How Brain Network Topology Is Reframing CNS Strategy
By Denis Katz, MD, MHA
Founder, Salience Clinical
From Chemistry to Systems
For much of its modern history, psychiatry has approached major depressive disorder (MDD) through the lens of neurochemistry primarily as dysregulation of serotonin, norepinephrine, and dopamine. This paradigm enabled important therapeutic advances, but its impact has plateaued.
A substantial proportion of patients fail to achieve sustained remission. Treatment resistance remains common, and relapse frequently follows initial response. These patterns suggest a structural limitation: incremental adjustments to synaptic signaling are no longer sufficient to meaningfully shift long-term outcomes at scale.
The next evolution in neuropsychiatry will not be defined by new molecules alone.
It will be defined by how we understand and intervene in brain systems.
An Architectural Model of Depression
Depression is increasingly understood not as a localized chemical imbalance, but as a disorder of large-scale brain network organization—how distributed systems coordinate, compete, and switch.
At the center of this model is the interaction among three core networks:
- Salience Network (SN): Identifies relevant internal and external stimuli and orchestrates transitions between other networks
- Central Executive / Frontoparietal Network (CEN/FPN): Supports cognitive control, working memory, and goal-directed behavior
- Default Mode Network (DMN): Governs self-referential thinking, memory, and internally directed attention
The key variable is not simply neural activity, but network dynamics how these systems interact over time and which dominates at critical moments.
In MDD, consistent findings include:
- Reduced functional separation between DMN and executive networks
- Aberrant coupling between salience and both DMN and CEN
- Impaired switching driven by salience network dysfunction
The result is a system that becomes stuck over-engaged in self-referential processing and under-engaged in adaptive cognitive control.
This is not just a chemical problem.
It is a coordination problem.
Why This Changes the Strategic Equation
Most current interventions whether pharmacologic or device-based modulate activity. Few are explicitly designed to restore balance across networks.
- Pharmacotherapies adjust synaptic signaling
- Neuromodulation tools influence regional excitability
But neither has traditionally been optimized to recalibrate network-level relationships: proportionality, timing, and switching behavior.
If depression reflects a failure of network coordination, then symptom improvement without restoring that coordination may explain why relapse remains so common.
The central question becomes:
Can we design therapies that restore network flexibility not just alter neurotransmission?
Toward Network-Level Therapeutics
A more mature model of CNS innovation is beginning to take shape, built around systems-level intervention:
- Connectivity-Based Stratification
Using imaging and electrophysiology to identify biologically distinct subtypes and predict treatment response - Network-Aware Endpoints
Incorporating measures of connectivity, flexibility, and switching into clinical trials alongside traditional scales - Closed-Loop Neuromodulation
Delivering stimulation guided by real-time biomarkers and network state - Plasticity-Focused Biologics
Developing therapies that enhance circuit adaptability and reshape dysfunctional connectivity patterns - Targeted Behavioral Interventions
Designing digital and cognitive approaches that actively rebalance executive control and reduce maladaptive internal focus
This is not a rejection of neurochemistry—it is its integration into a broader systems framework.
Synapses influence circuits. Circuits form networks. Networks define behavior.
Implications for Industry Strategy
1. Drug Development
Therapeutics should be evaluated not only on symptom change, but on their ability to modify connectivity patterns across key networks. Early-phase studies should incorporate these measures prospectively—not as exploratory afterthoughts.
2. Clinical Trial Design
Network biomarkers offer a path to:
- More precise patient selection
- Reduced heterogeneity
- Earlier detection of treatment response
Embedding these signals into inclusion criteria and endpoints can materially improve trial efficiency and interpretability.
3. Medical Affairs Positioning
The narrative of “chemical imbalance” is no longer sufficient. A network-based framework better reflects current science and provides a stronger foundation for integrated treatment approaches across drugs, devices, and digital therapeutics.
4. Durability of Response
Long-term remission may depend less on suppressing symptoms and more on restoring:
- Effective network switching
- Appropriate salience assignment
- Robust executive engagement
Therapies that address these dimensions are more likely to produce sustained benefit.
Operationalizing the Model
Consider a future treatment-resistant depression program:
A rapid-acting pharmacologic agent is combined with connectivity-guided neuromodulation. Patient selection is based on baseline network signatures. Success is defined not only by symptom reduction, but by normalization of interactions among key brain networks.
This is what it means to design for architecture—not just chemistry.
The Competitive Inflection Point
Several forces are converging:
- Network neuroscience has matured into actionable models
- Computational tools can extract predictive signatures from complex brain data
- Neuromodulation platforms are increasingly programmable and adaptive
- Multimodal datasets enable biologically grounded stratification
Organizations that integrate these capabilities into development strategy will define the next phase of CNS innovation.
Those that remain anchored in chemistry alone will face diminishing returns.
The Salience Clinical Perspective
Salience Clinical focuses on translating systems neuroscience into development strategy. We work at the intersection of network biology, clinical design, biomarker integration, and regulatory positioning.
Our objective is not theoretical insight it is execution.
Turning emerging science into decisions that improve probability of success.
Conclusion
The next breakthrough in psychiatry will not come from amplifying existing mechanisms.
It will come from understanding how the brain is organized and learning how to rebalance it.
Not louder interventions.
Better architectures.
