Picture the monitoring visit report from a time-sensitive ischemic stroke trial: treatment window listed as 4.5 hours, median time from symptom onset to treatment logged at six hours plus, and a deviation log that reads like a transcript of everything that was supposed to happen before the patient ever reached the infusion chair. The protocol said the site was ready. The site was not ready. And now a drug that may have worked is sitting behind a negative result.
That is the operational ghost haunting the THEIA trial. Published in The Lancet Neurology, the trial compared intravenous alteplase against oral aspirin for acute central retinal artery occlusion — an ocular ischemic stroke equivalent where irreversible retinal injury begins within minutes of vessel occlusion. The trial found no significant benefit for alteplase. A correspondence published in the same journal argues, credibly, that the result may reflect limitations in contemporary systems of care and trial execution rather than a definitive absence of therapeutic efficacy. That is not a defense of the drug. It is a diagnosis of the trial.
The Timeline the Protocol Assumes
Central retinal artery occlusion is among the most time-dependent conditions in all of vascular neurology. The treatment rationale for alteplase here mirrors its use in acute ischemic stroke: the earlier the intervention, the more salvageable tissue. Every hour of delay compresses the therapeutic window and dilutes any signal the drug might produce. A trial that cannot reliably deliver treatment within the protocol-specified window is, operationally, a different trial from the one the statistician powered.
This is the enrollment timeline problem that sponsors routinely underestimate in emergency and hyperacute indications. Site qualification visits check whether a site has the equipment, the staff credentials, and the regulatory approvals. What they rarely audit with sufficient rigor is the actual patient flow: How long does it take from ophthalmic emergency presentation to neurology or stroke team activation? Who calls whom, and in what order, at 11 PM on a Saturday? Sites I work with in stroke and vascular neurology routinely discover, at the first monitoring visit, that their theoretical pathway and their actual pathway diverge by 90 minutes or more — and in a four-and-a-half-hour window, that divergence is the entire margin.
ICH E6(R3), finalized in 2023, reinforces what GCP has always required: that investigator oversight of protocol-critical procedures must be substantive, not ceremonial. Section 5.5 of the revised guideline addresses sponsor responsibilities around site suitability in a way that should make any CTM nervous when the site’s “emergency pathway SOP” is a two-page document that has never been stress-tested on a live patient. The SOP existing and the SOP working are not the same artifact.
Deviation Logs Don’t Lie, Enrollment Plans Do
The correspondence authors point to systems-of-care limitations as a potential confounder of the THEIA efficacy signal. Operationally, that phrase covers a specific set of failures: sites selected without adequate triage infrastructure, enrollment windows that were aspirational rather than validated, and the compounding effect of screen failures driven by patients presenting outside the treatment window — not because the drug failed, but because the site couldn’t capture them in time.
Screen failure rates in hyperacute trials routinely run 40 to 60 percent above projections when sites have not formally mapped their patient pathway before first-patient-in. Every screen failure in a time-window trial carries a story: presented at hour five, transferred from a community hospital that wasn’t part of the network, referred by an ophthalmologist who didn’t know the trial existed. These are not random losses. They are the compounding product of site selection decisions made months earlier, when the feasibility questionnaire asked about “stroke team availability” but not about the median door-to-neurology-consult time on a weekend night.
FDA’s guidance on Oversight of Clinical Investigations, and the agency’s BIMO inspection framework, both treat protocol deviation frequency as a signal of site readiness failure — not just documentation failure. When a site’s deviation log shows repeated instances of delayed informed consent, missed assessment windows, or out-of-window procedures clustered in the first 60 days of enrollment, the root cause is almost never the coordinator. The root cause is a site qualification process that validated credentials and ignored workflow. A CAPA that writes corrective action at the coordinator level while leaving the patient pathway unexamined will not survive an effectiveness check.
That same CAPA logic applies here at the trial level. If the THEIA result prompts a sponsor to conclude that alteplase lacks efficacy in retinal artery occlusion and close the door on further development, they will have written a corrective action that addresses the outcome — a negative trial — without examining the root cause, which may be a systems failure that a better-designed and better-executed trial could overcome. FDA warning letters from BIMO inspections have cited this exact pattern: sponsors who accepted trial failure without conducting adequate root cause analysis of execution quality, foreclosing development pathways prematurely.
What Monday Morning Looks Like
For operators running any hyperacute or time-sensitive protocol right now, the THEIA execution questions are a pre-startup checklist that needs to happen before the site agreement is signed, not after first-patient-in. The site qualification visit must include a timed walk-through of the actual patient pathway — from emergency presentation to study drug administration — with the specific personnel who will be on call, not the clinical research coordinator who will be working banker’s hours. If that walk-through cannot be completed in time to meet the protocol window under realistic staffing conditions, the site is not qualified, regardless of what the feasibility questionnaire says.
For sponsors: the enrollment plan needs a screen failure assumption that reflects the time-window reality, not the optimistic scenario. In hyperacute vascular trials across our network, validated pathway sites — those that have completed a prospective patient flow audit prior to activation — show screen failure rates approximately 30 percent lower than sites activated on credential review alone. That gap translates directly into sample size adequacy and statistical power. A trial that is underpowered because its sites could not deliver patients within the protocol window will return a negative result that looks like a drug failure and reads, in the correspondence literature, exactly like THEIA.
The ICH E6(R3) framework gives sponsors the explicit authority to require enhanced site oversight for high-complexity or time-critical protocols. Use it before the first deviation, not after the fifteenth. Add a pre-startup pathway validation requirement to the site initiation visit checklist, document it in the TMF as a site-specific risk mitigation measure, and build the trigger for a focused monitoring visit if the first three enrollments show time-to-treatment outside the target window.
Alteplase may yet have a role in retinal artery occlusion. The biology has not changed. What needs to change is the infrastructure that surrounds the next trial — because a drug that arrives 90 minutes too late, at a site that was never operationally ready to deliver it on time, is not a drug that failed. It is a trial that did.
References
- The Lancet Neurology — “Gaps in system of care in acute central retinal artery occlusion” [Correspondence] (2026)
- The Lancet Neurology — Préterre C et al., “THEIA trial: intravenous alteplase versus oral aspirin for acute central retinal artery occlusion” (2026)
- ICH E6(R3) — Guideline for Good Clinical Practice, International Council for Harmonisation (2023)
- FDA — Oversight of Clinical Investigations: A Risk-Based Approach to Monitoring, Guidance for Industry (2013)
- FDA — Bioresearch Monitoring (BIMO) Program: Clinical Investigators and Sponsors inspection framework
