Measuring the Rewiring: Connectivity Biomarkers for the Next Generation of Psychiatric Drugs

Measuring the Rewiring: Connectivity Biomarkers for the Next Generation of Psychiatric Drugs

Over the past several years, a new generation of psychiatric therapeutics has moved through development. These drugs work through neuroplasticity, a mechanism that calls for next-generation assessment tools. Psychedelics and other neuroplastogens act on specific receptors, then drive changes in synaptic strength that reorganize entire brain circuits. The therapeutic premise is a lasting recovery through durable remodeling of dysfunctional brain networks.

At Lucerum, we have worked directly inside pharma programs facing this problem. The computational neuroscience methods needed to detect circuit-level change exist. What has been missing is a pipeline that turns them into something a clinical trial team can use — fast, standardized, and interpretable enough to inform a go/no-go decision. This is the gap we work in, across every phase of clinical development.

Neuroplastogenic effects are gradual, building over days to weeks, and persistent. A measure that tracks this trajectory in the brain directly adds to clinical outcomes that can, on their own, be unclear.

EEG is one such measure, long used as a functional readout of the brain, historically read through the power of its signal in different frequency bands, and an index of the balance between excitation and inhibition. This has served well for drugs that act quickly on neurotransmission. Neuroplastogens raise a different question: not just how active a region is, but how regions relate to one another over time. Connectivity measures are built on the signal phase relationship. Connectivity can change substantially, reflecting real reorganization between regions, while signal power can remain unchanged. Missing this metric means missing the earliest, and often most direct, evidence of neuroplastic activity.

Neuroplasticity and Circuit Remodeling

To see why connectivity is the measure that matters, it helps to step back to what neuroplasticity actually changes. Neuroplasticity is the brain's capacity to change the strength of its synaptic connections in response to experience or pharmacological intervention. When enough synapses across a region change together, the circuits built from them change too - a shift in how an entire population of connected regions cooperates.

We can measure this with functional connectivity metrics, which show the degree to which activity in different brain regions is synchronized, or share directional information flow over time. It is a readout of how the brain's networks are organized at a given moment, and how that organization shifts.

Drug Response Timing And Connectivity

Connectivity change can be detected on two useful timescales. Acute changes, visible within hours of drug administration, provide an early signal of target engagement, well before a clinical outcome could reasonably be expected. Changes that persist over days to weeks indicate that a new network configuration has stabilized. For early-phase trials, where the central question is whether a mechanism is engaging the brain at all, this acute signal is valuable: it gives a readout of biological effect before a program commits to a large efficacy study.

Spatial Scale of Neuroplastogen Effects

Beyond timing, connectivity can also be examined at two spatial levels. At the most granular level, pairwise connections between individual regions can be aggregated into a global density of rewiring, a diffuse, whole-brain signature and a meaningful first readout of a neuroplastogenic effect. Networks are the next level up: ensembles of regions that cooperate to support a specific function, such as emotional regulation. Tracking connectivity within one of these networks shows how rewiring is affecting a functionally specialized ensemble, tying back to a plausible clinical function.

Measuring Neuroplastic Effects in Healthy Volunteers vs. Patients

There is also a meaningful distinction between how this remodeling is expected to unfold in a healthy brain versus a disordered one. In a healthy brain, initial neuroplastic remodeling is expected to occur somewhat stochastically, but the specialized roles of established cognitive ensembles are preserved throughout, so that widespread reorganization does not come at the cost of specialized network function. In a brain affected by psychiatric or neurodegenerative illness, where functional dysconnectivity is often part of the underlying pathology, the same kind of neuroplastic intervention is expected to drive affected networks toward a more normalized, stable configuration, restoring functional stability that had been compromised. This distinction gives connectivity data from healthy volunteers and from patients different, complementary interpretations within the same drug program.

Connectivity change following administration of neuroplastogenic drugs has been demonstrated across multiple neuroimaging modalities, including fMRI and EEG, with EEG performing comparably to fMRI in detecting these effects, as work comparing simultaneously recorded EEG and fMRI under pharmacological challenge has shown. That convergence across modalities is part of why we consider connectivity a credible, generalizable biomarker.

Source-Localized qEEG Captures Connectivity

Measuring this rewiring requires turning the electrical signal recorded at the scalp into an estimate of activity inside the brain. Quantitative EEG tomography reconstructs current source density at each location across the cortex, an estimate of the underlying neuronal generators.

The magnitude of the reconstructed current source density gives a localized estimate of neuronal activation, the same excitation and inhibition signal historically read from EEG spectral power, now assigned to a specific brain region at high resolution. The phase relationship between the reconstructed time series at different locations gives functional connectivity, the coordination between regions that reflects circuit-level rewiring. Both measures come from the same EEG, so a single acquisition yields localized activation and connectivity together, region by region.

Clinical Trial Feasibility

This combination makes source-localized qEEG well suited to clinical neuroplasticity assessment. It is repeatable, portable, and sensitive enough to detect network reorganization that a conventional EEG readout would miss. Operationally, it adds little burden: recordings are short and can be run anywhere EEG is already collected. These are practical requirements for the multi-timepoint, dose-tracking design that early-phase neuroplastogen trials demand.

What Connectivity Metric to Use for Drug Efficacy?

Reliable source localization solves half of the problem; choosing the right way to report connectivity change is the other half. We have run into the same problem across program after program. An academic question about a specific circuit can be answered by isolating specific brain areas and proving a hypothesis with multiple methods; a drug development program cannot work that way. It needs a reproducible approach to neuroplastogen efficacy that can be computed quickly, that carries enough temporal and spatial specificity to identify whether the effect exists globally, and points to a plausible mechanism.

Current practice is fragmented across a wide range of connectivity metrics, applied inconsistently between studies, which has kept this class of biomarker largely confined to research settings.

Our answer has been to build composite connectivity metrics for this purpose. In addtion to reporting all connectivity changes in detail, we developed metrics that summarize, for each brain region, the density of rewired.

The result is a spatially resolved index of connectivity change across the whole brain that can be visualized as a three-dimensional map, tracked across drug administration timepoints, and compared across dose levels. It gives a trial team a single, interpretable readout of where and how strongly a drug is engaging brain networks, without parsing thousands of individual connections.

A Case in Point

One recent program illustrates how this comes together in practice. The drug, developed for major depressive disorder, had limited efficacy endpoints available from healthy volunteer studies, a common problem for centrally acting neuroplastogen compounds with no symptom to measure in an unaffected population. Source-localized qEEG connectivity closed that gap. In healthy volunteers, the drug produced clear, global changes in connectivity, with no corresponding change in conventional spectral power measures, confirming that a real neuroplastogenic effect was occurring even though standard EEG readouts showed nothing.

In patients, the same approach showed connectivity changes concentrated in networks associated with emotional regulation. These changes were dose-dependent and correlated with improvements in clinical outcome scores, consistent with a mechanism that restores synaptic plasticity and network stability in circuits known to be compromised in mood disorders.

This is the general approach we bring to a program: establish a connectome-level effect, localize it to specific networks, compare the healthy volunteer signal against the patient signal, quantify its extent and its dose relationship, and correlate it with clinical outcome. Each step builds toward a single, defensible answer to the question a clinical trial team needs answered: is this drug neuroplastic, how strong and where and how is it acting?

Connectome Biomarkers – Next Steps

Objectively quantifying drug-induced neural plasticity is a shift in how mechanism-based psychiatric drugs can be developed. Source-localized EEG connectivity offers the most practical near-term route to getting there: non-invasive, repeatable across a full trial timeline, and sensitive to network reorganization.

We built our approach on computational neuroscience methods that are already well established and accepted, then did the work of standardizing them into a consistent, repeatable approach that travels from one program to the next.

Regulatory bodies are moving quickly on new neurocomputational methods as well, and as this generation of neuroplastic drugs continues to mature, connectivity-based evidence is likely to become a standard part of how they are developed.

 REFERENCES

1.           O'Donnell P, Etkin A, Manji H. Brain Neuroplasticity Mechanisms in Psychiatric Illnesses and in the Development of Novel Treatments. American Journal of Psychiatry. 2026 Feb;183:98–107.

2.           Bulek D, BaDour S. From monoamine deficits to multiscale plasticity: twenty-five years of ketamine and the neurophysiology of depression. J Neurophysiol. 2026 Feb 13;135(2):406–13.

3.           Bhattacharjee M, Vlachos I, Kathpalia A, Hlinka J, Brunovský M, Barés M, et al. Causality analysis in major depressive disorder for early prediction of treatment outcomes with pharmacological and neuromodulation therapies. medRxiv. 2025.

4.           Forsyth A, McMillan R, Campbell D, Malpas G, Maxwell E, Sleigh J, et al. Modulation of simultaneously collected hemodynamic and electrophysiological functional connectivity by ketamine and midazolam. Human Brain Mapping. 2020;41(6):1472–94.

5.           Pascual-Marqui RD, Michel CM, Lehmann D. Low resolution electromagnetic tomography: a new method for localizing electrical activity in the brain. International Journal of Psychophysiology. 1994 Oct 1;18(1):49–65.

6.           Palmero-Soler E, Dolan K, Hadamschek V, Tass PA. swLORETA: a novel approach to robust source localization and synchronization tomography. Phys Med Biol. 2007 Mar;52(7):1783.

7.           Nolte G, Müller KR. Localizing and Estimating Causal Relations of Interacting Brain Rhythms. Front Hum Neurosci. 2010;4.

8.           Gomez C, Ruiz-Gomez SJ, Poza J, Maturana-Candelas A, Nunez P, Pinto N, et al. Assessment of EEG Connectivity Patterns in Mild Cognitive Impairment Using Phase Slope Index. 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC). 2018:263–6.