Gene Therapy’s Evidence Problem—Lessons From Recent FDA Decisions

Why It Matters

The ruling signals a tightening of FDA expectations for gene‑therapy data, increasing risk for developers that rely on surrogate endpoints. Strengthening early‑stage evidence could reduce costly late‑stage failures and accelerate patient access to transformative treatments.

Key Takeaways

• FDA rejected REGENXBIO's RGX‑121 due to unvalidated biomarker endpoint.
• Early preclinical models often fail to predict human outcomes for gene therapies.
• Regulators are tightening evidentiary standards for long‑lasting advanced therapeutics.
• Human‑derived models and computational tools are emerging to improve early evidence.

Pulse Analysis

The recent FDA refusal of REGENXBIO’s RGX‑121 underscores a growing tension between innovation and evidentiary rigor in the gene‑therapy space. While the agency continues to endorse flexible trial designs, it is increasingly scrutinizing surrogate endpoints that lack clear clinical validation. This shift reflects a broader regulatory trend: as therapies promise permanent biological changes, the burden of proof rises, forcing sponsors to demonstrate not just safety but durable, measurable benefit.

Underlying the regulatory push is a structural weakness in the drug‑development pipeline. Traditional animal models and simplified in‑vitro systems, designed for small‑molecule drugs, often cannot capture the nuanced human biology that gene and cell therapies target. Consequently, early signals may appear promising yet prove insufficient when translated to patients, leading to late‑stage setbacks that erode investor confidence and delay market entry. Companies that rely heavily on these legacy models risk misaligning trial design, dosing strategies, and endpoint selection with the realities of human disease.

To address these gaps, the industry is investing in human‑centric platforms such as organoids, organ‑on‑a‑chip, and large‑scale computational models. These technologies generate more predictive data, enabling developers to refine hypotheses before committing to costly clinical programs. By integrating multi‑modal human data early, firms can build a more robust evidentiary foundation, satisfy heightened FDA expectations, and ultimately accelerate the delivery of life‑changing gene therapies to patients. The next wave of innovation will hinge as much on refined evidence generation as on the therapeutic modalities themselves.