A site coordinator opens her study binder on a Monday morning and finds a protocol amendment notification. Nothing unusual — except this one does not change a visit schedule or a lab panel. It closes an entire treatment arm. The patients currently enrolled in that arm continue. The screen failure patients who were about to be randomized into it do not. Her CTMS still shows the arm as open. Her eISF has not been updated. Her PI has three questions she cannot answer yet. And the sponsor’s monitoring visit is in eleven days.
That scenario is coming to more neurology sites than most operational teams realize. The OCTOPUS trial — the Optimal Clinical Trials Platform for Progressive Multiple Sclerosis — is a multi-arm, multi-stage (MAMS) platform trial now in Stage 2, recruiting an additional 825 participants after completing Stage 1 with 375. Recruitment began in January 2023, and as of February 2026 the trial had exceeded 700 participants enrolled across a network targeting up to 30 UK sites. What OCTOPUS represents operationally goes well beyond its science: it is a preview of the site management complexity that follows when platform trial architecture, common in oncology, arrives in neurology.
The Operational Difference Nobody Budgets For
MAMS trials work by running multiple treatment arms against a shared control, with interim analyses determining which arms advance to the next stage and which are dropped. The statistical logic is sound. The site-level consequence is that your protocol document is a living object, and your staff must manage a trial that may not look the same in month 18 as it did at the SIV. That is a fundamentally different operational contract from what most site teams signed up for.
Consider what arm closure actually requires at the site level: a protocol deviation risk assessment, an IRB/IEC notification or amendment depending on the nature of the change, updated informed consent procedures if the arm change affects what subjects are consenting to, CTMS reconfiguration, revised eligibility screening logs, and retraining documentation for every coordinator who touches randomization. For a single-arm, fixed-design Phase 3 trial, those tasks occur once at amendment. In a MAMS platform trial, they occur every time the independent data monitoring committee issues a recommendation the sponsor acts on. Research on late-phase MAMS trials shows that out of 62 such studies, adoption in neurology significantly trails oncology, which means neurology site teams are building these operational muscles for the first time while running a live Phase 3 study.
The ICH E6(R3) framework places responsibility on sites to maintain inspection-ready documentation throughout the trial lifecycle, not at closeout. In a MAMS design, that means every arm transition event generates a documentation obligation: a dated record of when the arm closed, which enrolled subjects were affected, what communications went to the IRB, and what the PI signed off on. Sites that manage this through ad hoc emails and shared drive folders will have a painful experience when a BIMO inspection auditor asks to reconstruct the timeline of arm closure decisions.
The budget reality compounds the documentation burden. Most site budgets for platform trials are structured around a fixed per-visit, per-patient fee schedule negotiated at contract execution. When an arm closes mid-trial, the screen failure patients who were queued for that arm represent unrecovered costs: coordinator screening time, medical records review, eligibility confirmation, potentially a screening visit. Screen failure reimbursement clauses in standard site contracts rarely account for screen failures generated by arm closure rather than protocol ineligibility. That distinction matters when you are trying to reconcile a budget that assumed 18 months of a four-arm trial and delivered 14 months of three arms and one arm running at reduced capacity.
What Sponsors Miss When They Build the Governance Model
Platform trial governance — the charter, the DMC operating procedures, the statistical analysis plan amendments — is typically developed at the sponsor level, often in collaboration with an academic steering committee. What does not always make it into those governance documents is an operational annex that defines site-level obligations and timelines when the governance model makes a decision. OCTOPUS’s published protocol describes a statistically robust MAMS core design with prospective arm addition capability. What site operational teams need alongside that protocol is a clear SOP defining the lag time between an IDMC recommendation, a sponsor decision, IRB notification, and the moment a site coordinator can stop screening into a closing arm.
That lag time matters because it determines protocol deviation risk. If a site screens and randomizes a patient into an arm the sponsor has internally decided to close, but has not yet formally notified sites about, who owns the deviation? FDA’s 2019 guidance on adaptive designs for clinical trials requires that adaptive elements be prospectively planned and transparently documented, which puts the burden on sponsors to define communication timelines. But the guidance does not specify what a site’s grace period looks like operationally. That gap lives in your contract, and most contracts written for fixed-design trials do not address it.
The decentralized trial layer adds another variable. Neurology trials in particular have seen growing pressure to incorporate hybrid DCT elements: remote neurological assessments, ePRO-based symptom tracking, home nursing for infusions or sample collection. Research on DCT integration in MS trials shows that average patient dropout rates across clinical trials run at 30%, with remote participation models improving retention in neurological studies by reducing travel burden. In a MAMS design, a DCT component that works for Arm A may require different vendor configuration for Arm B, and may become unnecessary entirely when Arm C closes. The result is a vendor oversight matrix that shifts mid-trial, with site-level consequences for training, accountability logs, and TMF completeness.
Sites I work with that have run adaptive oncology trials describe the monitoring visit rhythm as fundamentally different from fixed-design studies. Risk-based monitoring plans built at activation become outdated when the risk profile of the trial changes with arm transitions. A CRA arriving for a monitoring visit six months after an arm closure should be working from an updated monitoring plan that reflects the current arms — but in practice, many sponsors do not update RBM triggers to account for the operational complexity of arm transition periods, leaving CRAs reviewing data against criteria that no longer match site reality.
What Sites and Sponsors Need to Change Before Activation
For sites being approached about platform trial participation, the pre-SIV feasibility conversation needs to include questions that most site teams are not trained to ask. What is the sponsor’s defined timeline from IDMC recommendation to site notification? What is the protocol deviation policy for patients screened during a transition window? What does the arm closure amendment process look like for central versus local IRB sites, and has the sponsor built those IRB notification timelines into the overall governance plan? What CTMS configuration will be required when an arm closes, and who is responsible for executing it — site IT, sponsor CRA, or a vendor? Getting answers to those questions during feasibility is worth more than any budget negotiation, because it surfaces the operational obligations before you have committed staff hours to a study that may restructure itself three times over its recruitment window.
For sponsors designing platform trials in neurology, the single highest-return investment at the site level is an operational annex to the protocol: a plain-language document, updated at each arm transition, that tells site coordinators exactly what changes, exactly when, and exactly who to call. OCTOPUS is targeting 30 sites. At any given site, the coordinator managing this trial is also managing two or three other protocols. The operational annex is not a courtesy — it is the difference between a site that executes arm transitions cleanly and one that generates a cluster of deviations you will be writing CAPA responses for in months 22 through 28.
The MAMS model will keep arriving in neurology. The sites that build the operational infrastructure now — the contract clauses, the arm-transition SOPs, the IRB amendment pre-templates, the budget reconciliation language for screen failures generated by arm closure — are the ones sponsors will call first when the next platform trial opens. The ones that do not will spend the back half of every platform study doing corrective work that should have been structural from the start.
References
- medRxiv Neurology — “Optimal Clinical Trials Platform for Progressive Multiple Sclerosis (OCTOPUS): protocol for an international, multi-arm, multi-stage, platform, randomized controlled, double-blind, phase 3 clinical trial”
- MS Trust — “OCTOPUS Trial: Progressive MS enrollment milestones and site network”
- PMC — “Multi-arm multi-stage (MAMS) trial adoption in late-phase oncology and neurology”
- FDA — “Adaptive Designs for Clinical Trials of Drugs and Biologics: Guidance for Industry”
- PMC — “Decentralized clinical trial integration in MS and neurology: dropout rates and retention outcomes”
