A scalable and equitable framework for target and patient prioritisation in rare disease antisense therapeutics

This paper presents the UPNAT framework, a scalable, disease-agnostic, and equitable system developed by UK experts to systematically prioritize patients and genetic targets for antisense oligonucleotide therapies within the NHS, aiming to standardize decision-making and bridge gaps in rare disease treatment.

The Gist

Imagine you are a master chef in a massive, public kitchen (the UK's National Health Service). Suddenly, you discover a new, magical ingredient: Antisense Oligonucleotides (ASOs). These are tiny, custom-made molecular tools that can fix broken recipes inside your cells. They are like "spellbooks" that tell a cell how to read its genetic instructions correctly again.

The problem? You have a pantry full of thousands of rare, broken recipes (genetic diseases), but you only have a limited amount of this magical ingredient, time, and money. You can't fix every single broken recipe at once.

The Big Question: How do you decide which broken recipes to fix first, and for which patients, without being unfair or making mistakes?

This paper introduces a new Decision Framework (a "Master Plan") created by a team of experts to answer that question. Here is how it works, explained simply:

1. The Problem: Too Many Choices, Not Enough Clarity

In the past, deciding who gets a new gene therapy was a bit like a game of "guess who." Doctors and scientists would look at a patient and say, "This looks promising!" or "Maybe not." But because everyone had their own way of thinking, some patients got lucky while others were left out, even if they had the same disease. It was chaotic and unfair.

2. The Solution: The "Four-Door" Filter

The authors built a system with four distinct doors (modules). Before a patient can get the "magic ingredient," their case must pass through these four checks. Think of it like a security checkpoint at an airport, but instead of checking for weapons, they are checking for "fixability."

• Door 1: The Disease Check (Is the recipe broken enough to matter?)

  • The Analogy: Is the broken recipe causing a disaster in the kitchen? Is there already a different tool that fixes it? If the disease is too mild, or if there's already a perfect cure, maybe we don't need this new magic ingredient yet.
  • The Goal: Make sure the disease is serious enough and that this specific tool is the best fit.

• Door 2: The Lab Check (Do we have a test kitchen?)

  • The Analogy: Before we cook the meal for a real person, can we test it in a small model? Do we have a "mini-kitchen" (lab models) that looks like the patient's body to test if the magic ingredient works?
  • The Goal: If we can't test it in the lab first, it's too risky to try on a human.

• Door 3: The Patient Check (Is the patient ready?)

  • The Analogy: Is the patient strong enough to handle the cooking process? Are they too sick for the treatment to help, or is the risk too high compared to the benefit?
  • The Goal: Ensure the treatment is safe and actually helpful for this specific person at this specific time.

• Door 4: The Code Check (Is the typo fixable?)

  • The Analogy: This is where they look at the actual genetic "typo." Is the mistake in a place where our magic ingredient can reach it? Some typos are in the middle of a word (easy to fix), while others are in a part of the book we can't touch.
  • The Goal: Confirm the genetic error is technically solvable.

3. How They Tested It

The team didn't just dream this up; they tested it.

• The "Gold Standard" Test: They ran their new system against diseases that already have approved drugs. The system correctly said, "Yes, these are good candidates," proving it works for known successes.
• The "Real World" Test: They applied it to real patients waiting for help.
  • Success Story: For a patient with a specific eye disease, all four doors opened. The system said, "Go ahead, this is a great candidate."
  • The "No" Story: For a patient with a severe brain and bone disease, the system said, "Stop." Even though the disease was serious, the patient was too sick to survive the treatment, or the genetic error was too hard to fix. This "No" was actually a good thing—it saved the patient from a dangerous, useless treatment.

4. Why This Matters

This framework is like a traffic light system for rare diseases.

• Green Light: "We have the science, the models, and the patient is ready. Let's develop a treatment."
• Red Light: "We can't fix this yet, or it's too risky. Let's wait or look for other options."
• Yellow Light: "We need more information before we decide."

The Bottom Line

This paper gives the UK (and the world) a fair, transparent, and repeatable rulebook for deciding who gets life-changing gene therapies. It stops the process from being a "lottery" and turns it into a structured, scientific decision.

It acknowledges that science changes. If we invent a better delivery method tomorrow, a patient who was once "Red Light" might become "Green Light" tomorrow. This system is designed to be updated, ensuring that as our technology gets better, we can help more people, faster, and more fairly.

Technical Summary

1. Problem Statement

The rapid advancement of Nucleic Acid Therapies (NATs), specifically Antisense Oligonucleotides (ASOs), has created a critical translational bottleneck. While ASOs offer flexible, gene-specific treatments for rare genetic diseases, the lack of a structured, transparent, and equitable framework for selecting which patients and targets (diseases, genes, or variants) to prioritize has led to:

• Inconsistent Decision-Making: Prioritization often relies on fragmented local expertise rather than standardized criteria.
• Scalability Issues: As genomic diagnostics identify more patients, healthcare systems (particularly the UK's NHS) struggle to manage the volume of potential candidates.
• Knowledge Gaps: Existing guidelines (e.g., N=1 Collaborative VARIANT guidelines) address variant eligibility but fail to integrate disease biology, functional model readiness, and patient-level risk-benefit analysis into a unified assessment.
• Equity Concerns: Without a disease-agnostic framework, access to life-saving therapies risks becoming dependent on geography or institutional resources rather than clinical need and biological feasibility.

2. Methodology

The authors developed the UPNAT (UK Platform for Nucleic Acid Therapies) Target Selection Framework through a multidisciplinary consensus process involving experts in genetics, clinical neurology, ophthalmology, and translational medicine.

• Framework Structure: The framework is modular and disease-agnostic, consisting of four distinct assessment modules:

  1. Disease Module: Evaluates biological tractability, disease mechanism, availability of natural history data, and existing effective therapies. It uses weighted scoring to distinguish between technical feasibility and broader prioritization considerations.
  2. Functional Model Module: Assesses the readiness of experimental systems (in vitro/in vivo models, functional assays) required to validate ASO efficacy and safety.
  3. Patient Module: Focuses on individual clinical suitability, prioritizing safety and risk-benefit analysis. It is primarily exclusion-focused to prevent inappropriate treatment of patients with advanced disease or unfavorable risk profiles.
  4. Variant Module: Integrates the N=1 Collaborative (N1C) VARIANT guidelines to assess genetic amenability to specific ASO mechanisms (splice correction, transcript knockdown, or upregulation).

• Assessment Process:

  • Two-Stage Workflow: Each module applies Step 1 (Exclusion Criteria) to filter out infeasible candidates early, followed by Step 2 (Scoring) to generate a prioritization score for those that pass.
  • Tiered Governance: Disease and functional model assessments are centralized and episodic (reusable across centers), while patient assessments are designed for iterative local application.

• Validation Strategy:

  • Calibration: The framework was retrospectively applied to 11 market-approved ASO therapies (8 gene-disease pairs) to ensure the criteria did not inappropriately exclude established successes.
  • Real-World Application: The framework was applied to a cohort of 26 patients referred for ASO consideration. Four specific case studies (Doyne honeycomb retinal dystrophy, Niemann-Pick type C, BANDDOS, and Centronuclear myopathy) were analyzed in detail to demonstrate the framework's ability to handle diverse decision scenarios.

3. Key Contributions

• First Disease-Agnostic Framework: This is the first end-to-end framework designed to systematically prioritize patients and targets for ASO development across the entire spectrum of rare genetic diseases, rather than focusing on a single condition.
• Integration of Biological and Clinical Dimensions: Unlike previous guidelines that focused solely on variant mechanics, this framework explicitly separates and then integrates biological feasibility (disease/variant/model) with clinical and system-level constraints (patient safety, delivery feasibility).
• Modular and Reusable Design: By decoupling disease-level assessments from individual patient assessments, the framework allows for efficient scaling. Once a disease/gene is assessed as "tractable," that data can be reused for multiple patients, reducing redundant expert review.
• Transparency and Reproducibility: The framework replaces implicit, informal decision-making with explicit scoring and exclusion criteria, enabling auditability and consistent application across the NHS.

4. Results

• Calibration with Approved Therapies: When applied to market-approved ASO targets, the framework correctly identified them as high-priority candidates. Notably, two conditions (Hereditary Angioedema and Hereditary Transthyretin Amyloidosis) were excluded at the Disease Module stage not due to biological infeasibility, but because effective alternative treatments already existed. This confirmed the framework's logic aligns with real-world resource allocation.
• Variant Triage: In the cohort of 26 patients, initial variant-level triage (using N1C guidelines) classified 10 as "Eligible" or "Likely Eligible."
  • Splice Correction: 2 variants were eligible (e.g., deep intronic variants in NPC1 and MTM1).
  • Transcript Knockdown: 7 variants were eligible (e.g., EFEMP1 in Doyne honeycomb retinal dystrophy).
  • Upregulation: No variants were deemed eligible, highlighting current limitations in this therapeutic modality.
• Case Study Outcomes:
  • Prioritization (DHRD/EFEMP1): High scores across all modules supported progression, aligning with ongoing clinical development.
  • Translational Constraints (NPC1): While biologically tractable, the framework highlighted constraints related to CNS delivery and the modest efficacy of existing therapies, supporting a cautious but active approach.
  • Exclusion (BANDDOS/CSF1R): Despite biological tractability, the patient was excluded due to rapid disease progression and unfavorable risk-benefit ratios, demonstrating the framework's ability to prevent futile interventions.
  • Conditional Progression (Centronuclear Myopathy/MTM1): High biological scores were tempered by technical delivery challenges (skeletal muscle/liver targeting), illustrating how the framework surfaces specific translational barriers.

5. Significance

• Equitable Access: The framework provides a mechanism to ensure that access to expensive, complex NATs is determined by transparent criteria rather than institutional bias or geographic location, a critical need for publicly funded systems like the NHS.
• Scalability for Rare Diseases: By enabling centralized assessment of disease biology and functional models, the framework reduces the burden on local clinicians, allowing the system to scale as genomic diagnosis rates increase.
• Future-Proofing: The modular design allows the framework to adapt to new ASO chemistries, delivery technologies, and other NAT modalities (e.g., siRNA, gene editing) without requiring a complete overhaul.
• Policy and Governance: It offers a model for regulatory bodies and healthcare systems to manage the "N-of-1" and small-cohort therapy landscape, balancing innovation with responsible, evidence-based clinical implementation.

In conclusion, the UPNAT framework transforms the prioritization of ASO therapies from an ad-hoc process into a structured, reproducible, and equitable system, bridging the gap between genomic discovery and clinical application for rare diseases.