The Therapeutic Landscape of Collapse

Benjamin Aaron Gustafsson AdultCognitiveDisease.com

Introduction: The Rationale for Phase-Specific Therapeutics

Part I: Phase I — The Silent Brainstem Erosion (Age 20–50)

Phase I Drugs

Olaparib (PARP Inhibitor)
Veliparib (PARP Inhibitor)
Rapamycin / Sirolimus (mTOR Inhibitor)
Urolithin A (Mitophagy Inducer)
Selegiline / Deprenyl (MAO-B Inhibitor)
Rasagiline (MAO-B Inhibitor)
Deferiprone (Iron Chelator)
ISRIB (Integrated Stress Response Inhibitor)
Edaravone (Free Radical Scavenger)
Idebenone (CoQ10 Analogue) Phase I Supplements
Nicotinamide Riboside (NR)
Nicotinamide Mononucleotide (NMN)
Niacin (Vitamin B³)
Coenzyme Q¹⁰ (Ubiquinol)
PQQ (Pyrroloquinoline Quinone)
Alpha-Lipoic Acid
Sulforaphane
Spermidine
Creatine
Astaxanthin

Part II: Phase II — The Hippocampal Bridgehead (Age 50–65)

Phase II Drugs Phase II Supplements

Part III: Phase III — The E/I Collapse (Age 65–75)

Phase III Drugs Phase III Supplements

Synthesis: Cross-Phase Interactions and Combination Strategies

References

Introduction

The Rationale for Phase-Specific

Therapeutics

Why One Drug Cannot Treat Three Diseases Alzheimer's disease has resisted pharmacological intervention for over three decades, with a failure rate in clinical trials exceeding 99 percent. The standard explanation for this extraordinary attrition has been that trials enroll patients too late, after neurodegeneration has become irreversible. While this explanation carries some truth, it obscures a more fundamental problem: the field has been treating Alzheimer's disease as a single entity when it is, in mechanistic terms, at least three sequential pathologies. A drug that targets amyloid-beta oligomers may be entirely irrelevant to the mitochondrial DNA damage accumulating in the locus coeruleus of a thirty-year-old, just as a mitophagy inducer would do nothing to rescue the excitatory-inhibitory circuit collapse unfolding in the cortex of a seventy-year-old. The Spectrum of Collapse framework argues that effective therapeutic strategy requires matching interventions to the correct phase of the neurodegenerative cascade. The Spectrum of Collapse, developed as part of the Organic Network Synthesis (ONS) methodology, proposes that Alzheimer's disease unfolds across three biologically distinct phases, each with its own dominant molecular machinery, its own vulnerable cell populations, and its own therapeutic windows. Phase I, the Silent Brainstem Erosion, spans roughly ages 20 to 50 and is characterized by the slow degradation of mitochondrial quality control in monoaminergic projection neurons. Phase II, the Hippocampal Bridgehead, covers approximately ages 50 to 65 and is defined by the convergence of lysosomal failure, tau propagation, ferroptosis in myelinating oligodendrocytes, and the phenotypic transition of microglia from homeostatic surveillants to disease-associated states. Phase III, the Excitatory/Inhibitory Collapse, unfolds from roughly age 65 to 75 and involves the destruction of perineuronal nets, the death of parvalbumin-positive fast-spiking interneurons, complement-mediated synapse elimination, and the inflammatory amplification loops that drive the final cortical collapse. This tripartite architecture has profound implications for drug development. If Alzheimer's disease is indeed three sequential pathologies rather than one, then a single therapeutic agent—no matter how potent—cannot address the full disease trajectory. Instead, we need a pharmacological strategy that evolves with the disease, deploying different agents at different biological moments. A PARP inhibitor that preserves NAD+ pools in the locus coeruleus during Phase I serves an entirely different biological purpose than a ferroptosis inhibitor that protects oligodendrocytes during Phase II, or an MMP-9 inhibitor that preserves perineuronal nets during Phase III. The therapeutic landscape of collapse demands what we might call 'temporal pharmacology'—the art of deploying the right drug at the right biological moment.

The Failure of Monolithic Therapeutics

The history of Alzheimer's drug development is a graveyard of monolithic strategies. The cholinesterase inhibitors—donepezil, rivastigmine, galantamine—were developed on the assumption that replacing lost acetylcholine would restore cognitive function. They provide modest symptomatic relief for six to twelve months, but they do nothing to address the upstream mechanisms that killed the cholinergic neurons in the first place. The amyloid hypothesis drove an entire generation of drugs—bapineuzumab, solanezumab, aducanumab, lecanemab, donanemab—aimed at clearing amyloid-beta plaques or oligomers. Even the most successful of these, lecanemab and donanemab, achieve plaque clearance yet deliver clinical benefits that are statistically significant but clinically marginal, slowing decline by roughly 25 to 35 percent over 18 months. The reason is architectural: by the time amyloid plaques are detectable on PET imaging, the upstream mitochondrial and lysosomal failures that set the cascade in motion have been operating for decades. The tau-targeting strategies—antisense oligonucleotides, tau immunotherapies, kinase inhibitors—face a similar temporal mismatch. Tau pathology in the entorhinal cortex and hippocampus is a Phase II phenomenon, a consequence of lysosomal failure and impaired autophagy rather than a root cause. Targeting tau without addressing the lysosomal dysfunction that drives its accumulation is analogous to mopping a floor while the faucet is still running. The repeated failure of tau vaccines in clinical trials—AADvac1, semorinemab, zagotenemab—reflects this fundamental misunderstanding of causal hierarchy. Even the NMDA receptor antagonist memantine, which addresses excitotoxicity in advanced disease, arrives too late in the cascade to prevent the upstream circuit damage that drives the excitatory-inhibitory imbalance. What the field lacks is not better drugs per se, but a coherent framework for matching existing and emerging drugs to the correct biological phase of the disease. Many of the compounds discussed in this document have already been tested in clinical trials—and many have failed. But they may have failed not because they are ineffective molecules, but because they were tested in the wrong patients at the wrong time. A PARP inhibitor tested in patients with moderate-to-severe Alzheimer's dementia is being deployed in Phase III of the cascade, when Phase I—the phase where PARP hyperactivation drives NAD+ depletion—ended decades earlier. The Spectrum of Collapse framework provides the biological logic for retesting many of these compounds in the correct temporal context.

The Architecture of This Document

This document systematically maps sixty pharmacological and nutraceutical agents to the three phases of the Spectrum of Collapse. For each phase, we present ten drugs (agents requiring a prescription or currently in clinical development) and ten supplements (compounds available over the counter or as dietary supplements). Each entry follows a standard format: mechanism of action, rationale within the specific phase, clinical trial history, and current status with known limitations. The goal is not to advocate for any particular therapeutic strategy, but to create a comprehensive reference that maps the existing pharmacological landscape onto the mechanistic architecture of the disease. Several important caveats apply. First, many of these compounds have pleiotropic effects that cross phase boundaries—rapamycin, for example, affects mitophagy (Phase I), autophagy and lysosomal function (Phase II), and neuroinflammation (Phase III). We have assigned each compound to its primary phase of relevance, but we note cross-phase effects where they are significant. Second, the clinical trial data cited here spans a wide range of quality, from large randomized controlled trials to small open-label pilot studies. We report the data as it exists, noting limitations in study design and statistical power where appropriate. Third, this document is a research reference, not clinical guidance. None of the phase-specific strategies outlined here have been validated in trials designed around the Spectrum of Collapse framework. The mapping is theoretical, grounded in mechanism, and intended to generate hypotheses for future investigation. The ultimate aspiration of this work is to shift the conversation in Alzheimer's drug development from 'which target?' to 'which target, when?' The history of cancer therapeutics offers an instructive parallel: the transformation from single-agent chemotherapy to stage-specific combination regimens—neoadjuvant, adjuvant, maintenance—was the breakthrough that turned many cancers from death sentences into manageable conditions. Alzheimer's disease may require a similar conceptual revolution. The drugs and supplements catalogued here are the raw materials for that revolution. What has been missing is the temporal architecture that tells us when to deploy them. The Spectrum of Collapse provides that architecture.

Phase I Overview: The Silent Brainstem Erosion

Phase I of the Spectrum of Collapse unfolds silently across the first half of the human lifespan, from approximately age 20 to age 50. Its epicenter is the brainstem, specifically the locus coeruleus and the dorsal raphe nucleus—the noradrenergic and serotonergic projection systems that innervate virtually the entire cortex. These neurons are uniquely vulnerable to mitochondrial stress for several reasons: they are unmyelinated or thinly myelinated, they maintain long, thin axons that reach broadly across the forebrain, they fire tonically without rest, and they rely on monoamine oxidase metabolism that generates hydrogen peroxide as a byproduct. The result is a cell population that experiences chronic oxidative stress from the moment it differentiates, and whose mitochondria accumulate damage relentlessly across the lifespan. The molecular cascade of Phase I involves several interconnected pathways. Oxidative DNA damage—particularly 8-oxoguanine lesions in mitochondrial DNA—activates poly(ADP-ribose) polymerase 1 (PARP-1), which consumes NAD+ to synthesize poly(ADP-ribose) chains. As PARP-1 hyperactivation depletes the cellular NAD+ pool, sirtuin-mediated mitochondrial quality control fails, mitophagy (the selective degradation of damaged mitochondria) becomes impaired, and a vicious cycle ensues in which damaged mitochondria generate more reactive oxygen species, causing more DNA damage, more PARP-1 activation, and further NAD+ depletion. In its most extreme form, this cycle triggers parthanatos—a PARP-1-dependent cell death pathway distinct from apoptosis—or SARM1-mediated Wallerian degeneration of the axon. The therapeutic targets in Phase I are therefore clear: preserve NAD+ pools, inhibit excessive PARP-1 activity, enhance mitophagy, scavenge reactive oxygen species, and protect mitochondrial electron transport chain function. The drugs and supplements mapped to Phase I in this document converge on these targets. PARP inhibitors (olaparib, veliparib) directly prevent NAD+ depletion by blocking the enzyme that consumes it. mTOR inhibitors (rapamycin) and mitophagy inducers (urolithin A) enhance the clearance of damaged mitochondria. MAO-B inhibitors (selegiline, rasagiline) reduce the oxidative burden generated by monoamine metabolism. Iron chelators (deferiprone) prevent Fenton chemistry that amplifies oxidative damage. Integrated stress response inhibitors (ISRIB) restore translational homeostasis in stressed neurons. Free radical scavengers (edaravone) and electron transport chain supports (idebenone) directly address the downstream consequences of mitochondrial dysfunction. On the supplement side, the NAD+ precursors (NR, NMN, niacin) replenish the depleted cofactor pool, while mitochondrial cofactors (CoQ, PQQ, alpha-lipoic acid) support electron transport ¹⁰ chain function and redox balance. Autophagy inducers (sulforaphane, spermidine) promote cellular housekeeping, while metabolic buffers (creatine) and lipophilic antioxidants (astaxanthin) provide additional neuroprotection. Part I

Phase I Drugs: Targeting the Silent

Brainstem Erosion

Pharmacological Agents for the Mitochondrial Crisis (Age 20–50)

Olaparib (Lynparza)

Phase I Supplements: Nutraceutical

Support for the Mitochondrial Crisis

Over-the-Counter Agents for NAD+ Repletion, Mitophagy, and Redox Balance

Nicotinamide Riboside (NR)

The Phase II rationale for Bacopa monnieri integrates its cholinergic enhancement with its anti-aggregation and anti-inflammatory activities. The cholinergic trophic withdrawal of Phase II involves both the degeneration of basal forebrain projection neurons and the reduced efficiency of cholinergic transmission at hippocampal synapses. Bacopa's dual action of inhibiting acetylcholinesterase (preserving available acetylcholine) and upregulating choline acetyltransferase (enhancing acetylcholine synthesis) addresses both aspects of this cholinergic deficit. Unlike pharmaceutical cholinesterase inhibitors, Bacopa's additional effects on A-beta aggregation and neuroinflammation provide multi-mechanism support that extends beyond symptomatic cholinergic enhancement to address the pathological processes driving Phase II progression. Within the Spectrum of Collapse framework, Bacopa's capacity to reduce A-beta aggregation is particularly relevant during Phase II when the PANTHOS mechanism is actively driven by intraneuronal A-beta accumulation. By reducing the formation of A-beta oligomers and protofibrils in the extracellular space, Bacopa may diminish the endosomal uptake of toxic aggregates that overwhelms the autophagy-lysosomal pathway in CA1 neurons. The LRP1 upregulation further supports A-beta clearance across the blood-brain barrier, reducing the overall brain A-beta burden during the Phase II window. The antioxidant induction through Nrf2 activation provides additional protection against the oxidative stress that drives ferroptotic oligodendrocyte death, positioning Bacopa as a comprehensive Phase II nutraceutical addressing cholinergic, proteinopathic, inflammatory, and oxidative mechanisms simultaneously.

Clinical Trial History

Bacopa monnieri has been evaluated in numerous clinical trials of cognitive function, primarily in healthy older adults and populations with age-associated memory impairment. A systematic review and meta-analysis by Kongkeaw et al. (2014) identified nine randomized controlled trials enrolling a total of 518 participants and concluded that Bacopa significantly improved attention, cognitive processing speed, and working memory compared to placebo (p < 0.001 for attention speed and p < 0.01 for cognitive processing). The most commonly tested formulation was a standardized extract containing 55 percent bacosides at doses of 300-450 mg daily for 12 weeks. The landmark study by Stough et al. (2001, published in Psychopharmacology) demonstrated significant improvements in visual information processing, learning rate, and memory consolidation with 300 mg daily for 12 weeks in healthy adults aged 40-65. Clinical trials specifically in Alzheimer's disease populations are more limited. A randomized, double-blind, placebo-controlled trial by Goswami et al. (2011) evaluated Bacopa 300 mg daily for six months in 39 patients with AD and reported significant improvements on MMSE and cognitive subscales compared to placebo. A larger trial by Sadhu et al. (2014) evaluated a Bacopa-containing polyherbal formulation in AD patients with improvements on cognitive assessments. These studies, while encouraging, are limited by small sample sizes and short durations. The BACOPA-CAIDE trial, evaluating Bacopa for dementia risk reduction in midlife adults with cardiovascular risk factors, represents the most rigorous ongoing investigation of Bacopa in an at-risk population. No Phase III Alzheimer's trial has been conducted.

Current Status and Limitations

Bacopa monnieri is widely available as a dietary supplement in standardized extract form, typically containing 20-55 percent bacosides. The herb has been used for centuries in Ayurvedic medicine and is generally well tolerated. The most common adverse effects include gastrointestinal symptoms (nausea, abdominal cramping, diarrhea), dry mouth, and fatigue. Bacopa may interact with thyroid medications through its effects on thyroid hormone synthesis, and it may potentiate the effects of cholinergic and serotonergic medications. The recommended dose for cognitive enhancement based on clinical trial evidence is 300-450 mg daily of standardized extract. The principal limitations of Bacopa within the Phase II framework include the lack of clinical trials specifically designed to evaluate disease modification in Alzheimer's disease, the short duration of existing studies relative to the Phase II timeframe, and the heterogeneity of commercial preparations. The in vitro anti-aggregation and LRP1 upregulation activities have not been confirmed in vivo in human subjects, and the brain bioavailability of bacosides following oral administration is insufficiently characterized. While the clinical evidence consistently supports cognitive enhancement in healthy older adults, the translation of these findings to disease-specific populations at defined phases of the Spectrum of Collapse requires targeted clinical trials with appropriate biomarker endpoints. Bacopa's greatest potential may lie in combination with other Phase II nutraceuticals, where its cholinergic enhancement complements the anti-ferroptotic effects of selenium and NAC and the neurotrophic effects of lion's mane. Part III

NAD+ depletion is the most pervasive metabolic deficit in the Spectrum of Collapse, driven by different mechanisms in each phase but converging on the same outcome: insufficient cofactor availability for sirtuin-mediated quality control, mitochondrial electron transport, and neuronal energy metabolism. In Phase I, NAD+ is consumed by hyperactivated PARP-1 responding to oxidative DNA damage. In Phase II, NAD+ depletion is compounded by reduced biosynthesis through the NAMPT pathway and increased consumption by activated immune cells. In Phase III, microglial CD38 upregulation creates an inflammatory NAD+ sink that drains the cofactor from the extracellular space and from surrounding neurons. A rational cross-phase NAD+ strategy would combine supply-side agents (NAD+ precursors: NR, NMN, or niacin from Phase I) with demand-side agents (PARP inhibitors from Phase I; CD38 inhibitors like apigenin from Phase III) to simultaneously increase production and reduce consumption. This bidirectional strategy has been validated preclinically by Camacho-Pereira et al. (2016), who showed that combining NR supplementation with CD38 inhibition produced synergistic increases in tissue NAD+ levels exceeding those achieved by either intervention alone. The clinical implementation of a cross-phase NAD+ strategy would involve initiating NAD+ precursor supplementation in Phase I (or even before Phase I, as a primary prevention measure), adding PARP-inhibitory compounds as biomarkers of PARP hyperactivation appear (elevated urinary PAR, declining blood NAD+ levels), and incorporating CD38 inhibitors (apigenin or pharmaceutical CD38 inhibitors in development) when markers of neuroinflammation signal the transition to Phase III. This phased approach to a single metabolic target—deploying different agents at different biological moments to address the dominant NAD+-consuming mechanism of each phase—exemplifies the temporal pharmacology that the Spectrum of Collapse framework demands.

The Anti-Inflammatory Gradient: Escalating With Disease

Progression

Neuroinflammation is present in all three phases of the Spectrum of Collapse, but its character changes dramatically across the disease trajectory. In Phase I, inflammation is largely cell-autonomous: individual neurons experience oxidative stress and activate intracellular stress responses (integrated stress response, unfolded protein response) that do not initially involve immune cells. In Phase II, microglia detect neuronal distress signals (released ATP, fractalkine shedding, complement opsonization) and transition from homeostatic surveillance to disease-associated phenotypes, producing moderate levels of pro-inflammatory cytokines that contribute to the microenvironment of neuronal vulnerability. In Phase III, fully post-homeostatic microglia engage in active destruction of neural infrastructure—degrading PNNs, stripping synapses, and producing NLRP3-driven IL-1β cascades that amplify the circuit collapse. A rational anti-inflammatory strategy would escalate intervention intensity to match the escalating inflammatory pathology. In Phase I, mild antioxidants and NRF2 activators (sulforaphane, astaxanthin, alpha-lipoic acid) may be sufficient to manage cell-autonomous oxidative stress without suppressing beneficial inflammatory responses. In Phase II, microglial modulators that promote homeostatic phenotype maintenance—targeting TREM2 signaling, CSF1R signaling, or P2Y12 expression—become appropriate as microglia begin to shift toward disease-associated states. In Phase III, potent anti-inflammatory agents targeting specific amplification mechanisms—NLRP3 inhibitors (MCC950/Inzomelid), anti-IL-1β antibodies (canakinumab), JAK inhibitors (baricitinib), and complement blockers (pegcetacoplan)—are warranted by the severity and specificity of the inflammatory assault on cortical circuits. The key principle is proportionality: immunosuppression carries infection risks that are disproportionate to the mild inflammation of Phase I but entirely justified by the destructive inflammation of Phase III. The supplement gradient follows the same logic. Phase I supplements with anti-inflammatory properties (alpha-lipoic acid, sulforaphane, astaxanthin) are mild antioxidants with minimal immunosuppressive risk. Phase II supplements (curcumin, omega-3 fatty acids, resveratrol) have moderate anti-inflammatory effects that modulate microglial phenotype without suppressing immune function. Phase III supplements (melatonin, PEA, luteolin, apigenin) have more targeted anti-inflammatory mechanisms—NLRP3 inhibition, NF-κB suppression, CD38 inhibition—that address the specific inflammatory machinery of the E/I collapse. This gradient from mild to potent, from broad to targeted, represents a phase-conscious approach to neuroinflammation management that is absent from current clinical practice.

The PNN Protection Cluster: A Phase III Priority

Among the Phase III targets, PNN protection represents the highest-priority combination strategy because PNN degradation is the initiating event of the E/I collapse cascade. If PNNs can be preserved, PV+ interneurons retain their protective ensheathment, and the downstream excitotoxic, ferroptotic, and phagoptotic death signals are significantly attenuated. The PNN protection cluster combines agents targeting different aspects of PNN maintenance: MMP inhibitors (sub-antimicrobial doxycycline, minocycline, EGCG, luteolin) to block enzymatic degradation; NLRP3 inhibitors (MCC950, melatonin, luteolin, apigenin) to reduce IL-1β-driven MMP upregulation; complement inhibitors (pegcetacoplan, vitamin D) to prevent ³ complement-mediated synapse elimination on PNN-denuded interneurons; and chondroitin sulfate supplementation to provide substrate for PNN biosynthetic repair. This four-pronged approach—inhibiting degradation, reducing the signals that drive degradation, preventing the consequences of partial degradation, and supporting repair—is analogous to the multi-target strategies used in cartilage preservation in orthopedic medicine. The combination of sub-antimicrobial doxycycline (pharmaceutical MMP inhibitor) with EGCG (supplement MMP inhibitor + iron chelator), melatonin (NLRP3 inhibitor), and chondroitin sulfate (PNN substrate) represents a pragmatic Phase III combination that uses readily available, well-tolerated agents to address the root mechanism of E/I collapse from multiple angles simultaneously. No clinical trial has tested this specific combination, but each component has an independent evidence base and safety profile that supports clinical investigation. 6.2 The Timing Problem: Why Phase-Matching Matters The Spectrum of Collapse framework makes a strong prediction: the same drug can be effective or ineffective depending on when it is deployed relative to the patient's biological phase. This prediction is testable, and indeed the existing clinical trial record provides several natural experiments that illuminate the timing problem. Memantine is FDA-approved for moderate-to-severe AD but shows no benefit in mild AD—consistent with the Spectrum prediction that NMDA antagonism is relevant only during Phase III, when excitotoxic signaling from disinhibited circuits is the dominant pathology. Minocycline showed preclinical efficacy in young APP mice but failed in mild AD patients—consistent with the prediction that microglial modulation is most relevant during the Phase II-III transition, before the post-homeostatic phenotype has been fully established. The anti-amyloid antibodies show modest benefits in amyloid-positive but cognitively normal or mildly impaired individuals—consistent with the prediction that amyloid-related pathology is a Phase II phenomenon best addressed before Phase III circuit collapse has begun. The timing problem creates a fundamental tension in clinical trial design. The standard regulatory pathway requires demonstrating efficacy in patients with diagnosed disease—but in the Spectrum framework, diagnosis typically occurs in Phase III, when the disease has already passed through the phases where many interventions would be most effective. Testing a Phase I agent (PARP inhibitor, mitophagy inducer) in Phase III patients is analogous to testing a fire prevention system in a building that has already burned down: the test will almost certainly fail, but the failure tells us nothing about the agent's potential for prevention. This temporal mismatch between the regulatory framework (which demands proof of benefit in symptomatic patients) and the biological reality (which demands intervention decades before symptoms) is one of the central obstacles to translating the Spectrum of Collapse framework into clinical practice. Adaptive platform trials offer a potential solution to this problem. Platforms such as DIAN-TU (Dominantly Inherited Alzheimer's Network Trials Unit) and A4 (Anti-Amyloid Treatment in Asymptomatic Alzheimer's) have demonstrated the feasibility of testing interventions in presymptomatic individuals—but these trials still treat Alzheimer's as a monolithic disease, enrolling amyloid-positive individuals regardless of their biological phase. A phase-stratified adaptive platform trial would use biomarker panels to assign each participant to a biological phase (I, II, or III), then randomize within each phase to phase-appropriate interventions. This design would allow simultaneous testing of multiple phase-specific hypotheses while maintaining the statistical power and efficiency advantages of a platform trial.

Phase Transition Biomarkers

The practical implementation of phase-stratified trials requires validated biomarkers that can distinguish the three phases in living patients. While no single biomarker panel has been validated for this purpose, the existing biomarker landscape provides candidate markers for each transition. The Phase I-to-II transition may be detectable through: declining plasma or CSF NAD+/NADH ratios (reflecting the metabolic exhaustion of Phase I), rising CSF sTREM2 (indicating microglial activation and transition to disease-associated phenotypes), appearance of CSF p-tau 217 (reflecting the tau propagation of early Phase II), and MRI evidence of hippocampal volume loss (indicating the engagement of the hippocampal bridgehead). The Phase II-to-III transition may be marked by: subclinical epileptiform activity on extended EEG monitoring (indicating emerging E/I imbalance), elevated CSF MMP-9 or CSPG fragments (indicating PNN degradation), rising CSF C3a or C5a (indicating complement activation), and elevated CSF IL-1β or ASC speck levels (indicating NLRP3 inflammasome activation). The development and validation of these phase-transition biomarker panels represents one of the most critical translational priorities identified by the Spectrum of Collapse framework. Without reliable phase assignment, phase-specific therapeutics cannot be deployed rationally, and clinical trials will continue to enroll biologically heterogeneous populations that dilute treatment effects. The precedent in oncology is instructive: molecular staging (ER/PR/HER2 status in breast cancer, EGFR/ALK/ROS1 status in lung cancer) transformed cancer therapeutics from empirical chemotherapy to precision medicine. The Spectrum of Collapse framework envisions a similar transformation in Alzheimer's therapeutics, from monolithic treatment paradigms to biologically informed, phase-specific intervention strategies. 6.3 Biomarker-Guided Intervention Selection Beyond phase assignment, biomarkers can guide the selection of specific agents within each phase. Not all patients in Phase III, for example, will have the same dominant pathology: some may show prominent NLRP3-driven inflammation (high CSF IL-1β, ASC), while others may show dominant complement-mediated synaptic stripping (high CSF C3a, low synaptic density on SV2A PET), and still others may present primarily with PNN degradation (high CSF MMP-9, CSPG fragments) or E/I imbalance (epileptiform activity on EEG). A biomarker-guided approach would select Phase III interventions based on the dominant mechanism in each individual: NLRP3 inhibitors for patients with inflammasome-dominant profiles, complement blockers for complement-dominant profiles, MMP inhibitors for PNN-degradation-dominant profiles, and SV2A modulators or NMDA antagonists for circuit-instability-dominant profiles. This individualized approach within each phase represents a second layer of precision beyond phase stratification. The analogy to oncology is again instructive: within a given cancer stage, molecular profiling determines whether the patient receives targeted therapy, immunotherapy, or conventional chemotherapy. In the Spectrum of Collapse framework, phase assignment determines the broad category of intervention (mitochondrial support for Phase I, lysosomal and microglial modulation for Phase II, E/I stabilization for Phase III), while within-phase biomarker profiling determines the specific agents selected. This two-level precision framework—phase plus mechanism—would represent a fundamental departure from the current one-size-fits-all approach to Alzheimer's treatment and would require the development of multi-analyte biomarker panels, standardized phase-assignment algorithms, and adaptive trial designs capable of testing multiple agents in biologically defined subpopulations. The development of SV2A PET tracers (such as ¹¹C-UCB-J) is particularly promising for Phase III biomarker development, as SV2A density provides a direct measure of synaptic density that can detect the synapse loss driven by complement stripping and PNN degradation. Combining SV2A PET with TSPO PET (a marker of microglial activation), amyloid PET, tau PET, and extended EEG monitoring would provide a multi-modal assessment of Phase III pathology that could guide intervention selection and monitor treatment response. The cost and complexity of this biomarker battery are significant barriers, but the oncology precedent demonstrates that complex molecular staging becomes cost-effective when it guides therapy toward agents that are more likely to be effective in the individual patient. 6.4 The Prevention Paradox: Phase I Interventions in

Asymptomatic People

The Spectrum of Collapse framework identifies Phase I—the Silent Brainstem Erosion, spanning ages 20 to 50—as the period of greatest therapeutic leverage. Interventions deployed during Phase I could prevent or delay the entire downstream cascade of Phase II and Phase III pathology. NAD+ precursor supplementation, mitophagy induction, MAO-B inhibition, and antioxidant support during Phase I could preserve brainstem monoaminergic neurons whose loss triggers the microglial transition of Phase II and the E/I collapse of Phase III. In theory, effective Phase I intervention could prevent Alzheimer's disease entirely—not by curing it, but by preventing the biological cascade from ever reaching the threshold where clinical symptoms emerge. This is the prevention paradox: the phase where therapeutic intervention could have the greatest impact is also the phase where it is hardest to justify, conduct, and measure. Phase I occurs in young to middle-aged adults who have no cognitive symptoms, no clinical diagnosis, and no obvious reason to take neuroprotective medications. Enrolling thirty-year-olds in a trial designed to prevent a disease that may not manifest for forty years faces insuperable practical challenges: the trial duration required to observe clinical endpoints, the cost of following thousands of participants for decades, the difficulty of maintaining adherence over multi-decade timeframes, and the ethical complexity of medicating healthy young adults for a disease they may never develop. Even with surrogate biomarker endpoints (NAD+ levels, mitochondrial function, brainstem volume), the signal-to-noise ratio in a healthy young population is likely to be low. The resolution of the prevention paradox may lie in the distinction between pharmaceutical and nutraceutical interventions. While deploying prescription PARP inhibitors or mTOR inhibitors in healthy thirty-year-olds faces legitimate safety and ethical concerns, many Phase I supplements—NAD+ precursors, CoQ, alpha-lipoic ¹⁰ acid, sulforaphane, spermidine—have excellent safety profiles and are already consumed by millions of people as dietary supplements. A population-level recommendation for NAD+ precursor supplementation beginning in midlife (age 40–50) would be a low-risk, potentially high-impact public health strategy that addresses Phase I pathology without the regulatory hurdles of pharmaceutical intervention. This approach accepts a modest effect size in exchange for broad population reach—the same logic that underlies recommendations for blood pressure management, cholesterol reduction, and physical activity as cardiovascular disease prevention strategies. The Spectrum of Collapse framework thus envisions a two-tier prevention strategy. The first tier is population-level: safe, accessible supplements (NAD+ precursors, mitochondrial cofactors, antioxidants) recommended for everyone from midlife onward, analogous to the recommendation for aspirin in cardiovascular disease prevention (though with considerably better safety profiles). The second tier is biomarker-guided: pharmaceutical interventions deployed in individuals identified through screening biomarkers as having active Phase I, II, or III pathology. The first tier catches the majority at low cost and low risk; the second tier provides targeted treatment to those at highest risk. Together, these tiers could constitute a comprehensive Alzheimer's prevention strategy that addresses the disease across its full temporal arc—from the first mitochondrial lesion in the locus coeruleus to the final excitotoxic cascade in the cortex. 6.5 Proposed Combination Architectures

Minimal Phase III Regimen

For patients identified as being in early Phase III (subclinical epileptiform activity, emerging PNN degradation biomarkers, preserved MMSE > 24), a minimal combination regimen might include: levetiracetam (125–250 mg twice daily) for E/I stabilization via SV2A modulation; sub-antimicrobial doxycycline (20 mg twice daily) for MMP inhibition and PNN protection; and melatonin (5–10 mg nightly) for NLRP3 suppression, circadian support, and GABAergic enhancement. This three-drug combination addresses the circuit, enzymatic, and inflammatory components of Phase III simultaneously using well-characterized, affordable, and generally well-tolerated agents. Each component has independent evidence of safety in elderly populations, and no pharmacokinetic interactions are expected between them. A supplement stack complementing this minimal pharmaceutical regimen could include: magnesium L-threonate (1.5 g daily) for endogenous NMDA modulation, EGCG (400 mg daily) for additional MMP inhibition and iron chelation, taurine (2–3 g daily) for GABAergic tone support, and apigenin (50–100 mg daily) for CD38 inhibition and NAD+ preservation. This supplement layer adds mechanistic depth to the pharmaceutical foundation without significant additional safety risk. The total cost of this combined regimen—including generic medications and quality supplements—would be substantially lower than the cost of anti-amyloid monoclonal antibodies, while potentially addressing a broader range of Phase III mechanisms.

Cross-Phase Transition Regimen

For patients identified as being at the Phase II-to-III transition—rising inflammatory biomarkers, early EEG changes, preserved cognition—a cross-phase regimen would include Phase II agents being tapered alongside Phase III agents being introduced. Phase II agents (e.g., ANAVEX 2-73 for sigma-1 receptor modulation, ferrostatin analogs for ferroptosis prevention, trehalose for autophagy enhancement) would continue as long as Phase II mechanisms remain active, while Phase III agents (levetiracetam, doxycycline, melatonin) would be introduced as E/I biomarkers emerge. This overlapping, transition-aware approach recognizes that phase boundaries are not sharp—Phase II and Phase III mechanisms coexist for years—and that the most dangerous period is the transition, when multiple pathological mechanisms are simultaneously active. The cancer therapeutics analogy is again instructive. In oncology, the transition from neoadjuvant to adjuvant to maintenance therapy is guided by staging, molecular profiling, and treatment response assessment. The same principles apply to the Spectrum of Collapse: transition from Phase II-targeted to Phase III-targeted therapy should be guided by biomarker evidence of evolving pathology, with overlapping regimens during the transition period to prevent gaps in coverage. The complexity of this approach—multiple drugs, multiple supplements, biomarker monitoring, dose adjustments—is admittedly formidable, but it reflects the biological complexity of a disease that has defied simpler strategies for over three decades. VII

Failed Trials Reinterpreted

Why Mechanistically Sound Drugs Failed When Deployed in the Wrong Phase The history of Alzheimer's drug development contains hundreds of clinical trial failures representing billions of dollars in investment and decades of scientific effort. The conventional narrative attributes these failures to insufficient understanding of disease biology, poor target selection, or inadequate clinical trial design. While these factors have certainly contributed, the Spectrum of Collapse framework offers a more specific diagnosis: many of these trials tested reasonable agents in the wrong biological phase. A drug targeting a Phase I mechanism will fail in Phase III patients not because the drug is ineffective, but because the pathology it targets has already run its course. Conversely, a drug targeting Phase III mechanisms will fail in Phase II patients because the pathology it addresses has not yet begun. The following reinterpretations are not post-hoc rationalizations but falsifiable predictions of the Spectrum framework, testable through re-analysis of existing trial data with phase-stratified biomarker panels. 7.1 Anti-Amyloid Antibodies: Wrong Phase? The anti-amyloid antibody program—encompassing bapineuzumab, solanezumab, aducanumab, lecanemab, and donanemab—represents the largest investment in Alzheimer's drug development history. Bapineuzumab (Pfizer/Johnson & Johnson) was the first anti-amyloid antibody to enter Phase III trials, targeting N-terminal epitopes on amyloid-β. Two large Phase III trials (301 and 302) enrolling over 4,500 patients with mild-to-moderate AD showed no significant cognitive benefit despite evidence of amyloid clearance on PET imaging (Salloway et al., 2014). Solanezumab (Eli Lilly), which targets soluble monomeric amyloid-β rather than plaques, failed three Phase III trials: EXPEDITION 1, EXPEDITION 2, and EXPEDITION 3. Even when restricted to patients with amyloid-positive mild AD (EXPEDITION 3), solanezumab showed no significant cognitive benefit (Honig et al., 2018). The Spectrum of Collapse interpretation of these failures centers on the relationship between amyloid accumulation and the phase architecture. In the Spectrum framework, amyloid-β accumulation begins during Phase I as a consequence of neuronal stress and impaired proteostasis, becomes pathologically significant during Phase II as it activates microglial transition to disease-associated phenotypes and triggers the lysosomal PANTHOS pathway, and becomes largely irrelevant by Phase III, when the E/I collapse is driven by PNN degradation, inflammasome activation, and complement stripping—mechanisms that are independent of ongoing amyloid production. Anti-amyloid antibodies tested in patients with mild-to-moderate AD are being deployed in Phase III, when the amyloid-related pathology of Phase II has already catalyzed the downstream mechanisms that are now driving the disease. Clearing amyloid plaques from a brain in Phase III is like disarming a bomb after it has already detonated: the damage was done by amyloid's earlier interaction with microglia, lysosomes, and complement, not by the plaques themselves. The modest successes of lecanemab and donanemab—which show statistically significant but clinically marginal benefits in early AD—are consistent with this interpretation. These newer antibodies are being tested in earlier-stage patients (MCI and early AD) who may be in Phase II, where amyloid-driven microglial activation is still ongoing and removable. The CLARITY AD trial for lecanemab (van Dyck et al., 2023) showed a 27 percent reduction in cognitive decline over 18 months, and the TRAILBLAZER-ALZ 2 trial for donanemab (Sims et al., 2023) showed a 35 percent reduction in patients with intermediate tau burden. The Spectrum framework predicts that these benefits will be greatest in patients who are still in Phase II (amyloid-driven microglial activation without E/I collapse) and will diminish or disappear in patients who have transitioned to Phase III. Biomarker-stratified re-analysis of these trials using Phase III markers (EEG epileptiform activity, CSF MMP-9, complement activation) could test this prediction. 7.2 Antioxidant Trials: Right Idea, Wrong Phase The antioxidant hypothesis of Alzheimer's disease—the notion that oxidative stress drives neurodegeneration and that antioxidant supplementation could prevent or slow the disease—has been tested in several large clinical trials with uniformly disappointing results. The PREADVISE trial (Prevention of Alzheimer's Disease by Vitamin E and Selenium; Kryscio et al., 2017) randomized 7,540 men aged 60 and older to vitamin E (400 IU daily), selenium (200 μg daily), both, or placebo, with a median follow-up of 5.4 years. The trial found no significant effect of vitamin E, selenium, or their combination on dementia incidence. Earlier, the Alzheimer's Disease Cooperative Study (ADCS) trial of vitamin E (2,000 IU daily) versus donepezil versus placebo in patients with mild cognitive impairment (Petersen et al., 2005) found no benefit of vitamin E on the rate of progression from MCI to AD over 36 months. The Spectrum of Collapse framework interprets these failures as a phase mismatch. Oxidative stress is indeed a critical pathological mechanism, but it is most relevant in Phase I, where it drives the mitochondrial DNA damage, PARP-1 hyperactivation, and NAD+ depletion that destroy brainstem monoaminergic neurons. By Phase II and Phase III—the phases at which the PREADVISE and ADCS trials enrolled patients—the oxidative damage that initiated the cascade has already been done. Giving vitamin E to a sixty-year-old with MCI is attempting to prevent a fire that started thirty years earlier. The relevant oxidative stress occurred in the locus coeruleus and dorsal raphe nucleus during ages 20–50, when these neurons were accumulating mitochondrial DNA damage and depleting their NAD+ reserves. Antioxidant supplementation during that period might have been effective; by age 60, the therapeutic window has closed. Moreover, the choice of antioxidant matters. Vitamin E (alpha-tocopherol) is a lipid-soluble membrane antioxidant that does not address the specific chemistry of Phase I oxidative stress—namely, mitochondrial superoxide production, Fenton chemistry from labile iron, and oxidative DNA damage leading to PARP activation. More targeted agents such as mitochondria-specific antioxidants (MitoQ, SS-31), iron chelators (deferiprone), or PARP inhibitors would have been more mechanistically appropriate for Phase I pathology. The antioxidant trials failed not only because they treated the wrong phase but because they used the wrong class of antioxidant—a double mismatch that virtually guaranteed failure regardless of patient selection or trial design. 7.3 Anti-Inflammatory Trials: Incomplete Mechanism

Understanding

The epidemiological observation that chronic NSAID use was associated with reduced Alzheimer's risk (in t Veld et al., 2001; Szekely et al., 2004) prompted a series of randomized prevention trials testing whether anti-inflammatory drugs could prevent cognitive decline. The ADAPT trial (Alzheimer's Disease Anti-inflammatory Prevention Trial) randomized 2,625 elderly individuals with a family history of AD to naproxen (220 mg twice daily), celecoxib (200 mg twice daily), or placebo. The trial was terminated early due to cardiovascular safety concerns with celecoxib and found no cognitive benefit from either NSAID (ADAPT Research Group, 2007). Tarenflurbil (Flurizan, Myriad Genetics), a gamma-secretase modulator with NSAID-derived anti-inflammatory properties, failed a Phase III trial of 1,649 patients with mild AD (Green et al., 2009). These failures effectively ended the 'inflammation hypothesis' era of Alzheimer's drug development. The Spectrum of Collapse reinterpretation reveals why NSAIDs were the wrong anti-inflammatory approach. NSAIDs primarily inhibit cyclooxygenase-1 and cyclooxygenase-2 (COX-1/COX-2), reducing prostaglandin synthesis. While prostaglandins contribute to neuroinflammation, they are not the dominant inflammatory mediators in Phase III of the Spectrum of Collapse. The key Phase III inflammatory mechanisms—NLRP3 inflammasome activation, IL-1β-mediated excitotoxic potentiation, complement-mediated synaptic stripping, MMP-driven PNN degradation—are largely independent of the COX pathway. NSAIDs do not inhibit NLRP3, do not block IL-1β, do not suppress complement activation, and do not inhibit MMPs. They target the wrong branch of the inflammatory cascade—addressing prostaglandin-mediated vasodilation and edema rather than the inflammasome-mediated cytokine and complement cascades that drive Phase III neurodestructive inflammation. The epidemiological protection observed with chronic NSAID use may reflect a different mechanism entirely: chronic COX-2 inhibition during Phase II may reduce prostaglandin-mediated blood-brain barrier breakdown, limiting the entry of peripheral immune cells and inflammatory mediators that accelerate the Phase II-III transition. This protective effect would not be captured in trials enrolling patients who are already in Phase III, where the BBB is extensively compromised and the dominant inflammatory mechanisms are CNS-intrinsic rather than peripherally-mediated. The failure of NSAID trials, like the failure of antioxidant trials, illustrates how incomplete mechanistic understanding—treating 'neuroinflammation' as a monolithic process rather than a phase-specific set of distinct inflammatory programs—leads to trials that are destined to fail. 7.4 Cholinesterase Inhibitors: Addressing a Phase II

Consequence

The cholinesterase inhibitors—donepezil (Aricept), rivastigmine (Exelon), and galantamine (Razadyne)—remain the most widely prescribed drugs for Alzheimer's disease despite providing only modest symptomatic benefit. These agents increase synaptic acetylcholine levels by inhibiting the enzymes that degrade it, partially compensating for the cholinergic deficit caused by the degeneration of basal forebrain cholinergic neurons (nucleus basalis of Meynert). Meta-analyses indicate that cholinesterase inhibitors produce improvements of approximately 2–3 points on the 70-point ADAS-Cog scale over 6–12 months, with benefits that diminish over time as the disease progresses (Birks, 2006). They do not modify disease progression by any biomarker measure. In the Spectrum of Collapse framework, the cholinergic deficit that these drugs address is a consequence of Phase II pathology—specifically, the vulnerability of basal forebrain cholinergic neurons to the lysosomal dysfunction, tau propagation, and inflammatory microenvironment that characterize the hippocampal bridgehead. The nucleus basalis of Meynert, like the hippocampal formation, is a Phase II target because its neurons depend on intact lysosomal function for the turnover of acetylcholine-containing synaptic vesicles and are vulnerable to tau pathology spreading from the entorhinal cortex. Cholinesterase inhibitors address the downstream consequence (reduced synaptic acetylcholine) without treating the upstream cause (lysosomal failure, tau accumulation, microglial assault on cholinergic projections). They are symptomatic treatments for a Phase II consequence, deployed in Phase III patients, using a mechanism that becomes progressively less effective as the cholinergic neurons that produce the acetylcholine they are trying to preserve continue to die. The Spectrum framework does not dismiss cholinesterase inhibitors as useless—their modest symptomatic benefit is real and valuable for patients and caregivers. But it contextualizes their limitation: they cannot modify the disease because they target a downstream consequence rather than an upstream mechanism. A Phase II-targeted strategy that preserved cholinergic neurons—through lysosomal rescue, tau reduction, and microglial modulation—would be expected to maintain cholinergic function naturally and eliminate the need for symptomatic cholinesterase augmentation. The continued reliance on cholinesterase inhibitors as the standard of care reflects the field's failure to develop effective Phase II interventions, not the inherent value of the cholinergic approach. 7.5 BACE Inhibitors: Removing Protective Biology The BACE (beta-site amyloid precursor protein cleaving enzyme) inhibitors represent perhaps the most dramatic failure in Alzheimer's drug development. Verubecestat (Merck), atabecestat (Janssen/Shionogi), lanabecestat (AstraZeneca/Eli Lilly), elenbecestat (Biogen/Eisai), and umibecestat (Novartis/Amgen) all entered clinical trials with strong preclinical data demonstrating robust reduction of amyloid-β production. Yet every one of these programs was terminated for futility or harm. The EPOCH trial of verubecestat in prodromal AD (Egan et al., 2019) showed that the drug reduced CSF amyloid-β by 60–80 percent but was associated with cognitive worsening, weight loss, and psychiatric adverse events including suicidality. The EARLY trial of atabecestat was terminated after elevated liver enzymes and cognitive worsening were observed (Novak et al., 2020). The Spectrum of Collapse framework offers a mechanistic explanation for why BACE inhibitors not only failed to help but actively worsened outcomes. BACE1 cleaves not only amyloid precursor protein but also numerous other substrates with important neurobiological functions, including neuregulin-1 (which regulates myelination and synaptic plasticity), CHL1 (a cell adhesion molecule involved in axonal guidance), and SEZ6 (seizure-related gene 6, which modulates excitatory synaptic transmission). Inhibiting BACE1 to reduce amyloid production simultaneously disrupts these protective biological processes. In the Spectrum framework, BACE inhibition during Phase II or III removes amyloid-β at a stage when its production may be serving a partially protective function—there is growing evidence that low concentrations of amyloid-β monomers have antimicrobial, neurotrophic, and synaptic-modulating properties—while simultaneously disrupting neuregulin-dependent myelination, CHL1-dependent axonal maintenance, and SEZ6-dependent excitatory balance. The cognitive worsening observed with BACE inhibitors is particularly instructive through the Phase III lens. By disrupting SEZ6 processing, BACE inhibition may alter the excitatory-inhibitory balance in cortical circuits that are already teetering on the edge of the E/I collapse. In a brain where PV+ interneurons are dying and inhibitory tone is failing, any additional perturbation of excitatory synaptic transmission—even one intended to be beneficial—can push circuits past the tipping point into pathological hyperexcitability. The BACE inhibitor story thus serves as a cautionary tale about the dangers of targeting a single molecule (amyloid-β) without understanding its multiple biological roles, and about the risks of deploying mechanistically disruptive agents in patients whose circuits are already in a fragile state of Phase III collapse. The broader lesson from BACE inhibitor failures is that the 'less is more' principle applies to Alzheimer's therapeutics: removing a pathological product is not the same as addressing the upstream process that produced it, and may be actively harmful if the product has dual roles that include protective functions. The Spectrum of Collapse framework redirects attention from amyloid removal (a downstream event) to the upstream processes—mitochondrial failure, lysosomal dysfunction, microglial phenotype shifts—that drove its overproduction in the first place. Addressing these upstream processes would naturally reduce amyloid accumulation while preserving the protective biology that BACE inhibitors inadvertently destroyed. VIII

The Validity Ledger

The argument above is only as strong as its weakest load-bearing joint, and the reader is owed an explicit accounting of where it stands on the ground and where it stands on inference.

Each claim below carries a tier and, where it is not settled, the observation that would settle it. 6 claims · 4 not yet settled

Strong (imported, established) — OmegAD found no effect of omega-3 supplementation on the primary cognitive endpoint in mild-to-moderate Alzheimer's disease.

A rigorous randomised trial of 204 patients, and the primary result is null. This paper's interest is in a prespecified very-mild subgroup, which is hypothesis-generating rather than confirmatory.

Strong (imported, established) — The anti-amyloid antibody programme — bapineuzumab through donanemab — represents the largest investment in Alzheimer's drug development, with a trial record now long enough to read.

A matter of public record, and the empirical base for the phase-timing argument that follows.

Moderate (inference, the paper's own claim) — Therapeutics should be matched to disease phase, because an intervention correct for one phase can be inert or harmful in another.

The organising argument of the review. It is assembled from mechanism rather than demonstrated by any trial that stratified on phase, because none has.

Weak (predicted, untested) — Anti-amyloid antibodies underperform because they are administered in the wrong phase, not because the target is wrong.

The sharpest and most consequential claim here, and it is a reinterpretation of existing trials rather than a finding from them.

What would settle it. Re-analysis of completed anti-amyloid trials with phase-stratified biomarker panels, testing whether effect size tracks phase at enrolment.

What would refute it. Effect size found flat across phase strata, which would mean timing is not what separates responders from non-responders.

Weak (predicted, untested) — Autophagy-flux inducers will help where lysosomal degradation is intact and will worsen outcomes where acidification has already failed.

A directional prediction with a clear clinical consequence, stated from mechanism.

What would settle it. A trial arm stratified on a marker of lysosomal acidification, testing for an interaction rather than a main effect.

Weak (predicted, untested) — DHA supplementation is net neuroprotective because GPX4 and allied antioxidant upregulation outweighs the peroxidation susceptibility that DHA-enriched membranes acquire.

Offered as the resolution of an apparent contradiction. The compensation is plausible and is observed in preclinical models; the net direction in humans is not established, and the clinical record is null.

Conclusion

From Monotherapy to Temporal Pharmacology The sixty agents catalogued in this document—thirty pharmaceutical and thirty nutraceutical—represent the raw materials for a fundamental reimagining of Alzheimer's therapeutics. The Spectrum of Collapse framework argues that the field's three-decade search for a single drug to treat Alzheimer's disease was misconceived from the outset, because Alzheimer's disease is not a single disease. It is a sequential cascade of three biologically distinct pathologies—mitochondrial exhaustion, lysosomal and inflammatory bridgehead formation, and excitatory-inhibitory circuit collapse—each with its own molecular machinery, its own vulnerable cell populations, and its own therapeutic targets. A PARP inhibitor cannot save a PV+ interneuron from complement-mediated phagoptosis. A complement inhibitor cannot preserve NAD+ pools in a locus coeruleus neuron. An MMP inhibitor cannot rescue a lysosome undergoing PANTHOS. Each agent has its moment of maximal relevance, and that moment is defined by the phase of the disease.

The Case for Phase-Specific Therapeutics

The central argument of this document is that effective Alzheimer's therapeutics requires phase-specificity: matching the right intervention to the right biological moment. This principle has transformed cancer medicine, where stage-specific and molecular-profile-specific treatment selection has turned many fatal cancers into manageable chronic diseases. The same transformation is possible in Alzheimer's disease, but it requires three capabilities that the field currently lacks. First, validated biomarkers for phase assignment—markers that can reliably distinguish patients in Phase I from those in Phase II, and those in Phase II from those in Phase III. Second, clinical trial designs that accommodate phase stratification—adaptive platform trials that test phase-appropriate agents in biologically defined subpopulations rather than clinically defined 'mild,' 'moderate,' and 'severe' categories that blur phase boundaries. Third, a regulatory framework that accepts biomarker endpoints for prevention trials—because Phase I interventions cannot be validated on a 40-year timeline waiting for clinical dementia endpoints. The development of these capabilities is not a scientific problem alone; it is an institutional and conceptual challenge. The pharmaceutical industry has been organized around the search for a single blockbuster drug—one molecule, one mechanism, one disease. The Spectrum of Collapse demands a different model: a portfolio of agents, each targeting a specific mechanism at a specific phase, deployed in combinations that evolve as the disease evolves. This model is closer to HIV/AIDS therapeutics (where combination antiretroviral therapy transformed a death sentence into a manageable chronic condition) or oncology (where multi-agent, stage-specific protocols have dramatically improved survival) than to the traditional single-drug-per-disease model that has dominated Alzheimer's research.

The Road Ahead: Adaptive Platform Trials With Phase

Stratification

The practical path forward begins with the design and implementation of a phase-stratified adaptive platform trial. Such a trial would enroll participants across the full biological spectrum of Alzheimer's disease—from amyloid-negative individuals with NAD+ depletion biomarkers (Phase I) to patients with established E/I imbalance (Phase III)—and assign them to treatment arms matched to their biological phase. Within each phase, multiple agents could be tested simultaneously against a shared control arm, with interim analyses allowing futile arms to be dropped and promising arms to be expanded. This design maximizes the information yield per enrolled participant while testing the core prediction of the Spectrum framework: that phase-matched treatments outperform phase-mismatched treatments. The biomarker panel for phase assignment would need to be validated prospectively, but candidate markers are available from existing research programs. Phase I assignment could use blood or CSF NAD+ metabolomics, PAR levels, and locus coeruleus MRI neuromelanin imaging. Phase II assignment could use amyloid PET, CSF p-tau 217, sTREM2, and ferritin. Phase III assignment could use overnight EEG (epileptiform activity), SV2A PET (synaptic density), CSF MMP-9, complement activation markers (C3a, C5a), and NLRP3 activation markers (IL-1β, ASC specks). The cost and logistical complexity of this biomarker battery are significant but not insurmountable, particularly in the context of a platform trial where the costs are distributed across multiple treatment arms and the biomarker validation itself becomes a scientific output of the trial. The Spectrum of Collapse framework also calls for a rethinking of clinical endpoints. Traditional cognitive endpoints (ADAS-Cog, CDR-SB) are insensitive to Phase I and early Phase II interventions, where the goal is to prevent progression rather than reverse existing deficits. Phase-specific biomarker endpoints—NAD+ restoration for Phase I, microglial phenotype markers for Phase II, E/I balance metrics for Phase III—could serve as surrogate endpoints that enable faster, smaller, and less expensive trials while maintaining biological relevance. The FDA's accelerated approval pathway, which allows approval based on surrogate endpoints reasonably likely to predict clinical benefit, provides a regulatory mechanism for this approach.

From Monotherapy to Informed Polypharmacy

The ultimate vision of the Spectrum of Collapse therapeutic strategy is an evolution from monotherapy to what might be called 'informed polypharmacy'—the deliberate combination of multiple agents, each addressing a specific mechanism at a specific phase, deployed in sequences and combinations guided by biomarker evidence of evolving pathology. This is not polypharmacy in the pejorative sense—the haphazard accumulation of medications driven by symptom management—but a coordinated, mechanism-driven therapeutic strategy analogous to the multi-agent protocols used in oncology and infectious disease. Each agent in the combination has a defined mechanistic role, a defined phase of relevance, and a defined set of biomarkers that justify its inclusion and monitor its effects. This vision is ambitious, but the components are largely available. Many of the agents catalogued in this document are already FDA-approved for other indications, generically available, and well-characterized in terms of safety. Others are available as dietary supplements with excellent safety profiles. What has been missing is not better drugs but a better framework for deploying them—a temporal architecture that specifies which targets matter when, which agents address which targets, and how to determine a patient's biological phase. The Spectrum of Collapse provides that architecture. The work that remains is translational: validating the phase-assignment biomarkers, designing the platform trials, building the clinical infrastructure for phase-specific treatment, and—perhaps most challengingly—shifting the culture of Alzheimer's drug development from the pursuit of a single silver bullet to the construction of a phase-aware therapeutic system. CHAPTER 7 THE STAKES COULD NOT BE HIGHER. China's Weaponization of Trade 2026:150-163. DOI 10.7312/cha-22239-010.

References

Naproxen and celecoxib do not prevent AD in early results from a randomized controlled trial. Neurology 2007;68(21):1800-1808. DOI 10.1212/01.wnl.0000260269.93245.d2.

Bhatt, D. K., et al. (2014). IL-1β upregulates MMP-9 expression in astrocytes via NF-κB. Glia, 62(7), 1124–1136. Birks, J. (2006). Cholinesterase inhibitors for Alzheimer's disease. Cochrane Database of Systematic Reviews, (1), CD005593. Bischoff-Ferrari, H. A., Vellas, B., Rizzoli, R., et al. (2020). Effect of vitamin D supplementation, omega-3 fatty acid supplementation, or a strength-training exercise program on clinical outcomes in older adults: The DO-HEALTH randomized clinical trial. JAMA, 324(18), 1855–1868. Braak, H., Thal, D. R., Ghebremedhin, E., & Del Tredici, K. (2011). Stages of the pathologic process in Alzheimer disease: Age categories from 1 to 100 years. Journal of Neuropathology & Experimental Neurology, 70(11), 960–969. Brotini, S., Schievano, C., & Bhatt, A. (2017). Ultra-micronized palmitoylethanolamide: A new adjunctive treatment in Alzheimer's disease? CNS & Neurological Disorders Drug Targets, 16(5), 540–548. Calabro, R. S., Naro, A., De Luca, R., et al. (2016). PEA-um in the treatment of mild cognitive impairment. Clinical Drug Investigation, 36(5), 401–405. Camacho-Pereira, J., Tarragó, M. G., Chini, C. C. S., et al. (2016). CD38 dictates age-related NAD decline and mitochondrial dysfunction through an SIRT3-dependent mechanism. Cell Metabolism, 23(6), 1127–1139.

Chen HQ, Jin ZY, Wang XJ, Xu XM, Deng L, Zhao JW. Luteolin protects dopaminergic neurons from inflammation-induced injury through inhibition of microglial activation. Neurosci Lett 2008;448(2):175-9. DOI 10.1016/j.neulet.2008.10.046.

D’Souza DC, Gil R, Cassello K, Morrissey K, Abi-Saab D, White J, et al.. IV glycine and oral d-cycloserine effects on plasma and CSF amino acids in healthy humans. Biological Psychiatry 2000;47(5):450-462. DOI 10.1016/s0006-3223(99)00133-x.
Heneka MT, Kummer MP, Stutz A, Delekate A, Schwartz S, Vieira-Saecker A, et al.. NLRP3 is activated in Alzheimer's disease and contributes to pathology in APP/PS1 mice. Nature 2013;493(7434):674-8. DOI 10.1038/nature11729.

Howard R, Zubko O, Bradley R, Harper E, Pank L, O'Brien J, et al.. Minocycline at 2 Different Dosages vs Placebo for Patients With Mild Alzheimer Disease: A Randomized Clinical Trial. JAMA Neurol 2020;77(2):164-174. DOI 10.1001/jamaneurol.2019.3762.

Monfort J, Pelletier JP, Garcia-Giralt N, Martel-Pelletier J. Biochemical basis of the effect of chondroitin sulphate on osteoarthritis articular tissues. Ann Rheum Dis 2008;67(6):735-40. DOI 10.1136/ard.2006.068882.

Parrish, K. E., Sarkaria, J. N., & Bhatt, W. F. (2015). The impact of efflux transporters on the brain distribution of olaparib. Journal of Pharmacology and Experimental Therapeutics, 355(2), 206–212. Petersen, R. C., Thomas, R. G., Grundman, M., et al. (2005). Vitamin E and donepezil for the treatment of mild cognitive impairment. New England Journal of Medicine, 352(23), 2379–2388. Reisberg, B., Doody, R., Steffens, D., et al. (2003). Memantine in moderate-to-severe Alzheimer's disease. New England Journal of Medicine, 348(14), 1333–1341. Ridker, P. M., Everett, B. M., Thuren, T., et al. (2017). Antiinflammatory therapy with canakinumab for atherosclerotic disease. New England Journal of Medicine, 377(12), 1119–1131. Robson, M. E., Im, S. A., Senkus, E., et al. (2017). OlympiAD: Olaparib for metastatic breast cancer in patients with a germline BRCA mutation. New England Journal of Medicine, 377, 523–533. Rosales-Corral, S. A., Acuna-Castroviejo, D., Coto-Montes, A., et al. (2012). Alzheimer's disease: Pathological mechanisms and the beneficial role of melatonin. Journal of Pineal Research, 52(2), 167–202. Salloway, S., Sperling, R., Fox, N. C., et al. (2014). Two phase 3 trials of bapineuzumab in mild-to-moderate Alzheimer's disease. New England Journal of Medicine, 370(4), 322–333. Seabrook, T. J., Jiang, L., Maier, M., & Bhatt, C. A. (2006). Minocycline affects microglia activation, amyloid-β deposition, and behavior in APP-tg mice. Glia, 53(7), 776–782. Shi, Q., Chowdhury, S., Ma, R., et al. (2017). Complement C3 deficiency protects against neurodegeneration in aged plaque-rich APP/PS1 mice. Science Translational Medicine, 9(392), eaaf6295. Shimoi, K., Okada, H., Furugori, M., et al. (1998). Intestinal absorption of luteolin and luteolin 7-O-β-glucoside in rats and humans. FEBS Letters, 438(3), 220–224. Sims, J. R., Zimmer, J. A., Evans, C. D., et al. (2023). Donanemab in early symptomatic Alzheimer disease: The TRAILBLAZER-ALZ 2 randomized clinical trial. JAMA, 330(6), 512–527. Singh, P., Gollapalli, K., Mangiola, S., et al. (2023). Taurine deficiency as a driver of aging. Science 380(6649), eabn9257. Slutsky, I., Abumaria, N., Wu, L. J., et al. (2010). Enhancement of learning and memory by elevating brain magnesium. Neuron, 65(2), 165–177. Szekely, C. A., Thorne, J. E., Zandi, P. P., et al. (2004). Nonsteroidal anti-inflammatory drugs for the prevention of Alzheimer's disease: A systematic review. Neuroepidemiology, 23(4), 159–169.

Tariot, P. N., Raman, R., Aisen, P. S., et al. (2011). Divalproex sodium in Alzheimer's disease: Lack of efficacy in a dose-controlled, parallel-group, double-blind trial (VALID). Archives of General Psychiatry, 68(9), 918–929. Tariot, P. N., Farlow, M. R., Grossberg, G. T., et al. (2004). Memantine treatment in patients with moderate to severe Alzheimer disease already receiving donepezil: A randomized controlled trial. JAMA, 291(3), 317–324. Theoharides, T. C., Asadi, S., & Panagiotidou, S. (2012). A case series of a luteolin formulation (NeuroProtek) in children with autism spectrum disorders. International Journal of Immunopathology and Pharmacology, 25(2), 317–323. van Dyck, C. H., Swanson, C. J., Aisen, P., et al. (2023). Lecanemab in early Alzheimer's disease. New England Journal of Medicine, 388(1), 9–21. Volt, H., García, J. A., Doerrier, C., et al. (2016). Same molecule but different expression: Aging and sepsis trigger NLRP3 inflammasome activation, a target of melatonin. Journal of Pineal Research 60(2), 193–205. Vossel, K. A., Ranasinghe, K. G., Beagle, A. J., et al. (2016). Incidence and impact of subclinical epileptiform activity in Alzheimer's disease. Annals of Neurology, 80(6), 858–870. Vossel, K. A., Tartaglia, M. C., Bhatt, D. B., et al. (2021). Effect of levetiracetam on cognition in patients with Alzheimer disease with and without epileptiform activity. JAMA Neurology, 78(11), 1345–1354. Wang, Y., Kim, N. S., Bhatt, A., et al. (2009). Poly(ADP-ribose) signals to mitochondrial AIF: A key event in parthanatos. Experimental Neurology, 218(2), 193–202. Xu, J., Wang, L. L., Dammer, E. B., et al. (2015). Melatonin for sleep disorders and cognition in dementia: A meta-analysis of randomized controlled trials. American Journal of Alzheimer's Disease & Other Dementias, 30(5), 439–447. Zhao, L., Wang, J. L., Liu, R., et al. (2013). Neuroprotective, anti-amyloidogenic and neurotrophic effects of apigenin in an Alzheimer's disease mouse model. Molecules, 18(8), 9949–9965.

Genes named on this page: NLRP3; PARP1, PARP-1; MAOA/MAOB, MAO-B, monoamine oxidase; GPX4; MMP9, MMP-9; mTOR, mTORC1; NGF, nerve growth factor; NFE2L2 (NRF2), Nrf2; PP2A; CD38; APP, amyloid precursor protein; NFKB1 (NF-κB), NF-κB, NF-kappaB, NFkappaB, NFκB, NF-kB; GSK3B, GSK-3, GSK3, GSK-3β, GSK3β, GSK-3beta, GSK3beta; C3; TNF; SirT1; IL6, IL-6; SV2A; HCAR2 (GPR109A), GPR109A, HCAR2; Fyn; BDNF; PSEN1, PS1; MMP2, MMP-2; PINK1; PYCARD, ASC; PRKAA (AMPK), AMPK; NMNAT2; TREM2, sTREM2; SST (somatostatin), somatostatin, SST; PRKN, Parkin; SARM1; TXN, Thioredoxin; C1QA, C1q; TFEB; NAMPT; EIF2S1 (eIF2α), eIF2; CYCS (cytochrome c), cytochrome c; IL18, IL-18; NOS2 (iNOS), iNOS; HSF1; APOE, apoe4; AKT1 (Akt), AKT; ULK1; SIRT3; PTGS2 (COX-2), Cox-2; BNIP3L (NIX), BNIP3L, NIX; OGDH (KGDH), ketoglutarate dehydrogenase, KGDH; JAK2; LRP1; BECN1, Beclin; MAPK14 (p38), p38; GRIN2B, NR2B, GluN2B; CAT (catalase), CATALASE; NTRK1 (TrkA), TrkA; HSPA1A (Hsp70), HSP70; CRP, C-reactive protein; EP300 (p300), EP300, p300; VDR, vitamin D receptor; STAT (1/3/6), STAT1, STAT3; HMOX1, heme oxygenase, HO-1; PVALB (parvalbumin), parvalbumin; CR3; PIK3CA (PI3K), PI3K; BCL2 (Bcl-2), Bcl-2, BCL2; CR1; PRKCA (PKC), PKC; HSP90AA1 (Hsp90), HSP90; Keap1; BACE1; MAP1LC3B (LC3), LC3; CASP1, caspase-1; FTH1/FTL (ferritin), ferritin; GSDMD, gasdermin; ADAMTS4, ADAMTS-4; ACHE, acetylcholinesterase; ABCB1 (P-gp), P-glycoprotein; CLU, clusterin; EIF2AK3 (PERK), PERK; CHAT (ChAT), choline acetyltransferase; Atg5; SLC1A2 (EAAT2/GLT-1), EAAT2, GLT-1; Atg7; SOD2; CHL1; MAPK8 (JNK), JNK; JAK1; ACAN (aggrecan), aggrecan; SQSTM1 (p62), p62; BCAN (brevican), brevican; SOD1; P2RY12, P2Y12; CSF1R; MAPT; ATF4; HDAC2; CX3CL1 (fractalkine), fractalkine; Mfsd2a; VCAN (versican), versican; CHI3L1 (YKL-40), YKL-40; NCAN (neurocan), neurocan; MMP3, MMP-3; MBP (myelin basic protein), myelin basic protein; GRIN2A, GluN2A; Cdk5; CCL2 (MCP-1), MCP-1, MCP1; NMRK1, NRK1; Il10, IL-10; IGF1, IGF-1; PDHA1 (PDH), pyruvate dehydrogenase; TFAM; NRG1 (neuregulin-1), neuregulin-1; AANAT; ALOX5, 5-LOX; NMRK2, NRK2; TSPO; NAPRT; PPP2R2A (B55α), PPP2R2A; FOXO3; HMGCR, HMG-CoA reductase; EGFR; NQO1.

Contents

Introduction: The Rationale for Phase-Specific Therapeutics

Part I: Phase I — The Silent Brainstem Erosion (Age 20–50)

Part II: Phase II — The Hippocampal Bridgehead (Age 50–65)

Part III: Phase III — The E/I Collapse (Age 65–75)

Synthesis: Cross-Phase Interactions and Combination Strategies

References

The Rationale for Phase-Specific

Therapeutics

The Failure of Monolithic Therapeutics

The Architecture of This Document

Phase I Overview: The Silent Brainstem Erosion

Phase I Drugs: Targeting the Silent

Brainstem Erosion

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Phase I Supplements: Nutraceutical

Support for the Mitochondrial Crisis

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

3

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

10

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action

Rationale Within Phase I

Clinical Trial History

Current Status & Limitations

Mechanism of Action