Denis Katz, MD, MHA
Founder, Salience Clinical

Executive Perspective

For decades, psychiatry has approached major depressive disorder (MDD) as a disorder of neurochemistry—an imbalance of serotonin, norepinephrine, or dopamine. This framework has enabled meaningful progress. Yet it has not delivered durable remission at scale.

Approximately one-third of patients remain treatment resistant. Among responders, relapse is common. Incremental advances in synaptic modulation are yielding diminishing returns.

The next transformation in neuropsychiatry will not be chemical.
It will be architectural.

By architecture, we mean the topology, coupling, and dynamic switching behavior of large-scale brain networks that govern cognition, emotion, and behavior.

From Chemistry to Systems Architecture

Advances in functional MRI, multimodal imaging, and computational neuroscience have fundamentally reshaped our understanding of depression. MDD is now consistently linked to dysfunction across three large-scale networks—the triple network:

• Salience Network (SN): Orchestrates detection of relevant stimuli and mediates switching between networks
• Central Executive / Frontoparietal Network (CEN/FPN): Enables cognitive control, working memory, and goal-directed behavior
• Default Mode Network (DMN): Supports self-referential thought and autobiographical processing, but when dysregulated, drives rumination

The critical variable is no longer signal amplitude alone—it is network interaction and dominance over time.

In MDD:

• DMN–CEN anticorrelation is reduced
• Salience network coupling is disrupted
• Network switching is inefficient or impaired

In treatment-resistant populations, these disturbances are amplified, reflecting deeper instability in system-level coordination.

Depression, in this model, is not simply a disorder of neurotransmitters.
It is a disorder of network orchestration.

The Strategic Gap: Modulating Signal vs. Engineering Systems

Current interventions largely operate at the level of signal modulation:

• Pharmacotherapy alters synaptic transmission
• Neuromodulation (rTMS, ECT, DBS) alters regional excitability

But these approaches were not designed to explicitly restore network proportionality, reciprocity, or switching dynamics.

We have optimized signal intensity.
We have not engineered system behavior.

This gap matters. If executive networks remain under-recruited and default/salience networks remain dominant or poorly regulated, short-term symptom relief may occur without durable system correction—contributing to relapse risk.

This reframes the core innovation challenge:

Can we design interventions that restore network balance and flexibility—not just neurotransmission?

Toward Network-Level Therapeutics

The next generation of precision psychiatry will be defined by network-informed intervention design:

• Imaging-based stratification: Defining patient subtypes based on connectivity patterns rather than symptom clusters
• Network-level endpoints: Measuring connectivity, flexibility, and switching alongside clinical outcomes
• Closed-loop neuromodulation: Real-time, biomarker-guided stimulation tailored to individual network states
• Plasticity-driven biologics: Agents that reshape circuit dynamics in addition to synaptic signaling
• Targeted behavioral and digital interventions: Designed to strengthen executive engagement and reduce maladaptive DMN dominance

This is not a departure from neurochemistry—it is its integration into a systems framework.

Synapses enable circuits. Circuits form networks. Networks define behavior.

Implications for Pharma and MedTech

1. Drug Development

Therapeutics must be evaluated not only by symptom reduction, but by their ability to:

• Restore connectivity between networks
• Improve switching efficiency
• Normalize network dominance patterns

Early incorporation of network biomarkers can sharpen mechanism differentiation and responder identification.

2. Clinical Trial Design

Connectivity-based stratification can:

• Reduce biological heterogeneity
• Improve signal detection
• Enable smaller, more efficient trials

Dynamic network metrics may also serve as early indicators of treatment trajectory—well before clinical endpoints mature.

3. Medical Affairs Narrative

The “chemical imbalance” paradigm is no longer sufficient. A network-based narrative:

• Aligns with modern neuroscience
• Enhances scientific credibility
• Supports integrated drug–device positioning

4. Durable Remission Strategy

Sustained outcomes will likely depend on restoring:

• Efficient salience gating
• Flexible network switching
• Robust executive network engagement

Architectural normalization—not transient suppression—will define long-term success.

Operationalizing the Architecture

A next-generation treatment-resistant depression program might:

• Combine a rapid-acting neuroplasticity agent
• Apply connectivity-guided rTMS targeting
• Measure success via both symptom change and restoration of DMN–CEN–SN balance

This is the shift from treating chemistry in isolation to engineering systems-level recovery.

The Competitive Inflection Point

We are entering a decisive phase in CNS innovation:

• Network neuroscience is now actionable
• Computational modeling can predict response from connectivity patterns
• Neuromodulation platforms are increasingly programmable and data-driven
• Multimodal datasets enable biologically grounded stratification

Organizations that embed network architecture into development strategy will define the next decade of therapeutics.

Those that remain chemistry-centric will face diminishing returns.

The Salience Clinical Perspective

Salience Clinical operates at the intersection of systems neuroscience, clinical development, and translational strategy.

Our focus is not theoretical insight—it is applied architecture.

We partner with sponsors to:

• Translate network science into development decisions
• Integrate biomarkers into trial design and endpoints
• Align scientific innovation with regulatory and commercial strategy

Closing Thought

The next transformation in psychiatry will not come from stronger drugs.

It will come from better models of the brain.

From chemistry to circuitry.
From circuitry to networks.
From networks to architecture.

And ultimately from architecture to durable remission.