ring, and the ring, along with a long list of chemical changes along it, is what lets enough of the molecule survive to be absorbed. In 2026 it became an approved tablet for lowering cholesterol, which makes it a useful place to look at what engineering a peptide into a pill settles and what it leaves open [1][3].

What the molecule blocks

Its target is PCSK9, a protein circulating in blood that shortens the working life of the receptors the liver uses to pull low density lipoprotein cholesterol out of circulation. Less of that protein at work means those receptors last longer and more cholesterol is cleared. Injected antibodies have been blocking it for years. Blocking it with something swallowed is the new part [1].

The trial

CORALreef Lipids was a multinational trial in which neither participants nor investigators knew who was receiving what. It enrolled adults who had already had a major cardiovascular event and whose low density lipoprotein cholesterol was 55 milligrams per decilitre or higher, and adults at risk of a first event whose level was 70 or higher. Of the 2,909 participants in the main analysis, 1,935 received the tablet and 969 received an inactive tablet, daily for 52 weeks. Average age was 63 years, 39.3 percent were women, and average low density lipoprotein cholesterol at the start was 96.1 milligrams per decilitre [1].

The main measurement was the percent change in that cholesterol at week 24. It fell by an average of 57.1 percent in the treated group, with a range of statistical uncertainty from 61.8 to 52.5 percent, and rose by 3.0 percent in the comparison group. The difference between the groups was 55.8 percentage points, with a range from 60.9 to 50.7, at a P value below 0.001. Non HDL cholesterol, apolipoprotein B and lipoprotein(a) also fell further than in the comparison group. The report states that the rate of side effects did not appear to differ [1].

The part that is new chemistry

A separate 2026 paper in Science is about how the molecule is made rather than what it does, and it is where the genuinely new science sits. Assembling a heavily modified ring shaped peptide at scale has historically taken a long chain of steps, many of them spent putting temporary chemical shields on parts of the molecule and taking them off again. The team reported building it with engineered enzymes that join the fragments and close the ring without those shields, and with crystallisations that remove a purification stage. They report cutting the number of steps by more than half against the previous best method [2].

Historically, many compelling therapeutic targets have been accessible only by injectable biologic drugs.
Klapars and colleagues, Science, 2026 [2]

That sentence is the reason the manufacturing paper matters beyond this one molecule. If rings of this kind can be made in half the steps and without shields, the argument that a peptide has to be an injection weakens for reasons that are industrial as much as biological [2].

What the label reveals about the gut

The approved United States labelling shows how partial the win over digestion is. The tablet is to be taken on an empty stomach in the morning with water, black coffee or plain tea, swallowed whole, and at least 30 minutes have to pass before any food or other drink. Those instructions exist because the molecule is still in a race with digestion. The ring wins it enough time, not unlimited time [3].

The indication section is careful in a way worth reading closely, because it separates what this medicine has shown from what other medicines have shown [3].

Cardiovascular outcomes trials have demonstrated that reducing LDL-C lowers the risk for major adverse cardiovascular events (MACE) in adults at increased risk when treated with statins or monoclonal antibody proprotein convertase subtilisin kexin type 9 (PCSK9) inhibitors as an add-on to statin therapy.
United States prescribing information for enlicitide, Indications and Usage [3]

The sentence points at trials of other medicines. It says that lowering this kind of cholesterol has been shown to reduce major cardiovascular events when the lowering was done with statins or with the injected antibodies. It does not say that this tablet has been shown to do it [3].

What it does not show

That is the open question. A 57.1 percent reduction in a laboratory value is a marker, and the thing a person actually wants to know is whether taking the medicine means fewer heart attacks and strokes. The trial built to answer it is registered as NCT06008756, with roughly 14,550 participants at high cardiovascular risk. Its main measurement is the time to a first major cardiovascular event, and the registry entry gives November 2029 as the estimated date for collecting it [4].

The published record also says nothing about how this tablet compares with the injected antibodies over years, or about anyone outside the trial population. What it does establish is narrower and still a first of its kind: a peptide, closed into a ring and manufactured by enzymes, lowered a cholesterol measurement over 52 weeks when swallowed once a day, and there is now an approved label written on the assumption that people will take it that way [1][3].

Sources

A weight loss pill cut average body weight 11.2 percent in a 72 week trial

Tue, 08 Sep 2026 00:00:00 +0000

The tablet is put together by ordinary chemistry instead of being built as a peptide chain, and two late stage trials tested it against an inactive tablet in 3,127 and 1,613 adults. No trial has put it against the injections.

What was studied: a once daily tablet built to switch on the same docking point as the injected weight loss medicines without being a peptide, in two late stage trials of 72 weeks covering 3,127 and 1,613 adults [1][2].

What it found: on the strongest amount tested, average weight reduction was 11.2 percent against 2.1 percent on an inactive tablet in adults with obesity, and 9.6 percent against 2.5 percent in adults who also had type 2 diabetes [1][2].

What it does not show: how the tablet compares with the injections, since no trial has put them against each other; the only published comparison is indirect, adjusted by statistics, and was written largely by employees of a company that makes a competing medicine [3].

Almost every weight loss medicine of the past decade is a peptide, a short chain of amino acids, and almost all of them are injected, because digestion takes chains like that apart. Orforglipron is neither. It is a small molecule, assembled by ordinary chemistry rather than built as a chain, and it is swallowed once a day. Two late stage trials of it have now been published, and together they describe both what the tablet did and what has never been tested [1][2].

Why the chemistry matters

The injected medicines in this field imitate a hormone called GLP-1, so they are shaped like it, which is to say they are chains. A chain has to go under the skin because the gut would digest it. A small molecule does not have that problem, and it can be made in a chemical plant rather than grown, which changes how much of it can exist. Orforglipron reaches the same docking point on the cell, the GLP-1 receptor, from a completely different starting material [1][2].

The trial in adults without diabetes

ATTAIN-1 enrolled 3,127 adults who had obesity and no diabetes, and ran for 72 weeks. Chance decided whether each participant received one of three amounts of the tablet or an inactive tablet, alongside advice on food and activity. The main measurement was percent change in body weight at week 72, counted in a way that keeps every participant in the group they were put into, whether or not they stayed on the tablet [1][4].

Average weight reduction was 7.5 percent on the smallest amount, 8.4 percent on the middle one and 11.2 percent on the largest, against 2.1 percent for the inactive tablet. Within the largest group, 54.6 percent of participants lost at least a tenth of their body weight, 36.0 percent lost at least 15 percent and 18.4 percent lost at least a fifth, against 12.9, 5.9 and 2.8 percent in the comparison group. Waist measurement, blood pressure, triglycerides and non HDL cholesterol also improved more than in the comparison group. Between 5.3 and 10.3 percent of the people taking the tablet stopped because of side effects, against 2.7 percent of the others [1].

The trial in adults who also have type 2 diabetes

ATTAIN-2 ran the same 72 weeks at 136 sites in ten countries and enrolled 1,613 adults who had obesity or overweight together with type 2 diabetes. Average weight reduction was 5.1 percent, 7.0 percent and 9.6 percent across the three amounts, against 2.5 percent for the inactive tablet. Ten deaths were reported during the trial, six among participants taking the tablet and four among those taking the inactive one [2].

That the figures are lower in the second trial will not surprise researchers, since weight tends to move less in people with type 2 diabetes on medicines of this kind. It is also the reason a single headline number attached to a medicine is usually the wrong thing to carry around. Same tablet, same length of trial, two populations, two different answers [1][2].

The comparison that has not been run

No trial has put this tablet against an injected medicine, or against a swallowed peptide. What exists instead is an indirect comparison published in Diabetes, Obesity and Metabolism in 2026. It took participant by participant data from a trial of a swallowed peptide and published summary figures from ATTAIN-1, then adjusted by statistics for differences between the two sets of participants in sex, body weight and blood sugar status [3].

It reported the swallowed peptide ahead on weight reduction by 3.2 percentage points under one count and 3.0 under the other, with ranges of uncertainty reaching 0.4 and 0.3 points at the near end, close to no difference at all. The authors state the caveat themselves [3].

These findings provide insight into the comparative effectiveness and tolerability in the absence of head-to-head clinical trials.
Michalak and colleagues, Diabetes, Obesity and Metabolism, 2026 [3]

Two facts belong next to that sentence. An indirect comparison rests on the assumption that the two trials were alike enough to line up, which no adjustment can fully guarantee. And the paper was written mostly by employees of the company that makes the comparison medicine, with paid contractors and one academic author. Neither fact makes the analysis wrong. Both are reasons it is not the same object as a trial in which the same people were assigned to one medicine or the other [3].

What it does not show

It does not show that the tablet works better or worse than the injections, because that trial does not exist. It does not show anything about heart attacks, strokes or years lived, because neither trial measured them and 72 weeks is too short to. It does not describe what happens to weight after the tablet stops. The registry entry for the first trial, which lists 3,127 participants, still shows the study as active and not recruiting, so parts of it continue past the results already printed [1][4].

What the published record does establish is narrower and still worth knowing. A medicine that is not a peptide, taken by mouth, reduced weight by an average of 11.2 percent over 72 weeks in adults with obesity and by 9.6 percent in adults who also had type 2 diabetes, in trials that compared it with nothing except an inactive tablet [1][2].

Sources

A review of 19 skin trials found only a small effect on wrinkles

Tue, 08 Sep 2026 00:00:00 +0000

The 19 randomised trials covered 1,341 adults, and the pooled wrinkle figure only just cleared the usual threshold. Seventeen of them tested something swallowed rather than a cream, and the reviewers say the trials were poorly blinded.

What was studied: 19 randomised trials of peptide preparations for signs of skin aging, covering 1,341 adults of average age 50.2 years, gathered from three databases and combined into one analysis in March 2026 [1].

What it found: a pooled mean difference of 0.27 on wrinkle measurements, with a range of uncertainty from 0.01 to 0.52 and a P value of 0.04, an effect the reviewers call modest and put down mostly to the trials of swallowed preparations rather than creams [1].

What it does not show: which preparation does what, since the review says the wrinkle methods were not consistently described; and almost nothing about creams, since 2 of the 19 trials tested one, covering 105 of the 1,341 participants [1].

A review published in Frontiers in Medicine in March 2026 gathered the randomised trials of peptide preparations for aging skin, found 19 of them covering 1,341 adults, and added their wrinkle results into one figure. The figure is small, and the work of reading the paper is understanding what sits underneath it [1].

How the review was assembled

The search covered three databases and returned 2,479 records: 1,680 from MEDLINE, 527 from CENTRAL and 272 from Web of Science. Nineteen randomised controlled trials survived the screening. Such a trial assigns each participant by chance to the thing being tested or to a placebo, an inactive substitute, so the comparison is not decided by who volunteered for what. The review was registered in advance with PROSPERO, the public register for reviews of this kind, and its authors declare no funding and no commercial relationships [1].

What the pooled figures were

Across the trials that measured wrinkles, the pooled mean difference was 0.27, with a range of statistical uncertainty from 0.01 to 0.52 and a P value of 0.04. That P value means a difference this large would be expected by chance about four times in a hundred if the preparations did nothing at all. The bottom of the range, 0.01, sits barely above zero, which is another way of saying the result only just cleared the line [1].

Two other measurements came out larger: skin hydration at a pooled mean difference of 5.80 and skin brightness at 2.40, both at P below 0.01. Elasticity and skin density moved the same way but did not clear the threshold. No trial reported a serious side effect [1].

Why 0.27 has no unit

A mean difference is expressed in whatever units the original trials used, and here they did not all use the same ones. Some read wrinkle depth off an instrument. Some had graders score photographs. The review states that the methods behind those measurements were not consistently described, and that many trials gave no detail on which of the two produced the number. So 0.27 is an average across scales that share neither a zero point nor a step size [1].

Variation between the trials stayed high. The review treats an I squared value above 50 percent as substantial, and in one subgroup comparison it reached 96 percent. I squared estimates how much of the spread between studies comes from real differences rather than chance, so a figure that high describes trials that are not measuring the same thing [1].

The blinding problem

Pooling validity is further compromised by methodological weaknesses in the included RCTs, particularly inadequate blinding and inconsistent outcome assessment methods.
Oral and topical peptides for skin aging, Frontiers in Medicine, March 2026 [1]

Blinding is whether the participant and the person scoring the result know which group someone is in. Where the outcome is how skin looks, an assessor who knows can score generously without meaning to, and a participant who knows can report a change that is not on the photograph. The review rated several trials unclear on this point, one as high risk, and two as high risk for missing data [1].

Seventeen swallowed, two applied

The split inside the review matters more than the headline. Seventeen of the 19 trials tested a preparation taken by mouth. Two tested one applied to the skin, and those two covered 105 of the 1,341 participants. The pooled wrinkle effect came from the oral trials; the topical ones showed a smaller effect that did not clear the threshold. A great deal of marketing rests on creams. In this pooled record, creams rest on two studies [1].

One figure in the paper is worth checking against itself. The summary box at the top gives the oral subgroup a mean difference of 0.35 at P equals 0.05. The abstract and the results section give 1.5 at P equals 0.01 for what reads as the same comparison. Both are printed in the same paper, which is a reason to read the results section rather than the box above it [1].

Who ran the trials underneath

The review's own funding is clean. The studies it gathers are a separate question, and skin papers published since 2024 show the pattern. A 2024 study followed 135 women aged 45 to 65 for three months on a collagen peptide supplement, 116 of whom completed. Its expert grader wrinkle score fell from 5.9 to 5.0 at P below 0.0001, and 83.6 percent improved. It was open label, meaning everyone knew what they were taking, and it had no comparison group at all. Its five authors are employees of the company behind the product [2].

A 2025 trial of a preparation made from pig placenta was built more carefully: 90 participants aged 35 to 60, randomised, double blind and placebo controlled across 12 weeks, reporting reductions in wrinkle depth and in water loss through the skin. Two of its authors work in the research division of the company that supplies the ingredient [3].

A 2026 paper pairing a laboratory skin model with a clinical study of a swallowed combination is franker still. Four of its authors are employees of the company that paid for it, and its funding statement says that company was involved in the study design, in collecting, analysing and interpreting the data, in writing the article, and in the decision to submit it for publication [4].

What would settle it

None of that makes a result false, and a paper that states who paid and what they did is behaving better than one that does not. What it sets is how much independent confirmation a finding needs before it counts. The reviewers put their position plainly: the available data are not sufficient to support definitive conclusions about how well these preparations work [1].

Trials of creams, blinded on both sides, measured the same way in every study and run by people with nothing riding on the answer, would move this record. Nineteen randomised trials sounds like a closed question. Read one study at a time, it is not [1].

Sources

A review found 20 studies of a copper peptide, only two of them in people

Tue, 08 Sep 2026 00:00:00 +0000

The review searched three databases through March 2026. Eighteen of the 20 studies were done in cells or animals, and the two randomised human trials, from 1992 and 2006, found no difference from their comparison groups.

What was studied: a systematic review in a plastic surgery journal that searched PubMed, Embase and Cochrane CENTRAL from each database's first record through March 2026 for studies of a single copper-binding peptide used on skin [1].

What it found: 20 studies met the criteria, 18 of them in cells or animals and 2 of them randomised trials in people; the laboratory work agreed on more structural protein being made, calmer inflammation signals and new blood vessel growth [1].

What it does not show: that the molecule changes skin in a person, since the two randomised human trials on the public record, from 1992 and 2006, found no difference from their comparison groups on any measurement taken by an instrument or a blinded grader [2][3].

One molecule sits behind most of the copper peptide products in skin care: glycyl-L-histidyl-L-lysine joined to a copper ion, written GHK-Cu. It has a large laboratory literature and a very small human one. A systematic review published in Aesthetic Surgery Journal in August 2026 counted both [1].

What the review counted

The search ran across PubMed, Embase and Cochrane CENTRAL from the first record each database holds through March 2026, following PRISMA, a published checklist that sets out how a systematic review should be run and reported. Twenty studies met the criteria. Eighteen were preclinical, meaning done in cells or in animals. Two were randomised controlled trials, in which people are assigned by chance to the thing being tested or to a comparison [1].

The preclinical studies agreed with each other. Across them the molecule increased production of type I collagen and of glycosaminoglycans, both structural material in skin; changed the activity of metalloproteinases, the enzymes that break that material down; encouraged new blood vessels to form; and lowered two inflammation signals, TGF beta and IL-6. The review also reports that microneedle and liposome delivery moved considerably more of it through skin than plain application did [1].

Nevertheless, the findings reported across studies were constrained by methodological variability and a limited number of well-designed clinical trials.
The Regenerative Potential of GHK-Cu in Aesthetic Medicine, Aesthetic Surgery Journal, August 2026 [1]

The randomised human record

Searching the indexed literature for randomised studies of this molecule in people returns two.

The older is from 1992, in the Journal of Vascular Surgery, and it is about wounds rather than wrinkles. Eighty six evaluable patients with venous leg ulcers, the slow healing open sores that follow poor blood return from the legs, were randomly assigned, with the assessors kept unaware of who got what, to a 0.4 percent copper tripeptide cream, a 1 percent silver sulfadiazine cream, or the inactive vehicle alone. The silver cream shrank ulcers more than either of the others. Between the copper peptide cream and the inert vehicle, the trial reports no difference [3].

The other is from 2006, in Archives of Facial Plastic Surgery, in skin resurfaced with a carbon dioxide laser. Thirteen patients completed it. They were randomised to an aftercare routine with or without the copper peptide, and redness was measured both by software and by evaluators kept unaware of the assignment. On redness the groups did not differ. On wrinkles and overall skin quality at 12 weeks every patient improved, and again the groups did not differ. One measurement did separate them: on a questionnaire the patients completed themselves, the copper peptide group rated their skin quality higher, at P equals 0.04 [2].

Thirteen people is a very small trial, and the single measurement that separated the groups is the one the patients scored themselves. That is the weakest kind of evidence in a study of appearance, and it is the finding most often repeated [2].

Where the new work is going

Research on this molecule since 2024 is busy and almost entirely preclinical. A 2026 paper in Biogerontology fed it to Caenorhabditis elegans, a roundworm about a millimetre long that is a standard laboratory model for aging, and reported a longer lifespan along with better movement, better resistance to heat and to oxidative stress, and less build up of the pigment lipofuscin. The authors trace the effect to mitochondrial function and to two stress response pathways called DAF-16 and SKN-1, and describe their work as the first mechanistic evidence that the molecule delays aging in that animal [4].

A worm is a long way from a face. The value of a study like that is the mechanism it maps, which is what tells a later human trial where to look. It is not evidence about skin, and the paper does not present it as such [4].

What a regulator has said

The United States Food and Drug Administration keeps a list of raw ingredients that facilities mixing medicines to order may want to use, and a category inside it for ingredients the agency judges may present significant safety risks. The injectable form of this molecule was placed in that category on 29 September 2023. The reason recorded is that injectable preparations containing it may carry a risk of immunogenicity, meaning the immune system reacting to the preparation itself, because the peptide can clump together and because related impurities may be present, and that there are limited data in humans to inform safety considerations [5].

That is not a finding of harm, and it says nothing about creams, which reach the market under a different part of the law. It is a statement that the agency does not have the human data it would want before an injectable version is mixed for patients [5].

What would settle it

The review's own recommendation is larger controlled trials using standardised preparations, amounts and delivery methods. Until such a trial exists, the accurate description of this record is a consistent laboratory case that has been tested against a comparison in people twice, in 86 patients and in 13, and that separated from it neither time on any measurement an instrument or a blinded grader took [1][2][3].

Sources

A higher amount barely changed the weight result in a 76 week trial

Tue, 08 Sep 2026 00:00:00 +0000

A late stage trial split 725 adults between a lower weekly injection, a higher one, and an inactive one. The higher amount moved the main measurement barely at all and raised gut side effects to nearly nine in ten.

What was studied: 725 adults with obesity and without diabetes, assigned by chance to a lower weekly amount of an injected weight loss drug, a higher weekly amount, or an inactive injection, for 76 weeks [1][2].

What it found: under the count the trial named in advance, average weight reduction was 12.2 percent on the lower amount and 13.0 percent on the higher one, against 5.4 percent in the comparison group, while stomach and gut side effects reached 80.9 percent and 89.7 percent of the two treated groups [1].

What it does not show: that the medicine does more or less than any other, since none was tested against it; what happens after 76 weeks; or that the higher amount earns the extra side effects, which the trial did not set out to answer [1][4].

A late stage trial of an injected weight loss drug tested two amounts of it against an inactive injection. Over 76 weeks the higher amount produced only a slightly larger average weight reduction than the lower one, while side effects in the stomach and gut were more common with it. The results appeared in the New England Journal of Medicine in August 2026 [1].

What the trial did

Seven hundred and twenty five adults took part. All had a body mass index of 30 or higher, or of 27 or higher with at least one weight related health problem, and none had diabetes. Chance decided which of three groups each person joined: a lower weekly injection of survodutide, a higher weekly injection of it, or placebo, an inactive substitute given so that investigators can separate what a medicine did from what would have happened anyway. Two hundred and forty one people went into the lower group, 242 into the higher group and 242 into the placebo group, and everyone also received counselling on food and activity. Average age was 47.1 years, average body mass index 37.9, and average weight 108.8 kilograms [1][2].

Survodutide acts on two receptors rather than one. Receptors are docking points on a cell surface, each shaped to fit a particular hormone. This molecule switches on the GLP-1 receptor, the one the familiar weight loss medicines act on, and the glucagon receptor, which researchers study for its role in energy use and in liver fat [1].

The numbers the journal printed

The main measurement was percent change in body weight from the start to week 76. Under the count the investigators named in advance as primary, the lower amount produced an average reduction of 12.2 percent, with a range of statistical uncertainty running from 13.6 to 10.8 percent. The higher amount produced 13.0 percent, with a range from 14.4 to 11.6. The placebo group came in at 5.4 percent, with a range from 6.9 to 4.0. A reduction of at least 5 percent was reached by 72.6 percent, 71.9 percent and 46.3 percent of the three groups in that order [1].

Two things in that list are worth sitting with. The first is that the higher amount barely separated from the lower one, and that slightly fewer of the people taking it reached the 5 percent mark. The second is the comparison group, which lost an average of 5.4 percent on an inactive injection and counselling. That is a lot, and it narrows the space the medicine had to work in [1].

Side effects moved the other way. Symptoms in the stomach and gut, meaning nausea, vomiting, diarrhoea and related trouble, were reported by 80.9 percent of the lower group, 89.7 percent of the higher group and 47.9 percent of the placebo group. The report describes them as typically mild to moderate. No deaths were reported [1].

The figure released in April, and the figure printed in August

Months before any of this reached a journal, the company that discovered the molecule published an announcement of the headline result [4].

Adults living with obesity or overweight, without type 2 diabetes, who were treated with survodutide experienced sustained weight loss of up to an average of 16.6% after 76 weeks using the efficacy estimand, a statistically significant decrease versus 3.2% in the placebo arm (p<0.0001).
Zealand Pharma, results announcement, April 2026 [4]

Nothing in that sentence is untrue, and it names its own count. An estimand is the exact question a trial's main number answers. The efficacy estimand asks what happens in people while they are actually taking the medicine. The treatment regimen estimand, the one this trial named as primary and the one the journal printed, keeps everyone in the group they were assigned to, including those who stopped early. Under the first count the two figures are 16.6 percent and 3.2 percent. Under the second they are 13.0 percent and 5.4 percent [1][4].

Both describe the same 725 people. The first answers what the medicine does when taken; the second answers what becomes of a group started on it. Neither is wrong, and a reader who meets only one has half the record [1][4].

A second trial, about the liver

A separate late stage trial of the same molecule, published in Nature Medicine in 2026, enrolled 216 adults who had obesity together with fatty liver disease linked to metabolic problems. One hundred and forty six received the medicine and 70 received placebo, for 48 weeks, and liver fat was measured by scan. A reduction of at least 30 percent in liver fat was reached by 84.2 percent of the treated group against 24.3 percent on placebo under the on treatment count, and by 68.5 percent against 28.6 percent under the count that keeps everyone as assigned [3].

The authors set out their own limits: 48 weeks is short, and participants came from only two countries, the United States and Spain. Liver fat on a scan is a measurement rather than an outcome, and whether less of it means fewer people develop liver failure later is a question this trial did not ask [3].

What it does not show

Nothing here compares the medicine against another medicine. The only comparison run was against an inactive injection, so any ranking against the injections already in use would have to come from a trial nobody has done. The registry entry shows the study started in November 2023 and finished collecting its main measurement in December 2025, and it stops at 76 weeks, which leaves year three unmeasured [1][2].

The population is narrow in the usual way: adults with obesity, no diabetes, near 108 kilograms on average at the start. And the side effect figures are the part of this record least likely to soften with familiarity. Close to nine in ten people on the higher amount reported symptoms in the stomach and gut, for a main result that barely moved away from the lower amount [1].

Sources

A weight loss drug group scored younger on body age tests in a trial

Mon, 07 Sep 2026 00:00:00 +0000

Blood stored during a 32 week trial in 84 adults was run through three tools that estimate biological age. All three scored the drug group younger, in a question put to the data only after the trial had ended.

What was studied: blood stored during a randomised 32 week trial of a weight loss drug in 84 adults living with HIV, tested again after the trial closed with tools that estimate biological age from chemical marks on DNA [1].

What it found: the drug group came out about 4.9 years younger on one tool and about 3.1 years younger on another, and a third put the rate those marks were building up about 9 percent lower than in the placebo group [1].

What it does not show: that anyone lived longer or felt better, that the pattern holds in other groups, or that the finding is settled at all, since the question was asked only after the trial had ended [1][2].

A trial designed to study body fat has produced a finding about aging. After the trial closed, investigators returned to blood that had been frozen along the way and ran it through tools that estimate how old a body looks, rather than how many years it has lived. The results were published in Nature Communications in May 2026 [1].

The trial the samples came from

The original study was a randomised controlled trial: participants were assigned by chance to receive either semaglutide or a placebo, an inactive substitute given so that investigators can separate what the drug did from what would have happened anyway. Eighty four adults living with HIV took part, all of them with lipohypertrophy, a build up of fat in places on the body where it does not normally gather. Forty five were assigned to semaglutide and thirty nine to placebo, and the study ran for 32 weeks with neither side knowing who had received which until it was over [1].

Its stated question was about fat. Aging entered later, when the team went back to stored samples and measured something the trial had never been built to answer. That distinction runs through everything below [1][2].

What the tools measure

The tools are called epigenetic clocks. Cells carry small chemical marks on their DNA that help decide which genes are switched on, and the pattern of those marks shifts with age in ways a statistical model can be trained to read. Fed a blood sample, such a model returns a number: an estimate of biological age, which can sit above or below the age on a person's birth certificate [1].

The analysis used several of these models rather than one, because they are not interchangeable. Each was trained on a different target, so each is really a separate estimate of the same loose idea. One of the three, DunedinPACE, reports a rate rather than a total: not how old the body looks, but how quickly the marks are piling up during the period being measured [1].

What the re-analysis reported

On PhenoAge, the semaglutide group came out about 4.9 years younger than the placebo group, with a P value of 0.004. On PCGrimAge the gap was about 3.1 years, at P equals 0.007. DunedinPACE, the rate measure, ran about 9 percent slower in the semaglutide group, at P equals 0.01. A P value of 0.004 means that a difference this large would be expected to appear by chance about four times in a thousand if the drug had done nothing at all [1].

The differences persisted after the investigators accounted for sex, body weight and inflammation, so weight reduction alone does not obviously explain them. What carries more weight than any single figure is the direction: every clock the team applied pointed the same way. The authors present the result as the first evidence from a randomised trial that a medicine of this kind can move these measures, in the population they studied [1][2].

Why the design caps what can be claimed

This was a post hoc analysis, meaning a question asked of the data after the fact rather than one the study was set up to answer. Nothing about that makes a result wrong. It does change how much it can carry: the size of the study was chosen for the fat measurements, no aging outcome was named in advance, and analyses of stored samples are where new hypotheses are generated rather than confirmed [1].

The population is narrow as well. These were adults living with HIV and a specific pattern of fat build up, which is not the general population, and 32 weeks is a short window in which to observe anything described as aging. Whether the same shift appears in other groups, or holds over years rather than months, is untested [1].

There is also the gap between a marker and an outcome. A clock reading is a laboratory measurement, not a record of health or of years lived. How tightly movement in these measures tracks anything a person would notice remains an open research question, and a trial that moved a clock is not the same as a trial that changed a life [1][2].

A separate question about the same family of medicines

Elsewhere, laboratory groups have turned to a separate question about the same class of medicine. At the University of Colorado Anschutz Cancer Center, researchers are testing whether these medicines act on immune cells directly, rather than only through appetite and blood sugar. The reasoning begins with a pattern visible across large populations, in which obesity travels with raised rates of several cancers, and works downwards to individual cells [3].

That work is at an early stage: experiments in cells, no published clinical trial, no figures worth quoting yet, and an application for federal funding to continue. It sits in this briefing for a single reason: the receptor these medicines act on appears on a wider range of cell types than the weight loss account ever needed, and that is the sort of observation that makes an aging result worth testing properly [3].

What would settle it

A trial that names biological age as its main measurement before it begins, in a broader group of participants, running long enough for the estimate to mean something. Until then the honest description of this result is a strong signal from a small trial, found by looking again at samples that had already been collected, in patients whose situation was unusual. Coverage that has reduced it to a claim that a weight loss medicine slows aging is reporting something the study did not test [1][4].

Sources

The biggest results for a three receptor drug are still unpublished

Mon, 07 Sep 2026 00:00:00 +0000

One mid stage trial has been through peer review, reporting an average weight reduction of 24.2 percent over 48 weeks. The much larger figures now in circulation come from a company announcement and a conference talk.

What was studied: an experimental molecule that acts on three hormone docking points, while earlier medicines of the kind reached one or two of them, tested for weight reduction in people with obesity [1][2].

What it found: in the published mid stage trial, people on the strongest amount tested lost an average of 24.2 percent of body weight over 48 weeks against 2.1 percent on placebo [1]; the large late stage trial reports either 28.3 percent or 25.0 percent over 80 weeks, depending on how people who stopped early are counted [2][3].

What it does not show: the late stage results have never been through peer review, the 104 week average has no comparison group behind it because the placebo group was moved onto the medicine for that stage, and no regulator anywhere has approved the drug [3][4].

Retatrutide has produced the largest weight reductions yet reported for a medicine of this kind. Almost none of that has appeared in a journal. Separating what has been published from what has been announced is most of the work of reading this record, and it is a useful exercise well beyond this one molecule [1][2][3].

What the molecule does differently

Medicines in this family work through receptors. Receptors are docking points on the cell surface, each shaped so that one particular hormone settles into it. Semaglutide acts on a single receptor, GLP-1. Tirzepatide acts on that one and on a second, GIP. Retatrutide is built to act on those two plus a third, the glucagon receptor. Three targets is what makes it, in the research literature, a triple agonist. An agonist is simply a molecule that turns a receptor on [1][7].

The glucagon receptor is the genuinely new part. It is studied for its role in how the body spends energy and handles fat in the liver, a route the one and two receptor medicines leave untouched. Whether that third target explains the trial figures is not established. Investigators describe the mechanism as incompletely understood, which is a fair summary of where the biology sits [1][6].

The one trial that has been published

Only one obesity study of the molecule has completed peer review, the process in which independent scientists examine a study before a journal will print it. It appeared in the New England Journal of Medicine in 2023. Over 48 weeks, participants receiving the strongest amount tested lost an average of 24.2 percent of their body weight, against 2.1 percent for participants given placebo, an inactive substitute [1].

That study was phase 2, the middle stage of human testing. Its job is to establish whether a molecule performs well enough, and at what strength, for a far larger study to be justified. Phase 2 asks whether the question deserves a bigger trial rather than settling the answer, and no other obesity trial of retatrutide has had its full results printed by a journal [1].

The large trial, announced rather than published

The larger trial is TRIUMPH-1, registered as NCT05929066, which enrolled 2,339 participants and ran for 80 weeks. The company that makes the molecule released its results in May 2026, and a scientific meeting heard them presented in June. No journal has published them. That describes where the evidence sits rather than criticising how the trial was run [2][3].

Lilly's triple agonist retatrutide delivered powerful weight loss in pivotal Phase 3 obesity trial
Eli Lilly and Company, results announcement, May 2026 [3]

A sponsor headline is a legitimate primary record of what a company says it found. It is not the same object as a peer reviewed paper, and the difference is worth holding onto while reading any of the numbers below [3][4].

Two counts, two headline figures

There are two defensible ways to count the trial's main result, and most coverage picks one of them and moves on. The first count treats participants as though all of them had stayed on the medicine for the full period; on that basis the strongest amount tested produced an average reduction of 28.3 percent against 2.2 percent for placebo. The second count keeps everyone as randomised, including those who stopped early, and gives 25.0 percent against 3.9 percent [2][3].

Both figures sit in the same announcement and neither is wrong. They answer different questions: the first asks what the molecule does when taken as intended, the second asks what becomes of a group of people offered it in practice. The company led with the higher figure while some coverage carried the lower one, which is why two headline numbers exist for one trial and a second source may not match the first [3][4].

The same announcement reports that 45.3 percent of that group lost at least 30 percent of their body weight, against 0.5 percent on placebo. A smaller set of participants continued to 104 weeks and reached an average of 30.3 percent. One caveat travels with that longer figure: everyone who had been on placebo was switched to the medicine for the extension, which leaves that stage with no comparison group at all [2][3].

What else the programme measured

The TRIUMPH programme looked beyond weight. Two further conditions were studied inside it: obstructive sleep apnea, where breathing halts and restarts repeatedly through the night, and osteoarthritis of the knee, in which the joint wears down and becomes painful and stiff. Among the sleep apnea participants, interruptions to breathing per hour of sleep fell by roughly 60 percent. Among the knee participants, a standard score for pain and function improved by about three quarters [2][6].

A separate diabetes trial, TRANSCEND-T2D-1, registered as NCT06354660, reported that 46 percent of its participants brought a long term measure of blood sugar below 5.7 percent by week 40. Every one of these figures carries the same label as the weight results: registered, announced, not yet published [2][5].

Where the record stands

Nowhere is retatrutide an approved medicine. Health Canada, the United States Food and Drug Administration and their counterparts have not evaluated it, and the sponsor has indicated that it does not expect to seek approval before late 2027. Specialists who have commented publicly describe a three receptor molecule as a real departure from what came before, while noting the same absence of approval [7].

The evidence, stated plainly: one published mid stage trial in a leading journal, one large late stage trial whose results exist as an announcement and a conference talk, two defensible ways of counting its main result, and an extension phase that no longer has a control group. Peer reviewed publication of the late stage data is the event that would change this record, and until it arrives the difference between the two kinds of source is the whole story [1][2][3].

Sources

Less thyroid eye disease on a newer blood sugar medicine, records show

Mon, 07 Sep 2026 00:00:00 +0000

Two matched groups of 173,618 patients with thyroid disease were followed through hospital records. Diagnoses ran about 18 percent lower at one year and operations about 58 percent lower, in a design that cannot show cause.

What was studied: hospital records from a United States network, covering people with thyroid disease who were starting one of two kinds of blood sugar medicine, matched into two comparable groups and followed for three years [1].

What it found: thyroid eye disease was diagnosed about 18 percent less often on the newer medicine at one year, and operations for it ran about 58 percent lower at one year and 46 percent lower at three [1][2].

What it does not show: that the medicines prevented anything, since nobody was assigned a treatment; nothing about how a thyroid works or the hormones it makes; and the same kind of medicine carries a thyroid warning that two national studies still disagree about [3][5][6].

Thyroid eye disease is a complication of an overactive thyroid. The immune attack driving the thyroid problem also reaches the soft tissue behind the eyes, which swells and pushes the eyeball forward. Vision can double, and severe cases need steroid medicines or surgery. A 2026 records study asked whether patients on a GLP-1 medicine are diagnosed with it less often, and found that they are [1].

What the study did

Investigators searched a large United States network of hospital records for people with thyroid disease who were beginning either a GLP-1 medicine or another kind of blood sugar medicine. Each patient was then paired with someone in the other group of similar age and sex, with a similar body mass index, similar blood sugar control, similar other illnesses and diabetes medicines, similar past radioactive iodine treatment and similar thyroid laboratory values. That left 173,618 patients on each side, 347,236 in total, alike on every factor the records held. What happened next was simply counted [1].

What it found

Diagnoses were less frequent on the GLP-1 side at each interval the investigators checked: roughly 18 percent lower after one year, 14 percent after two and 10 percent after three. In the paper's terms, the hazard ratio at one year was 0.82, with a range of statistical uncertainty from 0.76 to 0.88. A hazard ratio of 1.00 would mean no difference, and 0.82 means the event happened about 18 percent less often [1].

The difference was largest on the measures that mark a severe case. Operations ran about 58 percent lower at one year, a hazard ratio of 0.42 with a range from 0.36 to 0.49, and about 46 percent lower at three years. Steroid medicines given by mouth or by drip were also used less. The pattern held within the group whose thyroids were overactive, and it appeared for individual medicines as well as for the class as a whole [1][2].

The investigators kept their own conclusion narrow, and its wording carries more information than the percentages do. They report results consistent with GLP-1 signalling playing some protective part in eye inflammation, and say that trials assigning patients in advance would be needed before anyone could call it established. A strong lead, presented as a lead [1][2].

What a records study can and cannot carry

This was a retrospective cohort study, meaning it looked back at information already collected for other purposes. Nobody was assigned to a medicine. Doctors chose who received what, and the reasons behind those choices are not fully written down in a record. Matching makes two groups resemble each other on the factors that were measured and can do nothing at all about the factors that were not [1].

The population is also specific. Every person counted was living with thyroid disease and beginning a blood sugar medicine, so nothing here can be moved onto anyone outside that description. The condition studied is one narrow complication, and this work reports nothing about thyroid function or hormone levels [1].

The warning on the same record

Approved labels in this class carry a thyroid warning set inside a black box, the most serious warning a prescribing document carries in the United States. Rats and mice that received semaglutide across their whole lifetime developed more tumours in one particular kind of thyroid cell, and more of them again as exposure grew larger and longer [5].

It is unknown whether semaglutide causes thyroid C-cell tumors ... in humans, as the human relevance of semaglutide-induced rodent thyroid C-cell tumors has not been determined.
United States prescribing information for semaglutide, boxed warning [5]

The label still acts on that uncertainty. It bars the medicine outright for patients who have had medullary thyroid cancer or whose close relatives have, and for anyone with the inherited condition called MEN 2. In label language that is a contraindication, meaning do not give at all rather than give with care [5].

Two national studies that disagree

Two large database studies have since looked for the rodent finding in humans and reached opposite conclusions. A French national study found a higher rate of thyroid cancer where GLP-1 use had run for one to three years, a hazard ratio of 1.58 with a range from 1.27 to 1.95. A bigger Scandinavian study, following 145,410 new users against 291,667 people starting a different medicine, found no increase: a hazard ratio of 0.93 with a range from 0.66 to 1.31, wide enough to include no difference [6][7].

Part of the disagreement is design. Part is peculiar to thyroid cancer, much of which is found by accident during scans ordered for other reasons: a patient who has recently begun any new medicine tends to be examined more often, which can raise the count without raising the number of cancers. The question is open in both directions [6][7].

What has not been tested

A 2026 review gathered the wider evidence on these medicines and autoimmune thyroid disease, the conditions in which the immune system attacks the thyroid. It describes early reports of effects on thyroid size, thyroid function and immune activity, then states that dedicated trials in these conditions are lacking. A review tests nothing itself and is worth what the studies inside it are worth [3].

Laboratory work sits further back again. A team in Mainz worked with mice bred to develop a form of Graves' disease, treating them with a small ring shaped peptide copied from part of the receptor that the disease antibodies target. The animals improved on several measures, and a second laboratory working independently reproduced the thyroid result, which few animal findings ever get. It is still a mouse study, and findings in animals reach people only a minority of the time [4].

Set in order: a large records study that found an association worth testing properly, a review that says the trials do not exist yet, an animal experiment on an unrelated molecule, and a label warning two national studies cannot resolve. A trial assigning patients in advance would move any of it [1][3][6][7].