How AI Models Are Shaping Perceptions of Cystic Fibrosis Therapies

Key Findings from the LucidScope AI Visibility Analysis of Cystic Fibrosis Therapies

Vertex Pharmaceuticals is the undisputed clinical and commercial standard. Both GPT-5.5 and Gemini position Trikafta/Kaftrio and Alyftrek as the proven reference point in CFTR modulation, with challengers framed as promising but clinically unproven.
Tolerability and cost, not efficacy, now dominate the incumbent narrative. AI-generated discussion of Vertex has shifted toward side effects, mental health burden, access friction, and value.
AI models read the market as two separate risk lanes, not a head-to-head contest. Incumbent risk centers on incremental value and tolerability; challenger risk centers on foundational proof.
Source influence does not match market position. Company investor-relations domains and Reddit shaped what the models retrieved more than several regulatory and clinical sources.
Model choice changed the strategic takeaway. GPT-5.5 produced a source-grounded clinical read, while Gemini produced a more commercially dramatic, higher-risk interpretation.

Patients, physicians, investors, and industry stakeholders increasingly turn to AI systems for answers once sought through search engines. Questions such as “What is the best cystic fibrosis treatment?” or “What are the alternatives to Trikafta?” are now directed to platforms such as ChatGPT, Claude, Gemini, and Perplexity.

The answers these systems provide depend on the sources each model retrieves, the evidence it prioritizes, and how it interprets clinical data. As generative AI becomes part of healthcare information discovery, understanding AI perception is becoming increasingly important for biopharma companies.

To examine this dynamic in cystic fibrosis, LucidScope analyzed how leading AI models perceive Vertex’s approved portfolio alongside a field of investigational challengers: 4D Molecular Therapeutics (4D-710), Arcturus Therapeutics (ARCT-032), Sionna Therapeutics (SION-719, 451, 109), ReCode Therapeutics (RCT-2100), and Krystal Biotech (KB407). Download the CF AI Perception briefing: CF_Therapies_LucidScope_AI_Perceptions_Briefing

About the LucidScope AI Visibility Analysis

The assessment included:

GPT-5.5 and Gemini
Stakeholder-style prompts covering clinical data gaps, patient complaints, market access, insurance barriers, and cost-effectiveness
Evaluation of source retrieval patterns and citation behavior
Assessment of narrative framing and asset visibility

The objective was not to determine which therapy is clinically superior, but to understand how AI systems currently perceive and communicate information about these therapies. One methodological note: raw share-of-voice metrics were unreliable for this dataset and contradicted the models’ own generated text, so qualitative synthesis was prioritized over headline visibility numbers.

LucidScope AI Visibility Scorecard for Cystic Fibrosis Therapies

Therapy	Role	Evidence Maturity	How AI Frames It
Vertex (Trikafta/Kaftrio, Alyftrek)	Approved incumbent standard	High / established	Efficacy respected; scrutinized on tolerability and cost
Sionna (SION-719, 451, 109)	CFTR-modulator challenger	Pre-efficacy	Highly visible, but carries an “expensive add-on” narrative
4D-710 (4D Molecular Therapeutics)	Gene therapy challenger	Early, uncontrolled	Promising but dependent on redosing proof and AAV safety
ARCT-032 (Arcturus)	mRNA challenger	Early	Reduced by Gemini to a “no meaningful FEV1 improvement” shorthand
RCT-2100 (ReCode)	mRNA challenger	Limited, no public human efficacy	Skepticism around mucus delivery and nebulizer burden
KB407 (Krystal Biotech)	Gene therapy challenger	Limited, proof-of-concept	Molecular transduction does not equal clinical efficacy

Which Cystic Fibrosis Therapies Lead AI Visibility and AI Rankings?

Across both models, Vertex anchored the field, and the ranking tracked evidence maturity. An approved, standard-of-care portfolio sat at the center, while challengers were judged by how far they still are from clinical validation. For Vertex, the open efficacy questions were narrow, confined to the margins: pediatric extrapolation, rare-variant evidence, long-term outcomes, and residual disease. For the challengers, the questions were foundational, and both models kept returning to the same gating steps of delivery, durability, repeat dosing, safety, and whether molecular signals translate into clinical benefit.

Strategic Implication: Clinical Maturity Drives AI Positioning

Approval status and interpretable endpoints, not mechanism, decide who AI treats as proven. A novel modality earns visibility, but not credibility, until it clears human trials.

Why Do AI Models Focus on Tolerability When Evaluating Cystic Fibrosis Therapies?

The most striking finding was that the incumbent narrative has moved past efficacy. Rather than debating whether Vertex’s modulators work, AI systems concentrated on tolerability and real-world treatment burden.

For Trikafta/Kaftrio, models surfaced recurring online concerns around mental health, cognition and brain fog, GI effects, and liver-monitoring anxiety. For Alyftrek, they highlighted early-switcher complaints such as headaches and fatigue, plus uncertainty over whether those symptoms are transient. Gemini sharpened this into an emotive frame, presenting the incumbent experience as a choice between improved lung function and psychological burden.

For the investigational assets, there was little direct patient-complaint volume. Instead, sentiment centered on unmet need: patients who are modulator-ineligible, intolerant, poor responders, or still symptomatic want mutation-agnostic or add-on options.

Strategic Implication: Tolerability Narratives Now Shape Incumbent Perception

For a mature portfolio, the reputational battleground is patient experience, not efficacy. Whether that narrative helps or hurts depends on separating a product’s specific side-effect profile from the general burden of its drug class.

How AI Models Assess Market Access for Cystic Fibrosis Therapies

The models framed Vertex as dominant but economically scrutinized. High U.S. access is tightly controlled by payer friction and mandatory genotype documentation, and models flagged significant out-of-pocket risk and intense cost-effectiveness scrutiny for incremental benefits.

For challengers, two distinct access paths emerged. A mutation-agnostic path drew high payer confidence when targeting patients with no current options and a clear value proposition. An add-on path drew the opposite: both models highlighted a Sionna combination (SION-719 plus Trikafta) as a high-risk proposition, demanding proof of substantial incremental benefit to overcome payer resistance on combined cost and complexity.

Strategic Implication: Market Access Is Framed Around Economic Scrutiny

Efficacy alone does not secure access. The most defensible challenger position is serving patients the incumbent cannot reach, since an add-on layered onto the standard inherits its cost scrutiny.

How AI Models Differ in Their Evaluation of Cystic Fibrosis Therapies

Although both models reached similar high-level conclusions, their analytical postures diverged sharply.

GPT-5.5 was source-grounded and auditable. It balanced approved therapies against genetic and mRNA challengers, contextualized patient forums carefully, detailed trial endpoints, and cited primary regulatory sources, giving it low hallucination risk.

Gemini was market-oriented and dramatic. It framed the landscape as monopoly versus emerging threats, blended clinical disappointment with stock-market reaction, and made specific payer and pricing claims without explicit URLs, giving it high hallucination and auditability risk. Gemini also uniquely cast the ARCT-032 Phase 2 interim signal as a “failure to improve lung function.”

Strategic Implication: Model Choice Can Change the AI Narrative

The same prompts produced different strategic takeaways depending on the model. In a two-model environment, a single system like Gemini can set durable shorthand that hardens against an asset over time.

Which Sources Influence AI Perception of Cystic Fibrosis Therapies?

Why Visibility and Influence Are Not the Same

Vertex led the clinical narrative, but the most-cited domains were not Vertex-owned. GPT-5.5 drew on a defensible core of FDA.gov, ClinicalTrials.gov, SEC.gov, and DailyMed. Gemini drew on a periphery of media and speculative commentary such as BioSpace, Reddit, StockTitan, Substack, and SeekingAlpha. Company investor-relations domains, including investors.sionnatx.com and ir.krystalbio.com, were among the most influential sources shaping what the models retrieved.

Two structural traps emerged. In the IR echo chamber, early-stage narratives are controlled almost entirely by company press releases, which AI reads as promotional unless backed by peer-reviewed sources. In the Reddit sentiment trap, without robust real-world evidence on high-authority domains, models default to Reddit to answer “how do patients feel?”, skewing Vertex toward worst-case neuropsychiatric complaints.

Strategic Implication: Retrievability Can Shape AI Perception

Discoverability is now part of evidence strategy. What a model cannot easily retrieve, it effectively discounts, regardless of scientific quality.

Strategic Implications for Biopharma

In cystic fibrosis, AI perception now turns less on new data than on who publishes it, where, and in what form. The practical response splits by function:

Commercial and investor relations: move beyond the press release. Models penalize promotional domains and read investor-relations content as marketing unless a peer-reviewed source backs it, so early-stage and biomarker claims need to become structured clinical explainers tied to outcomes rather than milestones.
Medical affairs: occupy the authoritative zone the models already trust. Seeding FDA, DailyMed, and NIH-adjacent domains with peer-reviewed real-world evidence on long-term outcomes, side-effect management, and rare-variant efficacy is what displaces Reddit as the default answer on patient experience.
HEOR and market access: defend the economics from both directions. Challengers pursuing add-on positioning need published cost-benefit rationales before payer skepticism hardens into AI shorthand, while Vertex needs indexable payer-policy summaries so speculative pricing claims do not fill the vacuum.

The through-line is simple. The incumbent’s greatest exposure is a slowly compounding tolerability-and-cost narrative, and the challengers’ clearest opening is proving outcomes rather than mechanism for the patients standard-of-care therapy cannot yet reach.

Conclusion

Vertex currently leads AI perception of cystic fibrosis because of its approval status, established clinical evidence, and standard-of-care position. But that leadership has been reframed. AI models have effectively conceded the efficacy debate to Vertex, then reopened the conversation on tolerability, cost, and unmet need.

The challengers show the inverse: visibility driven by investor sources, running ahead of the clinical proof that would justify it. The lesson extends beyond cystic fibrosis. As answer engines shape how therapies are discovered and compared, the task for biopharma is no longer only to generate evidence, but to make it retrievable and consistently represented.

LucidScope helps organizations measure and improve how they are represented across leading AI platforms. Visit www.lucidscope.ai or contact info@lqventures.com to learn more.

FAQ

Which company leads AI perception in cystic fibrosis?

Vertex Pharmaceuticals. Both GPT-5.5 and Gemini treat Trikafta/Kaftrio and Alyftrek as the clinical and commercial standard in CFTR modulation, with all other assets framed as unproven challengers.

What is the main AI-driven risk to Vertex in cystic fibrosis?

A tolerability and cost narrative. Real-world discussion, heavily sourced from Reddit, concentrates on neuropsychiatric and cognitive side effects, shifting the story from efficacy toward patient burden and value.

Why do GPT-5.5 and Gemini give different answers?

They rely on different sources. GPT-5.5 uses regulatory and trial documents and cites them, while Gemini leans on media and investor forums and makes pricing claims without explicit URLs, giving it higher hallucination risk.

What is Generative Engine Optimization (GEO) in biopharma?

Generative Engine Optimization refers to improving how AI systems retrieve and represent information about a therapy, company, or disease area. For biopharma organizations, GEO involves ensuring that authoritative evidence is highly discoverable, consistently presented, and easy for AI systems to interpret.

Download the LucidScope AI Visibility Presentation

👉 Download the full LucidScope briefing on AI perception of cystic fibrosis therapies. Thirteen slides covering Vertex and five emerging challengers across clinical evidence, patient sentiment, market access, and source behaviour. CF_Therapies_LucidScope_AI_Perceptions_Briefing

📧 Contact us to get the full report: info@lqventures.com

LucidScope helps organizations measure and improve how they are represented across leading AI platforms. Visit www.lucidscope.ai to learn more.