5 Key Challenges in the Development of Gene Therapies

July 1, 2024

Two lab techs working with petri dishes test tubs in clean room

Over three decades of gene therapy (GT) clinical trials have yielded many innovative and promising treatments for pernicious genetic disorders but industry has shown slow development and dashed expectations. Often the desire for developing a needed treatment exceeds the sponsors understanding of the regulatory framework and respective healthcare models. This blog reviews the five key challenges in the development of gene therapies and recommendations to realize effective treatments.

The five challenges include:

  1. Lack of sufficient and relevant nonclinical efficacy and nonclinical safety concerns.
  2. Selecting the correct CDMO partner.
  3. Planning for GT treatment payor pre-approval / reimbursement.
  4. Agreeing on the correct clinical trial endpoint.
  5. Clinical site selection.

Let’s take a deeper dive into each challenge.

Lack of sufficient and relevant nonclinical efficacy and nonclinical safety concerns

Animal Models of Efficacy

Appropriate animal model selection is one of the first and most important steps to advance an investigational gene therapy (GT) into clinical trials. The nonclinical program should be tailored to the investigational product and the planned early-phase clinical trial to allow characterization of the product’s benefit/risk profile for the intended patient population.

The overall objectives of a nonclinical efficacy program for a GT product include:

  1. identification of a biologically active dose range
  2. recommendations for an initial clinical dose level, dose-escalation schedule, and dosing regimen
  3. establishment of feasibility and reasonable safety of the proposed clinical route of administration (ROA)

Nonclinical in vitro and in vivo proof-of-concept (POC) studies are recommended to establish feasibility and support the scientific rationale for administration of the investigational GT product in a clinical trial and finding the relevant animal model(s) to demonstrate efficacy/mode of action (MOA) and is a recurrent issue in GT nonclinical development. In vitro studies alone provide limited information on specificity and efficacy and are often inadequate for predicting pharmacological effects in humans and is recommended to obtain data from in vivo studies to best predict efficacy in humans. In vitro studies can be considered when relevant in vivo disease models are not available.

Factors that drive the nonclinical efficacy program include:

  1. GT product characteristics
  2. putative MOA
  3. target disease indication
  4. pediatric vs. adult population
  5. clinical ROA and dosing plan
  6. anatomic site of delivery

Efficacy studies should be performed using relevant animal species and models suitable to show that the nucleic acid sequence reaches its intended target (target organ or cells) and provides its intended function (level of expression and functional activity). Genetically modified animal models such as knockouts, transgenic or humanized models that express the gene target and recapitulate some if not all, disease pathophysiology are often used for GT testing.

It is known that many animal models do not completely mimic human diseases and viral vectors may infect various species differently. The testing of vectors in animal models often resemble the responses obtained in humans, but the larger size of humans in comparison to rodents presents additional challenges in the efficiency of delivery and penetration of tissue. To enhance the nonclinical-to-clinical translation, animal models of disease should reflect human pathophysiology as close as possible and often represents a key challenge. This ideally includes the molecular defects, biochemical abnormalities, pathology, functional changes, clinical signs and symptomatology, and the disease´s course of progression. To enhance the nonclinical-to-clinical translation, the primary and secondary measures should be aligned as closely as possible with those that can be used in human clinical trials (e.g., functional tests, biomarkers, imaging, etc.). Animal models of efficacy can also complement nonclinical safety evaluations including GLP regulatory toxicology studies with focus on identifying doses anticipated to be associated with a positive clinical benefit and favorable risk-benefit. Small animal models are easy to handle, and relatively cost-effective, and genetically modified models can be generated. However, small animal models only partially recapitulate human disease pathophysiology and cannot be used to estimate clinical dosages. It is critical to develop animal model(s) that recapitulates disease pathophysiology and has the most relevant anatomy and physiology but is also feasible from a practical standpoint.

If an appropriate animal model in which the transgene carried by the vector induces pharmacological effects similar to those in humans is not available because of species differences, it should be considered to conduct studies that uses a vector expressing a gene derived from the model animal that is homologous to the intended human gene. In this instance, it should be justified that the results obtained from those studies can be extrapolated to humans.

If the GT product will be used in a pediatric first clinic study, regulatory agencies often have concerns about animal models of efficacy being relevant as pediatric patients represents more than a minor increase over minimal risk, therefore demonstration of prospect of direct benefit (PDB) is often required. Nonclinical evidence to support a PDB is most important when clinical evidence of effectiveness is not available from adult subjects with the same disease. A strong argument needs to be presented to regulatory agencies that the conducted animal model(s) do indeed demonstrate a PDB. Often, regulatory agencies do not agree and recommend conducting clinical studies in adults (if possible) first prior to pediatric clinical studies. If no prior adult data is available, sponsors should provide a rationale as to why adults studies are not ethical, feasible or relevant. There are instances in which an in vivo animal study cannot be completed, for example, when a relevant disease model is not available. In this situation, the nonclinical package of studies sponsor believes confers PDB for early-stage pediatric enrollment should be discussed with the regulatory agencies in meetings such as INTERACT and Pre-IND.

In summary, a key challenge in the development of GT products is the development/availability of suitable nonclinical in vivo models of disease with robust phenotype translatability to human diseases. The design of clinical trials based on “proof-of-concept” experiments in animal models should anticipate substantially greater heterogeneity in the response to gene therapy in humans. Having relevant models allows for better characterization of biological activity, efficacy, pharmacology, and pharmacokinetics-pharmacodynamics relationships.

Nonclinical Safety Concerns

Toxicology studies for an investigational GT product should incorporate elements of the planned clinical trial (e.g., dose range, ROA, dosing schedule, evaluation endpoints, etc.) to the extent feasible. Study designs should be sufficiently comprehensive to permit identification, characterization, and quantification of potential local and systemic toxicities, their onset (i.e., acute or delayed) and potential resolution. In recent years, the number of clinical trials in which adeno associated virus (AAV) vectors have been used for in vivo gene transfer has steadily increased. The excellent safety profile, together with the high efficiency of transduction of a broad range of target tissues, has established AAV vectors as the current platform of choice for in vivo GT. Following administration of recombinant adeno-associated virus (rAAV)-based GT, the most common nonclinical toxicities include immunogenicity, hepatotoxicity, and neurotoxicity. Neurotoxicity, more specifically dorsal root ganglia (DRG) findings and regulatory agency’s view of this finding and potential clinical translation will be discussed. DRG toxicity involves axonal degeneration and swelling of DRG, spinal cord, and/or peripheral nerve have been observed in monkeys and mice given AAV-based gene therapies. The exact mechanism for these findings is unclear but evidence suggests that pathogenesis is possibly a consequence of increased vector genome transduction and/or transgene expression. Monkeys, mice, pigs, rats, and dogs all have all been shown to be susceptible to AAV induced DRG toxicity, however, sensitivity or relevance to humans is poorly understood. Presently, monkeys are considered the most relevant species due to the preponderance of studies conducted in this species and are the most sensitive species for studying vector induced DRG toxicity. The brains of non-human primates (NHPs) are very similar to humans in terms of anatomy, network organization and abilities so NHPs are well suited to study gene delivery to the brain.

The clinical translation of DRG toxicity seen in animals is unclear, and more clinical data and nonclinical studies are needed to better understand the potential mechanisms of toxicity, and clinically translatable safety biomarkers are needed. Despite many ongoing clinical trials with rAAV-based gene therapies, little evidence of CNS effects has been reported in humans with the relevance of monkey DRG toxicity being unclear. Potential risk should be monitored through periodic neurological examinations and other methods (monitoring of sensory motor neurons functionality) to mitigate the risk of DRG toxicity in clinical trials.

DRG toxicity is a common finding for GT products in non-human primates (NHP) and several factors are thought to contribute to the incidence and severity of DRG histopathology findings and include: 1) direct administration into the CSF via intra-cisterna magna (ICM) or intrathecal (IT) injection, and 2) administration of dose levels that are greater than 1 × 1013 vg/animal (approximately 1 × 1011 vg/g of brain weight). AAV does not cross the blood–brain barrier (BBB) readily in species larger than mice therefore intraparenchymal or IT routes of administration (RoA) are often used for neurodegenerative diseases. For example, DRG toxicity has been reported as a target organ in NHP administered ZOLGENSMA by the IT RoA, but not when administered by the IV route and the difference in response is thought to be due to the relatively high local vector concentrations in DRG neurons following IT administration, when compared to those achieved with IV dosing.

Observing this DRG toxicity in NHP often results in the inability of identifying a no-observed-adverse-effect level (NOAEL) in the pivotal GLP toxicity study and could potentially complicate clinical development. This current thinking is that the absence of a NOAEL for DRG toxicity should not preclude clinical development in the context of appropriate risk/benefit considerations and incidence/severity of DRG findings and clinical data is needed to understand the relevance of these nonclinical findings. An alternative to the nonclinical NOAEL approach is to potentially use the highest non-severely toxic dose (HNSTD) observed in the NHP GLP toxicity study to support a human starting dose. HNSTD could be a potentially a more suitable approach for setting the threshold value in communicating adversity for AAV-based gene therapies to regulatory agencies, especially in situations intended to treat disease indications that are considered severely debilitating or life-threatening or have high unmet medical need.

Different regulatory agencies have different opinions on this nonclinical toxicity in the context of selecting an appropriate human starting dose. Since rAAV vectors have been used for many of the currently approved GT products with acceptable safety profiles (lack of a clinical correlate for the DRG toxicity observed in NHP), sponsors are encouraged to present the nonclinical safety findings to regulatory agencies and provide rationale why the proposed clinical dose would be safe. Sponsors are recommended to consult with multiple regulatory agencies to gain agreement and align on the necessary studies, parameters measured and toxicity findings and if they will support the planned clinical study. For example, regulatory agencies may ask for longer than a 6-month readout in the GLP NHP toxicology study or recommend potential exploratory clinical biomarkers for the DRG toxicity (ex. circulating neurofilament light chain (NF-L)). Some regulatory agencies are more concerned about nonclinical DRG toxicity which influences the studies and endpoints required and could make it difficult to design a nonclinical toxicity package that is acceptable to all regions. One thing is clear is that DRG toxicity is an important “potential” human risk that will need to be considered, explained to regulatory agencies if observed in nonclinical studies and managed in clinical studies.

In summary, relevant animal models of disease are key to assessing variables related to the use of AAV vectors such as efficacy, dose, safety, and localization of transgene expression and are important before advancing into human trials.

Key questions to ask are:

  1. Which model best recapitulates human disease physiology?
  2. Which model best works for the targeted gene?

Animal models of efficacy for GT products provide important nonclinical evidence of feasibility and effectiveness and are needed to support planned clinical trials. Nonclinical safety studies for GT products should provide evidence to support clinical trials including the starting dose, potential target organs of toxicity and possible safety biomarkers, and to evaluate of risks-benefits of the design of the clinical trial. Communication with regulatory agencies such as the FDA at early stages of product development is recommended and could include INTERACT and Pre-IND meetings.

Selecting the correct CDMO Partner

Selection Methodology

Most GT innovators require a Contract Development Manufacturing Organization (CDMO) to develop gene therapy processing for clinical trials but with thousands of CDMO’s globally, how can you make sure that your gene therapy will be developed and manufactured according to your needs and requirements? The best practice is to utilize a five “C” methodology for selecting your CDMO partner.

  1. Capability. The capabilities of CDMO’s are as diverse as the number of CDMO’s. Identification of the GT specific capability requirements requires an effective process for creating and testing the GT. There can be gaps in the process in early phases, but a basic process flow is needed.
  2. Compliance . A strong compliance history with a compliance mindset demonstrated by overall policies, staffing and procedures which ensure compliance.
  3. Capacity. Ensuring that the CDMO’s capability is important but it doesn’t mean much if the CDMO does not have the capacity (either equipment or personnel) to produce the GT according to the sponsor’s schedule as well as delivering results and releasing materials in a timely manner.
  4. Customer service. A strong CDMO will have systems for planning, tracking, analysis, resolution, and reporting.
  5. Cost. Cost is always a factor in all development, but a well-structured pricing process creating transparency on the cost drivers including equipment, materials, staffing, facilities.


A CDMO’s GT capabilities need to be understood including receiving, storing, release, development, manufacturing, data systems and analytical testing. An understanding of the GT’s critical process parameters, the raw material requirements, data review and developmental testing methodologies. A limit on any one of these capabilities could delay development or implementation of a GT program. Also the costs to obtain systems such as control rate freezers, LN2 containment will need to be factored into the overall cost. Best in class capability reviews include a listing of the required capabilities, the CDMO’s capability and status including qualification status, procedural controls.


A CDMO’s compliance is important for multiple reasons including data integrity and focus on compliance without shortcuts. Reviewing the CDMO’s compliance history is a good starting point but additionally a review of deviation history CAPA overdue, training compliance will provide a clear picture of a CDMO’s likelihood of producing a compliant GT. Also, the quality system should be reviewed to ensure a holistic QMS program. The CDMO’s product profile needs to be reviewed to ensure that the GT can be produced in the same facility as other products and that a new product assessment is completed on any products which might be stored, produced, tested do not cause a cross-contamination issue.


Ensuring that the CDMO’s capability is important but it doesn’t mean much if the CDMO does not have the capacity (either equipment or personnel) to produce, analyze, hold and release the GT according to the sponsor’s schedule as well as delivering results and releasing materials in a timely manner. A review of the current CDMO product list of products utilizing the same equipment, warehousing and QC lab including volumes, frequency, phase of development are needed to gain an accurate understanding of when a new GT may be produced.

Customer Service

A strong CDMO will have systems for planning, tracking, analysis, resolution, and reporting. Most CDMO’s have dedicated project managers to support GT programs, but additional support may be required including logistical support for raw materials and finished products. GT’s have specific requirements based upon their CPP’s and CQAs. A strong customer service system will synergize the development rather than delay the program.


Cost is always a factor in all development, but a well-structured pricing process creates transparency on the cost drivers including equipment, materials, staffing, facilities. Transparency is created when the core drivers for each of the costs are understood and provided. Customer service should provide a cost breakdown aligned with the GT process being developed. Use caution when a CDMO provides a blanket cost, if the CDMO does not know the drivers (labor, materials, facilities, equipment time) the costs of the GT will be substantially increased vs a well-defined and understood model based on the GT specific requirements.


A complete report on the five "C"'s will provide realistic expectations, a list of gaps, timing, and costs to develop a GT. The report is a starting point for a technical transfer plan and a risk assessment allowing GT sponsors to establish a solid starting point. ProPharma has standard, customizable surveys, an up-to-date CDMO capabilities database of thousands of CDMO’s globally as well as contact lists to fast-track identification of the right GT CDMO for your product.

Planning for GT treatment payor pre-approval / reimbursement

Lifesaving GT treatments are not helpful if patients cannot afford them. Recent GT approvals which should have been celebrated were demonized when patient access was limited. Several payor options have been developed to support the life altering GT’s including Clinical Evidence and Outcomes Data, Health Economics and Value Proposition, Regulatory Landscape and Compliance, and others. Let’s review these further.

Clinical Evidence and Outcomes Data. Demonstrating the efficacy and safety of the gene therapy treatment through robust clinical evidence and outcomes data is essential to ensure payors will pre-approve a treatment. Strong clinical data not only supports the treatment's effectiveness but also assists payors in assessing its value and potential long-term benefits. Conclusive clinical trial endpoints which demonstrate quantifiable positive outcomes across the spectrum of the disorder as well as the patient population provides confidence for the value analysis.

Health Economics and Value Proposition: Highlighting the economic value of gene therapy is crucial. This involves detailing the treatment's cost-effectiveness, its impact on reducing long-term healthcare expenses (such as hospitalizations or ongoing treatments), and the improvement in patients' quality of life.

Regulatory Landscape and Compliance: Understanding the regulatory requirements and compliance standards set by regulatory bodies is imperative to ensure a viable payor model. Ensuring adherence to these guidelines not only facilitates the approval process but also builds credibility and trust with payors.

Engagement with Payors and Payers: Establishing open communication channels with payors early in the process is key. This involves educating them about the treatment, discussing pricing models, and addressing any concerns they may have regarding coverage and reimbursement.

Coding and Billing Strategies: Developing clear coding and billing strategies that align with existing healthcare reimbursement systems is crucial for smooth payment processes. This includes understanding appropriate billing codes, reimbursement rates, and navigating reimbursement mechanisms such as Medicare or private insurers.

Patient Access and Affordability Programs: Implementing patient access programs, financial assistance, and reimbursement support can alleviate the financial burden on patients. These initiatives can range from co-pay assistance to providing access to patient support services.

Agreeing on the correct clinical trial endpoint

A critical success factor for clinical trials can be difficult to select the most appropriate one. The inability to choose the right endpoint can lead to difficulty in interpreting findings, wasted resources, and decreased value in the overall results (Moher et all, 2016). The endpoint selected must be useful to the patient community, payors, and the clinicians. For gene therapy studies, the studies are often designed to have small sample size especially in the rare disease population. With the small number of patients included in the study, there is a risk that the benefit of treatment will not be significant enough.

In the last few years, the industry has seen a few clinical trials where the primary endpoint was not met.

Clinical site selection

Considering that more than 50% of clinical trials have difficulty in obtaining the required sample size, it is important to consider selecting the appropriate sites at the start of the clinical trial. With gene therapy studies often associated with rare diseases, the number of sites required per patient is often higher than non-rare disease studies. According to estimates, 80% of clinical trials do not meet their enrollment targets leading to extended timelines (Desai, 2020) and causing costly delays. Selecting the right clinical trials sites is imperative to the success of the clinical trials.

Feasibility: To overcome the additional complexities associated with running clinical trials in gene therapy, it is imperative to start with a comprehensive site feasibility. The feasibility must contain all areas important to running the clinical trial. It will help in understanding in the US if a site can utilize a Central IRB or utilize accelerated clinical trial agreements. Does the site have staff with the appropriate credentials and experience to run the clinical trial?

Prior experience: The ability to understand the past performance in recruitment is imperative. Utilizing powerful tools that look at past performance across various clinical trials but especially participation in previous gene therapy trials.

Adaptive approaches: An emerging trend in gene therapy clinical trials in easing patient and caregiver burden is creating adaptive approaches. This can range from home healthcare nurses to bespoke models. Careful selection of regionally located sites for treatment can be made, but reduce the patient and/or caregiver travel by allowing for follow up after administration to be completed at sites closer to the patient’s residence.

Required Equipment: During feasibility the equipment at the clinical trial site is often evaluated. Depending on the route of administration used to deliver the gene therapy, specialized equipment may be required. Understanding early in the process both the availability of equipment needed and the site familiarity with it can lead to better site selection.

Experience with Assessments: As previously mentioned, the clinical trial endpoints offer validity to the clinical trial. When the primary or secondary endpoint is a specific outcome assessment, does the trial site have familiarity with the administration?

Ineffective clinical site selection is the number one cause of gene therapies failing to complete phase 2 studies. Identification of sites with capabilities, experience, qualifications to ensure that the correct clinical patients are enrolled, monitored, and results are reported efficiently, effectively and in a compliant manner for your specific program.

Key Takeaways for the Development of Your Gene Therapy Treatment

Developing effective gene therapies involves navigating a complex landscape filled with significant challenges. From the necessity of selecting relevant animal models for nonclinical studies to the intricacies of choosing the right CDMO partner, each step requires meticulous planning and execution. Addressing nonclinical safety concerns, particularly around the use of adeno-associated virus vectors, demands rigorous testing and strategic consultation with regulatory bodies. Furthermore, the financial and logistical aspects of gene therapy development, including payor pre-approval and reimbursement strategies, must be carefully managed to ensure patient access to these groundbreaking treatments.

Achieving the correct clinical trial endpoint and selecting suitable clinical sites are equally crucial, as these decisions directly impact the success of trials and, ultimately, the viability of the gene therapy. Each of these challenges underscores the importance of a comprehensive, strategic approach to gene therapy development.

If you are navigating the complexities of gene therapy development, ProPharma offers unparalleled expertise and support. Our team can assist you in overcoming these challenges, from preclinical research through to clinical trial management and regulatory consultation. Contact ProPharma today to learn how we can help you advance your gene therapy program and bring life-changing treatments to patients in need.

Works cited

  • Moher D., Glaszious Chalmers I: increasing value reducing waste in biomedical research: who's listening? Lancet. 2016;387:1573–1586.


How to Prepare for Laboratory Partner Selection during CBD Product Development

The interest in developing consumer products or therapies derived from Cannabis or CBD is continuously growing. As these new products come to market, there is increasing need to comply with...

7 Critical Factors for Successful Selection of CDMO for Cell and Gene Therapy Manufacturing

Developing, optimising, and manufacturing Advanced Therapy Medicinal Products (ATMP's), such as Cell and Gene Therapy (CAGT) products is extremely complex. The choice of a reliable Contract...

Person holding up a ben with a note with a question mark

Critical Process Parameters and Critical Quality Attributes: Why Does the Selection Process Take so Long?

Quick, can you name the top 10 Critical Process Parameters (CPPs) and Critical Quality Attributes (CQAs) of your bio-pharmaceutical manufacturing process? Well if you hesitated don't feel bad. Most...