The Lancet does not issue Expressions of Concern casually. When The Lancet published one on May 18, 2026, citing findings from a regulatory inspection at King’s College Hospital NHS Foundation Trust, it placed a formal question mark over one of the most significant Parkinson’s disease trials in recent memory: the Phase 3 exenatide once-weekly study led by Nirosen Vijiaratnam and colleagues, originally published in February 2025. The inspection, which formed part of a broader regulatory review, identified department-wide concerns relating to trial conduct and oversight, not isolated to a single protocol deviation or a single coordinator acting outside their training. Department-wide. That phrase carries regulatory weight that the clinical operations community should not skim past.

Exenatide had earned its moment. The Vijiaratnam trial enrolled 194 participants aged 25 to 80 years across six UK research hospitals, randomised 1:1 to receive extended-release exenatide 2 mg or placebo by subcutaneous injection, with the aim of demonstrating disease-modifying activity in Parkinson’s disease. No approved disease-modifying therapy for Parkinson’s currently exists. According to a January 2024 landscape analysis, there were 136 active Phase 1 through Phase 3 trials evaluating drug therapies for the disease, and not one had crossed the finish line with a label that says “modifies the underlying pathology.” The exenatide program, building on GLP-1 receptor agonist biology already validated in metabolic disease, represented a plausible and well-funded bet on that gap.

Now the data supporting that bet sits under a compliance cloud.

What the Inspection Actually Signals

The MHRA’s Good Clinical Practice inspections framework applies to all organisations involved in clinical trials of medicinal products in the UK, and it does not distinguish between academic medical centres and commercial sponsors. An NHS trust running a Phase 3 trial operates under the same GCP obligations as a top-ten pharma company. When an inspection at a single site yields department-wide findings rather than individual procedural lapses, the regulatory interpretation is straightforward: systemic oversight failure, not human error.

Sponsors and clinical operations leaders have long treated the “multicentre” design as a risk-distribution mechanism. The logic runs that if one site fails, five others carry the data. But that assumption collapses when the compromised site holds a meaningful fraction of 194 participants, or when the site in question is King’s College Hospital, a major London academic centre whose involvement lent the trial scientific credibility from day one. The old operating assumption was that academic-anchor sites could be managed with lighter monitoring intensity because their internal governance systems were presumed robust. The MHRA findings at King’s College Hospital directly contradict that assumption. What sponsors should assume now: investigator site reputation does not substitute for a documented, risk-calibrated monitoring plan executed against it.

This matters beyond exenatide. A Phase 2 liraglutide trial in Parkinson’s disease, enrolling 63 patients, had already reported initial results through the Van Andel Institute, establishing a GLP-1 in neurodegeneration signal worth pursuing. If the exenatide Phase 3 data cannot be cleanly defended through regulatory review, the entire GLP-1 neurodegeneration hypothesis loses its best evidentiary anchor at the worst possible time, just as sponsors are making pipeline decisions about whether to advance similar programs into pivotal studies.

The Monitoring Failure Hiding in Plain Sight

Look at the inspection timeline against the trial calendar. The study was published in February 2025. The MHRA inspection findings were communicated to The Lancet in May 2026, more than a year after publication. Regulatory inspections at trial sites are not surprise events in the abstract; they are part of a structured oversight process that GCP requires sponsors and principal investigators to anticipate and prepare for. The question clinical operations teams should be asking is not whether the inspection found something, but why department-wide governance concerns were not surfaced earlier, either by the sponsor’s own monitoring visits, by the trial steering committee, or by the NHS Trust’s internal research governance mechanisms.

Under the MHRA’s GCP inspection framework, sponsors bear direct responsibility for the oversight of their investigator sites. The framework does not offer a safe harbour that says “we relied on the site’s internal systems.” That responsibility includes ensuring that source data verification is conducted with appropriate frequency, that site staff training is current, that protocol deviations are captured and escalated, and that oversight of those activities is documented at the monitoring level. If King’s College Hospital had systemic conduct concerns across a department, those concerns should have generated deviation reports, follow-up correspondence, and potentially a protocol amendment or site pause during the trial’s active period. An Expression of Concern issued more than 15 months after publication suggests those signals were either not generated or not acted upon.

King’s College Hospital has prior regulatory history worth noting. An FDA inspection of the Denmark Hill site conducted from February 19 to February 23, 2001, resulted in a VAI (Voluntary Action Indicated) classification, indicating objectionable conditions that did not meet the threshold for regulatory action at that time. A single VAI from 25 years ago proves nothing in isolation. But the pattern of a site with identifiable historical compliance findings participating in a pivotal trial without heightened monitoring frequency is a risk management question that sponsor teams and ethics committees should be prepared to answer.

What Clinical Ops Leaders Must Do Now

If you are managing an active multicentre trial at UK NHS sites, this inspection outcome changes your monitoring calculus immediately. Pull your site selection documentation and confirm that governance capability assessments, not just investigator CVs and patient throughput numbers, were part of your feasibility criteria. Review your monitoring visit logs for each site and confirm that source data verification coverage is documented against a pre-specified risk-based threshold, consistent with the principles in your protocol and your oversight plan. If any site has generated a pattern of minor deviations that were individually coded as low-risk, aggregate them now. Department-wide findings in MHRA inspections do not emerge from nowhere; they emerge from patterns that monitoring reports can often detect if the reviewer is looking for systemic signals rather than incident-level compliance.

For sponsors with GLP-1 programs targeting neurodegeneration specifically, the exenatide Expression of Concern creates a competitive intelligence problem as much as a regulatory one. Regulators reviewing a future NDA or MAA in this space will scrutinise the exenatide precedent as the closest comparable dataset. If that dataset is materially impaired, sponsors cannot simply cite it to establish clinical plausibility; they must build their own evidence foundation without the support of a validated Phase 3 anchor trial.

Watch for two signals over the next 90 days. First, whether The Lancet proceeds to retraction or correction, or whether the Expression of Concern resolves through additional documentation from the trial team, as those two outcomes carry very different downstream consequences for the exenatide regulatory dossier. Second, whether the MHRA publishes any formal inspection classification for King’s College Hospital publicly, which would indicate whether the findings rose to the level of Critical or Major GCP deficiencies under their classification scheme. A Critical finding at a pivotal Phase 3 site would almost certainly require the trial sponsor to notify the relevant regulatory authorities in any jurisdiction where this data has been submitted or is being considered, and would force a formal data integrity assessment before that dossier could move forward. The GLP-1 neurodegeneration hypothesis is scientifically credible and commercially significant. Whether it survives its best clinical test depends on what those inspection documents actually say.

References

  1. The Lancet — “Expression of Concern: Exenatide once a week versus placebo as a potential disease-modifying treatment for people with Parkinson’s disease in the UK”
  2. PubMed — Vijiaratnam et al., Phase 3 exenatide Parkinson’s trial, 194 participants, six UK research hospitals
  3. MHRA — “Clinical trials for medicines: Good clinical practice inspections”
  4. FDA Clinical Investigator Inspection List — King’s College Hospital Denmark Hill, VAI outcome, February 2001
  5. Van Andel Institute — Liraglutide Phase 2 Parkinson’s trial, 63 patients enrolled
  6. PubMed — Parkinson’s disease treatment pipeline, 136 active Phase 1-3 trials as of January 2024
Website |  + posts

Moe Alsumidaie, MBA, MSF, is founder and Chief Editor of Vanguard Publications, which publishes Clinical Trial Vanguard, Pharma Vanguard and BullScope, and Head of Research at CliniBiz. He has two decades in clinical trial operations and data science, with earlier roles at Genentech, Abbott Vascular and Stanford University Medical Center, and is a guest lecturer in clinical trial sciences at Rutgers University.