Navigating Drug Approvals

FDA rolls out generative AI tool ‘Elsa’ to speed up reviews and streamline regulatory tasks >> 💊The FDA is rolling out Elsa, a secure generative AI tool that helps staff accelerate clinical reviews, summarize adverse events, compare drug labels, and even generate code for internal systems 💊Elsa is built on a large language model and housed in a high-security GovCloud environment, ensuring sensitive regulatory data stays in-house and not trained on by external models 💊Early results from pilot testing with FDA scientific reviewers were positive, leading to the accelerated, under-budget deployment across all centers (original target launch date was June 30th) 💊Elsa’s debut is seen as the first step in a broader AI integration strategy that will expand to include advanced analytics and further generative AI use cases 💊FDA leadership is positioning AI as a lever to boost performance without compromising scientific rigor, describing Elsa as a tool that “enhances and optimizes the potential of every employee.” 💊Elsa launches amid a proposed 4% FDA budget cut and loss of up to 3,500 staff, potentially helping offset pressure on review timelines #digitalhealth #ai #pharma

Earlier this month, the U.S. Food & Drug Administration announced a major step toward integrating Generative AI across the agency — a move that could reshape how new medicines, devices, and diagnostics are evaluated.   The potential benefits are compelling. AI could streamline parts of the review process, reduce administrative burden, and enable faster, more consistent decision-making. For example, the FDA will use its GenAI tool, Elsa, to accelerate clinical protocol reviews, compare drug labels, summarize adverse events, identify high-priority inspection targets, and more. These applications could play a meaningful role in supporting the FDA’s mission of bringing safe, effective medicines to patients – potentially faster and more efficiently. Of course, with this opportunity comes responsibility. The agency oversees some of the most sensitive data and high-stakes decisions in healthcare. As AI becomes more embedded in regulatory workflows, a few principles will be critical: ◆ 𝗔𝗜 𝘀𝗵𝗼𝘂𝗹𝗱 𝗿𝗮𝗶𝘀𝗲 𝘁𝗵𝗲 𝗯𝗮𝗿. It should help ‘supercharge’ reviewers and strengthen the quality and consistency of reviews. ◆ 𝗛𝘂𝗺𝗮𝗻 𝗼𝘃𝗲𝗿𝘀𝗶𝗴𝗵𝘁 𝗶𝘀 𝗲𝘀𝘀𝗲𝗻𝘁𝗶𝗮𝗹. AI can and should support decision-making, but experienced reviewers will still need to be at the helm. ◆ 𝗧𝗿𝗮𝗻𝘀𝗽𝗮𝗿𝗲𝗻𝗰𝘆 𝗯𝘂𝗶𝗹𝗱𝘀 𝘁𝗿𝘂𝘀𝘁. Clear, proactive communication about how tools are trained and used will help bolster confidence across industry and the public. ◆ 𝗗𝗮𝘁𝗮 𝘀𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗺𝘂𝘀𝘁 𝗯𝗲 𝘂𝗻𝗰𝗼𝗺𝗽𝗿𝗼𝗺𝗶𝘀𝗶𝗻𝗴. Protecting proprietary and patient-related information, of course, has to remain a top priority. It’s encouraging to see the FDA taking such a forward-looking, measured approach — one that mirrors how many of us in the field, including our team at Recursion, are approaching AI: test, learn, improve, and scale. This is both an exciting and consequential moment for the industry. Done right, AI can help supercharge the regulatory review process while upholding the scientific rigor and trust that define the FDA. I’ll be watching closely — and optimistically — to see how this evolves over the months ahead! #GenerativeAI #ResponsibleAI #FDANews #RegulatoryAffairs #DrugDevelopment

🚨 𝗕𝗿𝗲𝗮𝗸𝗶𝗻𝗴 𝗻𝗲𝘄𝘀 𝗳𝗿𝗼𝗺 𝗙𝗗𝗔 Drug approval no longer starts with a trial. It starts with a mechanism. The FDA has introduced the “𝗽𝗹𝗮𝘂𝘀𝗶𝗯𝗹𝗲 𝗺𝗲𝗰𝗵𝗮𝗻𝗶𝘀𝗺” 𝗽𝗮𝘁𝗵𝘄𝗮𝘆: a new route to approve bespoke therapies when classic trials are impossible.  Think N-of-1 gene editing for ultra-rare, often fatal childhood diseases. 𝟱 𝗲𝗹𝗲𝗺𝗲𝗻𝘁𝘀 𝘁𝗵𝗲 𝗙𝗗𝗔 𝘄𝗮𝗻𝘁𝘀 𝘁𝗼 𝘀𝗲𝗲 𝗯𝗲𝗳𝗼𝗿𝗲 𝗮𝗽𝗽𝗿𝗼𝘃𝗮𝗹 1️⃣ 𝗖𝗹𝗲𝗮𝗿 𝗺𝗼𝗹𝗲𝗰𝘂𝗹𝗮𝗿 𝗰𝗮𝘂𝘀𝗲 A single, well-defined genetic or molecular defect driving the disease. 2️⃣ 𝗧𝗵𝗲𝗿𝗮𝗽𝘆 𝗱𝗲𝘀𝗶𝗴𝗻𝗲𝗱 𝘁𝗼 𝗳𝗶𝘅 𝘁𝗵𝗮𝘁 𝗱𝗲𝗳𝗲𝗰𝘁 The product must target the precise mechanism: the edit, splice, or RNA change that corrects the biology. 3️⃣ 𝗪𝗲𝗹𝗹-𝘂𝗻𝗱𝗲𝗿𝘀𝘁𝗼𝗼𝗱 𝗻𝗮𝘁𝘂𝗿𝗮𝗹 𝗵𝗶𝘀𝘁𝗼𝗿𝘆 Strong historical data showing how the disease progresses without treatment, so real benefit is clear. 4️⃣ 𝗘𝘃𝗶𝗱𝗲𝗻𝗰𝗲 𝗼𝗳 𝘁𝗮𝗿𝗴𝗲𝘁 𝗲𝗻𝗴𝗮𝗴𝗲𝗺𝗲𝗻𝘁 Proof the therapy does what it’s meant to do, through biomarkers, biopsies, or validated non-animal models. 5️⃣ 𝗠𝗲𝗮𝗻𝗶𝗻𝗴𝗳𝘂𝗹 𝗰𝗹𝗶𝗻𝗶𝗰𝗮𝗹 𝗯𝗲𝗻𝗲𝗳𝗶𝘁 Improvements strong enough that they cannot be dismissed as noise. A single patient can serve as their own control. If a platform shows success in several different patients, even with unique bespoke edits, the FDA can move toward platform-level authorization, not just case-by-case exemptions. 👉 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 𝗳𝗼𝗿 𝗽𝗮𝘁𝗶𝗲𝗻𝘁𝘀  For families facing ultra-rare genetic diseases, the old logic was brutal: too rare for trials → no drug → no options. This pathway changes that:  • From “too rare to study” → “biologically defined and actionable.”  • From isolated compassionate-use miracles → a structured regulatory route.  • From decade-long timelines → months from design to first dosing in the most urgent pediatric cases. And it does not end at approval:  • Long-term real-world evidence  • Ongoing monitoring for off-target edits, immune issues, developmental risks  • Registries to track durability and outcomes 👉 𝗪𝗵𝗼 𝗯𝗲𝗻𝗲𝗳𝗶𝘁𝘀 𝗳𝗶𝗿𝘀𝘁?  • Infants with lethal monogenic diseases  • Ultra-rare disorders with a single known driver mutation  • Small, genetically defined subsets in oncology and immunology 𝗜𝘁’𝘀 𝗮 𝗯𝗶𝗼𝗹𝗼𝗴𝘆-𝗳𝗶𝗿𝘀𝘁, 𝗽𝗹𝗮𝘁𝗳𝗼𝗿𝗺-𝗯𝗮𝘀𝗲𝗱 𝗮𝗽𝗽𝗿𝗼𝗮𝗰𝗵 𝗯𝘂𝗶𝗹𝘁 𝗳𝗼𝗿 𝗰𝗼𝗻𝗱𝗶𝘁𝗶𝗼𝗻𝘀 𝘄𝗵𝗲𝗿𝗲 𝗹𝗮𝗿𝗴𝗲 𝘁𝗿𝗶𝗮𝗹𝘀 𝘄𝗶𝗹𝗹 𝗻𝗲𝘃𝗲𝗿 𝗯𝗲 𝗽𝗼𝘀𝘀𝗶𝗯𝗹𝗲. It’s a design guide for how to build your next platform and IND package. Because for many families, this is the first time “𝘆𝗼𝘂𝗿 𝗯𝗶𝗼𝗹𝗼𝗴𝘆 𝗶𝘀 𝘂𝗻𝗶𝗾𝘂𝗲” doesn’t automatically mean “𝘆𝗼𝘂’𝗿𝗲 𝗼𝗻 𝘆𝗼𝘂𝗿 𝗼𝘄𝗻.”

I recently brought together two passions: the power of large language models (#LLMs) - primarily using Google's NotebookLM for this effort - and the challenge of getting early drug development right. With LLMs, I analyzed FDA “Other Action Letters” — Complete Response and refusal-to-file letters, all thankfully and recently published under FDA's OpenFDA. These previously unaccessible documents reveal why promising drugs stumble. The analysis surfaced five recurring barriers that derail early clinical programs: flawed trial design, weak data integrity, gaps in safety characterization, CMC/device issues, and lack of regulatory dialogue. The memo I’ve written distills these lessons for biotech CEOs and translational leaders. It also highlights the value of FDA’s transparency in making this data public — a resource the whole sector should be learning from. #FDA #regulatoryscience #earlydrugdevelopment #science #innovation #newdrugs #access4patients #OpenFDA

As anticipated, #FDA took another “radical transparency” step today: revising its policy of disclosing decision letters, known as complete response letters (CRLs), issued when FDA determines a new drug application (NDA or BLA) is not approvable. Eighty-nine recent, redacted and previously undisclosed CRLs were published– a far cry from the overblown announcement of 200+ CRLs published last month... which were already public. ► What’s new and striking is that “the agency will promptly release newly issued [and redacted] CRLs”, including for unapproved new drugs. ► In contrast, FDA historically released CRLs only post-approval in online ‘action packages’ or in response to FOIA requests. ► Sponsors will need to pivot in expectation that CRLs will be much more rapidly and broadly public – and hope redactions continue to be thorough and complete. ► For other stakeholders, the change could better illuminate deficiencies in manufacturing and CMC, or failed strategies in trial design and conduct, and #drugsafety concerns. Citing the President’s “Gold Standard Science” Executive Order, FDA is interpreting FOIA and the FFDCA provisions governing disclosure differently and more expansively. Unfortunately, FDA is again promulgating major policies via press release, rather than through the Federal Register, rendering it impossible for stakeholders to respond on the administrative record. Finally, recall Peter Lurie’s and FDA’s 2015 BMJ study that looked at 61 CRLs available exclusively to the Agency, finding that “press releases are generally an incomplete source of reasons for FDA nonapproval of applications” and that they "are incomplete substitutes for the detailed information contained in complete response letters." Lurie also concluded that SEC-compliant disclosures’ “mentions of complete response letters were sometimes absent and, in general, less detailed than the letters.” It remains to be seen how sponsors will modify their communications or otherwise respond to FDA’s more timely disclosure of CRLs, which FDA intends as “providing complete context to investors and shareholders, and above all, restoring public trust.” #kendallsquarepolicy

What happens when AI models become as trustworthy as clinical data? The answer could redefine drug development entirely: Drug development has always followed the same path: discovery, preclinical, clinical, regulatory, launch. Each phase took years, each step consumed capital. AI is now reshaping that entire cycle. Not just how drugs are discovered, but how they’re tested, approved, and commercialized. 1. Discovery: Faster Target-to-Candidate Nearly 30% of new preclinical candidates now come from AI pipelines. Platforms like Atomwise, BenevolentAI, and Insilico Medicine combine genomics and chemistry data to find targets, design molecules, and predict interactions. Discovery timelines are dropping up to 40%, costs by 30%. Partnerships such as AstraZeneca’s $555M deal with Algen Biotech show the shift from single assets to scalable discovery systems. 2. Preclinical: Reducing Animal Testing The FDA Modernization Act (2022) authorized AI toxicity and organ-on-chip models as non-animal alternatives. By 2025, pilot IND programs now accept validated AI and organ-chip data, removing months from preclinical cycles. Earlier go/no-go decisions and better human relevance are driving faster IND readiness. 3. Clinical: From Months to Weeks AI accelerates recruitment, site selection, and adaptive design. Patient matching that once took months now happens in days. Digital-twin models reduce participants while maintaining statistical power. FDA’s 2025 guidance defines how “AI model credibility” must be proven: clear context, explainability, and monitoring. Experts expect clinical programs 30–50% shorter by 2030. Could regulators ever skip Phase 1 studies if AI models predict safety and dose outcomes? Not yet, but hybrid models pairing AI evidence with smaller human trials are emerging. 4. Regulatory and Launch AI already supports dossier preparation, evidence synthesis, and risk analysis. Regulators now expect transparent validation and lifecycle monitoring for any AI used in submissions. Commercially, AI drives forecasting, access strategy, and post-market analytics. By 2030, it may function as the operating system for commercialization itself. 5. Proprietary vs Open AI Pharma is dividing. Some build closed, proprietary models for control and IP protection. Others favor open frameworks for speed and collaboration. The likely future is hybrid: closed models refined on private data, open components for discovery and interoperability. Computational evidence is becoming as strategic as clinical data. 3 Signals for Executives: • Is your team treating AI as core infrastructure, not a pilot? • Are you investing in regulatory-grade model validation and lifecycle monitoring? • Is your platform strategy built for transparency and collaboration with regulators and partners? At Kybora.com, we help leaders navigate this transformation, aligning science, capital, & execution in the AI-enabled future of biopharma.

FDA just raised the bar for CAR-T approvals. That creates space for different approaches. The agency now requires randomized superiority trials for CAR-T therapies, moving away from single-arm studies. This affects Bristol Myers Squibb, Gilead, and others pursuing new indications or earlier lines of treatment. The shift adds years and significant cost to development timelines. Companies need head-to-head trials against standard of care, larger patient populations, and longer follow-up periods. But here's what I'm watching: this policy change mostly impacts CAR-T programs. While those developers navigate expensive comparative trials in solid tumors where penetration remains a challenge, alternative cell therapy approaches face a different calculus. The FDA's demand for superior efficacy data signals something important. It validates that the standard of care in hard-to-treat solid tumors needs better options, particularly where the blood-brain barrier and tumor penetration create fundamental biological obstacles. For companies working on mechanisms that address those penetration challenges directly, the competitive landscape just shifted. Established players will be occupied with multi-year trials while newer platforms advance through earlier development stages. The regulatory bar went up. But it didn't go up uniformly across all cell therapy modalities.

For the first time, FDA drug CRLs are public. Clearer decisions. Smarter development. Better outcomes. For decades, the FDA’s Complete Response Letters—detailed explanations for why therapies cannot be approved—were kept private. That changes today. In a bold shift toward transparency, the FDA will now publish redacted CRLs for submissions eventually approved. This means sponsors, investors, clinicians, and patients can finally see: ✔️ Why therapies failed approval ✔️ What gaps remain in safety, efficacy, or manufacturing ✔️ Where companies must improve to protect patients and advance science ⸻ 💡 Why now? The FDA is reframing how we learn from regulatory decisions: 🔎 Restoring trust through transparency 📈 Helping sponsors build better products ⚖️ Leveling the playing field for all stakeholders ⸻ 👥 Who benefits? ➡️ Patients: clarity on why medicines are delayed ➡️ Sponsors: fewer repeated mistakes ➡️ Investors: a clearer regulatory picture ➡️ Healthcare: stronger science and public trust ⸻ This raises the bar for transparency and for leadership in uncertain times. ✅ Will the industry rise to meet it? — Jose ⸻ 🔗 Read the FDA announcement: https://lnkd.in/gBYJNAxi 🔗 FDA Disclosures at 21 CFR 314.430: https://lnkd.in/gdBs-wcZ ♻️ Share this to help your network understand what this means for drug innovation and patient care.

FDA's "Radical Transparency" on Decision Letters: A Gamechanger for Real-World Evidence 📢 The FDA now publicly releases Complete Response Letters (CRLs) — the detailed letters explaining why a drug application was not approved — in real time on openFDA (https://lnkd.in/edaSmKGk). Over 300 CRLs are now available for anyone to read. 🔍 Why does this matter for real-world evidence? Because we can now see the FDA's actual reasoning, not just a press release summary. ⚠️ Case in point: REGENXBIO's RGX-121 gene therapy for MPS II (Hunter syndrome), which received a CRL on February 7, 2026 (https://lnkd.in/evRENVPp). The publicly visible letter revealed that FDA's concerns centered on whether the external natural history control group was sufficiently comparable to the treated population — a core real-world evidence challenge. The agency also questioned whether the surrogate endpoint was reasonably likely to predict clinical benefit, and suggested paths forward that included incorporating an on-study untreated control group. This for an ultrarare, life-threatening pediatric disease where traditional RCTs are extremely difficult to conduct. 📊 The implications for anyone designing RWE-supported submissions are clear: external control comparability must be rigorously demonstrated, not just assumed. Target trial emulation principles matter more than ever. And there remains a real tension between FDA's agency-level support for RWE and the evidentiary standards applied by review divisions — particularly for rare diseases where conventional trial designs may be impractical. 📖 These publicly available CRLs accelerate shared learning and build trust in FDA decision-making. For those of us committed to transforming drug development through real-world evidence, they are required reading. 💬 Does publishing CRLs help or hinder the adoption of RWE in regulatory submissions? #RealWorldEvidence #FDA #FDATransparency #RWE #DrugDevelopment #TargetTrialEmulation #RareDisease #ClinicalTrials #EvidenceGeneration #Pharma

The FDA just made it possible to approve therapies for diseases with fewer than 10 patients. It is called the Plausible Mechanism Framework. Here is what it means for patients with a rare condition but no drug research. 𝗧𝗵𝗲 𝗼𝗹𝗱 𝗽𝗿𝗼𝗯𝗹𝗲𝗺: Your genetic condition affects 12 patients globally. No control arm exists. No placebo makes sense. Traditional trials are impossible. So your therapy never gets developed. 𝗧𝗵𝗲 𝗻𝗲𝘄 𝗽𝗮𝘁𝗵𝘄𝗮𝘆: FDA now outlines approval based on four core pillars: • A well-supported mechanism of action • Natural history data • Target engagement proof • Reasonably likely surrogates No randomized trial required. This framework explicitly covers genome-editing products and RNA-based therapies like antisense oligonucleotides. But it is written broadly enough for any targeted product that addresses a known genetic, cellular, or molecular abnormality. 𝗧𝗵𝗲 𝗕𝗮𝗯𝘆 𝗞.𝗝. 𝗽𝗿𝗲𝗰𝗲𝗱𝗲𝗻𝘁: Senior FDA leadership previewed this in NEJM using Baby K.J. - a patient-specific, single-patient CRISPR trial - as the prototype. This draft guidance operationalizes that concept into a regulatory roadmap. 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗺𝗮𝘁𝘁𝗲𝗿𝘀 𝗻𝗼𝘄: FDA is using master protocols that let one application cover multiple mutation-specific variants. You prove the mechanism once. Then extend to related mutations without repeating the full evidentiary package. 𝗧𝗵𝗲 𝗰𝗼𝗺𝗺𝗲𝗿𝗰𝗶𝗮𝗹 𝘀𝗵𝗶𝗳𝘁: 30 million Americans have rare diseases. Most have private or ultra-rare variants historically deemed non-viable. This framework clarifies what evidence FDA needs when N is too small for traditional designs. It incentivizes development of individualized therapies by removing the "impossible trial design" barrier. 𝗪𝗵𝗮𝘁 𝘁𝗼 𝗱𝗼: If you are working on an ultra-rare program, read the draft guidance now. The FDA is telling you what they need to see. What are your thoughts surrounding the Plausible Mechanism Draft Guidance? Follow Dr. Suzanne Morgan for the latest in the FDA Guidance for Rare Disease Drug approval