Strengthening Postapproval Monitoring: FDA’s Draft Guidance on Cell & Gene Therapy Products

FDA recently released a draft guidance, titled “Postapproval Methods to Capture Safety and Efficacy Data for Cell and Gene Therapy Products,” that provides valuable direction and insight into how sponsors can (and should) approach long-term data collection once a cell and gene therapy (CGT) product is approved.

Postapproval monitoring of safety and effectiveness of CGT products is of critical importance due to the long-lasting effects and typically limited number of subjects treated in clinical trials.  Unlike most other treatments for chronic diseases, CGT products are designed to achieve very prolonged or permanent effects and can’t typically be discontinued once administered.  Additionally, some of these patients may be treated as children, and it is important to assess any developmental impacts as they age into adulthood.  Therefore, these products may need to be monitored over several years or, in some cases, the lifetime of the patient.  Monitoring of patients and collection of long-term safety and efficacy data will inform sponsors and FDA about treatment durability, delayed adverse events, and real-world safety/efficacy trends.

The draft guidance was born out of an April 27, 2023 virtual public listening meeting, which, in turn, was a Prescription Drug User Fee Act (VII) commitment.  The meeting sought to gather relevant datapoints and perspectives for CGT products relating to methods and strategies for capturing postapproval safety and efficacy data.

This draft guidance references a 2020 guidance, “Long Term Follow-Up After Administration of Human Gene Therapy Products,” throughout.  This guidance describes a range of follow-up periods recommended by the Agency based on the specific type of gene therapy and its general potential to cause delayed adverse events.  This ranges from “up to five years” for AAV vectors to fifteen years for integrating vectors; this long-term follow-up is important to capture in informed consent forms.  However, this length of follow-up has proven challenging in practice for a variety of reasons, including for companies that cease operating in that timeframe (a lot can happen in 15 years).  As such, this new draft guidance recommends that “sponsors should provide a plan for follow-up, including funding, in the event the sponsor ceases to operate the study before completion.”  The postapproval collection of data described in this draft guidance is, in large part, intended to augment the long-term follow-up from the clinical trials.

Focal Points from the Draft Guidance

The draft guidance highlights some general methods, including concerns and limitations, for collecting postapproval data.

• General real-world data (RWD) and real-world evidence (RWE) principles

As this data collection depends heavily on the use of RWD and RWE, the draft guidance refers companies to its numerous resources regarding the use of RWE.  The guidance suggests using existing data sources (e.g., electronic health records, claims data, registries) to generate RWE supporting ongoing safety and effectiveness monitoring.  When using RWE, important data governance practices include ensuring data quality and confidentiality, as well as transparent, auditable, and compliant processes (e.g., HIPAA, Part 11).  Sponsors are encouraged to engage FDA (CBER) early when designing studies using RWD to support submissions.

• RWD sources

EHRs, claims data, registries, and vital statistics are commonly used sources for RWD (FDA recently published a handy list of approvals and other labeling changes based on RWE, and the data sources in this list reflect this).  The draft guidance describes a variety of potential uses for these RWD sources, including conducting utilization studies to assess exposure and characteristics of patients or prescribers, assessment of rates of clinical outcomes, determinations of background rates of outcomes of interest (e.g., malignancies, cardiovascular complications) occurring in the absence of CGT exposure, and training AI and Natural Language Processing machine-learning models to develop computable phenotypes for safety or efficacy outcomes.

As these data sources were not originally created for regulatory safety/effectiveness studies, sponsors must carefully evaluate their design and suitability.  Key challenges include absence of evidence regarding important covariates, missing or unstructured data, coding lag or inconsistency, fragmented patient records (due to changes in provider or insurer), and statistical limitations in rare disease settings.  Due to these challenges, the draft guidance recommends that sponsors perform feasibility assessments, collect as much uninterrupted data as possible, deal transparently with missing data, and consider linking multiple data sources.

In contrast to the above sources, the draft guidance notes that registries are often “highly curated RWE resources that may be able to overcome common RWD limitations.”  Registries can collect longitudinal, structured, disease-specific data including lab/imaging, patient-reported outcomes, adherence data, biomarkers, etc.  However, there are limitations of registries as well, including potential selection bias (e.g., patients more engaged or with more severe disease might enroll) resulting in a data source that may not be representative of the broader patient population.  Sponsors should address these in their design, including by encouraging all patients to participate and reducing burdens to participants.

FDA highlights four situations that are particularly relevant for postapproval data collection with registries: (1) assessment of long-term durability of response; (2) growth and developmental milestone data for pediatric patients; (3) surveillance for malignancies; and (4) fertility and pregnancy outcomes-related data.

• Decentralized data collection

Borrowing practices from decentralized clinical trials, postapproval data collection can also be done remotely (e.g., in patients’ homes, local clinics, via telehealth).  Benefits include reduced administrative/logistical burden, increased enrollment, better retention, and broader patient access.  These should improve generalizability of study results.

Protocols should, among other things, specify how to collect safety/efficacy outcomes remotely, the role of local providers, how to track and document effectiveness outcomes, and how care will be provided for adverse events requiring urgent care.

Best Practices for Industry and Parting Thoughts

We are still in the early days of commercial cell and gene therapies, and there is much still unknown.  Additionally, there continue to be new innovations and new treatment modalities.  There are some unique considerations to these products that require a different regulatory approach in a variety of ways.  Clinical trials are very useful in collecting important information in a shorter-term context, but these products have simply not been around long enough to fully understand any longer-term effects, if any.  This draft guidance is FDA seeking as much information as possible to inform its understanding of how best to regulate these products, which we very much encourage.

Patient safety is the paramount concern, and it is crucially important to understand whether there are any potential delayed adverse events from such products.  Relatedly, the draft guidance reminds sponsors that IRB/informed consent requirements apply in a postapproval context to ensure patients are protected.

However, as time goes by, we hope to learn that these concerns are overblown and that such delayed adverse events occur very rarely, if at all.  We hope sponsors pay close attention to these FDA recommendations.   The draft guidance emphasizes that sponsors should reach out to CBER and work with the Agency to make the best use of data possible.  If we accumulate enough experience to gain this understanding, perhaps the burdensome (for both patients and industry) long-term follow-up requirements can be relaxed, as appropriate.