The press release lands on a Monday morning: three new sites activated, enrollment access expanded, patient reach improved. Sponsors send it to investors. CROs forward it to their project teams. Coordinators at the existing sites read it and wonder, quietly, whether anyone modeled what happens to their monitoring attention budget when three new locations go live simultaneously. On May 19, 2026, NKGen Biotech announced the activation of three additional U.S. sites for its Phase 2 study of troculeucel in moderate Alzheimer’s disease — Rutgers Health in Newark, New Jersey; Alzheimer’s DiseaseAlzheimer’s Disease Research Center in Albany, New York; and K2 Medical Research in Orlando, Florida. Three states. Three IRBs or central IRB submissions. Three site initiation visits. Three new coordinator onboarding cycles. Three parallel regulatory binders starting from scratch while the existing sites are already mid-protocol.
This is not a criticism of the decision. Mid-trial site expansion in Alzheimer’s research is often operationally necessary. A 2025 PMC analysis of Alzheimer’s trial recruitment estimated that out of approximately 90 million potentially eligible Americans across preclinical, prodromal, and mild Alzheimer’s populations, only around 11,000 successfully enrolled in clinical trials in 2020. The enrollment gap in this indication is structural, not logistical — which means adding sites is a legitimate response. But the operational cost of doing it mid-trial is real, it compounds, and it is almost always underestimated in the announcement.
The Activation Burden Nobody Budgets For
Adding a site to an ongoing trial requires a protocol amendment or at minimum a substantial modification notification under 21 CFR 312.30, which governs changes during the conduct of a clinical investigation. Expanding the number of investigational sites triggers an IND amendment submission. That clock starts before the new site coordinator has opened a single recruitment log. Before Rutgers Health in Newark sees its first screening patient, the sponsor’s regulatory team is already working a parallel submission track that the original study timeline never accounted for.
The IRB timeline is where mid-trial expansions most reliably bleed days. An NCBI analysis of multisite IRB review for NCI-sponsored studies found that the average time for a multisite healthcare system to approve a protocol ranged considerably by institution type — academic medical centers routinely running longer than community sites. Three new states means three institutional review cycles running in parallel, even when a central IRB is in play, because site-specific agreements and local institutional sign-off layers have not disappeared from the process. Across our network, the practical window from “site identified” to “first patient screened” in a mid-trial addition runs 60 to 90 days at the fast end when everything cooperates. When the site has competing protocol priorities — and a place like K2 Medical Research in Orlando, a high-volume research site, almost certainly does — that window stretches.
NKGen is operating on a constrained capital structure that makes timeline slippage particularly costly. In April 2026, the company amended its secured convertible loan agreement to secure an additional $607,200 in financing. That is not a war chest for aggressive site build-out. Every week of activation delay across three sites is a week of CRA travel budget, site liaison time, and sponsor project management hours that arrives before a single enrollment event justifies the spend. The startup burn on three simultaneous activations — SIV preparation, local lab qualification, IP accountability setup, coordinator GCP training documentation, eISF standing up — runs into real dollars before the protocol even reaches a new patient’s hands.
And the existing sites are watching.
What Spreading the Network Actually Does to Velocity
The assumption embedded in every site expansion announcement is additive: more sites mean more patients. The operational reality is multiplicative on the cost side and often only marginal on the enrollment side, at least in the short term. Sites I work with on neurodegenerative protocols routinely cite the first 90 days post-activation as net-negative for network velocity — the new sites are consuming CRA monitoring bandwidth, generating start-up queries and deviation documentation, and pulling CTM attention precisely when the established sites need intensified recruitment support to hit their own accrual windows.
The math on Alzheimer’s trial enrollment makes this friction especially consequential. Alzheimer’s trials carry some of the highest screen failure rates in the industry — estimates in moderate Alzheimer’s populations frequently run above 50% once cognitive staging, amyloid confirmation, and comorbidity exclusions are applied. If troculeucel‘s protocol requires amyloid PET or CSF confirmation, which is standard in this disease stage, the screening infrastructure at each new site needs to be operationally ready before the enrollment numbers move. Rutgers Health and the Alzheimer’s Disease Research Center in Albany both have relevant infrastructure, but standing up imaging referral pathways and reading agreements is not a Day 1 capability — it is a 30- to 60-day build even for experienced sites.
There is a counterintuitive dynamic that sponsors sometimes miss here: geographic diversification does not automatically address the bottleneck. If the constraint on enrollment velocity is amyloid confirmation throughput or specialist referral pipelines rather than site count, adding sites in Newark, Albany, and Orlando replicates the bottleneck across three new geographies simultaneously. Velocity Clinical Research demonstrated the additive model at its best — enrolling 5,435 participants across 35 sites in a Phase 3 vaccine trial, finishing 11 days ahead of schedule — but that model works in high-throughput indications where screening infrastructure is straightforward. Moderate Alzheimer’s disease is the opposite operating environment.
What Operators Need to Do Before the SIV Calendar Fills Up
For the CRO managing this expansion, the immediate operational priority is a pre-activation resource reconciliation — not a staffing plan on paper, but a documented assessment of CRA capacity against the combined monitoring load of existing plus new sites. Adding three sites mid-trial without a formal monitoring resource analysis is how existing sites start seeing 90-day gaps between visits. Each new site should have a named CRA assigned before the SIV is scheduled, not after the SIV report is filed.
For NKGen’s clinical operations team, the budget pressure creates a specific risk: the temptation to compress site initiation visits or defer coordinator training documentation to save time. ICH E6(R3) Section 5.2 places unambiguous responsibility on the sponsor for ensuring investigator qualification and site adequacy before trial conduct begins. A site that goes live under time pressure and resource constraint is a deviation waiting to be written. The more durable play — even under a $607,200 convertible loan constraint — is a staged activation that sequences new sites by readiness score, not by geography or announcement timing.
For site coordinators at the three new locations: expect the first 60 days to be heavier on documentation and lighter on patients than the activation timeline suggests. Use the pre-enrollment window to audit the screening log template against the protocol’s eligibility criteria before the first patient is referred, not after the first screen failure generates a query. The sites in Alzheimer’s trials that hold their enrollment rates are almost always the ones that resolved their screening workflow ambiguities before the IRB approval arrived, not during it.
The signal to watch over the next 90 days is whether NKGen’s existing sites hold their enrollment pace or flatten while the new sites ramp. If the network’s aggregate accrual rate holds steady through August, the expansion was executed cleanly. If it dips, the cause will almost certainly be found not in patient supply, but in the monitoring calendar.
References
- GlobeNewswire — “NKGen Biotech Expands Clinical Trial Site Network for Alzheimer’s Disease Program Across U.S. to Improve Patient Access and Support Enrollment” (May 19, 2026)
- PMC — “Advancing the science of recruitment for Alzheimer’s clinical trials: Challenges and opportunities”
- FDA — “Changes or Modifications During the Conduct of a Clinical Investigation; Final Guidance for Industry and CDRH Staff”
- NCBI — “Multisite Studies and IRB Review — Contemporary Issues for Protecting Patients in Cancer Research”
- TipRanks — “NKGen Biotech Secures Additional Convertible Financing and Warrants” (April 28, 2026)
- Velocity Clinical Research — Site network enrollment performance data
