Every Cure and Computational Pharmacophenomics: A New Field of Medicine

Nelson Advisors
17 hours ago
10 min read

The Structural Crisis in Modern Pharmacology and the Treatment Gap

The global landscape of clinical medicine is currently defined by a profound and systemic disparity between the identification of human pathologies and the availability of regulatory-approved therapeutic interventions. While modern diagnostic capabilities and genomic sequencing have allowed for the identification of approximately 18,000 distinct human diseases, the therapeutic arsenal remains remarkably narrow.

Statistical analysis of the current pharmacopeia reveals that approximately 14,000 of these diseases, representing more than 75% of recognised human pathologies, do not have a single FDA approved treatment. This therapeutic vacuum affects over 300 Million people worldwide, many of whom suffer from rare conditions that fail to attract the research and development (R&D) investment required by the traditional pharmaceutical model.

The traditional paradigm for drug discovery is characterised by a "slow, expensive, and siloed" architecture. The development of a novel molecular entity (NME) typically necessitates a financial commitment ranging from $1 Billion to $2 Billion and a temporal investment of 10 to 15 years. Furthermore, this model is fraught with risk, maintaining a staggering 90% failure rate as candidates progress from preclinical stages through clinical trials. Consequently, commercial pharmaceutical entities are incentivised to prioritise "blockbuster" drugs targeting high-prevalence conditions within wealthy markets, leaving the "ignorome", the vast space of untreated and under-researched diseases, largely unaddressed.

Every Cure, a nonprofit organisation co-founded by David Fajgenbaum, MD, MBA, MSc, represents a fundamental shift in this ecosystem. By utilising artificial intelligence (AI) and the burgeoning field of computational pharmacophenomics, Every Cure aims to systematically evaluate the untapped potential of the approximately 4,000 drugs already approved by the FDA. The organisational premise is that the solution to many of the 14,000 untreated diseases may already exist on pharmacy shelves, hidden by fragmented data and a lack of financial incentive for repurposing.

Comparative Metrics of Drug Discovery Paradigms

Feature	Traditional De Novo Discovery	Systematic AI-Driven Repurposing
Average Cost per Drug	$1 Billion – $2 Billion	<1% of Traditional Cost
Timeline to Clinical Use	10 – 15 Years	1 – 3 Years (Variable)
Risk Profile	90% Failure Rate	Lowered due to established safety data
Discovery Mechanism	Targeted Molecular Design	Systematic AI/ML Analysis
Data Structure	Proprietary and Siloed	Open-Source and Integrated
Focus	New Intellectual Property	Clinical Impact and Generic Drugs

The Genesis of Computational Pharmacophenomics: The David Fajgenbaum Precedent

The conceptual foundation of Every Cure is inextricably linked to the personal and clinical history of David Fajgenbaum. In 2010, as a third-year medical student at the University of Pennsylvania, Fajgenbaum transitioned from an elite athlete to a critically ill patient suffering from idiopathic multicentric Castleman disease (iMCD). iMCD is a lethal haematologic disorder characterised by a massive proinflammatory cytokine storm, resulting in systemic organ failure.

Fajgenbaum’s journey involved five near-fatal relapses during which he experienced complete failure of his liver and kidneys, and was twice read his last rites. At the time, the only FDA-approved treatment for iMCD was effective in only one-third of the patient population. Faced with the limits of existing clinical practice, Fajgenbaum applied a rigorous, research-first approach to his own case, founding the Castleman Disease Collaborative Network (CDCN) in 2012 to unify global research efforts.

Through proteomic and genomic analysis of his own biospecimens, Fajgenbaum identified that the $PI3K/Akt/mTOR$signaling pathway was highly upregulated during his disease flares. This discovery led to the hypothesis that sirolimus—an inexpensive drug used for 25 years to prevent organ transplant rejection, could be repurposed to block this specific pathway in iMCD patients.

Fajgenbaum began treating himself with sirolimus in 2014, leading to a period of remission that has lasted over 11 years. This success served as a proof-of-concept for the idea that "cures are hiding in plain sight" and provided the impetus for scaling this approach through the Every Cure initiative.

Pathological Hallmarks of Idiopathic Multicentric Castleman Disease (iMCD)

Biomarker/Pathway	Status in iMCD	Therapeutic Implication
$PI3K/Akt/mTOR$	Highly Upregulated	Target for Sirolimus
Interleukin-6 ($IL-6$)	Elevated	Primary cytokine driver
Vascular Endothelial Growth Factor ($VEGF$)	Upregulated	Contributes to angiogenesis/swelling
$JAK/STAT$ Signaling	Upregulated	Target for Ruxolitinib
$CXCL13$ Chemokine	Highly Upregulated	Involved in lymph node B-cell homing

The MATRIX Platform: Engineering Systematic Discovery

To transition from the serendipitous discovery that saved Fajgenbaum's life to a systematic method for all diseases, Every Cure developed the MATRIX platform. MATRIX is a high-accuracy predictive engine designed to evaluate the potential of every FDA-approved drug against every recognised disease simultaneously. The platform addresses the "siloed" nature of medical data by integrating hundreds of disparate datasets into a unified computational framework.

Technical Architecture and Data Flow

The MATRIX repository is structured as a monorepo, utilizing advanced data engineering tools to manage massive scale. The pipeline is built on the Kedro framework, which provides a rigorous structure for machine learning (ML) workflows, managing data catalogs, parameters, and versioning.

Data Ingestion and Knowledge Graph Construction

The platform ingests raw data from multiple massive biomedical knowledge graph (KG) sources, including RTX-KG2 and ROBOKOP. To manage the processing of millions of nodes and edges, Every Cure employs PySpark for distributed data cleaning and node normalisation. The resulting knowledge graph is stored in Neo4j, a graph database optimised for the complex relational queries required to map drug-disease associations.

Core Library/Service	Functionality	Technical Basis
matrix-fabricator	Declarative synthetic data generation for testing	Python-based tools.
matrix-gcp-datasets	Integration with Google Cloud and Spark utilities	PySpark and GCP SDK.
matrix-mlflow-utils	Experiment tracking and metric reporting	MLflow integration.
matrix-auth	Authentication and environment security	Identity-Aware Proxy (IAP).
FastAPI Services	Hosting supporting APIs (Synonymizer, MOA Visualizer)	High-performance Python API.

Machine Learning and Graph Embeddings

At the heart of the MATRIX pipeline is the generation of graph embeddings. These are numerical vector representations of the complex biological relationships within the knowledge graph. By training models on known "treats" relationships, where a drug is already approved for a disease, the system learns the topological patterns associated with therapeutic success. The model then applies these learned patterns to the "ignorome", the 75 Million drug-disease pairs that have not yet been clinically validated, to quantify the likelihood of efficacy.

Computational Pharmacophenomics: A New Field of Medicine

Every Cure has formalized this approach into a new medical field called computational pharmacophenomics. As detailed in the Lancet Haematology, this methodology is both drug-agnostic and disease-agnostic. Unlike traditional repurposing, which might start with a specific drug and look for a new disease (or vice versa), computational pharmacophenomics evaluates the entire matrix of possibilities simultaneously.

The Human-in-the-Loop Review System

A critical differentiator of the MATRIX platform is its "human-in-the-loop" refinement process. Purely algorithmic predictions in medicine can suffer from biological implausibility. Every Cure addresses this by integrating a medical review team that assesses over 1,000 unique opportunities monthly. Feedback from these clinicians and researchers is captured as structured data and used to retrain the ranking algorithms, fostering a continuously improving system.

Methodology	Data Input	Strategic Outcome
KGML-xDTD	Path-based MOAs, RTX-KG2, ROBOKOP	Explainable drug repurposing predictions.
LLM-Based Synthesis	Biomedical literature and clinical notes	reasoning-based evidence synthesis.
Real-World Evidence	Electronic health records, prescription data	Validation of non-obvious clinical links.
Proteomic Profiling	Quantitative plasma protein analysis	Identification of actionable signaling targets.

The use of Large Language Models (LLMs), such as Google’s Gemini 2.0, allows Every Cure to construct complex reasoning chains that "pressure test" predictions against the vast body of published medical literature. This ensures that high-ranking predictions are supported by mechanistic rationale and clinical feasibility.

Portfolio of Impact: Successes and Frontier Research

The efficacy of the Every Cure model is demonstrated by a growing portfolio of repurposed drugs that have already transitioned from algorithmic prediction to patient impact.

Case Study: POEMS Syndrome and Hospice Remission

In January 2024, Every Cure’s algorithms identified a high-potential combination of carfilzomib, cyclophosphamide, and dexamethasone for a patient with POEMS syndrome (Polyneuropathy, Organomegaly, Endocrinology, Monoclonal protein, and Skin changes).

The patient, who was entering hospice care after failing all standard treatments, was administered the drug combination based on the high score. Remarkably, the patient achieved remission and has remained stable for over two years. This case illustrates the power of evaluating drugs that are approved for related haematologic conditions (like Multiple Myeloma) against rare, phenotypically similar syndromes.

Case Study: Lidocaine in Oncology

The platform has also identified lidocaine, a common local anesthetic, as a "Frontier Explorer" for localized cancers, including breast and oropharyngeal squamous cell carcinoma. Laboratory evidence suggests that lidocaine may inhibit tumor growth and metastasis by modulating cellular signaling during surgery. Every Cure is currently supporting studies to determine if injecting lidocaine around a tumour before surgical excision can significantly reduce the risk of recurrence and death.

Case Study: DFMO and Bachmann-Bupp Syndrome

Every Cure is collaborating with the researchers who discovered Bachmann-Bupp Syndrome (BABS) to repurpose DL-alpha-difluoromethylornithine (DFMO). Originally developed for African sleeping sickness, DFMO has shown the potential to stabilise and improve outcomes for children with this rare developmental disorder.

Summary of Targeted Repurposing Opportunities

Drug	Category	Target Disease	Clinical Rationale
Sirolimus	Clinical Gem	iMCD	$mTOR$ inhibition in cytokine storm.
Adalimumab	Clinical Gem	Refractory iMCD	$TNF$ inhibition identified via proteomics.
Pembrolizumab	Clinical Gem	Metastatic Angiosarcoma	Immune-checkpoint inhibition in vascular malignancy.
Ruxolitinib	Clinical Gem	iMCD	$JAK/STAT$ signaling modulation.
Botox	Unsung Hero	Major Depressive Disorder	Glabellar injection impacts facial feedback loops.
Beta Blocker	Frontier Explorer	Rare Neurodegenerative Disease	Enhancement of lysosomal function.
Folinic Acid	Unsung Hero	Cerebral Folate Deficiency	Bypasses folate receptor antibodies.

The Economic and Institutional Ecosystem of Every Cure

The success of a nonprofit drug repurposing organisation depends on navigating a market that traditionally ignores non-patentable, generic medicines. Every Cure addresses this through a robust network of strategic partners and innovative funding mechanisms

Strategic Partnerships and Infrastructure

In 2024, Every Cure expanded its collaboration with Google Cloud to leverage the Gemini 2.0 AI infrastructure. This partnership allows the organisation to accelerate its data processing and utilise generative AI for large-scale evidence synthesis. Additionally, a partnership with Elsevier provides access to one of the world's most comprehensive repositories of biomedical data, allowing Every Cure to integrate proprietary insights that are typically unavailable to independent researchers.

Funding for these operations has been secured through historic federal and philanthropic commitments. The Advanced Research Projects Agency for Health (ARPA-H) awarded Every Cure a three-year, $48.3 million contract to specifically develop the AI tool for rare diseases. This was followed by a five-year, $60 million commitment from TED’s Audacious Project to ensure that predictions are successfully translated into clinical trials and laboratory studies.

Partner/Donor	Contribution Type	Impact
ARPA-H	Federal Contract ($48.3M)	Scaling of the MATRIX AI platform.
TED Audacious Project	Philanthropic Grant ($60M)	Clinical trial and lab validation funding.
Google Cloud	Technology and Infrastructure	Gemini 2.0 for reasoning-based AI.
Elsevier	Data and Expertise	High-quality biomedical data integration.
Chan Zuckerberg Initiative	Funding and Collaborative Network	Acceleration of rare disease research.
Flagship Pioneering	Strategic Partnership	Innovation in drug development models.

Future Roadmap: Toward Open-Source Drug Discovery

The ultimate objective of Every Cure is the democratisation of drug discovery. By 2026, the organisation has committed to publicly releasing putative efficacy scores for all 75 million potential drug-disease matches. This "Public Data Zone" will allow researchers, physicians, and patient advocacy groups to prioritise treatments for their specific communities without the need for proprietary software or expensive consulting.

The 2030 Impact Goals

The organisation has established an ambitious target to advance repurposed treatments for 15 to 25 diseases by 2030. This goal is designed to prove that systematic repurposing can fundamentally bridge the treatment gap for marginalised patient populations. Beyond 2030, the aim is to establish a permanent, global infrastructure where "no drug is left behind," and every approved medication is fully utilised for every disease it can possibly treat.

Analysis of Contemporary Alternatives: Repurposing vs. The "Blockbuster" Market

While Every Cure focuses on unlocking existing medicines, the commercial pharmaceutical market continues to pursue high-cost, novel molecular entities. Reports for 2026 identify several "high-impact" drugs to watch, such as Eli Lilly’s orforglipron for obesity and Novo Nordisk’s CagriSema. These drugs represent the cutting edge of metabolic and chronic disease research, often carrying projected sales in the billions of dollars.

However, the development of these drugs highlights the very gap Every Cure aims to fill. While these blockbusters address large-market conditions like Type 2 Diabetes and Obesity, they do nothing for the thousands of ultra-rare diseases that lack any commercial appeal. Every Cure’s model of using AI to find uses for generic drugs like sirolimus or dexamethasone represents a vital "safety net" for the 14,000 diseases that the traditional system ignores.

2026 Clinical Outlook: Repurposing vs. Novel Entities

Drug Category	Example	Market Incentive	Accessibility
Novel Blockbusters (2026)	Orforglipron (Eli Lilly)	High Profitability / Patent	Expensive / Restricted
Repurposed Generics (Every Cure)	Sirolimus / Lidocaine	Patient Impact / Non-Profit	Inexpensive / Broad
Targeted Rare Disease NMEs	Voyxact (Otsuka)	Niche Profitability	High Cost
AI-Discovered Repurposed	Adalimumab for iMCD	Clinical Necessity	Immediate via Off-Label

Conclusion: A Paradigm Shift in Global Health Equity

The structural failure of the current drug development model is not a result of a lack of scientific curiosity, but a consequence of misaligned incentives and fragmented information. The "slow, expensive, and siloed" nature of drug discovery has left 14,000 diseases untreated, not because they are biologically incurable, but because they are commercially unprofitable.

Every Cure and the field of computational pharmacophenomics offer a viable alternative. By utilising AI to scan the "world's biomedical knowledge," Every Cure is effectively mapping the unknown space of human biology—the "ignorome"—and identifying therapeutic links that have existed for decades but remained invisible. The success of David Fajgenbaum’s "patient-scientist" model serves as the ultimate validation: a drug approved for one purpose can be the difference between life and death for an entirely different pathology if the data is viewed through a systematic, all-vs-all lens.

As the organization moves toward its 2026 goal of open-sourcing 75 million drug-disease scores, the medical community is on the verge of a new era of "collaborative discovery." By removing the profit motive from the identification of generic drug uses and leveraging the computational power of the MATRIX platform, Every Cure is not just searching for new cures—it is fundamentally redefining what it means to discover one. The mission to save lives by "unlocking the hidden potential of existing drugs" is a testament to the belief that in the face of 14,000 untreated diseases, the answers are not always in the future; sometimes, they are already on the shelf.

Nelson Advisors > European MedTech and HealthTech Investment Banking

Nelson Advisors specialise in Mergers and Acquisitions, Partnerships and Investments for Digital Health, HealthTech, Health IT, Consumer HealthTech, Healthcare Cybersecurity, Healthcare AI companies. www.nelsonadvisors.co.uk

Nelson Advisors regularly publish Thought Leadership articles covering market insights, trends, analysis & predictions @ https://www.healthcare.digital

Nelson Advisors publish Europe’s leading HealthTech and MedTech M&A Newsletter every week, subscribe today! https://lnkd.in/e5hTp_xb

Nelson Advisors pride ourselves on our DNA as ‘Founders advising Founders.’ We partner with entrepreneurs, boards and investors to maximise shareholder value and investment returns. www.nelsonadvisors.co.uk

#NelsonAdvisors #HealthTech #DigitalHealth #HealthIT #Cybersecurity #HealthcareAI #ConsumerHealthTech #Mergers #Acquisitions #Partnerships #Growth #Strategy #NHS #UK #Europe #USA #VentureCapital #PrivateEquity #Founders #SeriesA #SeriesB #Founders #SellSide #TechAssets #Fundraising #BuildBuyPartner #GoToMarket #PharmaTech #BioTech #Genomics #MedTech