## Skeptic Position: Round 4
### The Biomarker Gap: Why All Current Neuroplasticity Enhancement Strategies Remain Clinically Unactionable
As the debate has progressed, all participants have appropriately converged toward more empirically grounded frameworks—from abandoning quantum microtubule hypotheses to acknowledging the "permissive versus causal" ambiguity in metabolic coupling claims. The domain expert's network-level integration model represents a sophisticated synthesis. However, I argue that **the field's most fundamental limitation remains unaddressed: we lack validated biomarkers that can measure neuroplasticity engagement in individual human subjects**. Without such biomarkers, none of the proposed mechanisms—whether NMDA co-agonist enhancement, astrocytic metabolic coupling, or microglial modulation—can be rationally optimized for clinical use.
The NMDA co-agonist pathway exemplifies this problem. Despite decades of research and multiple Phase II/III trials (bitopertin, BI 425809, rapastinel, sodium benzoate), we cannot determine whether these compounds are actually enhancing neuroplasticity in treated patients. We measure cognitive performance on specific tests (MATRICS, WCST, CPT-IP), but these endpoints conflate drug effects on attention, motivation, working memory capacity, and actual learning/encoding processes. A compound could improve test performance through anxiolytic effects or psychomotor stimulation without engaging plasticity mechanisms at all. The fundamental question—"Is this drug making synapses more plastic?"—remains unanswerable with current tools. This explains why BI 425809 failed to separate from placebo in recent schizophrenia trials (NCT03859973): we cannot tell whether the drug hit its intended target in a functionally meaningful way.
The astrocyte-metabolic hypothesis faces even steeper biomarker challenges. Lactate levels in blood or CSF provide only crude proxies for the spatiotemporal dynamics of astrocytic metabolic priming at specific synapses. The spacing effect and metabolic recovery windows are real phenomena, but we cannot measure when an individual human's astrocytes are "primed" versus depleted. This means the therapeutic prescription—"enhance astrocyte metabolism to open plasticity windows"—cannot be operationalized. We lack any validated imaging or fluid biomarker that reflects the relevant biology.
**The Polytherapeutic Trap**: The domain expert's proposal for combination therapies targeting complementary pathways (NMDA enhancement + microglial modulation + metabolic support) is conceptually appealing but practically dangerous without biomarkers. Drug combinations increase the risk of adverse effects multiplicatively while the efficacy signal remains unmeasurable. We cannot determine whether two plasticity-enhancing compounds act synergistically, additively, or antagonistically at the circuit level. Testing all reasonable combinations empirically in clinical trials is prohibitively expensive and ethically questionable without better preclinical markers.
**Proposed Resolution**: I propose that the field should prioritize biomarker development as a prerequisite—not a parallel effort—to mechanism-focused drug development. This includes: (1) developing PET ligands for microglial activation states; (2) establishing MR spectroscopy protocols for synaptic versus astrocytic metabolic signatures; (3) validating EEG/fMRI metrics that specifically track learning-related plasticity rather than general arousal; (4) creating computational models that predict individual responses to plasticity enhancement based on baseline characteristics. Until we can measure what we're trying to modulate, all therapeutic claims remain speculative.
**Supporting Evidence**: The NMDA enhancement literature is littered with compounds that showed robust preclinical efficacy but failed in humans—not because the mechanism was wrong, but because we cannot verify target engagement in the living human brain (PMID: 32302636). Similarly, human neuroimaging studies of astrocyte function remain technically challenging, with most evidence deriving from animal models or indirect metabolic proxies (PMID: 33297842). The microglial field faces analogous challenges—CX3CR1 polymorphisms associate with neurological outcomes (PMID: 28796213), but we cannot image microglial states in living humans to guide therapeutic decisions.
**Confidence Score**: 0.75 — I am highly confident that the biomarker gap is the field's most critical limitation, though I acknowledge this is somewhat self-defeating: my argument suggests we cannot currently resolve which mechanism is "correct" precisely because we lack measurement tools.
**Key Caveats**: My argument could be dismissed as paralyzing—biomarker development is itself challenging and slow. One might counter that we should continue mechanism-focused research while developing biomarkers in parallel. However, the history of neuroplasticity pharmacology suggests this approach has yielded diminishing returns: numerous well-validated mechanisms have failed in clinical translation, indicating the bottleneck is measurement, not mechanism identification.
gue that **the field's most fundamental limitation remains unaddressed: we lack validated biomarkers that can measure neuroplasticity engagement in individual human subjects**. Without such biomarkers, none of the proposed mechanisms—whether NMDA co-agonist enhancement, astrocytic metabolic coupling, or microglial modulation—can be rationally optimized for clinical use.
The NMDA co-agonist pathway exemplifies this problem. Despite decades of research and multiple Phase II/III trials (bitopertin, BI 425809, rapastinel, sodium benzoate), we cannot determine whether these compounds are actually enhancing neuroplasticity in treated patients. We measure cognitive performance on specific tests (MATRICS, WCST, CPT-IP), but these endpoints conflate drug effects on attention, motivation, working memory capacity, and actual learning/encoding processes. A compound could improve test performance through anxiolytic effects or psychomotor stimulation without engaging plasticity mechanisms at all. T