The drug didn't work. A year later, the paper was flagged.

In 2024, a French trial reported that a drug that is used to treat diabetes appeared to stop symptoms from getting worse, a surprising result considering Parkinson's continually gets worse over time. Optimism for future developments with that class of drug rested on a larger trial of a very similar one, exenatide, which was published in early 2025.

That exenatide trial found no reportable benefit for Parkinson's. Then, in May 2026, the paper was flagged and placed under review.

A promising drug

Exenatide was considered to be one of the most promising treatments for Parkinson's.

A similar phase 2 trial published in 2017 tracked about 60 people for 48 weeks and found that those using exenatide ended up with symptoms around 4 points better on the motor scale than those on the placebo. This improvement also continued for a duration of around 12 weeks after the treatment ended, suggesting long-lasting results. This led to more follow-up research.

However, this was still a relatively small trial. This Phase 3 trial was designed to follow up on the hopeful results that the phase 2 trial provided.

What this trial did

The trial enrolled 194 Parkinson's patients across six different hospitals in the UK. Participants were randomly assigned to weekly dosages of either exenatide or a placebo. They followed these patients for 96 weeks, which was close to two years and almost twice the length of the earlier trial.

The primary endpoint was the motor examination score on the MDS-UPDRS. Prior to examination, participants had to withhold their usual Parkinson's medication overnight to prevent interference with the results, clearing out the effects of levodopa and revealing the true state of the disease without meds.

Their results

The exenatide did not stop symptoms from worsening, and the two groups got worse at almost the same rate.

The results slightly favored placebo, although the difference was insignificant and does not suggest anything meaningful. The p-value of 0.47 means that if the drug did nothing at all, this apparent gap in motor score would appear around half of the time. The confidence interval is more important to consider here. A negative means that the exenatide was better, whereas a positive favors placebo. The range included zero and is narrow enough to show that the trial was large enough to rule out the 2017 trial's possible benefit because there is no meaningful difference. It is important to note that the MDS-UPDRS scale has a range of 0 to 132, so this difference is close to negligible.

The research found that it simply did not benefit patients with Parkinson's.

The paper was flagged

On June 27th, 2026, more than a year after publication, The Lancet issued an expression of concern about the paper.

At one of the six sites, King's College Hospital NHS Foundation Trust, regulators found department-wide problems with how clinical trials held there were being conducted and overseen. Some findings were graded critical and others major, two serious categories in an inspection like this.

The Lancet asked the corresponding author's institution to investigate, attaching the notice alongside this request.

This expression of concern, however, does not mean that the study's result is wrong or should be disregarded. It simply raises a concern that has to be investigated.

Until this investigation is resolved, the result should be taken with a grain of salt since nobody can pin down how much of an influence that this questionable site had on the data.

Comparison of the two trials

At the moment, it is difficult to determine which study is closer to the truth.

The two trials differ in how patients were assessed. The French trial captured patients while they were on medication, possibly giving a more forgiving picture of symptomatic effect rather than the underlying state of disease that the exenatide trial looked for. Another difference is in duration. The 12 months for the lixisenatide trial could have showed a small early difference that disappears when the trial reaches a longer period closer to 96 weeks. It could also be a fundamental difference in the drugs or a false positive in the French trial, similar to what the 2017 exenatide trial might be.

Until this ongoing investigation into the 2025 trial is concluded, one more possibility is that the negative result itself is unreliable. Although this cannot be ruled out yet, this seems unlikely given that only one of the six sites is under scrutiny.

Why it matters

This hypothesis was able to reach a large-scale phase 3 trial after almost a decade. Although this result came out to be negative, it can benefit the research community because other groups can pivot away from ideas that have proven to not work and patients will stop enrolling in studies of something that will not help them.

The 2017 trial was conducted thoughtfully, but its sample size was small, taking data from only around 60 people. Small trials can produce uncertain estimates and can inflate excitement around positive results, even if they may not hold up in a larger trial, a relatively common phenomenon.

What's important to watch now is the impending conclusion of the investigation into the study. Once it is released, The Lancet will either retract the notice, correct the issue, or retract it altogether.